论文部分内容阅读
采用混合润滑模型分析了柴油机活塞环-缸套系统的润滑性能。选取影响润滑性能的5个主要参数进行正交试验,通过极差分析确定了各个因素对活塞环-缸套摩擦功耗影响的主次关系。建立遗传算法优化的BP神经网络,利用正交试验结果训练该神经网络,得到活塞环-缸套的摩擦功耗神经网络预测模型,然后利用该模型针对选取的5个主要参数进行了优化设计。结果表明,影响摩擦功耗的主要因素由强到弱为:缸套表面粗糙度、活塞环桶面高度、活塞环桶面偏移、活塞环表面粗糙度和活塞环轴向高度。运用正交试验和遗传算法优化的BP神经网络相结合的方法进行活塞环优化设计,试验工作量大大减少,预测精度较好,为活塞环-缸套摩擦学设计提供了便利。
The lubrication performance of the piston-ring-cylinder system of diesel engine is analyzed by the hybrid lubrication model. The five main parameters affecting the lubrication performance were selected for orthogonal test. The primary and secondary relationships between the factors affecting the frictional power consumption of the piston ring and the liner were determined by the range analysis. The BP neural network optimized by genetic algorithm is established. The neural network is trained by the orthogonal test results, and the friction power neural network prediction model of the piston ring - cylinder liner is obtained. Then the optimal design of the five main parameters is carried out by using this model. The results show that the main factors influencing the frictional power consumption are as follows: the surface roughness of the cylinder liner, the height of the barrel surface of the piston ring, the displacement of the piston ring, the surface roughness of the piston ring and the axial height of the piston ring. The optimal design of the piston ring is carried out by combining the orthogonal experiment with the BP neural network optimized by the genetic algorithm. The experimental workload is greatly reduced and the prediction accuracy is better, which provides a convenient tribological design for the piston ring - cylinder liner.