1 Introduction
Chilled water is a fundamental system in factory utilities. A chilled water pressure control system based on PLC and frequency converters offers advantages such as high automation, energy efficiency, safety, hygiene, and convenient maintenance; it utilizes Frofibus bus technology for high scalability; and the host computer control system features advanced functions such as dynamic process display, workflow management, and printing.
2 System Principle Design
2.1 System Overview
(1) Target design
The goal of the system design is to maintain stable water supply pressure in the pipeline network using PLC automatic control technology when the refrigerant water demand of production positions changes, so as to achieve the goals of energy conservation, emission reduction, and cost reduction.
(2) Scheme Design
Each chiller is equipped with two pumps. During normal operation, one pump operates in variable speed mode, while the other is on standby and can be put into operation at any time. Switching between the two pumps is done manually, and they are interlocked to prevent simultaneous operation. To ensure equal average operating time for all pumps, a timed pump replacement function is implemented. When this function is set, if a pump's continuous operating time exceeds the set value and the standby pump is in "rest" mode, the system will prompt for pump replacement, ensuring equal operating time for all pumps and extending their lifespan. In the event of a frequency converter failure, the system can automatically switch back to mains frequency to continue operation, ensuring uninterrupted water supply.
(3) Functional Design
The system features alarm functions, real-time monitoring, and data storage. Alarm displays include over-limit alarms and fault alarms. An over-limit alarm is generated when a preset monitored analog quantity exceeds a specified limit or when a switch quantity jumps to the alarm position. A fault alarm is generated when a preset monitored equipment or process malfunctions, or when the control system malfunctions. Once an alarm event occurs, the alarm signal is uploaded to the host computer, and a buzzer is simultaneously activated. Alarm records are displayed in different colors. The host industrial control computer stores real-time data such as the on/off status or fault status of each water pump, as well as real-time data on temperature, outlet pressure, regulating valve opening, and pump speed. It also allows for quick report retrieval and printing.
2.2 System Composition
This system design comprises three parts: a host computer, a local touchscreen, and a slave computer. The host computer displays the process flow diagram, parameter grouping diagram, equipment operating status, and dynamically displays data such as chilled water temperature, pressure, and pump speed. It also features high-speed historical data storage and retrieval, and alarm functions. The local touchscreen also dynamically displays chilled water temperature, pressure, and pump speed data. The slave PLC implements the automatic control of the chilled water process.
The lower-level computer system consists of a Siemens S-7200 PLC, an ABB frequency converter, a pressure sensor, a temperature sensor, an analog regulating valve, and other control equipment. For the PLC control part, since the system has 6 analog inputs and 4 analog outputs, an expansion unit is required. Therefore, the host computer is selected as one CPU224 PLC, plus two analog output modules EM232, and then an analog I/O module EM235 is added. The EM277 Frofibus-DP module is used to communicate with the host computer [1]. This module is used to receive commands from the host computer and upload alarm signals.
2.3 Control Principle
The system employs dual-path PID closed-loop control, adjusting the proportional valve and water pump speed based on pressure gauge readings to ensure stable refrigerant water pressure at the workstations and maximize system energy efficiency. The system principle block diagram is shown in Figure 1.
Figure 1 System schematic diagram
When the chiller is running, the control system controls the refrigerant water circulation pump to maintain a constant flow rate. At this time, the pump speed is set to maximum, and the pressure sensor detects the pipeline pressure, outputting a 4-20mA current signal to the PLC. This pressure feedback signal and the pressure setpoint signal are calculated by a fuzzy PID control program, outputting a control signal to the analog regulating valve. When the pressure is insufficient, the opening of the analog regulating valve is reduced to decrease refrigerant water backflow, thereby increasing the outlet pressure; conversely, the opening of the analog regulating valve is increased to increase refrigerant water backflow and decrease the outlet pressure. When the chiller stops running, i.e., the refrigerant water temperature reaches the set temperature, the control system automatically controls the refrigerant water pump to switch to variable flow constant pressure. At this time, the analog regulating valve closes, and the pressure feedback signal and the pressure setpoint signal are calculated by another fuzzy PID control program inside the PLC, outputting a speed control signal to the frequency converter. When the pressure is insufficient, the frequency converter increases the output frequency, the pump speed increases, the water supply increases, forcing the outlet pressure to rise. Conversely, the pump speed decreases, the water supply decreases, and the outlet pressure drops, thereby ensuring stable refrigerant water pressure. The system maintains a stable outlet pressure of 0.4 MPa, thus ensuring the operating efficiency of the refrigeration unit. The pressure regulation accuracy is ±5% of the set value, i.e., ±0.02 MPa, and the pressure can return to normal within 0.5-2 seconds after a change.
3. Fuzzy PID controller for chilled water temperature
3.1 Characteristics of Fuzzy PID Control
Classical PID closed-loop algorithms struggle to achieve control convergence in chilled water pressure regulation systems. Fuzzy PID control, on the other hand, utilizes the current control deviation, combined with changes in the dynamic characteristics of the controlled process, and based on practical experience with the specific process, determines control parameters through fuzzy rule reasoning according to certain control requirements or objective functions, thereby achieving system control.
Fuzzy control has a weak dependence on mathematical models and does not require the establishment of precise mathematical models of the process. Fuzzy control has a good control effect on the dynamic process of the system, but it cannot eliminate the static error of the system. Therefore, considering the respective characteristics of fuzzy control and PID control, applying a combination of PID control and fuzzy control to achieve step-by-step control of the system will yield good control results.
3.2 Fuzzy PID Control Process
Because the user's water demand is uncertain and the water pressure in the pipeline fluctuates greatly, it is difficult to determine the mathematical model for this system. However, fuzzy control does not require a precise mathematical model. Therefore, the pressure control algorithm is designed using fuzzy PID control [2]-[4].
Fuzzy PID control uses the error e and error change ec as inputs, which are then fuzzified and described using fuzzy language. Fuzzy control rules are used to determine the true value of the control quantity, and the output variable is u, which is a control current of 4–20 mA. The working process of the fuzzy controller can be described as follows: First, the input quantities of the fuzzy controller are converted into fuzzy quantities for use by the fuzzy control logic decision system. The fuzzy decision unit determines the fuzzy relation r according to the control rules, applies the fuzzy logic inference algorithm to obtain the fuzzy output quantity of the controller, and finally, the control quantity is precisely calculated to control the controlled object. The fuzzy PID control diagram is shown in Figure 2.
Figure 2. Fuzzy PID control block diagram
The fuzzy linguistic variables for pressure difference e, pressure difference change rate ec, and control quantity u are e, ec, and u, respectively. Their fuzzy linguistic values are {nb, nm, ns, zo, ps, pm, pb}, representing {negative large, negative medium, negative small, zero, positive small, positive medium, positive large}. Generally, the number of elements in the fuzzy universe of discourse is twice the size of the fuzzy linguistic vocabulary, so the fuzzy universe of discourse is {-6, -5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5, 6}. e and ec are fuzzified based on the scaling factors ke and kec.
Where n=6, the pipeline pressure variation range is 0.3～0.5MPa, and the set value is 0.4MPa, the basic universe of discourse for the error is obtained as e∈[-0.1, 0.1]. Empirically, it is known that under normal circumstances, the pressure variation will not exceed 0.05MPa/s, therefore the basic universe of discourse for the error variation is ec∈[-0.05, 0.05]. Thus, the scaling factors for the error e and the error increment ec are 60 and 120, respectively. Considering the principles of coverage, sensitivity, and robustness of the universe of discourse, the membership function of this system is selected as a triangular membership function.
Fuzzy control rules are the core of fuzzy control. They simulate human reasoning ability based on fuzzy concepts, essentially the process of using language to inductively summarize manual control strategies. Determining fuzzy control essentially involves summarizing control experience to derive a series of fuzzy conditional statements. These are represented by compound conditional statements such as: if
e=nlandec=nl
The formula `thenu=nl` optimizes both the dynamic and static characteristics of the system's output response. In this system, since `e` and `ec` each have 7 language input values, there are a total of 7 × 7 = 49 `if-then` statements, which can be summarized into a fuzzy control rule table, as shown in the attached table.
4. Conclusion
This paper designs an automatic chilled water control system based on PLC and frequency converter with remote monitoring function. It features fast and accurate response, convenient operation and maintenance, and high energy efficiency. The application of a fuzzy PID controller to this constant pressure control system overcomes the shortcomings of traditional PID control, improves the system's nonlinearity and large time lag characteristics, and enhances its robustness.
