Hybrid Neural Network Control Research Articles

Implementation of a hybrid neural network control technique to a cascaded MLI based SAPF

This paper presents a naïve back propagation (NBP) based icos∅\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$i\\cos \\emptyset$$\\end{document} technique implemented to a cascaded multilevel inverter (MLI) based shunt active power filter (SAPF). The recommended control algorithm is applied to extract the fundamental component of load current and to decide the compensating current reference for harmonic elimination. The performance of the SAPF using the proposed NBP-based icos∅\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$i\\cos \\emptyset$$\\end{document} technique is compared with another two classical control techniques, such as, id-iq\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$i_{d} - i_{q}$$\\end{document} technique and icos∅\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$i\\cos \\emptyset$$\\end{document} technique. The accuracy of the proposed control technique depends on the tuned estimation of active and reactive weights. The performance study of the proposed SAPF with the proposed control technique is investigated under non-linear conditions, with balanced and unbalanced loading conditions. The results reveal that the recommended SAPF is efficient enough to reduce the harmonics from the source current with smooth variation in DC link voltage. The effectiveness of the proposed method is validated by simulation using MATLAB Simulink, and the real-time results are also validated by the experimental setup.

Hybrid neural network control design for uncertain switched nonlinear systems with external disturbances: Application to ship maneuvering system

In this work, the problem of hybrid neural network control for a class of switched uncertain nonlinear systems in strict-feedback form is considered. To approximate unknown nonlinear functions, an improved whale optimization algorithm (IWOA)-based hybrid neural network controller is introduced. Based on hybrid neural network approximation ability, a new hybrid neural network controller is constructed via the backstepping technique and the common Lyapunov function (CLF) is used for the stability analysis of the proposed model. The suggested method ensures that all signals in the closed-loop system are semi-global uniform ultimate bounded (SGUUB), and the tracking error converges to a small neighborhood of zero. Finally, the proposed scheme is applied to a ship maneuvering system and a numerical example to verify its effectiveness.

Hybrid neural network control of uncertain switched nonlinear systems with bounded disturbance

AbstractIn this work, we focus on the design of a hybrid neural tracking controller for a class of uncertain switched nonlinear systems with bounded disturbance. A new tracking control model is constructed using adaptive particle swarm optimization (APSO) based neural network called hybrid neural network (HNN). Hybrid neural tracking controller is developed by combining the backstepping approach and neural network approximation ability along with complexity analysis. A common Lyapunov function (CLF) is used for the stability of the proposed model, to develop a CLF for the switched system, a virtual control function is developed via adaptive law and HNN. By appropriately choosing the design parameters, it has been proven that all closed‐loop signals are semi‐globally uniformly ultimately bounded and the tracking error converges to a small bounded region around the origin. Finally, two numerical examples and a real‐life example about one‐link manipulator systems demonstrate the effectiveness of the proposed hybrid controller.

Hybrid Neural Network Control for Uncertain Nonlinear Discrete-Time Systems with Bounded Disturbance

Hybrid Neural Network Control for Uncertain Nonlinear Discrete-Time Systems with Bounded Disturbance

Hybrid neural network controller for uncertain nonlinear discrete-time systems with non-symmetric dead zone and unknown disturbances

An adaptive hybrid neural control scheme is presented for uncertain non-linear discrete-time Systems (UNLDTS) with non-symmetric dead-zone input and unknown disturbances. This work aims to design an efficient control scheme for the proposed systems in the case of non-symmetric dead-zone as the input due to the presence of which the control of such systems becomes very complex and difficult. The system is converted into an n-step ahead predictor and an adaptive compensative term is introduced to overcome the non-symmetric dead-zone present in the system. A hybrid neural network controller is constructed for the control of the proposed class of systems. The designed controller is proved to be semi-globally uniformly ultimately bounded with the assistance of Lyapunov theory and the error is proved to approach very close to zero. The effectiveness of the proposed scheme is validated through two simulation examples and one example has been inspired from a real word system named continuous stirred tank reactor taken in discrete-time form which ensures the applicability of this controller in real-world applications.

Hybrid neural network control for nonlinear continuous time systems with time delays and dead zone input

SummaryIn this article, a hybrid neural controller is designed in the form of a hybrid neural network (HNN) to deal with the problem of nonlinear continuous‐time systems with time delays and dead‐zone input. The designed HNN comprises of a neural network (NN) and firefly algorithm (FA). The approximation of unknown functions is carried out by using HNN and based on transformed systems, an adaption law, virtual control law and optimal control law are designed. By using the Lyapunov method, it is proved that the designed controller is stable in the sense of semiglobally uniformly ultimate boundedness (SGUUB) and the system output tracks the reference signal with tracking error converging into a small neighborhood of zero. Complexity analysis of the proposed controller is discussed and two numerical examples are presented to prove the efficiency and effectiveness of the proposed model.

Adaptive Tracking Control of Cooperative Robot Manipulators With Markovian Switched Couplings

Many cooperative robotic systems have not only modeling heterogeneity and uncertainty but also switched couplings, causing control difficulties. Here, we develop a neural network adaptive control framework for cooperative robot manipulators with unknown Euler–Lagrange dynamics and Markovian switched couplings. Second-order Markovian switching networks are used for modeling such cooperative robotic systems, which admit a hybrid neural network control with a desired tracking performance. The hybrid neural network control scheme contains a distributed adaptive controller and a hybrid adaptation law, enabling learning in the closed-loop system. The position and velocity tracking errors are shown to be practically uniformly exponentially stable in the mean-square sense, respectively, guaranteeing the second-order practical tracking. The results also suggest that the neural weight evolves with practical convergence to the ideal, showing the effect of network structure on the adaptation capacity.

DSTATCOM employing hybrid neural network control technique for power quality improvement

A new hybrid control technique called gradient descent back propagation (GDBP)-based φ for a three phase two-level distribution static compensator (DSTATCOM) to perform the functions such as harmonic mitigation, power factor correction under reactive loads, which further reduces the DC link voltage across the self-supported capacitor of voltage source converter (VSC). The weighted value of fundamental active and reactive components of load currents are extracted using the proposed control technique to generate the reference source currents. Furthermore, these currents are used to trigger the VSC of the DSTATCOM. The effectiveness of this control technique is demonstrated through simulation using MATLAB/SIMULINK and sim power system tool boxes. The real-time implementation of DSTATCOM is also realised by real-time digital simulator. These results reveal the robustness of the proposed DSTATCOM as it is showing outstanding harmonic compensation capabilities under the various loading conditions and keeping the total harmonics distortion of the source current well <;5%, the limit imposed by IEEE-519 standard.

Hybrid Neural Network Controller Design for a Batch Reactor to Produce Methyl Methacrylate

Methyl methacrylate (MMA) production in an exothermic batch reactor provides a challenging problem for studying its dynamics behavior and temperature control. This work presents a neural network forward model (NN) to predict a concentration of methyl methacrylate, a jacket temperature and temperature profile in the reactor. An optimal NN model has been employed to predict state variables incorporating into a model predictive control (MPC) algorithm to determine optimal control actions. To control the temperature, neural network based control approaches: a neural network direct inverse control (NNDIC) and a neural network based model predictive control (NNMPC) have been formulated. In addition, a dynamic optimization approach has been applied to find out an optimal operating temperature to achieve maximizing the MMA product at specified final time. Simulation results have indicated that the NNMPC is robust and gives the best control results among the PID and NNDIC in all cases.

不整地移動ロボットのための環境適応型制御システム

This paper describes the development of an adaptive control system for an outdoor mobile robot. The adaptive control system is composed of an environment recognition system using a self-organizing map and hybrid-neural network controllers based on neural networks. The environment recognition system can recognize the environment in which the robot travels and can switch the hybrid-neural network controller. The hybrid-neural network controllers are tuned by experimental results for each environment. To evaluate the performance of target tracking and vibration suppression, an experiment using the wheeled mobile robot, “Zaurus”, was conducted in rough terrain. As a result, our proposed method could show less oscillatory motion in rough terrain and performed better than a well tuned PID controller.

The study of neural network-based controller for controlling dissolved oxygen concentration in a sequencing batch reactor

The design and development of the neural network (NN)-based controller performance for the activated sludge process in sequencing batch reactor (SBR) is presented in this paper. Here we give a comparative study of various neural network (NN)-based controllers such as the direct inverse control, internal model control (IMC) and hybrid NN control strategies to maintain the dissolved oxygen (DO) level of an activated sludge system by manipulating the air flow rate. The NN inverse model-based controller with the model-based scheme represents the controller, which relies solely upon the simple NN inverse model. In the IMC, both the forward and inverse models are used directly as elements within the feedback loop. The hybrid NN control consists of a basic NN controller in parallel with a proportional integral (PI) controller. Various simulation tests involving multiple set-point changes, disturbances rejection and noise effects were performed to review the performances of these various controllers. From the results it can be seen that hybrid controller gives the best results in tracking set-point changes under disturbances and noise effects.