基于Delphi-BP神经网络的装备保障能力评估

针对评估指标数量过多可能给评估过程及结果带来的弊端,在构建装备保障能力评估指标体系基础上,给出了基于Delphi法的指标体系筛选的方法步骤,通过计算累计贡献率,对指标体系进行了筛选,降低了评估模型的输入维度。建立了评估部队装备保障能力的三层BP神经网络模型,利用Matlab神经网络工具箱对网络进行了训练和仿真,结果显示误差小于10-3。最后,通过对比评估,验证了方法的有效性和正确性。

Evaluation of Equipment Supporting Capacity Based on Delphi-BP Neural Network

To solve the problem which is induced by excessive evaluation indexes ,this article builds index system of equipment supporting capacity,the method of filtering index system based on the method of Delphi is put forward and the index system is filtered by computing accumulative total contributing ratio which reduced the inputting dimensions of model. This paper builds three layers BP Neural Network model,the Neural Network is simulated by Matlab neural network toolbox which result error less than 10-3. In the end the model in contrast with traditional application of neural network method is validated.