一种基于BP神经网络算法PID控制器的研究与仿真

文中将BP神经网络的原理应用于参数辨识过程,结合传统的PID控制算法,形成一种改进型BP神经网络PID控制算法。该算法利用BP神经网络建立系统参数模型,能够跟踪被控对象的变化,取得较高的辨识精度。针对BP神经网络对权系初始值敏感的缺点,优化BP神经网络的初始权系数。通过BP算法修正BP网络自身权系数,实现PID参数的在线调整。仿真结果显示了该算法收敛速度快、精度高、鲁棒性强、稳定性好,表明了该算法的可行性与有效性。     

基于动态神经网络在线辨识的PID控制

针对工业控制领域中复杂非线性时变系统和传统RBF神经网络辨识PID控制的不足,提出了一种基于聚类结合算法的动态RBF神经网络在线辨识PID自适应控制方法.通过优化的动态RBF辨识神经网络更好地描述了控制对象的动态行为,获得PID参数在线调整信息,实现系统的智能控制.仿真结果表明,与常规RBF神经网络辨识的PID控制方法相比该方法具有较高的控制精度,较快的系统响应,较强的适应性和鲁棒性.     

不确定蔡氏电路混沌系统的神经网络优化控制

该文研究了不确定非线性蔡氏电路混沌系统的动态神经网络在线辨识和跟踪控制问题.利用无源技术得出梯度下降算法调整神经网络辨识器权值的稳定性定理,然后在辨识模型基础上设计局部优化控制器,将蔡氏混沌系统镇定到期望目标轨迹,并保证跟踪误差有界.数值仿真结果表明了所提出方法的有效性.     

输出不可量测非线性系统的神经模型参考自适应控制

该文针对被控对象输出不可量测的非线性系统,引入一个便于在线辨识的扩展神经网络模型,提出一种基于前馈-反馈结构的神经网络模型参考自适应控制方法。给出了具有全局收敛性的网络训练算法,并分析了控制系统的稳定性。仿真结果表明该控制方法是有效的,而且对网络初始权值的选取及被控对象特性参数的扰动都具有良好的鲁棒性。     

在线学习多重检测的可视对象跟踪方法

为了研究无约束环境下长时间可视跟踪问题,提出了一种在线学习多重检测的对象跟踪方法.该方法以随机蕨作为基础检测器结构,通过在线学习的方式,将目标对象的整体和局部表观,以及由场景学习中发掘的同步对象同时作为检测学习的基础数据,该检测器因而具备了对这多种对象的独立检测能力.由于其各个检测部分发挥了各自不同的作用,本文从测量的角度将检测器对这三种对象检测的结果进行融合,通过计算检测关于目标的配置概率进而确定目标位置,实现对象跟踪任务.基于真实视频序列的实验结果验证了本文方法的有效性和稳定性,以及较现有的跟踪方法在跟踪性能上的提高.     

基于模糊神经网络的传感器可信度实时获取

针对传感器在复杂环境中所测信息不完全准确的问题,提出了一种基于专家规则的零阶Sugeno模糊模型神经网络来获取传感器可信度的方法.神经网络经训练样本训练后,可以根据传感器状态和环境信息实时地得到传感器可信度.将该模型学习算法中的最小二乘识别器加以改进,并引入了遗忘因子,可以使该网络实现在线学习,不断更新网络参数.仿真结果表明该模糊神经网络可以有效地获得传感器可信度,且越小则网络在线学习能力越强.     

基于全数字模型辨识的高精度跟踪系统设计

目前工程上对光电跟踪系统的控制器设计多数采用基于被控对象数学模型的控制策略。模型的辨识是否准确,如何开展控制器设计成为系统的难点。利用DSP控制器对光电跟踪系统进行全数字方式的传递函数测试,在线处理测试数据,同时DSP控制器通过接收系统期望的特性参数,自动计算出校正参数,完成控制器框架设计。根据跟踪目标特性,结合跟踪误差特点及工程经验库,提出准速度/加速度补偿算法,优化控制器,该方法已成功应用于某光电跟踪系统。基于模型辨识设计的高精度控制器设计方法简单、实用、无附加成本,提高了系统跟踪精度,具有较好的使用效果。     

星地光通信中的精跟踪模拟实验研究

设计了一套简便易行的星地光通信中精跟踪演示系统方案,并详细介绍了精跟踪演示系统的组成.跟踪算法采用改进神经网络的PID算法,在线调整网络加权值.最后,针对不同频率的信标抖动进行了跟踪补偿模拟实验,实验结果表明,该精跟踪系统实现了高频信标光抖动跟踪补偿,具有很好的实用性.     

基于增加动量项的神经网络左逆张力辨识策略

以两电机张力控制系统为研究对象,为实现两电机张力控制系统的高性能控制及无传感器运行,张力的准确检测是其中的关键。文中提出了基于人工神经网络和左逆系统理论的两电机张力系统的一个新的识别方法。考虑到系统的参数是时变的和张力易受负载变化影响。神经网络左逆辨识被用在该系统中中,这是很容易实现的左逆模型。针对运用中传统BP网络的收敛速度慢,易陷入极小值的缺点,提出了增加动量项的改进神经网络左逆辨识策略,通过仿真模型对在负载扰动下两电机的张力进行辨识。仿真结果表明,辨识准确,策略可行。     

船用雷达目标跟踪的新方法

本文提出一种由广义小波神经网络实现船用雷达跟踪中航迹外推的自适应新方法.由S(sigmoid)函数构造的尺度函数和小波作为网络中神经元的激励函数,隐层节点数由小波分解次数和处理信号维数决定,输出层采用局部连接方式以解决多维信号的不利影响.理论证明,广义小波神经网络的鲁棒性在一定条件下优于BP网络.仿真表明,该方法的在线处理运算量不随所跟踪的运动目标模型的复杂性而增加,并且对变加速和急转弯运动目标具有较高的跟踪精度.     

Wavelet neural network control for induction motor drive using sliding-mode design technique

This paper addresses an adaptive observation system and a wavelet-neural-network (WNN) control system for achieving the favorable decoupling control and high-precision position tracking performance of an induction motor (IM) drive. First, an adaptive observation system with an inverse rotor time-constant observer is derived on the basis of model reference adaptive system theory to preserve the decoupling control characteristic of an indirect field-oriented IM drive. The adaptive observation system is implemented using a digital signal processor with a high sampling rate to make it possible to achieve good dynamics. Moreover, a WNN control system is developed via the principle of sliding-mode control to increase the robustness of the indirect field-oriented IM drive with the adaptive observation system for high-performance applications. In the WNN control system, a WNN is utilized to predict the uncertain system dynamics online to relax the requirement of uncertainty bound in the design of a traditional sliding-mode controller. In addition, the effectiveness of the proposed observation and control systems is verified by simulated and experimental results.     

A fuzzy neural network controller for parallel-resonant ultrasonicmotor drive

A newly designed driving circuit for the traveling-wave-type ultrasonic motor (USM), which consists of a push-pull DC-DC power converter and a current-source two-phase parallel-resonant inverter, is presented in this study. Moreover, since the dynamic characteristics of the USM are difficult to obtain and the motor parameters are time varying, a fuzzy neural network (NN) controller is proposed to control the USM drive system. In the proposed controller, a fuzzy model-following controller is implemented to control the rotor position of the USM, and an online trained NN with variable learning rates is implemented to tune the output scaling factor of the fuzzy controller. To guarantee the convergence of tracking error, analytical methods based on a discrete-type Lyapunov function are proposed to determine the desired variable learning rates. From the experimental results, accurate tracking response can be obtained by the proposed controller, and the influences of parameter variations and external disturbances on the USM drive also can be reduced effectively     

Implementation of LLCC-resonant driving circuit and adaptive CMAC neural network control for linear piezoelectric ceramic motor

In this paper, an adaptive cerebellar-model articulation computer (CMAC) neural network (NN) control system is developed for a linear piezoelectric ceramic motor (LPCM) that is driven by an LLCC-resonant inverter. The motor structure and LLCC-resonant driving circuit of an LPCM are introduced initially. The LLCC-resonant driving circuit is designed to operate at an optimal switching frequency such that the output voltage will not be influenced by the variation of quality factor. Since the dynamic characteristics and motor parameters of the LPCM are highly nonlinear and time varying, an adaptive CMAC NN control system is designed without mathematical dynamic model to control the position of the moving table of the LPCM drive system to achieve high-precision position control with robustness. In the proposed control scheme, the dynamic backpropagation algorithm is adopted to train the CMAC NN online. Moreover, to guarantee the convergence of output tracking error for periodic commands tracking, analytical methods based on a discrete-type Lyapunov function are utilized to determine the optimal learning-rate parameters of the CMAC NN. The effectiveness of the proposed driving circuit and control system is verified by experimental results in the presence of uncertainties, and the advantages of the proposed control system are indicated in comparison with a traditional integral-proportional position control system. Accurate tracking response and superior dynamic performance can be obtained due to the powerful online learning capability of the CMAC NN with optimal learning-rate parameters.     

基于反推的永磁同步电动机伺服系统的位置跟踪控制

永磁同步电动机工作性能优越，在当前交流伺服系统的驱动控制当中起着越来越重要的作用。为了实现永磁同步电动机的精确位置跟踪，把一种新颖非线性控制方法Backstepping应用于永磁同步电动机伺服系统控制器的设计。Backstepping控制器的设计以保证系统的全局一致渐近稳定为原则，因此该控制器不但可以保证系统的全局一致渐近稳定，而且系统具有快速跟踪，定位精确的特点。系统的设计能够有效降低转矩变化对位置跟踪性能的影响。最后通过Matlab仿真验证了系统设计的有效性和可行性。     

Fuzzy neural networks for identification and control of ultrasonicmotor drive with LLCC resonant technique

This paper demonstrates the applications of fuzzy neural networks (FNNs) in the identification and control of the ultrasonic motor (USM). First, the USM is derived by a newly designed high-frequency two-phase voltage-source inverter using LLCC resonant technique. Then, two FNNs with varied learning rates are proposed to control the rotor position of the USM. The USM drive system is identified by a fuzzy neural network identifier (FNNI) to provide the sensitivity information of the drive system to a fuzzy neural network controller (FNNC). A backpropagation algorithm is used to train both the FNNI and FNNC on-line. Moreover, to guarantee the convergence of identification and tracking errors, analytical methods based on a discrete-type Lyapunov function are proposed to determine the varied learning rates of the FNNs. In addition, the effectiveness of the FNN-controlled USM drive system is demonstrated by experimental results. Accurate tracking response can be obtained due to the powerful on-line learning capability of the FNNs. Furthermore, the influence of parameter variations and external disturbances on the USM drive system can be reduced effectively     

Current sensorless position–flux tracking controller for induction motor drives

An innovative indirect field-oriented output feedback controller for induction motor drives is presented. This solution is based on output feedback since only speed and position of the motor shaft are measured, while current sensors are avoided. This approach is suitable for low cost applications, where the position sensor cannot be removed to guarantee accurate position tracking.The proposed method provides global asymptotic tracking of smooth position and flux references in presence of unknown constant load torque. It is based on the natural passivity of the electromagnetic part of the machine and it guarantees asymptotic decoupling of the induction motor mechanical and electrical subsystems achieving at the same time asymptotic field orientation. Lyapunov analysis and nonlinear control design have been adopted to obtain good position tracking performances and effective torque–flux decoupling. The cascaded structure of the controller allows performing a constructive tuning procedure for speed and position control loops.Results of experimental tests are presented to demonstrate the tracking and robustness features of the proposed solution.     

Adaptive positioning control for a LPMSM drive based on adapted inverse model and robust disturbance observer

The adaptive robust positioning control for a linear permanent magnet synchronous motor drive based on adapted inverse model and robust disturbance observer is studied in this paper. First, a model following two-degrees-of-freedom controller consisting of a command feedforward controller (FFC) and a feedback controller (FBC) is developed. According to the estimated motor drive dynamic model and the given position tracking response, the inner speed controller is first designed. Then, the transfer function of FFC is found based on the inverse model of inner speed closed-loop and the chosen reference model. The practically unrealizable problem possessed by traditional feedforward control is avoided by the proposed FFC. As to the FBC, it is quantitatively designed using reduced plant model to meet the specified load force regulation control specifications. In dealing with the robust control, a disturbance observer based robust control scheme and a parameter identifier are developed. The key parameters in the robust control scheme are designed considering the effect of system dead-time. The identification mechanism is devised to obtain the parameter uncertainties from the observed disturbance signal. Then by online adapting the parameters set in the FFC according to the identified parameters, the nonideal disturbance observer based robust control can be corrected to yield very close model following position tracking control. Meanwhile, the regulation control performance is also further improved by the robust control. In the proposed identification scheme, the effect of a nonideal differentiator in the accuracy of identification results is taken into account, and the compromise between performance, stability, and control effort limit is also considered in the whole proposed control scheme.     

Robust control using neural network uncertainty observer for linearinduction motor servo drive

A robust controller, that combines the merits of integral-proportional (IP) position control and neural network (NN) observed technique, is designed for a linear induction motor (LIM) servo drive in this study. First, the secondary flux of the LIM is estimated using a sliding-mode flux observer on the stationary reference frame and the feedback linearization theory is used to decouple the thrust and the flux amplitude of the LIM. Then, the IP position controller is designed according to the estimated mover parameters to match the time-domain command tracking specifications. Moreover, a robust controller is formulated using the NN uncertainty observer, which is implemented to estimate the lumped uncertainty of the controlled plant, as an inner-loop force controller to increase the robustness of the LIM servo drive system. Furthermore, in the derivation of the online training algorithm of the NN, an error function is used in the Lyapunov function to avoid the real-time identification of the system Jacobian. In addition, to increase the speed and accuracy of the estimated flux, the sliding-mode flux observer is implemented using a 32 bit floating-point digital signal processor (DSP) with a high sampling rate. The effectiveness of the proposed control scheme is verified by both the simulated and experimental results     

基于SCIT算法的天气雷达回波风暴识别跟踪方法

基于雷达数据的风暴体识别、追踪及预警方法是最早出现的临近预报技术,其中对风暴的准确识别是进行风暴体追踪和预警的前提。本文借鉴SCIT(Storm Cell Identification and Tracking)算法对强风暴进行识别,根据"宁短勿长,特征相似"的原则匹配两时刻的风暴单体。通过风暴在过去两时刻的质心位置进行线性外推从而预报下一时刻风暴的位置。结果显示可以较好地识别强风暴并实现对识别出的风暴的大致跟踪。     

基于多分辨分析和WNN的模拟电路故障诊断

提出了一种基于多分辨分析和小波神经网络(WNN)相结合的模拟电路故障诊断方法。该方法利用了多分辨分析优异的时频特性,提取采集数据中的故障特征参数值,结合小波神经网络强大的非线性分类、学习、泛化能力及精度高、收敛速度快等特性,将得到的输入数据进行归一化处理作为小波神经网络的输入对其进行训练,并将训练的结果应用于滤波器电路故障诊断。结果表明,该方法实现了对故障模块的定位,是一种有效的模拟电路故障诊断方法。