基于动力总成-整车模型的发动机等效激励参数识别

摘要

The detection of engine excitation in the design of mounting systems is of great importance,but for the engine load identification,the priority is usually given to the inverse matrix method with the disadvantage of heavy workload,time-consuming,etc.A new method for the identification of engine excitation was proposed:a parametric model for the engine equivalent excitation was set up,which was then loaded on the model of a power train-full vehicle,the accelerations on the upper and lower mounts were acquired with the excitation calculated,an objective function was established according to the difference between the experimental accelerations and the simulation results,on this basis,the engine multidirectional excitation was finally identified by estimating the parameters of the equivalent excitation with a genetic algorithm.A car was analyzed by using the method proposed.The acceleration curve fitting is accurate and the load identification results are rather good.The mounting system was optimized in accordance with the load identified and its vibration isolation performance is greatly enhanced.%发动机激励的提取对于动力悬置系统的设计至关重要,但目前发动机载荷识别以逆矩阵法为主,存在工作量大、耗时较长等不足.提出了一种新的发动机激励识别的方法:建立发动机等效激励参数模型,并将该激励加载于动力总成-整车仿真模型上,从而计算出悬置上下支点振动加速度;以仿真与试验加速度的差异为目标,采用遗传算法,对等效激励的参数进行识别,从而识别出发动机多方向激励;通过引入随机因子保证发动机激励识别的准确性.运用该方法对某款轿车进行了分析,所得加速度曲线拟合度高,载荷识别结果较好;运用识别激励进行悬置优化,最终怠速频率下悬置系统的隔振性能大幅度提升.