Photovoltaic power generation optimization scheduling method based on interval prediction
By adopting an optimized scheduling method based on interval prediction in the photovoltaic power generation system, the scheduling difficulty problem caused by uncertainty and randomness in the photovoltaic power generation system is solved, and the system operation efficiency and stability are improved.
Patent Information
- Application Number
- CN202510080386.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-17
- Publication Date
- 2025-04-25
AI Technical Summary
Due to the influence of factors such as light intensity and temperature, photovoltaic power generation systems have great uncertainty and randomness, which makes it difficult to schedule and difficult to meet economic and environmental protection requirements.
The photovoltaic power generation optimization scheduling method based on interval prediction is adopted. By establishing multiple time intervals in a single adjustment cycle, and based on the prediction of the photovoltaic power intervals in each time interval, the sub-control strategies of each time interval are optimized in turn, and the relevant parameters are dynamically updated to improve the operating efficiency and stability of the photovoltaic system.
It improves the operating efficiency of the photovoltaic system and the stability of grid-connected operation, realizes accurate interval prediction of photovoltaic power, and provides reliable data support for photovoltaic power generation scheduling decisions.
Smart Images
Figure CN119891386A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of photovoltaic power generation, and in particular to a photovoltaic power generation optimization scheduling method based on interval prediction. Background Art
[0002] As we all know, photovoltaic power generation is pollution-free and has abundant light resources, which is conducive to the low-carbonization of the power industry. It plays an important role in improving global warming and alleviating the global energy crisis.
[0003] Photovoltaic power generation is affected by factors such as light intensity and temperature, and has great uncertainty and randomness. How to formulate a power dispatch plan under the condition of unstable photovoltaic power generation to meet economic and environmental requirements is a key issue that needs to be solved in photovoltaic power generation dispatch. Summary of the invention
[0004] The purpose of this application is: to solve the above technical problems, this application provides a photovoltaic power generation optimization scheduling method based on interval prediction, aiming to improve the operating efficiency of the photovoltaic system and ensure the stability of the photovoltaic system's grid-connected operation.
[0005] In some embodiments of the present application, multiple time intervals of a single regulation cycle are established, and based on the photovoltaic power interval prediction interval in each time interval, the sub-control strategy of each time interval is optimized in turn to improve the operating efficiency of the photovoltaic system and ensure the stability of the grid-connected operation of the photovoltaic system. At the same time, according to the sub-control strategy of the current time interval, subsequent related parameters are dynamically updated, thereby improving the stability of photovoltaic system regulation.
[0006] In some embodiments of the present application, multiple seed prediction models are trained based on historical parameters to construct an integrated interval prediction model, and based on the prediction error obtained in each cycle, the upper and lower limits of the photovoltaic power prediction interval are set through maximum likelihood estimation, thereby achieving accurate interval prediction of photovoltaic power and providing data support for photovoltaic power generation scheduling decisions.
[0007] In some embodiments of the present application, a photovoltaic power generation optimization scheduling method based on interval prediction is provided, comprising: Establish an interval forecasting model based on historical parameters; Establish a regulation cycle, and generate a photovoltaic power prediction interval within a single regulation cycle based on the interval prediction model and various characteristic index parameters; The battery status parameters are obtained, and a photovoltaic control strategy within a single regulation cycle is generated according to the battery status parameters and the photovoltaic power prediction interval.
[0008] In some embodiments of the present application, when establishing an interval prediction model based on historical parameters, it includes: Preprocess historical parameters; Generate multiple feature sample sets according to the preprocessing results; Generate training data packets and validation data packets based on all feature sample sets: Constructing multiple initial sub-prediction models; Iteratively train each sub-prediction model according to the training data package, and construct a first-level sub-prediction model sequence A according to the iteration results, A=(a1, a2…ai…am), where ai is the i-th first-level sub-prediction model; m is the number of first-level sub-prediction models; Construct a probability distribution function of prediction error based on the first-level sub-prediction model sequence A and the validation set data package; Generate a lower limit θ1 and an upper limit θ2 of the prediction error according to the preset confidence level, where θ1 is a negative value and θ2 is a positive value; An interval prediction model is established based on the first-level sub-prediction model series A and the upper and lower limits of the prediction error.
[0009] In some embodiments of the present application, when generating a photovoltaic power prediction interval within a single regulation cycle, the method includes: Obtain the change curve of each characteristic indicator in the current adjustment cycle; Set multiple time intervals according to the overall change curve; Establish a time interval sequence T of the current adjustment cycle, T = (t1, t2...ti...tn), where ti is the i-th time interval in the current adjustment cycle; n is the number of time intervals; Generate photovoltaic power prediction intervals for each time interval based on all change curves and interval prediction models; The photovoltaic power prediction interval sequence B of the current regulation period is established, B=(b1,b2…bi…bn), where bi is the photovoltaic power prediction interval in the i-th time interval.
[0010] In some embodiments of the present application, when generating photovoltaic power prediction intervals for each time interval, the method includes: According to the time interval sequence T, ti is set as the target time interval in sequence; Obtain the expected reference values of each characteristic indicator in the target time interval, and generate multiple point prediction values of photovoltaic power according to the first-level sub-prediction model sequence A; Establish a point prediction value sequence C, C = (c1, c2...ci...cm), where ci is the point prediction value of the photovoltaic power within the target time interval generated based on the i-th first-level sub-prediction model; Generate a first-level prediction value p; p=( βi*ci) / n; Among them, βi is the weight coefficient of the i-th first-level sub-prediction model; Generate the photovoltaic power interval b of the target time interval, b=(p+θ1,p+θ2); The photovoltaic power intervals of each time interval are generated in sequence.
[0011] In some embodiments of the present application, when generating a photovoltaic control strategy within a single regulation cycle, the method includes: According to the time interval sequence T, select t1 as the time interval to be planned; Obtain battery status parameters according to the start time node t1; Construct photovoltaic-battery control planning model; Substitute the battery status parameters, photovoltaic power interval b1 and user demand parameters in the current regulation cycle into the photovoltaic-battery planning model; Generate the optimal solution interval for time interval t1 and set the sub-control strategy within time interval t1.
[0012] In some embodiments of the present application, when generating a photovoltaic control strategy within a single regulation cycle, the following is also included: Select t2 as the time interval to be planned; According to the sub-control strategy within the time interval t1, the start time node of t1 is corrected to obtain the battery state parameters; Generate the battery status parameter at the start time of time interval t2 according to the correction result; Substitute the battery state parameters at the start of the time interval t2, the photovoltaic power interval b2 and the user demand parameters in the current regulation cycle into the photovoltaic-battery control planning model in turn; Generate the optimal solution interval of time interval t2, and set the sub-control strategy within time interval t2; Generate sub-control strategies for each time interval in sequence; Establish a sub-control strategy sequence D in the current adjustment cycle, D = (d1, d2...di...dn), where di is the sub-control strategy of the i-th time interval in the current adjustment cycle; The photovoltaic control strategy of the current regulation cycle is generated according to the sub-control strategy sequence D.
[0013] In some embodiments of the present application, further comprising: Establish multiple correction time nodes according to the end time nodes of each time interval in the current adjustment cycle; Obtaining actual state parameters of the battery according to the correction time node; Whether to correct the photovoltaic control strategy of the current regulation cycle is determined according to the actual state parameters of the battery.
[0014] In some embodiments of the present application, when determining whether to modify the photovoltaic control strategy in the current regulation cycle, the process includes: Generate the deviation evaluation value g of the storage battery at the current correction time node; g = αi * (fi - f'i) 2 ; where r is the number of state indicators of the storage battery; αi is the influence factor of the i-th state indicator; fi is the reference value set based on the actual state parameters of the storage battery at the current correction time node; f'i is the reference value set based on the expected state parameters of the storage battery at the current correction time node; Judge whether to generate a correction instruction according to the deviation evaluation value g.
[0015] In some embodiments of the present application, when judging whether to generate a correction instruction according to the deviation evaluation value g, it includes: Preset a first deviation evaluation value threshold G1 and a second deviation evaluation value threshold G2, and G1 < G2; If g < G1, no correction instruction is generated at the current correction time node; If G1 ≤ g < G2, a first-level correction instruction is generated at the current correction time node; If g ≥ G2, a second-level correction instruction is generated at the current correction time node.
[0016] Compared with the prior art, the beneficial effects of an optimized scheduling method for photovoltaic power generation based on interval prediction in an embodiment of the present application are as follows: Establish multiple time intervals for a single adjustment period, and based on the interval prediction intervals of photovoltaic power in each time interval, optimize the sub-control strategies for each time interval in turn, improve the operating efficiency of the photovoltaic system, ensure the stability of the grid-connected operation of the photovoltaic system, and at the same time dynamically update subsequent relevant parameters according to the sub-control strategy of the current time interval, thereby improving the stability of the photovoltaic system regulation.
[0017] Train multiple sub-prediction models based on historical parameters, construct an integrated interval prediction model, and set the upper and lower limits of the photovoltaic power prediction interval through maximum likelihood estimation based on the prediction error obtained in each cycle, so as to achieve accurate interval prediction of photovoltaic power and provide data support for photovoltaic power generation scheduling decisions. Brief Description of the Drawings
[0018] Figure 1 is a schematic flowchart of an optimized scheduling method for photovoltaic power generation based on interval prediction in a preferred embodiment of an embodiment of the present application. Detailed Embodiments
[0019] The following combines the drawings and embodiments to further describe the specific embodiments of the present application in detail. The following embodiments are used to illustrate the present application, but are not used to limit the scope of the present application.
[0020] In the description of the present application, it should be understood that the terms "center", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.
[0021] The terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, unless otherwise specified, "plurality" means two or more.
[0022] In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.
[0023] like Figure 1 As shown, a photovoltaic power generation optimization scheduling method based on interval prediction in a preferred embodiment of the present application includes: S101: Establishing an interval prediction model based on historical parameters; S102: Establishing a regulation cycle, generating a photovoltaic power prediction interval within a single regulation cycle according to the interval prediction model and various characteristic index parameters; S103: Obtain battery status parameters, and generate a photovoltaic control strategy within a single regulation cycle according to the battery status parameters and the photovoltaic power prediction interval.
[0024] Specifically, when establishing an interval forecasting model based on historical parameters, it includes: Preprocess historical parameters; Generate multiple feature sample sets according to the preprocessing results; Generate training data packets and validation data packets based on all feature sample sets: Constructing multiple initial sub-prediction models; Iteratively train each sub-prediction model according to the training data package, and construct a first-level sub-prediction model sequence A according to the iteration results, A=(a1, a2…ai…am), where ai is the i-th first-level sub-prediction model; m is the number of first-level sub-prediction models; Construct a probability distribution function of prediction error based on the first-level sub-prediction model sequence A and the validation set data package; Generate a lower limit θ1 and an upper limit θ2 of the prediction error according to the preset confidence level, where θ1 is a negative value and θ2 is a positive value; An interval prediction model is established based on the first-level sub-prediction model series A and the upper and lower limits of the prediction error.
[0025] Specifically, preprocessing of historical data refers to cleaning, filling, and removing outliers from historical data, and then normalizing the processed data.
[0026] Specifically, its characteristic indicators include but are not limited to temperature, relative humidity, total horizontal radiation Parameters that affect photovoltaic power include irradiance, scattered horizontal irradiance, wind direction, and daily rainfall.
[0027] Specifically, multiple initial sub-prediction models are constructed based on various prediction methods such as decision tree, support vector machine, KNN regression, random forest regression and bagging regression.
[0028] Specifically, the prediction errors of each initial sub-model in the verification process are collected by collecting the verification set data packets, and the probability distribution of the prediction error variables is obtained through maximum likelihood estimation, and the upper and lower limits of the photovoltaic power prediction interval are set according to the preset confidence level.
[0029] It can be understood that in the above embodiment, multiple seed prediction models are trained based on historical parameters to construct an integrated interval prediction model, and based on the prediction error obtained in each cycle, the upper and lower limits of the photovoltaic power prediction interval are set through maximum likelihood estimation, thereby achieving accurate interval prediction of photovoltaic power and providing data support for photovoltaic power generation scheduling decisions.
[0030] In a preferred embodiment of the present application, when generating a photovoltaic power prediction interval within a single regulation cycle, it includes: Obtain the change curve of each characteristic indicator in the current adjustment cycle; Set multiple time intervals according to the overall change curve; Establish a time interval sequence T of the current adjustment cycle, T = (t1, t2...ti...tn), where ti is the i-th time interval in the current adjustment cycle; n is the number of time intervals; Generate photovoltaic power prediction intervals for each time interval based on all change curves and interval prediction models; The photovoltaic power prediction interval sequence B of the current regulation period is established, B=(b1,b2…bi…bn), where bi is the photovoltaic power prediction interval in the i-th time interval.
[0031] Specifically, multiple time intervals are established according to the change curves of various characteristic indicators, and each characteristic indicator within a single time interval is maintained in a relatively stable range without drastic changes, thereby improving the accuracy of subsequent photovoltaic power prediction.
[0032] Specifically, when generating the photovoltaic power prediction interval for each time interval, it includes: According to the time interval sequence T, ti is set as the target time interval in sequence; Obtain the expected reference values of each characteristic indicator in the target time interval, and generate multiple point prediction values of photovoltaic power according to the first-level sub-prediction model sequence A; Establish a point prediction value sequence C, C = (c1, c2...ci...cm), where ci is the point prediction value of the photovoltaic power within the target time interval generated based on the i-th first-level sub-prediction model; Generate a first-level prediction value p; p=( βi*ci) / n; Among them, βi is the weight coefficient of the i-th first-level sub-prediction model; Generate the photovoltaic power interval b of the target time interval, b=(p+θ1,p+θ2); The photovoltaic power intervals of each time interval are generated in sequence.
[0033] In a preferred embodiment of the present application, when generating a photovoltaic control strategy within a single regulation cycle, it includes: According to the time interval sequence T, select t1 as the time interval to be planned; Obtain battery status parameters according to the start time node t1; Construct photovoltaic-battery control planning model; Substitute the battery status parameters, photovoltaic power interval b1 and user demand parameters in the current regulation cycle into the photovoltaic-battery planning model; Generate the optimal solution interval for time interval t1 and set the sub-control strategy within time interval t1.
[0034] Specifically, a single sub-control strategy includes the photovoltaic power generation power in the current time interval, the battery charging and discharging status and power.
[0035] Specifically, a photovoltaic-battery planning model is constructed by establishing multiple constraints, and the battery state parameters, photovoltaic power range and user demand parameters are substituted into each constraint to solve the problem, and the sub-control strategy within the current time interval is determined according to the optimal feasible solution.
[0036] Specifically, when generating a photovoltaic control strategy within a single regulation cycle, it also includes: Select t2 as the time interval to be planned; According to the sub-control strategy within the time interval t1, the start time node of t1 is corrected to obtain the battery state parameters; Generate the battery status parameter at the start time of time interval t2 according to the correction result; Substitute the battery state parameters at the start of the time interval t2, the photovoltaic power interval b2 and the user demand parameters in the current regulation cycle into the photovoltaic-battery control planning model in turn; Generate the optimal solution interval of time interval t2, and set the sub-control strategy within time interval t2; Generate sub-control strategies for each time interval in sequence; Establish a sub-control strategy sequence D in the current adjustment cycle, D = (d1, d2...di...dn), where di is the sub-control strategy of the i-th time interval in the current adjustment cycle; The photovoltaic control strategy of the current regulation cycle is generated according to the sub-control strategy sequence D.
[0037] It can be understood that in the above embodiment, by establishing multiple time intervals of a single adjustment cycle, and based on the photovoltaic power interval prediction interval in each time interval, the sub-control strategy of each time interval is optimized in turn, so as to improve the operating efficiency of the photovoltaic system and ensure the stability of the grid-connected operation of the photovoltaic system. At the same time, according to the sub-control strategy of the current time interval, the subsequent related parameters are dynamically updated, thereby improving the stability of the photovoltaic system regulation.
[0038] In a preferred embodiment of the present application, it also includes: Establish multiple correction time nodes according to the end time nodes of each time interval in the current adjustment cycle; Obtaining actual state parameters of the battery according to the correction time node; Whether to correct the photovoltaic control strategy of the current regulation cycle is determined according to the actual state parameters of the battery.
[0039] Specifically, when determining whether to modify the photovoltaic control strategy within the current regulation cycle, it includes: Generate a deviation evaluation value g of the battery at the current correction time node; g= αi*(fi-f'i)2 ; where r is the number of state indicators of the storage battery; αi is the influence factor of the i-th state indicator; fi is the reference value set based on the actual state parameters of the storage battery at the current correction time node; f'i is the reference value set based on the expected state parameters of the storage battery at the current correction time node; Judge whether to generate a correction instruction according to the deviation evaluation value g.
[0040] Specifically, when judging whether to generate a correction instruction according to the deviation evaluation value g, it includes: Preset a first deviation evaluation value threshold G1 and a second deviation evaluation value threshold G2, and G1 < G2; If g < G1, no correction instruction is generated at the current correction time node; If G1 ≤ g < G2, a first-level correction instruction is generated at the current correction time node; If g ≥ G2, a second-level correction instruction is generated at the current correction time node.
[0041] Specifically, the first deviation evaluation value threshold and the second deviation evaluation value threshold can be set according to historical parameters. The first-level correction instruction means obtaining the sub-control strategy in the next time interval and correcting it. The second-level correction instruction means reconstructing the adjustment period according to the current correction time node and regenerating the corresponding control strategy.
[0042] According to the first concept in this application, multiple time intervals of a single adjustment period are established, and based on the predicted intervals of the photovoltaic power intervals in each time interval, the sub-control strategies of each time interval are optimized in turn, so as to improve the operating efficiency of the photovoltaic system, ensure the stability of the photovoltaic system during grid-connected operation, and at the same time, according to the sub-control strategy of the current time interval, dynamically update the subsequent relevant parameters, thereby improving the stability of the photovoltaic system regulation.
[0043] According to the second concept in this application, multiple sub-prediction models are trained based on historical parameters to construct an integrated interval prediction model, and based on the prediction errors obtained in each cycle, the upper and lower limits of the photovoltaic power prediction interval are set through maximum likelihood estimation, so as to achieve accurate interval prediction of the photovoltaic power and provide data support for photovoltaic power generation scheduling decisions.
[0044] The above are only the preferred embodiments of this application. It should be noted that for those of ordinary skill in the art, without departing from the technical principles of this application, several improvements and substitutions can still be made, and these improvements and substitutions should also be regarded as the protection scope of this application.
Claims
1. A photovoltaic power generation optimization scheduling method based on interval prediction, characterized in that: include: Establish an interval forecasting model based on historical parameters; Establish a regulation cycle, and generate a photovoltaic power prediction interval within a single regulation cycle based on the interval prediction model and various characteristic index parameters; The battery status parameters are obtained, and a photovoltaic control strategy within a single regulation cycle is generated according to the battery status parameters and the photovoltaic power prediction interval.
2. The photovoltaic power generation optimization scheduling method based on interval prediction according to claim 1, characterized in that: When establishing an interval forecasting model based on historical parameters, it includes: Preprocess historical parameters; Generate multiple feature sample sets according to the preprocessing results; Generate training data packets and validation data packets based on all feature sample sets: Constructing multiple initial sub-prediction models; Iteratively train each sub-prediction model according to the training data package, and construct a first-level sub-prediction model sequence A according to the iteration results, A=(a1, a2…ai…am), where ai is the i-th first-level sub-prediction model; m is the number of first-level sub-prediction models; Construct a probability distribution function of prediction error based on the first-level sub-prediction model sequence A and the validation set data package; Generate a lower limit θ1 and an upper limit θ2 of the prediction error according to the preset confidence level, where θ1 is a negative value and θ2 is a positive value; An interval prediction model is established based on the first-level sub-prediction model series A and the upper and lower limits of the prediction error.
3. The photovoltaic power generation optimization scheduling method based on interval prediction according to claim 2 is characterized in that: When generating the PV power forecast interval within a single regulation cycle, it includes: Obtain the change curve of each characteristic indicator in the current adjustment cycle; Set multiple time intervals according to the overall change curve; Establish a time interval sequence T of the current adjustment cycle, T = (t1, t2...ti...tn), where ti is the i-th time interval in the current adjustment cycle; n is the number of time intervals; Generate photovoltaic power prediction intervals for each time interval based on all change curves and interval prediction models; The photovoltaic power prediction interval sequence B of the current regulation period is established, B=(b1,b2…bi…bn), where bi is the photovoltaic power prediction interval in the i-th time interval.
4. The photovoltaic power generation optimization scheduling method based on interval prediction according to claim 3 is characterized in that: When generating the PV power forecast interval for each time interval, it includes: According to the time interval sequence T, ti is set as the target time interval in sequence; Obtain the expected reference values of each characteristic indicator in the target time interval, and generate multiple point prediction values of photovoltaic power according to the first-level sub-prediction model sequence A; Establish a point prediction value sequence C, C = (c1, c2...ci...cm), where ci is the point prediction value of the photovoltaic power within the target time interval generated based on the i-th first-level sub-prediction model; Generate a first-level prediction value p; p=( βi*ci) / n; Among them, βi is the weight coefficient of the i-th first-level sub-prediction model; Generate the photovoltaic power interval b of the target time interval, b=(p+θ1,p+θ2); The photovoltaic power intervals of each time interval are generated in sequence.
5. The photovoltaic power generation optimization scheduling method based on interval prediction according to claim 4 is characterized in that: When generating a PV control strategy within a single regulation cycle, it includes: According to the time interval sequence T, select t1 as the time interval to be planned; Obtain battery status parameters according to the start time node t1; Construct photovoltaic-battery control planning model; Substitute the battery status parameters, photovoltaic power interval b1 and user demand parameters in the current regulation cycle into the photovoltaic-battery planning model; Generate the optimal solution interval for time interval t1 and set the sub-control strategy within time interval t1.
6. The photovoltaic power generation optimization scheduling method based on interval prediction according to claim 5, characterized in that: When generating the photovoltaic control strategy within a single adjustment period, it also includes: Select t2 as the time interval to be planned; Modify the start time node of t1 according to the sub-control strategy within the time interval t1 to obtain the battery state parameters; Generate the battery state parameters at the start moment of the t2 time interval according to the modification result; Successively substitute the battery state parameters at the start moment of the t2 time interval, the photovoltaic power interval b2, and the user demand parameters within the current adjustment period into the photovoltaic-battery control planning model; Generate the optimal solution interval for the time interval t2 and set the sub-control strategy within the time interval t2; Successively generate the sub-control strategies for each time interval; Establish a sequence D of sub-control strategies within the current adjustment period, D = (d1, d2... di... dn), where di is the sub-control strategy for the i-th time interval within the adjustment period; Generate the photovoltaic control strategy for the current adjustment period according to the sequence D of sub-control strategies.
7. The photovoltaic power generation optimization scheduling method based on interval prediction according to claim 6 is characterized in that: It also includes: Establish multiple correction time nodes according to the end time nodes of each time interval within the current adjustment period; Obtain the actual state parameters of the battery according to the correction time nodes; Judge whether to correct the photovoltaic control strategy for the current adjustment period according to the actual state parameters of the battery.
8. The photovoltaic power generation optimization scheduling method based on interval prediction according to claim 7, characterized in that: When judging whether to correct the photovoltaic control strategy within the current adjustment period, it includes: Generate the deviation evaluation value g of the battery at the current correction time node; g= αi*(fi-f'i) 2 ; Where r is the number of state indicators of the battery; αi is the influence factor of the i-th state indicator; fi is the reference value set based on the actual state parameters of the battery at the current correction time node; f'i is the reference value set based on the expected state parameters of the battery at the current correction time node; Judge whether to generate a correction instruction according to the deviation evaluation value g.
9. The photovoltaic power generation optimization scheduling method based on interval prediction according to claim 8, characterized in that: When judging whether to generate a correction instruction according to the deviation evaluation value g, it includes: Preset the first deviation evaluation value threshold G1 and the second deviation evaluation value threshold G2, and G1 < G2; If g < G1, no correction instruction is generated at the current correction time node; If G1 ≤ g < G2, a first-level correction instruction is generated at the current correction time node; If g ≥ G2, a second-level correction instruction is generated at the current correction time node.