Intelligent PID Controller for Real Time Automation of Microwave ...

More documents

Recommendations

Info

International Journal of Computer and Information Technology (ISSN: 2279 – 0764)Volume 02– Issue 04, July 2013Setpoint +(Tset) _Figure 9.errorGAP, I, D ParametersConventionalPID-ControlController outputReal Time MicrowaveBiodiesel ReactorANFIS Microwave BiodieselReactor ModelOnline updatingOnline errorGenetic tuning algorithm for PID parameters based on ANFISmodel_ANFIS output (T A)The GA starts with a random population being initialized.This can be easily achieved in LabVIEW using the randomnumber generator as shown in Fig.11. Three randompopulations each of the required size were initialized torepresent the proportional gain, derivative gain and the integralgain.The objective function is used to provide a measure of howindividuals have performed in the problem domain. In the caseof a minimization problem, the fit individuals will have thelowest numerical value of (E) which is given by (1). TheANFIS model is used to select the fittest individualscorresponding to the lowest values.III.CONTROLLER DESINGThe objective function used in this work is given in (1),(T set ) is the desired reactor temperature (set point), T A is theANFIS output temperature, and T Max , maximum ANFIS output.The GA routine is to minimize this fitness value of absoluteerror (E) as in Fig.10.Inlet oil flow rateInlet temperatureController outputPID ControllerP, I, D ParametersGeneticallyANFIS BiodieselModel-Out Temp.E=e+eo+essError? MinimumTA, TMAX, TSS (Steady state temperature)Figure 11.VI for initial random populationThe crossover operator is mainly responsible for the globalsearch property of the GA. The basic idea of the step is toproduce a new member in the population whose value isdetermined by two parents existing in the population.Crossover was implemented by randomly selecting two entriesand replacing the least fit of them by the weighted average ofthe two as shown in Fig.12.Figure 10.Fitness function block diagramE=e+e ss +e o (1)Wheree: error between the setpoint and ANFIS output.e=|T set -TA| (2)e ss : steady state errore ss =lim t e(t) (3)eo: overshoot errore o =|T set -T Max | (4)Figure 12.VI for crossoverT Max : Maximum ANFIS output temperaturewww.ijcit.com 812
After crossover is performed, mutation takes place. Themutation operator is used to inject new genetic material into thepopulation. Mutation randomly alters a variable with a smallprobability. The random number generator was used for bothrandom selection of a member from the current population andto replace it with a new random value as shown in Fig.13.International Journal of Computer and Information Technology (ISSN: 2279 – 0764)Volume 02– Issue 04, July 2013Figure 14.Intelligent PID controller at multiple setpoint trackingFigure 13.VI for mutationVI.EXPERIMENTAL RESULTThe PID controller is designed using LabVIEW. The inputof the controller represents the error between measuredtemperature and set point (required temperature). The output ofPID is the control signal applied to the biodiesel plant. Thecontroller working at (5s control sample interval) and(300ml/min flow rate). Fig.14 shows the controllerimplementation as follows:1- PID controller was tested to track the set pointbeginning from inlet oil temperature to requiredtemperature equal to 30°C.2- The controller was then subjected to a multiple setpoint tracking beginning from the nominaltemperature value of 30°C and changing up to35°C then down to 30°C.Figure 15. Comparison between intelligent PID controller andconventional PID controller (setpoint equal to 30°C)Fig.15 shows the controller comparison between thegenetically tuning PID controller and conventional PIDcontroller at set point tracking (required temperature equal to30°C). The disturbance was then introduced. The nominal feedin oil flow rate was reduced by 20% and 10% then rising by30% as shown in Fig.16.Figure 16. Disturbance rejections (change the oil flow rate -20%, -10%,+30%) for the Intelligent PID controllerwww.ijcit.com 813
Page 1 and 2: International Journal of Computer a
Page 3: International Journal of Computer a

Intelligent PID Controller for Real Time Automation of Microwave ...

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?