【摘要】 汽车噪声是影响汽车品质的重要因素,而后桥主减速器齿轮噪声是微车噪声的主要来源。齿轮故障诊断是一个复杂的非线性问题,目前只能靠人工经验来诊断故障。提出了改进的模拟退火——BP神经网络融合算法,该融合算法克服了BP算法易陷入局部极小的缺点,将其应用于主减速器品质诊断,研究结果表明优于传统算法,效果突出。更多还原

【Abstract】 The auto noise is the important factor affecting autos quality,and the rear axle main reduction gear noise is the important source of micro vehicle noise.At present,the gear failure diagosis can only be diagnosed by artificial experience for it is a complex non-linear problem.This paper proposed an improved simulation annealing BP neural network fusion algorithm,which overcame the outcoming of the BP algorithm easily falling into the local minimum.Applied in the main gear box quality diagnosis,the findings indicated that it surpassed the traditional algorithm,and the effect was splendid.更多还原