面向变速箱磨损状态评估的灰靶模型优化

Optimization of grey target model for evaluation of wear state of gearbox

为实现变速箱磨损状态的有效评估,提出一种灰靶模型的优化方法。传统灰靶模型中分辨系数的大小通常根据人为经验确定,缺少客观性,这将直接影响模型的分辨能力。该方法结合推土机变速箱磨损实例,提取铁谱分析和光谱分析双因素参数,构建适用于变速箱磨损状态评估的灰靶模型;然后基于非线性方式改进的粒子群算法对模型的分辨系数进行优化,得到优化后的灰靶模型。最后,通过与传统灰靶模型实例进行对比分析,验证了该优化方法对实现变速箱磨损状态评估的可靠性。

The selection of resolution coefficient can directly affect the resolving power of the grey target model,but the determination of resolution coefficient usually depends on human experience in traditional grey target model,lacking the objectivity.To evaluate the wear state of gearbox effectively,an optimization method of grey target model was proposed.Based on the case of bulldozer gearbox,two-parameter of Ferrography and spectral analysis were extracted to build the grey target model suitable for the wear state evaluation of gearbox.The improved particle swarm optimization algorithm based on nonlinear method was used to optimize the resolution coefficient of this model,and the optimized grey target model was obtained.Compared with the traditional grey target model,the reliability of the proposed optimization method in the wear state evaluation of gearbox was verified.