改进遗传BP算法在物流需求预测中的应用

Application of Improved Genetic BP Algorithm in Logistics Demand Forecasting

为了提高物流需求预测的准确性,针对现有遗传算法存在的局部最优解等问题,提出了改进遗传BP算法。首先,采用了双种群进化机制;其次,通过双极性压缩函数对适应度函数值进行改进;然后,设计了基于排挤方法的选择算子;最后,利用改进遗传算法对BP神经网络的初始权值和阀值进行优化。仿真实验表明,文章所提算法能够更加准确地对物流需求进行预测。

In order to improve the accuracy of logistics demand forecasting, for the local optimal solution of the existing genetic algorithm, a logistics forecasting algorithm based on improved genetic algorithm was proposed. Firstly, the double population evolution mechanism was adopted; secondly, the fitness function value was improved by the bipolar compression function; then, the selection operator based on crowding method was designed; finally, the initial weights and threshold of BP neural network was optimized by improved genetic algorithm. Simulation experiments show that the proposed algorithm can more accurately forecast the logistics demand.