基于改进TDABC法的PSO-BP模型在饮料产品生产成本预测中的应用

Application of PSO-BP Model in Prediction of Production Cost of Beverage Products Based on TDABC

为了准确预测饮料产品的生产成本,考虑规模经济购买导致的季节性闲置产能对生产成本的影响,采用时间驱动作业成本(Time Driven Activity-Based Costing,TDABC)法,归集季节性闲置产能成本,构建饮料产品的改进TDABC法生产成本核算模型,使产品单位生产成本的核算更加精确;结合误差反向传播(Error Back Propagation,BP)算法,以饮料产品单位材料费用、生产月份和子作业时间为影响因子,对饮料产品生产成本进行预测。针对BP算法易陷入局部极值和收敛慢等特点,引入粒子群优化(Particle Swarm Optimization,PSO)算法改进BP算法,提出了粒子群优化—误差反向传播(Particle Swarm Optimization-Error Back Propagation,PSO-BP)算法。实验表明,该算法具有较高的预测精度。

In order to accurately predict the production cost of beverage products,considering the impact of seasonal spare capacity caused by economies of scale purchase on production cost,based on TDABC(time-driven activity-based costing),the paper calculates and collects the cost of seasonal idle production capacity,establishes TDABC production cost accounting model to make the unit production cost of beverage products precise;Combined with BP(error back propagation neural network),the unit production cost of beverage products is predicted by using unit material cost,month and operation time of beverage products as influencing factors.In view of the characteristics of BP,such as easy to fall into local extreme value and slow convergence speed,PSO(particle swarm optimization)algorithm is introduced to improve BP,and PSO-BP algorithm is proposed.Experiments showed that the algorithm had higher prediction accuracy.