Production operation management method and system based on five-dimensional integration

Abstract

The present invention relates to the field of production management technology, and in particular to a production operation management method and system based on five-dimensional integration. The present invention collects production operation data of a power plant, including power generation data, equipment operation data, fuel consumption data, environmental monitoring data and fault data, evaluates eligibility according to the missing rate of the production operation data and determines the missing rate evaluation coefficient, thereby preprocessing the production operation data, performing nearest neighbor interpolation or missing data prediction based on a first time series prediction model, using a second time series prediction model to predict data within a future time period, and determining whether to perform a fault warning according to the proportion of fault data in the prediction result. The present invention overcomes the problem that the prior art cannot effectively identify and process low-quality data, affects the accuracy of prediction of the prediction model, and thus leads to poor accuracy in fault prediction in production operation.

Description

Technical Field

[0001] The present invention relates to the technical field of production management, and in particular to a production operation management method and system based on five-dimensional integration. Background Art

[0002] In the operation and management of modern power plants, it is crucial to achieve efficient, stable and sustainable production. In order to achieve this goal, it is particularly important to accurately collect, analyze and process the production and operation data of power plants. These data include power generation, equipment operating status, fuel consumption, environmental monitoring indicators and equipment failure information, which are of great significance for evaluating the operating efficiency, environmental impact and equipment health of power plants. However, due to various reasons, such as equipment failure, instability of monitoring systems or human errors, production and operation data are often missing or inaccurate, which affects the quality and subsequent processing of data. Traditional production and operation management methods often focus on the analysis and processing of single-dimensional data, such as focusing on the optimization of power generation or responding only to equipment failures. This method ignores the complexity of production and operation management and the interrelationship between data, resulting in incomplete management decisions and delayed response measures. In addition, the processing of missing data usually adopts a simple deletion or manual filling method, which not only increases the workload, but also may introduce additional errors, affecting the accuracy and reliability of data analysis.

[0003] Chinese patent application publication number: CN116955339A discloses a production and operation data asset management system, including database and resource planning: including the collection, processing, storage, transmission, use, conversion and evaluation of information resources to achieve integration of all levels of the system; data middle platform: including data collection platform, data governance platform and data sharing and exchange platform.

[0004] It can be seen that the existing technology has the following problems: it is unable to effectively identify and process low-quality data, which affects the accuracy of the prediction model and thus leads to poor accuracy in fault prediction in production operations. Summary of the invention

[0005] To this end, the present invention provides a production operation management method and system based on five-dimensional integration to overcome the problem that the prior art cannot effectively identify and process low-quality data, affecting the accuracy of prediction model predictions, thereby leading to poor accuracy of fault predictions in production operations.

[0006] To achieve the above objectives, on the one hand, the present invention provides a production operation management method based on five-dimensional integration, comprising:

[0007] Collecting production and operation data of the power plant at a preset frequency, wherein the production and operation data includes power generation data, equipment operation data, fuel consumption data, environmental monitoring data, and fault data;

[0008] Determine the eligibility of production and operation data based on the missing rate of production and operation data;

[0009] Determine a missing rate assessment coefficient for production operation data, and determine a preprocessing method for preprocessing the production operation data according to the missing rate assessment coefficient, so as to determine the nearest neighbor interpolation of missing data in different types of data according to the average time difference between the timestamps of adjacent missing data in different types of data, or predict the missing data according to the first time series prediction model;

[0010] Alternatively, a second time series prediction model is trained based on production operation data, and the trained second time series prediction model is used to predict the future duration, and whether to issue a fault warning is determined based on the ratio of the number of fault data predicted in the future duration to the number of predicted data in the future duration.

[0011] Further, determining the eligibility of the production and operation data includes determining that the production and operation data is unqualified under the condition that a missing rate of the production and operation data is greater than a preset missing rate.

[0012] Furthermore, the missing rate evaluation coefficient is calculated according to the following formula:

[0013]

[0014] Among them, F represents the missing rate assessment coefficient, A1 represents the missing rate of power generation data, A2 represents the missing rate of equipment operation data, A3 represents the missing rate of fuel consumption data, A4 represents the missing rate of environmental monitoring data, A5 represents the missing rate of fault data, and A0 represents the average missing rate of different types of data in production operation data.

[0015] Furthermore, the determining of the preprocessing method for preprocessing the production operation data includes extracting the timestamps of missing data of different types of data in the production operation data under the condition that the missing rate assessment coefficient is less than or equal to a preset missing rate assessment coefficient.

[0016] Furthermore, the preprocessing method for determining the preprocessing of the production operation data includes arranging the missing rates of different types of data in the production operation data from large to small under the condition that the missing rate assessment coefficient is less than or equal to the preset missing rate assessment coefficient, and extracting the timestamps of the missing data in the first K types of data.

[0017] Furthermore, the timestamp of missing data extracted from the first K types of data is included in the missing rates of different types of data arranged from large to small, and the missing rate difference between the missing rate of the current type of data and the missing rate of the next type of data is determined to obtain the missing rate of the current type of data with the maximum missing rate difference as the node, and the serial number of the node in the missing rates of different types of data arranged from large to small is used as the value of K.

[0018] Further, the determining of the nearest neighbor interpolation for the missing data in the different types of data includes performing the nearest neighbor interpolation for the missing data under the condition that the average time difference is greater than a preset average time difference.

[0019] Furthermore, the predicting of missing data according to the first time series prediction model includes extracting a section of data that meets preset conditions in the production operation data to train the first time series prediction model, and inputting the timestamp of the missing data and the non-missing data in other types of data at the same timestamp as the missing data into the trained first time series prediction model to predict the missing data.

[0020] Further, the determining whether to perform a fault warning includes determining whether to perform a fault warning under the condition that the ratio is greater than a preset ratio.

[0021] On the other hand, the present invention also provides a system for applying a production operation management method based on five-dimensional integration, comprising:

[0022] A data collection unit, which is used to collect production and operation data of the power plant;

[0023] A data quality inspection unit, which is connected to the data collection unit and is used to determine the eligibility of the production and operation data according to the missing rate of the production and operation data;

[0024] a data preprocessing unit connected to the data quality inspection unit, and used to determine a missing rate assessment coefficient of the production operation data, and determine a preprocessing method for preprocessing the production operation data according to the missing rate assessment coefficient, so as to determine the nearest neighbor interpolation of the missing data in the different types of data according to the average time difference between the timestamps of adjacent missing data in the different types of data, or predict the missing data according to the first time series prediction model;

[0025] A fault warning unit is respectively connected to the data quality inspection unit and the data preprocessing unit, and is used to train the second time series prediction model according to the production operation data, predict the future duration with the trained second time series prediction model, and determine whether to issue a fault warning based on the ratio of the number of fault data predicted in the future duration to the number of predicted data in the future duration.

[0026] Compared with the prior art, the beneficial effect of the present invention lies in that the present invention determines whether the production and operation data is qualified based on the comparison result of the missing rate of the production and operation data with the preset missing rate to ensure that the data used for further analysis and decision-making is more complete, which helps to improve the accuracy and reliability of data analysis, thereby improving the quality of early warning.

[0027] Furthermore, the present invention calculates the missing rate assessment coefficient of production and operation data according to the formula to determine the preprocessing method for preprocessing the production and operation data according to the missing rate assessment coefficient of the production and operation data, so as to more specifically solve the data missing problems in different situations and adopt different strategies, thereby improving the effectiveness and efficiency of data preprocessing.

[0028] Furthermore, by identifying the turning points where the difference in missing rates is most significant, the present invention can more accurately locate the areas where the severity of the data missing problem changes the most. Targeted processing helps to focus resources and attention on solving the most critical data missing problems, thereby improving the efficiency and effectiveness of data preprocessing.

[0029] Furthermore, the present invention determines whether to perform nearest neighbor interpolation on the missing data in different types of data based on the comparison result of the average time difference between the timestamps of adjacent missing data in different types of data with the preset average time difference, so as to significantly improve the integrity of the data and provide more continuous and complete data support for subsequent data analysis and decision-making.

[0030] Furthermore, the present invention completes the missing data through the time series prediction model to significantly improve the integrity of the data set and provide a more continuous and complete data basis for subsequent data analysis, decision support, etc. The time series prediction model can effectively utilize the time series characteristics of the data and accurately predict the missing data by analyzing the autocorrelation and trend of the data.

[0031] Furthermore, the present invention determines whether to perform a fault warning based on the comparison result between the ratio and the preset ratio to monitor abnormal changes in production operation data in real time, promptly discover signs that may lead to faults, thereby achieving early warning of faults and helping to take preventive measures to avoid or reduce losses caused by faults.

[0032] Furthermore, the data acquisition unit of the present invention can collect production and operation data of the power plant in real time to ensure the timeliness and integrity of the data. The data quality inspection unit can determine the eligibility of the data according to the missing rate of the production and operation data to improve the reliability and accuracy of the data. The data preprocessing unit can determine the preprocessing method for the production and operation data according to the missing rate assessment coefficient to ensure the integrity and consistency of the data. The fault warning unit can predict faults for the future duration by training the second time series prediction model, discover potential faults in advance, reduce the impact of faults on the production and operation of the power plant, and improve the operating efficiency and reliability of the power plant. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a flowchart of a production operation management method based on five-dimensional integration according to an embodiment of the present invention;

[0034] Figure 2 It is a logic block diagram of a production operation management method based on five-dimensional integration according to an embodiment of the present invention;

[0035] Figure 3 This is a schematic diagram of the structure of a system applying a five-dimensional integrated production operation management method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0036] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0037] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0038] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0039] See also Figure 1-Figure 2 As shown, Figure 1 It is a flowchart of a production operation management method based on five-dimensional integration according to an embodiment of the present invention; Figure 2 It is a logical block diagram of a five-dimensional production operation management method according to an embodiment of the present invention.

[0040] The embodiment of the present invention is based on a five-dimensional integrated production operation management method, including:

[0041] Step S1, collecting production and operation data of the power plant at a preset frequency T, wherein the production and operation data includes power generation data, equipment operation data, fuel consumption data, environmental monitoring data, and fault data;

[0042] Step S2, determining the eligibility of the production and operation data according to the missing rate L of the production and operation data;

[0043] Step S3, determining a missing rate assessment coefficient F of the production operation data, and determining a preprocessing method for preprocessing the production operation data according to the missing rate assessment coefficient F, so as to determine the nearest neighbor interpolation of the missing data in the different types of data according to the average time difference C between the timestamps of adjacent missing data in the different types of data, or predicting the missing data according to the first time series prediction model;

[0044] Alternatively, the second time series prediction model is trained according to the production operation data, and the future duration t is predicted using the trained second time series prediction model, and whether to issue a fault warning is determined based on the ratio Z of the number of fault data predicted within the future duration t to the number of predicted data within the future duration t.

[0045] In an embodiment of the present invention, power generation data includes but is not limited to daily power generation, monthly power generation and annual power generation, equipment operation data includes but is not limited to the output power of the generator set, the generator speed, and the generator temperature, fuel consumption data includes but is not limited to fuel type, fuel consumption, and fuel inventory, environmental monitoring data includes but is not limited to greenhouse gas emissions and wastewater emissions, fault data includes but is not limited to equipment maintenance records, equipment fault records and repair time, the preset frequency T is preferably 1HZ, and the future duration t is preferably 1 hour.

[0046] Specifically, the embodiment of the present invention determines whether the production and operation data is qualified according to the comparison result of the missing rate L of the production and operation data and the preset missing rate L0;

[0047] When L≤L0, the production operation data is determined to be qualified;

[0048] When L>L0, the production operation data is determined to be unqualified;

[0049] In the embodiment of the present invention, the preset missing rate L0 is set to 0.1, and those skilled in the art may adjust the preset missing rate L0 according to specific circumstances.

[0050] Specifically, the present invention determines whether the production and operation data is qualified based on the comparison result of the missing rate of the production and operation data with the preset missing rate to ensure that the data used for further analysis and decision-making is more complete, which helps to improve the accuracy and reliability of data analysis, thereby improving the quality of early warning.

[0051] Specifically, when it is determined that the production and operation data is unqualified, the missing rate assessment coefficient F of the production and operation data is calculated according to the following formula, set:

[0052]

[0053] Among them, A1 represents the missing rate of power generation data, A2 represents the missing rate of equipment operation data, A3 represents the missing rate of fuel consumption data, A4 represents the missing rate of environmental monitoring data, A5 represents the missing rate of fault data, and A0 represents the average missing rate of different types of data in production operation data.

[0054] Specifically, when the missing rate assessment coefficient F of the production operation data is calculated, the preprocessing method for preprocessing the production operation data is determined according to the comparison result of the missing rate assessment coefficient F and the preset missing rate assessment coefficient F0;

[0055] When F≤F0, it is determined to preprocess the production operation data in the first preprocessing manner;

[0056] When F>F0, it is determined to pre-process the production operation data in the second pre-processing manner;

[0057] Wherein, the first preprocessing method is to extract the timestamps of missing data of different types of data in the production operation data;

[0058] The second preprocessing method is to arrange the missing rates of different types of data in the production operation data from large to small, and extract the timestamps of the missing data in the first K types of data.

[0059] In an embodiment of the present invention, the preset missing rate assessment coefficient F0 is 0.054. The preset missing rate assessment coefficient F0 is obtained when the missing rate of any type of data in the production operation data is 0.5 and the missing rate L of the production operation data is 0.1 and the missing rate assessment coefficient is taken as 60%. Technical personnel in this field can adjust the preset missing rate assessment coefficient F0 according to specific circumstances.

[0060] Specifically, the present invention calculates the missing rate assessment coefficient of production and operation data according to the formula to determine the preprocessing method for preprocessing the production and operation data according to the missing rate assessment coefficient of the production and operation data, so as to more specifically solve the data missing problems in different situations and adopt different strategies, thereby improving the effectiveness and efficiency of data preprocessing.

[0061] Specifically, when it is determined to preprocess the production operation data in the second preprocessing method, the missing rate difference between the missing rate of the current type of data and the missing rate of the next type of data is determined among the missing rates of different types of data arranged from large to small, and the missing rate of the current type of data with the maximum missing rate difference is taken as the node, and the serial number of the node in the missing rates of different types of data arranged from large to small is taken as the value of K.

[0062] Specifically, by identifying the turning points where the difference in missing rates is most significant, the present invention can more accurately locate the areas where the severity of the data missing problem changes the most. Targeted processing helps to concentrate resources and attention on solving the most critical data missing problems, thereby improving the efficiency and effectiveness of data preprocessing.

[0063] Specifically, when the production operation data is preprocessed in the first preprocessing manner, it is determined whether to perform nearest neighbor interpolation on the missing data in the different types of data according to the comparison result of the average time difference C between the timestamps of adjacent missing data in the different types of data and the preset average time difference C0;

[0064] When C≤C0, it is determined that the nearest neighbor interpolation is not performed on the missing data;

[0065] When C>C0, it is determined to perform nearest neighbor interpolation on the missing data;

[0066] In the embodiment of the present invention, the preset average time difference C0 is set to 300% of the time difference between the timestamps of two adjacent production operation data. Those skilled in the art may adjust the preset average time difference C0 according to specific circumstances.

[0067] Specifically, the present invention determines whether to perform nearest neighbor interpolation on missing data in different types of data based on the comparison result of the average time difference between the timestamps of adjacent missing data in different types of data with the preset average time difference, so as to significantly improve the integrity of the data and provide more continuous and complete data support for subsequent data analysis and decision-making.

[0068] Specifically, when it is determined that the nearest neighbor interpolation is not performed on the missing data or the production operation data is preprocessed in the second preprocessing method, a section of data that meets the preset conditions in the production operation data is extracted to train the first time series prediction model, and the timestamp of the missing data and the non-missing data in other types of data at the same timestamp as the missing data are input into the trained first time series prediction model to predict the missing data.

[0069] In an embodiment of the present invention, the preset condition is that the number of a segment of data in the extracted production and operation data is preferably 10% of the total number of production and operation data, and the missing rate of the segment of data is less than or equal to 0.01, and the first time series prediction model is preferably an autoregressive moving average model (ARMA).

[0070] Specifically, the present invention completes missing data through a time series prediction model to significantly improve the integrity of the data set and provide a more continuous and complete data basis for subsequent data analysis, decision support, etc. The time series prediction model can effectively utilize the time series characteristics of the data and accurately predict the missing data by analyzing the autocorrelation and trend of the data.

[0071] Specifically, when it is determined that the production operation data is qualified or the nearest neighbor interpolation of the missing data or the prediction of the missing data is completed, it is determined whether to issue a fault warning according to the comparison result between the ratio Z and the preset ratio Z0;

[0072] When Z≤Z0, it is determined that no fault warning will be issued;

[0073] When Z>Z0, it is determined to carry out fault warning;

[0074] In the embodiment of the present invention, the preset ratio Z0 is set to 0.1. Those skilled in the art may adjust the preset ratio Z0 according to specific circumstances. The second time series prediction model preferably adopts a long short-term memory network (LSTM).

[0075] Specifically, the present invention determines whether to perform a fault warning based on the comparison result between the ratio and the preset ratio to monitor abnormal changes in production operation data in real time, promptly discover signs that may lead to faults, thereby achieving early warning of faults and helping to take preventive measures to avoid or reduce losses caused by faults.

[0076] Please continue reading Figure 3 As shown, Figure 3 This is a schematic diagram of the structure of a system applying a five-dimensional integrated production operation management method according to an embodiment of the present invention.

[0077] The embodiment of the present invention applies a system based on a five-dimensional production operation management method, including:

[0078] A data collection unit, which is used to collect production and operation data of the power plant;

[0079] A data quality inspection unit, which is connected to the data collection unit and is used to determine the eligibility of the production and operation data according to the missing rate L of the production and operation data;

[0080] a data preprocessing unit connected to the data quality inspection unit, for determining a missing rate assessment coefficient F of the production operation data, and determining a preprocessing method for preprocessing the production operation data according to the missing rate assessment coefficient F, so as to determine the nearest neighbor interpolation of the missing data in the different types of data according to the average time difference C between the timestamps of adjacent missing data in the different types of data, or to predict the missing data according to the first time series prediction model;

[0081] A fault warning unit is respectively connected to the data quality inspection unit and the data preprocessing unit, and is used to train the second time series prediction model according to the production operation data, and use the trained second time series prediction model to predict the future duration t, and determine whether to issue a fault warning based on the ratio Z of the number of fault data predicted within the future duration t to the number of predicted data within the future duration t.

[0082] Specifically, the data acquisition unit of the present invention can collect the production and operation data of the power plant in real time to ensure the timeliness and integrity of the data. The data quality inspection unit can determine the eligibility of the data according to the missing rate of the production and operation data to improve the reliability and accuracy of the data. The data preprocessing unit can determine the preprocessing method for the production and operation data according to the missing rate assessment coefficient to ensure the integrity and consistency of the data. The fault warning unit can predict faults for the future duration by training the second time series prediction model, discover potential faults in advance, reduce the impact of faults on the production and operation of the power plant, and improve the operating efficiency and reliability of the power plant.

[0083] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0084] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A production operation management method based on five-dimensional integration, characterized in that: include: Collecting production and operation data of the power plant at a preset frequency, wherein the production and operation data includes power generation data, equipment operation data, fuel consumption data, environmental monitoring data, and fault data; Determine the eligibility of production and operation data based on the missing rate of production and operation data; Determine a missing rate assessment coefficient for production and operation data, and determine a preprocessing method for preprocessing the production and operation data according to the missing rate assessment coefficient; The method of determining the preprocessing method for preprocessing the production operation data includes arranging the missing rates of different types of data in the production operation data from large to small under the condition that the missing rate evaluation coefficient is less than or equal to the preset missing rate evaluation coefficient, and extracting the timestamps of the missing data in the first K types of data; The timestamp of the missing data in the first K types of data is included in the missing rates of different types of data arranged from large to small, and the missing rate difference between the missing rate of the current type of data and the missing rate of the next type of data is determined, so as to obtain the missing rate of the current type of data with the maximum missing rate difference as a node, and the sequence number of the node in the missing rates of different types of data arranged from large to small is taken as the value of K; Determining the nearest neighbor interpolation for missing data in different types of data according to the average time difference between timestamps of adjacent missing data in different types of data, wherein determining the nearest neighbor interpolation for missing data in different types of data includes performing nearest neighbor interpolation on the missing data under the condition that the average time difference is greater than a preset average time difference; Or predicting the missing data according to the first time series prediction model, wherein predicting the missing data according to the first time series prediction model includes extracting a section of data that meets preset conditions in the production operation data to train the first time series prediction model, and inputting the timestamp of the missing data and non-missing data in other types of data at the same timestamp as the missing data into the trained first time series prediction model to predict the missing data; Alternatively, a second time series prediction model is trained based on production operation data, and the trained second time series prediction model is used to predict the future duration, and whether to issue a fault warning is determined based on the ratio of the number of fault data predicted in the future duration to the number of predicted data in the future duration.

2. The five-dimensional integrated production operation management method according to claim 1 is characterized in that: Determining the eligibility of the production operation data includes determining that the production operation data is unqualified under the condition that the missing rate of the production operation data is greater than a preset missing rate.

3. The five-dimensional integrated production operation management method according to claim 2 is characterized in that: The missing rate evaluation coefficient is calculated according to the following formula: ； Among them, F represents the missing rate assessment coefficient, A1 represents the missing rate of power generation data, A2 represents the missing rate of equipment operation data, A3 represents the missing rate of fuel consumption data, A4 represents the missing rate of environmental monitoring data, A5 represents the missing rate of fault data, and A0 represents the average missing rate of different types of data in production operation data.

4. The five-dimensional integrated production operation management method according to claim 3 is characterized in that: The method of determining the preprocessing method for preprocessing the production operation data includes extracting the timestamps of missing data of different types of data in the production operation data under the condition that the missing rate assessment coefficient is less than or equal to a preset missing rate assessment coefficient.

5. The five-dimensional integrated production operation management method according to claim 4 is characterized in that: The determining whether to perform a fault warning includes determining whether to perform a fault warning under the condition that the ratio is greater than a preset ratio.

6. A system using the five-dimensional integrated production operation management method according to any one of claims 1 to 5, characterized in that: include: A data collection unit, which is used to collect production and operation data of the power plant; A data quality inspection unit, which is connected to the data collection unit and is used to determine the eligibility of the production and operation data according to the missing rate of the production and operation data; a data preprocessing unit connected to the data quality inspection unit, and used to determine a missing rate assessment coefficient of the production operation data, and determine a preprocessing method for preprocessing the production operation data according to the missing rate assessment coefficient, so as to determine the nearest neighbor interpolation of the missing data in the different types of data according to the average time difference between the timestamps of adjacent missing data in the different types of data, or predict the missing data according to the first time series prediction model; A fault warning unit is respectively connected to the data quality inspection unit and the data preprocessing unit, and is used to train the second time series prediction model according to the production operation data, predict the future duration with the trained second time series prediction model, and determine whether to issue a fault warning based on the ratio of the number of fault data predicted in the future duration to the number of predicted data in the future duration.

Images