Intelligent control system design for UAV using a recurrent wavelet neural network

This paper aims to propose an efficient control algorithm for the unmanned aerial vehicle (UAV) motion control. An intelligent control system is proposed by using a recurrent wavelet neural network (RWNN). The developed RWNN is used to mimic an ideal controller. Moreover, based on sliding-mode approach, the adaptive tuning laws of RWNN can be derived. Then, the developed RWNN control system is applied to an UAV motion control for achieving desired trajectory tracking. From the simulation results, the control scheme has been shown to achieve favorable control performance for the UAV motion control even it is subjected to control effort deterioration and crosswind disturbance.     

Distributed cooperative learning algorithms using wavelet neural network

This paper investigates the distributed cooperative learning (DCL) problems over networks, where each node only has access to its own data generated by the unknown pattern (map or function) uniformly, and all nodes cooperatively learn the pattern by exchanging local information with their neighboring nodes. These problems cannot be solved by using traditional centralized algorithms. To solve these problems, two novel DCL algorithms using wavelet neural networks are proposed, including continuous-time DCL (CT-DCL) algorithm and discrete-time DCL (DT-DCL) algorithm. Combining the characteristics of neural networks with the properties of the wavelet approximation, the wavelet series are used to approximate the unknown pattern. The DCL algorithms are used to train the optimal weight coefficient matrix of wavelet series. Moreover, the convergence of the proposed algorithms is guaranteed by using the Lyapunov method. Compared with existing distributed optimization strategies such as distributed average consensus (DAC) and alternating direction method of multipliers (ADMM), our DT-DCL algorithm requires less information communications and training time than ADMM strategy. In addition, it achieves higher accuracy than DAC strategy when the network consists of large amounts of nodes. Moreover, the proposed CT-DCL algorithm using a proper step size is more accurate than the DT-DCL algorithm if the training time is not considered. Several illustrative examples are presented to show the efficiencies and advantages of the proposed algorithms.

     

Intelligent control of SVC using wavelet neural network to enhance transient stability

In order to enhance transient stability in a power system, a new intelligent controller is proposed to control a Static VAR compensator (SVC) located at center of the transmission line. This controller is an online trained wavelet neural network controller (OTWNNC) with adaptive learning rates derived by the Lyapunov stability. During the online control process, the identification of system is not necessary, because of learning ability of the proposed controller. One of the proposed controller features is robustness to different operating conditions and disturbances. The test power system is a two-area two-machine system power. The simulation results show that the oscillations are satisfactorily damped out by the OTWNNC.     

Robust control for nonlinear motor-mechanism coupling system using wavelet neural network

A robust controlled toggle mechanism, which is driven by a permanent magnet (PM) synchronous servo motor is studied in this paper. First, based on the principle of computed torque control, a position controller is developed for the motor-mechanism coupling system. Moreover, to relax the requirement of the lumped uncertainty in the design of a computed torque controller, a wavelet neural network (WNN) uncertainty observer is utilized to adapt the lumped uncertainty online. Furthermore, based on the Lyapunov stability a robust control system, which combines the computed torque controller, the WNN uncertainty observer and a compensated controller is proposed to control the position of the motor-mechanism coupling system. The computed torque controller with WNN uncertainty observer is the main tracking controller, and the compensated controller is designed to compensate the minimum approximation error of the uncertainty observer. Finally, simulated and experimental results due to a periodic sinusoidal command show that the dynamic behaviors of the proposed robust control system are robust with regard to parametric variations and external disturbances.     

Recurrent fuzzy network design using hybrid evolutionary learning algorithms

This paper proposes a recurrent fuzzy network design using the hybridization of a multigroup genetic algorithm and particle swarm optimization (R-MGAPSO). The recurrent fuzzy network designed here is the Takagi–Sugeno–Kang (TSK)-type recurrent fuzzy network (TRFN), in which each fuzzy rule comprises spatial and temporal sub-rules. Both the number of fuzzy rules and the parameters in a TRFN are designed simultaneously by R-MGAPSO. In R-MGAPSO, the techniques of variable-length individuals and the local version of particle swarm optimization are incorporated into a genetic algorithm, where individuals with the same length constitute the same group, and there are multigroups in a population. Population evolution consists of three major operations: elite enhancement by particle swarm optimization, sub-rule alignment-based crossover, and mutation. To verify the performance of R-MGAPSO, dynamic plant and a continuous-stirred tank reactor controls are simulated. R-MGAPSO performance is also compared with genetic algorithms in these simulations.     

ERP实施绩效的小波网络智能诊断

企业为了达到ERP 项目既定的目标,需要不断调整或重新策划ERP项目,必须定期对ERP的实施绩效进行诊断。提出了一种基于小波网络的ERP绩效智能诊断方法,在最小均方能量准则下,采用相应的共扼梯度学习算法求解子波函数线性组合的尺度和时延参数,以及神经网络的权值,通过实例应用给出了基于小波网络ERP绩效智能诊断的算法。     

Direct adaptive power system stabilizer design using fuzzy wavelet neural network with self-recurrent consequent part

This paper aims to propose a stable fuzzy wavelet neural-based adaptive power system stabilizer (SFWNAPSS) for stabilizing the inter-area oscillations in multi-machine power systems. In the proposed approach, a self-recurrent Wavelet Neural Network (SRWNN) is applied with the aim of constructing a self-recurrent consequent part for each fuzzy rule of a Takagi-Sugeno-Kang (TSK) fuzzy model. All parameters of the consequent parts are updated online based on Direct Adaptive Control Theory (DACT) and employing a back-propagation-based approach. The stabilizer initialization is performed using an approach based on genetic algorithm (GA). A Lyapunov-based adaptive learning rates (LALRs) algorithm is also proposed in order to speed up the stabilization rate, as well as to guarantee the convergence of the proposed stabilizer. Therefore, due to having a stable powerful adaptation law, there is no requirement to use any identification process. Kundur's four-machine two-area benchmark power system and six-machine three-area power system are used with the aim of assessing the effectiveness of the proposed stabilizer. The results are promising and show that the inter-area oscillations are successfully damped by the SFWNAPSS. Furthermore, the superiority of the proposed stabilizer is demonstrated over the IEEE standard multi-band power system stabilizer (MB-PSS), and the conventional PSS.     

Piezoelectric-actuated drop-on-demand droplet generator control using adaptive wavelet neural network controller

This paper presents the design, fabrication and control of a piezoelectric-type droplet generator which is applicable for on-line dispensing. Adaptive wavelet neural network (AWNN) control is applied to overcome nonlinear hysteresis inherited in the LPM. The adaptive learning rates are derived based on the Lyapunov stability theorem so that the stability of the closed-loop system can be assured. Unlike open-loop dispensing system, the system proposed can potentially generate droplets with high accuracy. Experimental verifications focusing on regulating control are performed firstly to assure the reliability of the proposed control schemes. Real dispensing is then conducted to validate the feasibility of the piezoelectric-actuated drop-on-demand droplet generator. In order to illustrate the effectiveness of the proposed method, experimental results obtained using the AWNN scheme are compared with their counterparts using traditional PID control. The results indicate that the proposed AWNN scheme not only outperforms PID control but also works well in developing the piezoelectric-actuated drop-on-demand dispensing system. The proposed dispensing system provides droplet chains with an averaged mass as small as 31.5 mg while the associated standard deviation is as low as 0.72%.     

基于神经网络的多层控制系统智能故障诊断

阐述了智能故障诊断技术的特点，针对复杂多层次控制系统，提出了一种分层模块化设计方法及模块间故障变量传播的搜索模式；探讨了基于神经网络与其它理论方法相结合的几种智能化诊断模式，对应用的方法，特点作了进一步分析。     

New nonleast-squares neural network learning algorithms forhypothesis testing

Hypothesis testing is a collective name for problems such as classification, detection, and pattern recognition. In this paper we propose two new classes of supervised learning algorithms for feedforward, binary-output neural network structures whose objective is hypothesis testing. All the algorithms are applications of stochastic approximation and are guaranteed to provide optimization with probability one. The first class of algorithms follows the Neyman-Pearson approach and maximizes the probability of detection, subject to a given false alarm constraint. These algorithms produce layer-by-layer optimal Neyman-Pearson designs. The second class of algorithms minimizes the probability of error and leads to layer-by-layer Bayes optimal designs. Deviating from the layer-by-layer optimization assumption, we propose more powerful learning techniques which unify, in some sense, the already existing algorithms. The proposed algorithms were implemented and tested on a simulated hypothesis testing problem. Backpropagation and perceptron learning were also included in the comparisons.     

Intelligent automatic landing system using time delay neural network controller

This paper presents an intelligent automatic landing system that uses a time delay neural network controller and a linearized inverse aircraft model to improve the performance of conventional automatic landing systems. The automatic landing system of an airplane is enabled only under limited conditions. If severe wind disturbances are encountered, the pilot must handle the aircraft due to the limits of the automatic landing system. In this study, a learning-through-time process is used in the controller training. Simulation results show that the neural network controller can act as an experienced pilot and guide the aircraft to a safe landing in severe wind disturbance environments without using the gain scheduling technique.     

An adaptive learning algorithm for a wavelet neural network

Abstract: An optimal online learning algorithm of a wavelet neural network is proposed. The algorithm provides not only the tuning of synaptic weights in real time, but also the tuning of dilation and translation factors of daughter wavelets. The algorithm has both tracking and smoothing properties, so the wavelet networks trained with this algorithm can be efficiently used for prediction, filtering, compression and classification of various non-stationary noisy signals.     

Adaptive neural complementary sliding-mode control via functional-linked wavelet neural network

Chaos control can be applied in the vast areas of physics and engineering systems, but the parameters of chaotic system are inevitably perturbed by external inartificial factors and cannot be exactly known. This paper proposes an adaptive neural complementary sliding-mode control (ANCSC) system, which is composed of a neural controller and a robust compensator, for a chaotic system. The neural controller uses a functional-linked wavelet neural network (FWNN) to approximate an ideal complementary sliding-mode controller. Since the output weights of FWNN are equipped with a functional-linked type form, the FWNN offers good learning accuracy. The robust compensator is designed to eliminate the effect of the approximation error introduced by the neural controller upon the system stability in the Lyapunov sense. Without requiring preliminary offline learning, the parameter learning algorithm can online tune the controller parameters of the proposed ANCSC system to ensure system stable. Finally, it shows by the simulation results that favorable control performance can be achieved for a chaotic system by the proposed ANCSC scheme.     

Integral variable structure control of nonlinear system using a CMAC neural network learning approach.

This work presents a novel integral variable structure control (IVSC) that combines a cerebellar model articulation controller (CMAC) neural network and a soft supervisor controller for use in designing single-input single-output (SISO) nonlinear system. Based on the Lyapunov theorem, the soft supervisor controller is designed to guarantee the global stability of the system. The CMAC neural network is used to perform the equivalent control on IVSC, using a real-time learning algorithm. The proposed IVSC control scheme alleviates the dependency on system parameters and eliminates the chattering of the control signal through an efficient learning scheme. The CMAC-based IVSC (CIVSC) scheme is proven to be globally stable inasmuch all signals involved are bounded and the tracking error converges to zero. A numerical simulation demonstrates the effectiveness and robustness of the proposed controller.     

基于PID型模糊神经网络的火电站汽包压力控制

为克服火电站燃料-汽包压力调节对象的非线性、时变和纯迟延特性, 采用含自回归神经元的PID型模糊神经网络作为汽包压力控制器, 进行协调控制系统的设计. 仿真研究表明, 这种初值易选、学习能力较强的模糊神经网络控制器可以克服该对象的时变性和随机性干扰, 大大改善控制品质.     

Intelligent control of photovoltaic system using BPSO-GSA-optimized neural network and fuzzy-based PID for maximum power point tracking

The maximum power point tracking (MPPT) technique is applied in the photovoltaic (PV) systems to achieve the maximum power from a PV panel in different atmospheric conditions and to optimize the efficiency of a panel. A proportional-integral-derivative (PID) controller was used in this study for tracking the maximum power point (MPP). A fuzzy gain scheduling system with optimized rules by subtractive clustering algorithm was employed for tuning the PID controller parameters based on error and error-difference in an online mode. In addition, an Elman-type recurrent neural network (RNN) was used for inverse identification of the PV system and for estimating the solar radiation intensity to determine the MPP voltage. The optimum number of neurons in the single hidden-layer of the RNN was determined by binary particle swarm optimization algorithm. The weights of this RNN were also optimized by using a hybrid method based on the Levenberg-Marquardt algorithm and gravitational search algorithm (GSA). In the proposed fitness function for optimization, both the RNN size and its convergence accuracy were considered. Thus, the algorithm for RNN optimization attempts to minimize both the structural complexity and the mean square error. Simulation results revealed superior performance of GSA in comparison with particle swarm, cuckoo, and grey wolf optimization algorithms. The performance of the proposed MPPT method was evaluated under four different ambient conditions. Our experimental results show that the proposed MPPT method is more efficient than the three competitive methods presented in recent years.     

Robust neural network control system design for linear ultrasonic motor

Linear ultrasonic motor (LUSM) has much merit, such as high precision, fast control dynamics and large driving force, etc.; however, the dynamic characteristic of LUSM is nonlinear and the precise dynamic model of LUSM is difficult to obtain. To tackle this problem, this study presents a robust neural network control (RNNC) system for LUSM to track a reference trajectory with L ₂ robust tracking performance. The developed RNNC system is composed of a neural network controller and a robust controller. The neural network controller is the principal controller used to mimic an ideal controller and the robust controller is adopted to achieve L ₂ robust tracking performance. The developed RNNC system is then applied to control an LUSM. Experimental results show that the developed RNNC system can achieve favorable tracking performance with unknown of LUSM model.