An improved genetic algorithm in shared bicycle parking point allocation

Abstract

Aiming to solve the problem of inadequate parking places for shared bicycles especially during peak hours, an improved genetic algorithm for parking point allocation is proposed in this paper. We integrate linear regression algorithm with the genetic algorithm to increase the direct of individual mutation, which leads to avoiding falling into local optimum. Meanwhile, we use linear regression to haste the convergence speed of genetic algorithm which ensures the new method can improve efficiency while allocating parking point. For the sake of carrying out the experiment accurately and conveniently, we use geohash to encode the locations of parking points and bicycles into short letters and numbers. According to the analysis of experimental results, it proves the improved algorithm is superior to the conventional method for parking point allocation.