A power grid meteorological data monitoring and evaluation method and system and storage medium

Abstract

The application discloses a kind of power grid meteorological data monitoring evaluation method, system and storage medium, it is related to power meteorological monitoring technical field, it solves the technical problem that the influence degree of the power loss of related equipment when operating and equipment state by meteorological anomaly is not considered in prior art, so that the efficiency of power generation is relatively low, leading to the imbalance of supply and demand of electric energy;Meteorological anomaly coefficient is generated according to environmental parameters and weather forecast;Equipment state score is generated according to power plant data and environmental data;Available power is generated according to equipment parameters and equipment state score and an adjustment scheme is generated accordingly;Alarm signal is generated according to meteorological anomaly coefficient and equipment state score, the monitoring results of multiple weather are quantified and future meteorological anomaly coefficient is calculated, the influence of weather on future power generation is considered, and the power supply of region is dynamically adjusted according to the demand data in supply area, improve the balance of supply and demand of electric energy and the accuracy and comprehensiveness of meteorological monitoring.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power weather monitoring, and specifically relates to a power grid weather data monitoring and evaluation method and system and a storage medium. BACKGROUND

[0002] Power grid weather monitoring is an important means to ensure the safe and stable operation of the power grid. It helps the power department to take preventive measures in time through real-time monitoring and early warning of the weather environment around the power grid equipment, reducing power grid failures and power outage accidents caused by weather disasters. Power grid equipment is exposed to complex and changeable weather environment for a long time, and is easily affected by weather disasters such as strong winds. These disasters not only may cause damage to the power grid equipment, but also may cause large-area power outage accidents, which will have a serious impact on social production and people's life.

[0003] The prior art (invention patent application with publication number CN117609802A) discloses a power grid weather monitoring data evaluation method, which comprises the following steps: A, first, collect the weather data of the power grid area, and pre-process the data; B, then, transmit the collected monitoring data to the processor module; C, the processor module transmits the weather monitoring data to the comparison module to compare with the weather monitoring data of the same period in the historical period, and outputs the comparison result; D, output the comparison result to the weather data prediction model to obtain the prediction data of the current period of the target to be evaluated; E, transmit the prediction data of the current period to the evaluation module for evaluation, and output the evaluation result.

[0004] The above-mentioned patent compares the data collected by multiple sensors with historical data to obtain the current weather abnormality; it lacks consideration of the influence degree of weather abnormalities on the power consumption of related equipment during operation and the state of the equipment, resulting in low efficiency of power generation and imbalance between supply and demand of electric energy; for example, frequent changes in wind direction make the wind blades rotate, affecting power generation and causing equipment wear and tear to intensify; therefore, the power grid weather data monitoring and evaluation system still needs to be further improved. SUMMARY

[0005] The present application aims to at least solve one of the technical problems existing in the prior art; for this purpose, the present application proposes a power grid weather data monitoring and evaluation method, system and storage medium, which is used to solve the technical problem that the prior art lacks consideration of the influence degree of weather abnormalities on the power consumption of related equipment during operation and the state of the equipment, resulting in low efficiency of power generation and imbalance between supply and demand of electric energy.

[0006] To achieve the above object, the first aspect of the present application provides a power grid meteorological data monitoring and evaluation system, comprising a data acquisition module, a data analysis module, a warning module and a database; the data acquisition module and the data analysis module are connected; the data analysis module and the warning module are connected; the data acquisition module, the data analysis module and the warning module are respectively connected with the database;

[0007] The data acquisition module: obtains power plant data and environmental data and demand data through a data acquisition device; the environmental data includes environmental parameters and weather forecasts; the power plant data includes power generation equipment ID and equipment parameters;

[0008] The data analysis module: generates a meteorological anomaly coefficient according to the environmental parameters and the weather forecast; generates an equipment state score according to the power plant data and the environmental data; obtains equipment parameters in real time, generates a power supply capacity according to the equipment parameters and the equipment state score; generates an adjustment scheme according to the power supply capacity and the demand data; generates an alarm signal according to the meteorological anomaly coefficient and the equipment state score;

[0009] The warning module: makes a prompt according to the alarm signal and contacts the management personnel;

[0010] The database is used to store the data of each module and store the historical data required for training the model.

[0011] Through the above steps, the monitoring data of various meteorological elements can be quantitatively processed, and the future meteorological anomaly coefficient can be calculated accordingly; in this process, the potential impact of meteorological conditions on future power generation is fully considered, and the actual electricity demand data in the supply area is combined to dynamically optimize the regional power supply scheme, which not only helps to improve the balance level of power supply and demand, but also effectively improves the accuracy and comprehensiveness of meteorological monitoring.

[0012] Further, the meteorological anomaly coefficient is generated according to the environmental parameters and the weather forecast, comprising:

[0013] Obtain the environmental parameters, equipment parameters and weather forecast in a time period T; the environmental parameters include wind speed, wind direction, temperature and humidity; the equipment parameters include equipment rotation frequency and equipment rotation amplitude; the weather forecast includes wind speed, temperature and humidity; the environmental parameters and weather forecast are represented as environmental data corresponding to different time periods;

[0014] Obtain historical environmental parameters, historical equipment parameters and their corresponding historical weather forecasts and prediction labels corresponding to a plurality of historical time periods;

[0015] Integrate the historical environmental parameters, historical equipment parameters and their corresponding historical weather forecasts corresponding to a plurality of historical time periods into a parameter prediction sequence;

[0016] inputting the predicted label and the parameter prediction sequence into a multi-modal prediction model to obtain predicted device parameters corresponding to a plurality of time periods; the multi-modal prediction model is constructed by an artificial intelligence model;

[0017] generate a weather anomaly coefficient according to the environmental data and the device parameters in the time period.

[0018] Further, the weather anomaly coefficient is generated according to the environmental data and the device parameters in the time period, comprising:

[0019] obtain the environmental data and the device parameters in the time period; the environmental data includes environmental parameters and weather forecasts; the device parameters include device parameters in the current time period and predicted device parameters in the future time period;

[0020] extract the wind speed FV, temperature WD and humidity SD in the environmental parameters and weather forecasts;

[0021] extract the device rotation frequency SZC and device rotation amplitude SZF in the device parameters;

[0022] construct an environmental impact function HYF(FV, WD, SD) through the nonlinear relationship between the wind speed FV, temperature WD and humidity SD and the weather anomaly coefficient; construct a device impact function SYF(SZC, SZF) through the nonlinear relationship between the device rotation frequency SZC and the device rotation amplitude SZF and the weather anomaly coefficient;

[0023] linearly weight and fuse the environmental impact function and the device impact function, and substitute a plurality of parameters in the time period into the calculation to obtain the weather anomaly coefficient corresponding to the time period.

[0024] Further, the device state score is generated according to the power plant data and the environmental data, comprising:

[0025] obtain the power plant data and the environmental data corresponding to a plurality of time periods; the power plant data includes power generation device ID and device parameters; the device parameters include start-stop frequency QT, rotation speed ZV and running time YS; the environmental data includes temperature WD and humidity SD;

[0026] construct a period score function ZPF(QT, ZV, YS, WD, SD) through the nonlinear relationship between the device parameters and the environmental data and the period device score;

[0027] input the corresponding device parameters and environmental data in the time period into the period score function to calculate the period device score;

[0028] divide a plurality of time periods according to the period device score to obtain a plurality of period clusters;

[0029] According to the formula The cluster period device score CZSP corresponding to each period cluster is calculated; wherein i represents the number of time periods in the period cluster; represents the synergy factor, ∈(0, 1), representing the amplification effect of the previous wear on the subsequent time period;

[0030] The device state score corresponding to the current monitoring time point is obtained by summing the cluster period device scores of several cluster periods before the current monitoring time point.

[0031] The present application first calculates the device state score for each time period, and then obtains the period device score; then, according to the influence degree of the period device score on the device state, cluster analysis is carried out on multiple time periods; after clustering, the cluster period device score of different time clusters is recalculated, and these scores are cumulatively summed, thereby obtaining the device state score corresponding to the current monitoring time point; considering that the device wear condition is more serious in the frequently changing time period, and the wear presents the characteristics of nonlinear accumulation, the scoring method adopted by the present application can make the device state score more in line with the actual situation, and significantly improve the accuracy of the device state score; and the accurate device state score also provides solid and reliable data support for subsequent power generation prediction.

[0032] Further, the division of the plurality of time periods according to the period device score to obtain a plurality of period clusters comprises the following steps:

[0033] Step one: obtaining a plurality of historical period device scores ZSP j and a plurality of time periods SZ j corresponding thereto;

[0034] Step two: creating a new period cluster list;

[0035] Step three: determining whether ZSP j is greater than the score threshold;

[0036] Yes, ZSP j is placed in the period cluster list, and the next time period is entered; step three is entered;

[0037] No, determine whether the period cluster list is an empty list; yes, ZSP j is placed in the period cluster list, and the next time period is entered; step two is entered; no, a new period cluster list is created, ZSP j is placed in the period cluster list, and the next time period is entered, step two is entered; wherein j represents the number corresponding to the time period.

[0038] Further, the generating the available power according to the device parameters and the device state score comprises:

[0039] obtaining a plurality of historical device parameters, historical device state scores, historical environment data, meteorological anomaly coefficients and prediction labels, and integrating the plurality of historical device parameters, historical device state scores, historical environment data and meteorological anomaly coefficients into a power prediction sequence;

[0040] inputting the prediction labels and the power prediction sequence into a multi-modal prediction model to obtain a predicted power generation; the multi-modal prediction model is constructed by an artificial intelligence model;

[0041] extracting the energy storage power in the power plant data in real time;

[0042] obtaining the available power by summing the energy storage power and the predicted power generation.

[0043] Further, the multi-modal prediction model is constructed by an artificial intelligence model, comprising:

[0044] obtaining historical prediction labels and corresponding historical parameter prediction sequences or historical power prediction sequences, and corresponding historical device parameters or historical power generation; the prediction labels include parameter prediction labels and power prediction labels;

[0045] dividing the plurality of historical parameter prediction sequences and historical device parameters, and the plurality of historical power prediction sequences and historical power generation into training data, validation data and test data corresponding to the prediction labels according to the prediction labels;

[0046] dividing the plurality of historical parameter prediction sequences and historical device parameters, and the plurality of historical power prediction sequences and historical power generation into training data, validation data and test data corresponding to the prediction labels according to the prediction labels; and performing data preprocessing on the training data, validation data and test data to obtain a training set, a validation set and a test set;

[0047] selecting two machine learning models as the basis models of the multi-task branch;

[0048] training the corresponding basis models through the respective training sets, and adjusting the learning rate and other hyperparameters on the respective validation sets to obtain respective pre-trained models;

[0049] verifying the respective pre-trained models on the respective test sets to finally obtain a multi-modal prediction model that inputs the prediction labels and the corresponding parameter prediction sequences or power prediction sequences, and outputs the predicted device parameters or predicted power generation.

[0050] Further, the generating the adjustment scheme according to the available power and the demand data comprises:

[0051] obtaining supply power and demand data; the demand data includes total demand power and regional demand power and regional parameters corresponding to a plurality of regions;

[0052] determining whether the supply power is greater than the total demand power; if yes, no operation is performed; if no, an electricity shortage early warning signal is generated, the maximum purchase power is obtained, the maximum supply power is obtained by summing the maximum purchase power and the supply power, it is determined whether the maximum supply power is greater than the total demand power; if yes, no operation is performed; if no, a regional power shortage alarm signal is generated; and the regional distribution power of the plurality of regions is generated according to the maximum supply power and the regional parameters; and the plurality of regions are powered according to the regional distribution power.

[0053] Further, the regional distribution power of the plurality of regions is generated according to the maximum supply power and the regional parameters, including:

[0054] obtaining the maximum supply power and the regional parameters; the regional parameters include building grades, population densities and special population quantities; the special population includes the elderly, the sick and the disabled;

[0055] extracting the required power of the buildings in the plurality of regions whose building grades are higher than a grade threshold;

[0056] obtaining the population densities RM and the special population quantities TS corresponding to the plurality of regions;

[0057] constructing a priority function YF(RM, TS) through a nonlinear relationship between the population densities and the special population quantities and the regional priorities; wherein the priority function is an increasing function;

[0058] inputting the population densities and the special population quantities of the plurality of regions into the priority function to calculate the regional priorities;

[0059] calculating the distributable power by difference between the maximum supply power and the sum of the required power of the plurality of buildings;

[0060] determining whether the distributable power is greater than 0; if yes, the regional priorities of the plurality of regions are normalized to obtain normalized regional priorities, and the distributable power of the plurality of regions is calculated by multiplying the distributable power and the normalized regional priorities; if no, a power failure alarm signal is generated;

[0061] calculating the regional distribution power by summing the distributable power of the region and the required power of the buildings in the region.

[0062] Further, the alarm signal is generated according to the meteorological anomaly coefficient and the equipment state score, including:

[0063] obtaining the meteorological anomaly coefficient and the equipment state score;

[0064] determining whether the meteorological anomaly coefficient is greater than a meteorological anomaly threshold value; yes, generating a meteorological warning signal; no, doing nothing;

[0065] determining whether the device state score is greater than a device retirement threshold value; yes, generating a device retirement warning signal; no, determining whether the device state score is greater than D times the device retirement threshold value; yes, generating a device maintenance early warning signal; no, doing nothing; wherein D represents a proportional coefficient, D∈(0, 1).

[0066] The present application monitors the meteorological anomaly coefficient and the device state score in real time, and once an abnormal state occurs, a warning signal is sent in time and corresponding measures are taken, so as to avoid time and resource waste caused by poor device state and improve the efficiency and comprehensiveness of the power grid meteorological data monitoring and evaluation system.

[0067] The second aspect of the present application provides a power grid meteorological data monitoring and evaluation method, comprising:

[0068] S0: obtaining power plant data, environmental data and demand data; the environmental data includes environmental parameters and weather forecast; the power plant data includes power generation equipment ID and device parameters;

[0069] S1: generating a meteorological anomaly coefficient according to the environmental parameters and the weather forecast; generating a device state score according to the power plant data and the environmental data;

[0070] S2: obtaining device parameters in real time, and generating available power according to the device parameters and the device state score;

[0071] S3: generating an adjustment scheme according to the available power and the demand data; generating a warning signal according to the meteorological anomaly coefficient and the device state score;

[0072] S4: making a prompt according to the warning signal, and contacting a management personnel.

[0073] Another aspect of the present application provides a computer readable storage medium, the computer readable storage medium stores a computer program, the computer program is executed by a processor to realize a power grid meteorological data monitoring and evaluation system according to the first aspect of the present application.

[0074] Compared with the prior art, the present application has the following beneficial effects:

[0075] 1. This application generates meteorological anomaly coefficients based on environmental parameters and weather forecasts; generates equipment status scores based on power plant data and environmental data; acquires equipment parameters in real time and generates available power based on equipment parameters and equipment status scores; generates adjustment plans based on available power and demand data; generates alarm signals based on meteorological anomaly coefficients and equipment status scores; quantifies the monitoring results of various meteorological phenomena and calculates future meteorological anomaly coefficients; considers the impact of meteorology on future power generation; and dynamically adjusts the power supply in the region based on demand data within the supply area, thereby improving the balance of power supply and demand and the accuracy and comprehensiveness of meteorological monitoring.

[0076] 2. This application uses the number of equipment rotations and the rotation amplitude of the equipment to represent the impact data of wind blades under frequent changes in wind direction. It also uses a pre-trained multimodal prediction model to predict the number of equipment rotations and the rotation amplitude of the equipment in future time periods. Furthermore, it combines weather forecasts to quantify the meteorological anomaly coefficient for future time periods, thereby improving the accuracy of meteorological anomaly calculations and meteorological forecasts.

[0077] 3. This application calculates the periodic equipment score by evaluating the equipment status within each time period. Then, based on the degree of influence of the periodic equipment score on the equipment status, several time periods are clustered, and the cluster periodic equipment scores for different time clusters are recalculated. The scores are then accumulated and summed to obtain the equipment status score corresponding to the current monitoring time point. Considering that severe wear within frequent time periods can lead to nonlinear accumulation of wear, the equipment status score is made more consistent with the actual situation, improving the accuracy of the equipment status score and providing strong data support for subsequent power generation prediction.

[0078] 4. This application first compares the available power supply of the power plant with the demand data. If the supply of the demand data can be met, there is no need to purchase electricity. Otherwise, if electricity purchase fails to achieve full coverage of all areas, priority will be given to supplying electricity to important buildings based on regional priority and special buildings in the region. Subsequently, the available power supply in each area will be dynamically allocated and corresponding alarm signals will be issued to take preventive measures in advance and reduce the occurrence of poor user experience due to sudden imbalances in power supply. Attached Figure Description

[0079] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0080] Figure 1 This is a schematic diagram illustrating the principle of a power grid meteorological data monitoring and evaluation system according to this application;

[0081] Figure 2 A flow chart is generated for the cycle cluster generation process of the present application;

[0082] Figure 3 A flow chart is generated for the adjustment scheme of the present application;

[0083] Figure 4 A flow chart is generated for the power grid meteorological data monitoring and evaluation method of the present application. DETAILED DESCRIPTION

[0084] The technical solutions of the present application will be described in detail below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0085] Please refer to Figure 1 The first aspect of the present application provides a power grid meteorological data monitoring and evaluation system, comprising: a data acquisition module, a data analysis module, a warning module and a database; the data acquisition module and the data analysis module are connected, the data analysis module and the warning module are connected, and the data acquisition module, the data analysis module and the warning module are respectively connected with the database;

[0086] The data acquisition module: obtains power plant data, environmental data and demand data through a data acquisition device; the environmental data includes environmental parameters and weather forecast; the power plant data includes power generation equipment ID and equipment parameters; the data acquisition device includes a plurality of sensors, etc.

[0087] The data analysis module: generates a meteorological anomaly coefficient according to the environmental parameters and the weather forecast, the meteorological anomaly coefficient being the degree of meteorological anomaly; generates an equipment state score according to the power plant data and the environmental data, the equipment state score being the score of the power generation equipment state; obtains the equipment parameters in real time, generates a power supply according to the equipment parameters and the equipment state score, the power supply being the power that the power plant can supply at different times; generates an adjustment scheme according to the power supply and the demand data, the adjustment scheme being the scheme for adjusting the power supply and the demand data; generates a warning signal according to the meteorological anomaly coefficient and the equipment state score;

[0088] The warning module: makes a prompt according to the warning signal and contacts the management personnel; the warning signal includes a meteorological warning signal, an equipment maintenance warning signal and a power shortage warning signal, etc.

[0089] The database is used to store the data of each module and store the historical data required for training the model.

[0090] The embodiment is used for monitoring meteorological abnormalities in a wind power plant. In the existing wind power plant, the wind blades are not fixed in position and do not change, but change according to the change of the wind direction. When the wind direction frequently changes, the wind blades will rotate to achieve the best power generation effect. The rotation times and rotation amplitudes of the wind blades can represent the change of the wind direction in a time period. The quantification of the abnormal conditions of the wind direction according to the rotation times and rotation amplitudes can make the meteorological monitoring more accurate.

[0091] In the embodiment, the meteorological abnormality coefficient is generated according to the environmental parameters and the weather forecast, including:

[0092] The environmental parameters, the device parameters and the weather forecast in a time period T are obtained. The time period T is set according to experience, which can be set to 1 hour or 1 day. In the embodiment, T is set to 3 hours. The environmental parameters include wind speed, wind direction, temperature and humidity. The device parameters include device rotation times and device rotation amplitudes. The weather forecast includes wind speed, temperature and humidity. The environmental parameters and the weather forecast are represented as environmental data corresponding to different time periods.

[0093] The historical environmental parameters, the historical device parameters, the corresponding historical weather forecast and the prediction labels of a plurality of historical time periods are obtained. The prediction labels include parameter prediction labels and power prediction labels.

[0094] The historical environmental parameters, the historical device parameters and the corresponding historical weather forecast of a plurality of historical time periods are integrated into a parameter prediction sequence.

[0095] The prediction labels and the parameter prediction sequence are input into a multi-modal prediction model to obtain prediction device parameters corresponding to a plurality of time periods. The multi-modal prediction model is constructed by an artificial intelligence model.

[0096] The meteorological abnormality coefficient is generated according to the environmental data and the device parameters in a time period.

[0097] The embodiment uses the device rotation times and the device rotation amplitudes as two indexes to represent the influence data of the wind blades when the wind direction frequently changes. The pre-trained multi-modal prediction model is used to accurately predict the device rotation times and the rotation amplitudes in a specific time period in the future. At the same time, the weather forecast information is combined to quantitatively analyze the meteorological abnormality coefficient in the future period, which effectively improves the accuracy of the meteorological abnormality calculation and the accuracy of the meteorological prediction.

[0098] In the embodiment, the meteorological abnormality coefficient is generated according to the environmental data and the device parameters in a time period, including:

[0099] obtaining environmental data and equipment parameters in a time period; the environmental data includes environmental parameters and weather forecast; the equipment parameters include equipment parameters in a current time period and predicted equipment parameters in a future time period;

[0100] extracting wind speed FV, temperature WD and humidity SD in the environmental parameters and the weather forecast;

[0101] extracting equipment rotation number SZC and equipment rotation amplitude SZF in the equipment parameters;

[0102] constructing an environmental impact function HYF(FV, WD, SD) through a nonlinear relationship between the wind speed FV, the temperature WD and the humidity SD and the meteorological anomaly coefficient; constructing an equipment impact function SYF(SZC, SZF) through a nonlinear relationship between the equipment rotation number SZC and the equipment rotation amplitude SZF and the meteorological anomaly coefficient;

[0103] The meteorological anomaly coefficient is calculated through a formula, which is: ; wherein, is a weight coefficient, ∈(0, 1), the specific value is set according to experience, if it is considered that the value calculated by the environmental impact function has a greater influence on the meteorological anomaly coefficient, then can be set larger, such as is set to 0.7; if it is considered that the value calculated by the environmental impact function has a smaller influence on the meteorological anomaly coefficient, then can be set smaller, such as is set to 0.3; in this embodiment, is set to 0.4;

[0104] The environmental impact function HYF(FV, WD, SD) can be expressed as: ; is expressed as a wind speed nonlinear attenuation coefficient, ∈(0, 1), the specific value is set according to experience, in this embodiment, is set to 0.7, is set to control the marginal effect of extreme wind speed; WMD is expressed as temperature deviation sensitivity, which is to control the exponential decay rate of temperature anomaly, the specific value is set according to experience, in this embodiment, WMD is set to 3℃; BW and BS are expressed as standard temperature and standard humidity, the specific values are set according to experience, in this embodiment, BW is set to 25℃ and BS is set to 70%; is expressed as maximum wind speed;

[0105] The equipment impact function SYF(SZC, SZF) can be expressed as ; is expressed as a rotation number sensitivity coefficient, ∈(0, 1), the specific value is set according to experience, and in the embodiment, the value is set to 0.5, The setting is to control the saturation speed of the hyperbolic tangent function. The setting is to control the saturation speed of the hyperbolic tangent function. , which is represented as a rotation amplitude attenuation coefficient, ∈(0, 1), the specific value is set according to experience, and in the embodiment, the value is set to 0.6, The setting is to control the saturation speed of the hyperbolic tangent function. The setting is to control the saturation speed of the hyperbolic tangent function. and , respectively represented as a device rotation frequency reference value and a device rotation amplitude reference value, the specific values are set according to experience.

[0106] The environmental impact function and the device impact function are linearly weighted and fused, and several parameters in a time period are substituted into the calculation to obtain the meteorological anomaly coefficient corresponding to the time period.

[0107] In the embodiment, the device state score is generated according to the power plant data and the environmental data, which includes:

[0108] Obtain power plant data and environmental data corresponding to several time periods; the power plant data includes a power generation device ID and device parameters; the device parameters include the start-stop frequency QT, the rotation speed ZV, and the running time YS; the environmental data includes the temperature WD and the humidity SD;

[0109] A period score function ZPF(QT, ZV, YS, WD, SD) is constructed through a nonlinear relationship between the device parameters and the environmental data and the period device score; and the period score function is an increasing function, and the specific representation of the function is: ; wherein, is represented as the maximum start-stop frequency, the specific value is set according to the design parameters of the fan, is represented as the rated rotation speed, is represented as the total length of the design life, BW and BS are represented as the standard temperature and the standard humidity, and the specific values are set according to experience, and in the embodiment, BW is set to 25℃ and BS is set to 70%;

[0110] The corresponding device parameters and environmental data in a time period are input into the period score function to calculate the period device score; since the period score function is known, the corresponding parameter values are substituted to calculate the period device score. The higher the values of the device parameters and the environmental data in the time period, the higher the period device score calculated, which represents the greater the degree of wear caused by the device in the period.

[0111] According to the periodic equipment score, a plurality of time periods are divided into a plurality of periodic clusters; the calculation of the periodic cluster is because the higher the damage caused by the adjacent periods, the more the overall cumulative wear will generally present a nonlinear superposition effect, resulting in a total wear degree significantly higher than the linear superposition value under normal circumstances;

[0112] According to the formula The cluster periodic equipment score CZSP corresponding to the plurality of periodic clusters is calculated; wherein i represents the number of the time period in the periodic cluster; represents the coordination factor, ∈(0, 1), the specific value is set according to experience, and in the embodiment, it is set to 0.15, The setting of is to represent the amplification effect of the previous wear on the subsequent time period; the higher the periodic equipment score corresponding to the earlier time period, and the more the number of time periods in the periodic cluster, the more the cluster periodic equipment score corresponding to the periodic cluster will gradually increase;

[0113] The equipment state score corresponding to the current monitoring time point is obtained by summing a plurality of cluster periodic equipment scores before the current monitoring time point.

[0114] Please refer to Figure 2 , in the embodiment, according to the periodic equipment score, a plurality of time periods are divided into a plurality of periodic clusters, including the following steps:

[0115] Step one: obtaining a plurality of historical periodic equipment scores ZSP j and a plurality of time periods SZ j corresponding thereto;

[0116] Step two: creating a new periodic cluster list; the new periodic cluster list is used to store a plurality of time periods;

[0117] Step three: judging whether ZSP j is greater than the score threshold value, the score threshold value is set according to experience; yes, ZSP j is placed in the periodic cluster list, and the next time period is entered; enter step three; no, judge whether the periodic cluster list is an empty list; yes, ZSP j is placed in the periodic cluster list, and the next time period is entered; enter step two; no, create a new periodic cluster list, place ZSP j in the periodic cluster list, and enter the next time period, enter step two; wherein j represents the number corresponding to the time period; through the above steps, all adjacent time periods with a periodic equipment score higher than the score threshold value can be converged, so that the equipment score can be close to the actual situation when the equipment score is calculated, and the accuracy of the equipment score is improved.

[0118] The available power generated according to the device parameters and the device state score in the embodiment includes:

[0119] Obtaining a plurality of historical device parameters, historical device state scores, historical environment data, meteorological anomaly coefficients and prediction labels, and integrating the plurality of historical device parameters, historical device state scores, historical environment data and meteorological anomaly coefficients into a power prediction sequence;

[0120] Inputting the prediction labels and the power prediction sequence into a multi-modal prediction model to obtain a predicted power generation; the multi-modal prediction model is constructed by an artificial intelligence model;

[0121] Real-time extraction of the energy storage power in the power plant data; the power obtained by the model prediction refers to the power that can be generated between the current time and the prediction time point, and the available power refers to all the power that can be generated at the prediction time point, which needs to be added with the power stored in the energy storage device in the power plant to obtain the available power;

[0122] Summing the energy storage power and the predicted power generation to obtain the available power.

[0123] The embodiment predicts the power generation in the future time through the multi-source data such as environment data, device state score and meteorological data, and through the pre-trained multi-modal prediction model, and sums the power generation with the energy storage power in the power plant to obtain the available power. The prediction of the power generation through the multi-source data makes the prediction of the power generation more accurate, and provides a good foundation for the subsequent power supply scheme.

[0124] The multi-modal prediction model in the embodiment is constructed by an artificial intelligence model, including:

[0125] Obtaining historical prediction labels and corresponding historical parameter prediction sequences or historical power prediction sequences, and corresponding historical device parameters or historical power generation; the prediction labels include parameter prediction labels and power prediction labels;

[0126] According to the prediction labels, the plurality of historical parameter prediction sequences and historical device parameters, and the plurality of historical power prediction sequences and historical power generation are divided into training data, verification data and test data corresponding to the parameter prediction labels, and training data, verification data and test data corresponding to the power prediction labels; data preprocessing is performed on the training data, verification data and test data to obtain a training set, a verification set and a test set; the ratio between the training set, the test set and the verification set is 7:2:1;

[0127] According to the prediction labels, the plurality of historical parameter prediction sequences and historical device parameters, and the plurality of historical power prediction sequences and historical power generation are divided into training data, verification data and test data corresponding to the parameter prediction labels, and training data, verification data and test data corresponding to the power prediction labels; data preprocessing is performed on the training data, verification data and test data to obtain a training set, a verification set and a test set; the ratio between the training set, the test set and the verification set is 7:2:1;

[0128] Two machine learning models are selected as the base models of the multi-task branch; both of the machine learning models adopt an LSTM model;

[0129] The corresponding base model is trained through the respective training set, and the learning rate and other hyperparameters are adjusted on the respective validation set to obtain the respective pre-trained model;

[0130] The respective pre-trained model is verified on the respective test set, and finally the input prediction label and the corresponding parameter prediction sequence or power prediction sequence are obtained, and the output is a multi-modal prediction model for predicting device parameters or predicted power generation.

[0131] Please refer to Figure 3 In the embodiment, the adjustment scheme generated according to the available power and demand data includes:

[0132] Obtain the available power and demand data; the demand data includes the total demand power and the regional power demand and regional parameters corresponding to a plurality of regions;

[0133] Determine whether the available power is greater than the total demand power; if yes, do not perform any operation; if no, generate a power shortage warning signal, obtain the maximum purchase power, and set the maximum purchase power by the power plant according to the cost, and obtain the maximum power supply by summing the maximum purchase power and the available power; determine whether the maximum power supply is greater than the total demand power; if yes, do not perform any operation; if no, generate a regional power shortage warning signal; generate a plurality of regional distribution power according to the maximum power supply and the regional parameters; and supply power to a plurality of regions according to the regional distribution power.

[0134] In the embodiment, the plurality of regional distribution power is generated according to the maximum power supply and the regional parameters, which includes:

[0135] Obtain the maximum power supply and the regional parameters; the regional parameters include building grades, population densities, and special population quantities; the special population includes the elderly, the sick, and the disabled; consider a special building such as a hospital, which has a higher corresponding building grade, and the building grade is evaluated by experts according to the impact of power shortage, and the power supply in buildings with higher building grades in the region is preferentially guaranteed;

[0136] Extract the power required by the buildings in the plurality of regions whose building grades are higher than the grade threshold;

[0137] Obtain the population density RM and the special population quantity TS corresponding to the plurality of regions;

[0138] Construct a priority function YF(RM, TS) through the nonlinear relationship between the population density and the special population quantity and the regional priority; wherein the priority function is an increasing function; and the priority function is specifically represented as: ; wherein, is an exponential coefficient, ∈ (0, 1), the specific value is set according to experience, and in the embodiment, the value is set to 0.8, is set to 0.8, The setting of reflects the marginal diminishing effect, that is, the higher the population density, the smaller the marginal increase in power demand per unit population, and the over-concentration of resources is avoided. is a proportionality coefficient, > 0, the specific value is set according to experience, and in the embodiment, the value is set to 2, is set to 2, The setting of is to control the strength of the synergistic effect, that is, when RM and TS exceed the reference value at the same time, the regional priority is significantly improved. and respectively represent the reference density and the reference number, and the specific values are set according to experience.

[0139] The population density and the number of special groups in a plurality of regions are input into the priority function to obtain the regional priority.

[0140] The distributable power is calculated by the difference between the maximum power supply and the sum of the power required by a plurality of buildings. The power required by buildings with high grades is preferentially supplied, and after the supply, the remaining distributable power is allocated according to the priority.

[0141] It is judged whether the distributable power is greater than 0; if yes, the regional priority is normalized to obtain the normalized regional priority, and the distributable power of a plurality of regions is calculated by multiplying the distributable power and the regional priority; if no, a power failure warning signal is generated.

[0142] The regional allocation power is obtained by summing the distributable power of a region and the power required by the buildings in the corresponding region.

[0143] In the embodiment, the power supply capacity of the power plant is first compared with the actual demand data; if the power supply capacity can meet the demand data, no additional power purchase operation is required; if the power supply capacity cannot meet the demand, after implementing the power purchase measures, there are still some regions that cannot realize power coverage, at this time, the power is preferentially provided to these regions and buildings according to the priority of the region and the importance of the special buildings in the region, so as to guarantee the power demand of important buildings; then, the available power of each region is dynamically allocated, and the corresponding warning signal is sent at the same time, so as to make preparations in advance, and reduce the probability of poor user experience caused by sudden imbalance of power supply.

[0144] In the embodiment, the alarm signal is generated according to the meteorological anomaly coefficient and the equipment state score, which includes:

[0145] The meteorological anomaly coefficient and the equipment state score are obtained.

[0146] determining whether the meteorological anomaly coefficient is greater than a meteorological anomaly threshold value, the meteorological anomaly threshold value being set according to experience; yes, generating a meteorological alarm signal; no, not doing any operation;

[0147] determining whether the device state score is greater than a device retirement threshold value, the device retirement threshold value being set according to experience; yes, generating a device retirement alarm signal; no, determining whether the device state score is greater than D times the device retirement threshold value, yes, generating a device maintenance early warning signal; no, not doing any operation; wherein D represents a proportional coefficient, D∈(0, 1), the specific value being set according to experience, in the embodiment, D is set to 0.6.

[0148] Please refer to Figure 4 The second aspect embodiment of the present application provides a power grid meteorological data monitoring and evaluation method, comprising:

[0149] S0: obtaining power plant data and environment data and demand data; the environment data includes environment parameters and weather forecast; the power plant data includes power generation equipment ID and device parameters;

[0150] S1: generating a meteorological anomaly coefficient according to the environment parameters and the weather forecast; generating a device state score according to the power plant data and the environment data;

[0151] S2: obtaining device parameters in real time, and generating available power according to the device parameters and the device state score;

[0152] S3: generating an adjustment scheme according to the available power and the demand data; generating an alarm signal according to the meteorological anomaly coefficient and the device state score;

[0153] S4: making a prompt according to the alarm signal, and contacting a management personnel.

[0154] Another aspect embodiment of the present application provides a computer readable storage medium, the computer readable storage medium stores a computer program, the computer program is executed by a processor to realize a power grid meteorological data monitoring and evaluation system according to the first aspect embodiment of the present application.

[0155] Some data in the above formula are calculated by removing the dimension and taking the numerical value, the formula is obtained by software simulation of a large amount of collected data to be closest to the real situation; the preset parameters and the preset threshold value in the formula are set by a person skilled in the art according to the actual situation or obtained by a large amount of data simulation.

[0156] The working principle of the present application: obtain power plant data and environmental data and demand data; generate meteorological anomaly coefficient according to environmental parameters and weather forecast; generate equipment state score according to power plant data and environmental data; obtain equipment parameters in real time, generate available power according to equipment parameters and equipment state score; generate adjustment scheme according to available power and demand data; generate alarm signal according to meteorological anomaly coefficient and equipment state score; make prompt according to alarm signal, contact management personnel, quantify the monitoring results of various weather, calculate future meteorological anomaly coefficient, consider the influence of weather on future power generation, dynamically adjust the power supply of the region according to the demand data in the supply area, improve the balance between supply and demand of electric energy and the accuracy and comprehensiveness of weather monitoring, avoid the problem that the prior art lacks consideration of the influence degree of meteorological anomaly on the power consumption of related equipment during operation and the state of the equipment, so that the efficiency of power generation is low, leading to the imbalance between supply and demand of electric energy.

[0157] The above embodiments are only used to illustrate the technical method of the present application and not to limit it. Although the present application has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical method of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical method of the present application.

Claims

1. A power grid meteorological data monitoring and evaluation system, characterized in that, include: Data acquisition module, data analysis module, early warning module, and database; The data acquisition module is connected to the data analysis module; the data analysis module is connected to the early warning module; the data acquisition module, the data analysis module, and the early warning module are each connected to the database. The data acquisition module acquires power plant data, environmental data, and demand data through data acquisition equipment; the environmental data includes environmental parameters and weather forecasts; the power plant data includes power generation equipment IDs and equipment parameters. The data analysis module generates meteorological anomaly coefficients based on environmental data and equipment parameters within a time period, including: acquiring environmental parameters, equipment parameters, and weather forecasts within a time period T; the environmental parameters include wind speed (FV), wind direction, temperature (WD), and humidity (SD); the equipment parameters include the number of equipment rotations (SZC) and the equipment rotation amplitude (SZF); the weather forecast includes wind speed (FV), temperature (WD), and humidity (SD); the environmental parameters and weather forecasts are represented as environmental data corresponding to different time periods; Acquire historical environmental parameters, historical equipment parameters, and their corresponding historical weather forecasts and prediction tags for several historical time periods; The historical environmental parameters, historical equipment parameters, and their corresponding historical weather forecasts for several historical time periods are integrated into a parameter prediction sequence. The predicted labels and parameter prediction sequences are input into a multimodal prediction model to obtain the predicted device parameters for several time periods; the multimodal prediction model is constructed using an artificial intelligence model. Acquire environmental data and equipment parameters within a time period T; the environmental data includes environmental parameters and weather forecasts; the equipment parameters include equipment parameters within the current time period and predicted equipment parameters for future time periods. Extract wind speed (FV), temperature (WD), and humidity (SD) from environmental parameters and weather forecasts; Extract the number of equipment rotations (SZC) and the equipment rotation amplitude (SZF) from the equipment parameters; The environmental impact function HYF(FV, WD, SD) is constructed based on the nonlinear relationship between wind speed FV, temperature WD, humidity SD and meteorological anomaly coefficient; the equipment impact function SYF(SZC, SZF) is constructed based on the nonlinear relationship between equipment rotation number SZC and equipment rotation amplitude SZF and meteorological anomaly coefficient. The environmental impact function HYF(FV, WD, SD) is expressed as follows: ; This is expressed as the nonlinear attenuation coefficient of wind speed. ∈(0,1); WMD represents temperature deviation sensitivity; BW and BS represent standard temperature and standard humidity; This represents the maximum wind speed; The equipment influence function SYF(SZC, SZF) is expressed as follows: ; This is expressed as the sensitivity coefficient based on the number of rotations. ∈(0,1); This is expressed as the rotation amplitude attenuation coefficient. ∈(0,1); and These are respectively represented as the reference value for the number of equipment rotations and the reference value for the equipment rotation amplitude; The meteorological anomaly coefficient corresponding to the time period is obtained by linearly weighting and fusing the environmental impact function and the equipment impact function, and substituting several parameters within the time period into the calculation; an equipment status score is generated based on power plant data and environmental data; equipment parameters are acquired in real time, and the available power is generated based on the equipment parameters and equipment status score; an adjustment plan is generated based on the available power and demand data; and an alarm signal is generated based on the meteorological anomaly coefficient and equipment status score. The process of generating equipment status scores based on power plant data and environmental data includes: Acquire power plant data and environmental data corresponding to several time periods; the power plant data includes generator ID and equipment parameters; the equipment parameters include start-stop count QT, rotational speed ZV, and operating time YS; the environmental data includes temperature WD and humidity SD. The periodic scoring function ZPF(QT, ZV, YS, WD, SD) is constructed by the nonlinear relationship between equipment parameters and environmental data and periodic equipment scores. ;in, This represents the maximum number of start-stop cycles. This is expressed as the rated speed. The total design life is represented by BW and BS, which represent standard temperature and standard humidity, respectively. The relevant equipment parameters and environmental data within the time period are input into the periodic scoring function to calculate the periodic equipment score; Several time periods are divided into several period clusters based on the periodic device score; According to the formula The cluster periodic device score (CZSP) corresponding to several periodic clusters is calculated; where i represents the number of the time period within the periodic cluster. Represented as a cooperating factor, ∈(0,1); The device status score corresponding to the current monitoring time point is obtained by summing the scores of several clusters of periodic devices prior to the current monitoring time point.

2. The power grid meteorological data monitoring and evaluation system according to claim 1, characterized in that, The process of dividing several time periods into several period clusters based on periodic device scores includes the following steps: Step 1: Obtain several historical periodic equipment scores (ZSP) j and its corresponding time periods SZ j ; Step 2: Create a new list of periodic clusters; Step 3: Determine the historical cycle equipment score (ZSP) j Is it greater than the scoring threshold? Yes, the historical cycle equipment score ZSP will be used. j Place it into the periodic cluster list and proceed to the next time period; proceed to step three; No, check if the cycle cluster list is empty. If yes, assign the historical cycle device score ZSP. j Add to the cycle cluster list and proceed to the next time cycle; proceed to step two; otherwise, create a new cycle cluster list and add the historical cycle device score ZSP. j Place it into the periodic cluster list and proceed to the next time period, then proceed to step two; where j represents the number corresponding to the time period.

3. The power grid meteorological data monitoring and evaluation system according to claim 1, characterized in that, The process of generating available power based on device parameters and device status scores includes: Acquire several historical equipment parameters, historical equipment status scores, historical environmental data, meteorological anomaly coefficients, and prediction labels, and integrate these historical equipment parameters, historical equipment status scores, historical environmental data, and meteorological anomaly coefficients into a power generation prediction sequence; The predicted power generation is obtained by inputting the predicted label and the power generation prediction sequence into a multimodal prediction model; the multimodal prediction model is constructed using an artificial intelligence model. Real-time extraction of energy storage capacity from power plant data; The available power is obtained by summing the stored power and the predicted power generation.

4. A power grid meteorological data monitoring and evaluation system according to claim 1 or claim 3, characterized in that, The multimodal prediction model is constructed using an artificial intelligence model, including: Obtain historical prediction tags and their corresponding historical parameter prediction sequences or historical power generation prediction sequences, as well as their corresponding historical equipment parameters or historical power generation; the prediction tags include parameter prediction tags and power generation prediction tags; Several historical parameter prediction sequences and historical equipment parameters, as well as several historical power prediction sequences and historical power generation, are divided into multi-task branches according to prediction labels. Several historical parameter prediction sequences and historical equipment parameters, as well as several historical power prediction sequences and historical power generation, are divided into training data, validation data, and test data corresponding to parameter prediction labels and power prediction labels, according to prediction labels. The training data, validation data, and test data are preprocessed to obtain training sets, validation sets, and test sets. Choose two machine learning models as the base models for the multi-task branching; Each model is trained on its own training set and its own pre-trained model is obtained by adjusting the learning rate and other hyperparameters on its own validation set. By validating their respective pre-trained models on their respective test sets, the input prediction labels and their corresponding parameter prediction sequences or power prediction sequences are finally obtained, and the output is a multimodal prediction model that predicts equipment parameters or power generation.

5. The power grid meteorological data monitoring and evaluation system according to claim 1, characterized in that, The process of generating an adjustment plan based on available power and demand data includes: Acquire available and demand data; the demand data includes total demand for electricity and regional electricity demand and regional parameters for several regions. Determine if the available power supply exceeds the total power demand; if yes, do nothing; if no, generate a low power warning signal, obtain the maximum purchase power, and sum the maximum power supply to obtain the maximum power supply. Determine if the maximum power supply exceeds the total power demand; if yes, do nothing; if no, generate a regional power shortage alarm signal; generate power allocation for several regions based on the maximum power supply and regional parameters; and supply power to several regions according to the regional power allocation.

6. The power grid meteorological data monitoring and evaluation system according to claim 5, characterized in that, The process of generating power allocation for several regions based on the maximum power supply and regional parameters includes: Obtain the maximum power supply and regional parameters; the regional parameters include building grade, population density, and number of special groups; the special groups include the elderly, patients, and disabled persons; Extract the electricity required for buildings with a building grade higher than the grade threshold in several regions; Obtain the population density RM and the number of special populations TS for several regions; A priority function YF(RM, TS) is constructed based on the nonlinear relationship between population density, the number of special population groups, and regional priority; where the priority function is an increasing function. The population density and number of special groups in several regions are input into the priority function to calculate the regional priority; The power supply can be allocated by calculating the difference between the maximum power supply and the sum of the power requirements of several buildings; Determine if the divisible power is greater than 0; if yes, normalize the priorities of several regions to obtain normalized region priorities, and calculate the divisible power of several regions by multiplying the divisible power by the region priorities; if no, generate a power outage alarm signal. The regional allocated electricity is obtained by summing the regional distributable electricity with the electricity required by the buildings in the corresponding region.

7. A method for monitoring and evaluating power grid meteorological data, applied to a power grid meteorological data monitoring and evaluation system according to any one of claims 1-6, characterized in that, include: S0: Acquire power plant data, environmental data, and demand data; the environmental data includes environmental parameters and weather forecasts; the power plant data includes power generation equipment IDs and equipment parameters; S1: Generate meteorological anomaly coefficients based on environmental data and equipment parameters within a time period; generate equipment status scores based on power plant data and environmental data; S2: Real-time acquisition of device parameters, and generation of available power based on device parameters and device status score; S3: Generate adjustment plans based on available power and demand data; generate alarm signals based on meteorological anomaly coefficients and equipment status scores.

Images