Rahul Malhotra Et Al IJCSET July 2011 Vol 1, Issue 6,315-319 Boiler .

Rahul Malhotra et alIJCSET July 2011 Vol 1, Issue 6,315-319Boiler Flow Control Using PID and FuzzyLogic ControllerRahul Malhotra1, Rajinder Sodhi2Department of Electronics & Communication Engineering1Bhai Maha Singh College of Engineering, Muktsar (PB)1Adesh Institute of Engineering & Technology, Faridkot (PB)1, 2non linear systems and is used for modeling complex systemswhere an inexact model exists or systems where ambiguity orvagueness is common. The fuzzy control systems are rulebased systems in which a set of fuzzy rules represent acontrol decision mechanism to adjust the effects of certainsystem stimuli. With an effective rule base, the fuzzy controlsystems can replace a skilled human operator. The rule basereflects the human expert knowledge, expressed as linguisticvariables, while the membership functions represent expertinterpretation of those variables.Abstract: Conventional Proportional Integral Controllers areused in many industrial applications due to their simplicity androbustness. The parameters of the various industrial processesare subjected to change due to change in the environment.These parameters may be categorized as steam, pressure,temperature of the industrial machinery in use. Various processcontrol techniques are being developed to control thesevariables. In this paper, the steam flow parameters of a boilerare controlled using conventional PID controller and thenoptimized using fuzzy logic controller. The comparative resultsshow the better results when fuzzy logic controller is used.Maximum overshoot for fuzzy logic controller is measured as9.35% as compared with 47.3% given by conventional PIDcontroller. Settling time for fuzzy logic controller and PIDcontroller is measured at 7.18 seconds and 10.14 secondsrespectively, which shows the superiority of fuzzy logiccontroller.II ler) is a generic control loop feedback mechanism(controller) widely used in industrial control systems – a PIDis the most commonly used feedback controller. A PIDcontroller calculates an "error" value as the differencebetween a measured process variable and a desired set point.The controller attempts to minimize the error by adjusting theprocess control inputs. In the absence of knowledge of theunderlying process, PID controllers are the best controllers.However, for best performance, the PID parameters used inthe calculation must be tuned according to the nature of thesystem – while the design is generic, the parameters dependon the specific system. The PID controller calculation(algorithm) involves three separate parameters, and isaccordingly sometimes called three-term control: theproportional, the integral and derivative values, denoted P, I,and D. The proportional value determines the reaction to thecurrent error, the integral value determines the reaction basedon the sum of recent errors, and the derivative valuedetermines the reaction based on the rate at which the errorhas been changing. The weighted sum of these three actionsis used to adjust the process via a control element such as theposition of a control valve or the power supply of a heatingelement. Heuristically, these values can be interpreted interms of time: P depends on the present error, I on theaccumulation of past errors, and D is a prediction of futureerrors, based on current rate of change. By tuning the threeconstants in the PID controller algorithm, the controller canprovide control action designed for specific processrequirements. The response of the controller can be describedin terms of the responsiveness of the controller to an error,the degree to which the controller overshoots the set pointand the degree of system oscillation. Note that the use of thePID algorithm for control does not guarantee optimal controlof the system or system stability. Some applications mayrequire using only one or two modes to provide theappropriate system control. This is achieved by setting thegain of undesired control outputs to zero. A PID controllerKeywords: boiler, fuzzy logic controller, PID controller.I INTRODUCTIONThe Proportional-Integral-Derivative (PID) controllers havebeen the most commonly used controller in process industriesfor over 50 years even though significant development havebeen made in advanced control theory. According to a surveyconducted by Japan Electric Measuring InstrumentManufacturers Association in 1989, 90 % of the control loopsin industries are of the PID type. The proportional actionadjusts controller output according to the size of the error, theintegral action eliminates the steady state offset and thefuture is anticipated via derivative action. These usefulfunctions are sufficient for a large number of processapplications and the transparency of the features lead to wideacceptance by the users. Strength of the PID controller is thatit also deals with important practical issues such as actuatorsaturation and integrator windup. PID controllers performwell for a wide class of processes and they give robustperformance for a wide range of operating conditions and areeasy to implement using analog or digital hardware.Moreover, due to process uncertainties, a more sophisticatedcontrol scheme is not necessarily more efficient than a welltuned PID controller.The concept of intelligent control lies with the fact thathuman intelligence is imbibed in to the controller architectureso that human behavior can be emulated in the controldecision. Human expert knowledge is based upon heuristicinformation gained in relation to the operation of the plant orprocess, and its inherent vagueness ("fuzziness") offers apowerful tool for the modeling of complex systems. Thefuzzy logic controller provides an algorithm, which convertsthe expert knowledge into an automatic control strategy.Fuzzy logic is capable of handling approximate informationin a systematic way and therefore it is suited for controlling315

Rahul Malhotra et alIJCSET July 2011 Vol 1, Issue 6,315-319will be called a PI, PD, P or I controller in the absence of therespective control actions. PI controllers are fairly common,since derivative action is sensitive to measurement noise,whereas the absence of an integral value may prevent thesystem from reaching its target value due to the controlaction.III FUZZY LOGIC BASED CONTROLLERFuzzy controllers are very simple conceptually. They consistof an input stage, a processing stage, and an output stage. Theinput stage maps sensor or other inputs, such as switches,thumbwheels, and so on, to the appropriate membershipfunctions and truth values. The processing stage invokes eachappropriate rule and generates a result for each, thencombines the results of the rules. Finally, the output stageconverts the combined result back into a specific controloutput value. The most common shape of membershipfunctions is triangular, although trapezoidal and bell curvesare also used, but the shape is generally less important thanthe number of curves and their placement. As discussedearlier, the processing stage is based on a collection of logicrules in the form of IF-THEN statements, where the IF part iscalled the "antecedent" and the THEN part is called the"consequent". Typical fuzzy control systems have dozens ofrules. Consider a rule for a thermostat: IF (temperature is"cold") THEN (heater is "high").Figure 2: Frequency domain analysis of the systemVI. PID CONTROLLER DESIGN AND TUNINGA feedback control system measures the output variable andsends the control signal to the controller. The controllercompares the value of the output signal with a referencevalue and gives the control signal to the final control elementvia the actuator.The characteristic equation obtained as belows 3 6 s 2 5s K cu 0Applying Routh criteria in eq (1) we get Kcu 30From auxiliary equation in routh criteria we get ω 2.03 andT 2.69TheequationofidealPIDcontrollerisIV PROBLEM FORMULATIONA boiler of a chemical plant is taken as a case study and thetemperature control of the boiler is achieved usingconventional PID controller and intelligent fuzzy logic basedcontroller. The comparison of both the controllerperformance is analyzed in this chapter.Set pointSet point of temperature 380 degree Celsius.t 1de(t ) u (t ) K c e(t ) e(t )dt d dt i 0 1 d s e( s )u ( s) K c 1 is 1 i s i d s 2 u ( s) K c e( s ) is V. MATHEMATICAL MODELING & CONTROLLERDESIGNThe basic conventional feedback controller is shown in figure1. In conventional PID controller the controller and theprocess are in series where as a feedback from the output isgiven to the input. The boiler of chemical plant ismathematically modeled using experimental data availableand the transfer function of the above system is achieved asPIDTherealPIDcontrolleris 1 i s 1 d s u ( s) K c e( s ) i s 1 d s The PID controller is traditionally suitable for second andlower order systems. It can also be used for higher orderplants with dominant second order behaviour. The ZieglerNichols (Z-N) methods rely on open-loop step response orclosed-loop frequency response tests. A PID controller istuned according to a table based on the process response test.According to Zeigler-Nichols frequency response tuningcriteriaK p 0.6 K cu , i 0.5T and d 0.125TProcessr(s)(1)c(s)For the PID controller in the heat exchanger, the values oftuning parameters obtained are Kp 32, τi 1.5, τd 0.29 andP 30, I 21.2, D 9Usually, initial design values of PID controller obtained byall means needs to be adjusted repeatedly through computersimulations until the closed loop system performs orcompromises as desired. This stimulates the development of“intelligent” tools that can assist the engineers to achieve thebest overall PID control for entire operating envelops.Figure 1: Block diagram of classical control architectureThe stability analysis of the system is done and the bode plotof the system is plotted which is shown in figure 3. The gainmargin is 20 db where as the phase margin is 56.2 .316