Progressive multi-stage early warning method and system for power enterprise production indexes

Abstract

The invention discloses a progressive multistage early warning method and system for power enterprise production indexes, and relates to the technical field of power enterprise production management, and the method comprises the steps: constructing a three-stage early warning mode composed of primary early warning, intermediate early warning and advanced early warning based on the power enterprise production indexes, and obtaining the progressive relation between two adjacent early warning modes; based on the threshold value and the judgment condition of each early warning mode, a first early warning mode is triggered, according to the progressive relation, a second early warning mode is triggered, the first early warning mode represents any one of three-level early warning modes, and the second early warning mode is different from the first early warning mode. According to the invention, through a multi-level and progressive early warning mechanism, comprehensive and accurate monitoring of power production indexes is realized, so that potential risks are found in advance, the probability of occurrence of production accidents is reduced, and safe production and efficient operation of power enterprises are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of production management of electric power enterprises, and in particular to a progressive multi-level early warning method and system for production indicators of electric power enterprises. Background Art

[0002] The traditional power company production index early warning mainly relies on a single limit early warning method. In this method, power companies will set a fixed early warning limit for each production index according to industry standards and internal corporate regulations. When the production index data exceeds this limit, the system will trigger an early warning to remind relevant personnel to take measures to deal with it. However, this method has obvious limitations. Since it only focuses on whether the production index exceeds the limit, but ignores the subtle trends and operating characteristics of the index changes, it often leads to problems such as delayed early warning and high false alarm rate.

[0003] In recent years, with the development of big data and machine learning technologies, some power companies have begun to try to use early warning methods based on statistical models. This method collects a large amount of production indicator data, uses machine learning algorithms to model and analyze the data, and thus predicts the changing trends and potential risks of production indicators. Although this method has improved the accuracy and timeliness of early warning to a certain extent, there are still some problems. For example, statistical models often require strict assumptions and preprocessing of data, and the stability and robustness of the model are also easily affected by factors such as data quality and changes in operating conditions.

[0004] Although the traditional single limit warning method and the warning method based on statistical model can meet the warning needs of power companies to a certain extent, they both have some obvious limitations and shortcomings.

[0005] 1. Traditional single limit warning method:

[0006] In terms of technology, this method mainly relies on fixed warning limits and lacks consideration of different working conditions and indicator change trends. Therefore, in practical applications, it is often difficult to capture subtle anomalies and potential risks in the production process, resulting in delayed warning.

[0007] In terms of performance, this method only focuses on whether the limit is exceeded, which is prone to false alarms and missed alarms. False alarms will increase unnecessary human and material resources, while missed alarms may cause serious production accidents.

[0008] 2. Early warning method based on statistical model:

[0009] Although this method can use big data and machine learning technology for modeling and analysis, the model construction and optimization process is often complicated and requires professional technicians to operate. In addition, the model has high requirements for data quality and integrity. Once there is a problem with the data, the warning effect of the model will be greatly reduced.

[0010] In terms of performance, the stability and robustness of the statistical model of this method are easily affected by factors such as changes in working conditions and data fluctuations. In practical applications, these factors may cause the early warning effect of the model to be unstable or even invalid. In addition, the cost of updating and maintaining the model is relatively high, which increases the operating costs of the enterprise.

[0011] Both the traditional single limit warning method and the warning method based on statistical models have obvious deficiencies in process or performance.

[0012] Therefore, it is urgent to propose a new progressive multi-level early warning technology for power enterprise production indicators to overcome the shortcomings of existing technologies, thereby improving the accuracy and timeliness of early warnings and providing strong support for the safe production and efficient operation of power enterprises. Summary of the invention

[0013] In order to solve the above problems, the purpose of the present invention is to provide a progressive multi-level early warning technology for the production indicators of power enterprises, which aims to comprehensively and accurately monitor the production indicators of power enterprises through a multi-level, progressive early warning mechanism, so as to timely discover potential risks, ensure the safe production and efficient operation of power enterprises, and thus improve the production management level of power enterprises and reduce the probability of production accidents, which is of great significance.

[0014] In order to achieve the above technical objectives, the present application provides a progressive multi-level early warning method for production indicators of power enterprises, comprising the following steps:

[0015] Based on the production indicators of power enterprises, a three-level warning model consisting of primary warning, intermediate warning and advanced warning is constructed to obtain the progressive relationship between two adjacent warning models;

[0016] Based on the threshold and judgment conditions of each warning mode, the first warning mode is triggered, and according to the progressive relationship, the second warning mode is triggered, wherein the first warning mode represents any one of the three-level warning modes, and the second warning mode is different from the first warning mode.

[0017] Preferably, in response to triggering the first warning mode, based on the input parameter data set of the power enterprise production index, an over-limit warning diagnosis is performed according to the regulation limit, and the state based on the regulation over-limit is preferentially identified to trigger the first warning mode.

[0018] Preferably, in response to triggering the second warning mode, while triggering the first warning mode, a fluctuation warning judgment based on the working conditions is performed for parameters that do not exceed the regulatory limit, potential anomalies in the production indicators are identified, and the second warning mode is triggered.

[0019] Preferably, in the process of triggering the second warning mode, after triggering the first warning mode or the second warning mode, for the parameters that have not undergone fluctuation warnings, it is detected whether the parameters have a degradation trend based on the parameter change rate. If the change rate has a degradation trend, based on the progressive relationship, a third warning mode different from the first warning mode or the second warning mode is triggered, wherein the third warning mode represents a trend warning, which is used to indicate that the system or equipment may be experiencing some kind of gradual failure or performance degradation.

[0020] Preferably, in the process of obtaining the production indicators of the power enterprise, the collected data is cleaned, format converted and missing value processed to ensure the accuracy and completeness of the data.

[0021] Preferably, in the process of triggering the second warning mode, for the production indicators that have not triggered the fluctuation warning, the adaptive algorithm is used to calculate their change rate, and the trend warning threshold is set according to the change rate as the basis for triggering the second warning mode.

[0022] Preferably, in response to the construction of the three-level early warning mode, a 64×3 abnormal state matrix of the production index parameters of the power enterprise is constructed according to the time dimension and abnormal category of the abnormal parameters corresponding to each early warning mode, and 64 different abnormal states are characterized in detail; when the state is judged to be normal, the edge side performs data compression processing according to the data compression model sent from the cloud or pre-configured; when the abnormal state is abnormal, the edge side node is automatically triggered to enter the high-frequency mode, the abnormal state is marked, and real-time data recording storage is started to retain the original high-frequency data for subsequent abnormal analysis and fault diagnosis tracing.

[0023] Preferably, when the status is judged to be normal, different storage strategies are adopted according to different production indicators of the power enterprise to meet the efficient retrieval requirements of real-time high-frequency data, historical data, abnormal event data, transition process data, test data, and sample data.

[0024] Preferably, in response to triggering the three-level warning mode, a warning response layer is set to execute warning notification, emergency response, personnel dispatch, equipment inspection, troubleshooting, feedback mechanism, effect evaluation and improvement measures, so as to formulate corresponding warning response strategies and disposal measures according to the warning level and type, and promptly convey the warning information to relevant personnel or systems so as to quickly respond to and deal with potential risks.

[0025] The present invention provides a progressive multi-level early warning system for production indicators of electric power enterprises, which is used to implement the above-mentioned progressive multi-level early warning method for production indicators of electric power enterprises, including:

[0026] The data processing module is used to construct a three-level warning mode consisting of primary warning, intermediate warning and advanced warning based on the production indicators of the power enterprise, and obtain the progressive relationship between two adjacent warning modes;

[0027] The early warning response module is used to trigger the first early warning mode based on the threshold and judgment conditions of each early warning mode, and trigger the second early warning mode according to the progressive relationship, wherein the first early warning mode represents any one of the three-level early warning modes, and the second early warning model is different from the first early warning mode.

[0028] The present invention discloses the following technical effects:

[0029] The present invention improves the accuracy and timeliness of early warning: by constructing a progressive multi-level early warning system, combined with real-time data and historical trends of production indicators, rapid response and precise positioning of abnormal situations can be achieved.

[0030] The present invention enhances the stability and robustness of the early warning system: by adopting advanced algorithms and models, the influence of operating condition changes and data fluctuations is considered, and the stability and reliability of the early warning system in complex environments are improved.

[0031] The present invention reduces the maintenance cost of the early warning system: by optimizing the algorithm structure and simplifying the operation process, the technical threshold and maintenance cost of the early warning system are reduced, which is convenient for power companies to promote and use in practical applications.

[0032] The present invention improves the production management level of power enterprises and ensures the safety and reliability of power supply. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0034] Figure 1 It is a schematic diagram of the system structure of the present invention;

[0035] Figure 2 It is a schematic diagram of the method flow of the present invention. DETAILED DESCRIPTION

[0036] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0037] like Figure 1-2 As shown, the present invention provides a progressive multi-level early warning method and system for production indicators of power enterprises. By constructing a progressive multi-level early warning system, combined with real-time monitoring data and historical data trends of production indicators, comprehensive monitoring and early warning of the power production process are achieved. The system includes four levels: data acquisition layer, data processing layer, early warning analysis layer and early warning response layer. Each level contains corresponding technical means and methods to ensure the accuracy and timeliness of the early warning system.

[0038] 1. Data collection layer:

[0039] Tools and equipment: Use advanced sensors and Internet of Things technology to collect real-time data on key indicators in the power production process, such as voltage, current, temperature, pressure, etc.

[0040] Data range: Ensure the accuracy and completeness of data collection. The frequency of data collection is set from every second to every minute according to production needs to capture subtle changes in the production process.

[0041] 2. Data processing layer:

[0042] Data cleaning: Use data cleaning algorithms, such as statistical-based methods and rule-based methods, to remove outliers, duplicate values, and missing values ​​to ensure data quality.

[0043] Data standardization: Standardize data of different dimensions, such as converting data into dimensionless Z scores or percentiles, for subsequent analysis.

[0044] Data storage: Use distributed database or cloud storage technologies, such as Hadoop and MongoDB, to ensure data reliability and scalability while supporting efficient data query and analysis.

[0045] 3. Early warning analysis layer:

[0046] Multi-level warning model: Build a progressive multi-level warning model, including three levels: primary warning, intermediate warning and advanced warning.

[0047] Primary warning: When the production indicator data exceeds the preset primary threshold, a primary warning is triggered to prompt relevant personnel to pay attention.

[0048] Intermediate warning: When production indicator data continues to be abnormal or exceeds the intermediate threshold, an intermediate warning is triggered, requiring relevant personnel to take preliminary measures.

[0049] Advanced warning: When production indicator data shows serious abnormalities or may cause production accidents, an advanced warning is triggered, requiring immediate emergency measures.

[0050] The warning thresholds and judgment conditions for each level are set based on historical data and expert experience, and adjusted according to actual conditions.

[0051] The three levels of warning, primary, medium and high, can be constructed and take effect separately.

[0052] Independence: Although the three levels of warning are progressive, the warning model of each level can be relatively independent when it is constructed. Each level of warning model can be constructed based on different data features, algorithms or thresholds to meet the warning needs of different levels.

[0053] Separate effects: In actual applications, when the production indicator data reaches a certain level of warning threshold, the warning model at that level will be triggered and generate corresponding warning information. The three levels of warning, primary, intermediate and high, can act separately, without waiting for the previous level of warning to be processed before triggering the next level of warning.

[0054] Progressive relationship: Although each level of warning can work independently, there is also a progressive relationship between them. When the primary warning is triggered, it means that the production indicator data has a slight abnormality and needs the attention of relevant personnel; when the intermediate warning is triggered, it may mean that the production indicator data continues to be abnormal or exceeds a more stringent threshold, and preliminary measures need to be taken; when the advanced warning is triggered, it may mean that the production indicator data has a serious abnormality or may cause a production accident, and emergency measures need to be taken immediately.

[0055] Thresholds and judgment conditions: The warning thresholds and judgment conditions for each level are set based on historical data and expert experience, and can be adjusted according to actual conditions.

[0056] Algorithm optimization: Use machine learning algorithms, such as support vector machines, random forests, neural networks, etc., to train and optimize early warning models to improve the accuracy and timeliness of early warnings.

[0057] Algorithm parameters: including learning rate, number of iterations, regularization parameters, etc., can be adjusted according to actual conditions.

[0058] Model evaluation: Evaluate the performance of the model through cross-validation, ROC curve and other methods to ensure the reliability of the early warning system.

[0059] Combined with the single limit warning method in the traditional mode, the fluctuation warning interval and adaptive threshold trend warning of the working condition characteristics are considered to construct a progressive multi-level warning mechanism. For example, for the input parameter data set, firstly, the limit-exceeding warning diagnosis is performed based on the regulation limit, and the state based on the regulation limit-exceeding is identified first, and the alarm level is the highest. Secondly, for the parameters that do not exceed the regulation limit, the fluctuation warning judgment based on the working condition is performed to more specifically identify the potential abnormalities of the current power enterprise production indicators. Finally, for the parameters that have not issued a fluctuation warning, although no alarm has occurred at the current calculation moment, based on the perspective of the parameter change rate, whether the parameter has a deterioration trend, if the change rate has a deterioration trend, a trend warning is issued. In this way, the progressive detection and judgment are carried out layer by layer, and the situation where no warning occurs in three rounds of detection is defined as a normal state.

[0060] In order to clearly understand the current operating status of the device, the parameters are defined as follows according to the abnormal situations at different times in the time dimension:

[0061] Historical anomaly: refers to the abnormal situation of the status parameter at the last moment.

[0062] Current abnormality: refers to the abnormality of the status parameters at the current moment.

[0063] Future anomaly: refers to the possibility of abnormal conditions in the state parameters in the next moment.

[0064] The device status at each moment can be divided into four states: over-limit abnormality, fluctuation abnormality, trend abnormality, and normal. The abnormality level identification is shown as follows.

[0065] Table 1 Abnormal level identification

[0066]

[0067] The normal state means that the operating parameters of the system or equipment fluctuate within the prescribed range, and the amplitude and frequency of fluctuations meet normal standards, without obvious abnormalities or failures.

[0068] Abnormal level: No abnormality, normal operation status.

[0069] An over-limit abnormality means that one or more operating parameters of a system or device exceed the preset upper and lower alarm limits. This usually means that the system or device may have a fault or abnormality that requires immediate attention and processing.

[0070] The abnormality level is high. Out-of-limit abnormalities are usually regarded as an emergency situation, and immediate measures must be taken to prevent the situation from deteriorating further.

[0071] Fluctuation abnormality means that the operating parameters of the system or equipment have fluctuated significantly in a short period of time, but have not exceeded the preset alarm upper and lower limits. This fluctuation may be caused by external interference, internal failure or system instability. The abnormality level is medium. Although the fluctuation abnormality has not reached the level of exceeding the limit, it still needs to be paid attention to in order to promptly discover and deal with potential failures or problems.

[0072] Trend anomaly refers to the obvious trend change of the operating parameters of the system or equipment over a period of time. This change may exceed the normal fluctuation range, but has not yet reached the level of over-limit. Trend anomaly usually means that the system or equipment may be experiencing some kind of gradual failure or performance degradation. The abnormality level is low (but pay attention to potential risks). Although trend anomaly is not as urgent as over-limit anomaly, it still needs to be taken seriously enough so that timely measures can be taken to prevent further aggravation of failure or performance degradation.

[0073] Further combining the time dimension and abnormal category of abnormal parameters, a 64x3 abnormal state matrix of power enterprise production index parameters is constructed to characterize 64 different abnormal states. When the state is judged to be normal, the edge side performs data compression processing according to the data compression model sent from the cloud or pre-configured. When the abnormal state is abnormal, the edge side node is automatically triggered to enter the high-frequency mode, mark the abnormal state, and start real-time data recording and storage, so as to retain the original high-frequency data for subsequent abnormal analysis and fault diagnosis and tracing.

[0074] Table 2 Early warning status matrix of production index parameters of power enterprises

[0075]

[0076]

[0077]

[0078] The multi-level early warning mechanism realizes dynamic adaptive storage. In the stable state, data will adopt different storage strategies according to the different production indicators of the power company to meet the efficient retrieval requirements of real-time high-frequency data, historical data (including various characteristic data), abnormal event data, transition process data, test data, sample data, etc. In abnormal events and transition processes, the original data and characteristic data are recorded continuously and at high density for easy playback and analysis. The recorded data includes the original waveform data directly collected by various sensors, monitoring communication data, characteristic values ​​obtained by calculation and analysis, etc., and stores continuous and uninterrupted data for 5 minutes before and after the event. In the case of early warning, alarm and system abnormality, various early warning and alarm data, communication status, operation actions, and system abnormality information will be recorded at the same time.

[0079] 4. Early warning response layer

[0080] Early warning notification: Early warning information will be sent to relevant personnel in real time through SMS, email, APP push, etc. to ensure timely delivery of early warning information.

[0081] Emergency response: Develop corresponding emergency response plans based on the warning level, including personnel dispatch, equipment inspection, troubleshooting, etc.

[0082] Personnel dispatch: According to the warning level and actual situation, relevant personnel are dispatched to the scene to handle the situation.

[0083] Equipment inspection: Conduct detailed inspections on equipment that may have problems to ensure that the equipment is operating normally.

[0084] Troubleshooting: Troubleshoot faults in the production process, find out the causes of the faults and take measures to repair them.

[0085] Feedback mechanism: Establish a feedback mechanism for early warning response, collect early warning effects and improvement measures, and continuously optimize the early warning system.

[0086] Effectiveness evaluation: Evaluate the effectiveness of the early warning system, including warning accuracy, response time, etc.

[0087] Improvement measures: According to the evaluation results and actual conditions, improve and optimize the early warning system to improve its performance and reliability.

[0088] The specific steps include:

[0089] Basic data collection and preprocessing: Collect various production index data of power enterprises in the production process, including but not limited to power generation, power supply, line loss rate, equipment utilization rate, etc. Preprocess the collected data, including data cleaning, format conversion, missing value processing, etc., to ensure the accuracy and completeness of the data.

[0090] Regulation limit over-limit warning: According to the power industry standards and internal enterprise regulations, the regulation limits of various production indicators are set. Real-time monitoring of production indicator data, when an indicator exceeds the regulation limit, an over-limit warning is immediately triggered and marked as the highest level of warning status (over-limit warning).

[0091] Operating condition fluctuation warning: For production indicators that have not triggered the over-limit warning, the fluctuation warning range is calculated by combining historical data and current operating condition characteristics. When the production indicator data fluctuates within the fluctuation warning range, the fluctuation warning is triggered and marked as the second highest level warning state (fluctuation warning).

[0092] Adaptive threshold trend warning: For production indicators that have not triggered fluctuation warnings, the adaptive algorithm is used to calculate their change rate, and the trend warning threshold is set according to the change rate. When the change rate of production indicator data exceeds the trend warning threshold, a trend warning is triggered and marked as a lower level warning state (trend warning).

[0093] Comprehensive judgment and warning output: Based on the comprehensive judgment results of the above three-level warning mechanism, the warning status of the current production index is output. At the same time, according to the definitions of historical anomalies, current anomalies and future warnings, the status of production indicators is divided and marked in the time dimension.

[0094] Early warning response and disposal: According to the warning level and type, formulate corresponding early warning response strategies and disposal measures. Communicate early warning information to relevant personnel or systems in a timely manner to quickly respond to and deal with potential risks.

[0095] In summary, the present invention has achieved significant technical innovations in terms of constructing a progressive multi-level early warning system, optimization based on machine learning algorithms, comprehensive consideration of operating condition changes and data fluctuations, flexible early warning response mechanisms, and data-driven decision support. These key technical points together constitute the core technical solution of the present invention, providing a strong guarantee for the safe production and efficient operation of power companies.

[0096] 1. Construction of a progressive multi-level early warning system:

[0097] The present invention constructs a progressive multi-level early warning system, which includes three levels: primary early warning, intermediate early warning and advanced early warning. Each level sets corresponding early warning thresholds and judgment conditions to ensure comprehensive monitoring and timely response to changes in production indicators.

[0098] Compared with the traditional single limit warning method, the progressive multi-level warning system can more accurately capture abnormal situations in the production process, avoid false alarms and missed alarms, and improve the accuracy and timeliness of the warning.

[0099] 2. Optimization based on machine learning algorithms:

[0100] At the early warning analysis layer, the present invention uses machine learning algorithms to train and optimize the early warning model. These algorithms can automatically adjust the early warning threshold and judgment conditions based on historical data and real-time data to adapt to different working conditions and production needs.

[0101] Through the optimization of machine learning algorithms, the early warning model can more accurately predict the changing trends and potential risks of production indicators, and improve the intelligence level and warning accuracy of the early warning system.

[0102] 3. Comprehensively consider changes in working conditions and data fluctuations:

[0103] In the early warning analysis process, the present invention fully considers the impact of working condition changes and data fluctuations on the early warning effect. By introducing technical means such as working condition identification and data smoothing, the stability and robustness of the early warning system in complex environments are improved.

[0104] This technical feature enables the early warning system to better adapt to the actual situation in the production process and reduce false alarms and missed alarms caused by changes in working conditions and data fluctuations.

[0105] 4. Flexible early warning response mechanism:

[0106] The present invention designs a set of flexible early warning response mechanisms, including early warning notification, emergency response and feedback mechanisms. These mechanisms can quickly start corresponding emergency plans according to different early warning levels and actual conditions to ensure production safety.

[0107] The flexible early warning response mechanism enables the early warning system to respond to abnormal situations in the production process in a timely manner and reduce the risk of production accidents. At the same time, through the establishment of a feedback mechanism, the performance and effect of the early warning system can be continuously optimized.

[0108] 5. Data-driven decision support:

[0109] The present invention also uses big data and artificial intelligence technology to provide data-driven decision support functions. By analyzing and mining production index data, it can provide a scientific basis for the production management and optimization of power companies.

[0110] Data-driven decision support functions enable power companies to more accurately understand the actual situation and problems in the production process, providing strong support for production optimization and management decisions.

[0111] The progressive multi-level early warning method for production indicators of power enterprises proposed in the present invention has shown significant advantages and effects in many aspects compared with existing early warning technologies. The progressive multi-level early warning method for production indicators of power enterprises proposed in the present invention has the advantages of significantly improving the accuracy and timeliness of early warning, enhancing the stability and robustness of the system, flexible early warning response mechanism and decision support, reducing energy consumption and raw material costs, and being beneficial to environmental protection and reducing labor intensity. These advantages make the present invention have broad application prospects and important practical value in the production management and optimization of power enterprises.

[0112] 1. Significant improvement in the accuracy and timeliness of early warning

[0113] Through the progressive multi-level warning system, the present invention can more accurately capture abnormal situations in the production process and issue warnings in the first place. Compared with the traditional single limit warning method, the warning accuracy of the present invention is improved by about 30%, and the warning response time is shortened by more than 50%.

[0114] In a simulated production environment, we conducted long-term monitoring and early warning tests on a variety of production indicators. The results show that when production indicators fluctuate abnormally, the present invention can issue an early warning within seconds to minutes, while traditional methods often take minutes to hours to detect abnormalities.

[0115] 2. Enhanced system stability and robustness

[0116] The present invention fully considers the impact of working condition changes and data fluctuations on the warning effect, and improves the stability and robustness of the warning system in complex environments by introducing technical means such as working condition identification and data smoothing. Even under harsh working conditions, the warning system of the present invention can maintain a high warning accuracy.

[0117] By statistically analyzing the production index data under different working conditions, we found that working condition changes and data fluctuations are one of the main reasons for false alarms and missed alarms. Therefore, the present invention adds a working condition identification module and a data smoothing algorithm to the early warning model to reduce the impact of these factors.

[0118] 3. Flexible early warning response mechanism and decision support

[0119] The present invention designs a set of flexible early warning response mechanisms, including early warning notification, emergency response and feedback mechanisms. These mechanisms can quickly start corresponding emergency plans according to different early warning levels and actual conditions to ensure production safety. At the same time, the present invention also provides data-driven decision support functions, providing a scientific basis for production management and optimization of power companies.

[0120] In actual applications, we simulated a variety of production abnormalities and tested the effectiveness of the early warning response mechanism. The results showed that when the early warning system issued an early warning, relevant personnel were able to receive the early warning notification within a few minutes and take corresponding measures according to the emergency plan. In addition, through data-driven decision support functions, we successfully identified potential problems in multiple production processes and put forward corresponding optimization suggestions.

[0121] 4. Reduction of energy consumption and raw material costs

[0122] Since the present invention can detect abnormal conditions in the production process in advance and take corresponding measures to prevent and control them, it can significantly reduce energy consumption and raw material costs in the production process. Compared with enterprises that do not adopt the present invention, enterprises that adopt the present invention reduce energy consumption by about 10% and save about 5% of raw material costs.

[0123] Abnormal conditions in the production process often lead to increased energy consumption and raw material costs. For example, equipment failure may lead to increased energy consumption, while quality problems in the production process may lead to waste of raw materials. By discovering these problems in advance and taking corresponding measures to prevent and control them, the present invention can effectively reduce energy consumption and raw material costs.

[0124] 5. It is beneficial to environmental protection and reduces labor intensity

[0125] The early warning system of the present invention can monitor the key indicator data of the production process in real time and discover potential environmental problems and safety hazards in advance. This helps enterprises to take timely measures to rectify and optimize, thereby reducing pollution and damage to the environment. At the same time, since the early warning system can automatically issue early warnings and activate emergency plans, it can significantly reduce the labor intensity and work pressure of relevant personnel.

[0126] The early warning system of the present invention has the characteristics of real-time monitoring, automatic early warning and intelligent decision-making, and can provide comprehensive production monitoring and management services for enterprises. These characteristics enable enterprises to deal with various problems and challenges in the production process more efficiently, thereby reducing pollution and damage to the environment, and reducing the labor intensity and work pressure of relevant personnel.

[0127] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0128] In the description of the present invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0129] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A progressive multi-level early warning method for production indicators of power enterprises, characterized in that: The following steps are involved: Based on the production indicators of power enterprises, a three-level early warning model consisting of primary warning, intermediate warning and advanced warning is constructed to obtain the progressive relationship between two adjacent early warning models; Based on the threshold and judgment condition of each warning mode, the first warning mode is triggered, and according to the progressive relationship, the second warning mode is triggered, wherein the first warning mode represents any one of the three-level warning modes, and the second warning mode is different from the first warning mode.

2. According to claim 1, a progressive multi-level early warning method for production indicators of a power enterprise is characterized by: In response to triggering the first warning mode, based on the input parameter data set of the power enterprise production index, an over-limit warning diagnosis is performed according to the regulation limit, and the status based on the regulation over-limit is preferentially identified to trigger the first warning mode.

3. According to claim 2, a progressive multi-level early warning method for production indicators of a power enterprise is characterized by: In response to triggering the second warning mode, while triggering the first warning mode, a fluctuation warning judgment based on the working conditions is performed for parameters that do not exceed the regulatory limit, potential anomalies in the production indicators are identified, and the second warning mode is triggered.

4. According to claim 3, a progressive multi-level early warning method for production indicators of a power enterprise is characterized by: In the process of triggering the second warning mode, after triggering the first warning mode or the second warning mode, for the parameters that have not caused fluctuation warnings, detect whether the parameters have a degradation trend based on the parameter change rate. If the change rate has a degradation trend, trigger the third warning mode different from the first warning mode or the second warning mode based on the progressive relationship, wherein the third warning mode represents a trend warning, which is used to indicate that the system or equipment may be experiencing some gradual failure or performance degradation.

5. According to claim 4, a progressive multi-level early warning method for production indicators of power enterprises is characterized by: In the process of obtaining the production indicators of power companies, the collected data are cleaned, format converted and missing value processed to ensure the accuracy and completeness of the data.

6. According to claim 5, a progressive multi-level early warning method for production indicators of a power enterprise is characterized by: In the process of triggering the second warning mode, for the production indicators that have not triggered the fluctuation warning, the adaptive algorithm is used to calculate their change rate, and the trend warning threshold is set according to the change rate as the basis for triggering the second warning mode.

7. A progressive multi-level early warning method for production indicators of a power enterprise according to claim 6, characterized in that: In response to the construction of the three-level warning mode, a 64×3 abnormal state matrix of power enterprise production indicator parameters is constructed according to the time dimension and abnormal category of the abnormal parameters corresponding to each warning mode, and 64 different abnormal states are characterized in detail; when the state is judged to be normal, the edge side performs data compression processing according to the data compression model sent from the cloud or pre-configured; When the abnormal state is abnormal, the edge node is automatically triggered to enter the high-frequency mode, the abnormal state is marked, and real-time data recording storage is started to retain the original high-frequency data for subsequent abnormal analysis and fault diagnosis tracing.

8. According to claim 7, a progressive multi-level early warning method for production indicators of a power enterprise is characterized by: When the status is judged to be normal, different storage strategies are adopted according to the different production indicators of the power company to meet the efficient retrieval needs of real-time high-frequency data, historical data, abnormal event data, transition process data, test data, and sample data.

9. A progressive multi-level early warning method for production indicators of a power enterprise according to claim 8, characterized in that: In response to triggering the three-level warning mode, the warning response layer is set to execute warning notification, emergency response, personnel dispatch, equipment inspection, troubleshooting, feedback mechanism, effect evaluation and improvement measures. It is used to formulate corresponding warning response strategies and disposal measures according to the warning level and type, and promptly convey the warning information to relevant personnel or systems to quickly respond to and deal with potential risks.

10. A progressive multi-level early warning system for production indicators of electric power enterprises, used to implement a progressive multi-level early warning method for production indicators of electric power enterprises as claimed in any one of claims 1 to 9, characterized in that: include: The data processing module is used to construct a three-level warning mode consisting of primary warning, intermediate warning and advanced warning based on the production indicators of the power enterprise, and obtain the progressive relationship between two adjacent warning modes; An early warning response module is used to trigger a first early warning mode based on the threshold and judgment conditions of each early warning mode, and to trigger a second early warning mode according to the progressive relationship, wherein the first early warning mode represents any one of the three levels of early warning modes, and the second early warning model is different from the first early warning mode.

Images