摘要
针对车辆安全辅助系统中对车辆运行状态识别率偏低的问题,采用小波神经网络对车辆运行状态进行识别。为了进一步提高模型的识别准确率以及减少训练时间,对样本进行主成分分析、卡尔曼滤波,最后利用遗传算法优化小波神经网络。通过对优化后的小波神经网络对数据进行训练与测试,测试结果表明在时间窗口1.8s时模型的识别率能达到91%以上,可以满足车辆安全辅助系统对于车辆状态识别的要求。
Aiming at the problem of low recognition rate of vehicle running state in vehicle safety assistant system,wavelet neural network is used to identify the vehicle running state.In order to improve the recognition accuracy and reduce the training time,the samples are normalized,principal component analysis,Kalman filter.Finally,the wavelet neural network is optimized by genetic algorithm.The data are trained and tested by the optimized wavelet neural network,the test results show that the recognition rate can reach above91%when the time window1.8s.It can meet the requirements of vehicle safety assistant system for vehicle state recognition.
作者
崔宇
黄晓梦
Cui Yu;Huang Xiaomeng(School of Automobile, Chang'an University, Shaanxi Xi'an 710064)
出处
《汽车实用技术》
2017年第8期112-114,共3页
Automobile Applied Technology
关键词
车辆运动状态
神经网络
卡尔曼滤波
主成分分析
遗传算法
vehicle status
neural network
Kalman filter
principal component analysis
genetic algorithm
