A PID Controller Design for an Air Blower System

Transcription

1 1 st International Conference of Recent Trends in Information and Communication Technologies A PID Controller Design for an Air Blower System Ibrahim Mohd Alsofyani *, Mohd Fuaad Rahmat, and Sajjad A. Anbaran Faculty of Electrical Engineering, Universiti Teknologi Malaysia UTM, Skudai 81310,Johor,Malaysia Abstract In this paper, PID controller is designed and applied on a nonlinear air blower system PT-326. Model of the system is estimated by using System Identification Toolbox in Matlab. This process began with collection of input and output data from experimental works. The data collected is used for model estimation. Auto-regressive with exogenous input (ARX) model is chosen as a model structure of the system. Based on the input and output data of the system, best fit criterion and correlation analysis of the residual is analyze to determine the adequate model for representing the PT-326 system. By using Ziegler-Nichols (ZN) tuning method and minimum square error (MSE) scheme, PID controller is designed for the model chosen through simulation in Simulink. In order to verify this controller, it is applied to the real time system and the performance of the system is monitored. The results obtained from both tuning methods show that the output of the system with the PID controller in simulation mode and experimental work is almost similar. The output of the system also tracked the input given successfully Keywords. PID; Air blower system 1 Introduction Temperature is an essential control variable like flow rate and motor velocity in thermal machines. For industrial applications, temperature needs to be finely controlled with consideration of equipment safety [1-4].The air blower system is a common process in our daily life where certain desired temperature is controlled. In industries such as pharmaceutical, ability to control temperature is crucial to ensure the quality of the product always within control. However, most of heating plants are complex with higher-order systems, which leads to unsatisfactory performance. The PT-326 air blower system is selected as a model system which needs to be maintained at a certain level of temperature. Therefore, model system has to be controlled by a suitable controller to achieve its desired temperature [5]. * Corresponding author: alsofyani@yahoo.com IRICT 2014 Proceeding 12 th -14 th September, 2014, Universiti Teknologi Malaysia, Johor, Malaysia

2 Ibrahim Mohd Alsofyani et. al. /IRICT (2014) In order to acquire the highest performance of the air blower system, a suitable controller needs to be designed. The controller design requires the best mathematical model of the system under control. Thus, a method of identifying the system needs to be chosen so that the best accuracy of the model can be obtained. Model identification of air blower system based on hardware-in the-loop simulation environment of Real-time Workshop(RTW) and system identification toolbox in Matlab has been proposed [5],[6]. Nonlinear hybrid controller composed of a proportional controller, a fuzzy controller and a classical PID controller for the model attained has been introduced. Similar work of system identification has been done in [7]. Experimental work on recursive identification represented by a discrete-time model in open-loop and closed-loop configurations is presented [8]. The unknown parameters of the system based on ARX model is estimated by using Recursive Least Square (RLS) method while model validation is verified by residual analysis. The PID controller parameters for a non-linear process quickly. To design a very efficient controller with high quality system performance, the system must be modeled in a proper way. The unknown system which has unknown parameters is called a black-box model. The mathematical modeling of this blackbox model system can be obtained using System Identification (SI) technique. The overall step of system identification procedure can be found in [9]. Only experimental approach is considered in this paper where the system model is referred as a black-box model (Section II). Tuning of PID controller parameters is done by suing minimum square error and Ziegler Nichols tuning methods (Section III). Closed loop simulation and performance analysis is included in this paper through Matlab simulation (Section III) and online implementation using Real-time Windows Target toolbox (Section IV). Finally, discussion and conclusion are drawn. 2 EXPERIMENTAL AND SIMULATION SETUP In this study, PT326 is used as the model system. It models common industrial situations in which temperature control is required. The process contained in the PT 326 involves air that is drawn from the atmosphere by a centrifugal blower, and is heated as it passes over a heater grid before being released into the atmosphere through a duct. The control objective is to maintain the temperature of the air at a desired level. Fig.1 shows the front panel of the PT 326 apparatus. Mathematical modeling is a description of a system in terms of equations. It can be divided into two parts; physical modeling and system identification. In this research, system identification technique is applied to attain the model of the system. Pseudo Random

3 Ibrahim Mohd Alsofyani et. al. /IRICT (2014) Binary Sequence (PRBS) is perturbed into the system through Simulink to collect temperature data using a bead thermistor placed in the flow at any of three positions along the duct. The collected input and output data is stored in workspace Matlab. This data is used for model estimation and validation part. Validation process is done in order to compare the estimated model output with the real output from the experiments [9]. By looking at the best fit criterion parameter, the validated model will be accepted [8]. Open-loop testing is done in simulation mode by inject step and sine input to the model obtained so that a suitable controller can be designed to improve the performance of the system. Fig.1. The air blower system (PT-326) PID controller is designed by inserting the calculated parameters in PID block in Simulink and examined the output result. In simulation mode, the PID controller is connected to the discrete transfer function model in Simulink block only. The output response is observed and recorded. It is followed by inserted the similar PID in a real-time system where it is located in the forward path of real-system.. A. System Identification Initially, system model must be determined before control technique is applied. The system modeling part is the most challenging and vital part in designing the control system of PT-326 [8]. In order to obtain a particular model for this system, the open loop identification experiment has been done using parametric approach. In this experiment, a system model is identified using data collected when the Pseudo Random Binary Sequence (PRBS) is perturbed into the system. From Fig. 2, there are 1023 samples of data with 0.07 seconds sampling interval. The PRBS input is generated in Matlab. The collection of data was performed by PCI-1711 interface card. The input-output data is then be analyzed by System

4 Ibrahim Mohd Alsofyani et. al. /IRICT (2014) Identification toolbox in Matlab [5]. From the set of input-output data in Fig. 2, it was divided into two parts. The first part is the training data and the second is for testing or validation data. Fig. 2. The input-output data set In this paper, the PT-326 system is modeled based on Autoregressive with exogenous input (ARX) model structure with sixth order. The best fit of output model is 93.42% as depicted in Fig.3. Its polynomial structure can be written asresults and Discussion Results of the experiments should be described and discussed in this section. A ( q ) y ( t ) B ( q ) u ( t ) e ( t ) 9 8 A ( z ) z z z z z (1) (2) B ( z ) z z z (3) Then, Loss function = and Akaike s Final Prediction Error (FPE) = Therefore, the air blower system PT-326 plant can be approximated modeled by this following equation: B ( q ) A ( q ) z z z z z z z z 5 (4)

5 Ibrahim Mohd Alsofyani et. al. /IRICT (2014) Meas Esti- Fig. 3. Measured and simulated model output Hence, based on this approximated plant model, conventional PID controller is tuned by MSE and ZN. The approximated plant gives a higher order model where an excess model order is usually represent the noise. Since the ARX model incorporate with noise in the system model, the model might be influenced by this noise [5]. Next, by observing the pole-zero plot of the model, all the zeros are inside the unit circle of the z-domain as shown in Fig.4. This is called minimum phase model. Since the system is assumed to be stable, all the poles will be inside the circle Fig. 4. Pole-zero plot

6 Ibrahim Mohd Alsofyani et. al. /IRICT (2014) B. System Identification In this study, the tuning values of K p, K i, K d are determined by using the conventional Ziegler-Nichols (ZN) tuning method and based on minimum square error (MSE) tuning scheme.for MSE, the performance measure that is used in this case is the Integral Square Error (ISE) given by: ISE = ((y d (t) - y(t)) 2 dt (5) where yd is the desired output (set point) while y is the actual output. This criterion, although not very selective, has been used because of the ease of computing the integral both analytically and experimentally. The MSE procedure used to optimize the controller parameters is summarized as follows: 1. Define the input design space, D, which consists of a set of initial values of the controller parameters 2. Obtain the ISE for the temperature for all the design space defined in TABLE I.. 3. Create the target data set, T, which consists of the normalized ISE for the temperture. 4. Increment the data corresponding for each PID parameter. 5. Find the minimum error of the output (estimated). The corresponding controller gains that minimized the output will be the gains to be verified in actual model simulation. 6. Repeat steps 1 to 6.if the controller parameter gains are not satisfactory. In this case, D is the set of discrete values given in Table I. Table 1. Controller Parameters Used For Simulations Initial and Large Data Sets D Kp Ki Kd {0.01,0.02,.,0.2} {0.1,0.2,...,0.7} {0.001,0.002,,0.01} 3 CLOSED- LOOP SIMULATION AND PERFORMANCE ANALYSIS Before the real process implementation, a simulation is carried out for each PID tuning method to verify the proposed controller designs. The aim of simulation is to give emphasis to the tuning of the conventional proportional-integral-derivative (PID) controller using the minimum square error and Ziegler Nichols methods. To insure stability, only closed loop controller is considered in this control system. The step input is applied to the system as a reference input with set point of 33.

7 Ibrahim Mohd Alsofyani et. al. /IRICT (2014) Fig. 5. Simulink block of the system and PID controller Fig.5 shows the Simulink block diagram with PID controller. The output response before implementing the PID controller can be represented in Fig.6. After tuning the PID using the Ziegler Nichols approach, the performance of the output response can be seen in Fig.7. From Fig.7, the overshoot of the system output is quite high with 7 seconds settling time. Even though the PID controller is widely used in industrial process, the tuning of PID parameters is a crucial issue in particular for the system s characteristic which has large time delay and high order system [9]. Commonly in industrial process, only an expert or experienced workers are able to monitor and tune the PID parameters based on their experience. Fig.6. Simulation process response without PID controller

8 Ibrahim Mohd Alsofyani et. al. /IRICT (2014) Fig.7. Simulation process response tuned by PID controller using ZN Therefore, in certain cases where there is deficient of experience with the processes, it is sometimes quite impossible to achieve a satisfactory performance. For these reason, it is desirable to introduce tuning methods for the PID controller. The minimum square error (MSE) is the PID tuning approach used as an alternative to the Ziegler Nichols (ZN) tuning method Fig.8 shows the output response of the PID controller using the MSE tuning method with no overshoot. Although the output response has no overshoot, this approach takes a longer settling time (7.7seconds) to accomplish the steady state. Fig.8. Simulation process response using MSE

9 Ibrahim Mohd Alsofyani et. al. /IRICT (2014) CONTROLLER IMPLEMENTATION IN THE REAL PT-326 SYSTEM In the previous section, two tuning PID controller methods have been designed via simulation. However, it was not enough to ensure that all the design controllers are exactly capable to control the real PT-326 system model until these tuning methods were implemented to perform on-line. This real system implementation is done using Real Time Windows Target (RTWT) toolbox in Matlab. Two blocks called Analog Output and Analog Input from RTWT connect the Simulink Matlab to the PT326 plant using data acquisition (DAQ) card PCI The controller will respond to the online process with 0.07 s sampling interval. The output of the controller will be fed into the Analog Output and the process output is generated from the Analog Input. Since only voltage is applicable in this RTWT toolbox, the output from the Analog Output need to be converted into temperature by multiply with constant, 3.3 as given in the previous section. The Simulink block diagram of the system with PID controller is represented in Fig.9. The system response before implementing the PID controller can be seen in Fig. 10. The system responses from both tuning PID methods; MSE and ZN are shown in Fig. 11, and Fig. 12, respectively. However, to satisfy the output, tuning parameter requires a little adjustment since the simulation tuning parameter is designed based on the approximated plant. Fig.9. Simulink block diagram of real plant implementation with PID

10 Ibrahim Mohd Alsofyani et. al. /IRICT (2014) Fig.10. Experimental process response from experiment without PID Fig.11 Experimental temperature process response using PID controller tuned by ZN

11 Ibrahim Mohd Alsofyani et. al. /IRICT (2014) Fig.12. Experimental temperature process response using PID controller tuned by MSE". By comparing the Fig. 11 and Fig. 12, it shows that the MSE and ZN tuning methods are capable to provide almost similar results as in the simulation. However, the MSE tuning provides no overshoot with almost similar settling time to the simulation. 5 Conclusion System identification technique has been successfully applied to the air blower system in order to produce the best linear discrete model of the system. PID controller is designed effectively for the system and applied in both simulation and real-time mode. The values of K p, K i and K d is determined by the Ziegler-Nichols and minimum square error tuning methods. The step input is injected to the system and the simulation result shows that the output tracked the input. The real-time experiments also proved the result where the output obtained is almost similar with the output response from simulation mode. References 1. Rehan M., Tahir F., Iqbal N. and Mustafa G., Modeling, simulation and decentral ized control of a nonlinear coupled three tank system, Proceedings of IEEE ICEE conference, Lahore, Pakistan, (2008) Chen H. Y. and Huang S. J., Adaptive neural network controller for the molten steel level control of strip casting processes, Journal of Mechanical Science and Technology, 24 (3) (2010)

12 Ibrahim Mohd Alsofyani et. al. /IRICT (2014) Park I., Heat transfer analysis during a curing process for UV nanoimprint lithography, Journal of Mechanical Science and Technology, 23 (4) (2009) Rehan M. and Iqbal N., Decentralized robust control of a MIMO system using parametric and non-diagonal interaction uncertainty modeling, Proceedings of IEEE ICEE conference, Lahore, Pakistan, (2008), Rahmat M.F., Mohd Subha N.A., K. Jusoff and N AbdulWahab, Fuzzy Logic Controller Design for a Small Scale Industrial Hot Air Blower Heating and Ventilation System, World Applied Science Journal (WASJ), Volume 9 Issue (10): page , ISSN , IDOSI Publications, Rahmat MF, Yeoh KH, Usman S and Abdul Wahab N, Modelling of PT326 Hot Air Blower Trainer Kit Using PRBS Signal and Cross Correlation Technique, Jurnal Teknologi-D, UTM Publisher, June 2005, volume 42, pp Rozali S Md, Rahmat MF, Zulfatman and Ghazali R, PID Controller Design for an Industrial Hydraulic Actuator with Servo System, IEEE SCOReD 2010, 8. Zulfatman, On-line Identification of an Electro-hydraulic System using Recursive Least Square, IEEE Conference on Robotics, Automation and Mechatronics, 2010, pp Lennart Anderson, U. J., Karl Henrik Johansson, and Johan Bengtsson "A manual for system identification."