A Traffic Management System for RealTime Traffic Optimisation in Railways

1
A Traffic Management System for Real-Time
Traffic Optimisation in Railways

• Research Framework
• COMBINE and COMBINE 2 Projects
• TMS Architecture
• Conflict Detection Resolution (CDR)
• Speed Profile Generator (SPG)
• TMS in Action
• Model for Conflict Resolution
• Model for Speed Regulation

1st International Seminar on Railway Operations
Modelling and Analysis - Delft, The Netherlands
June 8-10, 2005
Maura Mazzarello
2
Research Framework
3
(January 1999 - March 2001)

• (enhanced COntrol center for a Moving Block
  sIgnalling systEm) - TR 4004

• Architectures and algorithms for traffic
  management and control in railways equipped with
  moving block signalling
• Advanced Traffic Management System
• Conflict detection and resolution
• Real-time traffic optimisation
• In order to
• Improve throughput
• Improve performance (punctuality energy saving)
• Improve flexibility

4
(March 2002 - December 2003)

• (enhanced COntrol centres for fixed and Moving
  Block sIgNalling systEms 2) IST-2001-34705

Extends the research developed within COMBINE
project to large networks equipped with fixed and
moving block signalling systems. Real-time
traffic optimisation to provide improved service
and better exploitation of existing
infrastructures.
5
TMS Architecture

• Local level
• Automatic Conflict Detection Resolution Speed
  Profile Generation
• CDR optimal schedule
• punctuality
• SPG schedule execution
• speed control
• punctuality
• energy saving
• Network level
• Automatic Co-ordination (CM)
• global feasibility of local schedules

6
Co-ordination Manager

• Automatically computes whether local TMS plans
  are globally feasible or not
• Triggered by the generation of a local TMS plan
• Aggregate representation of the network
• Border graph
• Check for cycles
• Constraints to local areas

7
Conflict Detection Resolution

• Optimal train dispatching system, responsible
  for
• train scheduling
• route planning
• local re-routing
• Objective to set precedence rules among trains
  in order to fulfill targets and maximize
  punctuality

8
Speed Profile Generation

• Computes the optimal speed profile minimizing a
  cost function, which involves
• punctuality
• acceleration / deceleration
• maximum allowed speed
• Saves energy by limiting the instability of train
  speed profiles

9
TMS in Action
10
TMS in Action issues from real-time operation

• Process plan changes during a session
• Sudden infra or train degradations
• Equipped and not equipped trains
• Start-up in an active environment
• Advisory speeds not obeyed by train drivers
• Route bookings not properly implemented by
  dispatchers
• Communication delays and failures
• Errors in position and speed information from the
  trains
• Actual train characteristics differ from the
  expected ones

11
TMS in Action improvements for real-time issues

• Schedule windows
• Updates
• Route changes
• Co-operation between CDR and SPG
• Prediction
• Uncontrolled train model
• Driving tables

12
TMS in Action de Groene Golf pilot

• ProRail DTM in practice
• Pilot area Roosendaal-Dordrecht and
  BredaDordrecht, including the junction at Lage
  Zwaluwe
• June 2004
• TMS VPT
• TMS TT trains

13
Conflict Detection Resolution Model

• Alternative Graph Mascis Pacciarelli 2000

Variables starting time ti of operation oi ,
i1, , n
Operation
Node
N nodes F FIXED arcs A pairs of ALTERNATIVE
arcs
G(N,F,A)
A solution is feasible if and only if there are
no positive length cycles in the resulting
directed graphs
14
Train Scheduling Model
ti the time at which a given train enters a
given track element or block section B(i) node
i.

• Nodes
• Entry (exit) of each block section
• Switches
• Speed limitation borders
• Entry nodes Init node
• Exit node End node
• Position node

15
Train Scheduling Model
Blocking Constraint alternative arcs
Simple graph for two trains
16
Constraints in Train Scheduling
Minimum Speed Constraint
Passing Constraint
Stop and Departure Constraint
Connection Constraint
Out-of-Order Constraint
17
Train Scheduling Algorithm

• Create Plan
• Minimal Schedule Generation
• Graph Pre-Processing
• Graph Check
• Conflict Detection
• Conflict Resolution
• Re-routing
• Post-processing

18
Speed Regulation Model
The energy consumption is minimized when all
trains are operated at the lowest speeds
consistent with their required performance levels
The optimal speed profile must ensure the CDR
plan being executed in a safe and energy saving
manner
19
The Control Loop
The total delay
is the most important parameter to be taken into
account
20
Speed Regulation Cycle
Traffic perturbations

• Position/speed forecast
• Safety check
• Speed evaluation
• Route booking

Goals
Schedule update
Current train state (10 sec)
Advisory speed Route Booking
21
Conclusions and Future Works

• TMS core functions
• CDR stand-alone generates optimal schedules,
  given different traffic scenarios (capacity
  estimation and operational control strategies at
  bottlenecks)
• CDR SPG ( external modules) real-time
  flexible traffic handling (real-world pilot)
• TMS improvements
• TMS core functions
• TMS interfaces (to make the TMS much more
  independent of the chosen implementation of the
  surrounding systems and more focused on the core
  functions)