摘要
测量精度一直是影响车辆动态称重系统有效可靠性的主要因素。针对车辆动态称重系统测量精度较低这个问题,提出了一种基于鲸鱼优化(Whale Optimization Algorithm,WOA)算法和模拟退火(Simulated Annealing,SA)算法混合优化的BP神经网络(Back Propagation Neural Network)动态称重模型。首先,简单介绍了动态称重系统的结构和原理。然后,通过小波变换对动态称重系统的采样信号进行过滤重构处理,经过计算得到的动态车重、车速和轴数作为BP神经网络模型的输入参数。其次,建立了一个由WOSA算法优化的BP神经网络来预测实际车辆总重和轴重。最后,比较了WOSA算法优化的BP神经网络模型的预测能力并得出结论。仿真结果表明,WOSA-BP车辆动态称重模型收敛速度快,精度高,最大总重的相对误差为0.58%,最大轴重相对误差为6.73%。
Measurement accuracy has always been the main factor affecting the effective reliability of vehicle dynamic weighing systems.Aiming at the problem of low measurement accuracy of vehicle dynamic weighing system,a dynamic weighing model based on back propagation neural network and hybrid optimization of whale optimization algorithm(WOA)and simulated annealing(SA)algorithm is proposed.Firstly,the structure and principle of the dynamic weighing system are briefly introduced.Then,the sampling signal of the dy-namic weighing system is filtered and reconstructed by using wavelet transform,and the calculated dynamic vehicle weight,vehicle speed and number of axles are used as the input parameters of the BP neural network model.Secondly,a BP neural network optimized by the WOSA algorithm is established to predict the actual gross vehicle weight and axle load.Finally,the prediction ability of BP neural net-work model optimized by WOSA algorithm is compared and a conclusion is drawn.The simulation results show that the WOSA-BP vehi-cle dynamic weighing model has fast convergence speed and high accuracy,the relative error of the maximum gross weight is 0.58%,and the relative error of the maximum axle weight is 6.73%.
作者
袁科
许素安
富雅琼
徐红伟
YUAN Ke;XU Suan;FU Yaqiong;XU Hongwei(School of Mechanical and Electrical Engineering,China Jiliang University,Hangzhou Zhejiang 310018,China)
出处
《传感技术学报》
CAS
CSCD
北大核心
2024年第1期50-57,共8页
Chinese Journal of Sensors and Actuators
基金
国家自然科学基金项目(62373339)
浙江省自然科学基金项目(LZ24F030007)。
关键词
动态称重
BP神经网络
小波变换
鲸鱼优化算法
模拟退火算法
dynamic weighing
BP neural network
wavelet transform
whale optimization algorithm
simulated annealing algorithm