traffic light control using reinforcement learning daniel goldberg andrew elstein final presentation

16
Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Upload: owen-timpson

Post on 31-Mar-2015

217 views

Category:

Documents


3 download

TRANSCRIPT

Page 1: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Traffic Light Control Using Reinforcement Learning

Daniel Goldberg

Andrew Elstein

Final Presentation

Page 2: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

The Problem

• Traffic congestion is estimated to cost Americans $121 billion in lost productivity fuel, and other costs.

• Traffic Lights are imperfect and contribute to this

• Usually statically controlled

• A better method of controlling them can reduce waiting times significantly

Page 3: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Approach

• Implement a “Reinforcement Learning” (RL) algorithm to control traffic lights

• Create a simulation of traffic to tweak and test traffic light optimizations

Page 4: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Implementation• If minor adjustments

were made to the algorithm, it could operate within existing infrastructure

• Optimally, a camera system and would be added

Page 5: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Simulation

Insert picture of visualization

Page 6: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Simulation Structure• To build the simulation we created the follow Data

Structures:

• Cars

Position, Destination, Velocity, Map, Color

• Roads

• Lanes

• Individual Cells

• Intersection location matrix

• Intersections

• Position, Traffic Lights

• In total, the simulation is coded in

MATLAB with 3100 lines of code

Cars Struct

Page 7: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Simulation Dynamics

• Cars are spawned randomly

• They follow an randomly generated path to destination

• Cars follow normal traffic rules

• Road Cells are discretized to easily simulate traffic, only one car can exist in each road cell. Cars move ahead one or two cells in each time-step, depending on the car's max velocity and whether there is an open spot.

Page 8: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Demo

Page 9: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Reinforcement Learning

• Weiring - Multi-Agent Reinforcement Learning for Traffic Light Control

• It introduced an objective function to minimize or maximize a goal value

  

tl = traffic lightp = current position d = destinationL = light decision = discounting constant ‘ = next

Page 10: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Reinforcement Learning Theory

• Coordinating a system of lights to respond to current conditions can reap exceptional benefit

• The theory cleverly merges probability, game theory and machine learning to efficiently control traffic

• In our case, the expected value of each of a light’s possible states are calculated

• With this value function a game is played to maximize it, in turn minimizing waiting time

Page 11: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Results

Wrote a script to compare the smart algorithm to static On-Off-On-Off lights.

Our algorithm reduced average waiting time—and thus traveling time—for a system with any number of cars

Travelling time for our implementation was reduced by an average of 10%. There was a 15% reduction for sparse traffic systems from a static control, but only a 3% decrease for heavy congestion.

Page 12: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Results cont.

Page 13: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Extensions• Fairness-weighted objective:

•  

• ω = weighting constant

• t = current time

• ti = time of arrival for car i

•  

• If F(t) > 1, cars on road 1 get to go

• If F(t) < 1, cars on road 2 get to go

Page 14: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Further Extensions

• Car Path optimization and rerouting

• Model expansion to traverse an entire city

• Inter-traffic-light communication

• Retesting with increased computational resources for modeling accuracy and robustness

Page 15: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

RL In the News• Samah El-Tantawy, 2012 PhD recipient from the

University of Toronto, won the 2013 IEEE best dissertation award for her research in RL.

• Her RL traffic model showed reduced rush-hour travel times by as much as 26 percent and is working on monetizing her research with small MARLIN-ATSC (Multi-agent Reinforcement Learning for Integrated Network of Adaptive Traffic Signal Controllers) computers.

Page 16: Traffic Light Control Using Reinforcement Learning Daniel Goldberg Andrew Elstein Final Presentation

Challenges

• Difficult to understand data structures and how they would interact

• Object Oriented Approach vs. MATLAB’s index-based structures

• Understand how cars would interact with each other

• Understanding RL algorithm

• Adapting our model to use RL algorithm

• Limited computational resources