Artificial Neural Network Based Inverse Model Control Of A Nonlinear Process (original) (raw)
Related papers
Direct Inverse Neural Network Control of A Continuous Stirred Tank Reactor ( CSTR )
2009
In recent years, there has been a significant increase in the number of control system techniques that are based on nonlinear concepts. One such method is the nonlinear inverse model based control strategy. This method is dependent on the availability of the inverse of the system model. Since neural networks have the ability to model any nonlinear system including their inverses and their use in this control scheme is promising. In the present paper, direct inverse neural network control strategy for controlling the CSTR with van de vusse reaction is studied. The direct inverse NN control strategy utilizes the process inverse model as controller. For training the neural network, the process input-output data is generated by applying a pseudo random signal on a simulink model of the CSTR process. Then, the input-output data is divided into two parts for training & validation. Training is performed using the Levenberg-Marquardt method. Based on the SSE, the optimum number of hidden no...
International Journal of Engineering Research and Technology (IJERT), 2014
https://www.ijert.org/comparative-study-of-direct-inverse-neural-controller-with-conventional-pi-designed-at-lower-space-velocity-controller-for-an-isothermal-continuous-stirred-tank-reactor-with-input-multiplicities https://www.ijert.org/research/comparative-study-of-direct-inverse-neural-controller-with-conventional-pi-designed-at-lower-space-velocity-controller-for-an-isothermal-continuous-stirred-tank-reactor-with-input-multiplicities-IJERTV3IS071186.pdf In the present work, the Neural network (NN) based controller design has been implemented for a non-linear continuous stirred tank reactor processes with input multiplicities. Multilayer feed forward networks (FFNN) were used as direct inverse neural network (DINN) controllers. The training as well as testing database was created by perturbing the open loop process with pseudo random signals (PRS). Direct inverse neural network controller is analyzed to a continuous stirred tank reactor (CSTR) carrying out series and parallel reaction: A→B→C and 2A→D (Van de Vusse reaction) and exhibiting input multiplicities in the space velocity (i.e., manipulated variable), on the product concentration (B), (i.e. the controlled variable). Continuous Stirrer Tank Reactor which exhibits input multiplicities in space velocity on concentration. i.e., two values of space velocity will give the same value of concentration. The Performance of proposed direct inverse neural network controller and linear PI controller has been evaluated at lower and higher input. As the Neural network controller provides always the two values of space velocity for control action and by selecting the value at higher and lower to the operating point, it is found to give stable and faster responses than linear PI controller. Thus, direct inverse neural network control is found to overcome the control problems due to the input multiplicities at lower and higher input space velocities. It is interesting to note that the present neural network controller is giving superior performance like previously proposed nonlinear controller by authors (Reddy, G.P. and Chidambaram, M (1995)) to overcome the control problems due to input multiplicities.
Model based Controller Design using Real Time Neural Network Model and PSO for Conical Tank System
2020
Most of the industrial processes are nonlinear in nature and demanding an optimal control structure. Conventional controllers do not handle the nonlinear system behaviour effectively and they also have tuning associated problems. In this paper, an optimized new generation RTDA (Robustness, Set point tracking, Disturbance Rejection, Aggressiveness) controller is designed for a nonlinear conical tank system. The enhanced features of RTDA controller enables us to tune the parameters separately without affecting each other to obtain optimum performance where the other contemporary controllers fails to do so. The proposed scheme incorporates NARX (Nonlinear Autoregressive with Exogenous input) neural model in the RTDA controller design as it offers prior multi-step ahead predictions to predict the future plant behaviour. It requires multiple trials to determine the optimal or near optimal values for the tuning parameters for the NN based RTDA controller design and hence a highly skilled ...
Intelligent Controllers for Conical Tank Process
2014
Level control is an important control objective in process industries. Determining the optimal controller is vital, as it result in precise control of liquid level in the conical tank. The conventional PID controllers are used which will not provide a satisfactory control for various operating conditions. To overcome these difficulties, an intelligent controller is to be proposed. The objective of this project is to implement an intelligent controller for conical tank process. A Fuzzy Logic, Fuzzy PI and Neural Network controllers are implemented. Each controller is constructed based on the data collected from the process. The optimal control is identified as the Neural Network controller based on the performance indices such as settling time and overshoot.
IJERT-Intelligent Controllers for Conical Tank Process
International Journal of Engineering Research and Technology (IJERT), 2014
https://www.ijert.org/intelligent-controllers-for-conical-tank-process https://www.ijert.org/research/intelligent-controllers-for-conical-tank-process-IJERTV3IS20586.pdf Level control is an important control objective in process industries. Determining the optimal controller is vital, as it result in precise control of liquid level in the conical tank. The conventional PID controllers are used which will not provide a satisfactory control for various operating conditions. To overcome these difficulties, an intelligent controller is to be proposed. The objective of this project is to implement an intelligent controller for conical tank process. A Fuzzy Logic, Fuzzy PI and Neural Network controllers are implemented. Each controller is constructed based on the data collected from the process. The optimal control is identified as the Neural Network controller based on the performance indices such as settling time and overshoot.
Comparative Analysis of ANN based Intelligent Controllers for Three Tank System
International Journal of Intelligent Systems and Applications, 2016
Three tank liquid level control system plays a significant role in process industries and its behavior is nonlinear in nature. Conventional PID controller generally does not work effectively for such systems. This paper deals with the design of three intelligent controllers namely model predictive, model reference and NARMA-L2 controllers based on artificial neural networks for a three tank level process. These controllers are simulated using MATLAB/SIMULINK. The performance indices of intelligent controllers are compared based on the time domain specifications. The performance of NN predictive controller shows superiority over other controllers in terms of settling time.
Multivariable predictive control of a pressurized tank using neural networks
Neural Computing and Applications, 2006
The behavior of a multivariable predictive control scheme based on neural networks applied to a model of a nonlinear multivariable real process, consisting of a pressurized tank is investigated in this paper. The neural scheme consists of three neural networks; the first is meant for the identification of plant parameters (identifier), the second one is for the prediction of future control errors (predictor) and the third one, based on the two previous, compute the control input to be applied to the plant (controller). The weights of the neural networks are updated on-line, using standard and dynamic backpropagation. The model of the nonlinear process is driven to an operation point and it is then controlled with the proposed neural control scheme, analyzing the maximum range over the neural control works properly.
Soft Computing Technique and Conventional Controller for Conical Tank Level Control
Indonesian Journal of Electrical Engineering and Informatics (IJEEI), 2016
In many process industries the control of liquid level is mandatory. But the control of nonlinear process is difficult. Many process industries use conical tanks because of its non linear shape contributes better drainage for solid mixtures, slurries and viscous liquids. So, control of conical tank level is a challenging task due to its non-linearity and continually varying cross-section. This is due to relationship between controlled variable level and manipulated variable flow rate, which has a square root relationship. The main objective is to execute the suitable controller for conical tank system to maintain the desired level. System identification of the non-linear process is done using black box modelling and found to be first order plus dead time (FOPDT) model. In this paper it is proposed to obtain the mathematical modelling of a conical tank system and to study the system using block diagram after that soft computing technique like fuzzy and conventional controller is also used for the comparison.
Neural network inverse model-based controller for the control of a steel pickling process
Computers & Chemical Engineering, 2005
The present work investigates the use of neural network direct inverse model-based control strategy (NNDIC) to control a steel pickling process. The process is challenging due to the fact that the pH of effluent streams must be regulated accurately to protect aquatic and human welfare, and to comply with limits imposed by legislation. At the same time, the concentration of acid solution in the pickling step needs to be maintained at the optimum value in order to obtain the maximum reaction rate. Various changes in the open-loop dynamics are performed before implementation of the inverse neural network modeling technique. The optimal neural network architectures are determined by the mean squared error (MSE) minimization technique. The robustness of the proposed inverse model neural network control strategy is investigated with respect to changes in disturbances, model mismatch and noise effects. Simulation results show the superiority of the NNDIC controller in the cases involving disturbance, model mismatch and noise while the conventional controller gives better results in the nominal case.
Artificial Neural Network Model-Based Predictive Real-Time Control Of A Cascaded Two Tank System
2019
The development of reliable first principle models that totally describe the dynamic behaviour of nonlinear systems is a difficult and time-consuming task. This poses a major challenge in the development of nonlinear model-based controllers for industrial processes. Hence, an alternative approach which involves the use of artificial neural network (ANN) models for real-time predictive control of a cascaded two tank system housed in our laboratory is explored in this research work. To achieve this, the tank process was excited by well-designed input signals within a specified range to obtain real-time input-output data at a sampling time of 2s. The datasets obtained were used to fit recurrent neural network (RNN) and feedforward neural network (FFNN) models for the process. Thereafter, the models were used in the design of predictive controllers. The designed controllers were compiled and deployed to an Arduino microcontroller interfaced with the process to achieve real-time control. Validation results showed both models have good fits. The closed loop experimental results also showed good setpoint tracking performance for both controllers.