Wind-solar complementary power generation system and method based on intelligent monitoring

Abstract

The invention discloses a wind-solar complementary power generation system and method based on intelligent monitoring, belongs to the field of wind-solar complementary power generation, and is used for solving the problems of fixed parameter weight, lack of environment-equipment coupling analysis and large power generation capacity prediction deviation of a traditional system. Generating a standard matching curve template to match the current working condition, and monitoring a working condition fluctuation index to trigger updating; key equipment parameters are screened based on historical data and association degree, an environmental working condition scene association rule base is constructed, generating capacity fluctuation is judged through double references, and a core influence chain is traced; matching historical samples through Euclidean distance to calculate an initial generating capacity predicted value, constructing three types of correction coefficients of environmental deviation, equipment health and a core influence chain, and obtaining a comprehensive correction factor and a correction predicted value; the monitoring accuracy and the fluctuation attribution capability are improved, the prediction deviation is reduced, and the system operation strategy is ensured to adapt to the actual power generation demand.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of wind-solar complementary power generation, and particularly relates to a wind-solar complementary power generation system and method based on intelligent monitoring. BACKGROUND

[0002] The current wind-solar complementary power generation system obtains environmental and equipment data through sensors and data acquisition terminals, and a monitoring center adjusts a control strategy in combination with third-party weather data; however, the system has fixed weights of wind-solar collaborative monitoring parameters, does not consider the correlation factors of the environment and the equipment, and cannot correct the prediction with real-time data, thereby raising the following technical problems: When monitoring and analyzing the environmental and equipment data, the system has fixed weight distribution of environmental monitoring parameters, does not select key working condition adaptation parameters from the environmental monitoring parameters and adjust the priority in real time according to dynamic working condition scenes, so that the key correlation factors between the environment and the equipment are missed, the core factors affecting the power generation efficiency cannot be accurately identified, and the power generation efficiency of the system is lost under different working conditions; When analyzing the reasons for the power generation fluctuation, the coupling factors of the environmental data and the equipment data are not considered, so that it is difficult to accurately attribute the core problems of the power generation efficiency fluctuation. The coupling analysis capability of the environment and the equipment is lacking, so that the third-party weather prediction results cannot be dynamically corrected with historical real-time data, and the power generation prediction deviation is large, and the subsequent control strategy adjustment is also inconsistent with the actual power generation demand, and therefore the wind-solar complementary power generation system and method based on intelligent monitoring are proposed. SUMMARY

[0003] The purpose of the application is to provide a wind-solar complementary power generation system and method based on intelligent monitoring to solve the problems raised in the background.

[0004] To achieve the above purpose, the application provides the following technical scheme: a wind-solar complementary power generation system based on intelligent monitoring, comprising: A parameter priority control module: real-time monitoring of environmental data and equipment operation data, construction of an environmental feature matrix; calculation of the characteristic correlation degree of the environmental data and the power generation efficiency fluctuation, selection of key dimension reduction dimensions of the environmental working condition scene; generation of a standard matching curve template, establishment of an environmental working condition scene matching library to match the current environmental working condition scene; generation of a scene-key environmental parameter-importance weight mapping table, real-time monitoring of the working condition fluctuation index and triggering of the environmental working condition scene update; A coupling factor analysis module: analysis of key equipment parameters corresponding to the environmental working condition scene; construction of an environmental working condition scene correlation rule library; determination of the power generation fluctuation of the current working condition scene, if there is, selection of abnormal key equipment parameters and abnormal key environmental parameters, determination of initial factors, and tracing of a core influence chain report; Predictive dynamic correction module: Collects environmental parameter prediction data, matches historical samples to calculate the initial power generation prediction value; constructs environmental deviation correction coefficient, equipment health correction coefficient, and core impact chain correction coefficient, and performs comprehensive analysis to obtain a comprehensive correction factor to correct the initial prediction value.

[0005] The preferred process for reducing key dimensions in environmental scenarios is as follows: Real-time collection of environmental and equipment operation data; setting monitoring periods and collecting environmental data at each monitoring time point within those periods. An environmental feature matrix is ​​formed by arranging monitoring time nodes as rows and environmental data types as columns, according to the time-parameter correspondence. Obtain the environmental feature matrix and power generation efficiency fluctuation data of samples from the same period of near preset time under different environmental operating conditions, and calculate the feature correlation between each environmental data and efficiency fluctuation. The environmental parameters are sorted by absolute value of correlation to obtain an importance sequence. The top-ranked environmental parameters are selected as the key dimensionality reduction dimensions for this scenario.

[0006] Preferably, the specific process of generating a standard matching curve template and establishing an environmental condition scenario matching library to match the current environmental condition scenario is as follows: Obtain environmental monitoring data for each historical monitoring period of the same period sample, and reduce the dimensionality according to the key dimensionality reduction dimensions of the corresponding environmental working conditions to obtain dimensional feature values; Dimension feature curves are plotted with time as the horizontal axis and dimensional feature values ​​as the vertical axis. The dynamic time warping algorithm is used to group curves with similarity exceeding a preset threshold into the same feature group. The curves within the group are fitted to obtain representative curves, and the curve with the largest sample size is selected as the standard matching curve template for the scene. The environment and working condition scenario matching library is formed by mapping and storing the working conditions scenario, key dimensionality reduction dimensions, and standard templates. Traverse the matching library, reduce the dimensionality of the environmental feature matrix for the current monitoring period to generate the current feature curve, calculate its similarity with the standard template and take the average to obtain the overall similarity, and select the scene with the highest overall similarity as the environmental working condition scene for the current monitoring period.

[0007] Preferably, the specific process of generating a scenario-key environmental parameter-importance weight mapping table, real-time monitoring of the operating condition fluctuation index, and triggering environmental operating condition scenario updates is as follows: For each environmental scenario, the environmental parameters corresponding to each key dimensionality reduction dimension are recorded as key environmental parameters. The feature correlation of each key environmental parameter is normalized to obtain the importance weight. Organize the key environmental parameters and their corresponding importance weights for each scenario, form a scenario-key environmental parameter-importance weight mapping table, substitute it into the current working scenario for matching, and output its key environmental parameters and importance weights. The rate of change of key environmental parameters at each subsequent monitoring time point under the current operating condition is collected in real time. The rate of change is multiplied by the corresponding importance weight and summed to obtain the operating condition fluctuation index. If the index of n consecutive nodes exceeds the preset threshold, the environmental condition scenario update mechanism is triggered, and the condition scenario matching process is re-executed. n is the preset threshold for the number of consecutive monitoring time nodes.

[0008] Preferably, the key equipment parameters corresponding to the analyzed environmental operating conditions are analyzed; the specific process of constructing the environment-equipment association rule base is as follows: For each environmental scenario, based on the corresponding scenario-key environmental parameter-importance weight mapping table, each key environmental parameter is obtained and organized into a scenario-key environmental parameter list. Obtain historical equipment operation data of samples from the same period of a preset duration in this scenario, count the co-occurrence frequency of each equipment parameter and power generation efficiency fluctuation, select equipment parameters whose co-occurrence frequency is ≥ preset co-occurrence threshold as candidate equipment parameters, and compile them into a scenario-candidate equipment parameter list; Pair candidate device parameters with key environmental parameters in the scenario-key environmental parameter list to form key environmental parameter-candidate device parameter pairs. Calculate the response correlation to determine key device parameters and construct a scenario-key environmental parameter-key device parameter lookup table. Statistical parameters are used to verify association rules based on association direction and time series features. A scenario-key environmental parameter-association rule mapping table is generated. All scenario mapping tables are organized to form an environmental working condition scenario association rule library. The current working condition scenario is substituted into the library and matched with the corresponding scenario-key environmental parameter-association rule mapping table.

[0009] Preferably, the process of determining the power generation fluctuation in the current operating scenario, and if it exists, the specific process of screening abnormal key equipment parameters and abnormal key environmental parameters is as follows: Obtain real-time power generation efficiency data for the current monitoring period, and use the average power generation efficiency of the same period sample with a near-preset duration in the current operating condition scenario as the first benchmark; The average actual power generation efficiency of the same combination of key environmental parameters and key equipment parameters in the current monitoring period is used as the second benchmark. If the absolute value of the deviation between the real-time power generation efficiency and any benchmark exceeds the preset fluctuation threshold, and the deviation remains unchanged for a preset number of consecutive monitoring time points, it is determined that the power generation has suddenly fluctuated; the direction of fluctuation, the magnitude of the deviation, and the start time of the fluctuation are recorded to generate a result of the sudden fluctuation in power generation.

[0010] The preferred process for determining initial factors and tracing back to generate a report on the core impact chain is as follows: When it is determined that there is a sudden fluctuation in power generation in the current environmental working conditions, the normal operating range of key equipment parameters is preset, the real-time data of key equipment parameters within the preset time range before and after the start time of the sudden fluctuation in power generation is extracted, and the key equipment parameters whose real-time data exceeds the corresponding normal operating range are screened out and recorded as the initial screening abnormal key equipment parameters. The efficiency impact direction of each initially screened abnormal key equipment parameter is statistically analyzed. The corresponding parameter is retained as the abnormal key equipment parameter according to the direction of the sudden fluctuation in the current power generation. The magnitude of the deviation from the normal range is calculated and a list of abnormal key equipment parameters is generated. Substitute the parameters in the list into the scenario-key environmental parameter-association rule mapping table corresponding to the current environmental working condition scenario, match the corresponding key environmental parameter, if the change of the key environmental parameter meets the standard, mark it as an abnormal key environmental parameter, record the association relationship to generate an abnormal key environmental parameter-abnormal key equipment parameter association table, compare the appearance time of the parameters in the table according to the time sequence to determine the initial influencing factors, and generate a chain effect starting point record table. The initial influencing factors are used to identify the preliminary impact chain. The historical frequency ratio is verified by querying the rule library associated with environmental conditions and scenarios. If the criteria are met, it is considered a core impact chain, and a core impact chain report is generated.

[0011] Preferably, the specific process of collecting environmental parameter prediction data and matching it with historical samples to calculate the initial power generation prediction value is as follows: Collect environmental parameter prediction values ​​corresponding to the preset prediction duration for the current environmental working condition scenario output by a third-party platform; Call historical data under the current environmental conditions, including the historical environmental parameter sequence-actual power generation correspondence and the average power generation of the same environmental conditions in the same period of the past preset time. Based on the predicted environmental parameter sequence formed by the predicted environmental parameter values ​​output by the third-party platform, historical environmental parameter sequences of all historical sample groups are extracted from the historical data of the current working conditions, and the Euclidean distance between the two is calculated and denoted as the similarity distance. Historical sample groups with similarity distance ≤ preset similarity threshold are selected as matching sample groups, and their average actual power generation is calculated and denoted as the average power generation of the matching sample. To match the sample power generation mean with the historical average power generation under the same operating conditions, a weighted calculation is performed to obtain the initial power generation prediction value under the current environmental operating conditions.

[0012] Preferably, the specific process of analyzing and synthesizing correction factors to correct the initial predicted value is as follows: The system calls up the key environmental parameters and their corresponding importance weights for the current environmental working conditions, collects the real-time values ​​of the key environmental parameters in the current scenario, and retrieves the predicted values ​​for the corresponding time period from the short-term weather forecast data from third parties. By calculating the deviation rate of each key environmental parameter, the environmental deviation correction coefficient is obtained. Real-time collection of operational data of key equipment parameters in the current environmental working conditions; calculation of the health status of each key equipment parameter and taking the average value to obtain the equipment health correction coefficient; Obtain the core impact chain report within the preset period, and statistically analyze the frequency ratio of environmental factors and the corresponding actual power generation deviation rate, as well as the frequency ratio of equipment factors and the corresponding actual power generation deviation rate. Calculate the core impact chain correction coefficient based on the relationship between the two frequency ratios. If there is no core impact chain record within the preset period, the core impact chain correction coefficient is 1. By comprehensively analyzing the environmental deviation correction coefficient, the equipment health correction coefficient, and the core impact chain correction coefficient, a comprehensive correction factor is obtained. The initial power generation forecast under the current environmental conditions is multiplied by a comprehensive correction factor to obtain the corrected forecast value within a preset forecast period. The corrected forecast value is then sent to the personnel terminal to adjust the operation strategy so that the operation strategy matches the predicted power generation demand.

[0013] Compared with the prior art, the beneficial effects of the present invention are: (1) The wind-solar hybrid power generation system and method based on intelligent monitoring can monitor environmental and equipment data in real time through the parameter priority control module, calculate the feature correlation degree to screen key dimensionality reduction dimensions of environmental operating conditions, generate a matching library to dynamically match the current operating conditions, and also monitor the operating condition fluctuation index to trigger updates; avoid missing key correlation factors of environment and equipment, accurately identify the core parameters affecting power generation efficiency, and reduce power generation efficiency loss under different operating conditions.

[0014] (2) The wind-solar hybrid power generation system and method based on intelligent monitoring combines historical data to screen key equipment parameters, build an association rule base, use dual benchmarks to determine power generation fluctuations, and then use time series comparison of abnormal parameters to determine the initial factors and trace the core impact chain; solves the problem of difficulty in locating chain causes due to lack of coupling analysis, and provides accurate attribution basis for subsequent system regulation.

[0015] (3) The wind-solar hybrid power generation system and method based on intelligent monitoring calculates the initial prediction value by matching historical samples with Euclidean distance, and then generates a comprehensive correction factor by combining three types of correction coefficients: environmental deviation, equipment health, and core impact chain, to correct the prediction value and guide the adjustment of equipment parameters; solves the problems of large prediction deviation and disconnect between strategy and demand, and reduces power waste or insufficient supply. Attached Figure Description

[0016] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1; Please see Figure 1 The present invention provides a wind-solar hybrid power generation system and method based on intelligent monitoring, including: a parameter priority control module, a coupling factor analysis module and a prediction dynamic correction module; The parameter priority control module monitors environmental data and equipment operation data in real time, constructs an environmental feature matrix, calculates the correlation between environmental data and power generation efficiency fluctuations, and filters key dimensionality reduction dimensions for environmental operating conditions. It generates standard matching curve templates, establishes an environmental operating condition scenario matching library to match the current environmental operating condition scenario, and generates a scenario-key environmental parameter-importance weight mapping table. It monitors the operating condition fluctuation index in real time and triggers environmental operating condition scenario updates. The specific process is as follows: Based on a distributed sensor array, environmental data is monitored in real time, including: wind speed change gradient, instantaneous change rate of light intensity, temperature and humidity gradient, and transient rate of atmospheric pressure. Dedicated sensors are installed at the wind turbine, photovoltaic panel, and energy storage equipment to monitor equipment operation data in real time, including: blade stress, photovoltaic panel string voltage deviation, and energy storage battery charge / discharge rate. Set a monitoring period and collect environmental data at each monitoring time point within the monitoring period; An environmental feature matrix is ​​formed by arranging monitoring time nodes as rows and environmental data types as columns, according to the time-parameter correspondence. The system acquires environmental feature matrices and corresponding power generation efficiency fluctuation data for samples within a near-preset time range under different environmental conditions. These environmental conditions include: stable light wind, gusty cloudy, high temperature and strong sunlight. For each environmental operating scenario, the correlation degree between each environmental data and the power generation efficiency fluctuation characteristics under that scenario is calculated and denoted as the feature correlation degree. Further feature correlation is obtained as follows: From the samples of the same period with a near-preset duration in this working condition scenario, extract two types of data for the same historical monitoring period: time series data of a certain environmental parameter and time series data of power generation efficiency for the corresponding period. By analyzing time-series data on power generation efficiency, and statistically analyzing its dispersion (such as the extent to which the data deviates from the average efficiency), characteristic values ​​reflecting the degree of efficiency fluctuation during that period are obtained (the greater the fluctuation, the larger the characteristic value). The overall level (e.g., average value) and the dispersion (e.g., data dispersion degree) of environmental parameter time series data are statistically analyzed separately. If the efficiency fluctuation characteristic value tends to increase when environmental parameters increase, it indicates that the two are in a positive trend; if the efficiency fluctuation characteristic value tends to decrease when environmental parameters increase, it indicates that the two are in a negative trend; the stronger the synchronicity, the closer the correlation between the two. The statistical results of the above trend consistency are standardized to obtain a value in the range of -1 to 1, which is the feature correlation degree. The environmental data are sorted from largest to smallest based on the absolute value of their correlation to obtain a sequence of environmental parameter importance. A preset selection coefficient m is used to select the top m environmental parameters with the highest correlation from the environmental parameter importance sequence, which are then used as the key dimensionality reduction dimensions for this environmental scenario.

[0019] For each historical monitoring period in the same period sample, obtain the environmental monitoring data corresponding to each monitoring time node; For the environmental monitoring data corresponding to each monitoring time node, dimensionality reduction is performed according to the key dimensionality reduction dimensions corresponding to the environmental working scenario, and the values ​​of each key dimensionality reduction dimension at each monitoring time node are obtained and recorded as dimensional feature values. For each key dimension reduction dimension, the dimension feature value at each monitoring time node is marked on the coordinate system with the dimension feature value at each monitoring time node as the vertical axis and time as the horizontal axis. Then, based on the chronological order, adjacent marked data points are connected sequentially by a smooth curve to obtain the change curve of the key dimension reduction dimension in the historical monitoring period, which is denoted as the dimension feature curve. For each key dimension reduction, collect the dimension feature curve corresponding to that key dimension in all historical monitoring periods within the sample time range of the environmental working conditions scenario. The similarity between curves is calculated using a dynamic time warping algorithm, and curves with similarity higher than the corresponding preset threshold are grouped into the same feature group. For curves within the same feature group, calculate the arithmetic mean of the dimensional eigenvalues ​​of all curves at the same time node. Plot time on the horizontal axis and the mean on the vertical axis, and use the least squares method to fit a continuous curve, which is then used as the representative curve of the feature group. Select the representative curve with the largest sample size in the feature group as the standard matching curve template for the environmental working scenario (each key dimension of each scenario corresponds to a unique standard template). All standard matching curve templates corresponding to environmental operating scenarios are classified and stored according to the mapping relationship between operating scenario - key dimensionality reduction dimension - standard template, forming an environmental operating scenario matching library; each key dimensionality reduction dimension of each scenario in the environmental operating scenario matching library corresponds to a unique standard matching curve template. Based on the environmental feature matrix of the current monitoring period, the operating scenario of the current monitoring period is matched from the environmental operating scenario matching library, specifically as follows: Iterate through all environmental scenarios in the environmental scenario matching library: For each environmental scenario, based on the key dimensionality reduction dimensions of the environmental scenario, the environmental feature matrix of the current monitoring period is reduced in dimensionality, and the dimensionality feature values ​​of each key dimensionality reduction dimension in each time node of the current monitoring period are extracted. For each key dimension reduction dimension, generate the current feature curve corresponding to that dimension with the current dimension feature value on the vertical axis and time on the horizontal axis. The dynamic time warping algorithm is used to calculate the similarity between the current feature curve and the standard matching curve template of the key dimensionality reduction dimension; The average similarity of all key dimensionality reduction dimensions under the environmental working condition scenario is taken as the overall similarity between the environmental working condition scenario and the current monitoring data. From all environmental scenarios, the scenario with the highest overall similarity is selected as the current monitoring scenario.

[0020] For each environmental scenario, the environmental parameters corresponding to each key dimensionality reduction dimension of the environmental scenario are denoted as key environmental parameters. The feature correlation of each key environmental parameter is normalized to obtain the importance weight of each key environmental parameter; Obtain the key environmental parameters and their corresponding importance weights for each environmental scenario, and organize them to form a scenario-key environmental parameter-importance weight mapping table. The current working condition scenario is substituted into the scenario-key environmental parameter-importance weight mapping table for matching, and the key environmental parameters of the current working condition scenario and their corresponding importance weights are output; see the table below for reference:

[0021] For the current working conditions, the rate of change of each key environmental parameter (such as the magnitude of change of value per unit time) is collected in real time at each subsequent monitoring time point. At the same monitoring time point, the operating condition fluctuation index of that point is obtained by multiplying the rate of change of each key environmental parameter by its corresponding importance weight and summing the results. A preset operating condition fluctuation index threshold is set. If the operating condition fluctuation index exceeds the preset operating condition fluctuation index threshold for n consecutive monitoring time nodes, it indicates that the current scenario has changed significantly, triggering the environmental operating condition scenario update mechanism and re-executing the environmental operating condition scenario matching process. Here, n is the preset threshold for the number of consecutive monitoring time nodes.

[0022] It should be noted that by monitoring detailed environmental data such as wind speed gradient and instantaneous change rate of sunlight using a distributed sensor array, and combining this with data from specialized equipment such as wind turbine blade stress and photovoltaic panel string voltage deviation, the problem of generalized monitoring data in traditional systems can be avoided. Furthermore, by constructing an environmental feature matrix and calculating feature correlations to filter key dimensionality reduction dimensions of environmental operating scenarios, redundant parameters can be eliminated, focusing on core environmental parameters that significantly affect power generation efficiency fluctuations (such as the instantaneous change rate of sunlight under high temperature and strong sunlight scenarios, and the wind speed gradient under stable light wind scenarios). This solves the problem of fixed parameter weights in traditional systems that cannot adapt to dynamic operating conditions, laying a precise data foundation for subsequent analysis. A standard matching curve template is generated using a dynamic time warping algorithm, and an environmental condition scenario matching library is established. By comparing the similarity between the current feature curve and the standard template, the current environmental condition scenario can be accurately identified. At the same time, by monitoring the condition fluctuation index (combining the change rate of key environmental parameters and importance weights), if the index of n consecutive nodes exceeds the threshold, the condition is updated. This avoids the inaccurate control caused by the lag in condition identification in traditional systems and ensures that the system always carries out subsequent operations based on the current real condition. The scenario-key environmental parameter-importance weight mapping table generated by the module can provide the coupling factor analysis module with accurate key environmental parameters and weights, helping it to efficiently screen key equipment parameters and build an environment-equipment association rule base. At the same time, the dynamically updated operating condition information can also make the power generation prediction of the prediction and dynamic correction module more in line with the current operating conditions, improving the coordination and control accuracy of the entire system from the source, and reducing the power generation efficiency loss caused by improper parameter adaptation or misjudgment of operating conditions.

[0023] Coupling Factor Analysis Module: Based on historical equipment operation data of environmental operating scenarios, analyzes corresponding key equipment parameters; constructs an environmental operating scenario association rule base based on key equipment parameters and historical data; compares real-time efficiency with benchmark values ​​to determine the power generation fluctuation of the current operating scenario; if power generation fluctuation exists, it filters abnormal parameters and determines initial factors based on the reference table and mapping table of the current operating scenario, and traces and generates a core impact chain report. The specific process is as follows: For each environmental scenario, based on the corresponding scenario-key environmental parameter-importance weight mapping table, obtain the corresponding key environmental parameters and organize them into a scenario-key environmental parameter list. Acquire historical equipment operation data of samples from the same period for a near-preset duration under environmental conditions, and statistically analyze the co-occurrence frequency of fluctuations in each equipment parameter and power generation efficiency; Furthermore, the calculation method for the co-occurrence frequency of equipment parameter and power generation efficiency fluctuations is as follows: set the fluctuation judgment threshold of equipment parameters (i.e. the minimum amplitude of equipment parameters exceeding the normal range) and the fluctuation judgment threshold of power generation efficiency. The number of times the power generation efficiency synchronously reaches the fluctuation judgment threshold within a preset time window when the statistical equipment parameters reach the fluctuation judgment threshold is recorded as the co-occurrence count. The co-occurrence frequency of each equipment parameter is calculated using the formula: Co-occurrence frequency = Number of co-occurrences / Total number of times the equipment parameter reaches the fluctuation judgment threshold.

[0024] Select device parameters whose co-occurrence frequency is greater than or equal to the preset co-occurrence threshold as candidate device parameters, and organize them in the format of environmental condition scenario - candidate device parameter name - co-occurrence frequency to generate a scenario - candidate device parameter list; For each candidate device parameter in the scenario-candidate device parameter list, its direct physical relationship with key environmental parameters of the current environmental scenario is determined through data analysis to filter out key device parameters. The specific process is as follows: Pair the current candidate device parameters with each key environmental parameter in the scenario-key environmental parameter list to form key environmental parameter-candidate device parameter pairs; For each pair of key environmental parameters and candidate equipment parameters, historical time-series data are extracted from samples of the same period with a near-preset duration under that environmental condition scenario. The trigger strength threshold for this parameter pair is preset (i.e., the minimum change that the key environmental parameters need to reach). Preset response time window (i.e., the time range within which candidate device parameters may respond after a key environmental parameter reaches the trigger strength threshold). The total number of times that key environmental parameters reach the trigger intensity threshold, and the number of times that candidate device parameters reach the preset response amplitude within the response time window, are recorded as the number of valid responses. Calculate the response correlation for each parameter pair using the formula: Response correlation = Number of valid responses / Total number of responses; If the correlation between the candidate device parameter and any set of parameter pairs is greater than or equal to the preset response threshold, it is determined that the candidate device parameter has a direct physical interaction with the key environmental parameters of the current scene. If the correlation between the responses of all parameter pairs is less than the preset response threshold, it is determined that the candidate device parameter has no clear physical correlation with the key environmental parameters of the scene and is therefore eliminated. All candidate device parameters that are determined to have a direct physical interaction relationship are defined as key device parameters for this environmental working condition scenario; Organize the parameters according to the format of environmental scenario - key environmental parameter set - key equipment parameter - related parameter pair (including response correlation degree, trigger intensity threshold, response time window) to generate a scenario - key environmental parameter - key equipment parameter comparison table. Based on the scenario-key environmental parameter-key equipment parameter comparison table corresponding to the environmental operating conditions and historical data of samples from the same period of the near preset duration, environment-equipment association rules are constructed for the environmental operating conditions. The specific process is as follows: For each pair of key environmental parameters and key equipment parameters, extract the corresponding historical time-series data and analyze the consistency of their changing trends: If key environmental parameters change in a preset direction (e.g., upward direction), and key equipment parameters change in a direction that improves power generation efficiency (e.g., increased wind speed gradient → decreased blade speed deviation), then it is determined to be a positive correlation. If key environmental parameters change in a preset direction while key equipment parameters change in a direction that reduces power generation efficiency, then it is determined to be a reverse correlation. Organize them according to the format of environmental operating scenario - key environmental parameters - key equipment parameters - correlation direction to form a single parameter correlation trend table; For each pair of key environmental parameters and key equipment parameters that are determined to be positively / negatively correlated, two types of time-series characteristics are statistically analyzed: Timing delay value: that is, the average time difference of the synchronous response of key equipment parameters after the key environmental parameters change (consistent with the response time window of the parameter pair). Intensity threshold: The preset intensity threshold of the key environmental parameter that triggers a change in the response of the key equipment parameter (consistent with the trigger intensity threshold of the parameter pair). Organize the data according to the format of environmental operating scenario - key environmental parameters - key equipment parameters - time delay value - intensity threshold, and generate a related time series feature table; Based on the correlation direction of the single-parameter correlation trend table and the time-series characteristics of the correlation time-series characteristic table, a correlation rule is constructed for key environmental parameters changing according to intensity threshold → key equipment parameters changing according to correlation trend and time-series delay value. Specifically: Select samples from the same period of the current environmental working conditions for a preset verification time (such as samples from the same period of the past 3 months) as the verification set, substitute the above association rules into the verification set, and calculate the matching rate of the association rules. Furthermore, the specific calculation process of the matching rate is as follows: the number of events in the statistical verification set where the key environmental parameters reach the corresponding preset strength threshold and the key equipment parameters change in the associated direction within the time delay value range is recorded as the rule matching number; The total number of events in the statistical verification set that reach the corresponding preset intensity threshold is recorded as the total number of rule triggers. The matching rate of an association rule can be obtained using the formula: Match rate of association rule = Number of times the rule is matched / Total number of times the rule is triggered.

[0025] If the matching rate of an association rule is greater than or equal to the preset rule matching threshold, the association rule is retained, and the matching rate is used as the association weight of the association rule. If the matching rate of the association rule is less than the preset rule matching threshold, then adjust the preset strength threshold of the parameter pair and resubmit it into the validation set to calculate the matching rate until the matching rate of the association rule is greater than or equal to the preset rule matching threshold. The verified key environmental parameter-key equipment parameter association rules are organized into the format of environmental condition scenario-key environmental parameter-association rule (including association direction, time delay value, strength threshold, association weight) to generate a scenario-key environmental parameter-association rule mapping table. The scenario-key environmental parameter-association rule mapping tables for all environmental operating scenarios are numbered and organized to form an environmental operating scenario association rule library. Substitute the current working condition scenario into the environmental working condition scenario association rule library, and match the scenario-key environmental parameter-association rule mapping table corresponding to the working condition scenario. The process of acquiring real-time power generation efficiency data for the current monitoring period and determining whether there are sudden fluctuations in power generation is as follows: First benchmark: The average power generation efficiency of the same period sample with a near-preset duration in the current operating condition scenario is used as the benchmark value of the historical operating condition scenario with the same environment. Second benchmark: Based on the real-time data of key environmental parameters and key equipment parameters during the current monitoring period, query the average actual power generation efficiency corresponding to the same combination of key environmental parameters and key equipment parameters in the historical equipment operation data, and use this average value as the real-time predicted efficiency value. If the absolute value of the deviation between the real-time power generation efficiency and the first benchmark during the current monitoring period is greater than the preset fluctuation threshold, or the absolute value of the deviation between the real-time power generation efficiency and the second benchmark is greater than the preset fluctuation threshold, and the deviation is maintained for a preset number of consecutive monitoring time nodes, then the power generation during the current monitoring period is determined to be fluctuating suddenly. When power generation suddenly fluctuates, record the direction of the sudden fluctuation, the magnitude of the deviation (take the larger of the absolute values ​​of the deviations between the real-time power generation efficiency and the first and second benchmarks), and the start time of the fluctuation, and generate a result for determining the sudden fluctuation in power generation.

[0026] When a sudden fluctuation in power generation is detected, based on the scenario-key environmental parameter-key equipment parameter lookup table and the scenario-key environmental parameter-association rule mapping table corresponding to the current environmental operating conditions, the core impact chain of the sudden fluctuation in power generation is traced and formed. The specific process is as follows: For key equipment parameters, the normal operating range of key equipment parameters is preset, and real-time data of key equipment parameters within the preset time range before and after the start time of sudden fluctuation in power generation are extracted. Key equipment parameters whose real-time data exceeds the corresponding normal operating range are screened out and recorded as the initial screening abnormal key equipment parameters. For each critical equipment parameter identified in the initial screening, the historical percentage of times that the power generation efficiency changed towards a decrease or increase when the corresponding critical equipment parameter became abnormal was statistically analyzed. If the reduction ratio is greater than or equal to the preset efficiency impact threshold, the resulting efficiency reduction will be determined as the efficiency impact direction of the key equipment parameter of the initial screening anomaly. If the increase in percentage is greater than or equal to the preset efficiency impact threshold, then the increase in efficiency will be determined as the direction of the efficiency impact of the key equipment parameter of the initial screening anomaly. If the current power generation suddenly fluctuates and decreases, retain the initial screening abnormal key equipment parameters that affect efficiency and lead to the decrease in efficiency, and record them as abnormal key equipment parameters. If the current power generation suddenly fluctuates and increases, retain the initial screening abnormal key equipment parameters that affect efficiency in the direction of increasing efficiency, and record them as abnormal key equipment parameters. Using the formula: Exceeding normal range range = |Real-time data - Normal operating range boundary value| / Normal operating range boundary value × 100%, calculate the extent to which each abnormal critical equipment parameter exceeds the normal range. Organize the parameters into a list of abnormal critical equipment parameters in the format of "abnormal critical equipment parameters - magnitude of deviation from normal range - direction of fluctuation correlation". For each abnormal critical equipment parameter in the list of abnormal critical equipment parameters, substitute it into the scenario-critical environment parameter-association rule mapping table corresponding to the current environmental working scenario, and match the critical environment parameter corresponding to the abnormal critical equipment parameter. Obtain real-time data of the key environmental parameter before the sudden fluctuation in power generation. If the magnitude of its change reaches the strength threshold of the corresponding association rule in the mapping table, then mark the key environmental parameter as an abnormal key environmental parameter. Record the correlation between abnormal critical environmental parameters and abnormal critical equipment parameters (including the time delay value and intensity threshold of the corresponding correlation rules), and generate an abnormal critical environmental parameter-abnormal critical equipment parameter correlation table; see the table below for reference:

[0027] Compare the occurrence times of the abnormal critical environmental parameters and abnormal critical equipment parameters in the abnormal critical environmental parameters-abnormal critical equipment parameters association table in chronological order; If the occurrence time of the abnormal critical environmental parameter is earlier than the occurrence time of the abnormal critical equipment parameter, then the abnormal critical environmental parameter is determined to be the initial influencing factor. If the occurrence time of the abnormal critical equipment parameter is earlier than the occurrence time of the abnormal critical environmental parameter, then the abnormal critical equipment parameter is determined to be the initial influencing factor. Organize the data according to the format of environmental operating scenario - initial influencing factor type (environment / equipment) - initial influencing factor name - abnormal parameter occurrence sequence, and generate a chain reaction starting point record table; Starting with the initial influencing factors in the chain impact starting point record table, the initial impact chain is formed by sorting out the initial influencing factors → intermediate related parameters (abnormal key environmental parameters or abnormal key equipment parameters) → sudden fluctuations in power generation. Query the historical time-series data in the environmental operating condition scenario association rule base, and calculate the historical frequency percentage of the occurrence of the initial impact chain in the same period of samples with a preset time (i.e., the proportion of the number of times the initial impact chain occurs to the total number of the same type of sudden fluctuation in power generation). If the proportion of historical frequency is greater than or equal to the preset frequency threshold, then the initial impact chain is determined to be the core impact chain of sudden fluctuations in power generation. If the historical frequency percentage is less than the preset frequency threshold, then supplement the real-time data of key environmental parameters and key equipment parameters in other scenarios under the current environmental conditions, and re-execute the process of screening abnormal key equipment parameters → associating abnormal key environmental parameters → tracing the starting point of chain impact until the impact chain with the historical frequency percentage is greater than or equal to the preset frequency threshold is obtained. The core impact chain report is generated by organizing the information according to the following format: each link in the core impact chain (initial influencing factors, intermediate related parameters, and results of sudden fluctuations in power generation) - the time sequence relationship of each link - the correlation weight of the corresponding correlation rules - the historical frequency ratio.

[0028] It should be noted that by statistically analyzing the co-occurrence frequency of equipment parameters and power generation efficiency fluctuations, candidate equipment parameters that may be related are initially screened out. Then, by calculating the response correlation between key environmental parameters and candidate equipment parameters, and combining the physical interaction relationship, parameters without clear correlation are eliminated. Finally, key equipment parameters that are suitable for the current environmental operating conditions are determined, avoiding the problem of traditional systems "generalizing the monitoring of equipment parameters and failing to focus on core related equipment", thus laying a precise object foundation for subsequent coupling analysis. By statistically analyzing the correlation direction, time delay value, and intensity threshold of key environmental parameters and key equipment parameters, association rules are constructed. A validation set is introduced to calculate the matching rate. Rules with unsatisfactory matching rates are adjusted and optimized before being retained to ensure that the rules fit the actual working conditions. At the same time, an association rule library is formed according to the working scenario, which solves the problem of "no clear basis for environment-equipment association and chaotic analysis logic" in traditional systems, and provides traceable rule support for fluctuation attribution. By adopting a dual benchmark comparison of "historical benchmark value under the same operating conditions + real-time predicted efficiency value" and combining the judgment condition of "continuously preset number of monitoring nodes maintaining deviation", it avoids the shortcomings of a single historical benchmark being unable to reflect real-time operating condition changes, and also avoids misjudgment caused by short-term random deviations (such as not judging a single node deviation as a fluctuation), ensuring that the real sudden fluctuations in power generation can be accurately identified, providing reliable fluctuation judgment results for subsequent attribution analysis. The module starts by screening key equipment parameters for initial abnormalities, then identifies truly related abnormal parameters based on the direction of efficiency impact. It then matches these parameters with related key environmental parameters and determines the initial influencing factors through time-series comparison. Finally, it identifies the core impact chain verified by historical frequency (e.g., "abnormal instantaneous rate of change of sunlight → voltage deviation of photovoltaic panel strings → reduced power generation"). This process solves the problem of traditional systems that "can only detect fluctuations but cannot locate the source and intermediate links of the chain of impacts." It provides accurate attribution basis for subsequent parameter correction by the predictive dynamic correction module and adjustment of system operation strategies, helping to reduce power generation efficiency losses caused by unclear fluctuation attributions.

[0029] The predictive dynamic correction module collects third-party environmental parameter prediction data, calculates the initial power generation prediction value by matching historical samples using Euclidean distance, constructs environmental deviation correction coefficients, equipment health correction coefficients, and core impact chain correction coefficients, and obtains a comprehensive correction factor after comprehensive analysis to correct the initial prediction value. The specific process is as follows: Collect environmental parameter prediction values ​​for the current environmental conditions and scenarios corresponding to the preset prediction duration from third-party platforms, including: environmental parameter values ​​such as wind speed, light intensity, temperature, and humidity for the preset prediction duration in the future; Call historical data under the current environmental operating conditions, including: the correspondence between the historical environmental parameter sequence and the actual power generation in the same period of the near preset time, and the historical average power generation under the same operating conditions in this scenario; Based on the predicted environmental parameter sequence formed by the environmental parameter prediction values ​​of the current environmental working condition scenario with a preset prediction time output by a third-party platform, the historical environmental parameter sequence corresponding to all historical sample groups is extracted from the historical data of the current environmental working condition scenario. Calculate the Euclidean distance between the predicted environmental parameter sequence and each historical environmental parameter sequence, and record it as the similarity distance; A preset similarity threshold is set, and historical sample groups with a similarity distance less than or equal to the preset similarity threshold are defined as matching sample groups. The average power generation of the matched sample group is calculated by averaging the actual power generation of the matched sample group. After assigning preset weight coefficients to the average power generation of the matched samples and the average power generation under the same historical conditions, a weighted calculation is performed to obtain the initial power generation prediction value under the current environmental conditions. Call upon the key environmental parameters and their corresponding importance weights for the current environmental operating conditions; Real-time values ​​of key environmental parameters in the current working environment are collected, and forecast values ​​for the corresponding time period are retrieved from third-party short-term weather forecast data. For each key environmental parameter, the deviation rate P of each key environmental parameter is obtained using the formula: Deviation rate = |real-time value - predicted value| / predicted value × 100%. Using the formula: We obtain the environmental deviation correction coefficient k1, where i is the label of the key environmental parameter, R is the total number of key environmental parameters, wi is the importance weight corresponding to the i-th key environmental parameter, and Pi is the deviation rate corresponding to the i-th key environmental parameter. Among them, if the deviation between the real-time value and the predicted value is smaller, k1 is closer to 1 and the correction magnitude is smaller; if the deviation is larger, k1 deviates further from 1 and the correction magnitude is larger. Real-time collection of operational data for key equipment parameters in the current environmental operating conditions; for each key equipment parameter, the health status of the key equipment parameter is obtained using the formula: Health Status = Actual Operating Value / Average Normal Operating Range. The average health status of all key equipment parameters is statistically analyzed to obtain the equipment health correction coefficient k2. Among them, the closer the equipment is to the normal state, the closer k2 is to 1, and the smaller the correction to the predicted value; the more significant the equipment abnormality, the smaller k2 is, and the larger the correction magnitude. Obtain the core impact chain report within a preset period, and calculate the frequency percentage (HB) of environmental factors dominating the core impact chain (e.g., abnormal key environmental parameters being the initial impact factor); simultaneously calculate the average actual power generation deviation rate (HP) of the core impact chain dominated by environmental factors within the preset period; |(Actual power generation deviation rate = |(Actual power generation - Initial power generation) / Initial power generation)| The frequency percentage (SB) of equipment-driven factors (such as abnormal key equipment parameters as initial influencing factors) in the core impact chain is calculated; simultaneously, the average actual power generation deviation rate (SP) of the core impact chain dominated by equipment factors within the preset period is calculated. If the frequency percentage HB is greater than the frequency percentage SB, the core influence chain correction coefficient k3 is obtained using the formula: k3=1-(HP / 100); and then corrected according to the environmental deviation law. If the frequency percentage HB is less than the frequency percentage SB, use the formula: k3=1-(SP / 100) to obtain the core influence chain correction coefficient k3; then correct according to the equipment deviation law. If the frequency percentage HB equals the frequency percentage SB, the core influence chain correction coefficient k3 is obtained using the formula: k3=1-{(HP+SP) / (2×100)}; the correction is then made by combining the deviation patterns of the two. If there are no core impact chain records within the preset period, then k3=1 (no additional correction). By substituting the environmental deviation correction coefficient k1, equipment health correction coefficient k2, and core impact chain correction coefficient k3 calculated under the current environmental working conditions into the formula: K=k1×a1+k2×a2+k3×a3, the comprehensive correction factor K is obtained, where a1, a2, and a3 are preset weight coefficients. By multiplying the initial power generation forecast under the current environmental conditions by a comprehensive correction factor, the corrected forecast value within the preset forecast period is obtained. By sending revised forecast values ​​to personnel terminals, the operating parameters of wind turbines, photovoltaic panels, and energy storage devices in the wind-solar hybrid power generation system can be adjusted (e.g., when the predicted power generation increases, the charge-discharge rate of the energy storage device is increased to increase energy storage; when the predicted power generation decreases, the tilt angle of the photovoltaic panel is optimized to improve the utilization rate of sunlight), ensuring that the system operation strategy matches the predicted power generation demand.

[0030] It should be noted that, based on the sequence of predicted values ​​of third-party environmental parameters, the method matches historical sample groups under the current environmental conditions using Euclidean distance, selects matching samples with high similarity and calculates the average power generation, and then combines the weighted average power generation under the same historical conditions to obtain the initial predicted value. Compared with the traditional method that relies solely on third-party predictions or single historical data, this method utilizes the actual power generation patterns of similar historical scenarios and takes into account the overall average level of the operating conditions, effectively reducing the one-sidedness of the initial prediction and laying an accurate foundation for subsequent corrections. The three types of correction coefficients correspond to different key impact dimensions: the environmental deviation correction coefficient is calculated by the deviation rate between the real-time values ​​of key environmental parameters and the third-party predicted values, which can calibrate the errors caused by inaccurate weather forecasts; the equipment health correction coefficient is based on the average health of key equipment parameters, which can reflect the impact of the actual operating status of the equipment on power generation; and the core impact chain correction coefficient combines the dominant factors and deviation patterns of historical fluctuations within a preset period, further aligning with the historical fluctuation characteristics of the system. The three types of coefficients comprehensively cover "environmental prediction deviation, equipment status, and historical fluctuation patterns," solving the problem that traditional systems have a single correction dimension and cannot cope with complex variables. The revised forecast values ​​are sent to personnel terminals for targeted adjustments to the operating parameters of wind turbines, photovoltaic panels, and energy storage devices. This process transforms accurate forecasts into actionable control strategies, solving the problem of "disconnect between forecasting and control, and inability to adapt to demand" in traditional systems. It effectively reduces energy waste or supply shortages caused by inaccurate forecasts and improves the operating efficiency and supply-demand matching of wind-solar hybrid power generation systems.

[0031] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A wind-solar hybrid power generation system based on intelligent monitoring, characterized in that: Comprise: Parameter priority regulation module: real-time monitoring of environmental data and equipment operation data, build environmental characteristics matrix; Calculate the feature correlation between environmental data and power generation efficiency fluctuation, screen environmental working condition scene key dimension reduction dimension; Generate standard matching curve template, establish environment working condition scene matching library to match the current environment working condition scene; Generate scene-key environmental parameter-importance weight mapping table, real-time monitoring of working condition fluctuation index and trigger environment working condition scene update; Coupling factor analysis module: analyze the key equipment parameters corresponding to the environmental working condition scene; Build environmental working condition scene association rule base; Determine the power generation fluctuation of the current working condition scene, if there is, screen abnormal key equipment parameters and abnormal key environmental parameters, determine the initial factor, trace to generate core influence chain report; Predictive dynamic correction module: collect environmental parameter prediction data, match historical sample to calculate initial power generation prediction value; Build environmental bias correction coefficient, equipment health correction coefficient, core influence chain correction coefficient, and conduct comprehensive analysis to get comprehensive correction factor, correct the initial prediction value.

2. The smart monitoring based wind-solar hybrid power generation system as claimed in claim 1, wherein: The specific process of environmental working condition scene key dimension reduction dimension is: Real-time acquisition of environmental data and equipment operation data, set the monitoring period and collect the environmental data of each monitoring time node in the period; According to the time-parameter correspondence, arrange the environmental characteristics matrix with monitoring time node as row and environmental data type as column; Get the environmental characteristics matrix and power generation efficiency fluctuation data of the same period sample under different environmental working condition scenes within a preset time, calculate the feature correlation between each environmental data and efficiency fluctuation; Sort the environmental parameter importance sequence according to the absolute value of the correlation degree, select the top pre-set number of environmental parameters as the key dimension reduction dimension of this scene.

3. The smart monitoring based wind-solar hybrid power generation system as claimed in claim 2, wherein: The specific process of generating standard matching curve template and establishing environment working condition scene matching library to match the current environment working condition scene is: Get the environmental monitoring data of each historical monitoring period of the same period sample, reduce the dimension according to the key dimension reduction dimension of the corresponding environmental working condition scene, get the dimension characteristic value; Draw the dimension characteristic curve with time as horizontal axis and dimension characteristic value as vertical axis, use dynamic time warping algorithm to classify curves with similarity exceeding preset threshold into the same feature group, fit the representative curve of the curve in the group, select the sample with the largest amount as the standard matching curve template of this scene; Store according to the working condition scene-key dimension reduction dimension-standard template mapping, form the environment working condition scene matching library; Traverse the matching library, reduce the dimension of the current environmental characteristics matrix to generate the current characteristic curve, calculate the similarity with the standard template and take the average to get the overall similarity, select the scene with the highest overall similarity as the current monitoring period environment working condition scene.

4. The smart monitoring based wind-solar hybrid power generation system as claimed in claim 3, wherein: The specific process of generating scene-key environmental parameter-importance weight mapping table and real-time monitoring of working condition fluctuation index and triggering environment working condition scene update is: For each environmental working condition scene, the environmental parameters corresponding to each key dimension reduction dimension are recorded as key environmental parameters, and the feature correlation of each key environmental parameter is normalized to obtain the importance weight; The key environmental parameters of each scene and the corresponding importance weight are sorted to form a scene-key environmental parameter-importance weight mapping table, which is substituted into the current working condition scene matching to output the key environmental parameters and importance weight thereof; The key environmental parameter change rate of each monitoring time node after the current working condition is collected in real time, the change rate is multiplied by the corresponding importance weight, and the working condition fluctuation index is obtained by summation; If the index of n consecutive nodes exceeds the preset threshold, the environmental working condition scene updating mechanism is triggered, and the working condition scene matching process is re-executed, and n is a preset continuous monitoring time node quantity threshold.

5. The smart monitoring based wind-solar hybrid power generation system as claimed in claim 4, wherein: The specific process of analyzing the key equipment parameters corresponding to the environmental working condition scene is as follows: For each environmental working condition scene, based on the corresponding scene-key environmental parameter-importance weight mapping table, the key environmental parameters are obtained and sorted into a scene-key environmental parameter list; The historical equipment operation data of the same period sample within a preset time period under the scene is obtained, the co-occurrence frequency of each equipment parameter and power generation efficiency fluctuation is counted, the equipment parameters with a co-occurrence frequency greater than or equal to a preset co-occurrence threshold are selected as candidate equipment parameters, and a scene-candidate equipment parameter list is sorted; The candidate equipment parameters are paired with the key environmental parameters in the scene-key environmental parameter list to form a key environmental parameter-candidate equipment parameter parameter pair, the response correlation degree is calculated to determine the key equipment parameter, and a scene-key environmental parameter-key equipment parameter correspondence table is constructed; The parameter pair association direction and time sequence characteristics are counted, the association rules are verified by substituting the verification set, a scene-key environmental parameter-association rule mapping table is generated, and the mapping tables of all scenes are sorted to form an environmental working condition scene association rule library. The current working condition scene is substituted into the library, and the corresponding scene-key environmental parameter-association rule mapping table is matched.

6. The smart monitoring based wind-solar hybrid power generation system as claimed in claim 5, wherein: The specific process of determining the power generation fluctuation of the current working condition scene is as follows: The real-time power generation efficiency data of the current monitoring period is obtained, the average power generation efficiency in the same period sample within a preset time period under the current working condition scene is taken as the first reference; The actual power generation efficiency mean value of the same parameter combination is queried from the historical data based on the real-time data of the key environmental parameters and the key equipment parameters in the current monitoring period; If the absolute value of the deviation of the real-time power generation efficiency from any reference exceeds the preset fluctuation threshold, and the deviation is maintained for a continuous preset number of monitoring time nodes, it is determined that the power generation fluctuates suddenly; The fluctuation direction, deviation amplitude, and fluctuation start time are recorded to generate a power generation sudden fluctuation determination result.

7. The smart monitoring based wind-solar hybrid power generation system as claimed in claim 6, wherein: The specific process of determining the initial factor and tracing the core influence chain report is as follows: When it is determined that the current environmental working condition scene has a sudden power generation fluctuation, the normal operation range of the preset key equipment parameter is determined, the real-time data of the key equipment parameter within a preset time range before and after the sudden power generation fluctuation start time is extracted, and the key equipment parameter whose real-time data exceeds the corresponding normal operation range is selected as the initial screening abnormal key equipment parameter; Statistical efficiency of each primary screening abnormal key equipment parameters, according to the current power fluctuation direction of sudden reservation corresponding parameters as abnormal key equipment parameters, calculate the normal range of amplitude and generate abnormal key equipment parameters list; The list of parameters into the current environment working condition scene corresponding scene-key environmental parameters-correlation rule mapping table, match the corresponding key environmental parameters, if the key environmental parameter change amplitude standard is marked as abnormal key environmental parameters, record the correlation and generate abnormal key environmental parameters- abnormal key equipment parameters correlation table, according to the time sequence of the parameters to determine the initial impact factor, generate the chain of influence starting point record table; With the initial impact factor to comb the preliminary impact chain, query environmental working condition scene correlation rule library to verify the historical frequency ratio, up to standard is the core impact chain, and generate the core impact chain report.

8. The smart monitoring based wind-solar hybrid power generation system as claimed in claim 7, wherein: The specific process of collecting environmental parameter prediction data and matching historical samples to calculate the initial power generation prediction value is as follows: Collect the environmental parameter prediction value of the preset prediction time corresponding to the current environmental working condition scene output by the third party platform; Call the historical data under the current environmental working condition scene, including the historical environmental parameter sequence-actual power generation corresponding relationship of the scene in the same period sample within the preset time and the historical average power generation under the same environmental working condition scene; Take the predicted environmental parameter sequence formed by the environmental parameter prediction value output by the third party platform as the benchmark, extract the historical environmental parameter sequence of all historical sample groups from the current working condition historical data, calculate the Euclidean distance, and record it as the similarity distance; Select the historical sample group with similarity distance ≤ preset similarity threshold as the matching sample group, calculate its actual power generation average, and record it as the matching sample power average; Distribute the preset weight coefficient to the matching sample power average and the historical average power generation under the same working condition, and then calculate the weighted value to obtain the initial power generation prediction value under the current environmental working condition scene.

9. The smart monitoring based wind-solar hybrid power generation system as claimed in claim 8, wherein: The specific process of analyzing and correcting the initial prediction value is as follows: Call the key environmental parameters and corresponding importance weight of the current environmental working condition scene, real-time collect the real-time value of the key environmental parameters under the scene, and at the same time, call the predicted value of the corresponding period in the third party short-term weather prediction data, get the environmental deviation correction coefficient by calculating the deviation rate of each key environmental parameter; Real-time collect the running data of the key equipment parameters of the current environmental working condition scene, get the equipment health correction coefficient by calculating the health degree of each key equipment parameter and taking the average value; Get the core impact chain report within the preset period, and calculate the core impact chain correction coefficient according to the size relationship of the frequency ratio of the environmental factor and the corresponding actual power generation deviation rate average, the frequency ratio of the equipment factor and the corresponding actual power generation deviation rate average. If there is no core impact chain record within the preset period, the core impact chain correction coefficient is 1; Comprehensive analysis of environmental deviation correction coefficient, equipment health correction coefficient and core impact chain correction coefficient to get comprehensive correction factor; The initial power generation prediction value under the current environmental working condition scene is multiplied by the comprehensive correction factor to obtain a corrected prediction value within a preset prediction time length, and the corrected prediction value is sent to a personnel terminal to be used for regulating and controlling an operation strategy, so that the operation strategy is matched with the predicted power generation demand.

10. The wind-solar complementary power generation method based on intelligent monitoring is applied to the wind-solar complementary power generation system based on intelligent monitoring in any of claims 1-9, comprising: Step one: real-time monitoring of environmental data and equipment operation data, and construction of an environmental feature matrix; The characteristic correlation degree of the environmental data and the power generation efficiency fluctuation is calculated, the key dimension reduction dimension of the environmental working condition scene is screened, the standard matching curve template is generated, the environmental working condition scene matching library is established to match the current environmental working condition scene, the scene-key environmental parameter-importance weight mapping table is generated, the working condition fluctuation index is monitored in real time, and the environmental working condition scene update is triggered; Step two: analyzing the key equipment parameters corresponding to the environmental working condition scene; constructing an environmental working condition scene association rule library; determining the power generation fluctuation of the current working condition scene, if there is, screening the abnormal key equipment parameters and the abnormal key environmental parameters, determining the initial factors, and tracing to generate a core influence chain report; Step three: collecting environmental parameter prediction data, matching historical samples to calculate an initial power generation prediction value; constructing an environmental deviation correction coefficient, an equipment health correction coefficient, and a core influence chain correction coefficient, and performing comprehensive analysis to obtain a comprehensive correction factor to correct the initial prediction value.

Images