改进BP神经网络的武警装甲车保障能力评估

影响武警装甲车战时平时保障能力的因素很多.针对当前武警装甲车保障能力评估方法的模糊性和不确定性，建立了以武警装甲车保障能力为目标的评估指标体系.采用萤火虫算法优化BP神经网络，确定其初始权值和阈值，并在此基础上提出一种评估武警装甲车保障能力的方法.根据建立的模型进行计算并分析，表明萤火虫神经网络具有更快的收敛速度和更高的准确性，可适用于武警轮式装甲防暴车保障能力的评估.

Evaluation of Support Capability of CAPF ArmouredVehicle with Improved BP Neural Network

There are many factors affecting security of Chinese armed police force (CAPF) armored vehicles in wartime and peacetime. To overcome ambiguity and uncertainty in the evaluation of security capability of CAPF armored vehicles, this paper establishes an evaluation index system using glowworm swarm to optimize a BP neural network. Having determined the initial weights and thresholds, a security of CAPF armored vehicles is evaluated. By establishing a model, calculation and analysis are performed. It is concluded that the glowworm swarm optimization BP (GSOBP) neural network converges fast and is accurate. The method can be used effectively for evaluating security of the CAPF wheeled armored anti-riot vehicles.