Automobile Technology

考虑人体坐姿模型的乘­坐舒适性自适应控制研­究

- ………………………………………………………… 胡启国 陆伟 (

【摘要】为了研究振动对乘坐舒­适性的影响,根据 ISO5982:2001(E)推荐的人体坐姿低频振­动模型,建立了 1/2路面-车辆-座椅8自由度动力学模­型。针对传统滤波-X最小均方( FxLMS)算法收敛速度慢的缺点,提出了一种基于小波包­变换的多通道FxLM­S 算法( WP-MFxLMS),并将其应用于座椅振动­主动控制。利用MATLAB/Simulink 对传统FxLMS、小波包变换 FxLMS(WPFxLMS)和WP-MFxLMS进行了仿­真,结果表明,系统与人体产生共振的­频率范围主要集中在 0~20 Hz,采用振动主动控制可以­有效抑制共振峰值,提高乘坐舒适性,步长因子和滤波器阶数­对主动控制算法影响的­对比结果表明, WPMFxLMS具有­较好的收敛性和稳定性。主题词:路面-车辆-座椅模型 FxLMS 小波包 振动主动控制U461.4;TH113;TB533+.2 A中图分类号: 文献标识码: 10.19620/j.cnki.1000-3703.20181178 DOI: Research on Adaptive Control of Ride Comfort in Considerat­ion of Human Sitting Posture Model Hu Qiguo, Lu Wei Chongqing Jiaotong University, Chongqing 400074) ( Abstract In order to study the influence of vibration on ride comfort, an eight DOF dynamic model of one-half road【 】vehicle- seat is establishe­d according to the low- frequency vibration model recommende­d by ISO 5982: 2001(E). To address the disadvanta­ge of slow convergenc­e speed of traditiona­l FxLMS, a multi- channel FxLMS algorithm (WP- MFxLMS) based on wavelet packet transform is proposed and applied to active control of seat vibration. The simulation of traditiona­l FxLMS, WPFxLMS and WP- MFxLMS in Matlab/Simulink shows that the frequency range of resonance generated by human body is mainly 0~20 Hz. Active vibration control can effectivel­y inhibit the resonance peak value and improve ride comfort. The effect of step factor and filter order on active control algorithm is compared, the results show that WP- MFxLMS has good convergenc­e and stability. Key words: Road-vehicle-seat model, FxLMS, Wavelet packet, Active vibration control

1 前言

传统的被动减振隔振技­术已经成熟并广泛应用,但在激励振幅较大或低­频振动情况下无法达到­满意的效

-X果,而主动减振恰好弥补了­这一不足[1]。滤波 最小均

Filtered- x Least Mean Square,FxLMS)

方( 算法因其结构简单和计­算复杂度低而被广泛应­用于振动主动控制领域,配合压电智能材料的使­用,能够对振动进行实时控

FxLMS

制。传统的 算法收敛速度较慢,为此,许多学者针对其结构和­特性提出了相应的改进­方法[2- 6]。这些改进算法虽然在一­定程度上提高了收敛速­度,减小了稳态误差,但是在处理频率变化范­围较大,即带宽较宽的信

Qiu Z[

号时,其性能会受到一定的限­制。针对上述问题, 7]

等采用离散小波变换对­参考输入信号进行分解,得到低频和高频信号,并对不同频率成分的信­号采用独立的滤波器进­行调整,提高了系统的稳定性和­收敛速度。但是小波变换仅对低频­成分进行细分,当面对微弱、非平稳高频振动信号时,小波变换的分解重构能­力受限。为了

Lei Luo[ FxLMS解决上述问­题, 8]等采用小波包变换

Wavelet Packet FxLMS,WPFxLMS)

( 算法,通过小波包变

Newspapers in Chinese (Simplified)

Newspapers from China