TMCnet News
Study on Intelligent Temperature Controller Based on Expert Controlling [Sensors & Transducers (Canada)](Sensors & Transducers (Canada) Via Acquire Media NewsEdge) Abstract: This paper summarizes and analyses the problems existing in the tobacco leaf curing process: lag and Low accuracy of temperature control. The retrofit design of the temperature control system has been completed according to the expert control principle of the intelligent control. The design of the intelligent temperature controller has been done. Copyright © 2014 IFSA Publishing, S. L. Keywords: Expert controlling, Temperature, Intelligent controlling. 1. Introduction Automatic control technology has been widely used in industrial and agricultural production in our country. For example, in the process of tobacco baking, the automatic control of the temperature and humidity in the bam solves the difficult in traditional baking process in which the temperature and humidity is not easy to control and the labor intensity is big. For all this, temperature regulation of Automatic control was lag. High precision temperature control in existence using a variety of improved PID algorithm has the advantage of high accuracy, less overshoot. But the adjustment time of which is long and temperature rising is slowly. That still failed to make the various indicators of system reach the approving level, which is more used in the thermostat, only in water or oil as medium [1]. Intelligent control is a new kind of control method developed in recent years. It is suitable for uncertain production process which containing the complexity and incompleteness and vagueness of the known algorithms. Tobacco baking is an uncertain, nonlinear, time-varying distributed parameter system [2]. The expert control method in the intelligent control can achieve good results. This topic research the intelligent temperature controller based on expert control can make a real time adjustment and control of tobacco baking temperature based on tobacco experts provide baking process and the real-time detection of curing bam. 2. Expert Control (EC) and Fuzzy Control (FC) The main methods of intelligent control include Expert Control and Fuzzy Control and Neural Networks Control and HIC-Hierarchical Intelligent Control. Fuzzy Control is common method of intelligent control. But Fuzzy controller is based on fuzzy mies which are selected imperfectly because there has been not comprehensive to controlled process [3]. So the Fuzzy controller still cannot fully meet the requirements of flue-cured tobacco process. The combination of expert system and traditional control theory led to the expert control method. Expert system is a kind of intelligent computer program that contains knowledge and reasoning, which contains a certain level of domain experts' knowledge and experience, has the ability to solve specific problems. Expert Control also called the Expert intelligent Control is the combination of theory and technology of Expert system with Control theory, method and technology. It can realize the control of the system imitating the Expert's experience in unknown environments. Expert control is trying to "join" an experienced engineer on the basis of the traditional control to realize the function of control. Its main body is made of knowledge base and reasoning mechanism. Its reasoning output is based on the appropriate rules According to a certain strategy and acquisition and organization to domain knowledge (Prior experience, dynamic information, target, etc) [4]. Intelligent controller is made of the database, knowledge base, reasoning machine, learning link, the fixed link and the board, its structure diagram is shown in Fig. 1. Database stores all the moments of sampling values, deviation, control measure and its control rules, weighting coefficient of each of the control mies etc. Knowledge base is used to store the expertise provided by experts of domain. It is a collection or warehouse used to storage orderly knowledge. Reasoning mechanism is designed to complete functions of reasoning and decision-making of system. The blackboard describes the intermediate state of solving process of the problem and act as an information transmitter of every tacit knowledge. Self learning and correction link is responsible for supplement or modify the content of knowledge base according to the information online. 3. The Applications of Intelligent Controller Based on Expert Control Application of intelligent control to tobacco baking machinery "join" a control of an experienced engineer on the basis of the traditional control. 3.1 The Hardware Design of System Curing bam is chose of brick and tile civil curing bam that is now widely used also known as natural ventilation and air downward curing bam (hereinafter referred to as air downward curing bam). Adding a heat exchanger and hot air circulation device built the forced ventilation and hot air circulation system. The control objective is baking process provided by the expert. System consists of intelligent controller, temperature and humidity sensor, hot air circulation fan, combustion-supporting blower, the hot wind tunnel ventilation fan; skylight wet fan, etc. The C8051F310 MCU is as the core of the system (Fig.2). The DS1820 temperature sensors made by DALLAS Company complete the task for temperature data collection and conversion (Fig. 3). At the same time, the DMB12864B-01 n/W type intelligent LCD display is chose to make a real time monitoring of temperature change in curing bam. According to the best temperature control curve in the process of tobacco baking, using expert control method to control coal and air supply. Temperature sensor and humidity sensor connect the system through the microcontroller I/O port directly and send the measured temperature to the MCU. MCU send the measured temperature to the keyboard and display unit through data storage. Contrast the data with preset temperature. Sound and light alarm will be driven by MCU when the temperature beyond the upper and lower limit. And then the control coal or exhaust motor switch will be driven and act. Core circuit of intelligent controller include the single chip microcomputer communication module, peripheral temperature detection module, LCD display and keyboard module and motor drive module, etc. 3.2. The Software Design of System The software design of intelligent temperature controller can be divided into the monitoring of the main program, temperature measurement subroutine and display subroutine, control calculate way program, etc. The working process of the system is the following. Open after the power supply, SCM has formed a three-step baking process curve, if need to modify the next steps can be performed; If need to modify the each period according to the LCD display on the prompt modification temperature and time value. Then start running key, the timer TO and T1 and serial communication. TO is mainly used for single chip microcomputer, the timing of T1 is mainly used to control the sampling frequency. T1 data overflow began to adjust temperature measurement subroutine, compare temperature after send display subroutines and call control calculate way program, the output control amount. Curing bam is a nonlinear, time-varying distributed parameter system. It apparently is very difficult said its characteristics with accurate mathematical analytical formula. So it has limitations to try to use quantitative mathematical method to control this system, Control calculate way program should consider the combination of qualitative and quantitative methods. Intelligent control is just to use qualitative experience, learning the merits of logical reasoning ability and memory. It fully learned math and automatic control theory, combining qualitative and quantitative knowledge, in order to meet the control requirements [4]. A bam temperature closed loop control structure is shown in Fig.5. Hereyr (/) is the designed value of temperature; y(0 is the adjustment amount; «(/) is the controlled quantity; e(/) is the deviation rate of change defined by the following formula: e(t)=yAt)~y(t)' [yr(0> y(t), e(t), e(t) are the states of the observable space. Transition curve of y(/) was shown as the Fig. 6, which was obtained when entered yT ( t ) into the system at the step point. With the help of people empirical knowledge of temperature control, adjustment process of y ( t ) is as follows. Greatly reduce the control volume w(i) when y ( t ) is greater than the yT ( t ) ; Greatly increase the control volume w(i) when y(t) is far less thanyz (í) ; Determine the size of the control volume according to the variation tendency of the y(t) ; Increase the control volume u[t) if y(t) < y ( í ) and change of y(t) is stable; Increase, maintain or decrease control by considering deviation and the deviation change rate when change of y(í) has a tendency to decrease the deviation (BC curve) if y(í)<y (í) ; Greatly reduce the control volume u(t) when change of y(í) has a tendency to increase the deviation (CD curve) if y(t)>y ( í ) ; Reduce the control volume U(t) when change of y(¿) is stable (DE curve) if y(t) >y ( Í ) ; Increase, maintain or decrease control by considering deviation and the deviation change rate when change of y ( t ) has a tendency to decrease the deviation (EF curve) if y ( í ) > y ( Í ) ; Greatly increase the control volume w(i) when y(í) has a tendency to increase the deviation (FG curve) if y ( í ) < y (i). K(0 ét(i) ui(t)] (i = 1,2,***,#») In conclusion, people's experience of constant temperature value control main basis on deviation and the deviation change rate. So divide observable state space into some situations named as ( / = 1,2, * * *, m ) in this kind of knowledge expression of control process. Every situation is signified by a triple as: k(0 ¿,(0 «,-(*)] (1=1,2,***,#») In fact, every triple signify a set of production control rules of system as: K(0 èi(t)]^ui(t) (i = l,2, - ,m) This set of production control rules can be amended by self-learning. According people's experience of people's experience of constant temperature value control, the bigger the value of ( t ), the greater the corresponding control volume; On the contrary, control volume is small; the value change of ¿f(í) and corresponding control volume follow the same rule. The revised total output control amount is defined as following formula [4]: (ProQuest: ... denotes formulae omitted.) (1) In the formula: ß = 0.8 ~ 0.9 Using above formula can compile subroutine of control algorithm [4] (Fig. 7): 4. The Contrast of the Intelligent Controller and the Traditional Control Through testing and analysis the hardware of the system is superior. The data acquisition and processing is real-time and most functions are implemented by software. Intelligent controller has simple and compact structure, and it does not require an external clock reset circuit; Use 3.3 V power supply, the system power consumption is small; PCS are connected to the microcontroller through the adapter, which can be achieved online debugging and simulation without the emulator head. In order to ensure the changes in the environment in the real-time detection of curing bam, data acquisition and processing must meet certain time limit, so as to make real-time processing, against the happening of the accident. Because of the peripheral circuit is simple, the software can be modified at any time, so the system has strong adaptability. In addition, the system adopts modular structure design; in the case it is easy to add new modules and easy to extend the functions of the system without changing the structure. References [1]. L. X. Wang, Adaptive fuzzy systems and control: design and stability analysis, Prentice Hall, NJ, 1994. [2] . P. G. Harrald, M. Kamstra, Evolving artificial neural networks to combine financial forecasts, IEEE Transaction on Evolution Computers, Vol. 1, Issue 1, 1997, pp. 40-52. [3] . LIN Yan, XUE Yu-Zhi, The realization of high precision temperature control system, Control Engineering of China, Vol. 12, Issue 3, 2005, pp. 273-276. [4] . Wang Shunhuang, Shu Diqian, Intelligent control system and its application, China Machine Press, Beijing, 2005. Jian-Hui MA, Peng GUO, Kai MENG College of Mechanical Engineering, Henan Institute of Engineering, No.l Xianghe Road, Longhu, Xinzheng County, Zhengzhou City in Henan Province, 451191, China Tel:+ 8613598882067 E-mail: [email protected] Received: 5 May 2014 /Accepted: 27 May 2014 /Published: 31 May 2014 (c) 2014 IFSA Publishing, S.L. |