一种基于熵权多目标决策和人工神经网络的炼油企业绩效评价方法

An approach based on entropy-weighted technique for order preference by similarity to ideal solution and artificial neural network for oil refining enterprises performance evaluation

提出了一种基于熵权多目标决策的逼近理想解法(TOPSIS)和人工神经网络(ANN)的炼油企业绩效评价方法,以熵权TOPSIS的企业绩效评价结果作为学习样本,对神经网络进行训练、测试,进而对指标加以赋权,最终得到了企业绩效综合评判式,并将其用于炼油企业绩效评价。实例分析结果表明,该方法科学有效、实际可行,具有一定的智能性,为炼油企业绩效评价提供了一种新的途径。

An approach bssed on entropy-weighted technique for order preference by similarity to ideal solution （TOPSIS） method and artificial neural network （ANN） was proposed for oil refining enterprises performance evaluation. Using the resuits of entropy-weighted TOPSIS method as learning sample to train and test the artificial neural network, the weight of performance indicator and a synthetic evaluation formula were obtained. The oil refining enterprises performance evaluation was calculated by the formula. An example testifies the efficiency, practicability and intellectual ability of the method.