Aydogan Savran

Multifeedback-Layer Neural Network

The architecture and training procedure of a novel recurrent neural network (RNN), referred to as the multifeedback-layer neural network (MFLNN), is described in this paper. The main difference of the proposed network compared to the available RNNs is that the temporal relations are provided by means of neurons arranged in three feedback layers, not by simple feedback elements, in order to enrich the representation capabilities of the recurrent networks. The feedback layers provide local and global recurrences via nonlinear processing elements. In these feedback layers, weighted sums of the delayed outputs of the hidden and of the output layers are passed through certain activation functions and applied to the feedforward neurons via adjustable weights. Both online and offline training procedures based on the backpropagation through time (BPTT) algorithm are developed. The adjoint model of the MFLNN is built to compute the derivatives with respect to the MFLNN weights which are then used in the training procedures. The Levenberg-Marquardt (LM) method with a trust region approach is used to update the MFLNN weights. The performance of the MFLNN is demonstrated by applying to several illustrative temporal problems including chaotic time series prediction and nonlinear dynamic system identification, and it performed better than several networks available in the literature

Intelligent adaptive nonlinear flight control for a high performance aircraft with neural networks

This paper describes the development of a neural network (NN) based adaptive flight control system for a high performance aircraft. The main contribution of this work is that the proposed control system is able to compensate the system uncertainties, adapt to the changes in flight conditions, and accommodate the system failures. The underlying study can be considered in two phases. The objective of the first phase is to model the dynamic behavior of a nonlinear F-16 model using NNs. Therefore a NN-based adaptive identification model is developed for three angular rates of the aircraft. An on-line training procedure is developed to adapt the changes in the system dynamics and improve the identification accuracy. In this procedure, a first-in first-out stack is used to store a certain history of the input-output data. The training is performed over the whole data in the stack at every stage. To speed up the convergence rate and enhance the accuracy for achieving the on-line learning, the Levenberg-Marquardt optimization method with a trust region approach is adapted to train the NNs. The objective of the second phase is to develop intelligent flight controllers. A NN-based adaptive PID control scheme that is composed of an emulator NN, an estimator NN, and a discrete time PID controller is developed. The emulator NN is used to calculate the system Jacobian required to train the estimator NN. The estimator NN, which is trained on-line by propagating the output error through the emulator, is used to adjust the PID gains. The NN-based adaptive PID control system is applied to control three angular rates of the nonlinear F-16 model. The body-axis pitch, roll, and yaw rates are fed back via the PID controllers to the elevator, aileron, and rudder actuators, respectively. The resulting control system has learning, adaptation, and fault-tolerant abilities. It avoids the storage and interpolation requirements for the too many controller parameters of a typical flight control system. Performance of the control system is successfully tested by performing several six-degrees-of-freedom nonlinear simulations.

A multivariable predictive fuzzy PID control system

In this paper, a novel multivariable predictive fuzzy-proportional-integral-derivative (F-PID) control system is developed by incorporating the fuzzy and PID control approaches into the predictive control framework. The developed control system has two main units referred as adaptation and application parts. The adaptation part consists of a F-PID controller and a fuzzy predictor. The incremental control actions are generated by the F-PID controller. The controller parameters are adjusted with the predictive control approach. The fuzzy predictor provides the multi-step ahead predictions of the plant outputs. Therefore, the F-PID controller parameters are adjusted by minimizing the errors between the predicted plant outputs and reference trajectories over the prediction horizon. The fuzzy predictor is trained with an on-line training procedure in order to adapt the changes in the plant dynamics and improve the prediction accuracy. The Levenberg-Marquardt (LM) optimization method with a trust region approach is used to adjust both the controller and predictor fuzzy systems parameters. In the application part, an identical F-PID controller of the adaptation part is used to control the actual plant. The adjusted parameter values are transferred to this identical controller at each time step. The performance of the proposed control system is tested for both single-input single-output (SISO) and multiple-input multiple-output (MIMO) nonlinear control problems. The adaptation, robustness to noise, disturbance rejection properties together with the tracking performances are examined in the simulations.

A fuzzy model based adaptive PID controller design for nonlinear and uncertain processes.

We develop a novel adaptive tuning method for classical proportional-integral-derivative (PID) controller to control nonlinear processes to adjust PID gains, a problem which is very difficult to overcome in the classical PID controllers. By incorporating classical PID control, which is well-known in industry, to the control of nonlinear processes, we introduce a method which can readily be used by the industry. In this method, controller design does not require a first principal model of the process which is usually very difficult to obtain. Instead, it depends on a fuzzy process model which is constructed from the measured input-output data of the process. A soft limiter is used to impose industrial limits on the control input. The performance of the system is successfully tested on the bioreactor, a highly nonlinear process involving instabilities. Several tests showed the methods success in tracking, robustness to noise, and adaptation properties. We as well compared our systems performance to those of a plant with altered parameters with measurement noise, and obtained less ringing and better tracking. To conclude, we present a novel adaptive control method that is built upon the well-known PID architecture that successfully controls highly nonlinear industrial processes, even under conditions such as strong parameter variations, noise, and instabilities.

An adaptive recurrent fuzzy system for nonlinear identification

This paper describes the architecture and training procedure of a recurrent fuzzy system (RFS). The RFS is composed of a fuzzy inference system (FIS) and a delayed feedback connection. The recurrent property comes from feeding the FIS output back to the FIS input via an adjustable feedback parameter. Both the on-line and off-line training procedures based on the backpropagation-through-time (BPTT) algorithm have been investigated. The adjoint model of the RFS is obtained and used to compute the gradients. It is shown that the off-line training is insufficient to adapt to changes in system dynamics. So, an on-line training procedure is derived. In this procedure, a first in first out stack is used to store a certain history of the input-output data to perform a truncated BPTT algorithm. A quasi-Newton optimization method with a line search algorithm is used to adjust the RFS parameters. The performance of the developed RFS is demonstrated by applying to the identification of nonlinear dynamic systems. The simulation studies show that the proposed identification model has the ability to learn dynamics of highly nonlinear systems and compensate system uncertainties. The results are promising for the further application in the area of control and modeling.

Discrete state space modeling and control of nonlinear unknown systems

A novel procedure for integrating neural networks (NNs) with conventional techniques is proposed to design industrial modeling and control systems for nonlinear unknown systems. In the proposed approach, a new recurrent NN with a special architecture is constructed to obtain discrete-time state-space representations of nonlinear dynamical systems. It is referred as the discrete state-space neural network (DSSNN). In the DSSNN, the outputs of the hidden layer neurons of the DSSNN represent the systems (pseudo) state. The inputs are fed to output neurons and the delayed outputs of the hidden layer neurons are fed to their inputs via adjustable weights. The discrete state space model of the actual system is directly obtained by training the DSSNN with the input-output data. A training procedure based on the back-propagation through time (BPTT) algorithm is developed. The Levenberg-Marquardt (LM) method with a trust region approach is used to update the DSSNN weights. Linear state space models enable to use well developed conventional analysis and design techniques. Thus, building a linear model of a system has primary importance in industrial applications. Thus, a suitable linearization procedure is proposed to derive the linear state space model from the nonlinear DSSNN representation. The controllability, observability and stability properties are examined. The state feedback controllers are designed with both the linear quadratic regulator (LQR) and the pole placement techniques. The regulator and servo control problems are both addressed. A full order observer is also designed to estimate the state variables. The performance of the proposed procedure is demonstrated by applying for both single-input single-output (SISO) and multiple-input multiple-output (MIMO) nonlinear control problems.

Real-time implementation of mixing adaptive control on quadrotor UAVs

In this study, a novel multiple model adaptive control scheme is designed and implemented for quadrotor unmanned aerial vehicles (UAVs). The proposed scheme involves a mixing adaptive controller that blends a set of pre-designed linear quadratic controllers. A particular goal of the design is guaranteeing robustness in lateral motion against modeling uncertainties. The designed controller scheme is tested via real-time experiments on Quanser Qball-X4 UAVs. The experimental results verify the efficiency of the proposed scheme.

Hardware implementation of a wavelet neural network using FPGAs

In this paper, hardware implementation of a wavelet neural network (WNN) is described. The WNN is developed in MATLAB and implemented on a Field-Programmable Gate Array (FPGA) device. The structure of the WNN is similar to the radial basis function (RBF) network, except that here the radial basis functions are replaced by orthonormal scaling functions. The training of the WNN is simplified due to the orthonormal properties of the scaling functions. The performances of the proposed WNN are tested by applying for the function approximation, system identification and the classification problems. Because of their parallel processing properties, the FPGAs provide good alternative in real-time applications of the WNN. By means of the simple scaling function used in the WNN architecture, it can be favorable to multilayer feedforward neural network and the RBF Networks implemented on the FPGA devices.

A multilayer feedforward fuzzy neural network

This paper describes the architecture and learning procedure of a multilayer feedforward fuzzy neural network (FNN). The FNN is designed by replacing the sigmoid type activation function of the multilayer neural network (NN) with the fuzzy system (FS). The Levenberg-Marquardt (LM) optimization method with a trust region approach is adapted to train the FNN. Simulation results of a nonlinear system identification problem are given to show the validity of the approach.