Next Generation Decision Support System s for Railroad Planning Ravindra K. Ahuja Professor, University of Florida & President, Innovative Scheduling 1
Presentation Outline Overview Railroad Blocking Optimizer Train Scheduling Optimizer Locomotive Planning Optimizer Overview of Other Systems Lessons Learnt 2 2
Origin of Innovative Scheduling Started research on railroad planning and scheduling problems in 2000. The company started its formal operations in 2004 with a single employee and a Phase I grant from NSF’s Small Business Innovations Research (SBIR) Program. Received second SBIR Grant in 2005. Started commercialization of software immediately. Started forming development partnerships with companies to build products. The company now has about 20 full-time employees and 8-10 part-time employees. 3 3
Our Core Strength Ability to solve very complex decision problems efficiently: � Blocking problem � Train scheduling problem � Locomotive planning, simulation problems � Crew planning and scheduling problems Expertise in a variety of Operations Research techniques: � Linear programming � Integer programming � Network flows and discrete optimization � Several heuristic techniques � Simulation techniques Combine a variety of OR techniques to solve large-scale decision problems very efficiently. 4 4
Programming and IT Skills Programming Skills: � C+ + � Concert Technology, CPLEX � VB .NET and ASP .NET � Java � ESRI GIS programming for maps Decision Support Systems Building Skills � Excel-based applications � Desktop-based applications � Web-based applications Most of our solution engines are developed in C+ + / Java and packaged within web-enabled applications. 5 5
Our Railroad Decision Support Systems Innovative Railroad Blocking Optimizer (IRBO) Innovative Train Scheduling Optimizer (ITSO) Innovative Locomotive Planning Optimizer (ILPO) Innovative Locomotive Simulation Optimizer (ILSO) Innovative Crew Scheduling Optimizer (ICSO) Innovative Hump Yard Manager (IHYM) Innovative Network Flow Analyzer (INFA) Innovative Locomotive Shop Router (ILSR) Innovative Yard Simulation Optimizer (IYSO) 6 6
Presentation Outline Overview Railroad Blocking Optimizer Train Scheduling Optimizer Locomotive Planning Optimizer Overview of Other Systems Lessons Learnt 7 7
Railroad Planning and Scheduling Blocking Problem Service Design Plan Train Scheduling Locom otive Crew Yard Scheduling Scheduling Operations 8 8
Consolidation Problem Railroad blocking problem is essentially a consolidation problem, which is similar to that encountered in postal service design. Gainesville Gainesville A railroad block is like a mailbag in the postal service context. 9 9
Railroad Blocking Problem Destinations Destinations Yards Yards Origins Origins 10 10
The Railroad Blocking Model Network Blocks Railroad Railroad Blocking Blocking Shipment Block Model Model Shipments Assignments Constraints: Constraints: Objective Function: Objective Function: � Maximum number of � Distances traveled by blocks that can be build at shipments a node is limited. � Intermediate handlings of � Maximum volume of shipments shipments passing through a node is limited. 11 11
Literature Survey ABM (Algorithmic Blocking Model) by Carl Van Dyke [ 1986, 1988] Keaton [ 1989, 1992] Newton, Barnhart and Vance [ 1998] Barnhart and Vance [ 2000] The railroad blocking problem remained an unsolved problem until recently. 12 12
Our Contributions Multi-commodity flow network design and routing problem: � 3,000 nodes � 50,000 commodities � Over a million 0-1 network design variables � Hundreds of billions of integer flow variables We developed a very large-scale neighborhood (VLSN) search algorithm to solve this problem to near-optimality within one-two hours. Can also do incremental blocking and handle a variety of practical constraints. 13 13
Overview of the VLSN Search Algorithm 8 1 5 9 2 7 10 3 6 11 4 We reoptimize blocks at one node at a time assuming that blocks do not change at other nodes. We reoptimize all nodes one-by-one and keep performing passes over the nodes until the solution terminates to a local optimal solution. 14 14
Computational Results: Incremental Blocking % Savings in % New Blocks % Savings in I nterm ediate Car Miles Handlings 0 .9 % 0 .5 % 7 .9 % 1 .9 % 0 .5 % 1 0 .5 % 3 .8 % 0 .5 % 1 4 .1 % 9 .5 % 0 .6 % 1 9 .1 % Conclusion: Even small changes in the blocking plan can have significant impact on intermediate handlings. 15 15
Railroad Users Consulting Activities: � CSX Transportation in One Plan � Norfolk Southern in TOP II Plan � BNSF Railway in its current operating plan � Union Pacific in its Unified Plan Licensing: � Norfolk Southern � BNSF Railway Potential Future Clients: � Union Pacific � Canadian National � SNCF (France) � Deutsche Bahn (Germany) 16 16
Presentation Outline Overview Railroad Blocking Optimizer Train Scheduling Optimizer Locomotive Planning Optimizer Overview of Other Systems Lessons Learnt 17 17
18 7 6 0 1 6 Flow of Blocks on Trains 5 9 5 4 18 4 8 3 3 7 2 2 1 1
Train Schedule Design Problem Blocks Trains Block-to-Train Shipments Assignments Train Train Trip Plan Scheduling Scheduling Shipment-Block Optim izer Optim izer Assignments Balanced Crew Crew Assignment Balanced Locomotive Locomotive Assignment 19 19
Decision Variables Decision: � Train origins, destinations, and routes � Train days of operation and train times � Train block-to-train assignment by day of the week � Trip plans for all cars � Locomotive assignment � Crew assignment Constraints � Yard capacity constraints � Line capacity constraints � Train capacity constraints � Business rules 20 20
Contribution: Integration of Railroad Resources Constrained by Netw ork Capacity Constrained by Operating Rules Railcar Railcar I TSO I TSO I TSO Locom otive Crew Locom otive Crew We consider these three resources by maintaining three time-space networks. 21 21
Railcar Flow Network Ground Nodes Train 1 car car car car car Train 1 car car car car Train 2 Time car car car car car Train 2 car car car car Train 3 car car car car car Train 5 car car car car We construct the weekly time-space train network and flow railcars through this network. 22 22
Locomotive Flow Network Train 1 Train 4 Train 2 Train 5 Train 3 Train 6 We construct the weekly space-time train network and locomotives cycle through this network. 23 23
Crew Scheduling at US Railroads Each train requires a crew and changes crew at several locations as it travels from its origin to its destination. 1 2 3 4 5 6 7 1 2 3 4 5 6 7 24 24
Crew Flow Network Home Terminal Away Terminal We construct the weekly space-time Time 1 crew network and 3 crews cycle through this network. 4 5 6 We create a separate 7 network for each crew 8 district. 10 11 12 13 Train Arcs 14 Deadhead Arcs Rest Arcs 25 25
Constraints Yard Constraints � Number of trains originating at any node in each given time window is limited. � Number of trains terminating at any node in each given time window is limited. � Number of trains passing through each node in each given time window is limited. Track Constraints � Speed of a train on a track depends upon the type of train. � Number of trains passing through any corridor in any given time window is limited. � Satisfy headway constraints 26 26
Constraints (contd.) Train Capacity Constraints � The number of cars on any train is limited � The length of any train is limited � The weight-carrying capacity of any train is limited � No more than specified number of blocks per train � Number of stops of a train is limited Locomotive Constraints � Honor locomotive minimum connection times between trains � Provide number of locomotive based on train tonnages Crew Constraints � Honor crew minimum connection times between trains � Honor crew union rules related to work and rest 27 27
Objective Function Terms Car days Car Block miles swaps Train Loco miles cost Train Crew starts cost 28 28
Our Contribution Problem size: � Number of railcars: 125,000 � Number of locomotives: 2,000 – 4,000 � Number of crew districts: 300-400 � Number of crews: 4,000-6,000 We have developed a computer program to solve this problem within 1-2 hours on a laptop. Uses a variety of operations research techniques: � Construction heuristics � Network flows & Linear programming � Neighborhood search � Very large-scale neighborhood (VLSN) search 29 29
A Two-Stage Decomposition Process Train Details Train Route Optim ization Optim ization Train schedule with time Train schedule without time � Train routes � Train routes � Block-train assignment � Block-train assignment � Locomotive assignment � Locomotive assignment � Crew assignment � Crew assignment 30 30
Recommend
More recommend
