Method for performing auxiliary setting on basis of multi-dimensional loop state detection

Abstract

A method for performing auxiliary setting on the basis of multi-dimensional loop state detection. The method comprises: acquiring historical data, performing a quality check on the historical data, and then performing data segmentation (S1); calculating a valve saturation rate on the basis of the segmented historical data, so as to determine a current loop valve state (S2); calculating an excitation index on the basis of the loop valve state and a control mode, and determining a setting mode or calculating an automatic control rate on the basis of the excitation index (S3); calculating the automatic control rate on the basis of historical tag data of the control mode in the segmented historical data, and selecting a setting mode or calculating a stability rate on the basis of the automatic control rate (S4); calculating the stability rate on the basis of loop PV, SV and MV data in the segmented historical data, and determining a current control effect or performing loop oscillation detection on the basis of the stability rate (S5); and determining a setting mode on the basis of loop oscillation detection (S6). By means of performing loop diagnosis detection in terms of control performance, valve problem detection, disturbance, etc., a basis is provided for selecting a parameter setting mode.

Description

A method for auxiliary tuning based on multi-dimensional loop state detection Technical Field

[0001] The present invention relates to the field of automated control systems, and in particular to a method for auxiliary tuning based on multi-dimensional loop state detection. Background Art

[0002] With the development of automation technology, the number of controllers used in process industries has increased, and they are becoming increasingly complex. Various problems may arise after long-term operation of the controllers, necessitating timely loop optimization to ensure normal production operations. However, optimizing the loops or diagnosing and improving existing loop problems is extremely difficult for operators who maintain a large number of loops.

[0003] Existing technologies mostly perform single-dimensional loop status checks, providing little insight into subsequent loop maintenance and optimization. Furthermore, some companies have purchased tuning software to assist operators with loop optimization. This software offers a variety of tuning methods, but field operators lack the ability to accurately select the correct method. Using an inappropriate tuning method can result in parameters being unavailable, or the resulting parameters failing to achieve the desired results. Summary of the Invention

[0004] In order to overcome the shortcomings of the above technologies, the present invention provides a method for auxiliary tuning based on multi-dimensional loop status detection, including loop diagnostic detection in terms of control performance, valve problem detection, disturbance, etc., to help operators discover loop problems in a timely manner, and provide loop maintenance suggestions, and optimize the operation and maintenance of the loop in a targeted manner to improve operation and maintenance efficiency; at the same time, based on the loop status and the excitation index of the detection data segment, it indicates whether the current data segment has obvious characteristic information, providing a basis for the selection of parameter tuning method.

[0005] The technical solution adopted by the present invention to overcome its technical problems is to propose a method for auxiliary tuning based on multi-dimensional loop state detection, including: S1, obtaining historical data and segmenting the data after performing quality inspection on the historical data; S2, calculating the valve saturation rate based on the segmented historical data to determine the current loop valve state; S3, calculating the excitation index based on the loop valve state and the control mode, and determining the tuning method or calculating the automatic control rate according to the excitation index; S4, calculating the automatic control rate based on the bit number historical data of the control mode in the segmented historical data, and selecting the tuning method or calculating the stability rate according to the automatic control rate; S5, calculating the stability rate based on the loop PV, SV and MV data in the segmented historical data, and judging the current control effect or performing loop oscillation detection based on the stability rate; S6, determining the tuning method based on the loop oscillation detection.

[0006] Furthermore, the acquisition of historical data and data segmentation after quality inspection of the historical data specifically includes: S11, acquiring historical data of the loop preset time; S12, performing quality inspection on the loop PV / SV / MV historical data to confirm the quality good value rate, S13, if the good value rate is less than the preset good value threshold, the loop status detection cannot be performed; if the good value rate is greater than the preset good value threshold, the historical data is segmented, wherein the data with good quality code after segmentation is used for loop status detection.

[0007] Furthermore, the valve saturation rate is calculated based on the segmented historical data to determine the current loop valve status, specifically including: S21, calculating the valve saturation rate based on the upper and lower limits of the valve position value range and MV in the segmented historical data; S22, if the valve saturation rate is greater than the preset valve saturation rate threshold, the valve is abnormal; if the valve saturation rate is less than or equal to the preset valve saturation rate threshold, the valve is normal.

[0008] Furthermore, the valve saturation rate is equal to the valve upper limit saturation rate + the valve lower limit saturation rate. The valve upper limit saturation rate is calculated based on the number of MVs greater than or equal to the upper limit of the valve position value range and the total number of data. The valve lower limit saturation rate is calculated based on the number of MVs less than or equal to the lower limit of the valve position value range and the total number of data.

[0009] Furthermore, the excitation index is calculated based on the loop valve state and the control mode, and the adjustment method or the automatic control rate is determined according to the excitation index, specifically including: if the loop valve state is normal, the excitation index is calculated based on the control mode of the loop; if the excitation index is less than a preset excitation threshold, the parameter adjustment method based on the loop model is excluded and the parameters are positively formulated; if the excitation index is greater than or equal to the preset excitation threshold, the parameters are adjusted based on the loop model or the automatic control rate detection is continued.

[0010] Furthermore, the calculation of the excitation index based on the control mode of the loop specifically includes: if the control mode of the loop is manual mode, the excitation index is calculated according to the fluctuation area method of PV when MV changes; if the control mode of the loop is automatic mode, the excitation index is calculated according to the fluctuation area of ​​PV when SV changes.

[0011] Furthermore, the selection of the tuning method or calculation of the stability rate according to the automatic control rate specifically includes: if the automatic control rate is less than the preset automatic control threshold, parameter tuning is performed based on the loop model; if the automatic control rate is greater than or equal to the preset automatic control threshold, the stability rate of the loop is detected.

[0012] Furthermore, the stability rate is calculated based on the loop PV, SV and MV data in the segmented historical data, and the current control effect is judged or loop oscillation detection is performed based on the stability rate.

[0013] The beneficial effects of the present invention are:

[0014] 1. Conduct multi-dimensional and comprehensive detection of the circuit based on the circuit operation history data.

[0015] 2. Provide loop status detection suggestions by testing the actuator, control performance, whether the data itself has obvious characteristics and whether there are oscillations.

[0016] 3. Provide support for the selection of parameter tuning methods based on the loop vibration state and the excitation index of the detection data segment, helping operators to select the appropriate tuning method and obtain the best parameter tuning results. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] FIG1 is a flow chart of a method for performing auxiliary tuning based on multi-dimensional loop state detection according to an embodiment of the present invention;

[0018] FIG2 is a schematic diagram showing the principle of a method for performing auxiliary tuning based on multi-dimensional loop state detection according to an embodiment of the present invention.

[0019] FIG3 is a schematic diagram of calculating an incentive index according to an embodiment of the present invention;

[0020] FIG4 is a comparative schematic diagram of converting historical data in the time domain into frequency domain information according to an embodiment of the present invention. DETAILED DESCRIPTION

[0021] In order to further understand the present invention, some of the terms mentioned in the present invention are first explained:

[0022] PV: Process Variable, measured value;

[0023] SV:Set Variable, set value;

[0024] MV: Manipulated Variable, the output value after controller calculation in automatic mode;

[0025] Control Mode: Control Mode, manual and automatic mode of the loop;

[0026] Quality code: Tag Quality, a sign indicating whether the bit number is good.

[0027] In order to facilitate those skilled in the art to better understand the present invention, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments. The following is only exemplary and does not limit the scope of protection of the present invention.

[0028] As shown in FIG1 and FIG2 , a flow chart of a method for auxiliary tuning based on multi-dimensional loop state detection according to this embodiment includes the following steps.

[0029] S1, obtain historical data and perform data segmentation after quality inspection of the historical data;

[0030] In one embodiment of the present invention, historical data of a circuit operating over a period of time is obtained through a data platform. Each data item on the data platform carries a quality code to identify whether the data is good. Based on the quality code, the historical data of the circuit PV, SV, and MV are first quality-checked to determine the good value rate. The specific formula for the good value rate is shown in Formula (1).

[0031] If the good value rate is less than the preset good value threshold, it is considered that the current historical data segment has less valid data and the next loop status detection cannot be performed. It is recommended that the operator first check the data source to see if the communication is disconnected.

[0032] If the good value rate meets the quality requirement, that is, is greater than the preset good value threshold, the corresponding historical data will be segmented, and the segmented good quality code data will be used for the subsequent loop status monitoring process.

[0033] S2, calculates the valve saturation rate based on the segmented historical data to determine the current loop valve status.

[0034] In one embodiment of the present invention, the valve saturation rate is calculated based on the obtained historical data of the upper and lower limits of the valve position value range and the MV historical data to confirm whether the current loop valve is normal. The specific formulas are shown in formulas (2) to (4).

[0035] Valve saturation rate = valve upper limit saturation rate + valve lower limit saturation rate (4)

[0036] If the upper and lower limits of the valve position range are equal, only the upper limit of the valve saturation rate is calculated. If the valve saturation rate exceeds the preset valve saturation rate threshold, it is considered that there is a problem with the valve in the current circuit. Based on the current manual or automatic state of the circuit, appropriate suggestions are given. Whether in manual or automatic mode, on-site personnel are prompted to inspect the valve to determine if there is any abnormality. If the valve saturation rate is less than or equal to the preset valve saturation rate threshold, the subsequent incentive index calculation is performed.

[0037] S3, calculating the excitation index based on the loop valve state and the control mode, and determining the setting method and adjusting the loop data according to the excitation index until the excitation index is greater than the preset excitation threshold.

[0038] Some tuning methods require model identification based on loop operating data to derive the current system model, and then derive PID parameters based on the system model. Obtaining more accurate parameters requires an accurate model, and calculating an accurate model requires that the current loop operating data segment have distinct characteristics. When the excitation index is greater than the preset excitation threshold, the current data is considered to have distinct characteristics, allowing for model-based tuning.

[0039] In one embodiment of the present invention, whether the current loop set point or MV fluctuates and the magnitude of the fluctuation are determined mainly based on whether the control mode is automatic or manual.

[0040] The excitation index is calculated using the area method. For example, in manual control mode, as shown in Figure 3, the excitation index is calculated based on the PV fluctuation area (②-①-③) when the MV changes. ① represents the area outside the PV's steady state; ② represents the area within the steady state from the time the PV begins to change until it reaches steady state; and ③ represents the area where the PV changes in the opposite direction to the actual direction after the MV changes.

[0041] For example, if the control mode of the loop is automatic mode, the excitation index is calculated based on the fluctuation area of ​​PV when SV changes.

[0042] In yet other embodiments, the incentive index is adjusted accordingly based on empirical parameters.

[0043] If the current data features are not obvious, that is, the excitation index is less than the preset excitation threshold, model-based tuning is not recommended, and a tuning method without model identification should be used. If the current data features are obvious, proceed to the next step of calculation.

[0044] S4, calculating the automatic control rate based on the bit number historical data of the control mode in the segmented historical data, and selecting a setting method or calculating the stabilization rate according to the automatic control rate.

[0045] The control mode bit number historical data acquired from the data is also the part of the good quality code in the segmented historical data, where the control mode is the number of automatic ones. As shown in formula (5)

[0046] The "auto" count is the number of data with a good quality code corresponding to the control mode "auto" in the acquired control mode bit number historical data, and the "total data" is the total number of data with a good quality code in the control mode bit number historical data. If the automatic control rate is less than the preset automatic control threshold, model-based parameter tuning is performed directly. If the automatic control rate is greater than the preset automatic control threshold, the loop stability rate is further tested.

[0047] S5, calculates the stability rate based on the loop PV, SV and MV data in the segmented historical data, and judges the current control effect or performs loop oscillation detection based on the stability rate.

[0048] Select the stable judgment logic based on the size of the automatic control rate. If the automatic control rate is less than the preset automatic control rate threshold, it is necessary to judge and Determine whether the current point is in a stable state, otherwise according to the formula |pv[i]-sv[i]|≤Δpv and Determine whether the current point is in a stable state. Among them, Δpv and Δmv are set to appropriate values ​​according to the loop type. The weighted average is calculated based on historical data. Finally, the loop stability rate is obtained, as shown in formula (6).

[0049] The total number of data is 1000, and i represents the number from 0 to 1000. The number of points that meet the stable state is counted as the stable number.

[0050] If the loop stability rate is greater than or equal to the loop stability preset threshold, it can be considered that the control effect of the current parameters is already very good. It is recommended that on-site personnel can choose a parameter setting method based on fine-tuning of existing parameters or maintain the status quo.

[0051] If the loop stability rate is less than the loop stability preset threshold, the next step of loop oscillation detection will be carried out.

[0052] Generally, the loop stability preset threshold is set to above 80%.

[0053] S6, determining the tuning method based on loop oscillation detection.

[0054] Oscillation may be caused by external disturbances or poor loop parameters. If the loop oscillation index is large, there may be obvious and regular external disturbances. In this case, simply adjusting the PID parameters may not be effective. It is recommended that on-site personnel work with process personnel to check whether the current control strategy needs to be optimized.

[0055] If the loop's oscillation index is small, you can try fine-tuning the parameters using anthropomorphic parameter tuning methods. Specifically, using Fourier transform, historical data in the time domain is converted into frequency domain information. Prominent peaks appear when there are obvious regular external fluctuations. By observing the amplitude and location of these peaks in the frequency domain, the oscillation situation can be determined. As shown in Figure 4, a larger oscillation index indicates stronger oscillations.

[0056] The present invention utilizes a segment of loop operation data to perform loop status detection based on multiple dimensions, including first determining whether there is a problem with the actuator in the loop, and then determining whether the loop control performance is low and requires parameter tuning. At the same time, the loop oscillation area method and the excitation index of the data are calculated. The results of the loop status detection provide support for the selection of the parameter tuning method, helping the operator to select the tuning method and obtain the best parameter tuning result.

[0057] It should be noted that in other embodiments, the steps of the corresponding method are not necessarily performed in the order shown and described in this specification. In some other embodiments, the method may include more or fewer steps than those described in this specification. In addition, a single step described in this specification may be broken down into multiple steps for description in other embodiments, and multiple steps described in this specification may be combined into a single step for description in other embodiments.

Claims

1. A method for auxiliary setting based on multi-dimensional loop state detection, characterized in that, Including: S1, obtaining historical data, performing quality inspection on the historical data, and then segmenting the data; S2, calculating the valve saturation rate based on the segmented historical data to determine the current loop valve status; S3, calculating the excitation index based on the loop valve status and control mode, and determining the tuning method or calculating the automatic control rate according to the excitation index; S4, calculating the automatic control rate based on the tag number historical data of the control mode in the segmented historical data, and selecting the tuning method or calculating the smoothness rate according to the automatic control rate; S5, calculating the smoothness rate based on the loop PV, SV, and MV data in the segmented historical data, and judging the current control effect or performing loop oscillation detection based on the smoothness rate; S6, determining the tuning method based on the loop oscillation detection.

2. The method for auxiliary setting based on multi-dimensional loop state detection according to claim 1, characterized in that The obtaining of historical data, performing quality inspection on the historical data, and then segmenting the data specifically includes: S11, obtaining the historical data of the loop for a preset time; S12, performing quality inspection on the loop PV / SV / MV historical data to confirm the good value rate; S13, if the good value rate is less than the preset good value threshold, the loop status detection cannot be performed. If the good value rate is greater than the preset good value threshold, the historical data is segmented, and the data with good quality codes after segmentation is used for loop status detection.

3. A method for auxiliary setting based on multi-dimensional loop state detection according to claim 1, characterized in that, Calculating the valve saturation rate based on the segmented historical data to determine the current loop valve status specifically includes: S21, calculating the valve saturation rate based on the upper and lower limits of the valve position value range and MV in the segmented historical data; S22, if the valve saturation rate is greater than the preset valve saturation rate threshold, the valve is abnormal. If the valve saturation rate is less than or equal to the preset valve saturation rate threshold, the valve is normal.

4. A method for auxiliary tuning based on multi-dimensional loop state detection according to claim 3, characterized in that The valve saturation rate is equal to the upper valve saturation rate + the lower valve saturation rate. The upper valve saturation rate is calculated based on the number of MVs greater than or equal to the upper limit of the valve position value range and the total number of data. The lower valve saturation rate is calculated based on the number of MVs less than or equal to the lower limit of the valve position value range and the total number of data.

5. The method for auxiliary setting based on multi-dimensional loop state detection according to claim 3, wherein Calculating the excitation index based on the loop valve status and control mode, and determining the tuning method or calculating the automatic control rate according to the excitation index specifically includes: If the loop valve status is normal, calculate the excitation index based on the control mode of the loop; If the excitation index is less than the preset excitation threshold, select a parameter tuning method other than the method of tuning parameters based on the loop model; If the excitation index is greater than or equal to the preset excitation threshold, perform parameter tuning based on the loop model or continue with the automatic control rate detection.

6. The method for auxiliary setting based on multi-dimensional loop state detection according to claim 5, wherein The calculating of the excitation index based on the control mode of the loop specifically includes: If the control mode of the loop is the manual mode, calculate the excitation index according to the fluctuation area of PV when MV changes; If the control mode of the loop is the automatic mode, calculate the excitation index according to the fluctuation area of PV when SV changes.

7. A method for auxiliary tuning based on multi-dimensional loop state detection according to claim 5, characterized in that, Selecting the tuning method or calculating the smoothness rate according to the automatic control rate specifically includes: If the automatic control rate is less than the preset automatic control threshold, perform parameter tuning based on the loop model; If the automatic control rate is greater than or equal to the preset automatic control threshold, detect the smoothness rate situation of the loop.

8. A method for auxiliary tuning based on multi-dimensional loop state detection according to claim 5, characterized in that Calculate the stability rate based on the loop PV, SV, and MV data in the segmented historical data, and judge the current control effect or detect loop oscillation based on the stability rate. Select the stability judgment logic based on the self-control rate. The stability judgment logic is calculated based on the loop PV, SV, and MV data in the segmented historical data, and is used to judge whether each data point in the segmented historical data is in a stable state. Calculate the loop stability rate based on the number of data points in the stable state and the number of data in the segmented historical data. If the loop stability rate is greater than or equal to the loop stability preset threshold, no parameter tuning is required. If the loop stability rate is less than the loop stability preset threshold, loop oscillation detection is performed.

9. A method for assisting setting based on multi-dimensional loop state detection of data according to claim 5, characterized in that Determine the tuning method based on the loop oscillation detection, specifically including: If the oscillation index of the loop exceeds the preset vibration index threshold, tune the PID parameters in combination with the optimal control strategy. If the oscillation index of the loop is less than the preset vibration index threshold, use the parameter tuning method with anthropomorphic characteristics to tune the parameters.

Images