Optimization method for PID parameters of PID control system, and system thereof

Abstract

The present disclosure relates to an optimization method for PID parameters of a PID control system, and a system thereof. The method comprises: collecting closed-loop control data of a current feedback control loop; respectively filtering a PV curve and an MV curve, and respectively performing polynomial fitting on a filtered PV curve and a filtered MV curve; respectively preprocessing a first fitted curve and a second fitted curve; analyzing the first fitted curve having undergone abnormal peak and trough removal and the second fitted curve having undergone abnormal peak and trough removal to obtain feature information for evaluating a control effect of PID parameters; evaluating the control effect of the PID parameters on the basis of the feature information to obtain evaluation parameters of the control effect of the PID parameters; and optimizing a proportional band and an integral time in the PID parameters on the basis of a convergence factor and an overshoot factor, such that continuous optimization can be achieved, and the PID control effect can also be improved.

Description

A method for optimizing PID parameters in a PID control system and the system thereof.

[0001] Cross-references to related applications

[0002] This disclosure claims priority to Chinese Patent Application No. 202411805125.1, filed on December 10, 2024, entitled "A Method for Optimizing PID Parameters of a PID Control System and a System thereof", the entire contents of which are incorporated herein by reference. Technical Field

[0003] This disclosure relates to the field of PID control technology, and in particular to a method and system for optimizing PID parameters in a PID control system. Background Technology

[0004] For PID control systems, existing PID parameter optimization methods mainly utilize the internal model method, which uses step tests to obtain the dynamic response of the loop and optimize the PID parameters. However, this optimization method relies on model identification. Due to the characteristics and limitations of existing identification methods, this method often achieves higher accuracy in open-loop identification than in closed-loop identification, and it is dependent on step signals.

[0005] However, in actual production, considering the need for stable production, most loops do not meet the conditions for step testing, thus limiting existing optimization methods. Furthermore, the PID parameters obtained by existing PID optimization methods are often unrelated to the current control state, failing to fully utilize control information under the current PID parameter conditions and hindering continuous optimization. Additionally, some existing closed-loop PID parameter optimization methods rely on fixed strategies for qualitative analysis and optimization, resulting in accuracy deficiencies.

[0006] Public content

[0007] (a) Technical problems to be solved

[0008] In view of the above-mentioned shortcomings and deficiencies of the prior art, this disclosure provides a method and system for optimizing PID parameters of a PID control system, which solves the technical problems of the inability to continuously optimize and the poor control effect in the prior art.

[0009] (II) Technical Solution

[0010] To achieve the above objectives, the main technical solutions adopted in this disclosure include:

[0011] In a first aspect, embodiments of this disclosure provide a method for optimizing PID parameters in a PID control system. The optimization method includes: acquiring closed-loop control data of the current feedback control loop; wherein the closed-loop control data includes a PV curve and an MV curve; filtering the PV curve and MV curve respectively to obtain filtered PV curves and filtered MV curves, and performing polynomial fitting on the filtered PV curves and filtered MV curves respectively to obtain a first fitted curve corresponding to the filtered PV curve and a second fitted curve corresponding to the filtered MV curve; preprocessing the first fitted curve and the second fitted curve respectively to remove abnormal peaks and troughs from the first fitted curve and the second fitted curve, obtaining a first fitted curve and a second fitted curve after abnormal peak and trough removal; analyzing the first fitted curve and the second fitted curve after abnormal peak and trough removal to obtain feature information for evaluating the control effect of the PID parameters; evaluating the control effect of the PID parameters based on the feature information to obtain evaluation parameters for the control effect of the PID parameters; wherein the evaluation parameters include a convergence factor and an overshoot factor; and optimizing the proportional gain and integral time in the PID parameters based on the convergence factor and the overshoot factor.

[0012] In one possible embodiment, the closed-loop control data further includes the SV curve; the feature information includes, but is not limited to, the absolute phase difference between the first fitted curve after removing abnormal peaks and troughs and the second fitted curve after removing abnormal peaks and troughs, the change time of the filtered PV curve from the moment of change of the SV curve to the change value of the SV curve when the SV curve changes, the maximum overshoot when the SV curve changes, the average oscillation period when the SV curve does not change, and the oscillation amplitude when the SV curve does not change.

[0013] In one possible embodiment, when the SV curve changes, the control effect of the PID parameters is evaluated based on feature information to obtain evaluation parameters for the control effect of the PID parameters, including: comparing the maximum overshoot with the allowable control deviation corresponding to the current feedback control loop; if the maximum overshoot is less than or equal to the allowable control deviation corresponding to the current feedback control loop and the maximum overshoot is not equal to 0, then the time ratio of the adjustment time and change time corresponding to the current feedback control loop is calculated, and based on the target ratio range in which the time ratio is located, the convergence factor is determined to be the convergence factor corresponding to the target ratio range; if the maximum overshoot is greater than the allowable control deviation corresponding to the current feedback control loop, then the maximum PV adjustment value of the filtered PV curve is obtained, and the first ratio of the maximum PV adjustment value and the SV adjustment value is calculated, and then the first ratio is substituted into the formula overshoot factor overshoot = a - 0.5 to obtain the overshoot factor; where a represents the first ratio.

[0014] In one possible embodiment, when the SV curve changes, the control effect of the PID parameters is evaluated based on feature information to obtain evaluation parameters of the control effect of the PID parameters. This further includes: if the maximum overshoot is equal to 0, obtaining the PV adjustment value in the filtered PV curve that is closest to the SV adjustment value in the SV curve, and calculating a second ratio between the SV adjustment value and the PV adjustment value. Then, substituting the second ratio into the formula overshoot factor overshoot = b * 0.5, the overshoot factor is obtained; where b represents the second ratio.

[0015] In one possible embodiment, when the SV curve does not change, the control effect of the PID parameters is evaluated based on feature information to obtain evaluation parameters for the control effect of the PID parameters, including: determining that all peak amplitudes and all trough amplitudes of the filtered PV curve are within the target range; wherein, the target range is determined by the SV adjustment value in the SV curve and the allowable control deviation corresponding to the current feedback control loop; if it is determined that at least one peak amplitude or trough amplitude in the filtered PV curve exceeds the target range and the average oscillation period is greater than the steady-state threshold corresponding to the current feedback control loop, then the range of the absolute phase difference is determined, and the overshoot factor is determined based on the range of the absolute phase difference, and the convergence factor is determined based on the absolute phase difference and the oscillation amplitude.

[0016] In one possible embodiment, the range of the absolute phase difference is determined, and based on the range of the absolute phase difference, the overshoot factor is 0.5 if the absolute phase difference is less than 18 degrees; if the absolute phase difference is in the range of 18-72 degrees, the absolute phase difference is substituted into the formula overshoot factor overshoot = (c-18) / 108+0.5 to obtain the overshoot factor; where c represents the absolute phase difference; if the absolute phase difference c exceeds 72 degrees, the overshoot factor is 1.

[0017] In one possible embodiment, the convergence factor is determined based on the absolute phase difference and the oscillation amplitude, including: if the absolute phase difference is between 30 and 90 degrees, the absolute phase difference is substituted into the formula convergencespeed = (90-c) / 120 to obtain the convergence factor; where c represents the absolute phase difference; if the absolute phase difference is less than 30 degrees, the third ratio of the oscillation amplitude vmean to 2*error is calculated, and it is determined whether the third ratio is greater than 2; if the third ratio is less than or equal to 2, the third ratio is substituted into the convergence factor convergencespeed = f / 2 to obtain the convergence factor; where f is the third ratio; if the third ratio is greater than 2, the convergence factor is 1.

[0018] In one possible embodiment, when the SV curve changes, the proportional gain and integral time in the PID parameters are optimized based on the convergence factor and overshoot factor, including:

[0019] If the overshoot factor is less than 0.5 and the convergence factor is less than 0.5, the optimized scaling factor PB new And the optimized integration time Ti new The formula for calculating PB is as follows: new =(0.5+overshoot)×(0.5+convergencespeed)×PB old Ti new = (0.5 + overshoot) × Ti old ;

[0020] In the formula, PB old The scaling factor before optimization is shown; overshoot is the overshoot factor; convergencespeed is the convergence factor; Ti old The integration time before optimization;

[0021] Alternatively, if the overshoot factor is greater than or equal to 0.5 and the convergence factor is less than 0.5, then the optimized scaling factor PB... new And the optimized integration time Ti new The calculation formula is as follows: B new =(1.5-overshoot)×(0.5+convergencespeed)×PB old Ti new = [1 + 2 × (overshoot - 0.5)] × Ti old ;

[0022] Alternatively, if the overshoot factor is less than 0.5 and the convergence factor is greater than 0.5, then the optimized scaling factor PB... new And the optimized integration time Ti new The formula for calculating PB is as follows: new =[1+2×(convergencespeed-0.5)]×PB old Ti new =(1.5-convergencespeed)×(0.5+overshoot)×Ti old ;

[0023] Alternatively, if the overshoot factor is greater than or equal to 0.5 and the convergence factor is greater than or equal to 0.5, then the optimized scaling factor PB... new And the optimized integration time Ti new The formula for calculating PB is as follows: new=[1+2×(convergencespeed-0.5)]×PB old Ti new = [1 + 2 × (overshoot - 0.5)] × Ti old .

[0024] In one possible embodiment, when the SV curve does not change, the proportional gain and integral time in the PID parameters are optimized based on the convergence factor and overshoot factor, including:

[0025] If the overshoot factor is less than 0.5 and the convergence factor is less than 0.5, then the optimized scaling factor PB new And the optimized integration time Ti new The formula for calculating PB is as follows: new = (0.5 + convergencespeed) × PB old Ti new =(0.5+convergencespeed)×(0.5+overshoot)×Ti old ;

[0026] In the formula, PB old The scaling factor before optimization is shown; overshoot is the overshoot factor; convergencespeed is the convergence factor; Ti old The integration time before optimization;

[0027] Alternatively, if the overshoot factor is greater than or equal to 0.5 and the convergence factor is less than 0.5, then the optimized scaling factor PB... new And the optimized integration time Ti new The formula for calculating PB is as follows: new = (0.5 + convergencespeed) × PB old Ti new =[1+2×(overshoot-0.5)]×(0.5+convergencespeed)×Ti old ;

[0028] Alternatively, if the overshoot factor is less than 0.5 and the convergence factor is greater than 0.5, then the optimized scaling factor PB... new And the optimized integration time Ti new The formula for calculating PB is as follows: new =[1+2×(convergencespeed-0.5)]×PB old Ti new =(1.5-convergencespeed)×(0.5+overshoot)×Tiold ;

[0029] Alternatively, if the overshoot factor is greater than or equal to 0.5 and the convergence factor is greater than or equal to 0.5, then the optimized scaling factor PB... new And the optimized integration time Ti new The formula for calculating PB is as follows: new =[1+2×(convergencespeed-0.5)]×PB old Ti new = [1 + 2 × (overshoot - 0.5)] × Ti old .

[0030] In a second aspect, embodiments of this disclosure provide a PID control system, including a controller, the controller being configured to execute a method for optimizing PID parameters of the PID control system as described in any of the first aspects.

[0031] (III) Beneficial Effects

[0032] The beneficial effects of this disclosure are:

[0033] This disclosure provides a method and system for optimizing PID parameters in a PID control system. The method involves acquiring closed-loop control data from the current feedback control loop, then filtering the PV and MV curves to obtain filtered PV and MV curves. Polynomial fitting is then performed on the filtered PV and MV curves to obtain a first fitted curve corresponding to the filtered PV curve and a second fitted curve corresponding to the filtered MV curve. Subsequently, preprocessing is performed on both the first and second fitted curves to remove abnormal peaks and troughs from the first and second fitted curves, thus obtaining abnormal peaks. The first fitted curve after removing troughs and the second fitted curve after removing abnormal peaks and troughs are then analyzed to obtain characteristic information for evaluating the control effect of PID parameters. Based on this characteristic information, the control effect of the PID parameters is evaluated to obtain evaluation parameters, including convergence factors and overshoot factors. Subsequently, the proportional gain and integral time in the PID parameters are optimized based on the convergence and overshoot factors. Compared to existing solutions, this approach not only achieves continuous optimization but also improves the PID control effect.

[0034] To make the above-mentioned objects, features and advantages to be achieved by the embodiments of this disclosure more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0035] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings used in the embodiments of this disclosure will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this disclosure and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 shows a flowchart of a method for optimizing PID parameters of a PID control system provided in an embodiment of this disclosure;

[0037] Figure 2 shows a schematic diagram of a PID control system provided in an embodiment of this disclosure. Detailed Implementation

[0038] To better explain and facilitate understanding of this disclosure, the following detailed description of the disclosure is provided in conjunction with the accompanying drawings and specific embodiments.

[0039] To address the issues of poor control performance and stability in existing PID control loops, this disclosure provides a method and system for optimizing PID parameters in a PID control system. By utilizing information such as loop type characteristics and control trends under current PID parameter control, the PID control performance is evaluated. Based on the evaluation results, PID parameters are optimized. Since this PID parameter optimization method is based on existing PID parameters, it supports continuous iterative optimization of the closed-loop loop, gradually improving the PID control performance. Furthermore, it lowers the threshold for PID parameter optimization, enhancing the control performance and stability of the PID control loop.

[0040] To better understand the above technical solutions, exemplary embodiments of this disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be clearer and more thorough in understanding and will fully convey the scope of this disclosure to those skilled in the art.

[0041] To facilitate understanding of the embodiments of this disclosure, some terms involved in this disclosure are explained as follows:

[0042] “PV”: Measured value;

[0043] “SV”: Setting value;

[0044] “MV”: Valve position value;

[0045] "Average time interval tmean": It represents the average time between all adjacent peaks and troughs of the same curve.

[0046] It's important to note that in calculating the average time interval tmean, using half-waves (0 degrees and 180 degrees of a trigonometric function, not 360 degrees) provides higher accuracy than using 360 degrees. Of course, calculating using 360 degrees is also acceptable.

[0047] "Average amplitude vmean": It represents the average amplitude among all peaks and troughs of the same curve;

[0048] "Time interval t1": It represents the time interval between the current adjacent peaks and troughs of the same curve;

[0049] “Amplitude v1”: It represents the amplitude between the current adjacent peaks and troughs of the same curve;

[0050] "Waveform error time tmin": The waveform error time tmin is different for different types of feedback control loops. Please refer to Table 1 mentioned later for details.

[0051] "Change time tup": It represents the time from the moment the SV curve changes to the change value of the SV curve when the SV curve changes.

[0052] "Maximum overshoot": It represents the maximum overshoot of PV when there is a change in the SV curve;

[0053] "Allowable control deviation error": The allowable control deviation error is different for different types of feedback control loops. Please refer to Table 1 mentioned later for details.

[0054] "Adjust time": The adjustment time varies for different types of feedback control loops. Please refer to Table 1 mentioned later for details.

[0055] "Stable time": The steady time threshold varies for different types of feedback control loops. Please refer to Table 1 mentioned later for details.

[0056] "convergence speed": It is used to evaluate the current control effect and for PID parameter optimization;

[0057] "Overshoot factor": It is used to evaluate the current control effect and for PID parameter optimization;

[0058] “PB”: Proportionality;

[0059] “TI”: Integral Time.

[0060] First Embodiment

[0061] Please refer to Figure 1, which shows a flowchart of a method for optimizing PID parameters of a PID control system according to an embodiment of this disclosure. As shown in Figure 1, this optimization method can be executed by an electronic device, and the specific device of the electronic device can be set according to actual needs; this disclosure is not limited thereto. For example, the electronic device can be a computer or a server, etc. Specifically, the optimization method includes:

[0062] Step S110: Collect the closed-loop control data of the current feedback control loop. This closed-loop control data includes the PV curve, MV curve, and SV curve.

[0063] It should be understood that the type of feedback control loop can be set according to actual needs, and the embodiments disclosed herein are not limited thereto.

[0064] For example, the current feedback control loop can be a flow loop, a pressure loop, a level loop, a temperature loop, or other types of loops.

[0065] Step S120: Filter the PV curve and MV curve respectively to obtain the filtered PV curve and the filtered MV curve, and perform polynomial fitting on the filtered PV curve and the filtered MV curve respectively to obtain the first fitting curve corresponding to the filtered PV curve and the second fitting curve corresponding to the filtered MV curve.

[0066] Specifically, in order to better find the characteristics of the control curve, the first-order mean filtering method can be used to filter the PV curve and the MV curve respectively to obtain the filtered PV curve and the filtered MV curve. Then, polynomial fitting is performed on the filtered PV curve to obtain the first fitted curve, and polynomial fitting is performed on the filtered MV curve to obtain the second fitted curve.

[0067] Step S130: Preprocess the first fitting curve and the second fitting curve respectively to remove abnormal peaks and valleys in the first fitting curve and the second fitting curve, so as to obtain the first fitting curve and the second fitting curve after removing abnormal peaks and valleys.

[0068] Specifically, the information of all peaks and troughs in the first fitted curve is determined, and then the average time interval tmean and average amplitude vmean of all peaks and troughs in the first fitted curve are calculated. Also, the time interval t1 and amplitude v1 between the currently adjacent peaks and troughs in the first fitted curve are calculated, provided that... and In the case of an abnormal peak or valley, the current adjacent peak or valley is identified as an abnormal peak or valley, and the current adjacent peak or valley is removed.

[0069] In addition, if t1 < tmin, then the currently adjacent peaks and troughs are also determined to be abnormal peaks and troughs, and the currently adjacent peaks and troughs are removed. The waveform error time tmin varies for different types of feedback control loops, as detailed in Table 1 below.

[0070] Table 1: Loop Parameter Selection Table

[0071] As shown in Table 1, when the current feedback control loop type is a flow loop, the waveform error time tmin is 1 minute; when the current feedback control loop type is a pressure loop, a level loop, or a temperature loop, the waveform error time tmin is 5 minutes; and when the current feedback control loop type is another type, the waveform error time tmin is 3 minutes.

[0072] Furthermore, the adjustment time, steady-state threshold, and allowable control deviation in Table 1 are similar to the waveform error time tmin. The corresponding values ​​can be selected according to the type of feedback control loop mentioned in Table 1, and will not be repeated here.

[0073] It should be noted that although Table 1 contains multiple values, those skilled in the art should understand that these values ​​can be adjusted according to actual needs, and are not limited thereto. For example, although Table 1 states that the waveform error time tmin corresponding to the flow loop is 1 minute, those skilled in the art can adjust it to 2 minutes, etc., according to actual needs.

[0074] It should be noted that although the preprocessing of the first fitted curve is described using a coefficient of 1 / 4 as an example, those skilled in the art should understand that the coefficient can be adjusted to other coefficients, as long as the coefficient is greater than 0 and less than 0.5. This disclosure is not limited to this.

[0075] It should also be noted that the preprocessing process for the second fitted curve is similar to that for the first fitted curve, and will not be repeated here. For details, please refer to the relevant description of the first fitted curve.

[0076] Step S140: Analyze the first fitting curve and the second fitting curve after removing abnormal peaks and valleys to obtain characteristic information for evaluating the control effect of PID parameters.

[0077] It should be understood that the specific information included in the feature information can be set according to actual needs, and the embodiments disclosed herein are not limited thereto.

[0078] For example, this feature information includes, but is not limited to, the absolute phase difference between the first fitted curve after removing abnormal peaks and troughs and the second fitted curve after removing abnormal peaks and troughs, the change time of the filtered PV curve from the moment of change of the SV curve to the change value of the SV curve when there is a change in the SV curve, the maximum overshoot when there is a change in the SV curve, the average oscillation period when there is no change in the SV curve, and the oscillation amplitude when there is no change in the SV curve.

[0079] It should also be understood that the process of acquiring feature information can be set according to actual needs, and the embodiments disclosed herein are not limited thereto.

[0080] Optionally, the process for obtaining the absolute phase difference between the first fitted curve after removing the abnormal peaks and troughs and the second fitted curve after removing the abnormal peaks and troughs is as follows:

[0081] The peaks and troughs of the first fitted curve after removing abnormal peaks and troughs can be sequentially matched with the peaks and troughs of the second fitted curve after removing abnormal peaks and troughs. If any peaks and troughs cannot be matched, they can be deleted. For example, the first peak of the first fitted curve after removing abnormal peaks and troughs can be matched with the first peak of the second fitted curve, and the first trough of the first fitted curve after removing abnormal peaks and troughs can be matched with the first trough of the second fitted curve, and so on. Then, the absolute phase difference between the first and second fitted curves after removing abnormal peaks and troughs can be calculated. This absolute phase difference ranges from 0 to 90°; that is, it calculates the absolute phase difference between the peaks and troughs that correspond in the first and second fitted curves after removing abnormal peaks and troughs according to the above matching method.

[0082] Optionally, the process for obtaining the time tup of the filtered PV curve representing the change in the SV curve from the moment of change in the SV curve to the change value of the SV curve, and the maximum overshoot overmax when the SV curve changes, is as follows:

[0083] When the SV curve changes (i.e., the SV curve is not a straight line), the number of changes in the SV curve can be determined.

[0084] When the SV curve changes once (i.e., the SV curve includes two parallel straight lines with a connecting line between them), the time tup of the filtered PV curve from the moment of change of the SV curve (i.e., the moment when the first straight line and the connecting line of the two parallel straight lines meet) to the moment when the SV curve first reaches the moment of change of the SV curve (i.e., the moment when the second straight line and the connecting line of the two parallel straight lines meet) can be calculated, and the maximum overshoot overmax can be calculated.

[0085] When the SV curve changes multiple times, the multiple change times tup of the filtered PV curve from each change moment of the SV curve to the corresponding change value of the SV curve can be calculated. For example, when the SV curve changes twice, the first change time tup of the filtered PV curve from the first change moment of the SV curve to the first change value of the SV curve can be calculated, and the second change time tup of the filtered PV curve from the second change moment of the SV curve to the second change value of the SV curve can also be calculated. Subsequently, the average of the multiple change times tup can be calculated, and this average value can be used as the final change time tup, and the maximum overshoot overmax can be calculated.

[0086] It should be understood that the maximum overshoot (overmax) is the amount by which the adjustment exceeds the SV adjustment value. For example, if SV is changed from 10 to 15, PV, under the action of the controller, will start to rise from 10, reach 15, and may continue to rise to 16 before falling back to around 15. In this case, the overshoot is 16 - 15 = 1. The same logic applies to downward adjustment; if SV is changed from 10 to 5, PV will drop to a minimum of 4, and the overshoot is 1.

[0087] It should be noted that, in principle, only data segments with a single SV change are analyzed (a single data segment can contain at most one SV change). However, this method also supports use when SV changes multiple times. Furthermore, for multiple changes, the maximum overshoot (overmax) can be calculated separately, and then the average can be taken (this involves calculating the overshoot factor, which also needs to be calculated separately and then averaged). This will be explained later.

[0088] Optionally, the process for obtaining the average oscillation period and the oscillation amplitude when the SV curve does not change is as follows:

[0089] When the SV curve does not change (i.e., the SV curve is a straight line), the average oscillation period is 2*tmean and the oscillation amplitude is vmean.

[0090] It's important to note that when calculating the average time interval tmean earlier, for greater accuracy, we actually calculated half of the wave, that is, the time interval between the peak and trough of the PV wave (i.e., the positions of 0 degrees and 180 degrees in a trigonometric function, not 360 degrees). Therefore, a complete oscillation cycle requires multiplying this value by 2. Conversely, if calculated using 360 degrees, multiplying by 2 is not necessary.

[0091] Step 150: Evaluate the control effect of the PID parameters based on the feature information to obtain evaluation parameters for the control effect of the PID parameters. These evaluation parameters include a convergence factor and an overshoot factor.

[0092] It should be understood that the specific process for evaluating the control effect of PID parameters based on feature information can be set according to actual needs, and the embodiments disclosed herein are not limited thereto.

[0093] Optionally, when the SV curve changes, since different types of feedback control loops correspond to different allowable control deviations (error), the allowable control deviation (error) corresponding to the current feedback control loop can be determined by referring to Table 1. Furthermore, the magnitudes of the maximum overshoot (overmax) and the allowable control deviation (error) corresponding to the current feedback control loop can be compared.

[0094] If the maximum overshoot (overmax) is less than or equal to the allowable control deviation (error) corresponding to the current feedback control loop and the maximum overshoot (overmax) is not equal to 0, then the time ratio (adjusttime / tup) of the adjustment time (adjusttime) and change time (tup) corresponding to the current feedback control loop is calculated, and the convergence factor is determined based on the target ratio range in which the time ratio is located.

[0095] For example, when the maximum overshoot (overmax) is less than or equal to the allowable control deviation (error) corresponding to the current feedback control loop and the maximum overshoot (overmax) is not equal to 0, the time ratio (t) of the adjustment time (adjusttime) and change time (tup) corresponding to the current feedback control loop is calculated. Subsequently, if the time ratio (t) is determined to be in the range of 0.5-0.9, the convergence speed is too slow, and the convergence factor (convergencespeed) = (t-0.5)*1.25; if the time ratio (t) is determined to be less than 0.5, the convergence factor (convergencespeed) is 0; if the time ratio (t) is determined to be in the range of 1.1-1.5, the convergence factor (convergencespeed) = (t-1.1)*1.25+0.5; if the time ratio (t) is determined to be greater than 1.5, the time ratio (t) is treated as 1.5, and the convergence factor (convergencespeed) = (1.5-1.1)*1.25+0.5 = 1.

[0096] Furthermore, if the maximum overshoot (overmax) equals 0, indicating no overshoot, it suggests undershoot in the loop and slow adjustment. In this case, obtain the PV adjustment value in the filtered PV curve that is closest to the SV adjustment value in the SV curve, and calculate the second ratio (b) between the SV and PV adjustment values. This second ratio (b) must be between 0 and 1. Therefore, the overshoot factor (overshoot) = b * 0.5. For example, suppose the current PV and SV are both 10. Then, SV is adjusted to 15, and PV slowly increases, reaching a maximum of 13. In this case, the second ratio (b) is (13-10) / (15-10) = 0.6, and the overshoot factor (overshoot) = 0.6 * 0.5 = 0.3.

[0097] It should be noted that if SV is adjusted multiple times, the overshoot factor can be obtained for each adjustment of SV in the manner described above, and then the average of the multiple overshoot factors can be taken as the final overshoot factor.

[0098] Correspondingly, other situations are similar, specifically when SV is adjusted multiple times.

[0099] Furthermore, when the maximum overshoot (overmax) exceeds the allowable control deviation (error) corresponding to the current feedback control loop, overshoot exists. The maximum PV adjustment value of the filtered PV curve is obtained, and the first ratio (a) between the maximum PV adjustment value and the SV adjustment value is calculated. Therefore, the overshoot factor (overshoot) = a - 0.5. For example, if SV changes from 10 to 15, PV, under the control of the controller, will start rising from 10, reach 15, and may continue to rise to 18 before falling back to around 15. Thus, the maximum PV adjustment value is 18, the first ratio (a) = 1.2, and the overshoot factor (overshoot) = 1.2 - 0.5 = 0.7.

[0100] Optionally, when the SV curve remains unchanged, it is determined that all peak amplitudes and all trough amplitudes of the filtered PV curve are within the target range of [SV adjustment value - allowable control deviation error, SV adjustment value + allowable control deviation error]. If at least one peak amplitude or at least one trough amplitude exceeds the range of [SV adjustment value - allowable control deviation error, SV adjustment value + allowable control deviation error], then the maximum overshoot (overmax) and convergence factor (convergencespeed) are determined by the oscillation of PV and MV, specifically:

[0101] If the filtered PV curve lags behind the filtered MV curve in response, it is considered overshoot. The lag can be represented by the absolute phase difference c (in degrees). If the absolute phase difference c is less than 18 degrees, the overshoot factor overshoot = 0.5, meaning there is no overshoot. If the absolute phase difference c is in the range of 18-72 degrees, the overshoot factor overshoot = (c-18) / 108 + 0.5. If the absolute phase difference c exceeds 72 degrees, the overshoot factor overshoot equals 1.

[0102] Simultaneously, to address lag, PID parameters can be optimized by accelerating convergence. If the average oscillation period is greater than the steady-state threshold (stabletime) of the corresponding loop (see Table 1), for loops with an absolute phase difference between 30 and 90 degrees, the convergence speed is (90-c) / 120. Furthermore, for loops with an absolute phase difference less than 30 degrees and at least one peak amplitude or at least one trough amplitude outside the range of [SV adjustment value - allowable control deviation error, SV adjustment value + allowable control deviation error], it is considered an oscillation caused by excessively fast convergence. The third ratio f of the oscillation amplitude vmean to 2*error can be calculated, and the convergence speed is equal to the third ratio f / 2. Also, if the ratio f is greater than 2, it can be considered as 2, meaning that when the ratio f is greater than 2, the convergence speed is 1.

[0103] Step S160: Optimize the proportional gain and integral time in the PID parameters based on the convergence factor and overshoot factor.

[0104] It should be noted that in step S150, in some cases both (i.e., convergence factor and overshoot factor) will be calculated and evaluated, while in others only one of the indicators (i.e., convergence factor or overshoot factor) will be calculated and evaluated. In the latter case, the other indicator that is not included in the calculation and evaluation can be the default value of 0.5.

[0105] It should be understood that the specific process of optimizing the proportional gain and integral time in the PID parameters based on the convergence factor and overshoot factor can be set according to actual needs, and the embodiments disclosed herein are not limited thereto.

[0106] Alternatively, when the SV curve changes, it can be divided into the following four cases:

[0107] When the overshoot factor (overshoot) is less than 0.5 and the convergence factor (convergencespeed) is less than 0.5, the optimized scaling factor (PB) is... new And the optimized integration time Ti new The formula for calculating PB is as follows: new =(0.5+overshoot)×(0.5+convergencespeed)×PB old Ti new = (0.5 + overshoot) × Ti old ;

[0108] In the formula, PB old The scaling factor before optimization; Ti old The integration time before optimization.

[0109] Furthermore, when the overshoot factor is greater than or equal to 0.5 and the convergence factor is less than 0.5, the optimized scaling factor PB... new And the optimized integration time Ti new The formula for calculating PB is as follows: new =(1.5-overshoot)×(0.5+convergencespeed)×PB old Ti new = [1 + 2 × (overshoot - 0.5)] × Ti old .

[0110] Furthermore, when the overshoot factor (overshoot) is less than 0.5 and the convergence factor (convergencespeed) is greater than 0.5, the optimized scaling factor (PB) is... new And the optimized integration time Ti new The formula for calculating PB is as follows: new =[1+2×(convergencespeed-0.5)]×PB old Ti new =(1.5-convergencespeed)×(0.5+overshoot)×Ti old .

[0111] Furthermore, given that the overshoot factor (overshoot) is greater than or equal to 0.5 and the convergence factor (convergencespeed) is greater than or equal to 0.5, the optimized scaling factor (PB) is... new And the optimized integration time Ti new The formula for calculating PB is as follows: new =[1+2×(convergencespeed-0.5)]×PB old Ti new = [1 + 2 × (overshoot - 0.5)] × Ti old .

[0112] Alternatively, when the SV curve does not change, the following four cases can be distinguished:

[0113] Furthermore, when the overshoot factor (overshoot) is less than 0.5 and the convergence factor (convergencespeed) is less than 0.5, the optimized scaling factor (PB) is... new And the optimized integration time Ti new The formula for calculating PB is as follows: new= (0.5 + convergencespeed) × PB old Ti new =(0.5+convergencespeed)×(0.5+overshoot)×Ti old .

[0114] Furthermore, when the overshoot factor is greater than or equal to 0.5 and the convergence factor is less than 0.5, the optimized scaling factor PB... new And the optimized integration time Ti new The formula for calculating PB is as follows: new = (0.5 + convergencespeed) × PB old Ti new =[1+2×(overshoot-0.5)]×(0.5+convergencespeed)×Ti old .

[0115] Furthermore, when the overshoot factor (overshoot) is less than 0.5 and the convergence factor (convergencespeed) is greater than 0.5, the optimized scaling factor (PB) is... new And the optimized integration time Ti new The formula for calculating PB is as follows: new =[1+2×(convergencespeed-0.5)]×PB old Ti new =(1.5-convergencespeed)×(0.5+overshoot)×Ti old .

[0116] Furthermore, given that the overshoot factor (overshoot) is greater than or equal to 0.5 and the convergence factor (convergencespeed) is greater than or equal to 0.5, the optimized scaling factor (PB) is... new And the optimized integration time Ti new The formula for calculating PB is as follows: new =[1+2×(convergencespeed-0.5)]×PB old Ti new = [1 + 2 × (overshoot - 0.5)] × Ti old .

[0117] Therefore, by utilizing the above technical solution, the optimization method for PID parameters of the PID control system disclosed herein has the following advantages:

[0118] Compared with common internal model optimization methods, this disclosure does not rely on step information and will not affect the on-site production process.

[0119] It does not require modeling, does not rely on model accuracy, and uses data analysis methods to optimize PID, making full use of loop type characteristics and current PID parameters, etc.

[0120] The PID parameter optimization in this disclosure is based on the existing PID parameters and their control effects. The optimized PID parameters are dependent on the original PID parameters, making full use of the information of the current state of the loop. It will not generate a new PID parameter out of thin air, and will not produce significant changes in PID parameters during the optimization process, thus ensuring the safety of the optimization process.

[0121] The optimization process disclosed herein is divided into two main parts: evaluation and optimization. The optimization process is decoupled, allowing experienced users to incorporate their own understanding between evaluation and optimization for flexibility. In other words, the engineer's own understanding of the curve is used to replace steps S110 to S150. The entire method consists of two processes: evaluation and tuning. Steps S110 to S150 constitute the evaluation process. The complete process from steps S110 to S160 can be performed with a single click, simultaneously completing both evaluation and tuning (the logic is sequential, but from a user experience perspective, it's a one-time process). The evaluation of the curve shown in the results can also serve as a reference for engineers, who can then modify and re-tune it. Specifically, the following steps can be added between steps S150 and S160: Upon obtaining the evaluation parameters, display them to the user through the interface; if a modification instruction is received from the user, modify the evaluation parameters accordingly; if a confirmation instruction is received from the user, execute step S160.

[0122] It should be understood that the above-described method for optimizing PID parameters of the PID control system is merely exemplary. Those skilled in the art can make various modifications based on the above method, and the modified solutions also fall within the protection scope of this disclosure.

[0123] Second Embodiment

[0124] Please refer to Figure 2, which shows a schematic diagram of a PID control system 200 provided in an embodiment of this disclosure. Specifically, the PID control system 200 includes a controller for executing a method for optimizing PID parameters of the PID control system as described in the first embodiment.

[0125] Those skilled in the art will understand that embodiments of this disclosure can be provided as methods, systems, or computer program products. Therefore, this disclosure can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this disclosure can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0126] Those skilled in the art will understand that embodiments of this disclosure can be provided as methods, systems, or computer program products. Therefore, this disclosure can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this disclosure can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0127] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions.

[0128] It should be noted that the word "a" or "one" preceding a component does not preclude the existence of multiple such components. This disclosure can be implemented using hardware comprising several different components and using a suitably programmed computer. Among the listed devices, several of these devices may be embodied by the same hardware. The use of the terms "first," "second," "third," etc., is merely for convenience and does not indicate any order. These terms can be understood as part of the component names.

[0129] Furthermore, it should be noted that in the description of this specification, the terms "one embodiment," "some embodiments," "embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0130] Although preferred embodiments of this disclosure have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the technical solution should be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this disclosure.

[0131] Obviously, those skilled in the art can make various modifications and variations to this disclosure without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the technical solutions of this disclosure and their equivalents, then this disclosure should also include these modifications and variations.

Claims

1. A method for optimizing PID parameters in a PID control system, characterized in that, include: Collect closed-loop control data of the current feedback control loop; wherein, the closed-loop control data includes PV curve and MV curve; The PV curve and the MV curve are filtered respectively to obtain the filtered PV curve and the filtered MV curve. Polynomial fitting is then performed on the filtered PV curve and the filtered MV curve respectively to obtain the first fitting curve corresponding to the filtered PV curve and the second fitting curve corresponding to the filtered MV curve. The first fitted curve and the second fitted curve are preprocessed respectively to remove abnormal peaks and valleys in the first fitted curve and the second fitted curve, so as to obtain the first fitted curve and the second fitted curve after removing abnormal peaks and valleys. The first and second fitted curves after removing the abnormal peaks and troughs are analyzed to obtain characteristic information for evaluating the control effect of the PID parameters. The control effect of the PID parameters is evaluated based on the feature information to obtain evaluation parameters of the control effect of the PID parameters; wherein, the evaluation parameters include a convergence factor and an overshoot factor. The proportional gain and integral time in the PID parameters are optimized based on the convergence factor and the overshoot factor.

2. The optimization method according to claim 1, characterized in that, The closed-loop control data further includes the SV curve; the feature information includes, but is not limited to, the absolute phase difference between the first fitted curve after removing the abnormal peaks and troughs and the second fitted curve after removing the abnormal peaks and troughs, the change time of the filtered PV curve from the time of change of the SV curve to the change value of the SV curve when the SV curve changes, the maximum overshoot when the SV curve changes, the average oscillation period when the SV curve does not change, and the oscillation amplitude when the SV curve does not change.

3. The optimization method according to claim 2, characterized in that, When the SV curve changes, the evaluation of the control effect of the PID parameters based on the feature information yields evaluation parameters for the control effect of the PID parameters, including: Compare the maximum overshoot with the allowable control deviation corresponding to the current feedback control loop; If the maximum overshoot is less than or equal to the allowable control deviation corresponding to the current feedback control loop and the maximum overshoot is not equal to 0, then the time ratio of the adjustment time and the change time corresponding to the current feedback control loop is calculated, and based on the target ratio range in which the time ratio is located, the convergence factor is determined to be the convergence factor corresponding to the target ratio range. If the maximum overshoot is greater than the allowable control deviation corresponding to the current feedback control loop, then the maximum PV adjustment value of the filtered PV curve is obtained, and the first ratio of the maximum PV adjustment value to the SV adjustment value is calculated. Then, the first ratio is substituted into the formula overshoot factor overshoot = a - 0.5 to obtain the overshoot factor; where a represents the first ratio.

4. The optimization method according to claim 3, characterized in that, When the SV curve changes, the evaluation of the control effect of the PID parameters based on the feature information to obtain the evaluation parameters of the control effect of the PID parameters further includes: If the maximum overshoot is equal to 0, then obtain the PV adjustment value of the SV curve that is closest to the SV adjustment value in the filtered PV curve, and calculate the second ratio between the SV adjustment value and the PV adjustment value. Then, substitute the second ratio into the formula overshoot factor overshoot = b * 0.5 to obtain the overshoot factor; where b represents the second ratio.

5. The optimization method according to claim 2, characterized in that, When the SV curve does not change, the evaluation of the control effect of the PID parameters based on the feature information yields evaluation parameters for the control effect of the PID parameters, including: It is determined that the amplitudes of all peaks and all troughs of the filtered PV curve are within the target range; wherein, the target range is determined by the SV adjustment value in the SV curve and the allowable control deviation corresponding to the current feedback control loop; If it is determined that at least one peak amplitude or trough amplitude in the filtered PV curve exceeds the target range and the average oscillation period is greater than the steady-state threshold corresponding to the current feedback control loop, then the range of the absolute phase difference is determined, and the overshoot factor is determined based on the range of the absolute phase difference. At the same time, the convergence factor is determined based on the absolute phase difference and the oscillation amplitude.

6. The optimization method according to claim 5, characterized in that, Determining the range of the absolute phase difference, and based on the range of the absolute phase difference, includes: If the absolute phase difference is less than 18 degrees, the overshoot factor is 0.

5. If the absolute phase difference is within the range of 18-72 degrees, then substitute the absolute phase difference into the formula overshoot factor overshoot = (c-18) / 108 + 0.5 to obtain the overshoot factor; where c represents the absolute phase difference. If the absolute phase difference c exceeds 72 degrees, the overshoot factor is 1.

7. The optimization method according to claim 5, characterized in that, The determination of the convergence factor based on the absolute phase difference and the oscillation amplitude includes: If the absolute phase difference is between 30 and 90 degrees, then substitute the absolute phase difference into the formula convergencespeed = (90 - c) / 120 to obtain the convergence factor; where c represents the absolute phase difference. If the absolute phase difference is less than 30 degrees, calculate the third ratio of the oscillation amplitude vmean to 2*error, and determine whether the third ratio is greater than 2. If the third ratio is less than or equal to 2, then substitute the third ratio into the convergence factor convergencespeed = f / 2 to obtain the convergence factor; where f is the third ratio. If the third ratio is greater than 2, then the convergence factor is 1.

8. The optimization method according to claim 2, characterized in that, When the SV curve changes, the optimization of the proportional gain and integral time in the PID parameters based on the convergence factor and the overshoot factor includes: If the overshoot factor is less than 0.5 and the convergence factor is less than 0.5, the optimized scaling factor PB new And the optimized integration time Ti new The calculation formula is as follows: PB new =(0.5+overshoot)×(0.5+convergencespeed)×PB old ; Ti new =(0.5+overshoot)×Ti old ; In the formula, PB old The scaling factor before optimization is denoted as 'Ti'; overshoot is the overshoot factor; convergencespeed is the convergence factor; Ti old The integration time before optimization; Alternatively, if the overshoot factor is greater than or equal to 0.5 and the convergence factor is less than 0.5, then the optimized scaling factor PB... new and the optimized integration time Ti new The calculation formula is as follows: PB new =(1.5-overshoot)×(0.5+convergencespeed)×PB old ; Ti new =[1+2×(overshoot-0.5)]×Ti old ; Alternatively, if the overshoot factor is less than 0.5 and the convergence factor is greater than 0.5, then the optimized scaling factor PB... new and the optimized integration time Ti new The calculation formula is as follows: PB new =[1+2×(convergencespeed-0.5)]×PB old ; Ti new =(1.5-convergencespeed)×(0.5+overshoot)×Ti old ; Alternatively, if the overshoot factor is greater than or equal to 0.5 and the convergence factor is greater than or equal to 0.5, then the optimized scaling factor PB... new and the optimized integration time Ti new The calculation formula is as follows: PB new Two [1 + 2×(convergence speed - 0.5)] × PB old ; Ti new =[1+2×(overshoot-0.5)]×Ti old 。 9. The optimization method according to claim 2, characterized in that, When the SV curve remains unchanged, the optimization of the proportional gain and integral time in the PID parameters based on the convergence factor and the overshoot factor includes: If the overshoot factor is less than 0.5 and the convergence factor is less than 0.5, then the optimized scaling factor PB new And the optimized integration time Ti new The calculation formula is as follows: PB new =(0.5+convergencespeed)×PB old ; Ti new =(0.5+convergencespeed)×(0.5+overshoot)×Ti old ; In the formula, PB old The scaling factor before optimization is denoted as 'Ti'; overshoot is the overshoot factor; convergencespeed is the convergence factor; Ti old The integration time before optimization; Alternatively, if the overshoot factor is greater than or equal to 0.5 and the convergence factor is less than 0.5, then the optimized scaling factor PB... new and the optimized integration time Ti new The calculation formula is as follows: PB new =(0.5+convergencespeed)×PB old ; Ti new =[1+2×(overshoot-0.5)]×(0.5+convergencespeed)×Ti old ; Alternatively, if the overshoot factor is less than 0.5 and the convergence factor is greater than 0.5, then the optimized scaling factor PB... new and the optimized integration time Ti new The calculation formula is as follows: PB new =[1+2×(convergencespeed-0.5)]×PB old ; Ti new =(1.5-convergencespeed)×(0.5+overshoot)×Ti old ; Alternatively, if the overshoot factor is greater than or equal to 0.5 and the convergence factor is greater than or equal to 0.5, then the optimized scaling factor PB... new and the optimized integration time Ti new The calculation formula is as follows: PB new =[1+2×(convergencespeed-0.5)]×PB old ; Ti new =[1+2×(overshoot-0.5)]×Ti old 。 10. A PID control system, characterized in that, Includes a controller for performing a method for optimizing PID parameters of a PID control system as described in any one of claims 1 to 9.

Images