SOME STUDIES ON ARTIFICIAL INTELLIGENCE BASED FRACTIONAL ORDER PID CONTROLLER

Abstract

Process control in industry is improving gradually with the innovations and implementation of new technology. Different control techniques are being used for process control. Proportional Integral and Derivative (PID) controller is employed in every facet of industrial automation. In any of control application, controller design is the most important part. There are different types of controller architectures available in control literature. The applications of PID controller span from small industries to high technology industries. Designing a PID controller to meet gain and phase margin specifications is a well-known design technique. If the gain and phase margin are not specified carefully, then the design may not be optimum in the sense that a large phase margin (more robust) that could give better performance. This research outlines the development and design of an infrared radiation heating profile controller. An attempt has been made to theoretically analyze the system, design of the Controller, their simulation, and real-time implementation of an infrared ceramic heating profile controller. The Controller has been subjected to comparative testing with a proportional control model to observe its performance and validate its effectiveness. PID controllers of this nature that are commercially available either lack the functionality of this unit or are too expensive to implement for research purposes. This unit has been designed with cost-effectiveness in mind but still meets the standards required for an industrial controller. Heating profiles are necessary and useful tools for the proper processing of a host of materials. The Controller developed in this research is able to meet a level of a fair degree of accuracy and track a heating profile. The results confirm that this programmable control model will be advantageous and a valuable tool in temperature regulation. This means that intensive studies into the effects of infrared radiation on materials are now feasible. Research of this nature could possibly expand the application of infrared as a heating mechanism. Although tests were conducted on this Controller, they are not meant to serve as an exhaustive analysis. The conclusions of these simulations do reveal the benefit of such a v controller. More rigorous investigation is suggested as a subject for further study. System identification of this nonlinear process is done using black box model, which is identified to be nonlinear and approximated to be a First Order plus Dead Time (FOPDT) model. In order to obtain an accurate mathematical expression of the IR heater used in this research, a step response test of the IR heater has been completed. This method of testing has been done in accordance with the Ziegler-Nichols, Astrom Hügglund, and Cohen-Coon methods. Simulation of the obtained transfer function, using Mat Lab software, showed good agreement. Although the transfer function represented a first-order model with transportation lag, the simulated results reflected an acceptable accuracy. An exhaustive study has been done on different PID controller tuning techniques. The PID controller of the model has been designed using the classical method, and the results have been analyzed. A compromise has been made between robustness and tracking performance of the system in the presence of time delay. The results of the simulation indicate the validity of the study. Integer order PID controller (IOPID) based on Bode plot and Nyquist plot has been designed. The results illustrate that the IOPID controllers have the capability of minimizing the control objectives better than the previously designed controllers (Ziegler-Nichols, Astrom-Hügglund, and Cohen-Coon method-based Controller).With the change in temperature occurs, the oscillations of the controlled system outputs are eliminated and the output steady state errors become very small. The results demonstrate that the IOPID controller is stable and it suppresses the cost function (Maximum overshoot, Rise Time, Settling time and Peak time) even in case of significant disturbances. IOPID controller has also been designed using Bode plot and Nyquist plot for high gain system. The results have shown that the responses of Ceramic IR heater temperature profile have been reduced to very small value and prove that the IOPID controller is still stable and it suppresses the cost function even if significant disturbances have occurred. Fuzzy Logic controller-based model reference has been designed. Its implementation indicates that the proposed Controller suppresses the output of the controlled system. The results illustrate that the proposed Controller only slightly vi improves the performance of the cost function. The various AI techniques (GA) and Soft Computing (bio inspired) based algorithm (BFO, ACO) for PID controller offers several advantages. These methods can be used for higher order process models in complex problems. Approximations that are typical to classical tuning rules are not needed. Compared to conventionally tuned system, GA, PSO, BFO and ACO tuned system provides good steady state response and performance indices. The genetic approaches can achieve better temperature control with smaller settling time, overshoot and undershoot, and zero steady error. The control signal changes more frequently and with larger magnitude as the genetic algorithms are stochastic in nature. The PSO has an additional unique advantage that it adapts any change in system conditions, and obtains different system dynamics accurately in a short time period. It is a random search method but if combined with an artificial intelligence features, it tracks required system dynamics accurately in short time (small number of iterations). The BFO based Controller has the advantage of a better closed loop time constant, which enables the Controller to act faster with a balanced overshoot and settling time. The response of the conventional Controller is more sluggish than the BFO based Controller. Compared to conventionally tuned system, BFO tuned system has better steady state response and performance indices. Ant- Colony algorithm (ACO) has no special requirements on the characteristics of optimal designing problems, which has a fairly good universal adaptability and a reliable operation of program with ability of global convergence. Simulation results show the controlled system has satisfactory response and the proposed method has an effective tuning strategy. ACO shows better performance for PID controller parameter tuning of the considered control system. The simulation results show that the proposed method achieve minimum tracking error and estimate the parameter values with high accuracy. The work presents tuning method for fractional order proportional Integral and derivative controllers (FOPID) for the first order plus time delay (FOPTD) class of systems based on gain and phase margin. Techniques such as fractional order PID controller design and the results of their application to real-world system vii have been presented. A comparative study has been done using different control techniques to analyze the performance of different controllers. First, the conventional PID controller is implemented as primary Controller. The performance of PID, IOPID, Fuzzy Logic Controller, and Artificial Intelligence based PID, Bio inspired based PID controller and FOPID controllers have been examined. It has been concluded that the overall performance of the FOPID-based Controller is better than other controllers. In real-time implementation, the performance of the process control includes the time required by the heater to be settled on the initial set-up temperature. The rising of temperature is slow due to the resistance heating element used in ceramic infrared heaters. So the settling time is very high. The results obtained by simulation and real-time implementation with fractional order PID controller show overall better performance( rise time , settling time, peak time and peak overshoot) in comparison with other designed and implemented Controllers embedded with ceramic infrared systems. Further stability problems of fractional order system with leakage delay and distributed delay with hybrid feedback controller have been solved (with examples) using the Mittag-Leffler function and Lyapunov direct method and proved Global Mittag-Leffler stability of fractional order system of the proposed model which implies faster convergence rate of the network model which represents the stability of the system. This work performs a small-scale test measuring controller performance so that it serves as a platform for future research efforts leading to the real-life implementation of a Ceramic Infrared Heater Temperature control system.