Pradeep Kumar Juneja

Effect of dead time approximation on controller performance designed for a second order delayed model

The process models for many industrial processes can be represented by a Second Order Plus Dead Time system. In the present analysis, the selected Second Order plus Dead Time model is converted in to First Order plus Dead Time model using Taylors Series and Skogestad dead time approximation techniques. Controllers have been designed using Internal Model Control method, with various values of desirable closed loop time constants. A comparative study has been performed for closed loop responses for the original and the dead time approximated systems.

Controller design and its performance analysis for a delayed process model

The Dynamics of most of the industrial processes exhibits delay in a first order process model commonly known as first order plus dead time process model. In the present investigation, a second order plus dead time process model is converted into first order plus dead time process model with two model order reduction techniques namely Skogestads and Taylors series approximation and open loop responses have been compared. PID controllers for both the FOPDT process models have been designed and compared using Ziegler Nichols, Chein Hrones Reswick and Wang Juang Chan tuning techniques.

Effect of implementing different PID algorithms on controllers designed for SOPDT process

The most important control objective in a control system analysis is to design an optimal controller for the known process model. In the present analysis, various PID controllers based on different tuning methods have been designed for a selected Second Order plus Dead Time model. Also different algorithms of PID controllers such as parallel, parallel with derivative filter, series, series with derivative filter and cascade form are implemented. The steady state and dynamic characteristics of closed loop responses for the designed controllers are compared and analyzed.

Comparative analysis of controllers designed for pure integral complex delayed process model

In present analysis, controllers based on different tuning techniques for a pure integral process with time delay are designed for different approximation of the time delay. The controllers are compared for the approximated process for the set point tracking capability. The set point tracking capability is determined on the basis of transient as well as the steady state analysis of a step response.

Neuro-Fuzzy Controller Design for MIMO Boiler Turbine Process

The plant selected for control system analysis is a 2 × 2 MIMO system with high multivariable interaction having two manipulated variables, the position of the governor valve and the fuel flow rate and two controlled variables, generated electric power and the steam pressure. A neuro-fuzzy controller has been designed using ANFIS in MATLAB for a boiler turbine plant. This technique is a hybrid of neural network and fuzzy logic techniques.

Dead Time Compensation in Sugar Crystallization Process

The input to the sugar factory is the sugarcane billets and the output is the crystal sugar. There are many subprocesses with significant multivariable interaction involved within this process which is cane preparation, juice extraction by crushing mill, heating, clarification and filtration, evaporation, and crystallization. In the present paper, the Smith predictor is designed using MATLAB in order to compensate the dead time present in heat exchanger system of the crystallization process and its performance is compared to that of a conventional PI controller with no dead time compensation.

Model Order Reduction of a Higher Order Model of pH Neutralizer of Sugar Mill

Several industrial chemical processes are found tobe FOPDT (first order plus dead time). The presence of deadtime makes the transfer function to be irrational and makesthe simulation and controller design very complex. In thepresent work an FOPDT model of the pH neutralizationprocess which is found in sugar mills is considered as a subjectof study and analysis. Having obtained its Padéapproximation, its order has been reduced using Square RootBalance Truncation Method and then for this reduced ordermodel, Ziegler-Nichols(ZN), Internal model control (IMC)and optimized IMC using Genetic Algorithm(GA-IMC)based controllers are designed and their performance is compared.

Rationalizing the Mathematical Model of Boiler Turbine Process with Dead Time

More or less every chemical industrial process suffers from some amount of inherent time delay which is highly undesirable. The presence of this dead time term makes the transfer function model irrational. In the present work, a 2x2 transfer function model of an industrial boiler turbine process with dead time is considered, which after decoupling has been rationalized using Padé approximation. The first, second and third order approximants are determined and validated by comparing open loop step responses.

Design and Analysis of Control Systems for Heat Exchanger System of Sugar Mill

In sugar factory, after clarification andevaporation, the extracted juice from the crushing mill issubjected to the crystallization process. Mass transfer andevaporation process are involved in the crystallizationprocess. Vacuum keeps the temperature at low value forminimizing the formation of colour and sucrose inversion. In the present work, the heat exchanger system, deployed incrystallization process of sugar manufacturing industry, isconsidered and control system is designed for it. For controlsystem design, Ziegler-Nichols (ZN), Internal model control(IMC) and Linear quadratic Gaussian (LQG) tuning methodsare adopted using MATLAB and finally their controlperformances are compared.

Design of modified Smith predictor for dead time compensation for SOPDT process

Process industry many time used second order plus dead time process as an industrial process. Time taken between input and output of a phenomena in an industry is called dead time also known as transportation lag. Smith predictor is a dead time compensation technique which is used to compensate dead time. Smith predictor compare process model and actual process of the system until error between them zero, but there are some limitations of smith predictor if actual process and process model is not match. Modified Smith predictor is used to achieve stability and zero steady state error to step load response.