Photovoltaic power station short power prediction and supervision system based on AI technology

Abstract

The present invention discloses a photovoltaic power station short power prediction and supervision system based on AI technology. The present invention relates to the technical field of photovoltaic power stations and solves the problem that the original determination method is still slightly insufficient in accuracy. The present invention is based on the associated data of the same trend change segment, locks the temperature change curve associated with the corresponding trend segment, and combines several groups of temperature change curves by confirming them one by one to lock the overall temperature change curve, and then locks the specific abnormal temperature based on the set standard temperature, and then locks the associated impedance change based on the relevant numerical value of the abnormal temperature. When predicting power generation, the impedance factor is taken into account, the predicted standard power generation is locked, and thus the corresponding power generation is locked. This can not only make the predicted power generation more accurate, but also display it in the form of intervals, and simultaneously show its coverage, so as to achieve better numerical accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power stations, and in particular to a photovoltaic power station short power prediction and supervision system based on AI technology. Background Art

[0002] The core component of a photovoltaic power station consists of multiple solar cells. Solar cells convert sunlight into direct current through the photoelectric effect.

[0003] Common types of solar panels include monocrystalline silicon, polycrystalline silicon and thin-film solar panels. Monocrystalline silicon panels have higher efficiency, but the cost is relatively high; polycrystalline silicon panels have lower cost and slightly lower efficiency than monocrystalline silicon; thin-film solar panels are flexible and lightweight, and are suitable for some special occasions.

[0004] The application with publication number CN105404937A discloses a method and system for short-term power prediction of a photovoltaic power station, which obtains historical meteorological data recorded at the location of the photovoltaic power station within a time period, establishes a prediction model of light intensity and component backplane temperature; uses the ratio of power value to light intensity in the historical data as a correction coefficient, and constructs a correction coefficient function with backplane temperature as an independent variable; obtains the weather forecast value of the future time period at the location of the photovoltaic power station, substitutes the weather forecast value into the prediction model of light intensity and component backplane temperature respectively to obtain light intensity prediction value and backplane temperature prediction value; obtains correction coefficient prediction value through backplane temperature prediction value according to the correction coefficient function; and calculates power prediction value according to the ratio of power value to light intensity as a correction coefficient. Therefore, the method and system for short-term power prediction of a photovoltaic power station can accurately predict the short-term power of a photovoltaic power station.

[0005] When predicting the actual short-term power of a photovoltaic power station, the light intensity is generally determined based on specific weather conditions, and then the related power generation is locked based on the determined light intensity. In the original determination method, the temperature changes caused by the light intensity are not fully considered. The changing temperature will cause the impedance of the photovoltaic components to change, thereby affecting the corresponding power generation. Therefore, the original determination method is still slightly insufficient in accuracy and urgently needs to be improved. Summary of the invention

[0006] In view of the deficiencies in the prior art, the present invention provides a photovoltaic power station short power prediction and supervision system based on AI technology, which solves the problem that the original determination method is still slightly insufficient in accuracy.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: A photovoltaic power station short power prediction and supervision system based on AI technology, comprising:

[0008] The associated data acquisition terminal acquires the solar radiation intensity data associated with a set period from the weather data, and transmits the acquired solar radiation intensity data to the temperature data analysis terminal;

[0009] The temperature data analysis end generates a solar radiation intensity change curve belonging to the set period based on the solar radiation intensity data associated with the set period, and then confirms the temperature change data curve belonging to the set period based on the photovoltaic panel temperature data at the current moment. The specific method is as follows:

[0010] Based on the different solar radiation intensity data corresponding to different moments in the set period, a solar radiation intensity curve is generated. The horizontal coordinate axis of this curve is the timeline of the set period, and the vertical coordinate axis is the radiation intensity. The initial moment of the timeline is the current moment, and the temperature data of the photovoltaic panel corresponding to the current moment is calibrated as Wd;

[0011] Lock the same-frequency trend segment from the solar radiation intensity curve: calibrate the radiation intensity corresponding to the initial point in the curve as Q1, then determine the radiation intensity Q2 corresponding to the subsequent adjacent points, determine the intensity difference Qz = Q2-Q1, and determine a set of selection ranges based on the intensity difference Qz: Qz±Y1, where Y1 is a preset value, confirm the intensity difference between the adjacent points that appear in succession, calibrate the intensity difference belonging to this selection range as the same type of value, and divide the corresponding partial segments between the points associated with the same type of value and the initial point into the same same-frequency trend segment;

[0012] After the first group of same-frequency trend segments are confirmed, the end point of this same-frequency trend segment is used as the second group of initial points, and the subsequent same-frequency trend segments are confirmed in turn in the same way;

[0013] Based on the selection range associated with each group of same-frequency trend segments, the middle value of the selection range is taken as the characteristic value T of the same-frequency trend segment. i , where i represents different same-frequency trend segments;

[0014] For the first group of same-frequency trend segments: Based on the calibrated temperature data Wd, the characteristic value (T i ×C1) as the change trend, confirm the temperature change data curve belonging to the first group of same-frequency trend segments, where C1 is a preset fixed coefficient factor;

[0015] For the subsequent same-frequency trend segment: confirm the temperature data corresponding to the initial point of the same-frequency trend segment according to the confirmed temperature change data curve, and then use the characteristic value (T i ×C1) as the change trend, confirm the temperature change data curve belonging to this same frequency trend segment;

[0016] The temperature variation data curves associated with each group of same-frequency trend segments are combined to obtain the temperature variation data curve belonging to the set period;

[0017] The impedance data analysis end, based on the temperature change data curve determined in this set period, divides the temperature change data curve into standard divisions according to the set standard temperature, determines the abnormal temperature, and locks the impedance change caused by the abnormal temperature to determine its impedance range. The specific method is as follows:

[0018] Based on the determined temperature change data curve and the set standard temperature Bz, the standard temperature is the preset temperature;

[0019] The temperature data in the temperature change data curve that is higher than the standard temperature is calibrated as abnormal temperature, and the abnormal temperature corresponding to each different point is calibrated as YC k , where k represents different points;

[0020] Identify abnormal temperature YC from historical completion data k The impedance parameter associated with it is then confirmed from the impedance parameter, and its minimum and maximum values ​​are locked to this YC k The impedance range of

[0021] Then, based on the abnormal time corresponding to the abnormal temperature, the impedance interval corresponding to the abnormal time is transmitted to the associated power determination terminal;

[0022] The associated power determination end confirms the specific power generation generated at the corresponding time based on the solar radiation intensity data associated with different times within the set period, and then adjusts the power generation at the abnormal time based on the impedance interval corresponding to the abnormal time, so as to lock the total power generation, and then locks its power generation power based on the total power generation and the total radiation intensity; the specific method is as follows:

[0023] The different radiation intensity data corresponding to different moments are calibrated as FSq, where q represents different moments, and the specific power generation FDq corresponding to the corresponding moment is obtained by using FDq×C3=FDq, where C3 is the set power generation factor;

[0024] Determine the standard impedance Z of the photovoltaic panel, and then based on the impedance interval [Jmin, Jmax] corresponding to the abnormal moment, based on the percentage increase of different impedance values ​​in the impedance interval [Jmin, Jmax] compared to Z, reduce the FDq corresponding to the abnormal moment by the corresponding percentage, perform relevant processing on the FDq associated with the abnormal moment, and confirm the adjusted power generation value interval [FDqmin, FDqmax] corresponding to the abnormal moment in turn;

[0025] Based on the specific power generation FDq confirmed at different times and the power generation value interval [FDqmin, FDqmax] adjusted at the abnormal time, the power generation value interval [FFmin, FFmax] belonging to this set period is determined by summing them up;

[0026] Based on this power generation value range and the total radiation intensity ZF confirmed within this set period, the minimum power generation power GLmin is determined by: GLmin=FFmin÷ZF, and the maximum power generation power GLmax is determined by GLmax=FFmax÷ZF. The locked power generation power range [GLmin, GLmax] is transmitted to the verification center.

[0027] Preferably, it also includes:

[0028] The verification center locks the corresponding actual power generation based on the actual power generation generated in the corresponding set period, and then verifies the actual power generation with the power generation range based on the determined power generation range. The specific method is as follows:

[0029] The actual power generation generated in the set period is calibrated as ZL, and the actual power generation GG is determined by ZL÷ZF=GG;

[0030] Compare the actual generated power GG with the generated power range [GLmin, GLmax]:

[0031] If GG∈[GLmin, GLmax], no processing is performed;

[0032] If GG<GLmin, an abnormal power generation signal is generated and displayed through the signal terminal;

[0033] If GG>GLmax, a prediction deviation signal is generated and displayed through the signal terminal.

[0034] The present invention provides a photovoltaic power station short power prediction and supervision system based on AI technology. Compared with the prior art, it has the following beneficial effects:

[0035] The present invention confirms the corresponding radiation data through the generated solar radiation intensity data, and then locks the related same trend change segment based on the numerical change of the corresponding radiation data, and locks the temperature change curve associated with the corresponding trend segment based on the associated data of the same trend change segment. By confirming one by one, several groups of temperature change curves are combined to lock the overall temperature change curve, and then based on the set standard temperature, the specific abnormal temperature is locked, and then based on the related numerical value of the abnormal temperature, the related impedance change is locked. When predicting the power generation, the impedance factor is taken into account, the predicted standard power generation is locked, and thus the corresponding power generation is locked. This can not only make the predicted power generation more accurate, but also display it in the form of intervals, and simultaneously show its coverage, so as to achieve better numerical accuracy.

[0036] For the determined power generation range, by identifying the actual power generation, evaluating the accuracy of the power generation range prediction, and simultaneously generating corresponding correlation signals for display, external relevant personnel can correlate the signals and take timely response measures, thus achieving better overall supervision effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of the principle framework of the present invention;

[0038] Figure 2 This is a schematic diagram of abnormal signal generation according to the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] First embodiment

[0041] See also Figure 1 , the present application provides a photovoltaic power station short power prediction and supervision system based on AI technology, including an associated data acquisition end, a temperature data analysis end, an impedance data analysis end, an associated power determination end, a verification center and a signal end, wherein the associated data acquisition end is electrically connected to the temperature data analysis end or the associated power determination end input node, the temperature data analysis end is electrically connected to the impedance data analysis end input node, and the impedance data analysis end is electrically connected to the associated power determination end input node, and the associated power determination end, the verification center and the signal end are electrically connected in sequence from the output node to the input node;

[0042] Among them, the associated data acquisition end obtains the solar radiation intensity data associated with the set period from the weather data, and transmits the obtained solar radiation intensity data to the temperature data analysis end. The solar radiation intensity data can be directly obtained from the weather data. The set period is a specified time period following the current moment, which is a set of preset periods and is prepared in advance by relevant personnel based on experience;

[0043] The temperature data analysis end generates a solar radiation intensity change curve belonging to the set period based on the solar radiation intensity data associated with the set period, and then confirms the temperature change data curve belonging to the set period based on the photovoltaic panel temperature data at the current moment. Specifically, different solar radiation intensities will cause the corresponding photovoltaic panels to produce corresponding temperature changes, and the temperature changes are related to the temperature and radiation intensity of the original photovoltaic panels. Therefore, in order to analyze the subsequent specific power generation, it is necessary to fully confirm the specific impact of the temperature data on the internal impedance of the photovoltaic panel. The specific method of confirming the temperature change data curve is:

[0044] Based on the different solar radiation intensity data corresponding to different moments in the set period, a solar radiation intensity curve is generated. The horizontal coordinate axis of this curve is the timeline of the set period, and the vertical coordinate axis is the radiation intensity. The initial moment of the timeline is the current moment, and the temperature data of the photovoltaic panel corresponding to the current moment is calibrated as Wd;

[0045] Lock the same-frequency trend segment from the solar radiation intensity curve: calibrate the radiation intensity corresponding to the initial point in this curve as Q1, and then determine the radiation intensity Q2 corresponding to the subsequent adjacent points, determine the intensity difference Qz = Q2-Q1, and determine a set of selection ranges based on this intensity difference Qz: Qz±Y1, where Y1 is a preset value, and its specific value is determined by the operator based on experience. Confirm the intensity difference between the subsequent adjacent points, calibrate the intensity difference belonging to this selection range as the same type of value, and divide the corresponding partial segments between the points associated with the same type of value and the initial point into the same same-frequency trend segment. For example: the radiation intensity corresponding to the initial point of the proposed radiation intensity curve is 50, and the radiation intensities corresponding to the subsequent consecutive points are 51, 52, 53, 55, 57, 58, 60 respectively. , 59, 52, so the confirmed intensity difference Qz=51-50=1, among which Y1 takes the value of 1, so the confirmed selection range is: [0, 2], and the subsequent differences are: 1, 1, 2, 2, 1, 2, -1, -7, so the points associated with the first group of same-frequency trend segments are: 50, 51, 52, 53, 55, 57, 58, 60, 59, and the subsequent point 52 does not belong to this same-frequency trend segment. The points associated with group 59 are selected as the initial points for the second group of same-frequency trend segments to be confirmed. The confirmed difference is -7, so the confirmed selection range is [-8, -6]. Based on this selection range, the subsequent same-frequency trend segments are selected one by one. If there is no corresponding point, the numerical segment between 59 and 52 can be directly used as the corresponding same-frequency trend segment;

[0046] After the first group of same-frequency trend segments are confirmed, the end point of this same-frequency trend segment is used as the second group of initial points, and the subsequent same-frequency trend segments are confirmed in turn in the same way;

[0047] Based on the selection range associated with each group of same-frequency trend segments, the middle value of the selection range is taken as the characteristic value T of the same-frequency trend segment. i , where i represents different same-frequency trend segments;

[0048] For the first group of same-frequency trend segments: Based on the calibrated temperature data Wd (that is, the temperature corresponding to the initial moment, that is, the current moment), the characteristic value (T i ×C1) as the change trend, confirm the temperature change data curve belonging to this first group of same-frequency trend segments, where C1 is a preset fixed coefficient factor, and its specific value is determined by the operator based on experience. Specifically, after its change trend is confirmed, the proposed change trend is 1, and its temperature data is 20. This same-frequency trend segment is associated with five amplitude data points, so the change of its temperature data is 20-21-22-23-24, then this temperature data change curve corresponds to this same-frequency trend segment, and the temperature data 24 corresponding to the end point of this temperature data change curve is the temperature data corresponding to the initial point of the next group of same-frequency trend segments, and then according to the change trend corresponding to the next group of same-frequency trend segments, the temperature change curves of subsequent same-frequency trend segments can be confirmed in turn;

[0049] For the subsequent same-frequency trend segment: confirm the temperature data corresponding to the initial point of the same-frequency trend segment according to the confirmed temperature change data curve (the temperature data of the initial point is the temperature corresponding to the end point of the previous set of temperature change data curve), and then use the characteristic value (T i ×C1) as the change trend, confirm the temperature change data curve belonging to this same frequency trend segment;

[0050] The temperature variation data curves associated with each group of same-frequency trend segments are combined to obtain the temperature variation data curve belonging to the set period.

[0051] Second embodiment

[0052] After the temperature data of this set period is confirmed, the associated changes of the impedance data can be determined based on the temperature data of this set period, so as to confirm the change of the impedance data, and determine the difference change of the power generation based on the specific change of the impedance data;

[0053] The impedance data analysis end, based on the temperature change data curve determined in the set period, divides the temperature change data curve into standard divisions by the set standard temperature, determines the abnormal temperature, and locks the impedance change caused by the abnormal temperature to determine its impedance range, wherein the specific method of determination is:

[0054] Based on the determined temperature change data curve and the set standard temperature Bz, the standard temperature is a preset temperature, and a temperature higher than this standard temperature will cause the internal impedance of the photovoltaic panel to change;

[0055] The temperature data in the temperature change data curve that is higher than the standard temperature is calibrated as abnormal temperature, and the abnormal temperature corresponding to each different point is calibrated as YC k , where k represents different points;

[0056] Identify abnormal temperature YC from historical completion data k The impedance parameter associated with it is then confirmed from the impedance parameter, and its minimum and maximum values ​​are locked to this YC k The impedance range of

[0057] Based on the abnormal time corresponding to the abnormal temperature, the impedance interval corresponding to the abnormal time is transmitted to the associated power determination terminal.

[0058] The associated power determination end confirms the specific power generation generated at the corresponding time based on the solar radiation intensity data associated with different times within the set period, and then adjusts the power generation at the abnormal time based on the impedance interval corresponding to the abnormal time, so as to lock the total power generation, and then locks its power generation power based on the total power generation and the total radiation intensity. The specific method of locking is:

[0059] The different radiation intensity data corresponding to different moments are calibrated as FSq, where q represents different moments, and the specific power generation FDq corresponding to the corresponding moment is obtained by using FDq×C3=FDq, where C3 is the set power generation factor, which is prepared in advance by relevant operators based on experience;

[0060] Determine the standard impedance Z of the photovoltaic panel, and then based on the impedance interval [Jmin, Jmax] corresponding to the abnormal moment, based on the percentage increase of different impedance values ​​in the impedance interval [Jmin, Jmax] compared to Z, reduce the FDq corresponding to the abnormal moment by the corresponding percentage, perform relevant processing on the FDq associated with the abnormal moment, and confirm the adjusted power generation value interval [FDqmin, FDqmax] corresponding to the abnormal moment in turn;

[0061] Based on the specific power generation FDq confirmed at different times (different times here do not include abnormal times) and the power generation value interval [FDqmin, FDqmax] adjusted at abnormal times, they are summed to determine the power generation value interval [FFmin, FFmax] belonging to this set period;

[0062] Based on this power generation value range and the total radiation intensity ZF (the sum of radiation intensity data at several moments) confirmed within this set period, the minimum power generation value GLmin is determined by: GLmin = FFmin ÷ ZF, and the maximum power generation value GLmax is determined by GLmax = FFmax ÷ ZF. The locked power generation range [GLmin, GLmax] is transmitted to the verification center;

[0063] Specifically, when the corresponding moment is abnormal, it means that the specific power generation situation will be affected, thereby causing the corresponding power generation to be reduced. Because the impedance is the corresponding impedance interval, the corresponding power generation value interval is associated. After the specific change considerations, the power generation value can be locked in a more accurate power generation, which can not only make the predicted power generation more accurate, but also display it in the form of an interval, and simultaneously show its coverage, which can achieve better numerical accuracy.

[0064] Third embodiment

[0065] The verification center locks the corresponding actual power generation based on the actual power generation generated in the corresponding set period, and then verifies the actual power generation with the power generation interval based on the determined power generation interval. The signal end generates and displays a verification signal based on the verification result.

[0066] Among them, combined Figure 2 The specific method of verification is:

[0067] The actual power generation generated in the set period is calibrated as ZL, and the actual power generation GG is determined by ZL÷ZF=GG;

[0068] Compare the actual generated power GG with the generated power range [GLmin, GLmax]:

[0069] If GG∈[GLmin, GLmax], no processing is performed;

[0070] If GG<GLmin, an abnormal power generation signal is generated and displayed through the signal terminal for relevant personnel to view;

[0071] If GG>GLmax, a prediction deviation signal is generated through the signal terminal and displayed for relevant personnel to view. Based on the displayed prediction deviation signal, relevant personnel make numerical adjustments to the coefficient factors used in the system.

[0072] Fourth embodiment

[0073] The specific implementation process of this embodiment includes all the implementation processes of the above three groups of embodiments.

[0074] Some of the data in the above formulas are dimensionless and numerically calculated. Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0075] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. The short power prediction and supervision system of photovoltaic power station based on AI technology is characterized by: include: The associated data acquisition terminal acquires the solar radiation intensity data associated with a set period from the weather data, and transmits the acquired solar radiation intensity data to the temperature data analysis terminal; The temperature data analysis end generates a solar radiation intensity change curve belonging to the set period based on the solar radiation intensity data associated with the set period, and then confirms the temperature change data curve belonging to the set period based on the photovoltaic panel temperature data at the current moment, wherein the temperature change data curve is used to represent the correlation change relationship between the solar radiation intensity and the photovoltaic panel temperature; The impedance data analysis end, based on the temperature change data curve determined in the set cycle, divides the temperature change data curve into standard divisions according to the set standard temperature, determines the abnormal temperature, and locks the impedance change caused by the abnormal temperature to determine its impedance range; The associated power determination end confirms the specific power generation generated at the corresponding moment based on the solar radiation intensity data associated with different moments within the set period, and then adjusts the power generation at the abnormal moment based on the impedance interval corresponding to the abnormal moment, so as to lock the total power generation, and then locks its power generation power based on the total power generation and the total radiation intensity.

2. The photovoltaic power station short power prediction and supervision system based on AI technology according to claim 1 is characterized in that: The specific method of confirming the temperature change data curve at the temperature data analysis end is: Based on the different solar radiation intensity data corresponding to different moments in the set period, a solar radiation intensity curve is generated. The horizontal coordinate axis of this curve is the timeline of the set period, and the vertical coordinate axis is the radiation intensity. The initial moment of the timeline is the current moment, and the temperature data of the photovoltaic panel corresponding to the current moment is calibrated as Wd; Lock the same-frequency trend segment from the solar radiation intensity curve: calibrate the radiation intensity corresponding to the initial point in the curve as Q1, then determine the radiation intensity Q2 corresponding to the subsequent adjacent points, determine the intensity difference Qz = Q2-Q1, and determine a set of selection ranges based on the intensity difference Qz: Qz±Y1, where Y1 is a preset value, confirm the intensity difference between the adjacent points that appear in succession, calibrate the intensity difference belonging to this selection range as the same type of value, and divide the corresponding partial segments between the points associated with the same type of value and the initial point into the same same-frequency trend segment; After the first group of same-frequency trend segments are confirmed, the end point of this same-frequency trend segment is used as the second group of initial points, and the subsequent same-frequency trend segments are confirmed in turn in the same way; Based on the selection range associated with each group of same-frequency trend segments, the middle value of the selection range is taken as the characteristic value T of the same-frequency trend segment. i , where i represents different same-frequency trend segments; For the first group of same-frequency trend segments: Based on the calibrated temperature data Wd, the characteristic value (T i ×C1) as the change trend, confirm the temperature change data curve belonging to the first group of same-frequency trend segments, where C1 is a preset fixed coefficient factor; For the subsequent same-frequency trend segment: confirm the temperature data corresponding to the initial point of the same-frequency trend segment according to the confirmed temperature change data curve, and then use the characteristic value (T i ×C1) as the change trend, confirm the temperature change data curve belonging to this same frequency trend segment; The temperature variation data curves associated with each group of same-frequency trend segments are combined to obtain the temperature variation data curve belonging to the set period.

3. The photovoltaic power station short power prediction and supervision system based on AI technology according to claim 2 is characterized in that: The impedance data analysis end determines the impedance interval in the following specific manner: Based on the determined temperature change data curve and the set standard temperature Bz, the standard temperature is the preset temperature; The temperature data in the temperature change data curve that is higher than the standard temperature is calibrated as abnormal temperature, and the abnormal temperature corresponding to each different point is calibrated as YC k , where k represents different points; Identify abnormal temperature YC from historical completion data k The impedance parameter associated with it is then confirmed from the impedance parameter, and its minimum and maximum values ​​are locked to this YC k The impedance range of Based on the abnormal time corresponding to the abnormal temperature, the impedance interval corresponding to the abnormal time is transmitted to the associated power determination terminal.

4. The photovoltaic power station short power prediction and supervision system based on AI technology according to claim 3 is characterized in that: The specific method of locking the generated power at the associated power determination end is: The different radiation intensity data corresponding to different moments are calibrated as FSq, where q represents different moments, and the specific power generation FDq corresponding to the corresponding moment is obtained by using FSq×C3=FDq, where C3 is the set power generation factor; Determine the standard impedance Z of the photovoltaic panel, and then based on the impedance interval [Jmin, Jmax] corresponding to the abnormal moment, based on the percentage increase of different impedance values ​​in the impedance interval [Jmin, Jmax] compared to Z, reduce the FDq corresponding to the abnormal moment by the corresponding percentage, perform relevant processing on the FDq associated with the abnormal moment, and confirm the adjusted power generation value interval [FDqmin, FDqmax] corresponding to the abnormal moment in turn; Based on the specific power generation FDq confirmed at different times and the power generation value interval [FDqmin, FDqmax] adjusted at the abnormal time, the power generation value interval [FFmin, FFmax] belonging to this set period is determined by summing them up; Based on this power generation value range and the total radiation intensity ZF confirmed within this set period, the minimum power generation power GLmin is determined by: GLmin=FFmin÷ZF, and the maximum power generation power GLmax is determined by GLmax=FFmax÷ZF. The locked power generation power range [GLmin, GLmax] is transmitted to the verification center.

5. The photovoltaic power station short power prediction and supervision system based on AI technology according to claim 4 is characterized in that: Also includes: The verification center locks the corresponding actual power generation based on the actual power generation generated within the corresponding set period, and then verifies the actual power generation with the power generation range based on the determined power generation range.

6. The photovoltaic power station short power prediction and supervision system based on AI technology according to claim 5 is characterized in that: The specific method for the verification center to verify the actual power generation and the power generation range is: The actual power generation generated in the set period is calibrated as ZL, and the actual power generation GG is determined by ZL÷ZF=GG; Compare the actual generated power GG with the generated power range [GLmin, GLmax]: If GG∈[GLmin, GLmax], no processing is performed.

7. The photovoltaic power station short power prediction and supervision system based on AI technology according to claim 6 is characterized in that: If GG<GLmin, an abnormal power generation signal is generated and displayed through the signal terminal.

8. The photovoltaic power station short power prediction and supervision system based on AI technology according to claim 6 is characterized in that: If GG>GLmax, a prediction deviation signal is generated and displayed through the signal terminal.

Images