基于径向基小波神经网络的装备保障方案评价模型

针对装备保障方案评价的实际问题,利用小波函数的多分辨率分析和逐层逼近能力,建立了基于径向基小波神经网络的装备保障方案评价方法。通过对装备保障方案中评价参数的量化,建立评价参数的样本库。将各种性能指标输入到训练好的网络,客观地评价保障方案的性能。以陆军防空旅装备保障方案为例,组织保障领域的专家对评价参数进行综合研讨,得到量化的参数值,然后使用新方法对其进行评价。仿真结果表明,径向基小波神经网络具有良好的非线性映射能力和泛化性能,能够比较准确地对装备保障方案做出评价。

Evaluating for Equipment Support Plan Based on Radial Wavelet Basis Function Neural Network

Aiming at evaluating the equipment support plan, this paper lead into wavelet function with multi-resolution approximation and a new method based on radial wavelet basis function neural network is proposed. Training samples is obtained by quantifying evaluating parameters. Then the testing data as the input of the trained network is introduced to objectively evaluating the whole effect of the equipment support plan. The simulation results indicate that radial wavelet basis function neural network owns nonlinear mapping capacity and generalization, and this method can evaluate the equipment support plan efficiently.