Micro-Evolution Algorithms for Solving the Dynamic Location Problem of Customized Bus Stops

Abstract

Customized bus is one of the new public transit services by applying GPS, Internet+ and other new technologies which involves in the bus stops location and bus routing problem. In the Customized Bus Stops Location Problem (CBSLP), passengers specify transportation requests with their origins, destinations, and desired pickup or delivery times through internet in real time. The Customized bus operator or transit company will determine dynamically the bus stops from their candidate stops to best accommodate the demand and try to minimize the walking distance of passengers from their homes (or offices) to the bus pickup stops. This paper addresses the dynamic location problem of customized bus stops with maximum walking distance constraints. The micro-evolution algorithm (MEA) is designed based on gene structure evolution method for the solution of the problem. The traditional Primal-Dual Mapping Operator is improved by MEA to capture the changed environment information. Simulation tests show that the MEA performs better than conventional GA both in quality and efficiency in dynamic environment. Thus, an improving tool for better decision of customized bus system is provided.