A new energy index evaluation and early warning analysis method based on multi-source heterogeneous traffic

Abstract

The present application relates to the technical field of index examination and early warning, and particularly relates to a new energy index examination and early warning analysis method based on multi-source heterogeneous flow.The present application screens out target examination indexes from multiple examination indexes which are easily covered by normal fluctuations of data sources through index correlation inundation degree, and then determines abnormal tendency index set of the target examination indexes according to parameter mismatch characteristics and data asynchronous deviation characteristics.On this basis, the interlocking influence range is demarcated according to the shared data source correlation relationship between abnormal index items, and whether the interlocking influence range exists batch out-of-control risk tendency is determined through abnormal diffusion degree, so as to early identify and early warn and push in the initial stage of shared data source abnormality, eliminate the influence of the identification blind area of early abnormal signals of data sources by the conventional standard judgment, and improve the identification timeliness of new energy index examination and early warning and the safety of power station operation.

Description

Technical Field

[0001] This invention relates to the field of indicator assessment and early warning technology, and in particular to a new energy indicator assessment and early warning analysis method based on multi-source heterogeneous flow. Background Technology

[0002] With the rapid development of the new energy industry, the large-scale construction and grid-connected operation of new energy power generation facilities such as photovoltaic power plants have placed higher demands on the monitoring of power plant operation status, control of performance indicators, and early warning of anomalies. Currently, the monitoring systems for new energy power plants contain traffic data from multiple sources and of different types, including telemetry, remote signaling, and remote adjustment data from remote control, AGC, AVC, and expanded AGC services, as well as link status, message data, power prediction data, and resource data, forming a typical multi-source heterogeneous traffic characteristic.

[0003] In existing technologies, the assessment of new energy indicators often adopts a single-dimensional data statistics or independent indicator analysis model, which has the following drawbacks: First, multi-source heterogeneous flow data lacks a unified collection, cleaning, and fusion mechanism, and data from different businesses and in different formats cannot be effectively linked, resulting in large deviations in the calculation of assessment indicators; Second, indicator assessment and early warning analysis are independent of each other, and early warning triggering often relies on a single judgment of indicator thresholds, lacking correlation analysis of heterogeneous flow data, resulting in insufficient timeliness and accuracy of early warnings; Finally, existing methods are difficult to adapt to the differentiated needs of multiple types of assessment indicators, and cannot dynamically adjust assessment parameters and early warning thresholds according to the actual operation scenarios of power plants, resulting in poor flexibility and practicality.

[0004] Chinese Patent Publication No. CN115860163A discloses a method and system for evaluating the deviation of new energy power generation based on system operation indicators. Based on historical operating data, the method determines the monthly market prices for positive and negative deviations of the power system; based on the day-ahead market clearing results, it obtains the winning output curves and clearing prices of new energy power plants; based on actual operating conditions, it obtains the actual load curve of the power grid and the actual output curve of the new energy power plants; based on the obtained predicted and actual curves, it calculates the power system load deviation and the power generation output deviation of the new energy power plants; when the deviation trends of the power system load deviation and the power generation output deviation of the new energy power plants are opposite, the corresponding new energy power plants are assessed. This disclosure incorporates the comparison of deviation trends into the construction of the deviation assessment model, and links the assessment cost to the actual dispatching situation of the power system, avoiding excessive penalties for new energy power plants and allowing them to reasonably share the system dispatching costs.

[0005] However, the following problems still exist in the existing technology. In the assessment and early warning of new energy indicators, the existing analysis methods cannot make full use of the correlation between multi-source heterogeneous flow data, making it difficult to identify abnormal trends in indicators in advance. This reduces the timeliness and accuracy of assessment and early warning, as well as the reliability of new energy power plant operation. Summary of the Invention

[0006] To address this, the present invention provides a new energy indicator assessment and early warning analysis method based on multi-source heterogeneous flow, which overcomes the problems of existing analysis methods being unable to fully utilize the correlation of multi-source heterogeneous flow data, making it difficult to identify abnormal trends in indicators in advance, and failing to provide effective decision support for power plant operation and maintenance.

[0007] To achieve the above objectives, this invention provides a method for assessing and early warning analysis of new energy indicators based on multi-source heterogeneous flow, comprising: Several target assessment indicators are determined based on the inundation degree of indicator correlation; among them, the inundation degree of indicator correlation is determined based on the number of data sources for the assessment indicators and the historical fluctuation range of each data source. Based on the parameter mismatch characteristics and asynchronous data deviation characteristics of the data sources corresponding to the target assessment indicators, an abnormal tendency indicator set is determined; among them, the parameter mismatch characteristics are determined based on the response relationship between the measured values ​​and predicted values ​​of each data source, and the asynchronous data deviation characteristics are determined based on the change direction of each data source and the related data group from the same source. The scope of the cascading impact is determined based on the shared data source among the abnormal tendency indicators within the abnormal tendency indicator set. The extent of abnormal diffusion is used to determine whether there is a risk of mass loss of control within the scope of the chain reaction; whereby the extent of abnormal diffusion is determined based on the number of shared data sources and the number of abnormal tendency indicators within the scope of the chain reaction. In response to the potential for large-scale loss of control, early warnings will be sent to abnormal indicators within the scope of the chain reaction.

[0008] Furthermore, the process of determining several target performance indicators includes, If the inundation level of an assessment indicator exceeds the baseline inundation level, then the assessment indicator will be designated as the target assessment indicator.

[0009] Furthermore, the process of determining the inundation level associated with the indicators includes, Determine the number of data sources for the performance evaluation indicators; Determine the highest and lowest values ​​of each data source within the historical period, and determine the average of the differences between the highest and lowest values ​​of each data source as the historical fluctuation range of the data source. The product of the number of data sources for the assessment indicator and the historical fluctuation range is determined as the indicator correlation inundation degree of the assessment indicator.

[0010] Furthermore, the process of determining the set of anomalous tendency indicators includes, If the parameter mismatch characteristics of the target assessment indicator exceed the benchmark parameter mismatch degree, or the asynchronous deviation characteristics of the data exceed the benchmark deviation degree, then the target assessment indicator will be identified as an abnormal tendency indicator and included in the abnormal tendency indicator set.

[0011] Furthermore, the process of determining the parameter mismatch characteristics includes, The data sources are determined to be the measured and predicted values ​​at each sampling time. The sampling moment when the deviation between the measured value and the predicted value exceeds the benchmark deviation value is recorded as the mismatch moment. The ratio of the number of mismatch moments to the total number of sampling moments is determined as the parameter mismatch ratio of the data source; The maximum value of the parameter mismatch ratio in each data source corresponding to the target assessment indicator is determined as the parameter mismatch feature.

[0012] Furthermore, the process of determining the asynchronous deviation characteristics of the data includes, The ratio of the number of data sources in the same source associated data group whose change direction is different from the main change direction to the total number of data sources in the same source associated data group is determined as the asynchronous deviation ratio; The sampling moment when the asynchronous deviation ratio exceeds the baseline deviation value is recorded as the anomaly moment; The ratio of the number of aberration moments to the total number of sampling moments is determined as the asynchronous deviation characteristic of the data; Data sources of the same data type constitute a homogeneous related data group.

[0013] Furthermore, the process of determining the main direction of change includes, Determine the number of data sources with positive change direction and the number of data sources with negative change direction in the same source-related data group at each sampling time; The direction of change with the highest number is determined as the main direction of change at the sampling time.

[0014] Furthermore, the process of determining the scope of the cascading impact includes, Indicators of abnormal tendencies that share a data source are grouped into the same cascading effect range.

[0015] Furthermore, the process of determining whether there is a tendency for a large number of items to spiral out of control includes, If the abnormal diffusion exceeds the baseline abnormal diffusion, it is determined that there is a tendency for the chain reaction to go out of control in batches.

[0016] Furthermore, the process of determining the anomalous diffusion degree includes, The ratio of the number of abnormal tendency indicators within the scope of the chain reaction to the number of shared data sources is determined as the abnormal diffusion degree.

[0017] This invention relates to the field of performance evaluation and early warning technology, and in particular to a method for early warning analysis of new energy performance indicators based on multi-source heterogeneous flow. This invention filters out target performance indicators from multiple performance indicators by using the index correlation flooding degree, identifying those whose anomalies are easily masked by normal fluctuations in data sources. Then, based on parameter mismatch characteristics and data asynchronous deviation characteristics, it determines a set of anomaly tendency indicators for the target performance indicators. Furthermore, it delineates the cascading impact range based on the shared data source correlation between anomaly indicator items, and determines whether there is a risk of batch loss of control within the cascading impact range through anomaly diffusion degree. This allows for early warning pushes at the initial stage of shared data source anomalies, eliminating the blind spot in the identification of early anomaly signals from data sources in conventional compliance judgments, improving the timeliness of new energy performance indicator early warning and the safety of power plant operation.

[0018] In particular, this invention determines several target assessment indicators based on the inundation degree of indicator correlation. In the actual operation of the new energy indicator assessment and early warning system, the calculation of each assessment indicator relies on multi-source heterogeneous data such as remote sensing, AGC, AVC, and power prediction, and each data source has a normal fluctuation range of varying degrees. When an early abnormal signal appears in a shared data source, because the signal strength is not yet sufficient to make the final assessment indicator break through the target line, the ex-post judgment based on the comparison with the target line cannot trigger an early warning. The abnormal signal continues to accumulate at the data source level without being detected. By the time the indicator value finally exposes the problem, the abnormality at the data source level has lasted for a long time, and the best intervention window has been missed. In this process, the more data sources associated with the assessment indicator, the lower the contribution of the abnormal signal of a single data source to the final indicator value; the greater the historical fluctuation amplitude of each data source, the easier it is for early abnormal signals to be misjudged as normal fluctuations and ignored. Both of these factors lead to a longer accumulation time required for the abnormal signal to push the indicator value to break through the target line, increasing the risk of delayed problem identification. This application constructs the index correlation submergence degree by multiplying the number of data sources by the historical fluctuation amplitude, quantifies the risk of delayed identification of abnormal signals of each assessment indicator, and selects the assessment indicators whose early abnormal signals are most easily ignored as target assessment indicators. This lays the foundation for subsequent assessments based on parameter mismatch characteristics and asynchronous data deviation characteristics, thereby improving the targeting of early identification of abnormal signals and the ability of the new energy indicator assessment and early warning system to prevent and control potential risks.

[0019] In particular, this invention determines the set of abnormal tendency indicators based on the parameter mismatch characteristics and asynchronous deviation characteristics of the data source corresponding to the target assessment indicators, thereby eliminating the blind spot of conventional benchmark judgment in identifying early abnormal signals of the data source. In the actual operation of the new energy indicator assessment and early warning system, the calculation of each assessment indicator relies on multi-source heterogeneous data. When an early abnormal signal appears in a shared data source, conventional methods can only detect the problem after the final indicator value exceeds the target. To address this, this application directly captures abnormal signals at the data source level: the response deviation between the measured value and the predicted value of a normal data source should be within the normal range. However, when a data source is abnormal, this response relationship is disrupted, and the measured value frequently deviates from the predicted value. Based on this, the degree of continuous deviation of the measured value from the predicted value is quantified as the parameter mismatch ratio, and the maximum value of the parameter mismatch ratio in each data source corresponding to the target assessment indicator is taken as the parameter mismatch feature of that indicator. At this time, any decrease in accuracy in any data source can be captured, avoiding the dilution of single-point anomalies by other normal data sources due to averaging and missed detection. At the same time, in the same source-related data group, each data source is affected by the same physical and meteorological factors, and its change direction shows consistent synchronicity at most sampling times. However, when a data source is abnormal, its change direction will frequently deviate from the collective change trend of most data sources in the group, forming an asynchronous deviation feature of the data. This application captures parameter mismatch anomalies at the physical response level and asynchronous deviation anomalies at the statistical population level from the target assessment indicators, accurately identifying abnormal tendency indicators. This provides a reliable identification basis for defining the scope of subsequent chain effects and determining the risk of batch runaway, thereby improving the accuracy and timeliness of early identification of abnormal signals.

[0020] In particular, this invention determines the scope of cascading impact based on the shared data source relationships among the abnormal tendency indicators within an abnormal tendency indicator set. In the actual operation of the new energy indicator assessment and early warning system, the calculation of each assessment indicator relies on multi-source heterogeneous data, and multiple assessment indicators often share the same data source. When an abnormality occurs in a shared data source, this abnormality not only affects a single assessment indicator but also propagates to other assessment indicators that rely on that data source through the physical bridge of the shared data source, forming a cascading impact. If multiple abnormal tendency indicators share a data source, the continuous deterioration of any data source simultaneously affects multiple related indicators through the sharing relationship, forming a cascading risk of batch loss of control. This application determines the cascading relationship based on whether there is a shared data source among the abnormal tendency indicators, classifying each abnormal tendency indicator with a shared data source relationship into the same cascading impact range. This aggregates isolated indicator anomalies into a cascading risk assessment area, eliminating the blind spot of existing technologies that cannot perceive the cascading impact caused by shared data sources due to isolated analysis of each assessment indicator. This provides a precise analytical scope for subsequent batch loss of control risk tendency determination based on the degree of abnormality diffusion.

[0021] In particular, this invention determines whether there is a risk of batch loss of control within the scope of a chain reaction based on the degree of anomaly diffusion. This application introduces anomaly diffusion, quantified as the ratio of the number of anomalous tendency indicators within the scope of a chain reaction to the number of shared data sources. This ratio measures the concentration of the risk transmission path; a higher anomaly diffusion indicates more anomalous tendency indicators supported by a single shared data source, resulting in a more concentrated impact. If any shared data source continues to deteriorate, its impact can simultaneously spread to multiple anomalous tendency indicators, increasing the risk of batch loss of control. By quantifying the ratio of anomaly diffusion, this application elevates isolated indicator anomalies to a systemic risk assessment within the scope of a chain reaction. Before the physical process of batch indicator loss of control is triggered, it completes an advance judgment of the chain risks caused by shared data source anomalies, providing accurate decision-making basis for early warning pushes, thereby improving the ability of the new energy indicator assessment and early warning system to resist the risk of batch indicator loss of control. Attached Figure Description

[0022] Figure 1 This is a system block diagram of a new energy indicator assessment and early warning analysis system based on multi-source heterogeneous flow, as an embodiment of the invention. Figure 2 A schematic diagram illustrating the steps of the new energy indicator assessment and early warning analysis method based on multi-source heterogeneous flow, as an embodiment of the invention; Figure 3 A logical decision diagram for determining the abnormal tendency index set in an embodiment of the invention; Figure 4 This is a logic diagram for determining whether there is a risk of mass runaway in an embodiment of the invention. Detailed Implementation

[0023] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0024] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0025] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the term "connection" should be interpreted broadly. 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 an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0026] Please see Figure 1The diagram shown is a system block diagram of a new energy indicator assessment and early warning analysis system based on multi-source heterogeneous flow, according to an embodiment of the invention. The system of the present invention includes... The flooding index preliminary screening module is used to determine several target assessment indicators based on the correlation between the indicators and the flooding degree. The abnormal indicator screening module is used to determine the set of abnormal tendency indicators based on the parameter mismatch characteristics and asynchronous deviation characteristics of the data source corresponding to the target assessment indicators. The chain scope delineation module is used to determine the chain impact scope based on the shared data source among the abnormal tendency indicators within the abnormal tendency indicator set; The batch out-of-control judgment module is used to determine whether there is a risk of batch out of control within the scope of the chain reaction based on the degree of abnormal spread; The early warning information push module is used to push early warnings to abnormal tendency indicators within the scope of the chain reaction, in response to the risk of batch loss of control.

[0027] Please see Figure 2 The diagram illustrates the steps of a new energy indicator assessment and early warning analysis method based on multi-source heterogeneous flow, as described in an embodiment of the invention. The method of the present invention includes: Several target assessment indicators are determined based on the correlation between indicators and the degree of flooding. Based on the parameter mismatch characteristics and asynchronous deviation characteristics of the data sources corresponding to the target assessment indicators, a set of abnormal tendency indicators is determined. The scope of the cascading impact is determined based on the shared data source among the abnormal tendency indicators within the abnormal tendency indicator set. Determine whether there is a risk of mass loss of control in the scope of the chain reaction based on the degree of abnormal diffusion; In response to the potential for large-scale loss of control, early warnings will be sent to abnormal indicators within the scope of the chain reaction.

[0028] The process of determining several target assessment indicators in this embodiment includes, If the inundation level of an assessment indicator exceeds the baseline inundation level, then the assessment indicator will be designated as the target assessment indicator.

[0029] In this embodiment, the baseline flooding level is a reference value used to distinguish the degree of risk of delayed identification of abnormal signals of assessment indicators. As an example, those skilled in the art can calculate the average flooding level associated with each assessment indicator in the evaluated power plant and use this average value as the baseline flooding level. The specific value can be set according to the power plant's control requirements for the sensitivity of early warning of assessment indicators, which does not constitute a limitation of the present invention and will not be elaborated further here.

[0030] The process of determining the index-related submersion degree in this embodiment includes, Determine the number of data sources for the performance evaluation indicators; Determine the highest and lowest values ​​of each data source within the historical period, and determine the average of the differences between the highest and lowest values ​​of each data source as the historical fluctuation range of the data source. The product of the number of data sources for the assessment indicator and the historical fluctuation range is determined as the indicator correlation inundation degree of the assessment indicator.

[0031] In this embodiment, the highest and lowest values ​​of each data source within the historical period are obtained by retrieving the historical database of the new energy indicator assessment and early warning system. A complete time period that can cover different weather conditions and operating conditions is selected as the statistical window. The duration of the statistical window ensures that the historical fluctuation range can reflect the real fluctuation range of each data source under different scenarios. The specific period duration can be flexibly set according to the actual operating characteristics of the power plant and the data source type, which will not be elaborated here.

[0032] It is understood that the number of data sources for the assessment indicators can be directly read from the data source configuration information of the new energy indicator assessment and early warning system, or obtained by querying the relationship table between the data source and the assessment indicator. It is only necessary to obtain the total number of data sources that the assessment indicator depends on. The specific acquisition method does not constitute a limitation of the present invention, and will not be elaborated here.

[0033] This application determines target performance indicators by using correlation inundation. The more data sources an performance indicator is associated with, the lower the contribution of anomaly signals from a single data source to the final indicator value. Furthermore, the greater the historical fluctuation range of each data source, the easier it is for early anomaly signals to be misjudged as normal fluctuations and ignored. Both of these factors prolong the cumulative time required for anomaly signals to drive the indicator value to break through the target threshold, increasing the risk of delayed problem identification. By using correlation inundation to screen performance indicators whose early anomaly signals are most easily overlooked as target performance indicators, this application improves the targeting and efficiency of early anomaly signal identification, and enhances the ability of the new energy indicator performance assessment and early warning system to resist the risk of batch indicator runaway.

[0034] Please see Figure 3 As shown, it is a logic decision diagram for determining the set of abnormal tendency indicators in an embodiment of the invention. The process of determining the set of abnormal tendency indicators in this embodiment includes, If the parameter mismatch characteristics of the target assessment indicator exceed the benchmark parameter mismatch degree, or the asynchronous deviation characteristics of the data exceed the benchmark deviation degree, then the target assessment indicator will be identified as an abnormal tendency indicator and included in the abnormal tendency indicator set.

[0035] In this embodiment, the baseline parameter mismatch is a reference value used to determine whether the measured value of the data source continuously deviates from the predicted value. As an example, those skilled in the art can select the statistical distribution of the parameter mismatch ratio at each sampling time under long-term historical normal operating conditions of the data source, and use its mean as the baseline parameter mismatch. The specific value can be flexibly set according to the actual operating requirements of the power plant, which will not be elaborated here.

[0036] In this embodiment, the baseline deviation is a reference value used to determine whether the direction of change of the data source within a co-sourced data group deviates from the mainstream trend. As an example, those skilled in the art can select the statistical distribution of the asynchronous deviation ratio of the co-sourced data group at each sampling time under long-term historical normal operating conditions, and use its mean as the baseline deviation. The specific value can be flexibly set according to the actual operating requirements of the power plant, which will not be elaborated here.

[0037] In the actual operation of the new energy indicator assessment and early warning system, the calculation of each assessment indicator relies on multi-source heterogeneous data. When an early abnormal signal appears in a shared data source, conventional methods can only detect the problem after the final indicator value exceeds the target. To address this, this application directly captures abnormal signals at the data source level: the response deviation between the measured value and the predicted value of a normal data source should be within the normal range. However, when a data source is abnormal, this response relationship is disrupted, and the measured value frequently deviates from the predicted value. Based on this, the degree of continuous deviation of the measured value from the predicted value is quantified as the parameter mismatch ratio, and the maximum value of the parameter mismatch ratio in each data source corresponding to the target assessment indicator is taken as the parameter mismatch feature of that indicator. At this time, any decrease in accuracy in any data source can be captured, avoiding the dilution of single-point anomalies by other normal data sources due to averaging and missed detection. At the same time, in the same source-related data group, each data source is affected by the same physical and meteorological factors, and its change direction shows consistent synchronicity at most sampling times. However, when a data source is abnormal, its change direction will frequently deviate from the collective change trend of most data sources in the group, forming an asynchronous deviation feature of the data. This application captures parameter mismatch anomalies at the physical response level and asynchronous deviation anomalies at the statistical population level from the target assessment indicators, accurately identifying abnormal tendency indicators. This provides a reliable identification basis for defining the scope of subsequent chain effects and determining the risk of batch runaway, thereby improving the accuracy and timeliness of early identification of abnormal signals.

[0038] The process for determining parameter mismatch characteristics in this embodiment includes: The data sources are determined to be the measured and predicted values ​​at each sampling time. The sampling moment when the deviation between the measured value and the predicted value exceeds the benchmark deviation value is recorded as the mismatch moment. The ratio of the number of mismatch moments to the total number of sampling moments is determined as the parameter mismatch ratio of the data source; The maximum value of the parameter mismatch ratio in each data source corresponding to the target assessment indicator is determined as the parameter mismatch feature.

[0039] In this embodiment, each sampling time is a point in time within a continuous period segment intercepted from the current time, during which data corresponding to the target assessment indicators are collected at intervals. The interval between adjacent sampling times is determined by the inherent sampling frequency of the data acquisition device, and the total number of sampling times is the total number of sampling times contained within the continuous period. The specific duration of this continuous period can be set according to the data source type and the actual operating characteristics of the power plant, with the principle of sufficient accumulation of statistical data. This does not constitute a limitation of the present invention and will not be elaborated further.

[0040] In this embodiment, the benchmark deviation value is a reference value for determining whether the measured value of the data source deviates from the predicted value. As an example, those skilled in the art can select the statistical distribution of the deviation values ​​of the data source at each sampling time under long-term historical normal operating conditions, and determine the upper limit of the deviation fluctuation under normal conditions as the benchmark deviation value. The specific value of the benchmark deviation value can be set according to the actual operating requirements of the power plant, which does not constitute a limitation on the present invention, and will not be elaborated here.

[0041] It should be noted that the data sources corresponding to the assessment indicators may include multiple data types, and the physical meaning and normal fluctuation range of each data source are different. Therefore, each data source corresponds to its own benchmark deviation value.

[0042] It is understandable that the measured values ​​from each data source at each sampling time can be directly read from the data acquisition and monitoring control system of the new energy power plant, and the predicted values ​​can be obtained from the corresponding power prediction system or the prediction curve issued by the dispatch. It is only necessary to complete the synchronous acquisition of the measured values ​​and predicted values ​​of each data source within the preset statistical window, which will not be elaborated further.

[0043] The process for determining the asynchronous deviation characteristics of data in this embodiment includes, The ratio of the number of data sources in the same source associated data group whose change direction is different from the main change direction to the total number of data sources in the same source associated data group is determined as the asynchronous deviation ratio; The sampling moment when the asynchronous deviation ratio exceeds the baseline deviation value is recorded as the anomaly moment; The ratio of the number of aberration moments to the total number of sampling moments is determined as the asynchronous deviation characteristic of the data; Data sources of the same data type constitute a homogeneous related data group.

[0044] In this embodiment, the parameter mismatch characteristics at each sampling time are consistent with those in the previous embodiment, and will not be repeated here.

[0045] It is understandable that a group of data with the same source consists of data sources that belong to the same data type, and the data sources within the same group represent the same physical quantity.

[0046] In this embodiment, the direction of change of each data source is determined by comparing the value at the current sampling time with that at the previous sampling time: an increase in value indicates a positive direction, a decrease indicates a negative direction, and no change indicates a zero direction.

[0047] In this embodiment, the benchmark deviation value is a reference value for determining whether the direction of change between data sources within the same source-related data group has deviated. As an example, those skilled in the art can select the statistical distribution of the asynchronous deviation ratio of the same source-related data group at each sampling time under long-term historical normal operating conditions, and determine the mean of the asynchronous deviation ratio in the statistical distribution as the benchmark deviation value. Its specific value can be set according to the actual operating requirements of the power plant, which does not constitute a limitation on the present invention, and will not be elaborated here.

[0048] It should be noted that if the target assessment indicator corresponds to multiple related data groups from the same source, the asynchronous deviation characteristics of the data in each related data group from the same source should be determined, and the maximum value of the asynchronous deviation characteristics of the data in each related data group from the same source should be determined as the asynchronous deviation characteristics of the data in the target assessment indicator.

[0049] The process of determining the main direction of change in this embodiment includes, Determine the number of data sources with positive change direction and the number of data sources with negative change direction in the same source-related data group at each sampling time; The direction of change with the highest number is determined as the main direction of change at the sampling time.

[0050] In this embodiment, each sampling time is a time point within a preset statistical window where data is collected at intervals from each data source corresponding to the target assessment indicator. The change in value of each data source at the current sampling time relative to the previous sampling time is judged: an increase in value is positive, a decrease in value is negative, and no change in value is zero. When determining the main direction of change, only the number of positive and negative data sources is counted; zero indicates that the data source has not undergone substantial change and is not included in the direction statistics. The direction of change with the highest number is determined as the main direction of change at that sampling time.

[0051] It should be noted that if the number of positive and negative directions is the same, it is determined that there is no dominant change direction at that sampling moment, and that sampling moment is directly recorded as the moment of change.

[0052] The process of determining the scope of the cascading effect in this embodiment includes, Indicators of abnormal tendencies that share a data source are grouped into the same cascading effect range.

[0053] In this embodiment, the shared data source is a data source that is used by multiple abnormal tendency indicators in the calculation.

[0054] Please see Figure 4As shown, this is a logic diagram for determining whether there is a risk of mass runaway from control according to an embodiment of the invention. The process for determining whether there is a risk of mass runaway from control in this embodiment includes: If the abnormal diffusion exceeds the baseline abnormal diffusion, it is determined that there is a tendency for the chain reaction to go out of control in batches.

[0055] In this embodiment, the baseline anomaly diffusion degree is a reference value for determining whether there is a tendency for mass loss of control within the scope of the cascading impact. As an example, those skilled in the art can use the value corresponding to the critical state when the number of anomaly tendency indicators within the scope of the cascading impact is equal to the number of shared data sources as the baseline anomaly diffusion degree. Exceeding the baseline anomaly diffusion degree indicates that the risk transmission path has formed an intertwined network, and there is a tendency for mass loss of control. The specific value of the baseline anomaly diffusion degree can be flexibly set according to the actual operating requirements of the power plant, and does not constitute a limitation of the present invention, and will not be elaborated further here.

[0056] The process for determining abnormal diffusion in this embodiment includes, The ratio of the number of abnormal tendency indicators within the scope of the chain reaction to the number of shared data sources is determined as the abnormal diffusion degree.

[0057] Understandably, when it is determined that there is a risk of mass loss of control in the scope of the chain reaction, the system can automatically generate an early warning message containing the name of the abnormal tendency indicator, the scope of the chain reaction, and the risk type. This message will be sent to the operation and maintenance personnel's terminals via in-site messages, SMS, or APP push notifications. It is sufficient to remind them to check the relevant data sources in a timely manner, which will not be elaborated further.

[0058] In particular, this invention determines whether there is a risk of batch loss of control within the scope of a chain reaction based on the degree of anomaly diffusion. This application introduces anomaly diffusion, quantified as the ratio of the number of anomalous tendency indicators within the scope of a chain reaction to the number of shared data sources. This ratio measures the concentration of the risk transmission path; a higher anomaly diffusion indicates more anomalous tendency indicators supported by a single shared data source, resulting in a more concentrated impact. If any shared data source continues to deteriorate, its impact can simultaneously spread to multiple anomalous tendency indicators, increasing the risk of batch loss of control. By quantifying the ratio of anomaly diffusion, this application elevates isolated indicator anomalies to a systemic risk assessment within the scope of a chain reaction. Before the physical process of batch indicator loss of control is triggered, it completes an advance judgment of the chain risks caused by shared data source anomalies, providing accurate decision-making basis for early warning pushes, thereby improving the ability of the new energy indicator assessment and early warning system to resist the risk of batch indicator loss of control.

[0059] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles 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 all fall within the scope of protection of the present invention.

Claims

1. A method for assessing and early warning analysis of new energy indicators based on multi-source heterogeneous flow, characterized in that, include, Several target assessment indicators are determined based on the inundation degree of indicator correlation; among them, the inundation degree of indicator correlation is determined based on the number of data sources for the assessment indicators and the historical fluctuation range of each data source. Based on the parameter mismatch characteristics and asynchronous data deviation characteristics of the data sources corresponding to the target assessment indicators, an abnormal tendency indicator set is determined; among them, the parameter mismatch characteristics are determined based on the response relationship between the measured values ​​and predicted values ​​of each data source, and the asynchronous data deviation characteristics are determined based on the change direction of each data source and the related data group from the same source. The scope of the cascading impact is determined based on the shared data source among the abnormal tendency indicators within the abnormal tendency indicator set. The extent of abnormal diffusion is used to determine whether there is a risk of mass loss of control within the scope of the chain reaction; whereby the extent of abnormal diffusion is determined based on the number of shared data sources and the number of abnormal tendency indicators within the scope of the chain reaction. In response to the potential for large-scale loss of control, early warnings will be sent to abnormal indicators within the scope of the chain reaction.

2. The method for assessing and early warning analysis of new energy indicators based on multi-source heterogeneous flow as described in claim 1, characterized in that, The process of determining several target performance indicators includes: If the inundation level of an assessment indicator exceeds the baseline inundation level, then the assessment indicator will be designated as the target assessment indicator.

3. The method for assessing and early warning analysis of new energy indicators based on multi-source heterogeneous flow as described in claim 2, characterized in that, The process of determining the correlation between the index and the flooding degree includes, Determine the number of data sources for the performance evaluation indicators; Determine the highest and lowest values ​​of each data source within the historical period, and determine the average of the differences between the highest and lowest values ​​of each data source as the historical fluctuation range of the data source. The product of the number of data sources for the assessment indicator and the historical fluctuation range is determined as the indicator correlation inundation degree of the assessment indicator.

4. The method for assessing and early warning analysis of new energy indicators based on multi-source heterogeneous flow as described in claim 1, characterized in that, The process of determining the abnormal tendency index set includes, If the parameter mismatch characteristics of the target assessment indicator exceed the benchmark parameter mismatch degree, or the asynchronous deviation characteristics of the data exceed the benchmark deviation degree, then the target assessment indicator will be identified as an abnormal tendency indicator and included in the abnormal tendency indicator set.

5. The method for assessing and early warning analysis of new energy indicators based on multi-source heterogeneous flow as described in claim 4, characterized in that, The process of determining parameter mismatch characteristics includes, The data sources are determined to be the measured and predicted values ​​at each sampling time. The sampling moment when the deviation between the measured value and the predicted value exceeds the benchmark deviation value is recorded as the mismatch moment. The ratio of the number of mismatch moments to the total number of sampling moments is determined as the parameter mismatch ratio of the data source; The maximum value of the parameter mismatch ratio in each data source corresponding to the target assessment indicator is determined as the parameter mismatch feature.

6. The method for assessing and early warning analysis of new energy indicators based on multi-source heterogeneous flow as described in claim 4, characterized in that, The process of determining the asynchronous deviation characteristics of the data includes: The ratio of the number of data sources in the same source associated data group whose change direction is different from the main change direction to the total number of data sources in the same source associated data group is determined as the asynchronous deviation ratio; The sampling moment when the asynchronous deviation ratio exceeds the baseline deviation value is recorded as the anomaly moment; The ratio of the number of aberration moments to the total number of sampling moments is determined as the asynchronous deviation characteristic of the data; Data sources of the same data type constitute a homogeneous related data group.

7. The method for assessing and early warning analysis of new energy indicators based on multi-source heterogeneous flow as described in claim 6, characterized in that, The process of determining the main direction of change includes, Determine the number of data sources with positive change direction and the number of data sources with negative change direction in the same source-related data group at each sampling time; The direction of change with the highest number is determined as the main direction of change at the sampling time.

8. The method for assessing and early warning analysis of new energy indicators based on multi-source heterogeneous flow as described in claim 1, characterized in that, The process of determining the scope of the chain reaction includes, Indicators of abnormal tendencies that share a data source are grouped into the same cascading effect range.

9. The method for assessing and early warning analysis of new energy indicators based on multi-source heterogeneous flow as described in claim 1, characterized in that, The process of determining whether there is a risk of mass loss of control includes: If the abnormal diffusion exceeds the baseline abnormal diffusion, it is determined that there is a tendency for the chain reaction to go out of control in batches.

10. The method for assessing and early warning analysis of new energy indicators based on multi-source heterogeneous flow as described in claim 9, characterized in that, The process of determining the abnormal diffusion degree includes, The ratio of the number of abnormal tendency indicators within the scope of the chain reaction to the number of shared data sources is determined as the abnormal diffusion degree.

Images