Early warning method and device in equipment operation process

Abstract

The invention relates to an early warning method and device in an equipment operation process. According to the scheme provided by the invention, the normal operation interval can dynamically change according to the change of the actually acquired characteristic variable, namely, the interval can change according to the actual equipment operation parameters, so that efficient and accurate alarm and early warning are realized, the monitoring efficiency of a control system is improved, and the safe and stable operation of the equipment is ensured. Moreover, according to the scheme of the invention, a statistical method and a model construction method such as a machine learning algorithm are flexibly applied, and an intelligent early warning technology capable of automatically identifying and adapting to different working conditions of equipment is constructed in combination with historical operation data, so that the defects of an equipment warning function in the aspects of practicability, timeliness and the like are overcome; and the requirements of efficient monitoring, operation and maintenance of an industrial process control system are met.

Description

Technical Field

[0001] The present application relates to the technical field of industrial process control systems, and in particular to an early warning method and device during equipment operation. Background Art

[0002] At present, with the development of technology, the scale and number of monitoring equipment in industrial process control systems such as thermal power, hydropower, and chemical industry are gradually increasing, bringing new challenges to the operation and maintenance of equipment. Since some important operating parameters of equipment, such as equipment temperature, pressure, vibration parameters, etc., can intuitively reflect the operating status of equipment, when equipment is abnormal, problems can be discovered in time by monitoring and analyzing these important operating parameters, thereby avoiding accidents, suppressing the scope of impact of problems, and improving the safety and stability of system operation.

[0003] However, due to the lack of analysis and evaluation methods for the operating parameters of these important equipment, usually only simple threshold values ​​are set in the monitoring system to achieve the limit-crossing alarm. In order to reduce the number of false alarms in the system, the threshold values ​​are often set too large, resulting in the fault having occurred when the alarm is issued, and the effect of alarm prevention and suppression of the problem is not achieved. At the same time, since the equipment has a variety of working conditions during operation, the simple threshold value cannot adapt to various working conditions, and does not make full use of historical data to achieve refined alarms and early warnings, which makes it difficult to meet the needs of industrial process control systems for monitoring and operation and maintenance. Summary of the invention

[0004] In view of the problems existing in the prior art and to improve the safety of equipment operation, the inventor believes that the better solution is the intelligent alarm and early warning technology based on the identification of equipment operation status. Therefore, this application proposes an intelligent early warning technology for equipment operation parameters, which flexibly uses statistical methods and model building methods such as machine learning algorithms, and combines historical operation data to build an intelligent early warning technology that can automatically identify and adapt to different equipment working conditions, making up for the shortcomings of the equipment alarm function in terms of practicality and timeliness, and meeting the needs of efficient monitoring and operation and maintenance of industrial process control systems.

[0005] According to a first aspect of the present application, there is provided an early warning method during equipment operation, characterized in that it includes:

[0006] Determining characteristic variables from equipment operating parameters;

[0007] Constructing a standard value estimation model for the key operating parameter based on the sample data of the key operating parameter and the sample data of the characteristic variable;

[0008] Using a clustering algorithm on the sample data of the characteristic variable to identify the equipment operating condition to which the sample data of the key operating parameter corresponding to the sample data of the characteristic variable belongs;

[0009] Calculating the standard deviation of the sample data of the key operating parameters belonging to the same equipment operating condition, and determining the standard deviation of the key operating parameters corresponding to each equipment operating condition;

[0010] Inputting the characteristic variables acquired in real time into the standard value estimation model to determine the real-time standard value of the key operating parameter;

[0011] Determine the real-time standard deviation of the key operating parameter according to the current equipment operating condition to which the real-time acquired characteristic variable belongs and the standard deviation of the key operating parameter corresponding to each equipment operating condition; and

[0012] A normal operating range of the key operating parameter is determined according to the real-time standard value and the real-time standard deviation.

[0013] Wherein, determining the characteristic variable from the equipment operating parameters includes:

[0014] Determining the correlation between the key operating parameters and the non-key operating parameters in the equipment operating parameters; and

[0015] A preset number of parameters are determined from the non-key operating parameters as characteristic variables according to the correlation.

[0016] Wherein, determining the correlation between the key operating parameters and the non-key operating parameters in the equipment operating parameters includes:

[0017] Each non-key concern operating parameter is traversed, and the correlation between the key concern operating parameter and the non-key concern operating parameter is calculated.

[0018] Wherein, determining the correlation between the key operating parameters and the non-key operating parameters in the equipment operating parameters includes:

[0019] (a) determining the first delay moment among the set number of delay moments as the current delay moment, wherein the delay moment is the delay time between the non-focused operating parameter and the focused operating parameter;

[0020] (b) for any selected parameter among the non-focused operating parameters, calculating the correlation between the selected parameter and the focused operating parameter at the current delay time;

[0021] (c) if the current delay time is not the last delay time among the set number of delay times, determine the next delay time as the current delay time, and return to step (b);

[0022] (d) when the current delay time is the last delay time, determining the delay time corresponding to the maximum correlation among the set number of correlations as the set delay time for calculating the correlation between the selected parameter and the key operating parameter; and

[0023] (e) calculating the correlation between the selected parameter and the key operating parameter according to the set delay time.

[0024] The step of using a clustering algorithm on the sample data of the characteristic variables to identify the equipment operating condition to which the sample data of the key operating parameters corresponding to the sample data of the characteristic variables belong includes:

[0025] Using a clustering algorithm on the sample data of the characteristic variables to determine the equipment operating conditions corresponding to the sample data of the characteristic variables; and

[0026] According to the equipment operating conditions corresponding to the sample data of the characteristic variables, the equipment operating condition to which the sample data of the key operating parameter corresponding to the sample data of the characteristic variables belongs is determined.

[0027] The step of using a clustering algorithm on the sample data of the characteristic variables to determine the equipment operating conditions corresponding to the sample data of the characteristic variables includes:

[0028] (f) randomly selecting samples of the number of types of equipment operating conditions from the sample data of the characteristic variables as current cluster centers;

[0029] (g) calculating the distance between the sample data of the feature variable and each of the current cluster centers of the number of categories;

[0030] (h) classifying each of the sample data of the feature variable into the cluster class to which the current cluster center closest to the feature variable belongs, so as to form the number of cluster classes;

[0031] (i) calculating the centroid of each cluster according to the sample data of the characteristic variable contained in each of the clusters of the number of categories, and obtaining the centroids of the number of categories;

[0032] (j) when the difference between any one of the number of centroids of the categories and the corresponding current cluster center exceeds a preset threshold, the number of centroids of the categories is determined as the current cluster center, and the process returns to step (g); and

[0033] (k) When the difference between any one of the number of centroids of the type and the corresponding current cluster center does not exceed a preset threshold, the number of centroids of the type is determined as the cluster center corresponding to the equipment operating condition.

[0034] The early warning method is characterized in that it also includes:

[0035] Obtaining the measured value of the key operating parameter;

[0036] When the measured value exceeds the normal operating range, an alarm signal is output.

[0037] According to a second aspect of the present application, there is provided an early warning device during equipment operation, characterized in that it comprises:

[0038] A characteristic variable determination module, used for determining characteristic variables from equipment operation parameters;

[0039] An estimation model building module, used to build a standard value estimation model of the key operating parameter according to the sample data of the key operating parameter and the sample data of the characteristic variable;

[0040] An operating condition identification module, used for applying a clustering algorithm to the sample data of the characteristic variable to identify the equipment operating condition to which the sample data of the key operating parameter corresponding to the sample data of the characteristic variable belongs;

[0041] A working condition standard deviation determination module, used to calculate the standard deviation of the sample data of the key operating parameters belonging to the same equipment working condition, and determine the standard deviation of the key operating parameters corresponding to each equipment working condition;

[0042] A real-time standard value determination module, used for inputting the characteristic variables acquired in real time into the standard value estimation model to determine the real-time standard value of the key operating parameter;

[0043] A real-time standard deviation determination module, configured to determine the real-time standard deviation of the key operating parameter according to the current equipment operating condition to which the characteristic variable acquired in real time belongs and the standard deviation of the key operating parameter corresponding to each equipment operating condition; and

[0044] An operating range determination module is used to determine the normal operating range of the key operating parameter according to the real-time standard value and the real-time standard deviation.

[0045] According to a third aspect of the present application, there is provided an electronic device, comprising:

[0046] Processor; and

[0047] A memory storing computer instructions, which, when executed by the processor, causes the processor to execute the method described in the first aspect.

[0048] According to a fourth aspect of the present application, a non-transitory computer storage medium is provided, storing a computer program, which, when executed by multiple processors, enables the processors to execute the method described in the first aspect.

[0049] According to the early warning method and device during the equipment operation process provided by the present application, according to the correlation between the key focus operation parameters and the non-key focus operation parameters, the characteristic variables in the non-key focus operation parameters are determined, and a standard value estimation model for the key focus operation parameters is constructed. Through the equipment working conditions corresponding to the non-key focus operation parameters, the equipment working conditions to which the sample data of the corresponding key focus operation parameters belong are identified, and the standard deviation of the sample data of the key focus operation parameters belonging to the same equipment working conditions is determined; then, the characteristic variables obtained in real time are input into the standard value estimation model to determine the real-time standard value of the key focus operation parameters, and the real-time standard deviation of the key focus operation parameters is determined; finally, according to the real-time standard value and the real-time standard deviation, the normal operating range of the key focus operation parameters is determined. In this way, when the measured value of the key focus operation parameter obtained exceeds this range, an alarm is issued. According to the solution of the present application, it is possible to dynamically calculate in real time the reasonable operating range of key operating parameters under various operating conditions, realize refined intelligent alarms, and ensure the accuracy of standard deviations by introducing clustering and mathematical statistics techniques. The accuracy of expected values ​​is ensured through data correlation analysis and model building methods such as machine learning algorithms, thereby achieving accurate, real-time and efficient parameter alarms and early warnings, improving the monitoring efficiency of the control system, and ensuring safe and stable operation of equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without exceeding the scope of protection required by the present application.

[0051] Figure 1 It is a flowchart of an early warning method during equipment operation according to an embodiment of the present application.

[0052] Figure 2 It is a flowchart of an early warning method during equipment operation according to another embodiment of the present application.

[0053] Figure 3 It is a schematic diagram of an early warning device during equipment operation according to an embodiment of the present application.

[0054] Figure 4 It is a schematic diagram of an early warning device during equipment operation according to another embodiment of the present application.

[0055] Figure 5 It is a structural diagram of an electronic device provided by this application. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0057] According to one aspect of the present application, a method for early warning during the operation of a device is provided. Figure 1 As shown, the method includes the following steps.

[0058] Step S101, determining characteristic variables from equipment operating parameters.

[0059] In industrial process control systems, there are many types and quantities of equipment operating parameters. Among these equipment operating parameters, some operating parameters, such as the temperature, pressure, vibration parameters of the equipment, etc., can more intuitively reflect the operating status of the equipment. When the equipment is abnormal, the problem can be discovered in time by monitoring and analyzing these operating parameters. These parameters reflecting the operating status of the equipment are set as key operating parameters, and other operating parameters that are not key operating parameters are set as non-key operating parameters. It can be understood by those skilled in the art that the key operating parameters can be set by the technicians based on their experience and actual needs.

[0060] According to some embodiments, step S101 can be divided into two steps: step 1: determining the correlation between the key focus operating parameters and the non-key focus operating parameters in the equipment operating parameters; and step 2: determining a preset number of parameters from the non-key focus operating parameters as feature variables based on the correlation.

[0061] According to some embodiments, in actual calculation, it is necessary to focus on the historical data of the operating parameters and form samples Y in chronological order. i , where i = 1…n is the sampling number in chronological order. The non-key operating parameters in the equipment operation are set as independent variables, and the historical data of each variable is organized in the same chronological order to obtain the sample Where j=1…m is the jth independent variable.

[0062] According to a specific embodiment, Pearson correlation analysis may be used to screen out key independent variables that affect key operating parameters as feature variables.

[0063] The calculation of Pearson correlation analysis is shown in equation (1):

[0064]

[0065] Where: r is the Pearson correlation coefficient, x i , Y i , i=1…n are two groups of samples used to analyze the degree of correlation, is the average of the two groups of samples.

[0066] In this way, for any one of the focused operating parameters, its correlation with any one of the non-focused operating parameters is calculated.

[0067] Therefore, step 1 may include: traversing each non-key focus operating parameter, and calculating the correlation between the key focus operating parameter and the non-key focus operating parameter.

[0068] For ease of understanding, in order to calculate the correlation between the key focus operating parameters and the non-key focus operating parameters, the Pearson correlation analysis is used in the above embodiment. However, those skilled in the art will understand that as long as the correlation between the key focus operating parameters and the non-key focus operating parameters can be calculated, other correlation calculation methods or algorithms may also be used, which falls within the scope of this application.

[0069] After obtaining the correlation between the key concern operating parameters and the non-key concern operating parameters, the higher the correlation, the closer the relationship with the key concern operating parameters, and the greater the ability to affect the key concern operating parameters. In this way, in the determined correlation, one or more non-key concern operating parameters with the highest correlation are determined as characteristic variables. Among them, the number of characteristic variables can be set artificially. For each key concern operating parameter, a predetermined number of non-key concern operating parameters with the highest correlation can be determined as characteristic variables, that is, each key concern operating parameter (such as temperature, pressure, vibration) has a corresponding characteristic variable.

[0070] According to some embodiments, considering that after some non-focused operating parameters as independent variables change in actual operation, it takes a delay period of time to observe the change of the focused operating parameter Y, so in the calculation With Y i When calculating the correlation, the influence of the time domain needs to be considered. In the present invention, the method of traversing and taking the largest value is used for calculation.

[0071] Assuming that the delay time t=0…k (i.e., corresponding to k+1 delay moments), the correlation between the j-th non-focused operating parameter and the focused operating parameter under the delay time t is calculated respectively, as shown in equation (2):

[0072]

[0073] in: is the Pearson coefficient between the jth non-focused operating parameter as an independent variable and the focused operating parameter Y under the delay t condition, is the sample mean of the jth non-focus operating parameter as an independent variable.

[0074] The optimal delay time t and the optimal Pearson correlation coefficient of each non-key focus operating parameter as an independent variable and the key focus operating parameter Y in the time domain t=0…k can be calculated by equation (2). Thus, a preset number of non-key focus operating parameters with the greatest correlation with the key focus operating parameter Y can be selected from all non-key focus operating parameters as characteristic variables, and the relationship between each characteristic variable and the key focus operating parameter Y in the time domain can be obtained.

[0075] Thus, step 1 may specifically include the following steps:

[0076] (a) determining the first delay moment among the set number of delay moments as the current delay moment, wherein the delay moment is the delay time between the non-focused operating parameter and the focused operating parameter;

[0077] (b) for any selected parameter among the non-focused operating parameters, calculating the correlation between the selected parameter and the focused operating parameter at the current delay time;

[0078] (c) if the current delay time is not the last delay time among the set number of delay times, determine the next delay time as the current delay time, and return to step (b);

[0079] (d) when the current delay time is the last delay time, determining the delay time corresponding to the maximum correlation among the set number of correlations as the set delay time for calculating the correlation between the selected parameter and the key operating parameter; and

[0080] (e) calculating the correlation between the selected parameter and the key operating parameter according to the set delay time.

[0081] Step S102: constructing a standard value estimation model of the key operating parameter according to the sample data of the key operating parameter and the sample data of the characteristic variable.

[0082] According to some embodiments, during the operation of the industrial process control system, a lot of historical data will be generated about the parameters, that is, there are sample data of characteristic variables and sample data of corresponding key operating parameters. The sample data of characteristic variables and the sample data of corresponding key operating parameters are used as inputs, for example, a standard value estimation model for key operating parameters can be constructed using a support vector regression machine (SVR). After the standard value estimation model is constructed, the real-time standard value of the key operating parameter can be determined based on the characteristic variables obtained in real time.

[0083] According to some embodiments, the process of constructing the standard value estimation model of the key operating parameter according to the support vector regression machine (SVR) is as follows, wherein the support vector regression expression is shown in equation (3):

[0084]

[0085] Among them, α i is the Lagrange multiplier; is the support vector coefficient; K(x,x i ) is the kernel function, b is the distance between the hyperplane and the origin, and the Gaussian radial basis kernel function (RFB) is used, and its calculation formula is shown in equation (4):

[0086]

[0087] Among them, x is the input variable, x i is the support vector, σ is the kernel function width parameter, and ‖…‖ is the norm calculation.

[0088] When constructing the standard value estimation model of the key operating parameters, the input variables are the characteristic variables selected in step S101, and the output is the key operating parameters, that is:

[0089]

[0090] Among them, y is the key operating parameter of the monitoring system, such as temperature, pressure, vibration and other signals; is the characteristic variable selected in step S101, wherein j represents the number of the characteristic variable, and jt represents the optimal delay time of the jth characteristic variable.

[0091] According to the above input-output model, combined with the system historical data, a sample for training can be constructed. The estimation model of the key operating parameters can be obtained by training with the support vector regression machine. Since the output of the model uses the measured values ​​of the equipment operating parameters, the result obtained by substituting the trained model into the real-time values ​​of each characteristic variable is actually the expected value (i.e., standard value) of the key operating parameter. This machine algorithm model is a single-output expected value modeling, which is easy to implement using various machine learning algorithms.

[0092] Those skilled in the art will appreciate that, in addition to the support vector regression algorithm used above, other machine learning algorithms may also be used. For example, feedforward neural networks, long short-term memory neural networks, extreme learning machines and other machine learning algorithms may also be used for model training.

[0093] Step S103, using a clustering algorithm on the sample data of the characteristic variable to identify the equipment operating condition to which the sample data of the key operating parameter corresponding to the sample data of the characteristic variable belongs;

[0094] Step S104, calculating the standard deviation of the sample data of the key operating parameters belonging to the same equipment operating condition, and determining the standard deviation of the key operating parameters corresponding to each equipment operating condition.

[0095] In step S102, the input-output model and historical data have been combined to construct a sample for model training. The sample includes two parts: input and output. The input is a matrix of dimension m×n. The output is a vector of dimension m×1 In the process of identifying and dividing the working conditions, the matrix Since each key operating parameter has its corresponding characteristic variable, after determining the equipment operating condition to which the corresponding characteristic variable belongs, the equipment operating condition to which the corresponding key operating parameter belongs can be known. Thus, in step S103, firstly, the sample data of the characteristic variable adopts a clustering algorithm to determine the equipment operating conditions to which the sample data of the characteristic variable respectively correspond, and then, according to the equipment operating conditions to which the sample data of the characteristic variable respectively correspond, the equipment operating conditions to which the sample data of the key operating parameter corresponding to the sample data of the characteristic variable belongs are determined.

[0096] Clustering algorithms include but are not limited to K-means clustering, AP clustering, spectral clustering, etc. Below, K-means clustering is taken as an example to illustrate the clustering calculation process.

[0097] 1) Initialization: determine the number of equipment operating conditions. For example, if k equipment operating conditions are to be classified, the matrix Randomly select k samples as the centers of the initial clusters;

[0098] 2) Cluster the samples. For the selected cluster center, calculate the distance from all samples to each center, such as the Euclidean distance, and cluster each sample into the class of the center closest to it to form k cluster classes, which constitute the clustering result;

[0099] 3) Calculate the cluster center after clustering, calculate the centroid of each class, that is, the mean of the samples in each class, as the new cluster center;

[0100] 4) Then re-execute steps 2) and 3) until the clustering result no longer changes, that is; if the difference between any one of the centroids and the corresponding current cluster center exceeds the preset threshold, return to execute steps 2) and 3). If the difference between any one of the centroids and the corresponding current cluster center does not exceed the preset threshold, the centroid obtained by clustering is the cluster center, and the k centroids obtained are the cluster centers of the k equipment working conditions.

[0101] Thus, step S103 may specifically include the following steps:

[0102] (f) randomly selecting samples of the number of types of equipment operating conditions from the sample data of the characteristic variables as current cluster centers;

[0103] (g) calculating the distance between the sample data of the feature variable and each of the current cluster centers of the number of categories;

[0104] (h) classifying each of the sample data of the feature variable into the cluster class to which the current cluster center closest to the feature variable belongs, so as to form the number of cluster classes;

[0105] (i) calculating the centroid of each cluster according to the sample data of the characteristic variable contained in each of the clusters of the number of categories, and obtaining the centroids of the number of categories;

[0106] (j) when the difference between any one of the number of centroids of the categories and the corresponding current cluster center exceeds a preset threshold, the number of centroids of the categories is determined as the current cluster center, and the process returns to step (g); and

[0107] (k) When the difference between any one of the number of centroids of the type and the corresponding current cluster center does not exceed a preset threshold, the number of centroids of the type is determined as the cluster center corresponding to the equipment operating condition.

[0108] Clustering calculation can be performed on the matrix The samples in the cluster are classified and the cluster centers and cluster radii are obtained, which realizes the identification and division of equipment working conditions and converts the corresponding output vector According to the matrix The working conditions are classified to obtain the output sample set under various equipment working conditions. At this time, for each sample set By performing statistical calculations, the standard deviation of the key operating parameters under various working conditions can be obtained.

[0109] According to some embodiments, the standard deviation is calculated as shown in equation (6):

[0110]

[0111] Where: s i is the standard deviation of the key operating parameters under condition i, y ij is the sample set data focusing on the operating parameters under working condition i, is the average value of the key operating parameters under condition i.

[0112] Step S105, inputting the characteristic variables acquired in real time into the standard value estimation model to determine the real-time standard value of the key operating parameter;

[0113] Step S106, determining the real-time standard deviation of the key operating parameter according to the current equipment operating condition to which the characteristic variable obtained in real time belongs and the standard deviation of the key operating parameter corresponding to each equipment operating condition;

[0114] Step S107: determining a normal operating range of the key operating parameter according to the real-time standard value and the real-time standard deviation.

[0115] During the real-time operation of the system, the standard value estimation model of the key operating parameters constructed in step S102 is used to obtain and construct the input feature variables in real time. Substitute it into the standard value estimation model to obtain the standard value of the key operating parameter at the current moment At the same time, the input feature variable is substituted into each cluster center and cluster radius obtained in step S103 for calculation, the working condition to which the input feature variable belongs at the current moment is identified, and the standard deviation of the key operating parameters under the working condition is obtained. Once the standard value and standard deviation are known, the confidence interval at a certain confidence level can be calculated, that is, the normal operating range of the operating parameters can be focused on.

[0116] According to some embodiments, the confidence interval range calculation method is shown in equation (7):

[0117]

[0118] Wherein, α is (1-confidence level), for example, if the confidence level is 90%, then α = 1-0.9 = 0.1; is the Z value, which can be obtained by looking up the standard normal distribution table; n is the number of samples, that is, the number of samples in the sample set under various working conditions obtained in step S104.

[0119] Figure 2 is a flow chart of an early warning method during equipment operation according to another embodiment of the present application. Figure 1 compared to, Figure 2 Steps S201 to S207 and Figure 1 The steps S101 to S107 are the same as above, except that: Figure 2 The method shown also includes the following steps:

[0120] Step S208, obtaining the measured value of the key operating parameter;

[0121] Step S209: outputting an alarm signal when the measured value exceeds the normal operating range.

[0122] In the process of obtaining the normal operating range of the key operating parameter through steps S201 to S207, the measured value of the key operating parameter is also obtained (for example, through measurement). It is determined whether the measured value of the key operating parameter is within the normal operating range. If it is within the range, it indicates that the system is operating normally. Otherwise, it indicates that the system is operating abnormally and an alarm signal is output.

[0123] In the process of obtaining the normal operating range of the key operating parameters through steps S201 to S207, the normal operating range will dynamically change according to the changes in the characteristic variables actually obtained, that is, this range will change according to the actual equipment operating parameters, thereby achieving efficient and accurate alarms and early warnings, improving the monitoring efficiency of the control system, and ensuring safe and stable operation of the equipment.

[0124] According to another aspect of the present application, an early warning device for equipment operation is provided. Figure 3 As shown, the device includes the following modules.

[0125] The characteristic variable determination module 301 is used to determine the characteristic variable from the equipment operation parameters.

[0126] In industrial process control systems, there are many types and quantities of equipment operating parameters. Among these equipment operating parameters, some operating parameters, such as the temperature, pressure, vibration parameters of the equipment, etc., can more intuitively reflect the operating status of the equipment. When the equipment is abnormal, the problem can be discovered in time by monitoring and analyzing these operating parameters. These parameters reflecting the operating status of the equipment are set as key operating parameters, and other operating parameters that are not key operating parameters are set as non-key operating parameters. It can be understood by those skilled in the art that the key operating parameters can be set by the technicians based on their experience and actual needs.

[0127] According to some embodiments, module 301 may include: a correlation determination unit, used to determine the correlation between the key focus operating parameters and the non-key focus operating parameters in the equipment operating parameters; and a characteristic variable determination unit, used to determine a preset number of parameters from the non-key focus operating parameters as characteristic variables based on the correlation.

[0128] According to some embodiments, in actual calculation, it is necessary to focus on the historical data of the operating parameters and form samples Y in chronological order. i , where i = 1…n is the sampling number in chronological order. The non-key operating parameters in the equipment operation are set as independent variables, and the historical data of each variable is organized in the same chronological order to obtain the sample Where j=1…m is the jth independent variable.

[0129] According to a specific embodiment, Pearson correlation analysis may be used to screen out key independent variables that affect key operating parameters as feature variables.

[0130] In this way, for any one of the focused operating parameters, its correlation with any one of the non-focused operating parameters is calculated.

[0131] Therefore, the correlation determination unit may be used to: traverse each non-key focus operating parameter, and calculate the correlation between the key focus operating parameter and the non-key focus operating parameter.

[0132] For ease of understanding, in order to calculate the correlation between the key focus operating parameters and the non-key focus operating parameters, the Pearson correlation analysis is used in the above embodiment. However, those skilled in the art will understand that as long as the correlation between the key focus operating parameters and the non-key focus operating parameters can be calculated, other correlation calculation methods or algorithms may also be used, which falls within the scope of this application.

[0133] After obtaining the correlation between the key concern operating parameters and the non-key concern operating parameters, the higher the correlation, the closer the relationship with the key concern operating parameters, and the greater the ability to affect the key concern operating parameters. In this way, in the determined correlation, one or more non-key concern operating parameters with the highest correlation are determined as characteristic variables. Among them, the number of characteristic variables can be set artificially. For each key concern operating parameter, a predetermined number of non-key concern operating parameters with the highest correlation can be determined as characteristic variables, that is, each key concern operating parameter (such as temperature, pressure, vibration) has a corresponding characteristic variable.

[0134] According to some embodiments, considering that after some non-focused operating parameters as independent variables change in actual operation, it takes a delay period of time to observe the change of the focused operating parameter Y, so in the calculation With Y i When calculating the correlation, the influence of the time domain needs to be considered. In the present invention, the method of traversing and taking the largest value is used for calculation.

[0135] Thus, the relevance determination unit may be specifically configured to:

[0136] Determine the first delay moment among the set number of delay moments as the current delay moment, wherein the delay moment is the delay time between the non-focused operating parameter and the focused operating parameter;

[0137] For any selected parameter among the non-focused operating parameters, calculating the correlation between the selected parameter and the focused operating parameter at the current delay time;

[0138] If the current delay time is not the last delay time among the set number of delay times, determining the next delay time as the current delay time;

[0139] In the case where the current delay time is the last delay time, determining the delay time corresponding to the maximum correlation among the set number of correlations as the set delay time for calculating the correlation between the selected parameter and the focused operating parameter; and

[0140] According to the set delay time, the correlation between the selected parameter and the key operating parameter is calculated.

[0141] The estimation model building module 302 is used to build a standard value estimation model of the key operating parameter according to the sample data of the key operating parameter and the sample data of the characteristic variable.

[0142] According to some embodiments, during the operation of the industrial process control system, a lot of historical data will be generated about the parameters, that is, there are sample data of characteristic variables and sample data of corresponding key operating parameters. The sample data of characteristic variables and the sample data of corresponding key operating parameters are used as inputs, for example, a standard value estimation model for key operating parameters can be constructed using a support vector regression machine (SVR). After the standard value estimation model is constructed, the real-time standard value of the key operating parameter can be determined based on the characteristic variables obtained in real time.

[0143] According to the above input-output model, combined with the system historical data, a sample for training can be constructed. The estimation model of the key operating parameters can be obtained by training with the support vector regression machine. Since the output of the model uses the measured values ​​of the equipment operating parameters, the result obtained by substituting the trained model into the real-time values ​​of each characteristic variable is actually the expected value (i.e., standard value) of the key operating parameter. This machine algorithm model is a single-output expected value modeling, which is easy to implement using various machine learning algorithms.

[0144] Those skilled in the art will appreciate that, in addition to the support vector regression algorithm used above, other machine learning algorithms may also be used. For example, feedforward neural networks, long short-term memory neural networks, extreme learning machines and other machine learning algorithms may also be used for model training.

[0145] The operating condition identification module 303 is used to use a clustering algorithm on the sample data of the characteristic variable to identify the equipment operating condition to which the sample data of the key operating parameter corresponding to the sample data of the characteristic variable belongs;

[0146] The operating condition standard deviation determination module 304 is used to calculate the standard deviation of the sample data of the key operating parameters belonging to the same equipment operating condition, and determine the standard deviation of the key operating parameters corresponding to each equipment operating condition.

[0147] In module 302, the input-output model and historical data have been combined to construct a sample for model training. The sample includes two parts: input and output. The input is a matrix of dimension m×n. The output is a vector of dimension m×1 In the process of identifying and dividing the working conditions, the matrix Since each key operating parameter has its corresponding characteristic variable, after determining the equipment operating condition to which the corresponding characteristic variable belongs, the equipment operating condition to which the corresponding key operating parameter belongs can be known. Thus, in module 303, firstly, the sample data of the characteristic variable adopts a clustering algorithm to determine the equipment operating conditions to which the sample data of the characteristic variable respectively corresponds, and then, according to the equipment operating conditions to which the sample data of the characteristic variable respectively corresponds, the equipment operating conditions to which the sample data of the key operating parameter corresponding to the sample data of the characteristic variable belongs are determined.

[0148] Clustering algorithms include but are not limited to K-means clustering, AP clustering, spectral clustering, etc. Below, K-means clustering is taken as an example to illustrate the clustering calculation process.

[0149] 1) Initialization: determine the number of equipment operating conditions. For example, if k equipment operating conditions are to be classified, the matrix Randomly select k samples as the centers of the initial clusters;

[0150] 2) Cluster the samples. For the selected cluster center, calculate the distance from all samples to each center, such as the Euclidean distance, and cluster each sample into the class of the center closest to it to form k cluster classes, which constitute the clustering result;

[0151] 3) Calculate the cluster center after clustering, calculate the centroid of each class, that is, the mean of the samples in each class, as the new cluster center;

[0152] 4) Then re-execute steps 2) and 3) until the clustering result no longer changes, that is; if the difference between any one of the centroids and the corresponding current cluster center exceeds the preset threshold, return to execute steps 2) and 3). If the difference between any one of the centroids and the corresponding current cluster center does not exceed the preset threshold, the centroid obtained by clustering is the cluster center, and the k centroids obtained are the cluster centers of the k equipment working conditions.

[0153] Thus, module 303 may be specifically used for:

[0154] According to the number of types of equipment operating conditions, randomly select samples of the number of types from the sample data of the characteristic variables as current cluster centers;

[0155] Calculate the distance from the sample data of the feature variable to each of the current cluster centers of the number of categories;

[0156] Classify each of the sample data of the characteristic variable into the cluster class to which the current cluster center closest to the characteristic variable belongs, so as to form the number of cluster classes;

[0157] Calculate the centroid of each clustering class according to the sample data of the characteristic variable contained in each of the clustering classes of the number of types, and obtain the centroids of the number of types;

[0158] In the case where the difference between any one of the number of centroids of the types and the corresponding current cluster center exceeds a preset threshold, determining the number of centroids of the types as the current cluster center; and

[0159] When the difference between any one of the number of centroids of the types and the corresponding current cluster center does not exceed a preset threshold, the number of centroids of the types are determined as the cluster centers corresponding to the equipment operating conditions.

[0160] Clustering calculation can be performed on the matrix The samples in the cluster are classified and the cluster centers and cluster radii are obtained, which realizes the identification and division of equipment working conditions and converts the corresponding output vector According to the matrix The working conditions are classified to obtain the output sample set under various equipment working conditions. At this time, for each sample set By performing statistical calculations, the standard deviation of the key operating parameters under various working conditions can be obtained.

[0161] A real-time standard value determination module 305 is used to input the characteristic variables acquired in real time into the standard value estimation model to determine the real-time standard value of the key operating parameter;

[0162] A real-time standard deviation determination module 306, configured to determine the real-time standard deviation of the key operating parameter according to the current equipment operating condition to which the characteristic variable obtained in real time belongs and the standard deviation of the key operating parameter corresponding to each equipment operating condition;

[0163] The operating range determination module 307 is used to determine the normal operating range of the key operating parameter according to the real-time standard value and the real-time standard deviation.

[0164] During the real-time operation of the system, the standard value estimation model of the key operating parameters constructed in module 302 is used to obtain and construct the input characteristic variables in real time. Substitute it into the standard value estimation model to obtain the standard value of the key operating parameter at the current moment At the same time, the input characteristic variables are substituted into the cluster centers and cluster radii obtained in module 303 for calculation, the working condition to which the input characteristic variables belong at the current moment is identified, and the standard deviation of the key operating parameters under the working condition is obtained. Once the standard value and standard deviation are known, the confidence interval at a certain confidence level can be calculated, that is, the normal operating range of the operating parameters can be focused on.

[0165] Figure 4 Schematic diagram of an early warning device during equipment operation according to another embodiment of the present application. Figure 3 compared to, Figure 4 Modules 401 to 407 and Figure 3 Modules 301 to 307 are the same except that: Figure 4 The device shown also includes the following modules:

[0166] The measured value acquisition module 408 is used to obtain the measured value of the key operating parameter;

[0167] The alarm signal output module 409 is used to output an alarm signal when the measured value exceeds the normal operating range.

[0168] In the process of obtaining the normal operating range of the key operating parameter through modules 401 to 407, the measured value of the key operating parameter is also obtained (for example, through measurement). It is determined whether the measured value of the key operating parameter is within the normal operating range. If it is within the range, it indicates that the system is operating normally. Otherwise, it indicates that the system is operating abnormally, and an alarm signal is output.

[0169] In the process of obtaining the normal operating range of the key operating parameters through modules 401 to 407, the normal operating range will dynamically change according to the changes in the characteristic variables actually obtained, that is, this range will change according to the actual equipment operating parameters, thereby achieving efficient and accurate alarms and early warnings, improving the monitoring efficiency of the control system, and ensuring safe and stable operation of the equipment.

[0170] According to the early warning method and device during the equipment operation process provided by the present application, according to the correlation between the key focus operation parameters and the non-key focus operation parameters, the characteristic variables in the non-key focus operation parameters are determined, and a standard value estimation model for the key focus operation parameters is constructed. Through the equipment working conditions corresponding to the non-key focus operation parameters, the equipment working conditions to which the sample data of the corresponding key focus operation parameters belong are identified, and the standard deviation of the sample data of the key focus operation parameters belonging to the same equipment working conditions is determined; then, the characteristic variables obtained in real time are input into the standard value estimation model to determine the real-time standard value of the key focus operation parameters, and the real-time standard deviation of the key focus operation parameters is determined; finally, according to the real-time standard value and the real-time standard deviation, the normal operating range of the key focus operation parameters is determined. In this way, when the measured value of the key focus operation parameter obtained exceeds this range, an alarm is issued. According to the solution of the present application, it is possible to dynamically calculate in real time the reasonable operating range of key operating parameters under various operating conditions, realize refined intelligent alarms, and ensure the accuracy of standard deviations by introducing clustering and mathematical statistics techniques. The accuracy of expected values ​​is ensured through data correlation analysis and model building methods such as machine learning algorithms, thereby achieving accurate, real-time and efficient parameter alarms and early warnings, improving the monitoring efficiency of the control system, and ensuring safe and stable operation of equipment.

[0171] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0172] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present application.

[0173] In the several embodiments provided in the present application, it should be understood that the disclosed device can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of the units, which is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the device or unit can be electrical connection or other forms.

[0174] See also Figure 5 , Figure 5An electronic device is provided, comprising a processor and a memory. The memory stores computer instructions, and when the computer instructions are executed by the processor, the processor executes the computer instructions to achieve the following Figure 1 and Figure 2 The method and refinement scheme shown.

[0175] It should be understood that the above device embodiments are only illustrative, and the device disclosed in the present invention can also be implemented in other ways. For example, the division of units / modules described in the above embodiments is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units, modules or components can be combined, or can be integrated into another system, or some features can be ignored or not executed.

[0176] In addition, unless otherwise specified, each functional unit / module in each embodiment of the present invention may be integrated into one unit / module, each unit / module may exist physically separately, or two or more units / modules may be integrated together. The above-mentioned integrated unit / module may be implemented in the form of hardware or in the form of a software program module.

[0177] If the integrated unit / module is implemented in the form of hardware, the hardware may be a digital circuit, an analog circuit, etc. The physical implementation of the hardware structure includes but is not limited to transistors, memristors, etc. If not otherwise specified, the processor or chip may be any appropriate hardware processor, such as a CPU, a GPU, an FPGA, a DSP, an ASIC, etc. If not otherwise specified, the on-chip cache, the off-chip memory, and the memory may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc.

[0178] If the integrated unit / module is implemented in the form of a software program module and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a memory and includes several instructions for a computer electronic device (which can be a personal computer, a server or a network electronic device, etc.) to perform all or part of the steps of the method described in each embodiment of the present disclosure. The aforementioned memory includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk, etc. Various media that can store program codes.

[0179] The present application also provides a non-transitory computer storage medium storing a computer program. When the computer program is executed by a plurality of processors, the processors execute the following Figure 1 and Figure 2 The method and refinement scheme shown.

[0180] The embodiments of the present application are described in detail above. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and its core idea of ​​the present application. At the same time, changes or deformations made by those skilled in the art based on the ideas of the present application, the specific implementation methods and the scope of application of the present application, all belong to the scope of protection of the present application. In summary, the content of this specification should not be construed as a limitation on the present application.

Claims

1. An early warning method during equipment operation, characterized in that: include: Determining characteristic variables from equipment operating parameters; Constructing a standard value estimation model for the key operating parameter based on the sample data of the key operating parameter and the sample data of the characteristic variable; Using a clustering algorithm on the sample data of the characteristic variable to identify the equipment operating condition to which the sample data of the key operating parameter corresponding to the sample data of the characteristic variable belongs; Calculating the standard deviation of the sample data of the key operating parameters belonging to the same equipment operating condition, and determining the standard deviation of the key operating parameters corresponding to each equipment operating condition; Inputting the characteristic variables acquired in real time into the standard value estimation model to determine the real-time standard value of the key operating parameter; Determine the real-time standard deviation of the key operating parameter according to the current equipment operating condition to which the characteristic variable obtained in real time belongs and the standard deviation of the key operating parameter corresponding to each equipment operating condition; as well as A normal operating range of the key operating parameter is determined according to the real-time standard value and the real-time standard deviation.

2. The early warning method according to claim 1, characterized in that: Determining characteristic variables from equipment operating parameters includes: Determining the correlation between the key operating parameters and the non-key operating parameters in the equipment operating parameters; and A preset number of parameters are determined from the non-key operating parameters as characteristic variables according to the correlation.

3. The early warning method according to claim 2, characterized in that: The determining of the correlation between the key operating parameters and the non-key operating parameters in the equipment operating parameters includes: Each non-key concern operating parameter is traversed, and the correlation between the key concern operating parameter and the non-key concern operating parameter is calculated.

4. The early warning method according to claim 2, characterized in that: The determining of the correlation between the key operating parameters and the non-key operating parameters in the equipment operating parameters includes: (a) determining the first delay moment among the set number of delay moments as the current delay moment, wherein the delay moment is the delay time between the non-focused operating parameter and the focused operating parameter; (b) for any selected parameter among the non-focused operating parameters, calculating the correlation between the selected parameter and the focused operating parameter at the current delay time; (c) if the current delay time is not the last delay time among the set number of delay times, determine the next delay time as the current delay time, and return to step (b); (d) when the current delay time is the last delay time, determining the delay time corresponding to the maximum correlation among the set number of correlations as the set delay time for calculating the correlation between the selected parameter and the key operating parameter; and (e) calculating the correlation between the selected parameter and the key operating parameter according to the set delay time.

5. The early warning method according to claim 1, characterized in that: The adopting a clustering algorithm to the sample data of the characteristic variable to identify the equipment operating condition to which the sample data of the key operating parameter corresponding to the sample data of the characteristic variable belongs includes: Using a clustering algorithm on the sample data of the characteristic variables to determine the equipment operating conditions corresponding to the sample data of the characteristic variables; and According to the equipment operating conditions corresponding to the sample data of the characteristic variables, the equipment operating condition to which the sample data of the key operating parameter corresponding to the sample data of the characteristic variables belongs is determined.

6. The early warning method according to claim 5, characterized in that: The adopting a clustering algorithm to the sample data of the characteristic variables to determine the equipment operating conditions corresponding to the sample data of the characteristic variables respectively includes: (f) randomly selecting samples of the number of types of equipment operating conditions from the sample data of the characteristic variables as current cluster centers; (g) calculating the distance between the sample data of the feature variable and each of the current cluster centers of the number of categories; (h) classifying each of the sample data of the feature variable into the cluster class to which the current cluster center closest to the feature variable belongs, so as to form the number of cluster classes; (i) calculating the centroid of each cluster according to the sample data of the characteristic variable contained in each of the clusters of the number of categories, and obtaining the centroids of the number of categories; (j) when the difference between any one of the number of centroids of the categories and the corresponding current cluster center exceeds a preset threshold, the number of centroids of the categories is determined as the current cluster center, and the process returns to step (g); and (k) When the difference between any one of the number of centroids of the type and the corresponding current cluster center does not exceed a preset threshold, the number of centroids of the type is determined as the cluster center corresponding to the equipment operating condition.

7. The early warning method according to any one of claims 1 to 6, characterized in that: Also includes: Obtaining the measured value of the key operating parameter; When the measured value exceeds the normal operating range, an alarm signal is output.

8. An early warning device during equipment operation, characterized in that: include: A characteristic variable determination module, used for determining characteristic variables from equipment operation parameters; An estimation model building module, used to build a standard value estimation model of the key operating parameter according to the sample data of the key operating parameter and the sample data of the characteristic variable; An operating condition identification module, used to use a clustering algorithm on the sample data of the characteristic variable to identify the equipment operating condition to which the sample data of the key operating parameter corresponding to the sample data of the characteristic variable belongs; A working condition standard deviation determination module, used to calculate the standard deviation of the sample data of the key operating parameters belonging to the same equipment working condition, and determine the standard deviation of the key operating parameters corresponding to each equipment working condition; A real-time standard value determination module, used for inputting the characteristic variables acquired in real time into the standard value estimation model to determine the real-time standard value of the key operating parameter; A real-time standard deviation determination module, used to determine the real-time standard deviation of the key operating parameter according to the current equipment operating condition to which the characteristic variable obtained in real time belongs and the standard deviation of the key operating parameter corresponding to each equipment operating condition; as well as An operating range determination module is used to determine the normal operating range of the key operating parameter according to the real-time standard value and the real-time standard deviation.

9. An electronic device, characterized in that: The method comprises a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the method according to any one of claims 1 to 7 when executing the computer program in the memory.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images