COP97: Traffic Signal Control Optimization

An implementation of the Controlled Optimization of Phases (COP) algorithm based on Sen & Head's work on traffic signal optimization. This implementation uses Dynamic Programming (DP) to optimize traffic signal timing plans.

Overview

This project implements a traffic signal control algorithm that optimizes signal timing based on various performance criteria:

• Queue length minimization
• Stop minimization
• Delay minimization

The algorithm uses dynamic programming to find optimal green times for each phase while considering:

• Vehicle arrivals
• Queue formation and discharge
• Phase sequencing
• Red clearance intervals
• Minimum and maximum green times

Project Structure

cop97/
├── src/          # Source code
├── data/         # Input data files
├── scripts/      # Shell scripts
├── docs/         # Documentation
└── tests/        # Test files

Implementation Details

Main Components

• src/main.cpp: The primary implementation file
• Multiple version iterations (v0 through v5) in src/
• Shell scripts for execution and testing in scripts/
• Test data files in data/

Performance Indices

The implementation can optimize for different objectives:

1. QUEUES: Minimizes maximum queue length across all approaches
2. STOPS: Minimizes total number of vehicle stops
3. DELAY: Minimizes total vehicle delay

Key Features

• Dynamic Programming approach for phase optimization
• Flexible performance criteria selection
• Consideration of queue discharge rates
• Integration of vehicle arrival data
• Support for multiple phases
• Red clearance time handling

Building

Prerequisites

• C++ compiler
• CMake 3.10 or higher
• Shell environment for running test scripts

Build Instructions

mkdir build
cd build
cmake ..
make

Running

Using CMake build:

./build/cop97

Using scripts:

./scripts/run.sh         # Basic execution
./scripts/parallel.sh    # Parallel execution

Input Data

The program expects traffic data files in the data/ directory:

• data0.txt
• data1.txt
• data2.txt
• dataZ.txt

Algorithm Overview

The COP algorithm works through the following steps:

1. Initialize state variables and value function matrix
2. For each stage (phase):
   • Calculate feasible green times
   • Evaluate performance measures for each decision
   • Update value function
   • Store optimal decisions
3. Continue until convergence or maximum stages reached

Key optimization features:

• Queue length minimization
• Stop minimization
• Delay minimization
• Dynamic Programming approach
• Multiple optimization criteria
• Queue discharge consideration
• Vehicle arrival integration
• Multi-phase support
• Red clearance handling

References

Based on the seminal work: Sen, S., & Head, K. L. (1997). Controlled Optimization of Phases at an Intersection. Transportation Science, 31(1), 5-17.

Contributors

Original authors retain their copyright. This implementation is for academic and research purposes.

Status

This is a research implementation and should be considered experimental. Testing and validation are ongoing.