Using Generative Adversarial Network (GAN) in predicting people's travel based on socio-economic characteristics

Today, one of the most important issues discussed in the field of traffic is stationary transportation or parking lots. There are various types of parking lots, and in the present study, marginal parking lots have been examined. In the present study, marginal parking lots in inner-city streets of Rasht metropolis have been considered as a case study. Given that the purpose of the study is to investigate the effect of marginal parking on delay time at traffic nodes (signalized intersections), first, the current situation scenario for the signalized intersections of the study area is examined and its results are presented using simulation with IMSON software.Next, several scenarios such as demand reduction methods such as parking pricing and examining the structure of intersections, etc. are presented, and the simulation results of the presented scenarios are compared with the current situation scenario.The simulation results show that by eliminating marginal parking and freeing up lanes, travel time and delay time have been reduced to 50% of their current state, and the density of the route has reached about half of its original state, which has caused the speed of movement to increase and the flow to improve. Also, the pricing-based capping scenario improves latency by 25% due to reduced input and demand, and reduced traffic flow reduces speed and reduces stationary traffic, which reduces emissions by 20%.