Optimal Linear Unbiased Filtering with Polar Measurements for Target Tracking Λ

Size: px
Start display at page:

Download "Optimal Linear Unbiased Filtering with Polar Measurements for Target Tracking Λ"

Transcription

1 Optimal Linear Unbiased Filtering with Polar Measurements for Target Tracking Λ Zhanlue Zhao X. Rong Li Vesselin P. Jilkov Department of Electrical Engineering, University of New Orleans, New Orleans, LA 018 Phone: , Yunmin Zhu Department of Mathematics, Sichuan University, Chengdu, Sichuan 100, P. R. China Abstract In tracking applications, target dynamics is usually modeled in the Cartesian coordinates, while target measurements are directly available in the original sensor coordinates. Measurement conversion is widely used such that the Kalman filter in the Cartesian coordinates can be applied. A number of improved measurementconversion techniques have been proposed recently. However, they have fundamental limitations, resulting in performance degradation, as pointed out in Part III [] of a recent survey. This paper proposes a recursive filter that is theoretically optimal in the sense of minimizing the meansquare error among all linear unbiased filters in the Cartesian coordinates. The proposed filter is free of the fundamental limitations of the measurement-conversion approach. Results of an approximate implementation are compared with those obtained by two state-of-the-art conversion techniques. Simulation results are provided. Keywords: Target tracking, measurement conversion, Kalman filter, optimal linear filtering, filter credibility 1 Introduction In tracking applications, target dynamics is usually modeled in Cartesian coordinates, while the measurements are directly available in the original sensor coordinates, most often in spherical coordinates. Tracking in Cartesian coordinates using spherical measurements is mostly handled by measurement conversion method []. The basic idea is to transform the nonlinear measurements into a pseudolinear form in the Cartesian coordinates, and estimate the bias and covariance of the error of the converted measurement, and then use the Kalman filter framework to do tracking. A number of improved techniques have been proposed [,, 8,,, 1, 9, 10] by different ways of obtaining the bias and covariance of the measurement noise. Measurement Λ Research Supported in part by ONR grant N , NSF grant ECS-98, and NASA/LEQSF grant (001-)-01. models used in target tracking, including measurement conversion methods, are surveyed in []. The pros and cons of various measurement-conversion techniques have been revealed therein with succinct explanation. In particular, it was pointed out that these conversion methods have fundamental flaws, as elaborated in Section of this paper. The idea of developing an optimal linear filter for tracking with linear dynamics and nonlinear measurements was briefly discussed in []. In this paper, such an optimal linear filter is developed based on this idea. This development stems from the recognition that the Kalman filter is nothing else but a recursive BLUE (Best Linear Unbiased Estimator) for a linear system; furthermore, although the Kalman filter cannot optimally handle any nonlinear measurement directly, BLUE filter is still optimal for nonlinear measurements. This paper presents such a recursive BLUE filter for tracking with linear dynamics and polar (nonlinear) measurements. For simplicity, this paper deals only with the -dimensional case. Extensions to other cases will be reported in a forthcoming. The rest of the paper is organized as follows. In Section, we summarize the measurement-conversion method and explain its fundamental flaws. In Section we derive formulas of the recursive BLUE filter for nonlinear observation. In Section, the error analysis is given for the implementation of the optimal recursive BLUE filter. In Section, We compare the implementation of this new method with two state-of-the-art measurement-conversion techniques by simulation. Our conclusions are given in Section. Measurement-Conversion Approach The measured range and bearing are defined with respect to the true range r and bearing r m = r +~r; m = + ~ 1

2 the noise ~r and ~ are assumed to be independent with zero mean and standard deviations ff r and ff, respectively. The converted-measurement method converts the polar coordinates measurement into the Cartesian coordinates by x m = r m cos m =(r +~r) cos( + ~ ) y m = r m sin m =(r +~r) sin( + ~ ) and then transform the above nonlinear form into pseudolinear form with respect to (x; y), x c = x +~x c ; y c = y +~y c (1) where (x c ;y c ) is the converted (most often) debiased measurement, (x; y) = (r cos ; r sin ) is the true measurements and (~x c ; ~y c ) is the converted measurement noise. Observe that (1) just divides the transformed debiased measurement into true measurement (x; y) and measurement noise (~x c ; ~y c ). Note, however, that (~x c ; ~y c ) is actually dependent on the true measurement. The converted measurement method uses the Kalman filter based on (1) to do tracking. All converted measurement techniques differ from each other in the ways that the bias and the covariance of the measurement noise are calculated. Specific comparisons can be found in []. Basically, one way is to compute the mean and covariance conditioned on the measurement as E [(~x c ; ~y c ) jr m ; m ] and cov[(~x c ; ~y c ) jr m ; m ]. Another way is to compute them firstly conditioned on the true state and then conditioned on measurement as E [E [(~x c ; ~y c ) jr; ] jr m ; m ] and E [cov ((~x c ; ~y c ) jr; ) jr m ; m ]. They are referred to as fixed measurement and fixed truth approaches in [], respectively. As argued in [], however, we will call them measurement-conditioned () approach and nested conditioning () approach. From the above analysis, all converted measurement techniques have the following fundamental flaws. First, (~x c ; ~y c ) is state dependent; second, its covariance is estimated conditioned on the current measurement or state/measurement; third, the measurement noise sequence f(~x c ; ~y c ) k g is not white anymore. However, in the assumptions of Kalman filter, the measurement noise is independent of the state, its covariance is not conditioned on the current measurement, and it is white. These fundamental flaws were ignored or overlooked in these techniques. As a result, these techniques are by no means optimal. For example, when the measurement noise is dependent on state, the measurement prediction covariance formula of the Kalman filter should be modified to take into account the crosscorrelation between the state and the measurement noise. All these techniques fail to do such modification. Recursive BLUE Filter Notice that the Kalman filter is simply a recursive BLUE filter for a linear system. Nevertheless, BLUE does not require the observation model to be linear. A BLUE filter can be outlined below. Let z k and xk be the measurement and state at time k; z k1 stand for the sequence of all past measurements, and E Λ [yjz] stand BLUE estimator of y using z. Suppose at time k 1 the state estimator and covariance is ^xk1and P k1. The prediction part of a recursive state estimator is as follows. μxk = E Λ [xkjz k1 ] ~xk = xk μxk μz k = E Λ [z k jz k1 ] ~z k = z k μz k μp k = cov(~xk) S k = cov(~z k ) K k = cov(~xk; ~z k )S 1 k then the update is ^xk = μxk + cov(~xk; ~z k )S 1 k (z k μz k ) P k = P μ k cov(~xk; ~z k )S 1 k cov(~z k; ~xk) This recursive BLUE filter is optimal provided the process and measurement noises are white and uncorrelated, which is the case for the problem under consideration. This recursive BLUE filter requires the knowledge of the predicted state μxk, its error covariance P μ k, the predicted measurement μz k, its covariance cov(~z k ) and crosscovariance cov(~xk; ~z k ). In view of the above, a recursive BLUE filter is derived as follows. The key is to express the state in the Cartesian coordinates while keeping the measurement error in the polar coordinates. Because the target model is still linear, μxk and P μ k are the same as in Kalman filter. We still adopt the nonlinear observation form and calculate E Λ [z k jz k1 ]; cov(~z k ) and cov(~xk; ~z k ) analytically with only ^xk1 and P k1 given. First, transform the measurement in polar coordinates into the Cartesian coordinates: x m = (r +~r)cos( + ) ~ = x cos ~ y sin ~ x + px + y ~r cos ~ y px + y ~r sin ~ () y m = (r +~r) sin( + ) ~ = y cos ~ + x sin ~ y + px + y ~r cos ~ x + px + y ~r sin ~ () Note that since (r; ) and (~r; ~ ) are independent, (x; y) and (~r; ~ ) are also independent. The tracking system has the nonlinear observation: z k = h(xk;v k ) () where z k = [x m ;y m ] 0 k, x k = [x; _x; y; _y] 0 k, v k = [~r; ~ ] 0 k, and the white noise v k is independent of the state xk. 18

3 As pointed out above, μxk and μ P k are the same as in the Kalman filter. For brevity, we will drop the time index k and the conditioning on z k1. Now denote μxk and μ P k by μx =[μx μ _x μy μ _y] 0 μp = cov(~x) cov(~x; ~ _x) cov(~x; ~y) cov(~x; ~ _y) cov( ~ _x; ~x) cov( ~ _x) cov( ~ _x; ~y) cov (_x; ~ _y) cov(~y; ~x) cov(~y; ~ _x) cov(~y) cov(~y; ~ _y) cov( ~ _y; ~x) cov( ~ _y; ~ _x) cov( ~ _y; ~y) cov( ~ _y) What remains is to calculate E Λ [z k jz k1 ]; cov(~z k ) and cov(~x k ; ~z k ). Notice that E Λ [(~r k ; ~ k ) 0 jz k1 ]=E[(~r k ; ~ k ) 0 ] since the noise sequence [~r; ] ~ 0 k is white. Assuming ~r ο N (0;ffr ) and ~ ο N (0;ff ) are independent, we have Thus E[cos ~ ]=e ff = = 1 E[sin ~ ]=0 E[cos ~ ]= 1 (1 + eff )= E[sin ]= 1 (1 eff )= E[sin ~ cos ~ ]=0 μx m = E Λ [x m ] E [~r] =0 = E Λ [x cos ~ y sin ~ x + p x + y ~r cos ~ y p x + y ~r sin ] ~ = E Λ [x] E Λ [cos ~ ] = 1 μx () μy m = E Λ [y m ] = E Λ [y cos ~ + x sin ~ y + p x + y ~r cos ~ x + p x + y ~r sin ] ~ = E Λ [y] E Λ [cos ~ ] = 1 μy () Since the BLUE estimator is unbiased and orthogonal to its error, it can be easily shown that cov(~x) = E[(x μx)(x μx) 0 ] = E[xx 0 ] E[μxμx 0 ] cov(~x; ~z) = E[(x μx)(z μz) 0 ] = E[xz 0 ] E[μxμz 0 ] cov(~z~z 0 ) = E[(z μz)(z μz) 0 ] = E[zz 0 ] E[μz μz 0 ] NOTE μx and μz are the predicted state and measurement BLUE estimator here. The crosscovariance between measurement z and state x is calculated as follows. First, E [xz 0 ]=E xx m _xx m yx m _yx m m _ m yy m _yy m Substituting () and () into the above yields Then E [xz 0 ]= 1 E cov(~x; ~z) = E [xz 0 ] E[μxμz 0 ] = 1 E x _x _ y x _y _yy x cov(~x) = 1 cov( ~ _x; ~x) cov(~y; ~x) cov( ~ _y; ~x) x _x _ y x _y _yy x 1E cov(~x; ~y) cov( ~ _x; ~y) cov(~y) cov( ~ _y; ~y) The measurement prediction covariance is Since cov (~z) = E [zz 0 ] E[μz μz 0 ] cov(~xm ) cov(~x m ; ~y m ) = cov(~y m ; ~x m ) cov(~y m ) μx μxμy μx μ _x _xμy μxμy μy μx μ _y μ _y μy () E[x m ] = E[x cos ~ y sin ~ x + p x + y ~r cos ~ y p x + y ~r sin ] ~ and similarly = E[x cos ~ + y sin ~ + x x + y ~r cos ~ + y x + y ~r sin ] ~ = 1 (1 + eff )E x Λ + 1 (1 eff )E y Λ + 1 ff r + 1 ff r eff E x y x + y = E x Λ + E y Λ + 1 ff r + 1 ff r eff E x y x + y E[y m ] = E y Λ + E x Λ + 1 ff r + 1 ff r eff E y x x + y E[x m y m ] = e ff E [] +ff r eff E x + y 19

4 we have cov(~x m )=E[x m ] E[μx m μx 0 m ] (λ λ 1 )/λ λ /λ = E x Λ + E y Λ + 1 ff r + 1 ff r eff E x y x + y 1 E[μx ] = cov(~x) + cov(~y) + 1 ff r + μy +( 1 )μx + 1 x y ff r eff E x + y +( 1 )(E[μx ] μx )+ (E μy Λ μy ) (8) ß cov(~x) + cov(~y) + 1 ff r + μy +( 1 )μx + 1 x y ff r eff E x + y cov(~y m )=E[x m ] E[μx m μx 0 m ] = cov(~y) + cov(~x) + 1 ff r + μx +( 1 )μy + 1 y x ff r eff E x + y (9) +( 1 )(E μy Λ μy )+ (E[μx ] μx )(10) ß cov(~y) + cov(~x) + 1 ff r + μx +( 1 )μy + 1 y x ff r eff E x + y Let = e ff 1. Then cov(~x m ; ~y m )=E[x m y m ] E[^x m ^y m ] = e ff cov(~x; ~y) +ffr eff E x + y ß (11) + μxμy + (E[μxμy] μxμy) (1) e ff cov(~x; ~y) +ffr eff E x + y + μxμy (1) With (8), (10) and (1), the optimal recursive BLUE filter has the state update below ^x = μx + cov(~x; ~z)cov 1 (~z)(z μz) (1) P = P μ cov(~x; ~z)cov 1 (~z) cov(~z; ~x) (1) Error Analysis The two terms in (8) ( 1 )(E[μx ] μx )+ (E μy Λ μy ) are dropped in (9). We justify this approximation as follows. As we know, (μx E[μx ]) is usually much smaller σ θ Fig. 1: Ration of ( 1 ) and than μx in a tracking problem. Besides, the ratio of ( 1 ) and are as Fig. 1. Usually in a tracking problem, the bearing noise standard deviation ff is below 0:rad, so the contribution of (E[μx ] μx ) and (E μy Λ μy ) to cov(~x m ) will be greatly decreased by ( 1 ) and compared with the term cov(x). For example, when ff =0:rad, ( 1 ) = 0:0% and =%. Therefore, the dropped term is negligible compared with other terms. A similar justification can be applied to (11) and (1). So this approximation is highly accurate. It seems that the two terms E h y x x +y i and E h i x +y involved in cov(~z) have either no closed form or a complex expression. We approximate them by Taylor series expansion. The Taylor series expansion of a function f (x; y) at point (x 0 ;y 0 ) is f (x; y) ßf (x 0 ;y 0 )+(x x 0 (x 0;y 0 +(y y 0 (x 0;y 0 ρ + 1 (x x 0 f (x 0 ;y 0 +(x x 0 )(y y 0 f (x 0 ;y 0 ff +(y y 0 f (x 0 ;y 0 (1) Expanding y x x +y and x +y by Taylor series at (μx; μy) and then taking expectation yields y x E x + y ß μy μx μx +μy + 1! ( μy μy μx (μx +μy ) cov(x) + 8μxμy μx μy cov(x; (μx +μy ) y) ) μx μx μy (μx +μy ) cov(y) (1) 10

5 E x + y ß μxμy μx +μy + 1! ( μxμy μx μy (μx +μy ) cov(x) + μx μy μx μy cov(x; y) (μx +μy ) + μxμy μy μx (μx +μy ) cov(y) ) (18) It is reasonable to expect that this approximation is not crude at all since the denominator has higher order than the numerator in the differential terms and (μx +μy ) is usually much bigger than cov(~x),cov(~y) and cov(~x; ~y). Once (1) and (18) are calculated, our proposed recursive BLUE filter can be implemented straightforwardly. This implementation is a highly accurate approximation to the optimal BLUE filter. The above analysis will be verified in the next section. We can see that actual estimation errors are almost always perfectly consistent with the filter s computed covariance in all simulation results. Simulation and Comparison In this section, we compare two state-of-the-art conversion techniques with our proposed method. We choose the measurement-conditioned () approach and one of the nested conditioning () approaches, called fixed measurement and additive fixed truth approach II in [], respectively. See appendix for their formulas. We denote these two approaches by and, respectively. For convenience of comparison and computation, we use scenarios similar to that of [], but simplify it to two dimensions by eliminating elevation. The scenario is as follows. A two-dimensional Cartesian x-y space with a single sensor locates at the origin. Target sampling time is one sec. The coordinates (x; y) of the target object at time zero are determined by random draws from two independent, Gaussian distributions with means 0 km and 00 km, respectively, and common standard deviation km. The target moves at a constant high velocity (including direction and speed), whose components _x and _y; determined by a random draw from a Gaussian distribution, have means 1000m/s and 0m/s, respectively, and a common standard deviation 0:1 km/s. The sensor s independent measurement errors have standard deviations ff r = m and ff =10millirad. Following [], all the filters are initialized with an effectively infinite initial state error covariance matrix, and a highly inaccurate initial state estimate. The tracking period begins 100 sec after time zero and continues for 100 sec. We will compare the average normalized estimation error squared (ANEES) and root-mean-square error (RMSE) of position of these three techniques by increasing the process noise, and then increasing the measurement noise. The ANEES is defined by μψ = 1 Nn NX i=1 (xi^xi) 0 P 1 i (xi^xi) (19) where (xi^xi) is the -dimensional vector of state estimation errors in sample run i, and N is the total number of samples used in the test. If the estimation error and estimated covariance is consistent, the expectation of ANEES is 1 []. Because the target moves at a constant high velocity, the velocity RMSEs of those three filters have little difference in all simulations and will not be present. The performance of the three filters in different cases can be summarized as follows. Case 1 No process noise; measurement errors ff r = m and ff =10millirad; 100 runs Fig. : ANEES (Case 1) Fig. : Position RMSE (Case 1) 11

6 Fig. compares the ANEES of the and techniques with the proposed filter. The proposed one is much more consistent than the other two. Furthermore, the ANEES of the proposed filter is very close to the ideal unity, which indicates the estimator is almost perfectly consistent. Fig. compares the position RMSE of the three techniques. For convenience, we separate the transient part and steadystate part. It is not easy to tell which one is better in the transient part. However, our method is much more accurate than the other two after transient. Case Process noise Q = 10 I; measurement errors ff r =m and ff =10millirad; 100 runs filter has also smaller position RMSE than the other two after transient. Case Measurement errors ff r = 0m and ff = 100millirad; process noise Q =10 I; 100 runs Fig. : ANEES (Case ) 1. x Fig. : ANEES (Case ) Fig. : Position RMSE (Case ) In this case, the process noise is increased. As is clear from Figs. and, the ANEES of our proposed filter is very close to unity and is better than the other two. Our Fig. : Position RMSE (Case ) In this case, the measurement noise is increased. The simulation results are plotted in Figs. and. Our filter is much more accurate and credible. In particular, the ANEES of our filter is still very close to the ideal unity. Case Measurement errors ff r = 00m and ff = 00millirad; process noise Q =10 I; 100 runs In this case, the measurement noise is increased greatly. The simulation results in Figs. 8 and 9 show that ANEES of our filter is still very credible even though position RMSE 1

7 Fig. 8: ANEES (Case ) Fig. 10: ANEES (Case ) x x Fig. 9: Position RMSE (Case ) Fig. 11: Position RMSE (Case ) is rather large. For this problem, where bearing error is dominant, the tracking performance depends mainly on ff (in fact, changing ff r =0mtoff r =00m has virtually no impact). Case The coordinates (x; y) of the target object at time zero are determined by random draws from two independent, Gaussian distributions with means -0 m and 00 m, respectively, and common standard deviation m. The target moves at a constant velocity (including direction and speed), whose components _x and _y; determined by a random draw from a Gaussian distribution, have means m/s and 0m/s, respectively, and common standard deviation 0.1 m/s. The sensor s independent measurement errors have standard deviations ff r = m and ff = 0millirad. process noise Q =0:1 I. 100 runs In this case, our filter in this short range application still has a good performance as showed in Figs. 10 and 11.This indicates the approximation of the implementation is not sensitive to the scenario. Similar observations were made for many other cases not presented here. In the simulation conducted, the standard implementation of the and conversion techniques sometimes exhibited numerical problems; we had to recourse to the numerically more robust Joseph form for the covariance computation. On the contrary, we did not experience any numerical problem with the approximate implementation of our proposed filter. The three filters have essentially the same computational complexity. 1

8 Conclusion An optimal recursive filter in the sense of BLUE has been presented for linear dynamics with nonlinear (polar) measurements. This filter operates entirely in the Cartesian coordinates but is free of the fundamental flaws of the measurement-conversion method. The simulation-based comparison of an approximate implementation of the proposed filter with two state-of-the-art measurement-conversion techniques has demonstrated the following. In terms of estimation errors and filter credibility, the proposed filter outperforms significantly the two measurement-conversion techniques in all cases tested; in particular, the proposed filter is almost always perfectly credible in that the actual estimation errors are consistent with the filter s self-assessment. By abandoning the Kalman filter framework and using optimal BLUE filter in the implementation, the good performance is achieved without increasing the computational complexity. Moreover, the optimal filter framework can be extended to the D case and the RUV coordinate systems. Extensions to these cases will be reported in a forthcoming paper. Appendix The two measurement-conversion techniques provide the following converted measurement in the Cartesian coordinates: ffl Measurement-conditioned () (referred to as fixedmeasurement approach in []): x m = r m cos m A; y m = r m sin m A C ~x = 0:R 1 T 1 R 0 cos m A C ~y = 0:R 1 T R 0 sin m A C ~x~y = cos m sin m (R 1 A R 0 A ) ffl Nested-conditioning () (referred to as additive fixed-truth approach II in []): where x m = r m cos m S; y m = r m sin m S C ~x = R 0 cos m S(S A) +0:R 1 T 1 C ~y = R 0 sin m S(S A) +0:R 1 T C ~x~y = cos m sin m (R 0 S(S A) +R 1 A ) [] D. Lerro and Y. Bar-Shalom. Tracking with Debiased Consistent Converted Measurements vs. EKF. IEEE Trans. Aerospace and Electronic Systems, AES- 9():10110, July 199. [] X. R. Li and V. P. Jilkov. A Survey of Maneuvering Target Tracking Part III: Measurement Models. In Proc. 001 SPIE Conf. on Signal and Data Processing of Small Targets, vol., pages, San Diego, CA, USA, 001. [] X. R. Li, Z. Zhao, and V. P. Jilkov. Estimator s Credibility and Its Measures. In Proc. IFAC 1th World Congress, Barcelona, Spain, July 00. [] M. Miller and O. Drummond. Coordinate Transformation Bias in Target Tracking. In Proceedings of SPIE Conference on Signal and Data Processing of Small Targets 1999, pages 09, SPIE Vol [] M. D. Miller and O. E. Drummond. Comparison of Methodologies for Mitigating Coordinate Transformation Bias in Target Tracking. In Proc. 000 SPIE Conf. on Signal and Data Processing of Small Targets, vol. 08, pages 1, Orlando, Florida, USA, April 000. [] L. Mo, X. Song, Y. Zhou, and Z. Sun. An Alternative Unbiased Consistent Converted Measurements for Target Tracking. In Proceedings of SPIE: Acquisition, Tracking, and Pointing XI, Vol. 08, pages 08 10, 199. [8] L. Mo, X. Song, Y. Zhou, Z. Sun, and Y. Bar- Shalom. Unbiased Converted Measurements for Tracking. IEEE Trans. Aerospace and Electronic Systems, AES-():1010, [9] P. Suchomski. Explicit Expressins for Debiased Statistics of D Converted Measurements. IEEE Trans. Aerospace and Electronic Systems, AES- (1):80, [10] M. Toda and R. Patel. Performance Bounds for Continuous-Time Filters in the Presence of Modeling Errors. IEEE Trans. Aerospace and Electronic Systems, AES-1():9190, Nov A = e 0:ff ; R 0 = r m ; R 1 = r m + ff r R = r m +ff r ; S =1 eff + e 0:ff T 1 = 1+e ff cos m ; T =1 e ff cos m References [1] Y. Bar-Shalom and X. R. Li. Multitarget-Multisensor Tracking: Principles and Techniques. YBS Publishing, Storrs, CT,

Fisher Information Matrix-based Nonlinear System Conversion for State Estimation

Fisher Information Matrix-based Nonlinear System Conversion for State Estimation Fisher Information Matrix-based Nonlinear System Conversion for State Estimation Ming Lei Christophe Baehr and Pierre Del Moral Abstract In practical target tracing a number of improved measurement conversion

More information

Consistent Unbiased Linear Filtering with Polar Measurements

Consistent Unbiased Linear Filtering with Polar Measurements Consistent Unbiased Linear Filtering Polar Measurements Dietrich Fränken Data Fusion Algorithms & Software EADS Deutschl GmbH D-8977 Ulm, Germany Email: dietrich.fraenken@eads.com Abstract The problem

More information

Recursive LMMSE Filtering for Target Tracking with Range and Direction Cosine Measurements

Recursive LMMSE Filtering for Target Tracking with Range and Direction Cosine Measurements Recursive Filtering for Target Tracing with Range and Direction Cosine Measurements Zhansheng Duan Yu Liu X. Rong Li Department of Electrical Engineering University of New Orleans New Orleans, LA 748,

More information

A Novel Maneuvering Target Tracking Algorithm for Radar/Infrared Sensors

A Novel Maneuvering Target Tracking Algorithm for Radar/Infrared Sensors Chinese Journal of Electronics Vol.19 No.4 Oct. 21 A Novel Maneuvering Target Tracking Algorithm for Radar/Infrared Sensors YIN Jihao 1 CUIBingzhe 2 and WANG Yifei 1 (1.School of Astronautics Beihang University

More information

Systematic Error Modeling and Bias Estimation

Systematic Error Modeling and Bias Estimation sensors Article Systematic Error Modeling and Bias Estimation Feihu Zhang * and Alois Knoll Robotics and Embedded Systems, Technische Universität München, 8333 München, Germany; knoll@in.tum.de * Correspondence:

More information

Lecture 4: Least Squares (LS) Estimation

Lecture 4: Least Squares (LS) Estimation ME 233, UC Berkeley, Spring 2014 Xu Chen Lecture 4: Least Squares (LS) Estimation Background and general solution Solution in the Gaussian case Properties Example Big picture general least squares estimation:

More information

Sliding Window Test vs. Single Time Test for Track-to-Track Association

Sliding Window Test vs. Single Time Test for Track-to-Track Association Sliding Window Test vs. Single Time Test for Track-to-Track Association Xin Tian Dept. of Electrical and Computer Engineering University of Connecticut Storrs, CT 06269-257, U.S.A. Email: xin.tian@engr.uconn.edu

More information

Distributed estimation in sensor networks

Distributed estimation in sensor networks in sensor networks A. Benavoli Dpt. di Sistemi e Informatica Università di Firenze, Italy. e-mail: benavoli@dsi.unifi.it Outline 1 An introduction to 2 3 An introduction to An introduction to In recent

More information

Track-to-Track Fusion Architectures A Review

Track-to-Track Fusion Architectures A Review Itzhack Y. Bar-Itzhack Memorial Symposium on Estimation, Navigation, and Spacecraft Control, Haifa, Israel, October 14 17, 2012 Track-to-Track Architectures A Review Xin Tian and Yaakov Bar-Shalom This

More information

Optimal Linear Estimation Fusion Part VI: Sensor Data Compression

Optimal Linear Estimation Fusion Part VI: Sensor Data Compression Optimal Linear Estimation Fusion Part VI: Sensor Data Compression Keshu Zhang X. Rong Li Peng Zhang Department of Electrical Engineering, University of New Orleans, New Orleans, L 70148 Phone: 504-280-7416,

More information

Optimal Linear Estimation Fusion Part I: Unified Fusion Rules

Optimal Linear Estimation Fusion Part I: Unified Fusion Rules 2192 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL 49, NO 9, SEPTEMBER 2003 Optimal Linear Estimation Fusion Part I: Unified Fusion Rules X Rong Li, Senior Member, IEEE, Yunmin Zhu, Jie Wang, Chongzhao

More information

The Scaled Unscented Transformation

The Scaled Unscented Transformation The Scaled Unscented Transformation Simon J. Julier, IDAK Industries, 91 Missouri Blvd., #179 Jefferson City, MO 6519 E-mail:sjulier@idak.com Abstract This paper describes a generalisation of the unscented

More information

Heterogeneous Track-to-Track Fusion

Heterogeneous Track-to-Track Fusion Heterogeneous Track-to-Track Fusion Ting Yuan, Yaakov Bar-Shalom and Xin Tian University of Connecticut, ECE Dept. Storrs, CT 06269 E-mail: {tiy, ybs, xin.tian}@ee.uconn.edu T. Yuan, Y. Bar-Shalom and

More information

4 Derivations of the Discrete-Time Kalman Filter

4 Derivations of the Discrete-Time Kalman Filter Technion Israel Institute of Technology, Department of Electrical Engineering Estimation and Identification in Dynamical Systems (048825) Lecture Notes, Fall 2009, Prof N Shimkin 4 Derivations of the Discrete-Time

More information

Maneuvering target tracking using an unbiased nearly constant heading model

Maneuvering target tracking using an unbiased nearly constant heading model Maneuvering target tracking using an unbiased nearly constant heading model P.A. Kountouriotis EEE Department Imperial College London SW7 2AZ, UK Email: pk201@imperial.ac.uk Simon Maskell QinetiQ St. Andrews

More information

in a Rao-Blackwellised Unscented Kalman Filter

in a Rao-Blackwellised Unscented Kalman Filter A Rao-Blacwellised Unscented Kalman Filter Mar Briers QinetiQ Ltd. Malvern Technology Centre Malvern, UK. m.briers@signal.qinetiq.com Simon R. Masell QinetiQ Ltd. Malvern Technology Centre Malvern, UK.

More information

Generalized Linear Minimum Mean-Square Error Estimation

Generalized Linear Minimum Mean-Square Error Estimation Generalized Linear Minimum Mean-Square Error Estimation Yu Liu and X. Rong Li Department of Electrical Engineering University of New Orleans New Orleans, LA 7148, U.S.A. Email: {lyu2, xli}@uno.edu Abstract

More information

A PCR-BIMM filter For Maneuvering Target Tracking

A PCR-BIMM filter For Maneuvering Target Tracking A PCR-BIMM filter For Maneuvering Target Tracking Jean Dezert Benjamin Pannetier Originally published as Dezert J., Pannetier B., A PCR-BIMM filter for maneuvering target tracking, in Proc. of Fusion 21,

More information

ENGR352 Problem Set 02

ENGR352 Problem Set 02 engr352/engr352p02 September 13, 2018) ENGR352 Problem Set 02 Transfer function of an estimator 1. Using Eq. (1.1.4-27) from the text, find the correct value of r ss (the result given in the text is incorrect).

More information

State Estimation for Nonlinear Systems using Restricted Genetic Optimization

State Estimation for Nonlinear Systems using Restricted Genetic Optimization State Estimation for Nonlinear Systems using Restricted Genetic Optimization Santiago Garrido, Luis Moreno, and Carlos Balaguer Universidad Carlos III de Madrid, Leganés 28911, Madrid (Spain) Abstract.

More information

X t = a t + r t, (7.1)

X t = a t + r t, (7.1) Chapter 7 State Space Models 71 Introduction State Space models, developed over the past 10 20 years, are alternative models for time series They include both the ARIMA models of Chapters 3 6 and the Classical

More information

Incorporating Track Uncertainty into the OSPA Metric

Incorporating Track Uncertainty into the OSPA Metric 14th International Conference on Information Fusion Chicago, Illinois, USA, July 5-8, 211 Incorporating Trac Uncertainty into the OSPA Metric Sharad Nagappa School of EPS Heriot Watt University Edinburgh,

More information

Lecture Note 1: Probability Theory and Statistics

Lecture Note 1: Probability Theory and Statistics Univ. of Michigan - NAME 568/EECS 568/ROB 530 Winter 2018 Lecture Note 1: Probability Theory and Statistics Lecturer: Maani Ghaffari Jadidi Date: April 6, 2018 For this and all future notes, if you would

More information

Nonlinear Estimation Techniques for Impact Point Prediction of Ballistic Targets

Nonlinear Estimation Techniques for Impact Point Prediction of Ballistic Targets Nonlinear Estimation Techniques for Impact Point Prediction of Ballistic Targets J. Clayton Kerce a, George C. Brown a, and David F. Hardiman b a Georgia Tech Research Institute, Georgia Institute of Technology,

More information

ECE531 Lecture 11: Dynamic Parameter Estimation: Kalman-Bucy Filter

ECE531 Lecture 11: Dynamic Parameter Estimation: Kalman-Bucy Filter ECE531 Lecture 11: Dynamic Parameter Estimation: Kalman-Bucy Filter D. Richard Brown III Worcester Polytechnic Institute 09-Apr-2009 Worcester Polytechnic Institute D. Richard Brown III 09-Apr-2009 1 /

More information

9 Multi-Model State Estimation

9 Multi-Model State Estimation Technion Israel Institute of Technology, Department of Electrical Engineering Estimation and Identification in Dynamical Systems (048825) Lecture Notes, Fall 2009, Prof. N. Shimkin 9 Multi-Model State

More information

NOISE ROBUST RELATIVE TRANSFER FUNCTION ESTIMATION. M. Schwab, P. Noll, and T. Sikora. Technical University Berlin, Germany Communication System Group

NOISE ROBUST RELATIVE TRANSFER FUNCTION ESTIMATION. M. Schwab, P. Noll, and T. Sikora. Technical University Berlin, Germany Communication System Group NOISE ROBUST RELATIVE TRANSFER FUNCTION ESTIMATION M. Schwab, P. Noll, and T. Sikora Technical University Berlin, Germany Communication System Group Einsteinufer 17, 1557 Berlin (Germany) {schwab noll

More information

Unified Optimal Linear Estimation Fusion Part I: Unified Models and Fusion Rules

Unified Optimal Linear Estimation Fusion Part I: Unified Models and Fusion Rules Unified Optimal Linear Estimation usion Part I: Unified Models usion Rules Rong Li Dept of Electrical Engineering University of New Orleans New Orleans, LA 70148 Phone: 04-280-7416, ax: -30 xliunoedu unmin

More information

Tracking an Accelerated Target with a Nonlinear Constant Heading Model

Tracking an Accelerated Target with a Nonlinear Constant Heading Model Tracking an Accelerated Target with a Nonlinear Constant Heading Model Rong Yang, Gee Wah Ng DSO National Laboratories 20 Science Park Drive Singapore 118230 yrong@dsoorgsg ngeewah@dsoorgsg Abstract This

More information

L06. LINEAR KALMAN FILTERS. NA568 Mobile Robotics: Methods & Algorithms

L06. LINEAR KALMAN FILTERS. NA568 Mobile Robotics: Methods & Algorithms L06. LINEAR KALMAN FILTERS NA568 Mobile Robotics: Methods & Algorithms 2 PS2 is out! Landmark-based Localization: EKF, UKF, PF Today s Lecture Minimum Mean Square Error (MMSE) Linear Kalman Filter Gaussian

More information

Sampling strong tracking nonlinear unscented Kalman filter and its application in eye tracking

Sampling strong tracking nonlinear unscented Kalman filter and its application in eye tracking Sampling strong tracking nonlinear unscented Kalman filter and its application in eye tracking Zhang Zu-Tao( 张祖涛 ) a)b) and Zhang Jia-Shu( 张家树 ) b) a) School of Mechanical Engineering, Southwest Jiaotong

More information

SELECTIVE ANGLE MEASUREMENTS FOR A 3D-AOA INSTRUMENTAL VARIABLE TMA ALGORITHM

SELECTIVE ANGLE MEASUREMENTS FOR A 3D-AOA INSTRUMENTAL VARIABLE TMA ALGORITHM SELECTIVE ANGLE MEASUREMENTS FOR A 3D-AOA INSTRUMENTAL VARIABLE TMA ALGORITHM Kutluyıl Doğançay Reza Arablouei School of Engineering, University of South Australia, Mawson Lakes, SA 595, Australia ABSTRACT

More information

Tactical Ballistic Missile Tracking using the Interacting Multiple Model Algorithm

Tactical Ballistic Missile Tracking using the Interacting Multiple Model Algorithm Tactical Ballistic Missile Tracking using the Interacting Multiple Model Algorithm Robert L Cooperman Raytheon Co C 3 S Division St Petersburg, FL Robert_L_Cooperman@raytheoncom Abstract The problem of

More information

Kalman Filter. Predict: Update: x k k 1 = F k x k 1 k 1 + B k u k P k k 1 = F k P k 1 k 1 F T k + Q

Kalman Filter. Predict: Update: x k k 1 = F k x k 1 k 1 + B k u k P k k 1 = F k P k 1 k 1 F T k + Q Kalman Filter Kalman Filter Predict: x k k 1 = F k x k 1 k 1 + B k u k P k k 1 = F k P k 1 k 1 F T k + Q Update: K = P k k 1 Hk T (H k P k k 1 Hk T + R) 1 x k k = x k k 1 + K(z k H k x k k 1 ) P k k =(I

More information

01 Probability Theory and Statistics Review

01 Probability Theory and Statistics Review NAVARCH/EECS 568, ROB 530 - Winter 2018 01 Probability Theory and Statistics Review Maani Ghaffari January 08, 2018 Last Time: Bayes Filters Given: Stream of observations z 1:t and action data u 1:t Sensor/measurement

More information

A Sufficient Comparison of Trackers

A Sufficient Comparison of Trackers A Sufficient Comparison of Trackers David Bizup University of Virginia Department of Systems and Information Engineering P.O. Box 400747 151 Engineer's Way Charlottesville, VA 22904 Donald E. Brown University

More information

CS229 Lecture notes. Andrew Ng

CS229 Lecture notes. Andrew Ng CS229 Lecture notes Andrew Ng Part X Factor analysis When we have data x (i) R n that comes from a mixture of several Gaussians, the EM algorithm can be applied to fit a mixture model. In this setting,

More information

Minimum Necessary Data Rates for Accurate Track Fusion

Minimum Necessary Data Rates for Accurate Track Fusion Proceedings of the 44th IEEE Conference on Decision Control, the European Control Conference 005 Seville, Spain, December -5, 005 ThIA0.4 Minimum Necessary Data Rates for Accurate Trac Fusion Barbara F.

More information

The Kernel-SME Filter with False and Missing Measurements

The Kernel-SME Filter with False and Missing Measurements The Kernel-SME Filter with False and Missing Measurements Marcus Baum, Shishan Yang Institute of Computer Science University of Göttingen, Germany Email: marcusbaum, shishanyang@csuni-goettingende Uwe

More information

ESTIMATOR STABILITY ANALYSIS IN SLAM. Teresa Vidal-Calleja, Juan Andrade-Cetto, Alberto Sanfeliu

ESTIMATOR STABILITY ANALYSIS IN SLAM. Teresa Vidal-Calleja, Juan Andrade-Cetto, Alberto Sanfeliu ESTIMATOR STABILITY ANALYSIS IN SLAM Teresa Vidal-Calleja, Juan Andrade-Cetto, Alberto Sanfeliu Institut de Robtica i Informtica Industrial, UPC-CSIC Llorens Artigas 4-6, Barcelona, 88 Spain {tvidal, cetto,

More information

A Comparison of the EKF, SPKF, and the Bayes Filter for Landmark-Based Localization

A Comparison of the EKF, SPKF, and the Bayes Filter for Landmark-Based Localization A Comparison of the EKF, SPKF, and the Bayes Filter for Landmark-Based Localization and Timothy D. Barfoot CRV 2 Outline Background Objective Experimental Setup Results Discussion Conclusion 2 Outline

More information

Optimal Linear Estimation Fusion Part VII: Dynamic Systems

Optimal Linear Estimation Fusion Part VII: Dynamic Systems Optimal Linear Estimation Fusion Part VII: Dynamic Systems X. Rong Li Department of Electrical Engineering, University of New Orleans New Orleans, LA 70148, USA Tel: (504) 280-7416, Fax: (504) 280-3950,

More information

Testing Estimator s Credibility Part I: Tests for MSE

Testing Estimator s Credibility Part I: Tests for MSE Testing Estimator s Credibility Part I: Tests for MSE X. Rong Li Zhanlue Zhao Department of Electrical Engineering University of New Orleans New Orleans, LA 748, USA xli@uno.edu, zhanluezhao@gmail.com

More information

A FIXED-LAG SMOOTHING SOLUTION TO OUT-OF-SEQUENCE INFORMATION FUSION PROBLEMS

A FIXED-LAG SMOOTHING SOLUTION TO OUT-OF-SEQUENCE INFORMATION FUSION PROBLEMS COMMUNICATIONS IN INFORMATION AND SYSTEMS c 2002 International Press Vol. 2, No. 4, pp. 325-348, December 2002 001 A FIXED-LAG SMOOTHING SOLUTION TO OUT-OF-SEQUENCE INFORMATION FUSION PROBLEMS SUBHASH

More information

ESTIMATION THEORY. Chapter Estimation of Random Variables

ESTIMATION THEORY. Chapter Estimation of Random Variables Chapter ESTIMATION THEORY. Estimation of Random Variables Suppose X,Y,Y 2,...,Y n are random variables defined on the same probability space (Ω, S,P). We consider Y,...,Y n to be the observed random variables

More information

1 Kalman Filter Introduction

1 Kalman Filter Introduction 1 Kalman Filter Introduction You should first read Chapter 1 of Stochastic models, estimation, and control: Volume 1 by Peter S. Maybec (available here). 1.1 Explanation of Equations (1-3) and (1-4) Equation

More information

Target tracking and classification for missile using interacting multiple model (IMM)

Target tracking and classification for missile using interacting multiple model (IMM) Target tracking and classification for missile using interacting multiple model (IMM Kyungwoo Yoo and Joohwan Chun KAIST School of Electrical Engineering Yuseong-gu, Daejeon, Republic of Korea Email: babooovv@kaist.ac.kr

More information

State-space Model. Eduardo Rossi University of Pavia. November Rossi State-space Model Financial Econometrics / 49

State-space Model. Eduardo Rossi University of Pavia. November Rossi State-space Model Financial Econometrics / 49 State-space Model Eduardo Rossi University of Pavia November 2013 Rossi State-space Model Financial Econometrics - 2013 1 / 49 Outline 1 Introduction 2 The Kalman filter 3 Forecast errors 4 State smoothing

More information

X. F. Wang, J. F. Chen, Z. G. Shi *, and K. S. Chen Department of Information and Electronic Engineering, Zhejiang University, Hangzhou , China

X. F. Wang, J. F. Chen, Z. G. Shi *, and K. S. Chen Department of Information and Electronic Engineering, Zhejiang University, Hangzhou , China Progress In Electromagnetics Research, Vol. 118, 1 15, 211 FUZZY-CONTROL-BASED PARTICLE FILTER FOR MANEUVERING TARGET TRACKING X. F. Wang, J. F. Chen, Z. G. Shi *, and K. S. Chen Department of Information

More information

A Tree Search Approach to Target Tracking in Clutter

A Tree Search Approach to Target Tracking in Clutter 12th International Conference on Information Fusion Seattle, WA, USA, July 6-9, 2009 A Tree Search Approach to Target Tracking in Clutter Jill K. Nelson and Hossein Roufarshbaf Department of Electrical

More information

EKF/UKF Maneuvering Target Tracking using Coordinated Turn Models with Polar/Cartesian Velocity

EKF/UKF Maneuvering Target Tracking using Coordinated Turn Models with Polar/Cartesian Velocity EKF/UKF Maneuvering Target Tracking using Coordinated Turn Models with Polar/Cartesian Velocity Michael Roth, Gustaf Hendeby, and Fredrik Gustafsson Dept. Electrical Engineering, Linköping University,

More information

Solutions to Homework Set #6 (Prepared by Lele Wang)

Solutions to Homework Set #6 (Prepared by Lele Wang) Solutions to Homework Set #6 (Prepared by Lele Wang) Gaussian random vector Given a Gaussian random vector X N (µ, Σ), where µ ( 5 ) T and 0 Σ 4 0 0 0 9 (a) Find the pdfs of i X, ii X + X 3, iii X + X

More information

Design of Nearly Constant Velocity Track Filters for Brief Maneuvers

Design of Nearly Constant Velocity Track Filters for Brief Maneuvers 4th International Conference on Information Fusion Chicago, Illinois, USA, July 5-8, 20 Design of Nearly Constant Velocity rack Filters for Brief Maneuvers W. Dale Blair Georgia ech Research Institute

More information

A New Nonlinear State Estimator Using the Fusion of Multiple Extended Kalman Filters

A New Nonlinear State Estimator Using the Fusion of Multiple Extended Kalman Filters 18th International Conference on Information Fusion Washington, DC - July 6-9, 2015 A New Nonlinear State Estimator Using the Fusion of Multiple Extended Kalman Filters Zhansheng Duan, Xiaoyun Li Center

More information

Kalman Filtering. Namrata Vaswani. March 29, Kalman Filter as a causal MMSE estimator

Kalman Filtering. Namrata Vaswani. March 29, Kalman Filter as a causal MMSE estimator Kalman Filtering Namrata Vaswani March 29, 2018 Notes are based on Vincent Poor s book. 1 Kalman Filter as a causal MMSE estimator Consider the following state space model (signal and observation model).

More information

Quaternion based Extended Kalman Filter

Quaternion based Extended Kalman Filter Quaternion based Extended Kalman Filter, Sergio Montenegro About this lecture General introduction to rotations and quaternions. Introduction to Kalman Filter for Attitude Estimation How to implement and

More information

The Kalman Filter. Data Assimilation & Inverse Problems from Weather Forecasting to Neuroscience. Sarah Dance

The Kalman Filter. Data Assimilation & Inverse Problems from Weather Forecasting to Neuroscience. Sarah Dance The Kalman Filter Data Assimilation & Inverse Problems from Weather Forecasting to Neuroscience Sarah Dance School of Mathematical and Physical Sciences, University of Reading s.l.dance@reading.ac.uk July

More information

Parameterized Joint Densities with Gaussian Mixture Marginals and their Potential Use in Nonlinear Robust Estimation

Parameterized Joint Densities with Gaussian Mixture Marginals and their Potential Use in Nonlinear Robust Estimation Proceedings of the 2006 IEEE International Conference on Control Applications Munich, Germany, October 4-6, 2006 WeA0. Parameterized Joint Densities with Gaussian Mixture Marginals and their Potential

More information

ECE534, Spring 2018: Solutions for Problem Set #5

ECE534, Spring 2018: Solutions for Problem Set #5 ECE534, Spring 08: s for Problem Set #5 Mean Value and Autocorrelation Functions Consider a random process X(t) such that (i) X(t) ± (ii) The number of zero crossings, N(t), in the interval (0, t) is described

More information

UCSD ECE153 Handout #34 Prof. Young-Han Kim Tuesday, May 27, Solutions to Homework Set #6 (Prepared by TA Fatemeh Arbabjolfaei)

UCSD ECE153 Handout #34 Prof. Young-Han Kim Tuesday, May 27, Solutions to Homework Set #6 (Prepared by TA Fatemeh Arbabjolfaei) UCSD ECE53 Handout #34 Prof Young-Han Kim Tuesday, May 7, 04 Solutions to Homework Set #6 (Prepared by TA Fatemeh Arbabjolfaei) Linear estimator Consider a channel with the observation Y XZ, where the

More information

ECE 541 Stochastic Signals and Systems Problem Set 9 Solutions

ECE 541 Stochastic Signals and Systems Problem Set 9 Solutions ECE 541 Stochastic Signals and Systems Problem Set 9 Solutions Problem Solutions : Yates and Goodman, 9.5.3 9.1.4 9.2.2 9.2.6 9.3.2 9.4.2 9.4.6 9.4.7 and Problem 9.1.4 Solution The joint PDF of X and Y

More information

Gaussian Processes. Le Song. Machine Learning II: Advanced Topics CSE 8803ML, Spring 2012

Gaussian Processes. Le Song. Machine Learning II: Advanced Topics CSE 8803ML, Spring 2012 Gaussian Processes Le Song Machine Learning II: Advanced Topics CSE 8803ML, Spring 01 Pictorial view of embedding distribution Transform the entire distribution to expected features Feature space Feature

More information

Previously on TT, Target Tracking: Lecture 2 Single Target Tracking Issues. Lecture-2 Outline. Basic ideas on track life

Previously on TT, Target Tracking: Lecture 2 Single Target Tracking Issues. Lecture-2 Outline. Basic ideas on track life REGLERTEKNIK Previously on TT, AUTOMATIC CONTROL Target Tracing: Lecture 2 Single Target Tracing Issues Emre Özan emre@isy.liu.se Division of Automatic Control Department of Electrical Engineering Linöping

More information

A Theoretical Overview on Kalman Filtering

A Theoretical Overview on Kalman Filtering A Theoretical Overview on Kalman Filtering Constantinos Mavroeidis Vanier College Presented to professors: IVANOV T. IVAN STAHN CHRISTIAN Email: cmavroeidis@gmail.com June 6, 208 Abstract Kalman filtering

More information

E190Q Lecture 11 Autonomous Robot Navigation

E190Q Lecture 11 Autonomous Robot Navigation E190Q Lecture 11 Autonomous Robot Navigation Instructor: Chris Clark Semester: Spring 013 1 Figures courtesy of Siegwart & Nourbakhsh Control Structures Planning Based Control Prior Knowledge Operator

More information

% error. % error in the difference between simulated and predicted error. process noises fixed at 5. observation noises: 0.05 > 0.5.

% error. % error in the difference between simulated and predicted error. process noises fixed at 5. observation noises: 0.05 > 0.5. A comparison of fixed gain IMM against two other filters Eric Derbez & Bruno Remillard Laboratoire de recherche en probabilités et statistique Université du Québec a Trois-Rivi eres Trois-Rivi eres (Québec)

More information

A Probability Review

A Probability Review A Probability Review Outline: A probability review Shorthand notation: RV stands for random variable EE 527, Detection and Estimation Theory, # 0b 1 A Probability Review Reading: Go over handouts 2 5 in

More information

Lecture 25: Review. Statistics 104. April 23, Colin Rundel

Lecture 25: Review. Statistics 104. April 23, Colin Rundel Lecture 25: Review Statistics 104 Colin Rundel April 23, 2012 Joint CDF F (x, y) = P [X x, Y y] = P [(X, Y ) lies south-west of the point (x, y)] Y (x,y) X Statistics 104 (Colin Rundel) Lecture 25 April

More information

A DARK GREY P O N T, with a Switch Tail, and a small Star on the Forehead. Any

A DARK GREY P O N T, with a Switch Tail, and a small Star on the Forehead. Any Y Y Y X X «/ YY Y Y ««Y x ) & \ & & } # Y \#$& / Y Y X» \\ / X X X x & Y Y X «q «z \x» = q Y # % \ & [ & Z \ & { + % ) / / «q zy» / & / / / & x x X / % % ) Y x X Y $ Z % Y Y x x } / % «] «] # z» & Y X»

More information

Robotics 2 Target Tracking. Kai Arras, Cyrill Stachniss, Maren Bennewitz, Wolfram Burgard

Robotics 2 Target Tracking. Kai Arras, Cyrill Stachniss, Maren Bennewitz, Wolfram Burgard Robotics 2 Target Tracking Kai Arras, Cyrill Stachniss, Maren Bennewitz, Wolfram Burgard Slides by Kai Arras, Gian Diego Tipaldi, v.1.1, Jan 2012 Chapter Contents Target Tracking Overview Applications

More information

UNIVERSITY OF OSLO Fysisk Institutt. Tracking of an Airplane using EKF and SPF. Master thesis. Rune Jansberg

UNIVERSITY OF OSLO Fysisk Institutt. Tracking of an Airplane using EKF and SPF. Master thesis. Rune Jansberg UNIVERSITY OF OSLO Fysisk Institutt Tracking of an Airplane using EKF and SPF Master thesis Rune Jansberg May 31, 2010 Problem Description Tracking of an Airplane using EKF and SPF. The theory behind

More information

State Estimation by IMM Filter in the Presence of Structural Uncertainty 1

State Estimation by IMM Filter in the Presence of Structural Uncertainty 1 Recent Advances in Signal Processing and Communications Edited by Nios Mastorais World Scientific and Engineering Society (WSES) Press Greece 999 pp.8-88. State Estimation by IMM Filter in the Presence

More information

An algorithm for robust fitting of autoregressive models Dimitris N. Politis

An algorithm for robust fitting of autoregressive models Dimitris N. Politis An algorithm for robust fitting of autoregressive models Dimitris N. Politis Abstract: An algorithm for robust fitting of AR models is given, based on a linear regression idea. The new method appears to

More information

Multiple Model Cardinalized Probability Hypothesis Density Filter

Multiple Model Cardinalized Probability Hypothesis Density Filter Multiple Model Cardinalized Probability Hypothesis Density Filter Ramona Georgescu a and Peter Willett a a Elec. and Comp. Engineering Department, University of Connecticut, Storrs, CT 06269 {ramona, willett}@engr.uconn.edu

More information

Lecture Outline. Target Tracking: Lecture 7 Multiple Sensor Tracking Issues. Multi Sensor Architectures. Multi Sensor Architectures

Lecture Outline. Target Tracking: Lecture 7 Multiple Sensor Tracking Issues. Multi Sensor Architectures. Multi Sensor Architectures Lecture Outline Target Tracing: Lecture 7 Multiple Sensor Tracing Issues Umut Orguner umut@metu.edu.tr room: EZ-12 tel: 4425 Department of Electrical & Electronics Engineering Middle East Technical University

More information

A New Nonlinear Filtering Method for Ballistic Target Tracking

A New Nonlinear Filtering Method for Ballistic Target Tracking th International Conference on Information Fusion Seattle, WA, USA, July 6-9, 9 A New Nonlinear Filtering Method for Ballistic arget racing Chunling Wu Institute of Electronic & Information Engineering

More information

Name of the Student: Problems on Discrete & Continuous R.Vs

Name of the Student: Problems on Discrete & Continuous R.Vs Engineering Mathematics 05 SUBJECT NAME : Probability & Random Process SUBJECT CODE : MA6 MATERIAL NAME : University Questions MATERIAL CODE : JM08AM004 REGULATION : R008 UPDATED ON : Nov-Dec 04 (Scan

More information

Extended Object and Group Tracking: A Comparison of Random Matrices and Random Hypersurface Models

Extended Object and Group Tracking: A Comparison of Random Matrices and Random Hypersurface Models Extended Object and Group Tracking: A Comparison of Random Matrices and Random Hypersurface Models Marcus Baum, Michael Feldmann, Dietrich Fränken, Uwe D. Hanebeck, and Wolfgang Koch Intelligent Sensor-Actuator-Systems

More information

A new unscented Kalman filter with higher order moment-matching

A new unscented Kalman filter with higher order moment-matching A new unscented Kalman filter with higher order moment-matching KSENIA PONOMAREVA, PARESH DATE AND ZIDONG WANG Department of Mathematical Sciences, Brunel University, Uxbridge, UB8 3PH, UK. Abstract This

More information

Introduction to Unscented Kalman Filter

Introduction to Unscented Kalman Filter Introduction to Unscented Kalman Filter 1 Introdution In many scientific fields, we use certain models to describe the dynamics of system, such as mobile robot, vision tracking and so on. The word dynamics

More information

Model-Based Diagnosis of Chaotic Vibration Signals

Model-Based Diagnosis of Chaotic Vibration Signals Model-Based Diagnosis of Chaotic Vibration Signals Ihab Wattar ABB Automation 29801 Euclid Ave., MS. 2F8 Wickliffe, OH 44092 and Department of Electrical and Computer Engineering Cleveland State University,

More information

Using the Kalman Filter to Estimate the State of a Maneuvering Aircraft

Using the Kalman Filter to Estimate the State of a Maneuvering Aircraft 1 Using the Kalman Filter to Estimate the State of a Maneuvering Aircraft K. Meier and A. Desai Abstract Using sensors that only measure the bearing angle and range of an aircraft, a Kalman filter is implemented

More information

A Comparison of Multiple-Model Target Tracking Algorithms

A Comparison of Multiple-Model Target Tracking Algorithms University of New Orleans ScholarWors@UNO University of New Orleans heses and Dissertations Dissertations and heses 1-17-4 A Comparison of Multiple-Model arget racing Algorithms Ryan Pitre University of

More information

EEL 5544 Noise in Linear Systems Lecture 30. X (s) = E [ e sx] f X (x)e sx dx. Moments can be found from the Laplace transform as

EEL 5544 Noise in Linear Systems Lecture 30. X (s) = E [ e sx] f X (x)e sx dx. Moments can be found from the Laplace transform as L30-1 EEL 5544 Noise in Linear Systems Lecture 30 OTHER TRANSFORMS For a continuous, nonnegative RV X, the Laplace transform of X is X (s) = E [ e sx] = 0 f X (x)e sx dx. For a nonnegative RV, the Laplace

More information

A NEW FORMULATION OF IPDAF FOR TRACKING IN CLUTTER

A NEW FORMULATION OF IPDAF FOR TRACKING IN CLUTTER A NEW FRMULATIN F IPDAF FR TRACKING IN CLUTTER Jean Dezert NERA, 29 Av. Division Leclerc 92320 Châtillon, France fax:+33146734167 dezert@onera.fr Ning Li, X. Rong Li University of New rleans New rleans,

More information

Ross Bettinger, Analytical Consultant, Seattle, WA

Ross Bettinger, Analytical Consultant, Seattle, WA ABSTRACT DYNAMIC REGRESSION IN ARIMA MODELING Ross Bettinger, Analytical Consultant, Seattle, WA Box-Jenkins time series models that contain exogenous predictor variables are called dynamic regression

More information

A Study of Covariances within Basic and Extended Kalman Filters

A Study of Covariances within Basic and Extended Kalman Filters A Study of Covariances within Basic and Extended Kalman Filters David Wheeler Kyle Ingersoll December 2, 2013 Abstract This paper explores the role of covariance in the context of Kalman filters. The underlying

More information

A comparative study of multiple-model algorithms for maneuvering target tracking

A comparative study of multiple-model algorithms for maneuvering target tracking A comparative study of multiple-model algorithms for maneuvering target tracing Ryan R. Pitre Vesselin P. Jilov X. Rong Li Department of Electrical Engineering University of New Orleans New Orleans, LA

More information

Tracking of Extended Object or Target Group Using Random Matrix Part I: New Model and Approach

Tracking of Extended Object or Target Group Using Random Matrix Part I: New Model and Approach Tracing of Extended Object or Target Group Using Random Matrix Part I: New Model and Approach Jian Lan Center for Information Engineering Science Research School of Electronics and Information Engineering

More information

State-space Model. Eduardo Rossi University of Pavia. November Rossi State-space Model Fin. Econometrics / 53

State-space Model. Eduardo Rossi University of Pavia. November Rossi State-space Model Fin. Econometrics / 53 State-space Model Eduardo Rossi University of Pavia November 2014 Rossi State-space Model Fin. Econometrics - 2014 1 / 53 Outline 1 Motivation 2 Introduction 3 The Kalman filter 4 Forecast errors 5 State

More information

Kalman Filter. Man-Wai MAK

Kalman Filter. Man-Wai MAK Kalman Filter Man-Wai MAK Dept. of Electronic and Information Engineering, The Hong Kong Polytechnic University enmwmak@polyu.edu.hk http://www.eie.polyu.edu.hk/ mwmak References: S. Gannot and A. Yeredor,

More information

State Estimation of Linear and Nonlinear Dynamic Systems

State Estimation of Linear and Nonlinear Dynamic Systems State Estimation of Linear and Nonlinear Dynamic Systems Part I: Linear Systems with Gaussian Noise James B. Rawlings and Fernando V. Lima Department of Chemical and Biological Engineering University of

More information

Randomized Unscented Kalman Filter in Target Tracking

Randomized Unscented Kalman Filter in Target Tracking Randomized Unscented Kalman Filter in Target Tracking Ondřej Straka, Jindřich Duník and Miroslav Šimandl Department of Cybernetics, Faculty of Applied Sciences, University of West Bohemia, Univerzitní

More information

Parameter Estimation

Parameter Estimation Parameter Estimation Tuesday 9 th May, 07 4:30 Consider a system whose response can be modeled by R = M (Θ) where Θ is a vector of m parameters. We take a series of measurements, D (t) where t represents

More information

Ground Moving Target Parameter Estimation for Stripmap SAR Using the Unscented Kalman Filter

Ground Moving Target Parameter Estimation for Stripmap SAR Using the Unscented Kalman Filter Ground Moving Target Parameter Estimation for Stripmap SAR Using the Unscented Kalman Filter Bhashyam Balaji, Christoph Gierull and Anthony Damini Radar Sensing and Exploitation Section, Defence Research

More information

UCSD ECE 153 Handout #46 Prof. Young-Han Kim Thursday, June 5, Solutions to Homework Set #8 (Prepared by TA Fatemeh Arbabjolfaei)

UCSD ECE 153 Handout #46 Prof. Young-Han Kim Thursday, June 5, Solutions to Homework Set #8 (Prepared by TA Fatemeh Arbabjolfaei) UCSD ECE 53 Handout #46 Prof. Young-Han Kim Thursday, June 5, 04 Solutions to Homework Set #8 (Prepared by TA Fatemeh Arbabjolfaei). Discrete-time Wiener process. Let Z n, n 0 be a discrete time white

More information

Solutions to Homework Set #5 (Prepared by Lele Wang) MSE = E [ (sgn(x) g(y)) 2],, where f X (x) = 1 2 2π e. e (x y)2 2 dx 2π

Solutions to Homework Set #5 (Prepared by Lele Wang) MSE = E [ (sgn(x) g(y)) 2],, where f X (x) = 1 2 2π e. e (x y)2 2 dx 2π Solutions to Homework Set #5 (Prepared by Lele Wang). Neural net. Let Y X + Z, where the signal X U[,] and noise Z N(,) are independent. (a) Find the function g(y) that minimizes MSE E [ (sgn(x) g(y))

More information

The Kalman Filter. An Algorithm for Dealing with Uncertainty. Steven Janke. May Steven Janke (Seminar) The Kalman Filter May / 29

The Kalman Filter. An Algorithm for Dealing with Uncertainty. Steven Janke. May Steven Janke (Seminar) The Kalman Filter May / 29 The Kalman Filter An Algorithm for Dealing with Uncertainty Steven Janke May 2011 Steven Janke (Seminar) The Kalman Filter May 2011 1 / 29 Autonomous Robots Steven Janke (Seminar) The Kalman Filter May

More information

Simultaneous state and input estimation of non-linear process with unknown inputs using particle swarm optimization particle filter (PSO-PF) algorithm

Simultaneous state and input estimation of non-linear process with unknown inputs using particle swarm optimization particle filter (PSO-PF) algorithm Simultaneous state and input estimation of non-linear process with unknown inputs using particle swarm optimization particle filter (PSO-PF) algorithm Mohammad A. Khan, CSChe 2016 Outlines Motivations

More information

Tracking of Extended Objects and Group Targets using Random Matrices A New Approach

Tracking of Extended Objects and Group Targets using Random Matrices A New Approach Tracing of Extended Objects and Group Targets using Random Matrices A New Approach Michael Feldmann FGAN Research Institute for Communication, Information Processing and Ergonomics FKIE D-53343 Wachtberg,

More information