A method of pid loop performance rating and diagnosis based on wave variability coefficient

Abstract

The application provides a PID loop performance rating and diagnosis method based on fluctuation coefficient of variation, which is used to solve the problem of PID control loop performance decline and parameter re-tuning in industrial production. The method collects process variable (PV), set value (SV), controller output (MV) and mode point (MODE) data through DCS system, and calculates core performance indexes such as fluctuation coefficient of variation (CV), relative performance index (RPI) and autocorrelation coefficient after data preprocessing. The fluctuation coefficient of variation CV is a dimensionless quantity, which can effectively quantify the data volatility and overcome the limitations of traditional evaluation methods in multi-dimensional scenarios. The system rates the loop according to these indexes, which is divided into four levels: excellent, good, general and poor, and provides fault diagnosis and optimization suggestions for the general or poor loop. The method also includes establishing a loop rating history library to track the optimization effect. The application improves the real-time performance evaluation accuracy and diagnosis efficiency of the PID loop, and is suitable for the field of industrial automation control.

Description

Technical Field

[0001] This invention relates to the field of industrial production control technology, and in particular to a method for performance rating and diagnosis of PID loops based on fluctuation variation coefficient. Background Technology

[0002] Statistics show that over 95% of existing automatic control loops employ PID control strategies. While these controllers often exhibit good performance during the initial commissioning phase, as operating conditions fluctuate and process parameters change, the PID parameters gradually become inadequate for the new conditions. If these parameters are not adjusted in a timely manner, the loop's control performance will gradually deteriorate, impacting production efficiency.

[0003] Currently, most factories use the automation rate of the control loop as the standard to judge the control performance of a loop. Some use the Harris index and the minimum variance method for evaluation. The former can only assess the current automation level of the workshop and cannot judge whether the control performance of a single loop is good or bad. The latter evaluates from multiple perspectives through performance index calculation, but because the algorithm has certain defects, it cannot compare multi-dimensional situations and needs to refer to the average value, which is easily affected when there are multiple average values.

[0004] Furthermore, for loops with poor control performance and requiring parameter tuning, engineers often struggle to obtain effective information and fail to identify problems promptly. Maintenance is frequently delayed until production efficiency is affected, leading to economic losses. Therefore, it is essential to periodically conduct performance analysis, rating, and diagnosis of PID control loops in the DCS system using external methods.

[0005] Given the current situation, it is essential to propose a new method for evaluating the performance of PID loop control in industrial production, given the poor performance of existing PID loop control. Summary of the Invention

[0006] This invention proposes a PID loop performance rating and diagnosis method based on the fluctuation coefficient of variation, which overcomes the shortcomings of existing methods in multi-dimensional scenarios. This method collects relevant data from the PID control loop through a DCS system, performs data preprocessing, calculates core performance indicators, and diagnoses based on the loop rating results, providing engineers with a basis for process optimization.

[0007] To achieve the above objectives, the specific implementation plan is as follows:

[0008] Data collection

[0009] Based on the DCS system, process variable (PV), setpoint (SV), controller output (MV), and mode point (MODE) data of the control loop to be evaluated are collected in real time to ensure the real-time nature and comprehensiveness of the data.

[0010] Data preprocessing

[0011] The collected data is filtered to remove defective values, and missing data is supplemented to ensure data quality.

[0012] Performance index calculation

[0013] Based on the preprocessed data, the following three core performance indicators are calculated to rate the PID loop:

[0014] Fluctuation coefficient of variation (CV)

[0015] The coefficient of variation (CV) is obtained by calculating the ratio of the standard deviation to the mean of the process variable. The formula for calculating CV is as follows:

[0016]

[0017] PV mean Err represents the average value of PV. sd The standard deviation represents the error.

[0018] The formula for calculating the autocorrelation coefficient (RPI) is:

[0019] Introducing the PV autocorrelation coefficient, denoted by ρ:

[0020]

[0021] The relative performance index (PV) is a good indicator of a controller's response speed relative to a reference (expected steady-state time). A higher PV value indicates a faster response, while a lower PV value indicates a slower response. A value close to 1 represents the expected response speed. To clearly express the degree of oscillation in the control loop, the autocorrelation coefficient of PV is chosen to represent the correlation between a set of data and its predecessor (the data itself).

[0022] PV relative performance index, expressed in P r express:

[0023]

[0024] Where Tr desired Tr represents the expected steady-state time. actual This represents the actual steady-state time.

[0025] Circuit rating

[0026] Based on the above performance indicators, a loop rating table is used to rate the control loops, classifying them into four levels: Excellent, Good, Average, and Poor. The weighting of the loop rating table primarily focuses on the coefficient of variation (CV) to ensure the accuracy of the ratings.

[0027] Fault diagnosis and optimization suggestions

[0028] For circuits with average or poor performance, the system automatically generates diagnostic suggestions based on performance indicators and a diagnostic information database. This database is established using field process experience and parameter thresholds. The system combines rating results and calculated data to automatically match fault codes and provide diagnostic opinions for engineers' reference.

[0029] Circuit rating history database

[0030] Establish a historical loop rating database to record the historical rating information of loops, so as to facilitate the comparison and evaluation of subsequent optimization effects.

[0031] Open-loop state determination

[0032] When the automatic control rate of the control loop is less than 60%, the system defaults to judging the loop as an open loop and does not participate in the rating.

[0033] Beneficial effects

[0034] The PID loop performance rating and diagnosis method based on fluctuation variation coefficient proposed in this invention has the following advantages:

[0035] Performance evaluation is performed using the dimensionless fluctuation coefficient of variation (CV), which overcomes the limitations of traditional methods in multi-dimensional scenarios.

[0036] Automated diagnostics and optimization suggestions reduce the workload of engineers and lower maintenance costs;

[0037] Custom performance thresholds can be set for different types of control loops to further improve rating accuracy;

[0038] The established historical database of loop ratings provides a reference for long-term control loop optimization.

[0039] Through the above steps, this invention effectively improves the performance evaluation accuracy and diagnostic efficiency of PID control loops, and can be widely used in the field of industrial control to help factories improve production efficiency and reduce economic losses. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a schematic diagram of the system architecture of the present invention;

[0042] Figure 2This is a schematic diagram of the process of the present invention;

[0043] Figure 3 This is a schematic diagram of some fault codes and diagnostic information of the present invention. Detailed Implementation

[0044] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should also be noted that, to make the embodiments more comprehensive, the following embodiments are the best and preferred embodiments, and those skilled in the art can use other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0045] It should be noted that the use of terms such as "an embodiment," "an embodiment," "an exemplary embodiment," and "some embodiments" in the specification indicates that the described embodiment may include a specific feature, structure, or characteristic, but not every embodiment necessarily includes that specific feature, structure, or characteristic. Furthermore, when a specific feature, structure, or characteristic is described in connection with an embodiment, implementing such a feature, structure, or characteristic in conjunction with other embodiments (whether explicitly described or not) should be within the knowledge of those skilled in the art.

[0046] Generally, terms can be understood at least partly from their use in context. For example, depending at least partly on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in a singular sense, or a combination of features, structures, or characteristics in a plural sense. Additionally, the term "based on" can be understood not necessarily to convey an exclusive set of factors, but rather, alternatively, depending at least partly on the context, to allow for the presence of other factors that are not necessarily explicitly described.

[0047] Example 1

[0048] To make the implementation scheme and advantages of this method clearer, the invention will be further explained below with reference to flowcharts and implementation examples:

[0049] In existing factory DCS systems, over 95% of the loops are PID control loops. With changes in operating conditions and unforeseen factors, control parameters gradually become incompatible, leading to a gradual deterioration in control performance. Engineers often fail to identify and resolve these issues in a timely manner, easily resulting in economic losses. By conducting real-time evaluation and rating of loop control effectiveness, problems can be diagnosed for loops with poor ratings, allowing for timely resolution of issues and optimization of production processes.

[0050] This invention provides a method for performance rating and diagnosis of PID loop control based on fluctuation coefficient of variation. This embodiment is applied to the synthesis furnace section of the chlor-alkali process, including:

[0051] The tag numbers (PV, SV, MV, MODE) of all required control loops are filtered and collected into the database via the OPC server.

[0052] A filter is set up to preprocess the data, removing bad values ​​and replacing them.

[0053] Before calculating performance metrics, threshold rules are set for the performance metrics as a rating reference.

[0054] Select a time interval for loop data. The timestamp should generally be at least 12 hours ago, the time interval should be no less than 6 hours, and the number of data for a single tag should be no less than 150. Otherwise, the calculation results will not be meaningful.

[0055] Performance metrics were calculated for the loop data within the aforementioned time interval:

[0056] Specifically, the error, average error, and average PV for each time period are calculated according to the data sampling frequency, and the steady-state time is calculated.

[0057] Introducing the PV fluctuation coefficient of variation as a key indicator, denoted by CV:

[0058]

[0059] PV mean Err represents the average value of PV. sd The standard deviation represents the error.

[0060] The coefficient of variation (CV) represents the proportion by which data fluctuates around the mean. A larger value indicates greater volatility, and since this volatility is measured against the mean, it eliminates the problem of dimensionality. The CV does not become meaningless simply because some data points are exceptionally large. In short, the CV value is a measure of relative volatility.

[0061] Introducing the PV relative performance index, using P r express:

[0062]

[0063] Where Tr desired Tr represents the expected steady-state time. actual This represents the actual steady-state time.

[0064] The relative performance index is a good measure of a controller's response speed relative to a benchmark (expected steady-state time). A higher value indicates a faster response, a lower value indicates a slower response, and a value close to 1 indicates the expected response speed.

[0065] Introducing the PV autocorrelation coefficient, denoted by ρ:

[0066]

[0067] To clearly express the degree of oscillation in the control loop, the autocorrelation coefficient of PV is chosen to represent the correlation between a set of data and the data before and after it (with itself).

[0068] Understandably, in addition to the three main performance indicators (KPIs), it is also necessary to calculate the over-limit conditions of process variable PV and controller output MV, as well as the change frequency of setpoint SV and mode point MODE, which will help in the analysis of loop conditions.

[0069] Understandably, prolonged out-of-limit process variable PV indicates poor controller performance; prolonged out-of-limit controller output MV and prolonged fully open or fully closed valve positions indicate excessive saturation, requiring inspection of valve dimensions and control parameters; frequent changes in setpoint SV indicate frequent fluctuations in operating conditions, or require inspection of the loop control strategy to see if it is cascade control; frequent changes in mode point MODE indicate frequent manual / automatic switching of the loop, requiring inspection of the control strategy.

[0070] Based on the performance index calculation results, taking the liquid level circuit type as an example, the circuit performance is rated by referring to the circuit rating table.

[0071]

[0072] As shown in the table above, the loop rating is mainly determined by P. r The three main performance indicators are ρ, CV, and P, with CV having the largest weight. r The weight is the same as ρ.

[0073] Understandably, a loop rating is only excellent or good when the fluctuation coefficient of variation (CV) meets the conditions; when the fluctuation coefficient of variation (CV) does not meet the conditions, the loop rating is average or poor, and diagnostic optimization is required.

[0074] The circuit is rated as excellent only when all three performance indicators meet the requirements.

[0075] When the fluctuation coefficient of variation (CV) meets the condition, the relative performance index (P) r When either the autocorrelation coefficient ρ or the other meets the condition, the loop rating is good.

[0076] When the fluctuation coefficient of variation (CV) does not meet the condition, the relative performance index (P) r When both the autocorrelation coefficient ρ and the condition are met, the loop rating is general, and the loop needs to be diagnosed and optimized.

[0077] When the fluctuation coefficient of variation (CV) does not meet the condition, the relative performance index (P) rWhen only one of the autocorrelation coefficients ρ satisfies the condition, or neither satisfies the condition, the loop rating is poor, and the loop needs to be diagnosed and optimized.

[0078] When the loop self-control rate is ≤60%, the loop is considered to be in an open-loop state and will not be included in the rating.

[0079] A fault code database and a corresponding diagnostic information database are established based on on-site processes, production experience, and parameter thresholds. Based on the analysis results of key performance indicators and tag number parameters, corresponding fault information is matched against the database, and corresponding diagnostic opinions are displayed.

[0080] The system automatically collects all loops rated as average and poor, along with their corresponding diagnostic information, and displays the data in reports on the server. This allows engineers to inspect instruments or adjust loop control parameters.

[0081] The optimized control loop is re-evaluated in the next cycle, and the new evaluation is compared with the historical evaluation to verify the optimization results.

[0082] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0083] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for performance rating and diagnosis of PID loops based on fluctuation coefficient of variation, characterized in that, Includes the following steps: S1. Collect process variables, set values, controller outputs, and mode point data of the control loop based on the DCS system; S2. Preprocess the collected data, including removing bad values ​​and data supplementation; S3. Calculate performance indicators based on the data, including the coefficient of variation, relative performance index, and autocorrelation coefficient; S4. Use the fluctuation variation coefficient as the core rating indicator to evaluate the performance level of the control loop, and combine the fluctuation variation coefficient with the relative performance index and autocorrelation coefficient to form a loop rating table. S5. For control loops of general and poor quality, output diagnostic suggestions based on the diagnostic information database; S6. Establish a historical database of loop ratings to record changes in loop ratings; The formula for calculating the fluctuation coefficient of variation is as follows: ； in Represents the average value of PV. The standard deviation of the error, the PV relative performance index, is used... express: ； in Indicates the expected steady-state time. This represents the actual steady-state time.

2. The method for performance rating and diagnosis of PID loops based on fluctuation coefficient of variation according to claim 1, characterized in that, The formula for calculating the autocorrelation coefficient is as follows: Introducing the PV autocorrelation coefficient, using express: ； The relative performance index can effectively evaluate the controller's response speed relative to a benchmark. A larger value indicates a faster response speed, while a smaller value indicates a slower response speed. A value close to 1 indicates the desired response speed. To clearly express the degree of oscillation in the control loop, the autocorrelation coefficient of PV is chosen to represent the correlation between a set of data.

3. The method for performance rating and diagnosis of PID loops based on fluctuation coefficient of variation according to claim 1, characterized in that, The loop rating table classifies control loops into four levels: excellent, good, average, and poor, based on the fluctuation variation coefficient (CV), relative performance index (RPI), and autocorrelation coefficient.

4. The method for performance rating and diagnosis of PID loops based on fluctuation coefficient of variation according to claim 1, characterized in that, The diagnostic information database includes fault codes and diagnostic information, and automatically outputs corresponding diagnostic opinions based on the calculation results of various performance indicators for engineers' reference.

5. The method for performance rating and diagnosis of PID loops based on fluctuation coefficient of variation according to claim 1, characterized in that, When the self-control rate of the control loop is ≤60%, it is automatically determined to be in an open-loop state and will not be included in the rating.

6. The method for performance rating and diagnosis of PID loops based on fluctuation coefficient of variation according to claim 1, characterized in that, Further measures include establishing a historical loop rating database to record historical rating information for each loop, providing optimization references.

Images