Abstract: This design addresses the boiler temperature control problem, applying process control theory, simulation technology, computer remote control, and configuration software to design a boiler temperature-flow cascade control system. First, a mathematical model of the boiler is established through experiments, obtaining the transfer function between boiler temperature and inlet water flow. Simulation of the theoretically designed control scheme yields a satisfactory response curve, providing a prerequisite for the implementation of the actual control system. Second, intelligent instruments are used as controllers to construct a field instrument process control system. Through parameter tuning, good field control performance is achieved. Third, an integral separation PID control algorithm is implemented.
Keywords: temperature, flow rate, PID control, cascade control system
Introduction
Modern process industries are developing towards large-scale and continuous production, leading to increasingly complex production processes. Meanwhile, the demands for production quality, economic efficiency, safety, reliability, and environmental protection are rising. Furthermore, production safety, reliability, and the economic benefits of production enterprises have become crucial indicators for evaluating the current level of automation control. Therefore, the continued use of conventional control instruments (analog and digital) is no longer sufficient to meet the control requirements of modern process industries. Due to the advantages of computers, such as high processing speed, high precision, large storage capacity, flexible programming, and strong communication capabilities, industrial control and digital controllers—process computer equipment—based on microprocessors and single-chip microprocessors are gradually replacing analog controllers and are playing a very widespread role in process control.
Introducing a computer into a control system allows for the full utilization of its computational, logical reasoning, and memory capabilities to accomplish various control tasks and implement complex control laws. In this system, since computers can only process digital signals, the setpoint and feedback quantity must first be converted into digital quantities by an A/D converter before being input into the computer. Once the computer receives the setpoint and feedback quantity, it performs calculations according to the deviation value and a certain control law (PID). The calculation result is then converted back into an analog signal by a D/A converter and output to the actuator, thereby completing the control of the system.
Early process control systems employed base-mounted instruments and some unit-combined instruments. The structure of these systems was mostly single-input, single-output. Process control theory was based on classical methods such as the frequency method and root locus method, aiming to maintain the stability of controlled parameters (temperature, level, pressure, flow) and eliminate major disturbances. Subsequently, more complex process control systems such as cascade control, ratio control, and feedforward control were gradually applied to industrial production. Pneumatic and electric unit-combined instruments also began to be widely used. Simultaneously, electronic and computer technologies were applied to process control, enabling direct digital control (DDC) and setpoint control (SPC).
Subsequently, modern control theory, comprised of system identification based on the least squares method, optimal control using maximization and dynamic programming as primary methods, and best estimation based on Kalman filtering theory, began to be applied to solve nonlinear, coupled, and time-varying problems in process control and production, providing a better theoretical foundation for industrial process control. Simultaneously, new distributed control systems (DCS) integrate computer technology, control technology, communication technology, fault diagnosis technology, and graphical display technology, ushering in a new era of integrated control and management in industrial automation. Today, industrial automation has entered the era of Computer Integrated Process Systems (CIPS), and is developing towards comprehensive factory automation based on intelligent control technology combining artificial intelligence, control theory, and operations research.
MATLAB also possesses strong functional extensibility, and along with its main system, it can be equipped with a wide variety of toolboxes to accomplish specific tasks. MATLAB has a rich set of functions for control system analysis and design. The MATLAB Control System Toolbox provides various algorithms for the analysis, design, and modeling of linear systems; the MATLAB System Identification Toolbox can identify and model unknown objects within a controlled system. MATLAB's Simulation Toolbox provides an interactive integrated environment for modeling, simulating, and analyzing dynamic systems. It intelligently builds and runs simulations using block diagrams instead of programs, adapting to linear and nonlinear systems; continuous, discrete, and hybrid systems; and single-task and multi-task discrete-event systems.
1. Establishment of the controlled object model
This system primarily controls boiler water temperature, with inlet water flow rate as a secondary control. The goal is to maintain a constant water temperature under a given heating power. Its flowchart is shown in Figure 1.1.
Figure 1.1 Flowchart for measuring the step response of the controlled object
The temperature sensor (main transmitter) converts the temperature signal into an electrical signal, compares it with the setpoint temperature, and sends it to the main controller. The main controller outputs a flow control value, compares it with the inlet water flow signal fed back from the flow transmitter (secondary transmitter), and inputs it to the flow regulator (secondary controller). The flow regulator controls the opening of the regulating valve to control the inlet water flow, thereby achieving setpoint control of the boiler water temperature. Its system block diagram is shown in Figure 1.2.
Figure 1.2 Block diagram of boiler water temperature and flow cascade control system
In control system design, it is necessary to establish a mathematical model for a suitable object in the controlled process. The mathematical model of the controlled object is an important basis for designing process control systems, determining control schemes, analyzing quality indicators, and tuning regulator parameters. The mathematical model (dynamic characteristics) of the controlled object refers to the mathematical expression of the functional relationship between the corresponding output (controlled variable) and the process under the action of various input variables (including control variables and disturbance variables).
In a water temperature-flow cascade control system, our concern is how to maintain a constant water temperature under a given electrothermal power. The influent flow rate is the controlled object of the system; we must measure and calculate its model to analyze the system's steady-state performance and dynamic characteristics, providing a basis for other design work.
2. Measure the step response curve of the controlled object.
In this design, the response curves and related parameters of the controlled object's temperature and flow rate after the input step signal are measured by experimental modeling.
In determining model parameters, the regulating valve can be controlled using the following two methods to apply a step signal to the controlled object:
By changing the opening of the regulating valve through an intelligent regulating instrument, a step signal input to the controlled object can be achieved.
Figure 2.1 Schematic diagram of water temperature-flow rate model measurement
(2) By building a human-machine dialogue window in the MCGS monitoring software, the opening degree of the regulating valve is changed to realize the step signal input of the controlled object.
The program is written as follows:
Traffic PV = PV1
Temperature pv = pt / measured value display output
If set=0 then
Output=6
When Endif / set is 0, a 6mA current is output to the regulating valve.
If set=1 then
Output=8
When Endif / set is 1, it outputs 8mA current to the regulating valve.
Where `set` is the external input signal, which can be set by a button, and `Output` is the output signal, the magnitude of which is the output current value, in mA. The input signal range of the electric regulating valve is a 4-20mA current signal. This allows for a step signal input to the electric regulating valve. A 6mA current corresponds to an opening degree of (6-4)/(20-4) = 12.5% for the electric regulating valve. An 8mA current corresponds to an opening degree of (8-4)/(20-4) = 25%. The flow measurement values before and after the step are 6.5 and 10.2, respectively. The step value is 10.2-6.5=3.7. The actual measured step is shown in Figure 2.2.
Figure 2.2 Step response curves obtained from the experiment
After a step signal is given, the temperature response curve gradually decreases to a stable state. To conform to common practice and facilitate processing, the data is converted into a curve that gradually increases to a stable state, using the first sampled value as the standard. The conversion method is y = 33.71 - x, where y is the processed data and x is the data before processing.
The test results showed that when the valve opening was 12.5, 25, 40, and 80 degrees, the corresponding flow rates measured by the sensor were 6.5, 10.2, 14.6, and 26.2, respectively.
3. Control System Block Diagram Design
A control system block diagram is a prerequisite for the implementation of a control system. It reflects the flow and control process of system information based on the specific process flow. This design employs cascade control. Considering the rapid changes in flow rate and its small time inertia, which should be suppressed quickly, flow rate is selected as the secondary controlled parameter. The secondary loop is a servo control, prioritizing speed; therefore, a P-regulator is used. The P-regulator outputs a signal to control the valve opening, changing the flow rate. The flow sensor sends the detection signal back to the P-regulator, forming negative feedback. This closed loop serves as the inner loop. Temperature changes relatively slowly and has a large time inertia, so it is used as the primary controlled parameter. The primary loop is a setpoint control, prioritizing accuracy; therefore, a PID regulator is employed.
The difference between the setpoint and the feedback signal from the temperature sensor is input to the main controller for PID calculation, thus achieving control. The output signal of the main controller serves as the setpoint for the inner loop. Its difference with the feedback signal from the flow sensor is sent to the P controller for calculation and output to control the regulating valve. Changes in flow rate affect the boiler temperature. The resulting control system block diagram is as follows:
Figure 3 Control System Block Diagram
4 Simulink Control System Simulation
Simulink can dynamically simulate the response of a constructed control system under various signals by simply rewriting the objects in the control system block diagram as transfer functions.
The transfer function of the analog PID controller, D(s) = U(s)/E(s) = Kp(1 + 1/TiS + TdS), can be understood as the sum of the same signal after being processed by proportional (fcn1 in Figure 4.3), integral (fcn2 in Figure 4.3), and derivative (fcn3 in Figure 4.3) operations respectively; the P controller is a pure proportional element (fcn4 in Figure 4.3); the boiler transfer function has been obtained (fcn in Figure 4.3); firstly, assuming the control valve is a pure proportional element (fcn5 in Figure 4.3), the following system diagram can be constructed, where the parameters of PID, P, and valve are not tuned:
Figure 4.1 Linear simulation of Simulink control system
Considering that in practical applications, valves have a dead zone (i.e., there may be flow at 0 degrees of opening or no flow at small degrees of opening, and when the valve reaches its maximum opening, the control signal, although continuing to increase, has lost its regulating effect), and combining the measured flow characteristics of the valve, the valve's transfer function is treated as a nonlinear element, resulting in the nonlinear system diagram:
Figure 4.2 Nonlinear simulation of Simulink control system
In Figure 4.2, the PID and P parameters have been tuned. The combination of the Saturation and Coulomb & Viscous friction components forms the flow characteristics of the valve. Saturation is the limiting component, with an upper limit of 100 and a lower limit of 0. Coulomb & Viscous friction is the viscous friction component, and its function is set as y = 0.30x + 2.9.
To demonstrate the advantages of cascade control, it is necessary to compare the anti-interference capabilities of cascade control systems with those of single-loop control systems.
The characteristic of cascade control is its strong ability to resist secondary interference. A step signal is added to the secondary loop of the cascade control system to simulate the interference of flow. At the same time, in order to compare the anti-interference effect of the data with that of the single-loop control system in the same graph, it is necessary to set up a working area B and store it as a sequence.
The single-loop control system uses a PID controller to directly control the flow rate. A step disturbance signal is introduced at the same location to affect the flow rate, and the simulation results are output to the working area C, stored as a sequence.
Only when the step response curve of a single-loop control system is approximately the same as that of a cascade control system, and the same disturbance signal is applied, are their anti-interference capabilities comparable. In the absence of a disturbance signal, the parameters of the single-loop control system are adjusted to make its response curve approximate the response curve of the cascade control system under the same step signal.
Then, an interference signal is added. For the simulations of the two control systems in Figures 4.6 and 4.7, the simulation time is set to 4000s and the sampling time is set to 20s. The step time of the interference signal is 2000s, with an initial step value of 0 and a final step value of 18.
The simulation results are stored in working areas B and C respectively. To compare the response data of the two systems in the same graph, the following Matlab program needs to be written:
x=0:20:4000;
plot(x,b,x,c)
grid
Figure 4.3 Comparison of anti-interference capabilities of the two control systems
In Figure 4.3, the green curve represents the response curve of the single-loop control system, and the blue curve represents the response curve of the cascade control system. First, zooming in on the step response portions of the two systems in the figure, it can be seen that the two control systems essentially overlap in the rising phase, with roughly the same settling time. The single-loop control system exhibits a larger overshoot in its step response. Considering various indicators, it can be concluded that the control effects of the two systems under a step signal are roughly the same. With the control effects of the two control systems being identical, adding a disturbance signal, the single-loop control system, under the influence of the disturbance signal, reaches a maximum deviation of 0.4, which is 2% of the steady-state value, and even shows a small fluctuation at the end of the curve. In contrast, the cascade control system, under the influence of the disturbance signal, has a maximum deviation of only about 0.1, indicating that the system remains in a characteristic stable state. The two systems demonstrate a significant difference in their anti-interference capabilities, fully proving that cascade control has a strong ability to suppress secondary disturbances.
5. Conclusion
This paper takes a boiler control system as the research object and designs a boiler temperature and flow cascade control system using process control theory, simulation technology, computer remote control, and configuration software. First, a mathematical model of the boiler is established through experiments to obtain the transfer function between boiler temperature and inlet water flow. Simulation of the theoretically designed control scheme yields a good response curve, providing a prerequisite for the implementation of the actual control system. Second, an intelligent instrument is used as the controller to build a field instrument process control system. Through parameter tuning, good field control performance is obtained. Third, an integral separation PID control algorithm is implemented. MATLAB simulation results show that the designed system has fast control response, high control accuracy, and good dynamic characteristics.
