1 Introduction With the continuous increase of China's population and the continuous development of industry and agriculture, the country's fresh water resources are becoming increasingly tense. In order to save water, China's petroleum, chemical, electric power, steel, metallurgical and other enterprises are gradually adopting a circulating water cooling system instead of the original DC cooling water system. In the process of circulation, circulating cooling water will inevitably lead to scaling, metal corrosion and microbial growth, which seriously affect the normal operation of industrial production. At the same time, the concentration of circulating cooling water increases, and the monitoring heat exchanger is very It simulates the operation of industrial field heat exchangers well, monitors corrosion and fouling conditions, and is widely used in water quality monitoring [1, 2]. 1 The system adopts a distributed control system structure, which consists of a host computer and field devices. The system host computer uses the W inCC configuration software to develop the monitoring interface, which can complete real-time monitoring and data dynamic display, abnormal alarm, report production, trend analysis and other management tasks, and is equipped with a 5.7-inch touch screen, making monitoring and control more convenient. Siemens PLC, intelligent monitoring instrument and field sensor distributed in the field, as the lower computer of the monitoring system, complete the real-time data collection of the heat exchanger and automatically control the quality indicators. The heat exchanger in this system is a double-chamber heat exchanger, which is divided into carbon steel test tubes and stainless steel test tubes. The process flow chart is shown in Figure 1. The monitoring points of this system include carbon steel pipe inlet temperature, carbon steel pipe outlet temperature, carbon steel pipe steam temperature, stainless steel pipe inlet temperature, stainless steel pipe outlet temperature, stainless steel pipe steam temperature, feed water pH, feed water conductivity, hydration conductivity, circulating water conductance. Rate, feed water turbidity, carbon steel corrosion rate, stainless steel tube corrosion rate, influent flow rate, steam pressure, electric valve opening. The pressure of the vapor is controlled using a self-operated pressure regulating valve. The flow valve is controlled by a 4~20mA signal, and the flow control is available in both automatic and manual modes. 4.1 Hardware Configuration The PLC part includes the central processing unit CPU 224 XP, the analog input module uses the EM231 and EM231RTD modules, and the analog output uses one analog output of the CPU 224 XP. The corrosion monitoring equipment is an FSY-3 corrosion online detector. It adopts flow meter, pH meter and conductivity measuring instrument produced by American Metro. The turbidity meter is a US hash turbidity meter. 4.2 System structure and working principle The instruments at each site measure the signals of each detection and output point (4~20mA) and send them to the analog input module of the PLC. The PLC accepts the signal from the instrument and according to the meter's range and the instrument's correction coefficient, it is converted to the host computer through the ladder program conversion calculation, and the W inCC software displays and processes. At the same time, the scene can be controlled according to the control mode set by the operator. The flow pump and solenoid valve perform hydration and sewage discharge work. Real-time communication with the W inCC configuration software via the Modbus protocol, mainly for monitoring inlet temperature, outlet temperature, steam temperature, dirt thermal resistance, deposition rate, and scale thickness. The structure of the monitoring system is shown in Figure 2. 5.1 PC monitoring software development The development of the PC monitoring software (Human Machine Interface) is based on Siemens' SIMATICW inCC configuration software, which is a powerful client/server system including all SCADA functions, using Microsoft SQL Server 2000 as its configuration data and archive. Data storage database, you can easily access archived data using ODBC, DAO, OLE-DB, WinCCOLE-DB and ADO. Powerful standard interfaces such as OLE, ActiveX and OPC can easily exchange data with other applications [3] ]. The system monitoring part includes the display of real-time curves and historical curves of the display of various system parameters, dirt thermal resistance, pH, temperature, etc. The main functions are as follows: (1) Pop-up main menu: including main interface, parameter display, flow control, curve display, setting device parameters, printing, timing and other function options. 5.2 Main function module design 5.2.1 Traffic Fuzzy-PID Controller Design There are two pipelines of carbon steel test tube and stainless steel test tube in the system. The flow of the two channels is obtained by weighted distribution of the total flow of the system. In order to ensure the stability of the flow of the two pipelines, the total flow of the system needs to be controlled. The real-time requirements are relatively high. Due to the interference factors such as friction and noise of the bubble and the control mechanism, it is found that the overshoot and the oscillation are large under the condition of the traditional PID flow control test, the effect is not satisfactory, the control precision is not up to the requirement and there is obvious hysteresis. The fuzzy PID controller is used to improve and design a fuzzy controller for the total flow of the system. With two-dimensional fuzzy control, the flow setpoint (set to 4 000 L/h) and the exact measured value deviate to e, the deviation change rate is ec, and E and EC are the fuzzy quantities after e and ec fuzzification, respectively. The controller structure is shown in Figure 3. Let the basic domain of the rate of change ec of the flow rate be [-15, +15], and the discrete domain Y of the selected EC be {-3, -2, -1, 0, 1, 2, 3}, the quantization of ec The factor Kec = 3/15 = 0.2. Let the basic domain of the control change u be [-10, +10], and the discrete domain Z of the selected U be {-4, -3, -2, -1, 0, 1, 2, 3, 4}, Then the quantization factor of u is Ku=10/4=2.5. The linguistic variables E, EC, and U select seven language values: NB, NM, NS, O, PS, PM, and PB. The membership functions of e, ec and u are all triangular, and NB, NM, NS, O, PS, PM, and PB are large, large, large, moderate, small, small, and small, respectively. The control rule is in the form of "ifeandecthenu" [4]. After obtaining the control rule table through the field control process and expert control experience, the fuzzy relation R is obtained according to the fuzzy inference method, and the fuzzy control query table is obtained. As shown in Table 1. Based on the designed fuzzy PID controller, the fuzzy PID algorithm is implemented in the PLC: (1) The initial value of the quantization scale factor is stored in the data storage area of ​​the PLC; (2) It is converted to 4~ by the flowmeter measurement. The 20mA current signal is sent to the EM231 analog input module for acquisition, and the actual flow value is output according to the flow range (0~5000L/h in this system). When calculating e(k) = given value - measured value, ec(k) = The actual results show that the system response speed is fast, the overshoot and oscillation amplitude are small, the time required to reach the steady state becomes shorter, and the lag problem is solved. 5.2.2 Range conversion of analog signals All analog inputs in this system are standard signals of 4~20mA. The digital signals converted by analog input modules are 6 400~32 000[5], and the conversion formula is: [(in1-6 400)×(Limit_h) -Limit_l) /25 600]+Limit_l where: in1———the digital signal value after the signal is converted; Limit_h, Limit_l—the upper and lower limits of the range correction. 5.2.3 Analog acquisition digital filtering Due to the large number of on-site interference sources, in order to reduce various noise interference as much as possible, the analog input signals such as temperature, flow, conductance and pH in the system are digitally filtered after being collected by the analog acquisition module EM231. . The processing flow is divided into two steps: (1) singular values ​​in the sampled data are eliminated. According to the error theory and data processing theory, the Robus criterion [6] is chosen to eliminate the singular value, and the sampling value is subject to the normal distribution, then the mathematical expectation of the sampled value. The dual-chamber heat exchanger monitoring system developed by PLC and WinCC has a friendly interface and comprehensive functions. According to the requirements of the system for flow control, the fuzzy PID controller is used to improve the flow control, the control precision is improved, and the hysteresis problem is improved. At present, the system has been applied to a petrochemical enterprise in Xinjiang. The system is stable and reliable after operation and runs well.
2, system introduction
3. System control requirements
4, system hardware design
5, system software design
(2) Flow chart and parameter display interface: It is used to reflect the running status of the system in real time and dynamically.
(3) Trend curve: The real-time curve and historical curve of the dirt thermal resistance, pH, temperature, etc. are drawn.
(4) Alarm function: alarms each parameter exceeding the safety line, reminds the fault type, reports the fault location, and records the history.
(5) School time function: Since the system needs to record the calculation time of the dirt thermal resistance of the carbon steel test tube and the stainless steel test tube from time to time, the real-time performance is high, in order to prevent the internal clock of the programmable controller from being delayed due to error or power failure. When the calibration is required, the internal clock of the programmable controller is automatically set to the PC clock when the calibration button is pressed.
When the system is in normal operation, the total flow set value SV = 4 000 L / h, the deviation e = SV-PV, PV is the actual measured value of the flow. Let the basic domain of the change e of the flow rate be [-60, +60], and the discrete domain X of the selected E be {-3, -2, -1, 0, 1, 2, 3}, the quantization factor of e Ke = 3/60 = 0.05.
The given value is 4 000 L/h, and t is the conversion time of the EM231 module, which is taken as 0.02 s. After calculation, e(k), ec(k) are stored in the PLC storage area; (3) The fuzzification process converts e(k) and ec(k) from its basic domain to the corresponding fuzzy theory through the quantization factor. (4) Query fuzzy control query table; (5) The precise control quantity parameter is output to the conventional PID control command in the PLC to control the flow.
6, the conclusion