BP神经网络的地面雷达全寿命周期费用估算

A Method of Estimating the LCC of Ground Radar Equipment Based on BP Nerving Network

为了对地面雷达装备全寿命周期费用做出合理,准确、有效的估算,必须选取合适的估算方法对其进行费用分析。针对一般常用的地面雷达装备全寿命费用估算方法存在费用估算工作量大、预测精度带有很强的主观性和随机不确定性、建模与仿真过程存在很大局限性等问题,从BP神经网络原理出发,提出一种基于BP神经网络的全寿命费用估算方法。该方法充分利用计算机软件资源,建立起地面雷达全寿命周期费用估算的神经网络模型,并结合实例估算。采用Matlab软件语言构造出典型网络的工具函数,根据类似地面雷达装备以往的费用数据作为样本,对费用数据进行训练,调整权值,最后根据估算好的相应数据估算实际费用。通过与传统全寿命周期费用估算方法相比较,证实该模型能够应用于新型地面雷达装备全寿命周期费用估算,减小运算工作量,提高估算精度。

In order to estimate the LCC of new pattern ground radar equipment more logically, exactly and effectively, a right method should be chosen in estimation. In allusion to that there are many difficulties in common methods of estimating the LCC of the ground radar equipment, such as the heavy estimation workload, the estimation precision with quite stronger subjectivity and uncertainty, and the existence of great limitations in the process of the model and emulation, and from BP nerving network, a method is proposed based on BP nerving network in estimating the LCC. By using this method, the full use of the computer software resources is done and a nerving network model of estimating the LCC of the ground radar equipment is set up. In this paper, the Maflab software is adopted to build up an implement function of typical network. Firstly, the author uses the expenditure data similar to the former one of the ground radar equipment as a sample. Then trains the data and modulates the coefficient. Finally, the author estimates the factual expenditure based on the corresponding data. With the estimating and analyzing example, and through comparing with the common estimation method, the result shows that the model can be applied to estimating the LCC of new pattem ground radar equipment and has a higher precision in estimation.