Neural Network Feedback Controller Research Articles

Dynamic Output Feedback and Neural Network Control of a Non-Holonomic Mobile Robot.

This paper presents the design and synthesis of a dynamic output feedback neural network controller for a non-holonomic mobile robot. First, the dynamic model of a non-holonomic mobile robot is presented, in which these constraints are considered for the mathematical derivation of a feasible representation of this kind of robot. Then, two control strategies are provided based on kinematic control for this kind of robot. The first control strategy is based on driftless control; this means that considering that the velocity vector of the mobile robot is orthogonal to its restriction, a dynamic output feedback and neural network controller is designed so that the control action would be zero only when the velocity of the mobile robot is zero. The Lyapunov stability theorem is implemented in order to find a suitable control law. Then, another control strategy is designed for trajectory-tracking purposes, in which similar to the driftless controller, a kinematic control scheme is provided that is suitable to implement in more sophisticated hardware. In both control strategies, a dynamic control law is provided along with a feedforward neural network controller, so in this way, by the Lyapunov theory, the stability and convergence to the origin of the mobile robot position coordinates are ensured. Finally, two numerical experiments are presented in order to validate the theoretical results synthesized in this research study. Discussions and conclusions are provided in order to analyze the results found in this research study.

Feedback Control Systems Stabilization Using a Bio-inspired Neural Network

The proportional–integral–derivative (PID) control systems, which have become a standard for technical and industrial applications, are the fundamental building blocks of classical and modern control systems. In this paper, a three-layer feed-forward neural network (NN) model trained to replicate the behavior of a PID controller is employed to stabilize control systems through a NN feedback controller. A novel bio-inspired weights-and-structure-determination (BIWASD) algorithm, which incorporates a metaheuristic optimization algorithm dubbed beetle antennae search (BAS), is used to train the NN model. More presicely, the BIWASD algorithm identifies the ideal weights and structure of the BIWASD-based NN (BIWASDNN) model utilizing a power sigmoid activation function while handling model fitting and validation. The results of three simulated trials on stabilizing feedback control systems validate and demonstrate the BIWASDNN model’s exceptional learning and prediction capabilities, while achieving similar or better performance than the corresponding PID controller. The BIWASDNN model is compared to three other high-performing NN models, and a MATLAB repository is accessible in public through GitHub to encourage and enhance this work.

Novel Adaptive Fuzzy Feedback Neural Network Controller for Narrowband Active Noise Control System

With the development of science and technology in recent years, many operating machines have become sources of noise affecting quality of life. Hence, the topic of noise diminution using an active noise control (ANC) system has attracted many researchers. This paper develops a new adaptive fuzzy feedback neural network controller (AFFNNC) to improve the performance for narrowband active noise control (NANC) systems. The proposed controller combines fuzzy inference and adaptive feedback neural network controllers that are based on the filtered-s least mean square (FSLMS) algorithm. The AFFNNC comprises five network layers, in which the output layer of the controller uses an adaptive algorithm to tune directly the parameters of filters without prior training. The computational complexity, convergence and stability of the AFFNNC are analyzed. Evaluations are performed on both linear and nonlinear NANC systems with a recorded noise signal that was obtained from a transformer. Numerical simulations confirm the efficiency of the proposed controller compared with other ANC controllers.

Constrained Training for Guided Feedback Controller

In this paper, we are focusing on a neural network feedback controller tailored to a lane following application. Besides a classical goal like avoiding a collision with obstacles, we aim at investigating an approach in order to guarantee not to leave the street at any point, since this is of major importance for the safety of the passengers. Unfortunately, guarantees like this are quite rare since the training of neural networks is often based on data, which only allows statistical conclusions. We derive a deterministic inequality equation guaranteeing this safety. Furthermore, we introduce a penalty strategy in order to integrate this inequality into a Reinforcement Learning algorithm.

Thermal comfort control via air conditioning system using fuzzy neural network feedback controller

<span>Despite their complexity and uncertainty, air conditioning systems should provide the optimal thermal conditions in a building. These controller systems should be adaptable to changes in environmental parameters. In most air conditioning systems, today, there are On/Off controllers or PID in more advanced types, which, due to different environmental conditions, are not optimal and cannot provide the optimal environmental conditions. Controlling thermal comfort of an air conditioning system requires estimation of thermal comfort index. In this study, fuzzy controller was used to provide thermal comfort in an air conditioning system, and neural network was used to estimate thermal comfort in the feedback path of the controller. Fuzzy controller has a good response given the non-linear features of air conditioning systems. In addition, the neural network makes it possible to use thermal comfort feedback in a real-time control.</span>

Vibration Control of a Flexible Robotic Manipulator in the Presence of Input Deadzone

In this paper, a neural network (NN) controller is designed to suppress the vibration of a flexible robotic manipulator system with input deadzone. The NN aims to approximate the unknown robotic manipulator dynamics and eliminate the effects of input deadzone in the actuators. In order to describe the system more accurately, the model of the flexible manipulator is constructed based on the lumping spring-mass method. Full state feedback NN control is proposed first and output feedback NN control with a high-gain observer is then devised to make the proposed control scheme more practical. The effect of input deadzone is approximated by a radial basis function neural network (RBFNN) and the unknown dynamics of the manipulator is approximated by another RBFNN. The proposed NN control is able to compensate for the estimated deadzone effect and track the desired trajectory. For the stability analysis, the Lyapunov's direct method is used to ensure uniform ultimate boundedness (UUB) of the closed-loop system. Simulations are given to verify the control performance of the NN controllers comparing with the proportional derivative (PD) controller. At last, the experiments are conducted on the Quanser platform to further prove the feasibility and control performance of the NN controllers.

A composite control method based on the adaptive RBFNN feedback control and the ESO for two-axis inertially stabilized platforms

Due to the nonlinearity and time variation of a two-axis inertially stabilized platform (ISP) system, the conventional feedback control cannot be utilized directly. To realize the control performance with fast dynamic response and high stabilization precision, the dynamic model of the ISP system is expected to match the ideal model which satisfies the desired control performance. Therefore, a composite control method based on the adaptive radial basis function neural network (RBFNN) feedback control and the extended state observer (ESO), is proposed for ISP. The adaptive RBFNN is proposed to generate the feedback control parameters online. Based on the state error information in the working process, the adaptive RBFNN can be constructed and optimized directly. Therefore, no priori training data is needed for the construction of the RBFNN. Furthermore, a linear second-order ESO is constructed to compensate for the composite disturbance. The asymptotic stability of the proposed control method has been proven by the Lyapunov stability theory. The applicability of the proposed method is validated by a series of simulations and flight tests.

Novel Decentralized Control of Power Systems With Penetration of Renewable Energy Sources in Small-Scale Power Systems

In this paper, the power grid with penetration of renewable energy sources is modeled as a multigenerator interconnected power network. The power grid includes distributed energy resources including conventional synchronous generators and renewable energy sources; here called renewable generators that are connected to the grid via grid-tie inverters (GTIs). With the proposed modeling, the GTI resembles a synchronous generator with excitation control. The modeling takes into account the dc-link capacitor stored energy as a dynamical state, in contrast with the available methods, and through an appropriate controller assures the stability of the dc link and the entire grid without needing an abundant-energy dc link. Next, the power grid comprising the synchronous and renewable generators is converted to decentralized control form with subsystems in Brunovsky canonical form whose interactions with the rest of the grid are unknown. A decentralized adaptive neural network (NN) feedback controller is proposed with quadratic update law to stabilize the rotor speed and dc-link voltage oscillations in asymptotic fashion in the presence of grid disturbances. The proposed controller is then simplified. Though the solar power interacting with conventional synchronous generators is considered in this paper, the proposed modeling and controller design can be applied to many other renewable energy systems. Simulation results on the IEEE 14-bus power system with penetration of solar power are provided to show the effectiveness of the approach in damping oscillations that occur after disturbances.

Adaptive output feedback neural network control of uncertain non-affine systems with unknown control direction

This paper deals with the problem of adaptive output feedback neural network controller design for a SISO non-affine nonlinear system. Since in practice all system states are not available in output measurement, an observer is designed to estimate these states. In comparison with the existing approaches, the current method does not require any information about the sign of control gain. In order to handle the unknown sign of the control direction, the Nussbaum-type function is utilized. In order to approximate the unknown nonlinear function, neural network is firstly exploited, and then to compensate the approximation error and external disturbance a robustifying term is employed. The proposed controller is designed based on strict-positive-real (SPR) Lyapunov stability theory to ensure the asymptotic stability of the closed-loop system. Finally, two simulation studies are presented to demonstrate the effectiveness of the developed scheme.

Hybrid control method for hysteresis non-linearity of magnetically controlled shape memory actuator

In order to eliminate the hysteresis non-linearity between the input and output of the magnetically controlled shape memory alloy (MSMA) actuator and improve the control accuracy, a hybrid control method is proposed. The proposed hybrid control method consists of two parts: a hysteresis inverse model feedforward controller and a neural network feedback controller. According to the hysteresis non-linearity between the input and output of MSMA actuator, a double sigmoid function (double S function) and a four layer feedforward neural network are used to establish the hysteresis model of MSMA actuator. With the same method, the hysteresis inverse model is established. A three layer feedforward neural network is used in the feedback controller. The experimental results show that the proposed control method can effectively improve the output accuracy of the MSMA actuator.

Adaptive dynamic output feedback neural network control of uncertain MIMO nonlinear systems with prescribed performance.

An adaptive dynamic output feedback neural network controller for a class of multi-input/multi-output affine in the control uncertain nonlinear systems is designed, capable of guaranteeing prescribed performance bounds on the system's output as well as boundedness of all other closed loop signals. It is proved that simply guaranteeing a boundedness property for the states of a specifically defined augmented closed loop system is necessary and sufficient to solve the problem under consideration. The proposed dynamic controller is of switching type. However, its continuity is guaranteed, thus alleviating any issues related to the existence and uniqueness of solutions. Simulations on a planar two-link articulated manipulator illustrate the approach.

Application of Fuzzy Neural Network in Ship Course Control

As the uncertainty of parameters and the random nature of disturbances that effect a ship’s course，one of the methods which can be used for designing a nonlinear ship course controller is the neural intelligent method. It is used here for designing a configuration of a nonlinear controller, which are then applied to ship course control. In the paper, we used fuzzy logic and neural network based on T-S model to design the controller. During the control design process, the structure of ship automatic steering and fuzzy neural network feedback controller were built. In order to enhance adaptive characteristics of the controller，the parameters of the network were optimized by a GA algorithm. The proposed course keeping method was tested in a simulation study involving a non-linear model of a ship. Simulation results show that the performance of the ship controller is valuable and easy to implement.

Adaptive Output Feedback NN Control of a Class of Discrete-Time MIMO Nonlinear Systems With Unknown Control Directions

In this paper, adaptive neural network (NN) control is investigated for a class of block triangular multiinput-multioutput nonlinear discrete-time systems with each subsystem in pure-feedback form with unknown control directions. These systems are of couplings in every equation of each subsystem, and different subsystems may have different orders. To avoid the noncausal problem in the control design, the system is transformed into a predictor form by rigorous derivation. By exploring the properties of the block triangular form, implicit controls are developed for each subsystem such that the couplings of inputs and states among subsystems have been completely decoupled. The radial basis function NN is employed to approximate the unknown control. Each subsystem achieves a semiglobal uniformly ultimately bounded stability with the proposed control, and simulation results are presented to demonstrate its efficiency.

Neural network solution for suboptimal control of non-holonomic chained form system

In this paper, we develop fixed-final time nearly optimal control laws for a class of non-holonomic chained form systems by using neural networks to approximately solve a Hamilton—Jacobi—Bellman equation. A certain time-folding method is applied to recover uniform complete controllability for the chained form system. This method requires an innovative design of a certain dynamic control component. Using this time-folding method, the chained form system is mapped into a controllable linear system for which controllers can systematically be designed to ensure exponential or asymptotic stability as well as nearly optimal performance. The result is a neural network feedback controller that has time-varying coefficients found by a priori offline tuning. The results of this paper are demonstrated in an example.

Approximate Adaptive Output Feedback Stabilization via Passivation of MIMO Uncertain Systems Using Neural Networks

An adaptive output feedback neural network controller is designed, which is capable of rendering affine-in-the-control uncertain multi-input-multi-output nonlinear systems strictly passive with respect to an appropriately defined set. Consequently, a simple output feedback is employed to stabilize the system. The controlled system need not be in normal form or have a well-defined relative degree. Without requiring a zero-state detectability assumption, uniform ultimate boundedness, with respect to an arbitrarily small set, of both the system's state and the output is guaranteed, along with boundedness of all other signals in the closed loop. To effectively avoid possible division by zero, the proposed adaptive controller is of switching type. However, its continuity is guaranteed, thus alleviating drawbacks connected to existence of solutions and chattering phenomena. Simulations illustrate the approach.

Output Feedback NN Control for Two Classes of Discrete-Time Systems With Unknown Control Directions in a Unified Approach

In this paper, output feedback adaptive neural network (NN) controls are investigated for two classes of nonlinear discrete-time systems with unknown control directions: 1) nonlinear pure-feedback systems and 2) nonlinear autoregressive moving average with exogenous inputs (NARMAX) systems. To overcome the noncausal problem, which has been known to be a major obstacle in the discrete-time control design, both systems are transformed to a predictor for output feedback control design. Implicit function theorem is used to overcome the difficulty of the nonaffine appearance of the control input. The problem of lacking a priori knowledge on the control directions is solved by using discrete Nussbaum gain. The high-order neural network (HONN) is employed to approximate the unknown control. The closed-loop system achieves semiglobal uniformly-ultimately-bounded (SGUUB) stability and the output tracking error is made within a neighborhood around zero. Simulation results are presented to demonstrate the effectiveness of the proposed control.

Adaptive Predictive Control Using Neural Network for a Class of Pure-Feedback Systems in Discrete Time

In this paper, adaptive neural network (NN) control is investigated for a class of nonlinear pure-feedback discrete-time systems. By using prediction functions of future states, the pure-feedback system is transformed into an n-step-ahead predictor, based on which state feedback NN control is synthesized. Next, by investigating the relationship between outputs and states, the system is transformed into an input-output predictor model, and then, output feedback control is constructed. To overcome the difficulty of nonaffine appearance of the control input, implicit function theorem is exploited in the control design and NN is employed to approximate the unknown function in the control. In both state feedback and output feedback control, only a single NN is used and the controller singularity is completely avoided. The closed-loop system achieves semiglobal uniform ultimate boundedness (SGUUB) stability and the output tracking error is made within a neighborhood around zero. Simulation results are presented to show the effectiveness of the proposed control approach.

ニューラルネットとオブザーバを用いた油圧サーボ系の制御

The purpose of this study is to improve performance of hydraulic servo systems using a state feedback Neural Network and observer. It is generally not easy to acquire high control performance in hydraulic servo systems, since they have an intense non-linearity. The state feedback Neural Network compensator is one of the available options to get over this inherent problem. However, to measure the entire state variables of the actual control object is generally difficult. Using an observer is an effective method for estimating missing state variables of the control object. In this paper, a state feedback Neural Network control system with observer is proposed for improving the control performance of hydraulic servo systems under the condition of reduced state variables. At first, two Neural Network control systems with observer are designed, and they are incorporated into the hydraulic motor system. Then, the angular velocity and position control performance of the two control systems are experimentally compared to determine the optimum structure of the control system. From the experimental result, suitable structure of the Neural Network with observer is confirmed. Finally, it is clarified that the proposed Neural Network control system has excellent performance.

A neural network solution for fixed-final time optimal control of nonlinear systems

In this paper, fixed-final time optimal control laws using neural networks and HJB equations for general affine in the input nonlinear systems are proposed. The method utilizes Kronecker matrix methods along with neural network approximation over a compact set to solve a time-varying HJB equation. The result is a neural network feedback controller that has time-varying coefficients found by a priori offline tuning. Convergence results are shown. The results of this paper are demonstrated on an example.

Nearly optimal neural network stabilization of bipedal standing using genetic algorithm

In this work, stability control of bipedal standing is investigated. The biped is simplified as an inverted pendulum with a foot-link. The controller consists of a general regression neural network (GRNN) feedback control, which stabilizes the inverted pendulum in a region around the upright position, and a PID feedback control, which keeps the pendulum at the upright position. The GRNN controller is also designed to minimize an energy-related cost function while satisfying the constraints between the foot-link and the ground. The optimization has been carried out using the genetic algorithm (GA) and the GRNN is directly trained during optimization iteration process to provide the closed loop feedback optimal controller. The stability of the controlled system is analyzed using the concept of Lyapunov exponents, and a stability region is determined. Simulation results show that the controller can keep the inverted pendulum at the upright position while nearly minimizing an energy-related cost function and keeping the foot-link stationary on the ground. The work contributes to bipedal balancing control, which is important to the development of bipedal robots.