Improvement of Genetic Algorithm and the Application in Computer Simulation Model of O2O Delivery Strategies

The popularity of the Internet has provided the foundation for the rapid development of Online to Offline (O2O) business. With the increase of the number of users, online food ordering platforms are facing more and more challenges and pressures. The efficiency of order processing directly determines the customer satisfaction with the platform and the comprehensive competitiveness of the platform. In this paper, The Genetic Algorithm (GA) is applied to solve the TSP problem to optimize the delivery path of riders, so as to improve the delivery speed of orders. By comparing the efficiency of GA and Dynamic Programming (DP) algorithm in a simulation model of O2O system developed by SUMO, we found that the dynamic programming will not be applicable when the number of TSP nodes is beyond a threshold, i.e., 10 in this case. To improve the efficiency further, multiple genetic algorithms were run in a manner of parallel computing for the distribution strategy of takeout orders, which means that an independent genetic algorithm serves for processing the TSP route of an individual rider.