Evaporative crystallization level-flow adaptive cascade control optimization system and method

Abstract

This invention discloses an adaptive cascade control optimization system and method for evaporation and crystallization liquid level-flow rate, relating to the field of automatic control technology for evaporation and crystallization processes. The system includes: a liquid level measuring instrument, a feed flow rate measuring instrument, a discharge flow rate control device, a feed regulating valve, and a controller. The method employs a cascade control structure, where the main loop uses the liquid level of the evaporation and crystallization tank as the controlled variable, and the secondary loop uses the feed flow rate of the feed pipeline as the controlled variable. The output of the main controller serves as the flow rate setpoint for the secondary loop. The method includes a valve action suppression strategy, an adaptive PID tuning strategy, and a feedforward-feedback composite control strategy. This invention, through the synergistic optimization of valve action suppression, adaptive PID tuning, and feedforward-feedback composite control, can fundamentally suppress liquid level oscillations, adapt to all operating conditions, improve the stability and accuracy of liquid level control, and enhance the system's anti-interference capability, ensuring smooth operation of the evaporation and crystallization process.

Description

Technical Field

[0001] This invention relates to the field of automatic control technology for evaporation and crystallization processes, and specifically to an adaptive cascade control optimization system and method for evaporation and crystallization liquid level-flow rate. Background Technology

[0002] Evaporation crystallization is a widely used unit operation in chemical, pharmaceutical, and environmental wastewater treatment fields. The liquid level control effect of its core equipment, the evaporation crystallizer, directly affects the evaporation efficiency, the quality of the crystallized product, and the safety of system operation. In mechanical vapor recompression evaporation crystallization systems, a cascade control structure with liquid level as the primary controlled variable and feed flow rate as the secondary controlled variable is commonly used. By adjusting the feed flow rate, the liquid level is kept stable, providing constant supersaturation conditions for the crystallization process.

[0003] Several control schemes for evaporation and crystallization processes have been disclosed in the prior art. For example, CN1330038A discloses an automatic control method for a sodium sulfate production line. This scheme uses self-tuning PID parameters to adjust the liquid level in the evaporation chamber, detects changes in the liquid level in the evaporation chamber through a dual-flange level transmitter, and uses a multi-parameter self-tuning PID control method to adjust the feed regulating valve. It utilizes the small inertia principle of the regulating valve to maintain the liquid level within a small fluctuation range. Another example is CN206414788U, which discloses an energy-saving evaporation, concentration, and crystallization system for salt solutions. This scheme achieves automatic adjustment and stable operation of the concentration and crystallization system through monitoring liquid level, pressure, and temperature signals and automatic feedback control of the valves.

[0004] However, the following technical problems still exist in practical applications: the evaporation crystallization process has large inertia, large hysteresis, and nonlinear characteristics (such as the evaporation intensity changing with temperature and concentration). Especially during the system startup phase and low-load operation, the feed regulating valve is prone to viscous-slip phenomenon in the small opening range due to nonlinear factors such as friction, resulting in frequent valve vibration and continuous liquid level oscillation, with significant differences in the dynamic characteristics of the controlled object; when the demand of the crystallization system changes and causes fluctuations in the discharge flow rate, the feedback control response is lagging, resulting in a large deviation in liquid level and slow recovery; although the existing self-tuning PID control method can adjust parameters according to process changes, it fails to effectively handle the nonlinear characteristics of the actuator under small openings. The theoretical valve opening command output by the control algorithm is mismatched with the actual valve action characteristics, making it difficult to fundamentally suppress the liquid level oscillation problem, affecting the stability of the crystallization process and product quality. To solve the above problems, experienced control engineers usually have to manually adjust the PID parameters repeatedly on site, which is not only time-consuming and laborious, but also difficult to achieve the optimal control state. For valve jitter problems, sometimes a simple method of setting a control dead zone is used, but this sacrifices control accuracy and makes it impossible to correct small deviations in time.

[0005] To address the aforementioned issues, there is an urgent need to find a liquid level control method for evaporation crystallizers that can effectively suppress the slight opening fluctuations of the regulating valve, in order to overcome the continuous liquid level oscillation defects caused by the nonlinear characteristics of the valve in the existing technology. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to overcome the above-mentioned defects of the prior art and provide a liquid level-flow adaptive cascade control optimization system and method for evaporation and crystallization. This system and method can suppress liquid level oscillation from the source, adapt to the operating characteristics of all working conditions, improve the stability and accuracy of liquid level control and the anti-interference ability of the system, and ensure the smooth operation of the evaporation and crystallization process by valve action suppression, adaptive PID tuning and feedforward-feedback composite control synergistic optimization.

[0007] The technical solution adopted by this invention to solve its technical problem is as follows: an adaptive cascade control optimization system for evaporation and crystallization liquid level-flow rate, the system comprising: A liquid level measuring instrument is installed on the evaporation crystallization tank to detect the liquid level in the evaporation crystallization tank in real time; A feed flow measurement instrument is installed on the feed pipeline to detect the flow rate of the feed pipeline in real time; Discharge flow control equipment is used to control the discharge flow of the evaporation crystallizer; A feed regulating valve is installed on the feed pipeline to regulate the feed flow rate; The controller is connected to the liquid level measuring instrument, the feed flow measuring instrument, the discharge flow control device, and the feed regulating valve, respectively. The controller contains, in the form of functional blocks, a valve action suppression module for implementing the valve action suppression strategy, an adaptive PID tuning module for implementing the adaptive PID tuning strategy, and a feedforward-feedback composite control module for implementing the feedforward-feedback composite control strategy.

[0008] By setting up a liquid level measuring instrument, a feed flow measuring instrument, a discharge flow control device, a feed regulating valve, and a controller integrating an adaptive cascade control method, a complete physical system can be constructed to implement the above control strategy, ensuring the coordinated operation of each functional module. The controller is programmed to implement a cascade control loop, a valve action suppression strategy, an adaptive PID tuning logic, and a feedforward-feedback composite strategy. Based on the liquid level measurement value, the feed flow measurement value, and the discharge control command, the controller calculates and outputs the control command for the feed regulating valve.

[0009] Preferably, the discharge flow control device is a discharge pump driven by a frequency converter, and the speed command output by the frequency converter is input to the controller as the main disturbance quantity for feedforward compensation model calculation.

[0010] By using the frequency converter-driven discharge pump as the discharge flow control device and its speed command as the input of the feedforward compensation model, the main source of interference on the discharge side can be introduced into the control system in real time, thus achieving targeted feedforward compensation.

[0011] Preferably, the controller is a distributed control system or a programmable logic controller (PLC), and the distributed control system or PLC stores computer program code for executing the method.

[0012] By employing a distributed control system or a programmable logic controller as the controller and storing the computer program code that executes the above method, the control method of this invention can be implemented using a mature control platform in the field of industrial control, ensuring the reliability and engineering applicability of the system.

[0013] The technical solution adopted by this invention to further solve its technical problem is as follows: An adaptive cascade control optimization method for evaporation crystallization liquid level-flow rate, applicable to the aforementioned adaptive cascade control optimization system for evaporation crystallization liquid level-flow rate, wherein the method adopts a cascade control structure, wherein the main loop uses the liquid level of the evaporation crystallizer as the controlled variable, the secondary loop uses the feed flow rate of the feed pipe as the controlled variable, and the output of the main controller is used as the flow rate setpoint of the secondary loop. The method includes the following strategies: Valve action suppression strategy: In the control signal path from the secondary circuit output to the feed regulating valve, a minimum opening limit value V with an absolute value greater than zero is set. min When the theoretical opening command V output by the control algorithm cmd When the absolute value of V is less than min The actual command V applied to the feed control valve out The value is: V out =sign(V cmd )×V min Where, sign is the sign function, used to return the theoretical opening instruction V. cmd The symbol, the V min Determined based on the actual opening degree of the feed regulating valve under the lowest stable flow rate of the system; Adaptive PID tuning strategy: Based on the operating stage or operating condition of the evaporation crystallizer, automatically switch the PID parameter groups of the main controller and the auxiliary controller, dividing the PID parameter groups of the main controller and the auxiliary controller into a first group of PID parameters corresponding to the start-up / low load stage and a second group of PID parameters corresponding to the normal / high load stage. Feedforward-feedback composite control strategy: Establish a feedforward compensation model between the main disturbance and the feed flow rate. The main disturbance includes at least the discharge flow rate setpoint or the inverter speed command of the discharge pump. When the discharge flow rate setpoint or the inverter speed command changes, the feedforward flow correction value is calculated through the feedforward compensation model and superimposed on the secondary loop setpoint to correct the final feed flow rate setpoint.

[0014] Preferably, in the valve action suppression strategy, the V min Determined through the following testing steps: When the evaporation and crystallization system is running stably, manually and slowly close the feed regulating valve until a critical point is observed where the feed flow rate is about to begin to decrease uncontrollably. Record the opening value of the feed regulating valve at this point, and use the absolute value of this opening value as the V. min ; The V min The value range is 3% to 10% of the full opening of the feed regulating valve.

[0015] To address the "stickiness" and "jittering" issues of control valves at extremely small openings due to nonlinear friction, a reasonable minimum operating opening threshold V is set before the control signal is output to the valve. min When the calculated absolute value of the theoretical valve opening command is less than this threshold, the valve is forced to open at V. min The amplitude of the movement can prevent the feed control valve from vibrating at high frequencies due to nonlinear friction in the small opening range, thus eliminating the root cause of liquid level oscillation at the actuator level and fundamentally improving the stability and lifespan of the actuator.

[0016] V is determined through preliminary testing. min The specific value is set and limited to 3% to 10% of the full opening. This ensures that the valve action suppression strategy effectively avoids small opening fluctuations without excessively restricting the normal adjustment range.

[0017] Preferably, in the adaptive PID tuning strategy, during the startup / low load phase, the proportional band used by the main controller is larger than that used during the normal / high load phase, and the integral time used by the main controller is longer than that used during the normal / high load phase.

[0018] By using a larger proportional band and a longer integral time during the startup / low load phase to enhance system stability, and a smaller proportional band and a shorter integral time during the normal / high load phase to improve control response speed, the controller parameters can be matched with the dynamic characteristics of the controlled object under different operating conditions, achieving a smooth control effect across the entire operating range.

[0019] Preferably, the adaptive PID tuning strategy further includes a PID parameter fine-tuning step based on the liquid level deviation change trend: When the liquid level deviation in the evaporation crystallizer is e L When the fluctuation remains within a preset small deviation range for a preset duration, reduce the proportional gain K of the main controller. p ; When a liquid level deviation e is detected L When there is a continuous unidirectional deviation and the absolute value of the deviation increases, increase the integral gain K of the main controller. i .

[0020] By reducing the proportional gain when the liquid level deviation fluctuates within a small range for a long time to smooth the control action, and increasing the integral gain when the liquid level deviates continuously in one direction and the deviation increases to accelerate the correction speed, the control action can be dynamically fine-tuned according to the liquid level change trend, further improving the precision and stability of liquid level control, and achieving more refined controller performance self-optimization.

[0021] Preferably, in the adaptive PID tuning strategy, the start-up / low load phase includes at least the initial running period after the evaporation crystallization system starts up, or the period when the feed flow rate is continuously lower than the preset load switching threshold; The normal / high load phase includes at least the period during which the feed flow rate is continuously higher than a preset load switching threshold.

[0022] Based on the identifiable system operating status (such as load rate and operating stage), the system automatically divides at least two typical operating conditions, namely the start-up / low load stage and the normal / high load stage, by comparing the feed flow rate with the preset load threshold. It abandons the single fixed parameter mode and can achieve accurate triggering and automated operation of PID parameter switching without manual intervention. It also configures targeted optimized control parameters for each operating condition, so that the control system always matches the characteristics of the controlled process.

[0023] Preferably, in the feedforward-feedback composite control strategy, the feedforward flow correction value ΔF is calculated through the feedforward compensation model. ff The feedforward flow correction value ΔF ff The feedback flow setpoint F output by the main controller sp-fb The values ​​are superimposed and used together as the final feed flow rate setpoint F for the secondary loop. sp The formula for calculating the final feed flow rate setpoint is: F sp =F base +F sp-fb +ΔF ff F base This is the basic feed flow rate set based on the production load.

[0024] By establishing a feedforward compensation model between the main disturbance and the feed flow rate, and superimposing the feedforward correction value and the feedback flow rate setpoint as the final setpoint of the secondary loop, the feed flow rate can be adjusted in advance to compensate before the disturbance of the discharge flow rate change affects the liquid level. This feedforward action, combined with the feedback control loop, forms a composite control structure, which significantly improves the system's ability to suppress the main disturbance.

[0025] Preferably, the feedforward compensation model is a proportional model: ΔF ff =K ff ×ΔD, or ΔF ff =K ff ×ΔS, where ΔD is the change in the discharge flow rate setpoint, ΔS is the change in the inverter speed command, and K ff Forward coefficients; The feedforward coefficient K ff The value was determined through system step response testing. It is the steady-state gain of the feed flow rate to the change in the discharge flow rate, i.e., the ratio of the change in feed flow rate ΔF to the step change in discharge flow rate ΔD. K ff =ΔF / ΔD. The two ΔD values ​​above refer to the same physical quantity and have the same meaning.

[0026] By adopting the proportional model ΔF ff =K ff ×ΔD is used as the feedforward compensation model, and the feedforward coefficient K is determined based on the system step response test. ff K represents the steady-state gain of the feed flow rate on the change in the discharge flow rate, i.e., the ratio of the change in feed flow rate ΔF to the step change in discharge flow rate ΔD. ff =ΔF / ΔD, which ensures the accuracy of feedforward compensation and the ease of engineering implementation.

[0027] The beneficial effects of this invention on the adaptive cascade control optimization system and method for evaporation and crystallization liquid level-flow rate are as follows: (1) This invention sets a valve action suppression strategy in the control signal path from the secondary circuit output to the feed regulating valve, and performs range limitation processing on the theoretical opening command output by the control algorithm. This can avoid the operating condition of the feed regulating valve in the small opening range from the actuator level, effectively eliminate the sticky sliding and high-frequency shaking phenomenon caused by nonlinear friction of the valve, and fundamentally solve the problem of continuous liquid level oscillation caused by the mismatch between the actual valve action characteristics and the output command of the control algorithm. It reduces the fluctuation amplitude by more than 70%, ensures the near-linear operation stability of the evaporation and crystallization process, provides constant supersaturation conditions for the crystallization process, and improves the quality stability and consistency of the crystallized product. At the same time, it can effectively avoid abnormal wear of internal components such as valve core and valve seat caused by high-frequency small stroke shaking of the regulating valve, and extend the service life of the equipment. (2) By adopting an adaptive PID parameter switching strategy based on operating conditions, combined with a parameter dynamic fine-tuning mechanism based on the liquid level deviation change trend and a feedforward-feedback composite control structure, this invention enables the control parameters to be accurately matched with the dynamic characteristics of the controlled object in the entire operating range of the evaporation crystallization system. This effectively adapts to the control requirements of the system from startup, low load to normal high load. At the same time, it can compensate for the main disturbances on the discharge side in advance, greatly improve the anti-interference capability and liquid level control accuracy of the system, reduce the cost of manual tuning and intervention during system operation, greatly shorten the system recovery time, ensure the control reliability and stability of the evaporation crystallization system in the entire operating range, reduce the excessive reliance on the personal experience of DCS control engineers, and is conducive to the standardization and application of the technology. Attached Figure Description

[0028] Figure 1 This is a structural block diagram of an embodiment of the system of the present invention; Figure 2 This is a schematic diagram of the input-output characteristics of the valve action suppression strategy according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the operating condition judgment and parameter switching logic of the adaptive PID tuning strategy of the main controller in an embodiment of the present invention (the operating condition judgment and parameter switching logic of the secondary controller is the same as that of the main controller). Figure 4 This is a flowchart of the overall control method according to an embodiment of the present invention; In the diagram, F pv The feed flow rate measurement is shown below. FT is the flow meter, LT is the level transmitter, FV is the control valve, and the discharge variable frequency pump is a discharge pump driven by a frequency converter. P A For the proportional band of the main controller under operating condition A, I A P is the integral time of the main controller under operating condition A. B For the proportional band of the main controller under operating condition B, I B is the integral time of the main controller under operating condition B, and RMS is the root mean square value. Detailed Implementation

[0029] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0030] Example: like Figure 1 As shown, the evaporation crystallizer liquid level-flow adaptive cascade control optimization system applicable to this embodiment is built on the mechanical vapor recompression evaporation crystallization process system. The hardware components of the system include an evaporation crystallizer, a liquid level measuring instrument, a feed flow measuring instrument, a discharge flow control device, a feed regulating valve, and a controller.

[0031] The liquid level measuring instrument adopts a dual-flange differential pressure liquid level transmitter, which is installed on the side wall of the evaporation crystallizer. The pressure taps of the liquid level measuring instrument are respectively set below the gas phase space and the lowest operating liquid level of the liquid phase in the evaporation crystallizer. It is used to detect the liquid level in the evaporation crystallizer in real time and transmit the detected liquid level analog signal to the controller in real time.

[0032] The feed flow measurement instrument is an electromagnetic flow meter, which is installed on the feed pipeline of the evaporation crystallizer. The front end of the feed pipeline is connected to the raw liquid storage tank to be treated, and the rear end is connected to the feed inlet of the evaporation crystallizer. The measurement range of the electromagnetic flow meter covers the minimum feed flow rate to the maximum design feed flow rate of the system. It is used to detect the feed flow rate in the feed pipeline in real time and transmit the detected flow rate analog signal to the controller in real time.

[0033] The discharge flow control equipment uses a centrifugal pump driven by a frequency converter, which is installed on the discharge pipeline of the evaporation crystallizer. The front end of the discharge pipeline is connected to the bottom discharge port of the evaporation crystallizer, and the rear end is connected to the subsequent crystallization solid-liquid separation system. The control terminal of the frequency converter is connected to the controller signal to receive the speed command output by the controller, adjust the operating speed of the discharge pump, and thus control the discharge flow of the evaporation crystallizer. At the same time, the frequency converter synchronously transmits the real-time output speed command to the controller as the input signal for feedforward compensation.

[0034] The feed regulating valve is a pneumatic diaphragm regulating valve, installed on the feed pipeline and located downstream of the feed flow measurement instrument. The signal receiving end of the feed regulating valve is connected to the analog output end of the controller to receive the opening command output by the controller, and adjust its own valve opening accordingly, thereby changing the feed flow in the feed pipeline.

[0035] The controller employs a programmable logic controller (PLC) or a distributed control system. It is equipped with analog input modules, analog output modules, digital input modules, and digital output modules. Signals from all field instruments and equipment are connected to the controller through their respective modules. The controller stores computer program code that implements this control method, enabling it to perform all control logic operations and output instructions. The controller contains functional blocks for implementing a valve action suppression strategy, an adaptive PID tuning strategy, and a feedforward-feedback composite control strategy.

[0036] It should be noted that the "strategy" in the technical solution of this invention refers to the logical combination of methods, while the "module" in this embodiment is the specific program functional unit that implements the strategy. The two are essentially corresponding.

[0037] like Figure 1As shown, the cascade control structure used in this embodiment includes a main controller and a secondary controller. The main controller is a level PID controller, and the secondary controller is a flow PID controller. The input signals to the main controller include the level setpoint and the real-time level measurement value detected by the level measuring instrument. The main controller performs PID calculations based on the deviation between the level setpoint and the measured level, and the output result serves as the feedback flow setpoint for the secondary loop. The input signals to the secondary controller include the final feed flow setpoint for the secondary loop and the real-time feed flow measurement value detected by the feed flow measuring instrument. The secondary controller performs PID calculations based on the deviation between the final feed flow setpoint and the measured feed flow, outputs a theoretical opening command to the valve action suppression module, and after processing, outputs an actual opening command to the feed regulating valve.

[0038] In this embodiment, the valve action suppression module, which implements the valve action suppression strategy, is configured as a functional block in the control signal path from the secondary loop to the feed regulating valve. All theoretical opening commands output by the secondary controller must be processed by this module before being output to the feed regulating valve. The module's computational logic is integrated within the controller and continuously executes during system operation. This module is implemented through software logic within the controller, requiring no additional hardware configuration.

[0039] As the fundamental hardware connection of the control system, a one-to-one signal connection channel is established between the controller's analog output module and the electric positioner of the feed control valve. The analog output module outputs a standard current signal of 4mA to 20mA, corresponding to the opening range of the feed control valve from -100% to 100%. Specifically, 4mA corresponds to -100% opening, 12mA corresponds to 0% opening, and 20mA corresponds to 100% opening. The opening from -100% to 0% corresponds to the closing direction adjustment of the feed control valve, and the opening from 0% to 100% corresponds to the opening direction adjustment of the feed control valve.

[0040] In the software logic implementation, a valve action suppression strategy function block is written in the controller's program development environment. The input variable of the function block is the theoretical opening command output by the secondary controller, and the output variable is the actual opening command applied to the feed control valve. The internal parameter of the function block is a preset minimum opening limit value. The execution cycle of the function block is consistent with the control cycle of the secondary controller; in this embodiment, it is set to 0.2s to ensure the real-time performance of the valve action suppression processing.

[0041] like Figure 2 As shown, the valve action suppression strategy is as follows: The standard valve opening command V output by the secondary loop flow controller... cmd The following processing is performed, and the processing logic is implemented using the following piecewise function mathematical formula: ; Among them, V outThis parameter represents the actual opening command applied to the feed control valve. It is an output variable of the valve action suppression strategy function block, corresponding to the opening value of the analog signal ultimately output by the controller to the feed control valve. Its value ranges from -100% to 100%, with the sign of the value indicating the valve's direction of action and the absolute value representing the valve's opening size. V cmd This parameter, representing the theoretical valve opening command output by the control algorithm, is an input variable of the valve action suppression strategy function block. It is the valve opening command output by the secondary controller after PID calculation, and its value ranges from -100% to 100%. V min This is the preset minimum opening limit value. This parameter is an internal parameter of the valve action suppression strategy function block. It is a fixed value with an absolute value greater than zero, determined through field testing. A typical value is approximately 5% of the valve's full stroke. `if` is a conditional judgment flag used to trigger the execution logic of the corresponding segment. `otherwise` is an other condition flag, referring to all operating conditions that do not meet the above two conditions.

[0042] The execution condition for the first logic segment is the theoretical opening instruction V. cmd Greater than zero and less than the minimum opening limit value V min At this time, the actual opening command V out The value of is equal to the minimum opening limit value V. min This logic segment corresponds to the small opening range in the forward opening direction of the valve. When the forward opening command output by the secondary controller is less than the minimum opening limit, the valve opening is forcibly limited to the minimum opening limit to prevent the valve from operating in the small forward opening range.

[0043] The execution condition for the second logic segment is the theoretical opening instruction V. cmd If the value is greater than the negative minimum opening limit but less than zero, then the actual opening command V is... out The value of is equal to the negative minimum opening limit. This logic segment corresponds to the small opening range in the reverse closing direction of the valve. When the reverse opening command output by the secondary controller is between the negative minimum opening limit and zero, the valve opening is forcibly limited to the negative minimum opening limit to prevent the valve from entering the reverse small opening range.

[0044] The execution condition for the third logic segment is any other operating condition that does not meet the above two conditions; in this case, the actual opening instruction V is executed. out The value of V is related to the theoretical opening command. cmd Completely consistent with the output, no additional processing is required; the output is directly sent to the feed control valve. This logic section covers the normal valve opening range, ensuring that the valve's adjustment action within the non-micro-opening range completely follows the control algorithm's output and does not affect the system's normal regulation performance.

[0045] like Figure 2 As shown in the figure, the horizontal axis represents the theoretical opening command V.cmd The vertical axis represents the actual output opening command V. out The figure shows the input-output characteristic curves for two scenarios: without a suppression strategy and with a suppression strategy. Without a suppression strategy, the input-output characteristic is a straight line with a slope of one, and the actual output opening is exactly the same as the theoretical command. With a suppression strategy, when the theoretical opening command is within the range between the positive and negative minimum opening limits, the actual output opening is limited to the positive and negative minimum opening limits, forming a suppression zone to prevent the valve from operating within a very small opening range.

[0046] Minimum opening limit value V min Through on-site testing, it was determined that the testing procedures should be conducted after the evaporation crystallization system has been installed and debugged, but before formal production. The testing process was carried out under stable system operation. During the test, the controller was switched to manual control mode, and the opening command was manually output to the feed regulating valve through the controller's human-machine interface to stabilize the system feed flow rate within 20% to 30% of the design rated flow rate, maintain a stable liquid level in the evaporation crystallizer, and keep the discharge flow rate constant. Subsequently, the feed regulating valve was manually and slowly closed, with each adjustment not exceeding 1% of the full opening. After each adjustment, stable operation was maintained for at least 30 seconds, and the trend of feed flow rate changes was continuously observed until a critical point was observed where the feed flow rate was about to begin an uncontrollable decrease. The opening value of the feed regulating valve corresponding to this critical point was recorded, and the absolute value of this opening value was taken as the minimum opening limit value V. min .

[0047] For example, in this embodiment, the minimum opening limit value V obtained through field testing is... min The value is 5% of the full opening of the feed control valve. This value falls within the range of 3% to 10% of the full opening, effectively avoiding valve vibration at small openings without excessively restricting the valve's normal adjustment range. After testing, this value is written into the controller's holding register as an internal parameter of the valve action suppression strategy function block.

[0048] Specifically, during the system's automatic operation, the valve action suppression module continuously executes, performing real-time condition judgment and corresponding processing on each theoretical opening command output by the secondary controller. When the theoretical opening command output by the secondary controller fluctuates slightly near zero, the actual output opening command is limited to the positive and negative minimum opening limit values, preventing the valve from operating in the small opening range. This eliminates the high-frequency jitter problem caused by the valve's nonlinear friction force at the actuator level, thereby preventing liquid level oscillation.

[0049] In this embodiment, the computational logic of the adaptive PID tuning module, which implements the adaptive PID tuning strategy, is integrated into the controller and is continuously executed during system operation. It automatically switches the PID parameter groups of the main controller and the auxiliary controller according to the operating stage or operating condition of the evaporation crystallizer, and can also fine-tune the PID parameters in real time according to the changing trend of the liquid level deviation.

[0050] Specifically, the adaptive PID tuning strategy is as follows: First, the PID parameter groups of the main controller and the slave controller are divided into two groups. The first group of PID parameters corresponds to the low load stage of startup, and the second group of PID parameters corresponds to the normal high load stage. Both groups of PID parameters are pre-tuned by the closed-loop tuning method before the system is put into production and stored in the controller's storage unit.

[0051] like Figure 3 As shown, the division and judgment logic of the operation phases are explained in detail first.

[0052] The low-load startup phase includes the initial running period after the evaporation and crystallization system starts up, or the period during which the feed flow rate is consistently lower than a preset load switching threshold. In this embodiment, the preset initial running period is the first 2 hours after the system's cold start, and the preset low-load switching threshold is 40% of the system's designed rated feed flow rate. When the system meets the condition that the running time after cold start is less than 2 hours, or the real-time feed flow rate is consistently lower than 40% of the designed rated feed flow rate, the controller determines that the system is in the low-load startup phase and automatically calls the first set of PID parameters.

[0053] The normal high-load phase includes periods when the feed flow rate is consistently higher than a preset load switching threshold. In this embodiment, the preset high-load switching threshold is 60% of the system's designed rated feed flow rate, and the duration of this determination is 30 minutes. When the system's real-time feed flow rate consistently exceeds 60% of the designed rated feed flow rate for 30 minutes, the controller determines that the system is in the normal high-load phase and automatically calls the second set of PID parameters.

[0054] Understandably, when the system's feed flow rate is in the range of 40% to 60%, the controller keeps the currently called PID parameter set unchanged to avoid frequent parameter switching due to small fluctuations in flow rate and to ensure the stability of the control process.

[0055] The following section provides a detailed explanation of the rules for setting the two sets of PID parameters.

[0056] The calculation formula for a PID controller is: ; Where u(t) is the output value of the PID controller at time t. For the main controller, this output value is the feedback flow setpoint; for the secondary controller, this output value is the theoretical opening command. K p For proportional gain, this parameter is inversely related to the proportional band; the larger the proportional band value, the smaller the proportional gain value, and the weaker the proportional effect. e(t) is the input deviation value of the PID controller at time t. For the main controller, this deviation is the difference between the liquid level setpoint and the liquid level measurement value; for the secondary controller, this deviation is the difference between the final feed flow rate setpoint and the feed flow rate measurement value. K i This parameter represents the integral gain, which is inversely related to the integration time. The larger the integration time, the smaller the integral gain, and the weaker the integration effect. K is the integral term of the deviation value over time, used to eliminate the steady-state deviation of the system. d This is the differential gain, which is directly proportional to the differential time. The larger the differential time value, the larger the differential gain value, and the stronger the differential effect. The rate of change of the deviation value at time t is the input of the differential term, used to predict the trend of deviation changes and apply control in advance.

[0057] Specifically, during the low-load startup phase, the proportional gain K of the main controller is greater than that used during the normal high-load phase. p Smaller, and the integral time used by the main controller is longer than that used during normal high-load phases, corresponding to the integral gain K of the main controller. i Smaller. This parameter setting method can reduce the intensity of control action during the low-load startup phase, avoiding overshoot and oscillation caused by the system's large inertia characteristics, and ensuring the smoothness of the system startup process.

[0058] During the normal high-load phase, the proportional band used by the main controller is smaller than that used during the startup low-load phase, corresponding to the proportional gain K of the main controller. p The integral gain is larger, and the integral time used by the main controller is shorter than that used during the low-load startup phase, corresponding to the integral gain K of the main controller. i Larger. This parameter setting method can enhance the response speed and anti-interference capability of the control action during normal high-load phases, ensuring the accuracy and speed of liquid level control.

[0059] For example, in this embodiment, the first set of PID parameters for the main controller has a proportional band of 300%, an integral time of 120s, and a derivative time of 5s. The second set of PID parameters for the main controller has a proportional band of 150%, an integral time of 60s, and a derivative time of 3s. The two sets of PID parameters for the slave controller are tuned according to the fast response characteristics of the flow loop. The first set of parameters has a proportional band of 120%, an integral time of 10s, and a derivative time of 0s; the second set of parameters has a proportional band of 80%, an integral time of 5s, and a derivative time of 0s.

[0060] like Figure 3 As shown, taking the main controller as an example, during system operation, it continuously collects real-time feed flow rate and system runtime, cyclically executes operating condition judgment logic, and switches the PID parameter group of the main controller according to the judgment result. Simultaneously, it can execute parameter fine-tuning logic based on the liquid level deviation change trend. The adaptive PID tuning logic of the secondary controller is the same as that of the main controller, and will not be described further here.

[0061] The following section provides a detailed explanation of the PID parameter fine-tuning steps based on the trend of liquid level deviation changes.

[0062] During system operation, the controller continuously collects liquid level measurements and calculates the liquid level deviation e. L Liquid level deviation e L The calculation formula is: e L =L sp -L pv ; Among them, e L L represents the liquid level deviation in the evaporation crystallizer, which is the difference between the set liquid level value and the measured liquid level value. sp The liquid level setpoint for the evaporation crystallizer is the target liquid level value required by the process, which is preset by the operator in the controller. pv This refers to the liquid level measurement value of the evaporation crystallizer detected in real time by the liquid level measuring instrument.

[0063] Specifically, when the liquid level deviation e in the evaporation crystallizer L When the fluctuation remains within a preset small deviation range for a preset duration, the controller reduces the proportional gain K of the main controller. p The operation is as follows. In this embodiment, the preset duration is set to 5 minutes, and the preset small deviation range is set to -1% to 1% of the liquid level range. The controller continuously calculates the liquid level deviation e over the past 5 minutes. L If the level deviation remains within a preset small deviation range for a preset time period, the controller will reduce the root mean square (RMS) value of the main controller. When the RMS value remains below 1% of the level range, the level deviation is determined to be in a small fluctuation state. At this time, the proportional gain K of the main controller will be adjusted.p Multiplying by a correction factor of 0.9 slightly reduces the proportional gain, weakens the proportional effect, avoids continuous small oscillations caused by excessive control action, and makes the control process smoother.

[0064] When the controller detects a liquid level deviation e L When there is a continuous unidirectional deviation and the absolute value of the deviation increases, the integral gain K of the main controller is increased. i The operation. In this embodiment, the criterion for determining continuous unidirectional deviation is that the liquid level deviation e L The controller maintains the same positive or negative sign for five consecutive control cycles, and the absolute value of the deviation in each control cycle is greater than the absolute value of the deviation in the previous cycle. When this condition is met, the controller will increase the integral gain K of the main controller. i Multiplying by a correction factor of 1.5 temporarily increases the integral gain, enhances the integral action, accelerates the elimination of steady-state deviation, and prevents the liquid level from continuously deviating from the set value.

[0065] Understandably, the parameter fine-tuning operation has upper and lower limit protection settings, and the proportional gain K... p With integral gain K i The adjustment range shall not exceed ±30% of the corresponding operating condition reference parameter to avoid excessive parameter adjustment leading to control instability. At the same time, when switching operating conditions, the fine-tuned parameters shall be automatically reset to the reference parameters of the corresponding operating condition to ensure the consistency of parameter switching.

[0066] In this embodiment, the computational logic of the feedforward-feedback composite control module, which implements the feedforward-feedback composite control strategy, is integrated into the controller and is continuously executed during system operation. By establishing a feedforward compensation model between the main disturbance and the feed flow rate, the disturbance on the discharge side is compensated in advance, thereby improving the system's anti-interference capability.

[0067] Specifically, the feedforward-feedback composite control strategy is as follows: In this embodiment, the main disturbance is the discharge flow rate setpoint, or the inverter speed command of the discharge pump can also be selected. Both types of disturbances can be directly read from the controller without the need for additional detection instruments. When the discharge flow rate setpoint or the inverter speed command changes, the feedforward flow correction value is calculated through the feedforward compensation model and superimposed on the secondary loop setpoint to correct the final feed flow rate setpoint. The feedforward compensation model adopts a proportional model, and the feedforward flow correction value ΔF is calculated through the feedforward compensation model. ff The feedforward flow correction value ΔF ff The feedback flow setpoint F output by the main controller sp-fb The values ​​are superimposed and used together as the final feed flow rate setpoint F for the secondary loop. sp .

[0068] The formula for calculating the final feed flow rate setpoint is: Fsp =F base +F sp-fb +ΔF ff ; Among them, F sp This is the final feed flow rate setpoint for the secondary loop, and the input setpoint for the secondary controller. F sp-fb The feedback flow setpoint output by the main controller is the value output by the main controller after PID calculation based on the liquid level deviation. ΔF ff F is the feedforward flow correction value, which is the correction value calculated by the feedforward compensation model based on the change in the discharge-side disturbance. base This is the basic feed flow rate set based on the production load.

[0069] The formula for calculating the feedforward flow correction value is: ΔF ff =K ff ×ΔD, or ΔF ff =K ff ×ΔS; Among them, K ff ΔS is the feedforward coefficient, a fixed proportional coefficient, predetermined through system step response testing. ΔD is the change in the discharge flow rate setpoint, which is the difference between the current discharge flow rate setpoint and the previous discharge flow rate setpoint. When the discharge pump inverter speed command is used as the disturbance, ΔS is the change in the inverter speed command.

[0070] Next, we will analyze the feedforward coefficient K. ff The testing and determination process will be explained in detail.

[0071] Feedforward coefficient K ff The test was conducted before the system was put into production, and the test process was carried out under stable system operation. During the test, the controller was switched to manual control mode, and the feed flow rate and discharge flow rate were manually adjusted to stabilize the liquid level in the evaporation crystallizer at the process set value, maintaining stable system operation. Subsequently, a step change of a fixed amplitude was given to the discharge flow rate set value, and the step change amount ΔD was recorded. Keeping other operating parameters unchanged, after the system reached a stable state again, the change in feed flow rate ΔF required to maintain a constant liquid level in the evaporation crystallizer was recorded, along with the feedforward coefficient K. ff The value of K is the ratio of the change in feed flow rate ΔF to the step change in discharge flow rate ΔD. ff =ΔF / ΔD.

[0072] For example, in this embodiment, the step change in the set value of the discharge flow rate during testing is 10m. 3 After the system stabilizes, the required change in feed flow rate to maintain a constant liquid level is 10 m³ / h. 3 / h, therefore the feedforward coefficient Kff The value is set to one. This value conforms to the principle of material conservation. When the discharge flow rate increases, the feed flow rate needs to increase by the same amount to maintain a stable liquid level. Feedforward compensation can adjust the set value of the feed flow rate simultaneously with the change in discharge flow rate, without waiting for the liquid level to change before feedback adjustment.

[0073] Specifically, during the automatic operation of the system, the controller continuously collects real-time data of the discharge flow rate setpoint, calculates the change in the discharge flow rate setpoint ΔD, and calculates the feedforward flow correction value ΔF through the feedforward compensation model. ff The feed flow rate is then superimposed on the feedback flow rate setpoint output by the main controller to obtain the final feed flow rate setpoint of the secondary loop, thus realizing feedforward-feedback composite control.

[0074] In this embodiment, the control cycle of the main controller is set to 1 second, and the control cycle of the secondary controller is set to 0.2 seconds. This conforms to the cascade control design principle that the flow loop is faster than the level loop, ensuring that the secondary loop can quickly eliminate interference from the feed side, and the main loop achieves precise control of the level. During system operation, the valve action suppression module, the adaptive PID tuning module, and the feedforward-feedback composite control module are executed synchronously to collaboratively achieve stable and precise control of the level in the evaporation crystallizer.

[0075] In such Figure 4 The flowchart of the overall control method of this embodiment of the invention clearly illustrates the collaborative workflow of the three core strategies within each control cycle. Its working principle is as follows: At the start of each control cycle, the controller first collects the real-time liquid level of the evaporation crystallizer, the actual feed flow rate of the feed pipeline, and the set value of the discharge flow rate (or the speed command of the discharge pump frequency converter). After the data collection is completed, the main controller (liquid level PID controller) calculates the deviation between the liquid level set value and the measured liquid level, and outputs a feedback flow set value. This value represents the direction and magnitude of the feed flow rate adjustment required to eliminate the current liquid level deviation.

[0076] Meanwhile, the feedforward compensation module calculates a feedforward flow correction value based on the change in the discharge flow rate setpoint (i.e., the change in the main disturbance). This correction value, based on a pre-tested and determined proportional relationship, is used to compensate for changes in the discharge flow rate in advance, preventing the disturbance from being transmitted to the liquid level before feedback adjustment. Subsequently, the feedback flow rate setpoint and the feedforward flow correction value are superimposed to form the final feed flow rate setpoint of the secondary loop.

[0077] The secondary controller (flow PID controller) targets the final feed flow rate setpoint and compares it with the measured feed flow rate. Through PID calculations, it obtains a theoretical valve opening command. This command reflects the valve opening size and direction required to track the set flow rate.

[0078] After the theoretical valve opening command is generated, it enters the valve action suppression module. This module first determines whether the absolute value of the theoretical opening command is less than the minimum opening limit value determined in advance through field testing. If the absolute value of the theoretical opening command is less than this limit value, it means that the controller is attempting to drive the valve to a small opening range that is prone to nonlinear frictional vibration. To avoid high-frequency valve vibration and actuator wear caused by this, the valve action suppression module forcibly sets the actual output valve opening command to the minimum opening limit value and maintains the same direction as the original command (i.e., forward or reverse opening); if the absolute value of the theoretical opening command is greater than or equal to the minimum opening limit value, the actual output command is consistent with the theoretical command. The processed actual opening command is sent to the feed regulating valve to drive the valve to actuate and regulate the feed flow rate, thereby completing the closed-loop control of the liquid level in the evaporation crystallizer.

[0079] Throughout the control cycle, the adaptive PID tuning module operates in parallel, continuously monitoring the system's operating status. This module first automatically determines whether the system is in a startup / low-load or normal / high-load phase based on indicators such as feed flow rate and system runtime, and accordingly switches the PID parameter sets of the main and auxiliary controllers. During the startup / low-load phase, a parameter set with a larger proportional band and longer integral time is used to ensure smooth system startup; during the normal / high-load phase, it switches to a parameter set with a smaller proportional band and shorter integral time to enhance the control system's rapid response and anti-interference performance. Furthermore, the module fine-tunes the main controller parameters based on the trend of liquid level deviation: when the liquid level deviation fluctuates within a small range for a preset time, the proportional gain of the main controller is appropriately reduced to suppress small oscillations that may be caused by excessive proportional gain; when a continuous unidirectional deviation in liquid level deviation is detected and the absolute value of the deviation continues to increase, the integral gain of the main controller is appropriately increased to accelerate the elimination of steady-state deviation and prevent the liquid level from deviating from the set value for an extended period.

[0080] All the above strategies are completed collaboratively within a control cycle, after which the controller waits for the start of the next cycle and repeats the entire process. By organically combining three major strategies—valve action suppression, adaptive PID tuning, and feedforward-feedback composite control—this method fundamentally solves the nonlinear jitter problem of the feed regulating valve at small openings, achieves automatic matching of controller parameters with different operating conditions, and significantly improves the ability to suppress interference on the discharge side, thereby achieving stable and precise control of the liquid level in the evaporation crystallizer across the entire operating range.

Claims

1. A level-flow adaptive cascade control optimization system for evaporation and crystallization, characterized in that, The system consists of: A liquid level measuring instrument is installed on the evaporation crystallization tank to detect the liquid level in the evaporation crystallization tank in real time; A feed flow measurement instrument is installed on the feed pipeline to detect the flow rate of the feed pipeline in real time; Discharge flow control equipment is used to control the discharge flow of the evaporation crystallizer; A feed regulating valve is installed on the feed pipeline to regulate the feed flow rate; The controller is connected to the liquid level measuring instrument, the feed flow measuring instrument, the discharge flow control device, and the feed regulating valve, respectively. The controller contains, in the form of functional blocks, a valve action suppression module for implementing the valve action suppression strategy, an adaptive PID tuning module for implementing the adaptive PID tuning strategy, and a feedforward-feedback composite control module for implementing the feedforward-feedback composite control strategy.

2. The adaptive cascade control optimization system for evaporation and crystallization liquid level-flow rate according to claim 1, characterized in that, The discharge flow control device is a discharge pump driven by a frequency converter. The speed command output by the frequency converter is used as the main disturbance input to the controller for feedforward compensation model calculation.

3. The adaptive cascade control optimization system for evaporation and crystallization liquid level-flow rate according to claim 1 or 2, characterized in that, The controller is a distributed control system or a programmable logic controller, and the distributed control system or programmable logic controller stores computer program code for executing the method.

4. A method for adaptive cascade control optimization of evaporation crystallization liquid level-flow rate, applicable to the adaptive cascade control optimization system for evaporation crystallization liquid level-flow rate as described in any one of claims 1 to 3, characterized in that, This method employs a cascade control structure, where the liquid level in the evaporation crystallizer is the controlled variable in the main loop, the feed flow rate in the feed pipeline is the controlled variable in the secondary loop, and the output of the main controller serves as the flow rate setpoint for the secondary loop. The method includes the following strategies: Valve action suppression strategy: In the control signal path from the secondary circuit output to the feed regulating valve, a minimum opening limit value V with an absolute value greater than zero is set. min When the theoretical opening command V output by the control algorithm cmd When the absolute value of V is less than min The actual command V applied to the feed control valve out The value is: V out =sign(V cmd )×V min Where, sign is the sign function, used to return the theoretical opening instruction V. cmd The symbol, the V min Determined based on the actual opening degree of the feed regulating valve under the lowest stable flow rate of the system; Adaptive PID tuning strategy: Based on the operating stage or operating condition of the evaporation crystallizer, automatically switch the PID parameter groups of the main controller and the auxiliary controller, dividing the PID parameter groups of the main controller and the auxiliary controller into a first group of PID parameters corresponding to the start-up / low load stage and a second group of PID parameters corresponding to the normal / high load stage. Feedforward-feedback composite control strategy: Establish a feedforward compensation model between the main disturbance and the feed flow rate. The main disturbance includes at least the discharge flow rate setpoint or the inverter speed command of the discharge pump. When the discharge flow rate setpoint or the inverter speed command changes, the feedforward flow correction value is calculated through the feedforward compensation model and superimposed on the secondary loop setpoint to correct the final feed flow rate setpoint.

5. The method for adaptive cascade control optimization of evaporation crystallization liquid level-flow rate according to claim 4, characterized in that, In the valve action suppression strategy, the V min Determined through the following testing steps: When the evaporation and crystallization system is running stably, manually and slowly close the feed regulating valve until a critical point is observed where the feed flow rate is about to begin to decrease uncontrollably. Record the opening value of the feed regulating valve at this point, and use the absolute value of this opening value as the V. min ; The V min The value range is 3% to 10% of the full opening of the feed regulating valve.

6. The method for adaptive cascade control optimization of evaporation crystallization liquid level-flow rate according to claim 4 or 5, characterized in that, In the adaptive PID tuning strategy, during the startup / low load phase, the proportional band used by the main controller is larger than that used during the normal / high load phase, and the integral time used by the main controller is longer than that used during the normal / high load phase.

7. The method for adaptive cascade control optimization of evaporation crystallization liquid level-flow rate according to any one of claims 4 to 6, characterized in that, The adaptive PID tuning strategy also includes a PID parameter fine-tuning step based on the liquid level deviation change trend: When the liquid level deviation in the evaporation crystallizer is e L When the fluctuation remains within a preset small deviation range for a preset duration, reduce the proportional gain K of the main controller. p ; When a liquid level deviation e is detected L When there is a continuous unidirectional deviation and the absolute value of the deviation increases, increase the integral gain K of the main controller. i .

8. The method for adaptive cascade control optimization of evaporation crystallization liquid level-flow rate according to any one of claims 4 to 7, characterized in that, In the adaptive PID tuning strategy, the start-up / low load phase includes at least the initial running period after the evaporation crystallization system starts up, or the period when the feed flow rate is continuously lower than the preset load switching threshold. The normal / high load phase includes at least the period during which the feed flow rate is continuously higher than a preset load switching threshold.

9. The method for adaptive cascade control optimization of evaporation crystallization liquid level-flow rate according to any one of claims 4 to 8, characterized in that, In the aforementioned feedforward-feedback composite control strategy, the feedforward flow correction value ΔF is calculated using the feedforward compensation model. ff The feedforward flow correction value ΔF ff The feedback flow setpoint F output by the main controller sp-fb The values ​​are superimposed and used together as the final feed flow rate setpoint F for the secondary loop. sp The formula for calculating the final feed flow rate setpoint is: F sp =F base +F sp-fb +ΔF ff F base This is the basic feed flow rate set based on the production load.

10. The method for adaptive cascade control optimization of evaporation crystallization liquid level-flow rate according to claim 9, characterized in that, The feedforward compensation model is a proportional model: ΔF ff =K ff ×ΔD, or ΔF ff =K ff ×ΔS, where ΔD is the change in the discharge flow rate setpoint, ΔS is the change in the inverter speed command, and K ff Forward coefficients; The feedforward coefficient K ff The value was determined through system step response testing. It is the steady-state gain of the feed flow rate to the change in the discharge flow rate, i.e., the ratio of the change in feed flow rate ΔF to the step change in discharge flow rate ΔD. K ff =ΔF / ΔD.

Images