Approximate Dynamic Programming in Rail Operations

Size: px
Start display at page:

Download "Approximate Dynamic Programming in Rail Operations"

Transcription

1 Approximate Dynamic Programming in Rail Operations June, 2007 Tristan VI Phuket Island, Thailand Warren Powell Belgacem Bouzaiene-Ayari CASTLE Laboratory Princeton University Warren B. Powell, Princeton University

2 Warren B. Powell 2007 Slide 2

3 Warren B. Powell 2007 Slide 3

4 Warren B. Powell 2007 Slide 4

5 Warren B. Powell 2007 Slide 5

6 A bit of history: History» Developed first locomotive optimization model based on the principles of approximate dynamic programming (the optimizing-simulator back then).» Implementation at Norfolk Southern Railroad as both a strategic and an operational planning system» 2002: New locomotive manager: Why in the world would anyone do a project like this? » Redeveloped modeling library» Advances in approximate dynamic programming» Used to develop ADP-based optimization model for car distribution » Locomotive modeling project restarted using new modeling library, new algorithms, new implementation strategy. Warren B. Powell 2007 Slide 6

7 Outline The locomotive planning problem Approximate dynamic programming Solving the subproblem Decomposition strategy Implementation Warren B. Powell 2007 Slide 7

8 Outline The locomotive planning problem Approximate dynamic programming Solving the subproblem Decomposition strategy Implementation Warren B. Powell 2007 Slide 8

9 Warren B. Powell 2007 Slide 9 Optimization models Normally, we would formulate a big optimization problem: 0 subject to : min = + = k t t k k t k t k t k t k t k t k T t k t t x x u x b x B x A x c (0,1) k x t Integer!

10 Optimization models Multicommodity flow formulation Warren B. Powell 2007 Slide 10

11 Optimization models Multicommodity flow formulation Warren B. Powell 2007 Slide 11

12 The challenge Real-world issues: The future is uncertain» Trains are added (and dropped) with 1-3 days notice.» Tonnage per train is not known until the last minute.» Equipment fails.» Trains arrive late. Locomotive assignments are complex» Locomotives are bundled in consists there is a penalty for breaking a consist. This requires that every locomotive be modeled individually.» Leader locomotives The lead locomotive has to have certain characteristics ranging from bulletproof glass to flush toilets.» Shop routing We need to route power toward shops for maintenance. Data is not perfect» Railroads are notorious for incomplete and imperfect data.» There are multiple explanations when the model does not behave as we would expect. Warren B. Powell 2007 Slide 12

13 The challenge Competing methodologies: Deterministic optimization» Problem is NP-complete» Heuristics provide high-quality overall solutions, but can produce quirky solutions when evaluated up close.» Puts equal weight on decisions now and in the future.» Models here and now and the future at the same level of detail. Simulation» Able to handle a very high level of detail, but» Does not attempt to provide the best possible solution.» Suffers from complex rules needed to make decisions. Stochastic programming» Explodes problem size Dynamic programming/markov decision processes» You have to be kidding Warren B. Powell 2007 Slide 13

14 Outline The locomotive planning problem Approximate dynamic programming Solving the subproblem Decomposition strategy Implementation Warren B. Powell 2007 Slide 14

15 Approximate dynamic programming The challenge of dynamic programming: ( { }) V ( S ) = max C ( S, x ) + E V ( S ) S t t t t t t+ 1 t+ 1 t x X Three curses Problem: Curse of dimensionality State space Outcome space Action space (feasible region) Warren B. Powell 2007 Slide 15

16 Action State Use weather report Do not use weather report Information Forecast rain.1 Forecast cloudy.3 Forecast sunny.6 Schedule game Cancel game State Action Schedule game Cancel game Schedule game Cancel game Schedule game Cancel game Rain.2 -$2000 Clouds.3 $1000 Sun.5 $5000 Rain.2 -$200 Clouds.3 -$200 Sun.5 -$200 Information Rain.8 -$2000 Clouds.2 $1000 Sun.0 $5000 Rain.8 -$200 Clouds.2 -$200 Sun.0 -$200 Rain.1 -$2000 Clouds.5 $1000 Sun.4 $5000 Rain.1 -$200 Clouds.5 -$200 Sun.4 -$200 Rain.1 -$2000 Clouds.2 $1000 Sun.7 $5000 Rain.1 -$200 Clouds.2 -$200 Sun.7 -$200 - Decision nodes - Outcome nodes

17 Transition function Traditional modeling of dynamics S = S ( S, x, W ) M t+ 1 t t t+ 1 System model Transition function New information Decision Current state of the system New state Warren B. Powell 2007 Slide 17

18 State variables New concept: The pre-decision state variable:» S t = The information required to make a decision x t» Same as a decision node in a decision tree. The post-decision state variable:» x S t = The state of what we know immediately after we make a decision.» Same as an outcome node in a decision tree. Warren B. Powell 2007 Slide 18

19 State variables Breaking down the system dynamics: Instead of modeling from pre- to pre- S = S ( S, x, W ) M t+ 1 t t t+ 1 we use one function to go from pre- to post- S = S ( S, x ) x M, x t t t with another function from post- to pre- S = S ( S, W ) MW, x t+ 1 t t+ 1 Warren B. Powell 2007 Slide 19

20 State variables Pre-decision: resources and demands S = ( R, D ) t t t x t Warren B. Powell 2007 Slide 20

21 Post-decision: State variables S = S ( S, x ) x M, x t t t Warren B. Powell 2007 Slide 21

22 New information: State variables x S t S = S ( S, W ) MW, x t+ 1 t t+ 1 W = ( Rˆ, Dˆ ) t+ 1 t+ 1 t+ 1 Warren B. Powell 2007 Slide 22

23 New pre-decision: State variables S t + 1 Warren B. Powell 2007 Slide 23

24 State variables Pre- and post-decision attributes for a discrete resource: t = 40 Pre-decision t = 40 Decision t = 40 Post-decision t = 50 Pre-decision City ETA Equip Dallas 41.2 Good Chicago - - Chicago 54.7 Good Chicago 56.2 Repair Warren B. Powell 2007 Slide 24

25 State variables Bellman s equations broken into stages: Optimization problem (making the decision): ( x( M, x )) V ( S ) = max C ( S, x ) + V S ( S, x ) t t x t t t t t t t» Note: this problem is deterministic! Schedule game Cancel game Simulation problem (the effect of exogenous information): {, } V ( S ) = E V ( S ( S, W )) S x x M W x x t t t t t t» Need to compute expectation. Rain.8 -$2000 Clouds.2 $1000 Sun.0 $5000 Challenge: What is ( M, x (, )) V S S x x t t t t Warren B. Powell 2007 Slide 25

26 Our general algorithm Step 1: Start with a post-decision state Step 2: Obtain Monte Carlo sample of and compute the next pre-decision state: Step 3: Solve the deterministic optimization using an approximate value function: ( ) n n n 1 M, x n vˆ t = max x Ct( St, xt) + Vt ( S ( St, xt) ) n to obtain. x t (, ( ) ) 1 S = S S W ω n M, W x, n n t t t, S x n t ( 1 n Wt ω ) Step 4: Update the value function approximation n x, n n 1 x, n n t 1( t 1) = (1 αn 1) t 1 ( t 1) + αn 1ˆ t V S V S v Simulation Optimization Statistics Step 5: Find the next post-decision state: S = S ( S, x ) xn, M, x n n t t t vˆ updates V ( S ) x t t 1 t 1 Warren B. Powell 2007 Slide 26

27 Iterative learning t Warren B. Powell 2007 Slide 27

28 Iterative learning Warren B. Powell 2007 Slide 28

29 Iterative learning Warren B. Powell 2007 Slide 29

30 Value function approximations Value function approximations: Linear (in the resource state): x V ( R ) = v R x t t ta ta a A Piecewise linear, separable: x x Vt ( Rt ) = Vta ( Rta ) a A Indexed PWL separable: ( x ) x Vt( Rt ) = Vta Rta ( featurest) a A Best when assets are complex, R ta which means that is small (typically 0 or 1). Best when assets are simple, R ta which means that may be larger. Helps to capture dependencies. Warren B. Powell 2007 Slide 30

31 Value function approximations Value function approximations: Ridge regression (Klabjan and Adelman) Benders cuts (from stochastic programming) V ( R ) t t ( ) x V ( R ) = V R R = θ R t t tf tf tf fa ta f F a A f x 1 x 0 What worked: nested separable approximations Warren B. Powell 2007 Slide 31

32 Coming in September, 2007?? Warren B. Powell 2007 Slide 32

33 Outline The locomotive planning problem Approximate dynamic programming Solving the subproblem Decomposition strategy Implementation Warren B. Powell 2007 Slide 33

34 Solving the subproblem Elements of the problem Here and now» Assigning the right number, and the right types of locomotives to trains.» Must consider variety of operational constraints Consist breakup Leader locomotives Future» We need to move power now to serve needs in the future» Future train movements consist of: Scheduled trains (but uncertain tonnages) Unscheduled trains We have to consider» Getting locomotives to shop» Random travel times, equipment failures Warren B. Powell 2007 Slide 34

35 Solving the subproblem Atlanta Locomotives Trains Yards Baltimore Jacksonville Warren B. Powell 2007 Slide 35

36 Solving the subproblem Horsepower Locomotives Consist-breakup costs Shop routing bonuses/ penalties Leader logic Warren B. Powell 2007 Slide 36

37 Solving the subproblem Horsepower Locomotives Locomotives coming in on same train Warren B. Powell 2007 Slide 37

38 Solving the subproblem Train reward function Horsepower Locomotives Warren B. Powell 2007 Slide 38

39 Solving the subproblem The train reward function Train reward Minimum Goal Overpowering Power Warren B. Powell 2007 Slide 39

40 Solving the subproblem Horsepower Locomotives Train may need horsepower. Solutions: = = = Warren B. Powell 2007 Slide 40

41 Solving the subproblem Horsepower Locomotives Train may need horsepower. Solutions: = = = Warren B. Powell 2007 Slide 41

42 Solving the subproblem Horsepower Locomotives Locomotive buckets Warren B. Powell 2007 Slide 42

43 Warren B. Powell 2007 Slide 43

44 Warren B. Powell 2007 Slide 44

45 Warren B. Powell 2007 Slide 45

46 Warren B. Powell 2007 Slide 46

47 Linear value function approximations 1990 s linear value function approximations Objective function Total reward Iteration Warren B. Powell 2007 Slide 47

48 Measuring performance Model vs. history (August, 2000) Percent History Model Setouts Swaps Nonpreferred consists Underpowered Overpowered Warren B. Powell 2007 Slide 48

49 Results from the 1990 s 1990 s Solving locomotive assignment problem using a hybrid LP-relaxation and local search heuristic Value of locomotives in the future we estimated using linear functions Slope of the value function were estimated using the dual variable from the LP relaxation. Implemented in production at Norfolk Southern Weaknesses: Linear value function approximations could be unstable. Local search heuristic would occasionally produce anomalies.» The fact that the solution was not as good was irrelevant.» Anomalies reduced confidence in the model.» Produced diagnostic problems, since it was not easy to identify why an odd locomotive assignment was due to data problem, model problem, coding issue, or the heuristic. Warren B. Powell 2007 Slide 49

50 The challenge That s not very good Warren B. Powell 2007 Slide 50

51 Solving the subproblem Horsepower Locomotives Locomotive buckets Warren B. Powell 2007 Slide 51

52 Solving the subproblem Horsepower Locomotives Locomotive buckets Warren B. Powell 2007 Slide 52

53 Status in : Local search heuristic has been replaced with Cplex.» We have no trouble solving the IP to optimality, without sacrificing our ability to handle all the operational constraints. Value function approximations:» First replaced linear with piecewise linear separable Separate PWL value function for each type of locomotive at each location. Works much better than linear, but not well enough. Occasionally would move 5 locomotives to a location that needed 4. Introduced nested, separable piecewise linear value function approximation. Warren B. Powell 2007 Slide 53

54 Nested separable nonlinear Nested separable, nonlinear The value of six-axle high-adhesion locomotives in Atlanta The value of locomotives in Atlanta Warren B. Powell 2007 Slide 54

55 Nested separable nonlinear n R t n R t,(6 HL, ATL ) n R t,(4 HL, ATL ), xn R t ( ) F R t n R t,(6 HH, ATL ) n R t,(6 HL, JAX ) n R t,(4 HL, JAX ) n R t,(6 HH, JAX ) Warren B. Powell 2007 Slide 55

56 Nested separable nonlinear n R t, xn R t ( ) F R t vˆn,(6 t HL, ATL ) vˆn,(4 t HL, ATL ) vˆn,(6 t HH, ATL ) ˆn,(6, ) v t HL JAX ˆn,(4, ) v t HL JAX ˆn,(6, ) v t HH JAX Warren B. Powell 2007 Slide 56

57 Updating the value function approximation Estimate the gradient at n R t ˆn ν t,(6 HL, ATL ) ( ) F R t n R t,(6 HL, ATL ) Warren B. Powell 2007 Slide 57

58 Updating the value function approximation Update the value function at xn, Rt 1 V ( R ) n 1 x t 1 t 1 ˆn ν t,(6 HL, ATL ) ( ) F R t xn, n Rt 1,(6 HL, ATL) R t,(6 HL, ATL ) Warren B. Powell 2007 Slide 58

59 Updating the value function approximation Update the value function at xn, Rt 1 V ( R ) n 1 x t 1 t 1 ˆn ν t,(6 HL, ATL ) xn, Rt 1,(6 HL, ATL) Warren B. Powell 2007 Slide 59

60 Updating the value function approximation Update the value function at xn, Rt 1 V ( R ) n 1 x t 1 t 1 V ( R ) n x t 1 t 1 xn, Rt 1,(6 HL, ATL) Warren B. Powell 2007 Slide 60

61 Nonlinear VFA 2007 Nested, separable nonlinear approximations Objective Objective Iteration Warren B. Powell 2007 Slide 61

62 Outline The locomotive planning problem Approximate dynamic programming Solving the subproblem Decomposition strategy Implementation Warren B. Powell 2007 Slide 62

63 Decomposition strategy Spatial decomposition alternatives: The entire company» Simultaneously assign locomotives to trains across the entire company at a point in time.» Makes it possible to assign power to trains in different locations. One terminal at a time» This was the strategy when we first used linear approximations.» Reflects tendency of dispatchers to handle one yard at a time.» Creates problems when close-by terminals are managed jointly. Regions» Decompose the company the way it is actually managed. Warren B. Powell 2007 Slide 63

64 Decomposition strategy Option 1: Optimize over entire company at each point in time t Warren B. Powell 2007 Slide 64

65 Decomposition strategy Option 2: Decompose by terminal t Warren B. Powell 2007 Slide 65

66 Decomposition strategy Option 2: Decompose by terminal t Warren B. Powell 2007 Slide 66

67 Decomposition strategy Option 2: Decompose by terminal t Warren B. Powell 2007 Slide 67

68 D q D q D q D q D q D q Norfolk Southern Warren B. Powell 2007 Slide 68

69 Warren B. Powell 2007 Slide 69

70 Decomposition strategy Option 3: Decompose by desk (region) t Warren B. Powell 2007 Slide 70

71 Outline The locomotive planning problem Approximate dynamic programming Solving the subproblem Decomposition strategy Implementation Warren B. Powell 2007 Slide 71

72 Implementation strategy Applications: Strategic planning» What is the impact of changes in fleet size and mix on train delay?» How do changes in shop locations affect maintenance routing?» How do changes in schedule affect train delay for a given locomotive fleet?» How do changes in operating policies affect performance? Tactical planning» How much power will you have at each terminal 1, 2, 3 days out?» Where do you anticipate being short power? Real-time planning» What train should a locomotive be assigned to in order to get it to shop on time? Warren B. Powell 2007 Slide 72

73 Strategic planning One supersource node for each type of locomotive Warren B. Powell 2007 Slide 73

74 Strategic planning One supersource node for each type of locomotive Warren B. Powell 2007 Slide 74

75 Strategic planning One supersource node for each type of locomotive Warren B. Powell 2007 Slide 75

76 Strategic planning One supersource node for each type of locomotive Warren B. Powell 2007 Slide 76

77 Strategic planning One supersource node for each type of locomotive Warren B. Powell 2007 Slide 77

78 Strategic planning 99.5 percent train coverage on a historical dataset Coverage Coverage Iteration Warren B. Powell 2007 Slide 78

79 Strategic planning Total train delay (mins) Change in fleet size Warren B. Powell 2007 Slide 79

80 Strategic planning Total train delay (mins) Change in fleet size Warren B. Powell 2007 Slide 80

81 Strategic planning Fleet sizing system: Status» Fleet sizing system has been delivered.» Full laboratory functionality.» Hope to have user acceptance by September, Features» Provides tradeoff between fleet size and train delay» Sensitive to train schedule, operating policies, transit reliability, shop location, fleet mix, Implementation» Runs as stand-alone system» Requires network, train schedule, business rules and parameters Warren B. Powell 2007 Slide 81

82 Operational forecasting Operational forecasting system Initial inventories Warren B. Powell 2007 Slide 82

83 Operational forecasting Operational forecasting system Initial inventories Warren B. Powell 2007 Slide 83

84 Operational forecasting Operational forecasting system: Status» Extensive calibration as part of the fleet sizing system.» Project will start fall, Features» Updates every 1-2 minutes.» Models locomotives and trains at a high level of detail» Able to incorporate uncertainty into forecasts of train movements, tonnages and transit times» Uses same core model as the strategic planning system Implementation» Not yet started.» Sensitive issues: Accuracy of locomotive snapshot Ability to model unscheduled trains Balancing user expectations for the accuracy of the forecast with the realities of messy railroad operations Warren B. Powell 2007 Slide 84

85 Real-time locomotive assignment Real-time system: Status» Requires reoptimizing a single subproblem in real-time» Project will start summer, Features» Estimated reoptimization time for one subproblem < 1 second.» Responds in real-time to user overrides» Subproblem talks to operational planning model Implementation» Not yet started.» Sensitive issues: Tight integration with existing user interface Warren B. Powell 2007 Slide 85

86 pf Warren B. Powell 2007 Slide 86

The Dynamic Energy Resource Model

The Dynamic Energy Resource Model The Dynamic Energy Resource Model Group Peer Review Committee Lawrence Livermore National Laboratories July, 2007 Warren Powell Alan Lamont Jeffrey Stewart Abraham George 2007 Warren B. Powell, Princeton

More information

Approximate Dynamic Programming: Solving the curses of dimensionality

Approximate Dynamic Programming: Solving the curses of dimensionality Approximate Dynamic Programming: Solving the curses of dimensionality Informs Computing Society Tutorial October, 2008 Warren Powell CASTLE Laboratory Princeton University http://www.castlelab.princeton.edu

More information

The Optimizing-Simulator: Merging Optimization and Simulation Using Approximate Dynamic Programming

The Optimizing-Simulator: Merging Optimization and Simulation Using Approximate Dynamic Programming The Optimizing-Simulator: Merging Optimization and Simulation Using Approximate Dynamic Programming Winter Simulation Conference December 5, 2005 Warren Powell CASTLE Laboratory Princeton University http://www.castlelab.princeton.edu

More information

We present a general optimization framework for locomotive models that captures different levels of detail,

We present a general optimization framework for locomotive models that captures different levels of detail, Articles in Advance, pp. 1 24 ISSN 0041-1655 (print) ISSN 1526-5447 (online) http://dx.doi.org/10.1287/trsc.2014.0536 2014 INFORMS From Single Commodity to Multiattribute Models for Locomotive Optimization:

More information

Approximate Dynamic Programming for High Dimensional Resource Allocation Problems

Approximate Dynamic Programming for High Dimensional Resource Allocation Problems Approximate Dynamic Programming for High Dimensional Resource Allocation Problems Warren B. Powell Abraham George Belgacem Bouzaiene-Ayari Hugo P. Simao Department of Operations Research and Financial

More information

Tutorial: Stochastic Optimization in Energy

Tutorial: Stochastic Optimization in Energy Tutorial: Stochastic Optimization in Energy FERC, Washington, D.C. August 6, 2014 Warren B. Powell CASTLE Labs Princeton University http://www.castlelab.princeton.edu Warren B. Powell, 2014 Slide 1 Mission

More information

Approximate Dynamic Programming in Transportation and Logistics: A Unified Framework

Approximate Dynamic Programming in Transportation and Logistics: A Unified Framework Approximate Dynamic Programming in Transportation and Logistics: A Unified Framework Warren B. Powell, Hugo P. Simao and Belgacem Bouzaiene-Ayari Department of Operations Research and Financial Engineering

More information

Dynamic Models for Freight Transportation

Dynamic Models for Freight Transportation Dynamic Models for Freight Transportation Warren B. Powell Belgacem Bouzaiene-Ayari Hugo P. Simao Department of Operations Research and Financial Engineering Princeton University August 13, 2003 Abstract

More information

Approximate Dynamic Programming in Transportation and Logistics: A Unified Framework

Approximate Dynamic Programming in Transportation and Logistics: A Unified Framework Approximate Dynamic Programming in Transportation and Logistics: A Unified Framework Warren B. Powell, Hugo P. Simao and Belgacem Bouzaiene-Ayari Department of Operations Research and Financial Engineering

More information

Some Fixed-Point Results for the Dynamic Assignment Problem

Some Fixed-Point Results for the Dynamic Assignment Problem Some Fixed-Point Results for the Dynamic Assignment Problem Michael Z. Spivey Department of Mathematics and Computer Science Samford University, Birmingham, AL 35229 Warren B. Powell Department of Operations

More information

Proper Security Criteria Determination in a Power System with High Penetration of Renewable Resources

Proper Security Criteria Determination in a Power System with High Penetration of Renewable Resources Proper Security Criteria Determination in a Power System with High Penetration of Renewable Resources Mojgan Hedayati, Kory Hedman, and Junshan Zhang School of Electrical, Computer, and Energy Engineering

More information

Using Static Flow Patterns in Time-Staged Resource Allocation Problems

Using Static Flow Patterns in Time-Staged Resource Allocation Problems Using Static Flow Patterns in Time-Staged Resource Allocation Problems Arun Marar Warren B. Powell Hugo P. Simão Department of Operations Research and Financial Engineering, Princeton University, Princeton,

More information

Dynamic Models for Freight Transportation

Dynamic Models for Freight Transportation Dynamic Models for Freight Transportation Warren B. Powell Belgacem Bouzaiene-Ayari Hugo P. Simao Department of Operations Research and Financial Engineering Princeton University September 14, 2005 Abstract

More information

Stochastic Programming in Transportation and Logistics

Stochastic Programming in Transportation and Logistics Stochastic Programming in Transportation and Logistics Warren B. Powell and Huseyin Topaloglu Department of Operations Research and Financial Engineering, Princeton University, Princeton, NJ 08544 Abstract

More information

Stochastic Integer Programming

Stochastic Integer Programming IE 495 Lecture 20 Stochastic Integer Programming Prof. Jeff Linderoth April 14, 2003 April 14, 2002 Stochastic Programming Lecture 20 Slide 1 Outline Stochastic Integer Programming Integer LShaped Method

More information

A Benders decomposition method for locating stations in a one-way electric car sharing system under demand uncertainty

A Benders decomposition method for locating stations in a one-way electric car sharing system under demand uncertainty A Benders decomposition method for locating stations in a one-way electric car sharing system under demand uncertainty Hatice Calik 1,2 and Bernard Fortz 1,3 1 Department of Computer Science, Université

More information

Multi-Area Stochastic Unit Commitment for High Wind Penetration

Multi-Area Stochastic Unit Commitment for High Wind Penetration Multi-Area Stochastic Unit Commitment for High Wind Penetration Workshop on Optimization in an Uncertain Environment Anthony Papavasiliou, UC Berkeley Shmuel S. Oren, UC Berkeley March 25th, 2011 Outline

More information

Duality in Robust Dynamic Programming: Pricing Convertibles, Stochastic Games and Control

Duality in Robust Dynamic Programming: Pricing Convertibles, Stochastic Games and Control Duality in Robust Dynamic Programming: Pricing Convertibles, Stochastic Games and Control Shyam S Chandramouli Abstract Many decision making problems that arise in inance, Economics, Inventory etc. can

More information

MARKOV DECISION PROCESSES (MDP) AND REINFORCEMENT LEARNING (RL) Versione originale delle slide fornita dal Prof. Francesco Lo Presti

MARKOV DECISION PROCESSES (MDP) AND REINFORCEMENT LEARNING (RL) Versione originale delle slide fornita dal Prof. Francesco Lo Presti 1 MARKOV DECISION PROCESSES (MDP) AND REINFORCEMENT LEARNING (RL) Versione originale delle slide fornita dal Prof. Francesco Lo Presti Historical background 2 Original motivation: animal learning Early

More information

On-line supplement to: SMART: A Stochastic Multiscale Model for the Analysis of Energy Resources, Technology

On-line supplement to: SMART: A Stochastic Multiscale Model for the Analysis of Energy Resources, Technology On-line supplement to: SMART: A Stochastic Multiscale Model for e Analysis of Energy Resources, Technology and Policy This online supplement provides a more detailed version of e model, followed by derivations

More information

Lecture 1. Stochastic Optimization: Introduction. January 8, 2018

Lecture 1. Stochastic Optimization: Introduction. January 8, 2018 Lecture 1 Stochastic Optimization: Introduction January 8, 2018 Optimization Concerned with mininmization/maximization of mathematical functions Often subject to constraints Euler (1707-1783): Nothing

More information

Approximate dynamic programming in transportation and logistics: a unified framework

Approximate dynamic programming in transportation and logistics: a unified framework EURO J Transp Logist (2012) 1:237 284 DOI 10.1007/s13676-012-0015-8 TUTORIAL Approximate dynamic programming in transportation and logistics: a unified framework Warren B. Powell Hugo P. Simao Belgacem

More information

A Benders Algorithm for Two-Stage Stochastic Optimization Problems With Mixed Integer Recourse

A Benders Algorithm for Two-Stage Stochastic Optimization Problems With Mixed Integer Recourse A Benders Algorithm for Two-Stage Stochastic Optimization Problems With Mixed Integer Recourse Ted Ralphs 1 Joint work with Menal Güzelsoy 2 and Anahita Hassanzadeh 1 1 COR@L Lab, Department of Industrial

More information

Logic, Optimization and Data Analytics

Logic, Optimization and Data Analytics Logic, Optimization and Data Analytics John Hooker Carnegie Mellon University United Technologies Research Center, Cork, Ireland August 2015 Thesis Logic and optimization have an underlying unity. Ideas

More information

Clearing the Jungle of Stochastic Optimization

Clearing the Jungle of Stochastic Optimization Clearing the Jungle of Stochastic Optimization Warren B. Powell Department of Operations Research and Financial Engineering Princeton University Prepared for Informs TutORials, 2014. DRAFT April 7, 2014

More information

Optimization Tools in an Uncertain Environment

Optimization Tools in an Uncertain Environment Optimization Tools in an Uncertain Environment Michael C. Ferris University of Wisconsin, Madison Uncertainty Workshop, Chicago: July 21, 2008 Michael Ferris (University of Wisconsin) Stochastic optimization

More information

Bayesian Active Learning With Basis Functions

Bayesian Active Learning With Basis Functions Bayesian Active Learning With Basis Functions Ilya O. Ryzhov Warren B. Powell Operations Research and Financial Engineering Princeton University Princeton, NJ 08544, USA IEEE ADPRL April 13, 2011 1 / 29

More information

Weather routing using dynamic programming to win sailing races

Weather routing using dynamic programming to win sailing races OSE SEMINAR 2013 Weather routing using dynamic programming to win sailing races Mikael Nyberg CENTER OF EXCELLENCE IN OPTIMIZATION AND SYSTEMS ENGINEERING AT ÅBO AKADEMI UNIVERSITY ÅBO NOVEMBER 15 2013

More information

A heuristic algorithm for the Aircraft Landing Problem

A heuristic algorithm for the Aircraft Landing Problem 22nd International Congress on Modelling and Simulation, Hobart, Tasmania, Australia, 3 to 8 December 2017 mssanz.org.au/modsim2017 A heuristic algorithm for the Aircraft Landing Problem Amir Salehipour

More information

A Distributed Decision Making Structure for Dynamic Resource Allocation Using Nonlinear Functional Approximations

A Distributed Decision Making Structure for Dynamic Resource Allocation Using Nonlinear Functional Approximations A Distributed Decision Making Structure for Dynamic Resource Allocation Using Nonlinear Functional Approximations Huseyin Topaloglu School of Operations Research and Industrial Engineering Cornell University,

More information

There has been considerable recent interest in the dynamic vehicle routing problem, but the complexities of

There has been considerable recent interest in the dynamic vehicle routing problem, but the complexities of TRANSPORTATION SCIENCE Vol. 38, No. 4, November 2004, pp. 399 419 issn 0041-1655 eissn 1526-5447 04 3804 0399 informs doi 10.1287/trsc.1030.0073 2004 INFORMS The Dynamic Assignment Problem Michael Z. Spivey,

More information

Stochastic Dual Dynamic Programming with CVaR Risk Constraints Applied to Hydrothermal Scheduling. ICSP 2013 Bergamo, July 8-12, 2012

Stochastic Dual Dynamic Programming with CVaR Risk Constraints Applied to Hydrothermal Scheduling. ICSP 2013 Bergamo, July 8-12, 2012 Stochastic Dual Dynamic Programming with CVaR Risk Constraints Applied to Hydrothermal Scheduling Luiz Carlos da Costa Junior Mario V. F. Pereira Sérgio Granville Nora Campodónico Marcia Helena Costa Fampa

More information

Improvements to Benders' decomposition: systematic classification and performance comparison in a Transmission Expansion Planning problem

Improvements to Benders' decomposition: systematic classification and performance comparison in a Transmission Expansion Planning problem Improvements to Benders' decomposition: systematic classification and performance comparison in a Transmission Expansion Planning problem Sara Lumbreras & Andrés Ramos July 2013 Agenda Motivation improvement

More information

Stochastic Integer Programming An Algorithmic Perspective

Stochastic Integer Programming An Algorithmic Perspective Stochastic Integer Programming An Algorithmic Perspective sahmed@isye.gatech.edu www.isye.gatech.edu/~sahmed School of Industrial & Systems Engineering 2 Outline Two-stage SIP Formulation Challenges Simple

More information

Distributed Optimization. Song Chong EE, KAIST

Distributed Optimization. Song Chong EE, KAIST Distributed Optimization Song Chong EE, KAIST songchong@kaist.edu Dynamic Programming for Path Planning A path-planning problem consists of a weighted directed graph with a set of n nodes N, directed links

More information

Outline. Relaxation. Outline DMP204 SCHEDULING, TIMETABLING AND ROUTING. 1. Lagrangian Relaxation. Lecture 12 Single Machine Models, Column Generation

Outline. Relaxation. Outline DMP204 SCHEDULING, TIMETABLING AND ROUTING. 1. Lagrangian Relaxation. Lecture 12 Single Machine Models, Column Generation Outline DMP204 SCHEDULING, TIMETABLING AND ROUTING 1. Lagrangian Relaxation Lecture 12 Single Machine Models, Column Generation 2. Dantzig-Wolfe Decomposition Dantzig-Wolfe Decomposition Delayed Column

More information

Stochastic Decision Diagrams

Stochastic Decision Diagrams Stochastic Decision Diagrams John Hooker CORS/INFORMS Montréal June 2015 Objective Relaxed decision diagrams provide an generalpurpose method for discrete optimization. When the problem has a dynamic programming

More information

MIDAS: A Mixed Integer Dynamic Approximation Scheme

MIDAS: A Mixed Integer Dynamic Approximation Scheme MIDAS: A Mixed Integer Dynamic Approximation Scheme Andy Philpott, Faisal Wahid, Frédéric Bonnans May 7, 2016 Abstract Mixed Integer Dynamic Approximation Scheme (MIDAS) is a new sampling-based algorithm

More information

Decision making, Markov decision processes

Decision making, Markov decision processes Decision making, Markov decision processes Solved tasks Collected by: Jiří Kléma, klema@fel.cvut.cz Spring 2017 The main goal: The text presents solved tasks to support labs in the A4B33ZUI course. 1 Simple

More information

Generation and Representation of Piecewise Polyhedral Value Functions

Generation and Representation of Piecewise Polyhedral Value Functions Generation and Representation of Piecewise Polyhedral Value Functions Ted Ralphs 1 Joint work with Menal Güzelsoy 2 and Anahita Hassanzadeh 1 1 COR@L Lab, Department of Industrial and Systems Engineering,

More information

Multi-Area Stochastic Unit Commitment for High Wind Penetration in a Transmission Constrained Network

Multi-Area Stochastic Unit Commitment for High Wind Penetration in a Transmission Constrained Network Multi-Area Stochastic Unit Commitment for High Wind Penetration in a Transmission Constrained Network Anthony Papavasiliou Center for Operations Research and Econometrics Université catholique de Louvain,

More information

Chapter 7 Forecasting Demand

Chapter 7 Forecasting Demand Chapter 7 Forecasting Demand Aims of the Chapter After reading this chapter you should be able to do the following: discuss the role of forecasting in inventory management; review different approaches

More information

MODELING DYNAMIC PROGRAMS

MODELING DYNAMIC PROGRAMS CHAPTER 5 MODELING DYNAMIC PROGRAMS Perhaps one of the most important skills to develop in approximate dynamic programming is the ability to write down a model of the problem. Everyone who wants to solve

More information

Agenda today. Introduction to prescriptive modeling. Linear optimization models through three examples: Beyond linear optimization

Agenda today. Introduction to prescriptive modeling. Linear optimization models through three examples: Beyond linear optimization Agenda today Introduction to prescriptive modeling Linear optimization models through three examples: 1 Production and inventory optimization 2 Distribution system design 3 Stochastic optimization Beyond

More information

Section #2: Linear and Integer Programming

Section #2: Linear and Integer Programming Section #2: Linear and Integer Programming Prof. Dr. Sven Seuken 8.3.2012 (with most slides borrowed from David Parkes) Housekeeping Game Theory homework submitted? HW-00 and HW-01 returned Feedback on

More information

Approximate Dynamic Programming for High-Dimensional Resource Allocation Problems

Approximate Dynamic Programming for High-Dimensional Resource Allocation Problems Approximate Dynamic Programming for High-Dimensional Resource Allocation Problems Warren B. Powell Department of Operations Research and Financial Engineering Princeton University Benjamin Van Roy Departments

More information

Weather Forecast Guidance and Impact on NAS Management August 3, 2016

Weather Forecast Guidance and Impact on NAS Management August 3, 2016 Friends and Partners of Aviation Weather Summer 2016 Meeting Weather Forecast Guidance and Impact on NAS Management August 3, 2016 Jim Enders and Kevin Johnston FAA Air Traffic Organization INTRODUCTION

More information

An Integrated Optimizing-Simulator for the Military Airlift Problem

An Integrated Optimizing-Simulator for the Military Airlift Problem An Integrated Optimizing-Simulator for the Military Airlift Problem Tongqiang Tony Wu Department of Operations Research and Financial Engineering Princeton University Princeton, NJ 08544 Warren B. Powell

More information

Mixed Integer Programming Solvers: from Where to Where. Andrea Lodi University of Bologna, Italy

Mixed Integer Programming Solvers: from Where to Where. Andrea Lodi University of Bologna, Italy Mixed Integer Programming Solvers: from Where to Where Andrea Lodi University of Bologna, Italy andrea.lodi@unibo.it November 30, 2011 @ Explanatory Workshop on Locational Analysis, Sevilla A. Lodi, MIP

More information

A Parallelizable and Approximate Dynamic Programming-Based Dynamic Fleet Management Model with Random Travel Times and Multiple Vehicle Types

A Parallelizable and Approximate Dynamic Programming-Based Dynamic Fleet Management Model with Random Travel Times and Multiple Vehicle Types A Parallelizable and Approximate Dynamic Programming-Based Dynamic Fleet Management Model with Random Travel Times and Multiple Vehicle Types Huseyin Topaloglu School of Operations Research and Industrial

More information

Section Notes 9. Midterm 2 Review. Applied Math / Engineering Sciences 121. Week of December 3, 2018

Section Notes 9. Midterm 2 Review. Applied Math / Engineering Sciences 121. Week of December 3, 2018 Section Notes 9 Midterm 2 Review Applied Math / Engineering Sciences 121 Week of December 3, 2018 The following list of topics is an overview of the material that was covered in the lectures and sections

More information

Anticipatory Freight Selection in Intermodal Long-haul Round-trips

Anticipatory Freight Selection in Intermodal Long-haul Round-trips Anticipatory Freight Selection in Intermodal Long-haul Round-trips A.E. Pérez Rivera and M.R.K. Mes Department of Industrial Engineering and Business Information Systems, University of Twente, P.O. Box

More information

Applying High Performance Computing to Multi-Area Stochastic Unit Commitment

Applying High Performance Computing to Multi-Area Stochastic Unit Commitment Applying High Performance Computing to Multi-Area Stochastic Unit Commitment IBM Research Anthony Papavasiliou, Department of Mathematical Engineering, CORE, UCL Shmuel Oren, IEOR Department, UC Berkeley

More information

Application of Monte Carlo Simulation to Multi-Area Reliability Calculations. The NARP Model

Application of Monte Carlo Simulation to Multi-Area Reliability Calculations. The NARP Model Application of Monte Carlo Simulation to Multi-Area Reliability Calculations The NARP Model Any power system reliability model using Monte Carlo simulation consists of at least the following steps: 1.

More information

is called an integer programming (IP) problem. model is called a mixed integer programming (MIP)

is called an integer programming (IP) problem. model is called a mixed integer programming (MIP) INTEGER PROGRAMMING Integer Programming g In many problems the decision variables must have integer values. Example: assign people, machines, and vehicles to activities in integer quantities. If this is

More information

Integrated Electricity Demand and Price Forecasting

Integrated Electricity Demand and Price Forecasting Integrated Electricity Demand and Price Forecasting Create and Evaluate Forecasting Models The many interrelated factors which influence demand for electricity cannot be directly modeled by closed-form

More information

How useful is the deterministic solution? Stein W. Wallace

How useful is the deterministic solution? Stein W. Wallace Joint work with Francesca Maggioni, Biju K. Thapalia, Michal Kaut, Teodor G. Crainic Outline The overall goal 1 The overall goal 2 3 4 Conclusions on VSS-related measures 5 The tradition In parts of OR

More information

The Optimizing-Simulator: An Illustration using the Military Airlift Problem

The Optimizing-Simulator: An Illustration using the Military Airlift Problem The Optimizing-Simulator: An Illustration using the Military Airlift Problem Tongqiang Tony Wu Warren B. Powell Princeton University and Alan Whisman Air Mobility Command There have been two primary modeling

More information

Interior-Point versus Simplex methods for Integer Programming Branch-and-Bound

Interior-Point versus Simplex methods for Integer Programming Branch-and-Bound Interior-Point versus Simplex methods for Integer Programming Branch-and-Bound Samir Elhedhli elhedhli@uwaterloo.ca Department of Management Sciences, University of Waterloo, Canada Page of 4 McMaster

More information

Network Flows. 6. Lagrangian Relaxation. Programming. Fall 2010 Instructor: Dr. Masoud Yaghini

Network Flows. 6. Lagrangian Relaxation. Programming. Fall 2010 Instructor: Dr. Masoud Yaghini In the name of God Network Flows 6. Lagrangian Relaxation 6.3 Lagrangian Relaxation and Integer Programming Fall 2010 Instructor: Dr. Masoud Yaghini Integer Programming Outline Branch-and-Bound Technique

More information

Lecture 1: March 7, 2018

Lecture 1: March 7, 2018 Reinforcement Learning Spring Semester, 2017/8 Lecture 1: March 7, 2018 Lecturer: Yishay Mansour Scribe: ym DISCLAIMER: Based on Learning and Planning in Dynamical Systems by Shie Mannor c, all rights

More information

Forecast Confidence. Haig Iskenderian. 18 November Sponsor: Randy Bass, FAA Aviation Weather Research Program, ANG-C6

Forecast Confidence. Haig Iskenderian. 18 November Sponsor: Randy Bass, FAA Aviation Weather Research Program, ANG-C6 Forecast Confidence Haig Iskenderian 18 November 2014 Sponsor: Randy Bass, FAA Aviation Weather Research Program, ANG-C6 Distribution Statement A. Approved for public release; distribution is unlimited.

More information

Traffic Flow Impact (TFI)

Traffic Flow Impact (TFI) Traffic Flow Impact (TFI) Michael P. Matthews 27 October 2015 Sponsor: Yong Li, FAA ATO AJV-73 Technical Analysis & Operational Requirements Distribution Statement A. Approved for public release; distribution

More information

Perhaps one of the most widely used and poorly understood terms in dynamic programming is policy. A simple definition of a policy is:

Perhaps one of the most widely used and poorly understood terms in dynamic programming is policy. A simple definition of a policy is: CHAPTER 6 POLICIES Perhaps one of the most widely used and poorly understood terms in dynamic programming is policy. A simple definition of a policy is: Definition 6.0.1 A policy is a rule (or function)

More information

IOE 202: lecture 14 outline

IOE 202: lecture 14 outline IOE 202: lecture 14 outline Announcements Last time... Value of information (perfect and imperfect) in decision analysis Course wrap-up: topics covered and where to go from here IOE 202: Operations Modeling,

More information

Perfect and Imperfect Competition in Electricity Markets

Perfect and Imperfect Competition in Electricity Markets Perfect and Imperfect Competition in Electricity Marets DTU CEE Summer School 2018 June 25-29, 2018 Contact: Vladimir Dvorin (vladvo@eletro.dtu.d) Jalal Kazempour (seyaz@eletro.dtu.d) Deadline: August

More information

Regularized optimization techniques for multistage stochastic programming

Regularized optimization techniques for multistage stochastic programming Regularized optimization techniques for multistage stochastic programming Felipe Beltrán 1, Welington de Oliveira 2, Guilherme Fredo 1, Erlon Finardi 1 1 UFSC/LabPlan Universidade Federal de Santa Catarina

More information

JOINT PRICING AND NETWORK CAPACITY SETTING PROBLEM

JOINT PRICING AND NETWORK CAPACITY SETTING PROBLEM Advanced OR and AI Methods in Transportation JOINT PRICING AND NETWORK CAPACITY SETTING PROBLEM Luce BROTCORNE, Patrice MARCOTTE, Gilles SAVARD, Mickael WIART Abstract. We consider the problem of jointly

More information

Probabilistic Planning. George Konidaris

Probabilistic Planning. George Konidaris Probabilistic Planning George Konidaris gdk@cs.brown.edu Fall 2017 The Planning Problem Finding a sequence of actions to achieve some goal. Plans It s great when a plan just works but the world doesn t

More information

High-dimensional Problems in Finance and Economics. Thomas M. Mertens

High-dimensional Problems in Finance and Economics. Thomas M. Mertens High-dimensional Problems in Finance and Economics Thomas M. Mertens NYU Stern Risk Economics Lab April 17, 2012 1 / 78 Motivation Many problems in finance and economics are high dimensional. Dynamic Optimization:

More information

Predicting MTA Bus Arrival Times in New York City

Predicting MTA Bus Arrival Times in New York City Predicting MTA Bus Arrival Times in New York City Man Geen Harold Li, CS 229 Final Project ABSTRACT This final project sought to outperform the Metropolitan Transportation Authority s estimated time of

More information

The Knowledge Gradient for Sequential Decision Making with Stochastic Binary Feedbacks

The Knowledge Gradient for Sequential Decision Making with Stochastic Binary Feedbacks The Knowledge Gradient for Sequential Decision Making with Stochastic Binary Feedbacks Yingfei Wang, Chu Wang and Warren B. Powell Princeton University Yingfei Wang Optimal Learning Methods June 22, 2016

More information

The fire and rescue vehicle location problem

The fire and rescue vehicle location problem The fire and rescue vehicle location problem Henrik Andersson Norwegian University of Science and Technology Tobias Andersson Granberg Linköping University Introduction The Swedish Civil Contingencies

More information

Asset Management Planning. GIS and Asset Management Integration Readiness Assessment

Asset Management Planning. GIS and Asset Management Integration Readiness Assessment Asset Management Planning GIS and Asset Management Integration Readiness Assessment With too little data, you won t be able to make any conclusions that you trust. With loads of data you will find relationships

More information

Stochastic Sequencing and Scheduling of an Operating Room

Stochastic Sequencing and Scheduling of an Operating Room Stochastic Sequencing and Scheduling of an Operating Room Camilo Mancilla and Robert H. Storer Lehigh University, Department of Industrial and Systems Engineering, November 14, 2009 1 abstract We develop

More information

Incorporating Demand Response with Load Shifting into Stochastic Unit Commitment

Incorporating Demand Response with Load Shifting into Stochastic Unit Commitment Incorporating Demand Response with Load Shifting into Stochastic Unit Commitment Frank Schneider Supply Chain Management & Management Science, University of Cologne, Cologne, Germany, frank.schneider@uni-koeln.de

More information

Introduction to optimization and operations research

Introduction to optimization and operations research Introduction to optimization and operations research David Pisinger, Fall 2002 1 Smoked ham (Chvatal 1.6, adapted from Greene et al. (1957)) A meat packing plant produces 480 hams, 400 pork bellies, and

More information

An Operational Evaluation of the Ground Delay Program (GDP) Parameters Selection Model (GPSM)

An Operational Evaluation of the Ground Delay Program (GDP) Parameters Selection Model (GPSM) An Operational Evaluation of the Ground Delay Program (GDP) Parameters Selection Model (GPSM) Lara Shisler, Christopher Provan Mosaic ATM Dave Clark MIT Lincoln Lab William Chan, Shon Grabbe NASA Ames

More information

Scheduling an Aircraft Repair Shop

Scheduling an Aircraft Repair Shop Proceedings of the Twenty-First International Conference on Automated Planning and Scheduling Scheduling an Aircraft Repair Shop Maliheh Aramon Bajestani and J. Christopher Bec Department of Mechanical

More information

CS6375: Machine Learning Gautam Kunapuli. Decision Trees

CS6375: Machine Learning Gautam Kunapuli. Decision Trees Gautam Kunapuli Example: Restaurant Recommendation Example: Develop a model to recommend restaurants to users depending on their past dining experiences. Here, the features are cost (x ) and the user s

More information

Capacity Planning with uncertainty in Industrial Gas Markets

Capacity Planning with uncertainty in Industrial Gas Markets Capacity Planning with uncertainty in Industrial Gas Markets A. Kandiraju, P. Garcia Herreros, E. Arslan, P. Misra, S. Mehta & I.E. Grossmann EWO meeting September, 2015 1 Motivation Industrial gas markets

More information

21. Set cover and TSP

21. Set cover and TSP CS/ECE/ISyE 524 Introduction to Optimization Spring 2017 18 21. Set cover and TSP ˆ Set covering ˆ Cutting problems and column generation ˆ Traveling salesman problem Laurent Lessard (www.laurentlessard.com)

More information

Decomposition methods in optimization

Decomposition methods in optimization Decomposition methods in optimization I Approach I: I Partition problem constraints into two groups: explicit and implicit I Approach II: I Partition decision variables into two groups: primary and secondary

More information

IBM Research Report. Stochasic Unit Committment Problem. Julio Goez Lehigh University. James Luedtke University of Wisconsin

IBM Research Report. Stochasic Unit Committment Problem. Julio Goez Lehigh University. James Luedtke University of Wisconsin RC24713 (W0812-119) December 23, 2008 Mathematics IBM Research Report Stochasic Unit Committment Problem Julio Goez Lehigh University James Luedtke University of Wisconsin Deepak Rajan IBM Research Division

More information

Optimal Large-Scale Air Traffic Flow Management

Optimal Large-Scale Air Traffic Flow Management Optimal Large-Scale Air Traffic Flow Management Hamsa Balakrishnan Massachusetts Institute of Technology Cambridge, MA 02139 hamsa@mit.edu Bala G. Chandran Resilient Ops LLC Winchester, MA 01890 bala.chandran@resilientops.com

More information

Planning a 100 percent renewable electricity system

Planning a 100 percent renewable electricity system Planning a 100 percent renewable electricity system Andy Philpott Electric Power Optimization Centre University of Auckland www.epoc.org.nz (Joint work with Michael Ferris) INI Open for Business Meeting,

More information

WeatherCloud Hyper-Local Global Forecasting All rights reserved. Fathym, Inc.

WeatherCloud Hyper-Local Global Forecasting All rights reserved. Fathym, Inc. WeatherCloud Hyper-Local Global Forecasting based on current forecast techniques EVOLVING FORECASTING TECHNOLOGY 1) The WeatherCloud backend forecast system allows for routing around hazardous weather

More information

CSE 417T: Introduction to Machine Learning. Final Review. Henry Chai 12/4/18

CSE 417T: Introduction to Machine Learning. Final Review. Henry Chai 12/4/18 CSE 417T: Introduction to Machine Learning Final Review Henry Chai 12/4/18 Overfitting Overfitting is fitting the training data more than is warranted Fitting noise rather than signal 2 Estimating! "#$

More information

INTEGER PROGRAMMING. In many problems the decision variables must have integer values.

INTEGER PROGRAMMING. In many problems the decision variables must have integer values. INTEGER PROGRAMMING Integer Programming In many problems the decision variables must have integer values. Example:assign people, machines, and vehicles to activities in integer quantities. If this is the

More information

What you should know about approximate dynamic programming

What you should know about approximate dynamic programming What you should know about approximate dynamic programming Warren B. Powell Department of Operations Research and Financial Engineering Princeton University, Princeton, NJ 08544 December 16, 2008 Abstract

More information

Applications. Stephen J. Stoyan, Maged M. Dessouky*, and Xiaoqing Wang

Applications. Stephen J. Stoyan, Maged M. Dessouky*, and Xiaoqing Wang Introduction to Large-Scale Linear Programming and Applications Stephen J. Stoyan, Maged M. Dessouky*, and Xiaoqing Wang Daniel J. Epstein Department of Industrial and Systems Engineering, University of

More information

Sample questions for COMP-424 final exam

Sample questions for COMP-424 final exam Sample questions for COMP-44 final exam Doina Precup These are examples of questions from past exams. They are provided without solutions. However, Doina and the TAs would be happy to answer questions

More information

Section Notes 8. Integer Programming II. Applied Math 121. Week of April 5, expand your knowledge of big M s and logical constraints.

Section Notes 8. Integer Programming II. Applied Math 121. Week of April 5, expand your knowledge of big M s and logical constraints. Section Notes 8 Integer Programming II Applied Math 121 Week of April 5, 2010 Goals for the week understand IP relaxations be able to determine the relative strength of formulations understand the branch

More information

Cover Page. The handle holds various files of this Leiden University dissertation

Cover Page. The handle  holds various files of this Leiden University dissertation Cover Page The handle http://hdl.handle.net/1887/39637 holds various files of this Leiden University dissertation Author: Smit, Laurens Title: Steady-state analysis of large scale systems : the successive

More information

Payments System Design Using Reinforcement Learning: A Progress Report

Payments System Design Using Reinforcement Learning: A Progress Report Payments System Design Using Reinforcement Learning: A Progress Report A. Desai 1 H. Du 1 R. Garratt 2 F. Rivadeneyra 1 1 Bank of Canada 2 University of California Santa Barbara 16th Payment and Settlement

More information

Generalization and Function Approximation

Generalization and Function Approximation Generalization and Function Approximation 0 Generalization and Function Approximation Suggested reading: Chapter 8 in R. S. Sutton, A. G. Barto: Reinforcement Learning: An Introduction MIT Press, 1998.

More information

Column Generation for Extended Formulations

Column Generation for Extended Formulations 1 / 28 Column Generation for Extended Formulations Ruslan Sadykov 1 François Vanderbeck 2,1 1 INRIA Bordeaux Sud-Ouest, France 2 University Bordeaux I, France ISMP 2012 Berlin, August 23 2 / 28 Contents

More information

Stochastic Unit Commitment with Topology Control Recourse for Renewables Integration

Stochastic Unit Commitment with Topology Control Recourse for Renewables Integration 1 Stochastic Unit Commitment with Topology Control Recourse for Renewables Integration Jiaying Shi and Shmuel Oren University of California, Berkeley IPAM, January 2016 33% RPS - Cumulative expected VERs

More information

An Approximate Dynamic Programming Algorithm for the Allocation of High-Voltage Transformer Spares in the Electric Grid

An Approximate Dynamic Programming Algorithm for the Allocation of High-Voltage Transformer Spares in the Electric Grid An Approximate Dynamic Programming Algorithm for the Allocation of High-Voltage Transformer Spares in the Electric Grid Johannes Enders Department of Operations Research and Financial Engineering Princeton

More information

MEDIUM-TERM HYDROPOWER SCHEDULING BY STOCHASTIC DUAL DYNAMIC INTEGER PROGRAMMING: A CASE STUDY

MEDIUM-TERM HYDROPOWER SCHEDULING BY STOCHASTIC DUAL DYNAMIC INTEGER PROGRAMMING: A CASE STUDY MEDIUM-TERM HYDROPOWER SCHEDULING BY STOCHASTIC DUAL DYNAMIC INTEGER PROGRAMMING: A CASE STUDY Martin Hjelmeland NTNU Norwegian University of Science and Technology, Trondheim, Norway. martin.hjelmeland@ntnu.no

More information