An Actor-based Approach to Commodity Transport Modelling 5. - PowerPoint PPT Presentation

******* Titel An Actor-based Approach to Commodity Transport Modelling 5. Conference on Applied Infrastructure Research Gernot Liedtke 6.-7.10.2006 1

Overview Motivation and goals Motivation and goals • • Principles of the actor-based approach • InterLOG simulation model • Selected scenarios • Conclusion • 2

Macroscopic approach Scalability Categorisations Production functions 1.) Generation 1.) Generation 2.) Distribution 2.) Distribution 3.) Mode choice 3.) Mode choice 4.) Vehicle and 4.) Vehicle and network models network models 3

Macroscopic approach Scalability Categorisations Production functions 1.) Generation 2.) Distribution Gravity model Distribution systems 3.) Mode choice 4.) Vehicle and network models 4

Macroscopic approach Scalability Categorisations 1.) Generation Production functions 1.) Generation 2.) Distribution 2.) Distribution Gravity model Distribution systems 3.) Mode choice 3.) Mode choice Disaggregate models Heterogeneity 4.) Vehicle and network models 5

Macroscopic approach Scalability Categorisations Production functions 1.) Generation 2.) Distribution Gravity model Distribution systems 3.) Mode choice Disaggregate models Heterogeneity 4.) Vehicle and Proportion of Spatial patterns network models empty running Consolidation Freight flows cannot be mapped in form of a matrix 6

Actor-based approach • Mapping of individual firms • Large number • Control parameters 7

Challenges for realising the actor-based approach Inter-sectoral flows Production rates Distance sensibility Shipper-carrier problem Vehicle dispatching Number of actors 8

**** Principles Motivation and goals • Principles of the actor-based approach • InterLOG simulation model • Selected scenarios • Analysis and outlook • 9

Principles of actor-based freight transport modelling Interaction structures: transport markets • Feed-back mechanism for relationship-stabilisation: • operational meso-structures Actors’ individual decision making : CLP-based • “decision support engines” 10

Interaction fields: transport markets time space Central warehouse Food factory Carrier’s depot 11

Relationship stabilisation: meso-structures How would these transports be 12 reflected in aggregate statistics?

Individual decision making: Constraint Logic Dispatcher’s task Depot • To find feasible solutions Vehicle • To find robust solutions • In an uncertain environment V 1 V 2 V 3 Perspective constraints S 1 S 3 Capacity constraints Configuration constraints Sequence constraints Properties of CLP Time-windows • Domain reduction/goals • Object orientation Legislation • Constraint propagation Service activities 13

**** InterLOG Motivation and goals • Principles of the actor-based approach • InterLOG simulation model • Selected scenarios • Conclusion • 14

Class structure “World” Companies Hauliers Network Regions Definition Relationships Lorries Tour-planning Contracts Tour plans problems Optimisation environment Solution 15

InterLOG simulation model: overview Company generation Generation modules Generation Generation Forwarder generation Set of attributed companies in space Distribution Distribution Sourcing module: choice of suppliers Commodity flows between companies Lot-size choice Lot-size choice Market Shippers models Market choice Market choice interaction module Forwarders models Lorry model Lorry model Network model Network model Attributed lorry trips on the road network Aggregation, verification 16

Actor-generation module Company-size distribution by 3-digit cpa-sector Employment data by region and sector 17

Sourcing module Output rates by 3-digital cpa-sector Inter-sectoral flows ([tonnes]) by transport sub-markets Sensibility to distance by transport sub-markets 18

Market interaction module Shipper • Cheapest forwarder States or • Minimisation of the retrieves total logistics cost Contracts/ Constraints Market exploration Calls for offers • Weight • Time • Distance Forwarder • Does the contract fit? • Daily and weekly planning Tariff determination/rejection Tour planning, 19

**** Scenarios Motivation and goals • Principles of the actor-based approach • InterLOG simulation model • Selected scenarios • Conclusion • 20

Scenario “with many actors” Characteristics � 1000 Companies � 200 hauliers � ~ 2000 shipment cases/week � ~ 1 •10 6 lorry km / week � ∅ 60 lorries / week on motorways � Prop. of empty running: ~ 30% Network nodes - Link without lorry-traffic load - Lorry traffic load 21

Policy measure sensitivity: Impacts of a toll Total mileage ↓ Shift to External effects ↑ federal roads Meso-structures ↑ Mutual adaptation ∑= ? Motorway toll Punishment of Efficiency ↑ inefficient hauliers Load factors ↑ Punishment of partial loads Stocks ↑ 22

Policy measure sensitivity: Impacts of a toll Total mileage 225000 220000 vehicle km 215000 210000 205000 No toll 200000 Toll 195000 14 16 18 20 22 24 26 28 30 Simulation period (weeks) 23

Policy measure sensitivity: Impacts of a toll Mileage empty runs 88000 No toll 86000 toll 84000 vehicle km 82000 80000 78000 76000 74000 72000 70000 14 16 18 20 22 24 26 28 30 32 Simulation period (weeks) 24

Policy measure sensitivity: Impacts of a toll Number of shipments per week 284 282 280 278 276 274 272 270 No toll 268 Toll 266 264 14 16 18 20 22 24 26 28 30 32 Simulation period (weeks) 25

**** Analysis and Outlook Motivation and goals • Principles of the actor-based approach • InterLOG simulation model • Selected scenarios • Conclusion • 26

Conclusion Microscopic approach to freight transport modelling • Data availability • Limitations and possibilities for further development • Interaction structures • 27

******* ENDE An Actor-based Approach to Commodity Transport Modelling 5. Conference on Applied Infrastructure Research Gernot Liedtke 28