Study of Regional Logistics Demand Forecasting methods based on Quantum Particle Swarm Optimization

This paper considers the regional logistics demand forecasting problem (RLDFP) in regional logistics planning issues with the feature of coexistent stabile subsystem and mutative subsystem. Development of stabile subsystem is smooth and can be forecasted with methods of trend extension, while mutative subsystem is change in step and can´t be forecasted with methods of trend extension. For this complicated problem, we present method framework combining quantitative and qualitative analysis. As for stabile subsystem, we propose quantum particle swarm optimization combination method (QPSOCM) which is based on quantitative analysis and can obtain optimized forecasting results. And as for mutative subsystem, we propose decomposition statistics method (DSM) which qualitatively analyses the components of the subsystem and then accumulated forecasting indicators. Computations show that solutions from QPSOCM are better than the traditional methods as far as the total deviation between the actual values and forecasting values is concerned.