摘要
针对汽车状态估计过程中噪声的统计特性难以确定的特性,提出一种遗传算法与扩展卡尔曼滤波相结合的车辆状态观测器,对车辆状态进行估计。以车辆非线性动力学模型为基础,使用扩展卡尔曼滤波对车辆的横摆角速度、质心侧偏角和纵向速度等状态参数进行估计的同时,结合在遗传算法中构建的适应度函数,实现对噪声自适应寻优,降低噪声的影响。利用Matlab/Simulink与Carsim仿真软件对所提方法进行验证,并与扩展卡尔曼滤波算法进行对比。结果表明:与扩展卡尔曼滤波相比,该改进的算法有效提高了对横摆角速度、质心侧偏角以及纵向速度估计的精确度,给车辆提供准确的状态信息的同时,利于保障车辆的稳定性。
The active safety technology of automobiles has always been the focus of attention and the key to active safety technology is to master the driving state of the vehicle,mainly including the yaw rate,longitudinal speed,lateral speed and centroid sideslip angle of the vehicle.Hence,an accurate acquisition of vehicle status is of great significance for ensuring vehicle driving safety.The state parameters of the vehicle can be obtained by sensors combined with filtering algorithm.The filtering algorithm can make up for the lack of sensor accuracy and improve the accuracy to acquire vehicle state parameters.A vehicle is a complex system and,when running,system process noise and measurement noise are actually in a changing process.Therefore,both the noise covariance matrix and the measurement noise covariance matrix of the system are constantly changing.Traditional filtering algorithm such as Kalman filtering generally assumes that the covariance is a fixed value,which will affect the accuracy of filtering,resulting in a large deviation of vehicle state estimation.Genetic algorithm is a computational model simulating natural selection and genetic mechanism of biological evolution.Genetic algorithm simulates the evolution of nature in a mathematical way,and expresses the phenomenon of“survival of the fittest”in nature with mathematical probability to solve the problem of optimal solution in the function.In view of the uncertain situation of the statistical characteristics of process noise and measurement noise in vehicle state estimation,a new vehicle state observer based on genetic algorithm and extended Kalman filter is proposed to estimate the vehicle state in the paper.First of all,based on the vehicle nonlinear dynamic model,the extended Kalman filter is used to estimate the state parameters of the vehicle,such as yaw rate,centroid sideslip angle and longitudinal speed.At the same time,by combining extended Kalman filter with the fitness function constructed in the genetic algorithm,the process noise and measurement noise are optimized according to the fitness function,the adaptive effect of noise is realized,and signal noise is cut down.The algorithm is compared and verified through joint simulation of Matlab/Simulink and Carsim.The results show that,compared with the traditional extended Kalman filter,the improved algorithm,with the help of genetic algorithm optimization,can effectively improve the estimation accuracy of yaw rate,centroid sideslip angle and longitudinal velocity,which not only provides accurate state information for the vehicle,but also ensures the stability of the vehicle.
作者
易鑫
陈勇
YI Xin;CHEN Yong(Mechanical Electrical Engineering School,Beijing Information Science and Technology University,Beijing 100192,China;Beijing Laboratory for New Energy Vehicle,Beijing 100192,China;Beijing Collaborative Innovation Center for Electric Vehicles,Beijing 100192,China)
出处
《重庆理工大学学报（自然科学）》
CAS
北大核心
2022年第12期1-8,共8页
Journal of Chongqing University of Technology：Natural Science
基金
科技创新服务能力建设-北京实验室建设项目(PXM2020_014224_000065)。
关键词
车辆动力学
状态估计
遗传算法
扩展卡尔曼滤波
自适应控制
vehicle dynamics
state estimation
genetic algorithm
extended Kalman filtering
adaptive control
