节点文献

车辆油气悬架电液伺服加载系统控制方法研究

Study on Control Methods for Electro-hydraulic Servo Loading System of Vehicle Hydro-pneumatic Suspension

【作者】 张辉

【导师】 韩俊伟;

【作者基本信息】 哈尔滨工业大学 , 机械电子工程, 2012, 博士

【摘要】 油气悬架系统具有功率密度高、安装空间小的优点,能提高车辆行驶时的舒适性和安全性,广泛应用于各种机动车及场内车辆中。油气悬架试验系统用于模拟重型车辆在不同路面条件和速度下悬架系统的动态响应,验证轮式车辆悬架及车桥的设计参数,提高悬架系统及车桥的可靠性。本文以哈尔滨工业大学电液伺服仿真及试验系统研究所承担的“油气悬架加载试验台”项目为背景,对油气悬架加载试验系统中采用的关键控制技术进行了理论研究和实验研究,研究成果对于提高系统控制性能及加载精度具有积极意义。文中建立了由电液伺服阀控缸系统和油气悬架组成的加载系统的数学模型,对油气悬架高度的非线性以及加载系统固有频率低、阻尼小的特点进行了分析。引入三状态控制策略,应用极点配置的方法对控制器参数进行设计,在保持系统稳定性的基础上拓展了系统的频宽,改善了系统的动态特性。设计了位置与加速度信号切换的输入滤波器,使系统保持性能不变的情况下,实现位置和加速度两种控制方式。通过仿真分析和实验测试,验证了三状态控制器的有效性和切换位置和加速度控制方式的可行性。主动驱动力加载结构是目前最为常用的加载结构,力传感器安装于加载系统与试验负载之间,检测加载时的弹性力和粘性力以及惯性力,其缺点是控制特性易受到负载刚度的影响。常规PID控制为了达到稳定性的要求,控制参数往往选取比较保守,系统性能受到制约。针对油气悬架的高度非线性,文中采取变结构的控制方法,选取含有位置、速度和加速度状态的滑模函数。利用指数趋近方式,使系统能够快速地稳定于平衡位置。基于Lyapunov稳定理论,设计了参数摄动和不确定量两种非线性环节的自适应律,结合自适应控制方式减小了滑模带的宽度,从而减小由于开关控制而引起的系统振动。仿真分析和实验验证了当负载刚度变化时,采用变结构滑模力加载控制策略的优势,为具有状态反馈的力加载控制系统提供了实用的控制方法。正弦信号是协调加载系统控制中常用的工程信号。在多通道同时对同一试件进行加载时,各通道响应正弦信号幅值和相位的差异会使实验结果与实际情况有着很大的差异。本文针系统对正弦信号的精确复现控制要求提出了基于单层双通道神经网络控制方法。采用LMS算法自动调整权值向量,使系统输出正弦波的幅值、相位与期望信号一致,从而实现正弦波信号的精确复现。通过对算法收敛性必要条件的分析,提出了归一化正弦波复现的控制方法。经仿真分析和实验研究,验证了该方法在收敛性和收敛速度方面的优势。时域波形复现与激振台架是实验室内再现路谱的必要条件。时域波形复现(TWR)的目的是将跑车场测试得到的车辆响应转化为激振台架的输入信号。传统时域波形复现方法采用频响函数迭代的方式,在非线性系统中应用时会变得不易收敛,耗费大量时间。本文提出了基于自适应逆控制的方法,降低迭代次数。应用递推最小二乘算法设计了能够实时辨识系统传递函数和阻抗函数,并在线调整系统逆函数的控制策略。经过仿真分析和实验,验证了控制算法的稳定性。根据加载系统的要求研制了包括液压源、机械部分、液压伺服加载部分和数字式控制系统在内的油气悬架加载试验台。采用Labview SIT快速控制原型技术设计了基于以上控制策略的数字式控制软件,对所设计控制策略进行了实验研究。实验结果验证了本文所提出的各种控制技术的有效性和先进性。

【Abstract】 The hydro-pneumatic suspension have been system used increasingly for variouson-road and off-road applications along with driving safety and comfort due to thehigh power density of the hydraulics and installation space. The hydro-pneumaticsuspension test system promises to simulate the dynamic responses of a heavy dutyvehicle suspension systems adapting to different terrains and speeds, validates thedesign parameters and improves reliability of the hydro-pneumatic suspension andvehicle axle in wheeled vehicles.This thesis investigates the control of a Hydro-pneumatic Suspension TestSystem developed by IEST of HIT. The development, modeling and algorithms ofpivotal control technology have been theoretically studied. The results are validatedthrough experimental activities done by a suspension bench test. The main resultsare valuable to improve the control performance and the precision of simulation.The test system constructed by a hydro-pneumatic damper and anelectro-hydraulic servo-valve controlled cylinder is modeled.The highly nonlinearcharacter of the hydro-pneumatic damper suspension system and low naturalfrequency and small damping of the loading system rre investigated. TheThree-states Control Strategy is developed for the test system, the controlparameters are designed via pole placement, which improves the dynamiccharacteristics such as extending the frequency width of system and keep the systemstable. An input filter switch between the position and acceleration signals isdesigned, and position and acceleration control schemes are implemented withoutchanging the control performance. Simulation and experimental results indicate thatthe three-state controller is effective and the method of switch control schemebetween position and acceleration controller can be achieved.The active force loading is a popular loading control method in which the forcesensor is fixed between cylinder and suspension system to measure the loading force,viscous force and inertia force. However, the disadvantage is that the controlperformance of test system may change along the stiffness fluctuation of thehydro-pneumatic damper suspension system. In common PID control,the gain ofcontroller is always low to keep system stability, thus the system characteristics arelimited. This thesis presented a variable structure controller for the highly nonlinearsystem. The therr-state control strategy including position, velocity, and acceleration information is selected as cascade function. The stability and robustness are enhancedby making use of exponential convergence. Based on the Lyapunov function, anadaptive control law is developed for the nonlinear system under parameterdisturbance and non-parametric uncertainty, which is also adopted to reduce thevibration of responses on cascade surface. Simulation results illustrate the advantageof variable structure force loading control strategy, it proposes a novel controlscheme for force loading system with states feedback.Sine signal is commonly used in control of a coordinated loading system. Whena sample is tested by multichannel at one time, the amplitude and phase of responsesignals from each channel are different. As a result, the output of test will contrary tothe expectation. This thesis introduces a control method based on single-layer&two-channel neural network to meet the requirement of an accurate recurrence control.A LMS algorithm is used to adjust the weight vector automatically and make theamplitude and phase of outputting sine signal consistent with the expected signal.Thus, the precise sine wave repetition is achieved. By analyzing the necessaryconditions for convergence of the algorithm, this thesis presents the normalizedsinusoidal recurrence control method. Advantages of this method in convergence andrapidity have been verified through the simulation and experimental analysis.Experiments on vibration test rigs in combination with time waveformreplication (TWR) can create repeatable testing conditions of the road. The objectiveof TWR is to determine the inputs to the shakers in order to obtain the same sensorreadings as during the field experiments. The classical TWR method consists of aniterative procedure that can be time consuming and divergence for a nonlinear system.This thesis presents a feedback approach based on adaptive inverse control, therebyreducing the number of iterations. By means of recursive least square (IRLS) method,a nonlinear system transfers function and its inverse function identification controllercan be designed. Experiments indicate that with the present control algorithm a betterstability can be achieved.The hydro-pneumatic suspension loading test-bed consists of a hydraulic source,a mechanical system, a hydraulic servo system and a digital control system. Based onLabView SIT (Simulation Interface Toolkit) fast control prototype technology, thedigital control software is developed to carry out the proposed control strategy.Experiments indicate that the present control schemes are effective and advanced.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络