The price of anarchy in transportation networks by estimating user cost functions from actual traffic data

摘要: We consider a large-scale road network in Eastern Massachusetts. Using real traffic data the form of spatial average speeds and flow capacity for each segment network, we convert speed to estimate origin-destination demand matrices network. Assuming that observed correspond user (Wardrop) equilibria different times-of-the-day days-of-the-week, formulate appropriate inverse problems recover per-road cost (congestion) functions determining route selection month time-of-day period. Then, system-optimum problem order find socially optimal flows investigate performance, terms total latency, under user-optimal policy versus system-optimal policy. The ratio these two quantities is defined as Price Anarchy (POA) quantifies efficiency loss selfish actions compared ones. Our findings contribute efforts smarter more efficient city.