Urban Bus Stop Site Selection using Multi-Layer Perceptron Neural Network and Error Back-Propagation Algorithm (Case Study: Kermanshah City)

Transportation has always been one of the most important factors, which affects the structure of cities. However, the development of a variety of motor vehicles and ever-increasing population changes, especially in the last century, have become one of the main urbanization problems. Considering the volume of inland traffic in Kermanshah City, the standard site selection of bus stations is one of the most important parameters that improves the proper operation of public transportation in the stations. This includes the reducing of Get-of and the Get-in time of the users as well as making less traffic jam of other vehicles. In this study, the multi-layer perceptron with the error-back propagation algorithm as one of the most important structures of the neural networks is used. Afterward, 15 map layers were used as the effective data to select bus station locations. In addition, 500 layers were prepared as network teaching points and 10 intermediate layers were determined. Then, the optimized site zones were obtained by implementing a network of susceptible zones. Eventually, the stations were able to be constructed by identifying the passages in the susceptible zones. It was also observed that susceptible passages are located near demographic, downtown, cultural and commercial centers.