论文部分内容阅读
为提高根据车辆动力学响应建立路面不平度时域模型的精度,对RBF神经网络的设计、输入神经网络的动力学响应参数和汽车车体测量点的位置进行了研究。基于拉格朗日第2方程建立了车身任意位置的5自由度振动模型,以滤波白噪声法建立的路面时域激励为车辆激励的输入和神经网络的理想输出,采用改进的人工鱼群算法(AFSA),针对车身测量点距质心的距离、待测量的动力学响应参数的类型和RBF神经网络的扩展系数建立了优化分析模型。提出了2种需测量的车辆动力学响应参数方案,以及各方案下车身测量点的具体位置。研究结果表明,2种方案的路面不平度时域激励估测精度均高于0.99。
In order to improve the accuracy of time domain model of road roughness based on vehicle dynamics response, the design of RBF neural network, the dynamic response of input neural network and the location of vehicle body measuring point were studied. Based on the Lagrange equation (2), a 5-DOF vibration model with arbitrary position of the vehicle body is established. The road surface excitation established by the filter white noise method is the input of the vehicle excitation and the ideal output of the neural network. An improved artificial fish swarm algorithm (AFSA). The optimization analysis model is established for the distance between the body measurement point and the center of mass, the type of dynamic response parameters to be measured and the expansion coefficient of RBF neural network. Proposed two kinds of vehicle dynamic response parameters to be measured program, as well as the specific location of the measurement points under the various programs. The results show that the accuracy of time-domain estimation of road roughness is higher than 0.99.