Neuro-control Scheme Research Articles

Optimal neuro-control strategy for nonlinear systems with asymmetric input constraints

In this paper, we present an optimal neuro-control scheme for continuous-time &#x0028 CT &#x0029 nonlinear systems with asymmetric input constraints. Initially, we introduce a discounted cost function for the CT nonlinear systems in order to handle the asymmetric input constraints. Then, we develop a Hamilton-Jacobi-Bellman equation &#x0028 HJBE &#x0029 , which arises in the discounted cost optimal control problem. To obtain the optimal neurocontroller, we utilize a critic neural network &#x0028 CNN &#x0029 to solve the HJBE under the framework of reinforcement learning. The CNN &#x02BC s weight vector is tuned via the gradient descent approach. Based on the Lyapunov method, we prove that uniform ultimate boundedness of the CNN&#x02BC s weight vector and the closed-loop system is guaranteed. Finally, we verify the effectiveness of the present optimal neuro-control strategy through performing simulations of two examples.

Application of lattice probabilistic neural network for active response control of offshore structures

The reduction of the dynamic response of an offshore structure subjected to wind-generated random ocean waves is of extreme significance in the aspects of serviceability, fatigue life and safety of the structure. In this study, a new neuro-control scheme is applied to the vibration control of a fixed offshore platform under random wave loads to examine the applicability of the proposed method. It is called the Lattice Probabilistic Neural Network (LPNN), as it utilizes lattice pattern of state vectors as the training data of PNN. When control results of the LPNN are compared with those of the NN and PNN, LPNN showed better performance in effectively suppressing the structural responses in a shorter computational time.

Neuro-control of fixed offshore structures under earthquake

An intelligent control technique using a neural network is proposed for seismic protection of offshore structures. Fluid-structure interaction was considered in developing controller and a training algorithm for the neuron-controller is presented. In the numerical example, the performance of the proposed neuron-controller was evaluated. Moreover, a neuron-controller is tested even when it is trained by using a linearized equation of motion for fluid structure interaction (FSI). Based on the examples, it can be concluded that the proposed neuro-control scheme can be used for offshore structures which have intrinsic nonlinearity due to FSI.

An Experimental Study of Semiactive Modal Neuro-control Scheme Using MR Damper for Building Structure

In this study, a semiactive modal neuro-control scheme which combines the modal neuro-control algorithm with a semiactive MR damper is proposed, and its effectiveness is experimentally verified through a series of shaking table tests. A modal neuro-control scheme uses modal coordinates as inputs of neuro-controller. Hence, it is more convenient to design the controller compared with conventional neuro-control schemes. A Kalman filter is introduced to estimate modal states from measurements. Moreover, the clipped algorithm is adopted to provide an appropriate command voltage to an MR damper. For shaking table tests, a scaled three-story shear building model is considered. Two types of semiactive modal neuro-controllers are trained with a reproduced El Centro earthquake for their own purposes. The performance of the proposed semiactive modal neuro-control scheme is compared with that of the passive-optimal case. In the experiments, the proposed semiactive modal neuro-controllers show better performance than the passive-optimal case, especially in adaptability over various excitations and reducing inter-story drifts as well as accelerations. Moreover, the proposed scheme can be designed for specific purpose which fulfills the designer's requirement (e.g., focusing on reducing inter-story drifts). Therefore, the proposed semiactive modal neuro-controller can be effectively used in reducing seismic responses of large engineering structures.

Fault-Tolerant Optimal Neurocontrol for a Static Synchronous Series Compensator Connected to a Power Network

This paper proposes a fault-tolerant optimal neurocontrol scheme (FTONC) for a Static Synchronous Series Compensator (SSSC) connected to a multi-machine benchmark power system. The dual heuristic programming technique and the radial basis function neural networks are used to design a nonlinear optimal neurocontroller (NONC) for the external control of the SSSC. Compared to the conventional external linear controller, The NONC improves the damping performance of the SSSC. The internal control of the SSSC is achieved by a conventional liner controller. A sensor evaluation and (missing sensor) restoration scheme (SERS) is designed by using the auto-associative neural networks and the particle swarm optimization technique. This SERS provides a set of fault-tolerant measurements to the SSSC controllers and therefore guarantees a fault-tolerant control for the SSSC. The proposed FTONC is verified by simulation studies in PSCAD/EMTDC environment.

Active control strategy of structures based on lattice type probabilistic neural network

A new neuro-control scheme for active control of structures having a basic structure similar to the Probabilistic Neural Network (PNN) is proposed. It utilizes the lattice pattern of state vector as the training data of PNN, and thus it is called the Lattice Probabilistic Neural Network (LPNN). Comparing the two schemes, PNN takes much time to obtain a control force in the application because it uses all the training patterns. This may delay the control action inevitably. However, in LPNN, the control force is calculated by using only the adjacent information of LPNN input, making the response of LPNN greatly faster than that of PNN. To investigate the general control capability of the proposed algorithm, one-story and three-story buildings under California, El Centro, and Northridge earthquakes are used as test models. Control results of the LPNN are compared with those of the conventional PNN, and these show that the structural responses have been suppressed effectively by the proposed algorithm.

격자 확률신경망 기법을 이용한 구조물의 능동 제어

A new neuro-control scheme for active control of structures is proposed. It utilizes lattice pattern of state vector as training data of probabilistic neural network(PNN). Therefore. it is the so-called lattice probabilistic neural network(LPNN). PNN makes control forces by using all the training patterns. Therefore, it takes much time to obtain a control force in application. This inevitably may delay the control action. However. control force of LPNN is calculated by using only the adjacent information of LPNN input. So, the response of LPNN is greatly faster than PNN. The proposed control algorithm is applied for three story building under California and El Centro earthquakes. Also, control results of the LPNN are compared with those of the conventional PNN. The structural responses have been suppressed effectively by the proposed algorithm.

A real-time neuro-adaptive controller with guaranteed stability

This paper presents a new model reference adaptive neuro-control scheme using feedforward neural networks with momentum back-propagation (MBP) learning algorithm. Training is done on-line to tune the parameters of the neuro-controller that provides the control signal. Noting that pre-learning is not required and the structure of overall system is very simple and straightforward. No additional controller or robustifying signal is required. Tracking performance is guaranteed via Lyapunov stability analysis. Both tracking error and neural network weights remain bounded. An interesting fact about the proposed approach is that it does not require a NN being capable of reconstructing globally model non-linearities.

Neuro-control of unmanned underwater vehicles

Unmanned underwater vehicles (UUVs) typically operate in uncertain and changing environments. Since the dynamics of UUVs are highly nonlinear and their hydrodynamic coefficients vary with different operating conditions, a high-performance control system of a UUV is needed to have the capacities of learning and adaptation to the variations in the UUV's dynamics. This paper presents the utilization of an adaptive neuro-control scheme as a controller for controlling a UUV in six degrees of freedom. No prior offline training phase and no explicit knowledge of the structure of the vehicle are required, and the proposed scheme exploits the advantages of both neural network control and adaptive control. Asymptotic convergence of the UUV's tracking errors and stability of the presented control system is guaranteed on the basis of the Lyapunov theory. In this paper, neural network architectures based on radial basis functions and multilayer structures have been used to evaluate the performance of the adaptive controller via computer simulation.

Robust neuro-H∞ controller design for aircraft auto-landing

A robust neuro-control scheme is presented for aircraft auto-landing under severe wind conditions and partial loss of control surfaces. In the scheme, a dynamic radial basis function network (RBFN) called minimal resource allocating network (MRAN), that incorporates a growing and pruning strategy, is utilize to aid an H/sub /spl infin// controller using a feedback-error-learning mechanism. The neural network uses only online learning and is not trained "a priori". Specifically, the performance of this neuro-controller for aircraft auto-landing in a microburst along with a partial loss of control effectiveness is analyzed and compared with other control schemes. Simulation studies show that the performance obtained by the neuro-H/sub /spl infin// control scheme is better than the other control schemes under failure and extreme wind conditions.

Indirect Adaptive Neurocontrol Scheme for a Static Compensator Connected to a Power System

An indirect adaptive neurocontrol scheme for a Static Compensator connected to a power system using two Artificial Neural Networks (ANNs) is presented in this paper. The ANNs are trained online and there is no need for offline data. The neurocontroller has a better performance in adaptively controlling the Static Compensator and damping the system transients, compared to conventional controllers. Preliminary results are provided to show the performance of the neurocontroller for large disturbances.

Neuro-controller design for nonlinear fighter aircraft maneuver using fully tuned RBF networks

In this paper, an on-line learning neuro-control scheme that incorporates a growing radial basis function network (GRBFN) is proposed for a nonlinear aircraft controller design. The scheme is based on feedback-error-learning strategy in which the neuro-flight-controller (NFC) augments a conventional controller in the loop. By using the Lyapunov synthesis approach, the tuning rule for updating all the parameters of the RBFN (weights, widths and centers of the Gaussian functions) is derived which ensures the stability of the overall system with improved tracking accuracy. The theoretical results are validated using simulation studies based on a nonlinear 6-DOF high performance fighter aircraft undergoing a high α stability-axis roll maneuver. Compared with a traditional RBFN where only the weights are tuned, a GRBFN with tuning of all the parameters can implement a more compact network structure with smaller tracking error.

Online learning in adaptive neurocontrol schemes with a sliding mode algorithm

The novel features of an adaptive PID-like neurocontrol scheme for nonlinear plants are presented. The controller tuning is based on an estimate of the command-error on its output by using a neural predictive model. A robust online learning algorithm, based on the direct use of sliding mode control (SMC) theory is applied. The proposed approach allows handling of the plant-model mismatches, uncertainties and parameters changes. The results show that both the plant model and the controller inherit some of the advantages of SMC, such as high speed of learning and robustness.

Implementation of indirect neuro-control for a nonlinear two-robot MIMO system

This paper presents identification and control designs using neural networks for a nonlinear two-robot multiinput multioutput (MIMO) system. The proposed neuro-controller is a combination of linear controllers and a neural-network controller, and is trained by an indirect neuro-control scheme. The proposed neuro-controller is implemented and tested on an IBM PC-based two 2-bar systems holding an object, and is applicable to many d.c.-motor-driven precision nonlinear two-robot MIMO systems. The algorithm and experimental results are described. The experimental results are shown to be superior to those of conventional control.

Neural Network-Based Speed Control of A Two-Mass-Model System

We present an application of artificial neural networks to the problem of controlling the speed of an elastic drive system. We derive a neural network structure to simulate the inverse dynamics of the system, then implement the direct inverse control scheme in a closed loop. The neural network learning is done on-line to adaptively control the speed to follow a stepwise changing reference. The experimental results with a two-mass-model analog board confirm the effectiveness of the proposed neurocontrol scheme.

Implementation of indirect neuro-control for nonlinear dynamic systems

This paper represents identification and control designs using neural networks for a class of nonlinear dynamic systems. The proposed neuro-controller is a combination of a linear controller and a neural network trained by an indirect neuro-control scheme. The proposed neuro-controller is implemented and tested on an IBM PC-based bar system, and is applicable to many dc-motor-driven precision servo mechanisms. The algorithm and experimental results are described. The experimental results are shown to be superior to those of conventional control.

On-Line Optimization of the Turning Process Using an Inverse Process Neurocontroller

This paper presents a feedback neurocontrol scheme that uses an inverse turning process model to synthesize optimal process inputs. The inverse process neurocontroller is implemented in a multilayer feedforward neural network. On-line adjustments of feed rate and cutting speed parameters are carried out based on a cost/quality performance index, estimated from force and vibration sensor measurements. Both non-adaptive and adaptive neurocontrol schemes are considered. The simulations and experimental investigations presented herein demonstrated the effectiveness of neural networks for controlling and optimizing turning operations. Applied to single point turning of a typical finishing cut, the final dimensions and surface finishes were found to be better by 40 and 80 percent respectively, while productivity was increased by 40 percent over the conditions proposed in machining data handbooks. This approach is also applicable to several other manufacturing processes.

Modeling and Control of Pressure in a Chloride Gas Plant by Neural Networks

Abstract This paper is concerned with an application of the neuro-PID controller architecture with a pressure control problem of a chemical plant which produces soda and chloride by electrolysis of salt water. In order to reduce the electrical cost, this plant adopts a switching method of electric current twice a day, which causes fluctuations of pressure inside electrolytic tanks. To stabilize the pressure in the tank, the convenient PID controllers have been devised. In this plant, we discuss the development of a backpropagation(BP) neural network(NN) control scheme to find PID gains. The simulation results indicate that neuro-control schemes are superior and more easily implemented.

Hybrid neural adaptive control for bank-to-turn missiles

A novel hybrid neural adaptive bank-to-turn (BTT) lateral autopilot is described for a short-range command-to-line-of-sight (CLOS) surface-to-air missile. This employs a multiinput-multioutput Gaussian radial basis function (RBF) network in parallel with a constant parameter, independently regulated lateral autopilot, to adaptively compensate for roll-induced cross-coupling time-varying aerodynamic derivatives and control surface constraints, in order to achieve consistent tracking performance over the flight envelope. The hybrid neural autopilot is evaluated in three dimensional (six-degree of freedom) simulation studies against realistic pitch acceleration and roll rate profiles generated from a typical CLOS guidance scenario, and its performance compared with a carefully designed gain scheduled autopilot. The results are found to be encouraging and clearly demonstrate the potential advantages of the neurocontrol scheme.

MIMO furnace control with neural networks

The development of a multilayered neural network control scheme for a multi-input multi-output (MIMO) furnace is discussed. The scheme is based on the back-error-propagation algorithm and uses neural net emulators and controllers. The neural network models are trained using only the input-output characteristics of the plant without the need for using any initial conventional controller or knowledge regarding dynamics. The scheme allows online learning of the neural net emulators and controllers whereby their performance can be further improved. The approach is applicable to a wide variety of open-loop stable systems. Experiments are conducted to see how well the neurocontrol scheme compares with two established control schemes implemented for the same process. Comparisons are made with respect to set-point changes, load disturbance rejection, parameter variations, and controller saturations. The experimental results show that the neurocontrol scheme has considerable robustness and performs better than the other two controllers.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>