【作者】 梁春英；

【导师】 纪建伟；

【作者基本信息】 沈阳农业大学 ， 农业电气化与自动化， 2013， 博士

【摘要】 电控液压系统以响应速度快、输出功率大、控制精度高的特点,适应大型变量施肥播种机的作业需求,在变量施肥控制系统中得到广泛应用。由于液压伺服系统存在时变参数以及外干扰引起的不确定性,有些参数难以用理论方法确定,而且随着工况的不同和压力流量等主参数的变化,这些参数的值也会发生变化。建立准确描述电液系统动态特性的数学模型是控制系统性能研究的前提和关键,本文提出一种系统参数辨识建模法,该方法以系统的试验输入输出数据为依据,融合参数估计和误差准则建立电液变量系统的数学模型,比以往机理分析建模法更能准确表达系统状态传递关系。针对常规PID控制规律采用线性定常组合方案,难于协调快速性和稳态性之间的矛盾,鲁棒性能欠佳,提出对参数摄动和外部干扰具有自适应能力的变结构控制策略,克服非线性和参数不确定对系统性能造成的不利影响。并针对变结构控制的抖振缺陷提出了模糊控制优化趋近过程,抑制抖振提高系统动态品质和稳态精度。本研究从非线性、时变液压系统参数辨识角度出发,探索基于系统辨识的电液比例变量施肥系统数学模型的途径,并采用具有较强鲁棒性的变结构控制结合智能控制,提高系统的鲁棒性和跟踪特性,进而提高变量施肥播种机的施肥精度。本研究主要完成了以下内容：1.利用系统辨识法建立电液比例变量施肥系统的数学模型。通过机理分析法,确定系统的最高工作频率,得到M序列的时钟周期Δt与序列长度Np。依据信号发生器产生的辨识输入信号M序列和数据采集仪获取的液压马达转速,选择一种模型类,利用最小二乘法,对系统进行模型参数的辨识,再由最终预报误差准则(FPE),进一步系统阶次。本研究根据系统实际工作状况,分别选择带控制量的自回归模型ARX和带控制量的平均自回归模型ARMAX,通过MATLAB编程获取系统的参数模型,并对比研究两类模型的拟合度,确定ARMAX4431为该电液变量施肥系统的数学模型。2.研究电液变量施肥系统的滑模变结构控制策略。利用变量施肥控制系统的辨识模型ARMAX4431,转换为状态空间模型,将模型转换为能控标准型,并通过坐标变换法和最优二次型法确定比例切换函数的系数,选择指数趋近律,利用广义滑模和系统模型确定滑模控制策略,实现电液变量施肥系统的滑模变结构控制。利用MATLAB编程对比研究了PID控制和滑模变结构控制(SMC)的阶跃跟踪及正弦跟踪特性。3.研究电液变量施肥系统的离散滑模变结构控制策略。计算机参与控制的电液变量施肥系统是一个离散时间系统,状态变量是采样周期的时间序列,趋近运动只能形成准滑动模态。利用MATLAB仿真研究等效控制离散变结构和离散指数趋近律变结构控制性能,结果可以看出系统状态在一定宽度的切换带内是形成准滑动模态运动,针对这一缺陷,提出变速指数趋近律削弱准滑动模态的抖振幅度,改善了准滑模稳态阶段的运动性能。考虑指数趋近律在系统稳态时,系统状态在切换带内往复运动,不能稳定于零点；变速指数趋近律渐进稳定于原点,但是在切换初期会产生幅度较大的抖振,综合二者优点研究组合趋近控制律,达到指数趋近的优化控制。4.研究电液变量施肥系统的模糊滑模变结构控制策略。模糊滑模控制,采用二维模糊滑模控制器,以滑模切换函数s及其导数s的模糊值为输入,通过模糊逻辑推理,根据距离切换面的位置调节切换增益消除抖振,模糊自适应滑模控制,为二输入单输出模糊控制器,取切换函数s(k)及其变化率ds(k)作为输入,fs(k)作为输出,模糊推理符号函数增益,消除抖振。5.试验研究电液变量施肥系统PID、滑模变结构和模糊滑模变结构跟踪性能和抗干扰性能,系统初始状态运行,采集系统的采样频率1000Hz,记录液压马达转速和控制器输出时域数据文件,试验数据进行处理,分别研究三种控制策略的阶跃跟踪性能和正弦跟踪性能,并对系统稳态误差和抗干扰性能进行对比研究。研究结果与结论如下：1.通过参数模型辨识法研究电液变量施肥控制系统的数学模型,采用最小二乘法辨识系统模型参数、最终预报误差准则确定系统的结构,得到本系统拟合程度好的模型为ARMAX4431。2.通过PID控制和滑模变结构控制(SMC)的跟踪特性和抗干扰特性对比研究,结果表明PID控制稳态性能好,但动态调节时间长、抗干扰能力差；SMC控制动态调节快速,鲁棒性好,但存在抖振,影响系统稳态性能。3.通过电液变量施肥系统离散滑模变结构性能研究,结果表明等效控制动态调节迅速,系统响应有振荡,并存在抖振；离散指数趋近律是准滑模运动,系统存在抖振,不能渐进稳定于零点；变速指数趋近律调节了切换带形状,系统状态渐进稳定于零点,但切换初始阶段存在大幅度抖动；组合控制综合二者优点,实现有限步长内趋近滑模切换面,并渐进稳定零点,削弱了抖振。4.通过切换增益和指数趋近律模糊滑模变结构控制性能的研究,结果表明模糊自适应滑模趋近律控制系统的跟踪性能更好,控制输出更稳定,系统的抖振更小。5.通过对PID控制,SMC控制、模糊SMC控制的动态特性、稳态特性和抗干扰特性的试验研究,结果表明模糊SMC在稳态特性和抗干扰特性方面都优于前两种控制,而且较好地抑制了系统抖振。因此,利用系统参数模型辨识法可以获得更精确的数学模型,变结构控制系统调节快速且无超调,但存在抖振,离散组合控制和模糊指数趋近律控制都可以削弱抖振,但模糊控制指数趋近律在跟踪性能方面优于组合控制,而且抑制抖振效果更优。更多还原

【Abstract】 Electrically controlled hydraulic system has fast response, large output power and high control precision, suitable for operation demand of large-scale variable fertilizer and seed drill, widely used in variable fertilization control system. Because the hydraulic servo system is uncertainty caused by time-varying parameters and external disturbance, some parameters are difficult to determine with theoretical method, and with the changes of main parameters such as pressure and flow conditions, the value of these parameters will change. The premise and key of control system is to establish an accurate mathematical model to describe the dynamic characteristics of electro-hydraulic system, pose the system identification modeling method, this method based on the input-output data of the system, fusing the parameter estimation and error criterion to establish the mathematical model of electro hydraulic variables system, which lays the foundation for the research of control algorithm. According to the law of conventional PID control which using the linear combination scheme, is difficult to coordinate the contradiction between rapidity and stability, and it’s robust performance is poor, the variable structure parameter perturbation and external disturbance with adaptive control strategy is given to overcome the nonlinearity and parameter uncertainty impact on system performance. Aiming at the defects of the chattering variable structure control is proposed to optimize the reaching process fuzzy control, suppress chattering and improve the system dynamic performance and steady-state precision. This study based on the nonlinear, time-varying parameter identification of hydraulic system perspective, and explore the way of system identification of the electro-hydraulic proportional variable fertilization system of mathematical model, and the variable structure robust control with intelligent control, to improve the system robustness and tracking features, and then to improve the variable fertilization seeder fertilization precision.Research contents:1. To establish the mathematical model of the electro-hydraulic proportional variable fertilization system based on system identification method. Through mechanism analysis, to determine the maximum working frequency of system clock cycle and the length of the sequence, M sequence. Based on the signal generator of the input signal for identification of M sequence and data acquisition instrument to obtain hydraulic motor speed, choosing a model class and using the least square method for identification of model parameters on the system, then from the final prediction error criterion (FPE), further to determine the optimal control system of electro-hydraulic variable fertilization model. In this study, according to the actual working condition, respectively choose autoregressive model ARX with controlled auto regressive moving average amount of ARMAX with control volume, identified model parameters on the system by least square method through MATLAB programming and the final prediction error (FPE) criterion for model structure identification, and a comparative study of the two model fitting degree, determined that ARMAX4431is the mathematical model of the electro-hydraulic variable fertilization system.2. Research on electro hydraulic variable fertilization system sliding mode variable structure control strategy. Using the ARMAX4431identification model of control system of variable rate fertilizer applicator, transformed into state space model, the model is converted into the control standard, and through the coordinate transformation method and optimal two quadratic coefficient method to determine the selection of proportion switching function, exponential rate, determine the sliding mode control strategy using the generalized sliding mode and system model, the sliding mode of electro-hydraulic variable fertilization system with variable structure control. Research on PID control and sliding mode variable structure control using MATLAB programming contrast (SMC) step tracking and sinusoidal tracking characteristic; control using MATLAB programming of PID, SMC control step tracking and sinusoidal tracking response characteristics.3. Discrete sliding mode variable structure control strategy of the electro-hydraulic variable fertilization system. Electro hydraulic variable fertilization system on computer is involved in the control of a discrete time system, the state variable is the time series of the sampling period, the formation of quasi sliding mode. By using MATLAB simulation study on the equivalent discrete variable structure control and discrete exponential reaching law of variable structure control performance, the system state switching in a certain width of the band is the formation of quasi sliding mode motion, to solve this problem, the variable exponent reaching law chattering amplitude of quasi sliding mode is mentioned to improve the motion performance of quasi sliding mode state stage. Considering the exponential reaching law in the state system, the system state in the reciprocating motion in band switching, can not be stable at zero; variable exponent reaching law asymptotically stable at the origin, but in early switching may produce buffeting larger amplitude, integrated two advantages of combined reaching control law, to optimize the control index approach.4. Fuzzy sliding mode of the electro-hydraulic variable fertilization system variable structure control strategy. Fuzzy sliding mode control, which use the two-dimensional fuzzy sliding mode controller, has input from the fuzzy sliding mode switching function5and its derivative values s(k), by the fuzzy logic reasoning, according to the position of the switching surface distance adjusting switch gain to eliminate the chattering, adaptive fuzzy sliding mode control, two input single output fuzzy controller, the switching function s(k) and its change rate ds(k) as input, the output fs(k) fuzzy reasoning, the sign function gain, eliminate chattering.5. Experimental study on electro hydraulic system of variable rate fertilizer PID, sliding mode variable structure and fuzzy sliding mode variable structure tracking performance and anti-jamming performance, the initial state of system operation, acquisition system sampling frequency of1000Hz, record the hydraulic motor speed and the controller output time domain data files, test data for processing, respectively, of three kinds of control strategy for step tracking performance and sinusoidal tracking performance, and compare the system steady-state error and anti disturbance performance.Results and conclusion:1. Through mathematical model parameter identification model of electro-hydraulic control system of variable rate fertilizer applicator, parameters, using least squares identification system model to determine the final prediction error criterion system structure, model fitting degree of the system for ARMAX4431.2. Through the contrast research of tracking performance and disturbance performance of PID control and SMC control, the result showed that PID control had better steady-state characteristics but long dynamically adjust time and poor disturbance performance; which SMC control had less dynamically adjust time and better robustness but system steady-state characteristics was impacted by chattering.3. Through the research of the equivalent control, exponential reaching law, variable exponent reaching law and the combination control of discrete sliding mode variable structure performance, results show that the equivalent control dynamically adjust quickly, the system response is oscillatory, and the presence of buffeting; discrete exponential reaching law is quasi sliding mode, system chattering, cannot be asymptotically stable zero; variable exponent reaching rate adjustment switch with shape, the state of the system is asymptotically stable at zero, but the initial stage of the switchover in large amplitude jitter; combined control integrated two advantages, realization of reaching sliding cutting surface finite step, and asymptotic stability of zero, chattering.4. The switching gain and index reaching study on fuzzy sliding mode variable structure control performance, the results show that the fuzzy adaptive sliding mode reaching law control system better tracking performance, control more stable output, buffeting smaller system.5. Based on the experimental study of PID,SMC,FSMC control, dynamic characteristics and steady-state characteristics of the anti interference characteristics, the results show that the fuzzy SMC control is better than that of the steady state characteristics and anti-interference characteristics before two, but also suppress chattering.Therefore, the use of system model parameter identification method can obtain more accurate mathematical model of variable structure control system to adjust quickly and without overshoot, but there are chattering, discrete combination control and fuzzy index reaching rate control can weaken chattering, but fuzzy control index reaching rate due to the combination control in tracking performance, and restrain the chattering effect is better.更多还原