本文提出了基于粒子群（PSO）的训练ANN 的新算法，以此为基础建立了对库存品进行ABC 分类的模型。新算法充分结合了PSO 与BP 两者的优势，在训练过程中能同时优化权值以及神经元log-Sigmoid函数。实验结果表明，新算法是企业库存信息管理系统中进行决策预测的一种可行方法。关键字：粒子群（PSO）算法；ABC 分类；神经网络；反向传播算法Abstract: This paper presents a new ANN training algorithm based on PSO for ABC classification of stock keeping units (SKUs). The new algorithm combines the strengths of BP training algorithm and PSO-Gain training algorithm; it can both optimize the weights of network and the log-sigmoid function of the network during the training process. The experiment results show that the new algorithm is a viable way for decision-making in information management system of enterprise inventory.Keywords: Particle swarm optimization; ABC classification; Neural network; Back propagation