基于BP神经网络的野战给水装备性能评估模型

Appraisal model of performance to field waterworks based on BP neural networks

为了评估野战给水装备的性能,方便部队选购最佳的净水器材,在前人研究的基础上,结合战场需求和当前水质净化技术研究的成果,提出了野战给水装备的指标体系,并利用BP(back-propagation)神经网络对野战给水装备的性能进行了建模,给出了一种新的算法。算例通过对5种野战给水装备的评价结果与专家们的评价结果相比较的方法,验证了模型的可靠性和可行性。得出了该模型能够作为“计算机评价专家”来代替人的评价的结论。可以将此模型应用到部队选购给水装备的实际工作中,方便采购人员对野战给水装备的性能进行横向比较,选出综合性能最高的装备。

Based on the researches of predecessors and in veiw of the needs of war and technology of purification, an index system of filed waterworks was offered and the appraisal model built with neural networks and a new algorithm presented in order to appraise the performance of field waterworks, and choose and purchase the best watercleaner conveniently. The appraisal results of five waterworks compared with those of the specialists prove that the model is credible and practical. This model can replace man and work as a computer-specialist. So this model can be used in purchasing waterworks for our troops, which makes buyers analyze and compare waterworks conveniently and choose good waterworks.