Alma Y. Alanis

Discrete-Time Adaptive Backstepping Nonlinear Control via High-Order Neural Networks

This paper deals with adaptive tracking for discrete-time multiple-input-multiple-output (MIMO) nonlinear systems in presence of bounded disturbances. In this paper, a high-order neural network (HONN) structure is used to approximate a control law designed by the backstepping technique, applied to a block strict feedback form (BSFF). This paper also includes the respective stability analysis, on the basis of the Lyapunov approach, for the whole controlled system, including the extended Kalman filter (EKF)-based NN learning algorithm. Applicability of the scheme is illustrated via simulation for a discrete-time nonlinear model of an electric induction motor.

Real-Time Discrete Neural Block Control Using Sliding Modes for Electric Induction Motors

This paper deals with real-time adaptive tracking for discrete-time induction motors in the presence of bounded disturbances. A high-order neural-network structure is used to identify the plant model, and based on this model, a discrete-time control law is derived, which combines discrete-time block-control and sliding-mode techniques. This paper also includes the respective stability analysis for the whole system with a strategy to avoid adaptive weight zero-crossing. The scheme is implemented in real time using a three-phase induction motor.

Discrete-Time High Order Neural Control

The objective of this work is to present recent advances in the theory of neural control for discrete-time nonlinear systems with multiple inputs and multiple outputs. The results that appear in each chapter include rigorous mathematical analyses, based on the Lyapunov approach, that guarantee its properties; in addition, for each chapter, simulation results are included to verify the successful performance of the corresponding proposed schemes. In order to complete the treatment of these schemes, the final chapter presents experimental results related to their application to a electric three phase induction motor, which show the applicability of such designs. The proposed schemes could be employed for different applications beyond the ones presented in this book. The book presents solutions for the output trajectory tracking problem of unknown nonlinear systems based on four schemes. For the first one, a direct design method is considered: the well known backstepping method, under the assumption of complete sate measurement; the second one considers an indirect method, solved with the block control and the sliding mode techniques, under the same assumption. For the third scheme, the backstepping technique is reconsidering including a neural observer, and finally the block control and the sliding mode techniques are used again too, with a neural observer. All the proposed schemes are developed in discrete-time. For both mentioned control methods as well as for the neural observer, the on-line training of the respective neural networks is performed by Kalman Filtering.

Real-Time Recurrent Neural State Estimation

A nonlinear discrete-time neural observer for discrete-time unknown nonlinear systems in presence of external disturbances and parameter uncertainties is presented. It is based on a discrete-time recurrent high-order neural network trained with an extended Kalman-filter based algorithm. This brief includes the stability proof based on the Lyapunov approach. The applicability of the proposed scheme is illustrated by real-time implementation for a three phase induction motor.

Real-time Discrete Backstepping Neural Control for Induction Motors

This brief focuses on real-time implementation, as applied to a three-phase induction motor, of results already published in 2007. The proposed controller is based on a high-order neural network, trained online using Kalman filter learning, to approximate a control law designed by the backstepping technique.

Electric load demand prediction using neural network trained by Kalman filtering

This work presents the application of recurrent multilayer perceptron neural networks to electric load demand prediction; the respective training is performed extended Kalman filtering. The goal is to obtain a 24 hours horizon, prediction for the electric load demand; data from the state of California, USA, is utilized.

High Order Neural Networks for wind speed time series prediction

In this paper, we propose a High Order Neural Network (HONN) trained with an extended Kalman filter based algorithm to predict wind speed. Due to the chaotic behavior of the wind time series, it is not possible satisfactorily to apply the traditional forecasting techniques for time series; however, the results presented in this paper confirm that HONNs can very well capture the complexity underlying wind forecasting; this model produces accurate one-step ahead predictions.

Inverse optimal neural control of blood glucose level for type 1 diabetes mellitus patients

Abstract In this paper, inverse optimal neural control for trajectory tracking is applied to glycemic control of type 1 diabetes mellitus (T1DM) patients. The proposed control law calculates the adequate insulin delivery rate in order to prevent hyperglycemia and hypoglycemia levels in T1DM patients. Two models are used: (1) a nonlinear compartmental model in order to obtain type 1 diabetes mellitus virtual patient behavior, and (2) a neural model obtained from an on-line neural identifier, which uses a recurrent neural network, trained with the extended Kalman filter (EKF); the last one allows the applicability of an inverse optimal neural controller. The proposed algorithm is tuned to track a desired trajectory; this trajectory reproduces the glucose absorption of a healthy person. The applicability of the proposed control scheme is illustrated via simulations.

Higher Order Wavelet Neural Networks with Kalman learning for wind speed forecasting

In this paper, a Higher Order Wavelet Neural Network (HOWNN) trained with an Extended Kalman Filter (EKF) is implemented to solve the wind forecasting problem. The Neural Network based scheme is composed of high order terms in the input layer, two hidden layers, one incorporating radial wavelets as activation functions and the other using classical logistic sigmoid, and an output layer with a linear activation function. A Kalman filter based algorithm is employed to update the synaptic weights of the wavelet network. The size of the regression vector is determined by means of the Lipschitz quotients method. The proposed structure captures more efficiently the complex nature of the wind speed time series. The proposed model is trained and tested using real wind speed data values.

DISCRETE-TIME REDUCED ORDER NEURAL OBSERVERS FOR UNCERTAIN NONLINEAR SYSTEMS

This paper focusses on a novel discrete-time reduced order neural observer for nonlinear systems, which model is assumed to be unknown. This neural observer is robust in presence of external and internal uncertainties. The proposed scheme is based on a discrete-time recurrent high order neural network (RHONN) trained with an extended Kalman filter (EKF)-based algorithm, using a parallel configuration. This work includes the stability proof of the estimation error on the basis of the Lyapunov approach; to illustrate the applicability, simulation results for a nonlinear oscillator are included.