Abstract: Supply chain collaborative management is coordinated through the planning，organization and control of human，material，financial and information of supply chain system to achieve their goals.How about their collaborative management effectiveness? It can be verified through the performance evaluation which is one of important links of collaborative management. This article sets up a supply chain collaborative management performance evaluation model based on balanced scorecard，heuristic attribute reduction and BP neural network，and gives a practical example，firstly reduces its index based on the balanced scorecard system，then inputs the reduction index to BP neural network for intelligent training，Finally，inputs the evaluated sample to the trained network BP and gets evaluation value.The evaluating result is matched with the actual result.

Key words: balanced scorecard, heuristic attribute reduction, BP neural network, supply chain collaborative management, performance evaluation

摘要： 供应链协同管理是通过对供应链系统的人力、物力、资金及信息等资源的规划、组织和控制，来实现其目标。供应链协同管理的效果如何，可通过绩效评估进行检验，他是协同管理的一个重要环节。使用平衡积分卡、粗糙集中启发式属性约简的方法和BP神经网络的理论，探索建立了平衡积分卡、启发式属性约简和BP神经网络相结合的供应链协同管理的绩效评估模型。并结合一个实例，首先对其基于平衡记分卡的指标体系进行了约简，然后将约简的评价指标输入到BP神经网络中进行智能训练，最后把评估的样本输入到训练好的BP网络中得出供应链协同管理的绩效评估值，评估结果与实际结果基本吻合。

关键词: 平衡积分卡, 启发式属性约简, BP神经网络, 供应链协同管理, 绩效评估