Big data analysis method and system based on artificial intelligence

Abstract

The present application relates to a kind of big data analysis method and system based on artificial intelligence, its method includes steps, there are several target wind turbine generators and several target photovoltaic generators, by big data end, all active power data of target wind turbine generators and target photovoltaic generators in historical period and the meteorological environmental data of same period are collected;All active power data and the meteorological environmental data of same period are denoised and labeled, label is used to identify the source of active power data and identify the collection time of active power data and meteorological environmental data;Establish a dynamic analysis model with variable parameters, all active power data and the meteorological environmental data of same period after denoising and adding label are analyzed by the dynamic analysis model with variable parameters, the parameter of dynamic analysis model with variable parameters is dynamically changed until meeting analysis condition in the process of analysis.

Description

Technical Field

[0001] This invention relates to a big data analysis method and system based on artificial intelligence. Background Technology

[0002] In related existing technologies, for example, patent document CN106446494B discloses a wind and solar power prediction technology based on wavelet packet-neural network. This technology first collects meteorological data and corresponding active power data over a three-month period before predicting wind and solar power. Based on this, linear regression calculations are performed on the relevant data to determine variables with strong correlations. In the application of this technology, historical data is collected within a fixed timeframe, such as three months. However, because the impact of climate factors on wind and solar power is complex, especially given its periodicity, simply relying on a fixed period as the historical data collection period cannot accurately reflect the periodicity of the impact. Therefore, subsequent calculations cannot obtain accurate climate influencing factors, i.e., they cannot obtain highly correlated climate influencing factors. Summary of the Invention

[0003] To overcome the shortcomings of existing technologies, this invention provides a big data analysis method and system based on artificial intelligence.

[0004] The technical solution adopted by this invention to solve its technical problem is:

[0005] Artificial intelligence-based big data analysis methods include the following steps:

[0006] Several target wind turbine generators and several target photovoltaic generators are selected. All active power data and concurrent meteorological environmental data of the target wind turbine generators and target photovoltaic generators during the historical period are collected through a big data platform. Noise is removed and tags are added to all active power data and concurrent meteorological environmental data. The tags are used to identify the source of the active power data and the collection time of the active power data and meteorological environmental data. A dynamic analytical model with variable parameters is established. This dynamic analytical model is used to analyze all the noise-removed and tagged active power data and concurrent meteorological environmental data. During the analysis process, the parameters of the dynamic analytical model with variable parameters are dynamically changed until the analysis conditions are met.

[0007] Furthermore, a dynamic analytical model with variable parameters,

[0008] Specifically

[0009] Where t is the quantified value of a specific factor in the meteorological environment data, w is a variable period parameter, P1 is the adjustment benchmark value of the active power of the wind turbine generator set, P2 is the adjustment benchmark value of the active power of the photovoltaic generator set, M1 is the statistical representative value of the wind turbine generator set, and M2 is the statistical representative value of the photovoltaic generator set.

[0010] The value of M1 is equal to the mean of all statistical values, and the value of M2 is also equal to the mean of all statistical values;

[0011] A dynamic analytical model with variable parameters is used to analyze all active power data after denoising and labeling, as well as meteorological environmental data from the same period. Specifically, a certain number of selectable values ​​are first assigned to w, P1, and P2: (w 1, w2,......w n ),(P1 1 P1 2 ,.......P1 n ), (P2 1 P2 2 ,.......P2 n The optional values ​​of w, P1, and P2 are sequentially input into a dynamic analytical model with variable parameters using a traversal method. During each input, the goodness of fit of the dynamic analytical model with variable parameters to t, M1, and M2 is calculated. If the goodness of fit of the dynamic analytical model with variable parameters to t, M1, and M2 meets a threshold, the current set of optional values ​​of w, P1, and P2 is retained and the input order is recorded. This process is repeated several times, lowering the goodness of fit threshold, to iterate the input of optional values ​​of w, P1, and P2 into the dynamic analytical model with variable parameters. Finally, several sets of optional values ​​of w, P1, and P2 that meet the threshold are obtained. The "meeting the analytical condition" refers to the final set of optional values ​​of w, P1, and P2 that meet the threshold, where the threshold is the minimum threshold.

[0012] Furthermore, the goodness of fit of the dynamic analytical model with variable parameters to t, M1, and M2 is calculated as follows:

[0013] Once the possible values ​​of w, P1, and P2 are determined, the following calculation is performed for each input t:

[0014] Then calculate the mean of all Q values, which represents the goodness of fit of the dynamic analytical model with variable parameters to t, M1, and M2.

[0015] Artificial intelligence-based big data analysis systems include:

[0016] The big data terminal is used to collect all active power data of the target wind turbine generator and the target photovoltaic generator during the historical period, as well as the meteorological environment data of the same period. It is also used to denoise and add tags to all active power data and meteorological environment data of the same period. The tags are used to identify the source of active power data and the collection time of active power data and meteorological environment data.

[0017] The analytical model building unit is used to establish a dynamic analytical model with variable parameters. The dynamic analytical model with variable parameters is used to analyze all active power data after denoising and labeling, as well as meteorological and environmental data of the same period. During the analysis process, the parameters of the dynamic analytical model with variable parameters are dynamically changed until the analysis conditions are met.

[0018] Furthermore, the system includes a processor, which is used to execute the functions of the big data terminal and the parsing model building unit.

[0019] Beneficial effects

[0020] This application constructs a dynamic analytical model with variable parameters and analyzes historical active power data and concurrent meteorological environmental data. During the analysis process, the optimal parameters are determined by calculating the goodness of fit. Based on this, the optimal variable period parameter w can be selected. By determining the variable period parameter w, the periodic variation characteristics of climate environmental data and active power data can be accurately reflected. Based on this, the optimal period for collecting historical data can be determined, and more comprehensive and accurate climate factors with stronger influence and correlation on active power can be obtained, providing a precise data foundation for power prediction. Detailed Implementation

[0021] This application discloses an artificial intelligence-based big data analysis method, including the following steps: setting several target wind turbine generators and several target photovoltaic generators; collecting all active power data and concurrent meteorological environmental data of the target wind turbine generators and target photovoltaic generators during historical periods through a big data terminal; denoising and labeling all active power data and concurrent meteorological environmental data, with labels used to identify the source of active power data and the collection time of active power data and meteorological environmental data; establishing a dynamic analysis model with variable parameters; and analyzing all denoised and labeled active power data and concurrent meteorological environmental data through the dynamic analysis model with variable parameters, dynamically changing the parameters of the dynamic analysis model with variable parameters until the analysis conditions are met. This application constructs a dynamic analytical model with variable parameters and analyzes historical active power data and concurrent meteorological environmental data. During the analysis process, the optimal parameters are determined by calculating the goodness of fit. Based on this, the optimal variable period parameter w can be selected. By determining the variable period parameter w, the periodic variation characteristics of climate environmental data and active power data can be accurately reflected. Based on this, the optimal period for collecting historical data can be determined, and more comprehensive and accurate climate factors with stronger influence and correlation on active power can be obtained, providing a precise data foundation for power prediction.

[0022] Preferably, a dynamic analytical model with variable parameters is used.

[0023] Specifically

[0024] Where t is the quantified value of a specific factor in the meteorological environment data, w is a variable period parameter, P1 is the adjustment benchmark value of the active power of the wind turbine generator set, P2 is the adjustment benchmark value of the active power of the photovoltaic generator set, M1 is the statistical representative value of the wind turbine generator set, and M2 is the statistical representative value of the photovoltaic generator set.

[0025] The value of M1 is equal to the mean of all statistical values, and the value of M2 is also equal to the mean of all statistical values;

[0026] A dynamic analytical model with variable parameters is used to analyze all active power data after denoising and labeling, as well as meteorological environmental data from the same period. Specifically, a certain number of selectable values ​​are first assigned to w, P1, and P2: (w 1, w2,......w n ),(P1 1 P1 2 ,.......P1 n ), (P2 1 P2 2 ,.......P2n The optional values ​​of w, P1, and P2 are sequentially input into a dynamic analytical model with variable parameters using a traversal method. During each input, the fit of the dynamic analytical model with variable parameters to t, M1, and M2 is calculated. If the fit of the dynamic analytical model with variable parameters to t, M1, and M2 meets a threshold, the input of a set of optional values ​​of w, P1, and P2 is retained and the input order is recorded. This process is repeated several times, lowering the fit threshold, to iterate the input of optional values ​​of w, P1, and P2 into the dynamic analytical model with variable parameters. Finally, several sets of optional values ​​of w, P1, and P2 that meet the threshold are obtained. The "meeting the analytical condition" refers to the final set of optional values ​​of w, P1, and P2 that meet the threshold, where the threshold is the minimum threshold.

[0027] The goodness of fit of the preferred dynamic analytical model with variable parameters to t, M1, and M2 is calculated as follows:

[0028] Once the possible values ​​of w, P1, and P2 are determined, the following calculation is performed for each input t:

[0029] Then calculate the mean of all Q values, which represents the goodness of fit of the dynamic analytical model with variable parameters to t, M1, and M2.

[0030] In addition, this application discloses a big data analysis system based on artificial intelligence, including,

[0031] The big data terminal is used to collect all active power data of the target wind turbine generator and the target photovoltaic generator during the historical period, as well as the meteorological environment data of the same period. It is also used to denoise and add tags to all active power data and meteorological environment data of the same period. The tags are used to identify the source of active power data and the collection time of active power data and meteorological environment data.

[0032] The analytical model building unit is used to establish a dynamic analytical model with variable parameters. The dynamic analytical model with variable parameters is used to analyze all active power data after denoising and labeling, as well as meteorological and environmental data of the same period. During the analysis process, the parameters of the dynamic analytical model with variable parameters are dynamically changed until the analysis conditions are met.

[0033] The AI-based big data analysis system includes a processor, which is used to execute the functions of the big data terminal and the analysis model building unit.

[0034] As is known from common technical knowledge, this invention can be implemented through other embodiments that do not depart from its spirit or essential characteristics. The disclosed embodiments described above are merely illustrative in all respects and are not the only ones. All modifications within the scope of this invention or its equivalents are included in this invention.

Claims

1. A big data analysis method based on artificial intelligence, characterized by, The steps comprise: setting a plurality of target wind turbine generators and a plurality of target photovoltaic generators; collecting, by a big data terminal, all active power data of the target wind turbine generators and the target photovoltaic generators in a historical period and meteorological environment data in the same period; All the active power data and the meteorological environment data in the same period are denoised and labeled, and the labels are used to identify the source of the active power data and the collection time of the active power data and the meteorological environment data; A dynamic analysis model with variable parameters is established, and all the denoised and labeled active power data and the meteorological environment data in the same period are analyzed by the dynamic analysis model with variable parameters, and the parameters of the dynamic analysis model with variable parameters are dynamically changed during the analysis until the analysis condition is met; The dynamic analysis model with variable parameters, In particular ; Wherein t is the quantitative value of a specific factor in the meteorological environment data, w is the variable period parameter, P1 is the adjustment reference value of the active power of the wind turbine generator, P2 is the adjustment reference value of the active power of the photovoltaic generator, M1 is the statistical representative value of the wind turbine generator, and M2 is the statistical representative value of the photovoltaic generator; The value of M1 is equal to the average of all statistical values, and the value of M2 is also equal to the average of all statistical values; All the active power data and the synchronous meteorological environment data after denoising and adding labels are analyzed by the dynamic analysis model with variable parameters, specifically, first, w, P1, P2 are all assigned with a certain amount of optional values: (w 1, w2,......w n ),(P1 1 ,P1 2 ,.......P1 n ),(P2 1 ,P2 2 ,.......P2 n ),w, P1, P2 are input into the dynamic analysis model with variable parameters in turn according to the traversal method, the fitting degree of the dynamic analysis model with variable parameters to t, M1, M2 is calculated in each input process; if the fitting degree of the dynamic analysis model with variable parameters to t, M1, M2 meets the threshold value, the optional values of w, P1, P2 of this time are retained and the input order is recorded, the fitting degree threshold value of the dynamic analysis model with variable parameters to t, M1, M2 is reduced, w, P1, P2 are input into the dynamic analysis model with variable parameters in turn for several times of traversal, the above operation is iterated, finally several groups of optional values of w, P1, P2 meeting the threshold value are obtained; the threshold value meeting the analysis condition refers to the minimum threshold value among the several groups of optional values of w, P1, P2 meeting the threshold value finally obtained. The fitting degree of the dynamic analysis model with variable parameters to t, M1 and M2 is calculated as follows: When the optional values of w, P1 and P2 are determined, for each input t, the following is calculated: ; then the mean of all Qs is calculated, the mean of all Qs being the degree of fit of the dynamic analytical model with variable parameters to t, M1, M2.

2. The system for operating the artificial intelligence-based big data analysis method of claim 1, characterized by, The system comprises: The big data terminal is used to collect all the active power data of the target wind turbine generators and the target photovoltaic generators in a historical period and meteorological environment data in the same period, and is also used to denoise and label all the active power data and the meteorological environment data in the same period, and the labels are used to identify the source of the active power data and the collection time of the active power data and the meteorological environment data; The analysis model construction unit is used to establish a dynamic analysis model with variable parameters, and all the denoised and labeled active power data and the meteorological environment data in the same period are analyzed by the dynamic analysis model with variable parameters, and the parameters of the dynamic analysis model with variable parameters are dynamically changed during the analysis until the analysis condition is met. 3.The artificial intelligence-based big data analysis system of claim 2, wherein, The system comprises a processor for executing the functions of the big data terminal and the analysis model construction unit.

Images