永磁同步电机神经网络自适应滑模控制器设计

设计了神经网络自适应滑模控制器.用RBF神经网络自动调整滑模控制器的切换项增益,无需建立包含参数摄动和干扰在内的整个系统的精确数学模型,有效提高了系统的稳定性和鲁棒性.采用Lyapunov稳定性理论证明了系统稳定性,并针对常值干扰、时变干扰和参数摄动情况分别进行了仿真与实验.与传统的PI控制相比,神经网络自适应滑模控制器具有更好的稳定性和抗干扰能力.     

Indirect model reference adaptive control with dynamic adjustment of parameters

The paper discusses in detail a new method for indirect model reference adaptive control (MRAC) of linear time-invariant continuous-time plants with unknown parameters. The method involves not only dynamic adjustment of plant parameter estimates but also those of the controller parameters. Hence the overall system can be described by a set of non-linear differential equations as in the case of direct control. Many of the difficulties encountered in the conventional indirect approach, where an algebraic equation is solved to determine the control parameters, are consequently bypassed in this method. The proof of stability of the equilibrium state of the overall system is found to be different from that used in direct control. Using Lyapunov's theory, it is first shown that the parameter errors between the parameter estimates of the identifier and the true parameters of the plant, as well as those between the actual parameters of the controller and their desired values, are bounded. Following this, using growth rates of signals in the adaptive loop as well as order arguments, it is shown that the error equations are globally uniformly stable and that the tracking (control) error tends to zero asymptotically. This in turn establishes the fact that both direct and indirect model reference adaptive schemes require the same amount of prior information to achieve stable adaptive control.     

感应电机的神经网络鲁棒自适应动态面控制

针对感应电机的高阶、强耦合的非线性,设计一种具有自适应性和鲁棒性的动态面控制律,以实现对感应电机的控制。首先通过变量代换,将五阶感应电机模型变为块三角形式;接着采用神经网络逼近,将模型转化为含有结构不确定性和参数不确定性的参数化模型。对于感应电机系统中的参数不确定性,在动态面控制方法的基础上设计参数自适应律,在线估计其未知参数;对于感应电机系统中的结构不确定性,设计鲁棒控制律,使得电机控制系统具有鲁棒性。最后通过仿真实现,以验证所提出的控制方法的有效性与实用性。     

模型不确定的下肢康复机器人轨迹跟踪自适应控制

为了实现康复训练过程中高精度的轨迹跟踪控制,针对下肢康复机器人的模型参数和外界干扰等不确定性因素对其轨迹跟踪造成严重影响,提出一种模型不确定的下肢康复机器人轨迹跟踪自适应控制方法。根据所提方案,设计了相应的轨迹跟踪自适应控制器;并进行了轨迹跟踪控制仿真实验对比分析,结果表明,计算力矩控制方法在系统模型不确定时,膝关节的最大角度跟踪误差高达11.3°,髋关节最大稳态误差4.6°;而轨迹跟踪自适应控制方法在模型不确定的情况下,髋关节和膝关节的角度跟踪稳态误差均收敛于零;轨迹跟踪自适应控制方法可以显著提高下肢康复机器人轨迹跟踪的精度。     

Computer simulation of electric motor drive systems including thepower electronic network

A method for studying a combined power electronic network/variable-speed AC electric motor drive system based on a simple numerical algorithm is presented. It can be applied to any machine that can be describing using a d-q-type model. An LCI-fed synchronous motor drive in the steady state is used as an illustrative example. A simple iterative technique is used to find the initial conditions in the time domain, making use of an infinite link inductance approximation to provide the first iteration. The predicted waveforms of voltage and current for a sample operating point are compared with their experimental equivalents (measured on a 3 kVA prototype drive system), and a good level of agreement is reached. The technique is sufficiently simple to have been implemented on a personal computer     

A class of adaptive regulators for robot manipulators

In this paper we propose a framework to adaptive regulator design of robot manipulators. A class of global regulators for robot is characterized for which we provide guidelines to derive adaptive versions. The class of regulators is described by control laws comprising the gradient of an artificial potential energy depending in a linear manner on the robot and payload unknown parameters plus a linear velocity feedback. We provide explicit sufficient conditions on the artificial potential energy which allow to obtain an adaptive regulator which yields a stable closed-loop system and global positioning. Using this framework we revise two previously proposed adaptive regulators and we suggest two new ones. © 1998 John Wiley & Sons, Ltd.     

同步电动机三自由度内模动态解耦控制

为了解决矢量控制只能实现电机转矩与磁链的静态解耦,不能实现两者的动态解耦的问题,探讨了多自由度内模解耦方案。由于一自由度内模解耦,当电机参数非同步变化时,电机定子电压方程仍存在耦合现象;而二自由度内模解耦,耦合程度虽有所减少,但动态解耦仍不完全;提出了三自由度内模解耦控制方案。该方案中通过前馈控制器、内模控制器、反馈滤波器的设计,从而实现动态全解耦。通过同步电动机内模解耦控制的仿真与实验结果证明了三自由度内模解耦控制方案能够实现动态完全解耦。     

MRAS感应电机定子电阻的在线辨识

针对感应电机定子电阻值受外界因素干扰而影响其矢量控制系统稳定性和控制精确度问题,提出了基于人工神经网络的定子电阻在线辨识方法.为了辨识定子电阻,将人工神经网络模型的定子电流估算值与实际测量电流值的误差反馈以调整神经网络的权值.借助MATLAB/SIMULINK搭建仿真系统,验证了定子电阻在线辨识的必要性.结果表明,该方法可以有效地对定子电阻进行在线辨识.     

An interim dynamic induction motor model for stability studies in the WSCC

An "interim" composite load model containing a static part and a dynamic part has been implemented in the Western Systems Coordinating Council (WSCC). The static part of the load model is about 80% of the total load and comprises of existing static load data from the WSCC members. The dynamic part is a default induction motor model for approximately 20% of the total load. This composite model is used for all loads in the full-scale WSCC grid models that are the basis for system interconnection studies. The model is designed primarily to capture the effects of dynamic induction motor loads for highly stressed north to south flow conditions during summer peaks in the WSCC.     

基于模型参考自适应滑模控制的直线电机速度环研究

为了获得直线电机较高的跟踪和鲁棒性能,在库仑摩擦模型补偿控制基础上,同时考虑动子质量、粘滞摩擦系数的时变性,提出了基于模型参考自适应滑模控制的直线电机速度环控制方案,将模型补偿、自适应补偿尚未能完全补偿的部分当作外界干扰,由滑模控制加以克服。仿真结果表明,该方案可以获得较高的速度跟踪性能和鲁棒性能,并且由于自适应控制的应用,减小了滑模控制引起的抖振。     

不确定机器人的间接模糊自适应轨迹跟踪控制

针对不确定的机器人系统轨迹跟踪控制问题,依据自适应模糊控制算法和H∞控制理论,提出了一种间接模糊自适应控制方法。该方法将自适应模糊控制和H∞控制理论相结合,用稹糊逻辑系统直接为机器人建模,不需对机器人的数学模型进行转换,并结合H∞控制理论,将建模误差和外部干扰衰减到预先规定的指标。基于Lyapunov方法,给出了学习自适应率,H∞跟踪特性的证明。对二自由度机器人的仿真结果表明了所提出的控制算法有效性。     

Design and evaluation of the LOPES exoskeleton robot for interactive gait rehabilitation.

This paper introduces a newly developed gait rehabilitation device. The device, called LOPES, combines a freely translatable and 2-D-actuated pelvis segment with a leg exoskeleton containing three actuated rotational joints: two at the hip and one at the knee. The joints are impedance controlled to allow bidirectional mechanical interaction between the robot and the training subject. Evaluation measurements show that the device allows both a "patient-in-charge" and "robot-in-charge" mode, in which the robot is controlled either to follow or to guide a patient, respectively. Electromyography (EMG) measurements (one subject) on eight important leg muscles, show that free walking in the device strongly resembles free treadmill walking; an indication that the device can offer task-specific gait training. The possibilities and limitations to using the device as gait measurement tool are also shown at the moment position measurements are not accurate enough for inverse-dynamical gait analysis.     

基于混合逻辑动态模型的电机驱动系统仿真

电机驱动系统是一个由离散的控制信号驱动连续状态变量随时间演化的典型的混杂系统。针对传统的基于数学模型的建模方法都是研究逆变器中的开关元件开路等结构性变化,不能用于描述系统中的参数变化,本文建立逆变器和电机的混合逻辑动态模型,并运用Matlab/Simulink对系统中的R、L、Vdc等参数变化进行仿真分析,仿真结果表明该模型能很好地描述电机驱动系统参数的变化。     

A robust adaptive control with unmodeled dynamic for HVDC transmission systems

Utilizing the feature of quick response of HVDC to improve the performance of AC/DC system has become the emphasis to be researched. This paper introduces firstly the principle of the robust adaptive control of nonlinear systems with unmodeled dynamics, then developed the robust adaptive additional control of HVDC with unmodeled dynamics of generator in order to improve stability of power system. The additional control of HVDC with unmodeled dynamics only uses the local signals and its design is simple, furthermore it can obviously improve the stability of power system in different operational conditions. Experimental results using the presented concepts obtained on single machine infinite bus model are also included. These results prove the efficiency of the control scheme. The design process of controller provided a new idea to design controller by use of simplified model. __________ Translated from Journal of Sichuan University (Engineering Science Edition), 2005, 37(5): 154–158 (in Chinese)     

基于Hopfield神经网络的直流传动系统模型参考自适应控制

本文将Hopfield神经网络引入直流传动系统的模型参考自适应控制,通过神经网络模型参考自适应控制器来给出常规双闭环调速系统的速度控制器外环动态输出,使速度控制具有对某些参数变化的一定程度的鲁棒性     

参数在线跟踪的直流传动双神经网络模型参考自适应控制

针对在“基于Hopfield神经网络的直流传动系统模型参考自适应控制“^[8]一文中指出的所提神经网络控制器在负载转矩未知和变化情况下的缺陷,将作者先前所提出的交流传动系统参数自动跟踪神经网络^[2]引入直流传动系统,构成具有参数在线跟踪功能的直流传动双神经网络模型参考自适应控制模式,进一步提高了系统的控制性能。     

机器人的神经网络鲁棒轨迹跟踪控制

考虑了一类具有外界扰动和参数不确定性机器人系统的轨迹跟踪鲁棒控制问题。提出了两种控制方法：第一种应用输入输出线性化方法以及Lyapunov函数法,推导出鲁棒输出跟踪控制器。所获得的控制器可确保系统输出按指数规律跟踪期望输出,同时相应闭环系统的状态一致最终有界。第二种方法在第一种控制方法的基础上,利用一个RBF神经网络自适应学习系统不确定性的未知上界,有效的克服了系统不确定性的影响,提高了控制精度。     

分布式发电配电网故障区间定位的自适应矩阵算法

准确定位分布式发电配电网故障区段是有效利用清洁能源的前提.基于馈线终端单元(FTU)的配电网故障定位矩阵算法,提出一种适合分布式发电配电网故障定位自适应算法.根据流过FTU的过电流及其方向,首先判断分布式电源是否投入运行和母线及分布式电源是否发生故障,初步确定配电网的故障区域;再根据故障区域结构和FTU过电流信息自适应构成故障矩阵,定位故障所在线路;为排除畸变故障信息的干扰,比较故障线路两端检测到的故障电流差值,提出了故障电流差值比较判据.算例验证了算法的准确性、高效性和良好的容错机制.     

Data mining of motor patterns recorded with wearable technology

Discusses utilizing self-organizing maps to manage large data sets gathered in the field. Dramatic advances in wearable technology suggest that wearable sensors and systems are close to deployment in the clinical field environment. Miniature, wireless sensors are becoming readily available due to the tremendous research efforts focused on this area as well as the great interest demonstrated by the private sector for the associated potential market of these devices. Wearable systems, including base stations and hand-held units, are becoming less cumbersome, with an ever-increasing capacity to receive and store vast amounts of data for extensive periods of time. Further, the growing availability of wireless LANs makes it possible to envision wearable systems from which data can be continuously uploaded, while hand-held units provide a means to communicate with the patient when needed.     

带神经网络转矩观测器的PMSM自适应前馈 PID控制器设计

推导了永磁同步电机(PMSM)的ARMA模型，由递归神经网络构成综合转矩观测器，从而把综合负载转矩视为可测干扰；结合极点配置自校正控制、PID控制和前馈控制思想，设计了PMSM的自适应前馈PID控制器。有效地解决了PMSM系统中参数变化和负载扰动等不确定性问题。仿真实验结果表明，该方法具有较强的鲁棒性。