基于小波能量熵的司机疲劳驾驶行为响应算法

Response Algorithm of Driver Fatigue Driving Behavior Based on Wavelet Energy Entropy

为实时、准确识别司机疲劳驾驶行为响应状态,保护司机驾驶过程中的人身安全,研究基于小波能量熵的司机疲劳驾驶行为响应算法。通过匹配追踪算法匹配司机驾驶时的方向盘转向信号,基于稀疏分解对司机方向盘转向信号进行去噪处理,获取纯净的司机方向盘转向信号;基于小波能量熵对去噪后的司机方向盘转向信号进行小波多分辨分析,获取信号的小波能量熵值,通过能量熵测度得到方向盘的修正响应行为,识别司机疲劳驾驶行为响应状态。仿真分析得出,在S形、双道路两种道路工况中,所提算法对驾驶熟练度存在差异的司机驾驶行为响应状态识别结果和实际响应状态相符,识别耗时低于0.5s,且有效提升司机方向盘转向信号信噪比。

In order to identify the driver’s fatigue driving behavior response state in real time and accurately and protect the driver’s personal safety during driving, the driver’s fatigue driving behavior response algorithm based on wavelet energy entropy was studied in the paper. According to the matching tracking algorithm, the steering signal of the driver was matched. Sparse decomposition was introduced to denoise the driver’s steering wheel steering signal in order to obtain the pure driver’s steering wheel steering signal. The denoised driver steering wheel steering signal was analyzed in detail via wavelet energy entropy, thus obtaining the wavelet energy entropy of the signal. According to the energy entropy measure, the modified response behavior of the steering wheel was obtained to identify the response state of the driver’s fatigue driving behavior. The simulation results show that in the two road conditions of S-shaped and dual Road, the recognition results of the response state of the driver’s driving behavior with different driving proficiency are consistent with the actual response state, the recognition time is less than 0.5 s, and the signal-to-noise ratio of the driver’s steering wheel is effectively improved.