Improvement of Swarm Intelligence Algorithm and Its Application in Logistics Network Routing

The Vehicle Routing Problem (VRP) is a key aspect of logistics network routing, and an excellent routing optimization strategy can effectively improve the service experience of users and reduce transportation costs. With the explosive growth of e-commerce, the logistics network routing system is faced with a huge number of demand points and customer-specified receiving time slots, which poses a huge challenge for routing optimization of large-scale delivery. To address the problems of high total cost and low effective vehicle utilisation in the logistics network routing system, this work introduces a new swarm intelligence method, the pigeon-inspired optimization (PIO), and improves it. Two improvement strategies are proposed to address the strengths and weaknesses of the PIO algorithm. Firstly, by combining the high swarm dispersion of the quantum evolutionary algorithm and the fast convergence of the PIO, the PIO method is upgraded by mixing the algorithms to achieve the effect of complementing each other’s strengths and improving the global exploration ability of the PIO algorithm; secondly, a Gaussian variation operator is added to the PIO algorithm to enhance its local exploitation capa-bility and prevent prematurity in order to retain the variety in future iterations. Each individual contains information on both client points and routes. The effectiveness of the improved pigeon flock intelligence optimisation algorithm is verified through test function simulations. The effectiveness and rationality of the improved PIO algorithm is verified in a case study based on the Solmon arithmetic example, which has some engineering application value. © 2023, Taiwan Ubiquitous Information CO LTD. All rights reserved.