AI-based electricity price tracking and sales methods and related equipment

Abstract

This invention relates to the technical field of energy storage devices, providing an AI-based method and equipment for electricity price tracking and sales, applied to photovoltaic energy storage devices. The method includes acquiring user electricity consumption habits, predicting the next day's solar radiation value and the current day's electricity price fluctuation curve, calculating the next day's photovoltaic power generation based on the next day's solar radiation value, and if the power generation exceeds the capacity of the photovoltaic energy storage device, predicting the peak electricity price period based on the price fluctuation curve, and analyzing the actual electricity demand range and actual electricity consumption time periods based on user electricity consumption habits to determine the optimal time for the photovoltaic energy storage device to sell electricity beyond the wall. Through intelligent electricity price prediction and user electricity demand analysis, optimal power allocation and flexible electricity sales for photovoltaic energy storage devices are achieved. This prioritizes user electricity demand while improving the problems of untimely response and missed energy storage discharge opportunities in situations with highly frequent electricity price fluctuations, complex grid response mechanisms, or uncertain user load behavior.

Description

Technical Field

[0001] This application relates to the field of energy storage devices, and in particular to methods and equipment for electricity price tracking and sales based on AI models. Background Technology

[0002] With the rapid development of distributed energy, the proportion of photovoltaic energy storage systems in the power system has been increasing year by year. Especially in industrial, commercial, and residential electricity consumption scenarios, users can achieve self-consumption of energy and grid connection of surplus electricity by installing photovoltaic modules and supporting energy storage equipment, which has become one of the important paths to promote the popularization of green energy. At the same time, with the increasing volatility and uncertainty of new energy output, how to improve the economic utilization efficiency of energy storage while ensuring the reliability of electricity supply has become a core issue in current energy management.

[0003] Among related technical approaches, electricity price tracking and sales methods based on static time strategies or rule engines allocate photovoltaic power generation through preset peak and off-peak periods or fixed thresholds. For example, some existing systems prioritize the use of energy storage devices to meet user loads during morning and evening peak hours based on pre-set electricity price thresholds or price levels, while the remaining power is sold or stored during off-peak periods or under capacity constraints. This improves user-side revenue and reduces electricity costs, achieving preliminary optimization of photovoltaic energy storage resource scheduling.

[0004] While the above technical solutions can achieve basic electricity price tracking and electricity sales control through rule-based time matching and electricity price threshold judgment, they suffer from problems such as untimely response and missed energy storage discharge opportunities when electricity prices fluctuate frequently, grid response mechanisms are complex, or user load behavior is uncertain. Summary of the Invention

[0005] To address the issues of delayed response and missed energy storage discharge opportunities in situations involving frequent electricity price fluctuations, complex grid response mechanisms, or uncertain user load behavior, this application provides an AI-based electricity price tracking and sales method and related equipment.

[0006] This invention provides an AI-based method for electricity price tracking and sales, applied to photovoltaic energy storage devices. The method includes: acquiring user electricity consumption habits and predicting the solar radiation value for the next day and the electricity price fluctuation curve for the current day; calculating the photovoltaic power generation for the next day based on the solar radiation value for the next day; if the photovoltaic power generation for the next day is higher than the capacity of the photovoltaic energy storage device, predicting the highest electricity price period based on the electricity price fluctuation curve for the current day, and analyzing the user's actual electricity demand range and actual electricity consumption period based on the user's electricity consumption habits; comparing the highest electricity price period with the electricity consumption period; if the highest electricity price period and the electricity consumption period are within a preset time interval, reserving enough electricity to meet the electricity demand range for the photovoltaic energy storage device, and selling the remaining electricity from the photovoltaic energy storage device through a wall; if the highest electricity price period and the electricity consumption period are not within a preset time interval, selling all the electricity from the photovoltaic energy storage device through a wall.

[0007] As a preferred embodiment, the steps of acquiring user electricity consumption habits and predicting the solar radiation value for the next day and the electricity price fluctuation curve for the current day include: acquiring historical electricity consumption data of users; performing time series analysis on the historical electricity consumption data of users to extract the frequency of electricity consumption, the magnitude of changes in electricity consumption, and periodic characteristics of users in different time periods to construct user electricity consumption habits; inputting the user electricity consumption habits and historical weather data into a preset time series prediction network to predict the user electricity consumption trend for the next 24 hours and the key meteorological parameters for the next day; inputting the key meteorological parameters into a preset solar radiation prediction sub-model to output the solar radiation value for the next day; using a feature extraction network to extract features from historical electricity price records of the power grid, holiday information, and demand response policies to construct electricity price fluctuation-related features; inputting the electricity price fluctuation-related features into a preset short-term price prediction model to output the electricity price fluctuation curve for the current day.

[0008] As a preferred embodiment, the step of inputting the electricity price fluctuation-related features into a preset short-term price prediction model and outputting the electricity price fluctuation curve for the day includes: inputting the electricity price fluctuation-related features into a preset short-term price prediction model to obtain a predicted hourly electricity price sequence, an electricity price volatility sequence, and a set of abnormal factors for high-frequency price fluctuations; performing a fluctuation window sliding analysis on the hourly electricity price sequence, and extracting local maximum points and slope information based on the analysis results, and using the local maximum points and slope information to construct the price fluctuation intensity and price trend change rate for each time period; weighting and fusing the price fluctuation intensity with the electricity price volatility sequence to generate a price fluctuation weight curve; identifying the cross-time points of the price trend change rate and the set of abnormal factors to generate a set of sensitive periods for price jumps; and performing joint cluster analysis on the price fluctuation weight curve and the set of sensitive periods to obtain the electricity price fluctuation curve for the day.

[0009] As a preferred embodiment, the step of calculating the photovoltaic power generation of the next day based on the solar radiation value of the next day includes: combining the solar radiation value of the next day with the area, tilt angle parameter, and conversion efficiency parameter of the photovoltaic panel equipment using a preset combined calculation formula to obtain the photovoltaic power generation for each time period of the next day; wherein, the combined calculation formula is as follows:

[0010] P t =I t ×A×η×cos(θ);

[0011] Wherein, P t Let I be the photovoltaic power generation at time t; t Let t be the solar radiation value at time t; A be the effective area of ​​the photovoltaic panel; η be the conversion efficiency of the photovoltaic panel; and θ be the tilt angle of the photovoltaic panel.

[0012] As a preferred embodiment, the step of predicting the highest electricity price period based on the daily electricity price fluctuation curve includes: standardizing the daily electricity price fluctuation curve to construct a time gradient curve reflecting the magnitude of electricity price changes; performing peak detection on the time gradient curve to identify candidate periods of rapid electricity price increases; extracting the price growth rate and duration corresponding to the candidate periods, and performing normalized aggregation analysis on the price growth rate and duration to obtain a period-level fluctuation intensity scoring matrix; inputting the fluctuation intensity scoring matrix into a preset price risk identification model to output a confidence-weighted price anomaly risk factor and a price increase rate change spectrum; performing hierarchical threshold analysis on the price anomaly risk factor to identify candidate periods of price surges; performing gradient reconstruction on the candidate periods of price surges and the price increase rate change spectrum to extract the main peak period; comparing the main peak period with a preset electricity price peak threshold to select the highest electricity price period when the price is higher than the average peak level.

[0013] As a preferred embodiment, the step of analyzing the user's actual electricity demand range and actual electricity consumption time period based on the user's electricity consumption habits includes: performing cluster modeling on the user's electricity consumption habits to obtain a typical daily electricity consumption template for the user; inputting the typical daily electricity consumption template for the user into a preset user behavior prediction model to predict the user's electricity consumption at different times in the next 24 hours; and performing segmented statistical analysis and sliding window regression analysis on the electricity consumption at different times to obtain the user's actual electricity demand range and actual electricity consumption time period.

[0014] As a preferred embodiment, the step of performing segmented statistical analysis and sliding window regression analysis on the electricity consumption in different time periods to obtain the user's actual electricity demand range and actual electricity consumption time period includes: performing segmented statistical analysis on the electricity consumption in different time periods to obtain statistical results; extracting the mean and deviation based on the statistical results; and identifying the basic electricity demand range and the high-intensity electricity demand range based on the mean and the deviation; fusing the basic electricity demand range and the high-intensity electricity demand range to generate a continuous non-overlapping electricity intensity curve; performing time-series fitting on the electricity intensity curve; and extracting the peak period, duration, and surge point based on the fitting results; constructing the user's actual electricity demand range based on the peak period, duration, and surge point; performing sliding window regression analysis on the electricity consumption in different time periods to extract stable load areas and sudden load areas; constructing a load change trend curve based on the stable load areas and the sudden load areas; and performing regression fitting and cross-analysis on the load change trend curve using the actual electricity demand range to obtain the user's actual electricity consumption time period.

[0015] This application also provides an AI-based electricity price tracking and sales system for photovoltaic energy storage devices, comprising: an acquisition module for acquiring user electricity consumption habits, predicting the solar radiation value of the next day and the electricity price fluctuation curve of the current day, and calculating the photovoltaic power generation of the next day based on the solar radiation value of the next day; a prediction module for predicting the highest electricity price period based on the electricity price fluctuation curve of the current day if the photovoltaic power generation of the next day is higher than the capacity of the photovoltaic energy storage device, and analyzing the user's actual electricity demand range and actual electricity consumption period according to the user's electricity consumption habits; a comparison module for comparing the highest electricity price period with the electricity consumption period, and if the highest electricity price period and the electricity consumption period are within a preset time interval, reserving enough electricity to meet the electricity demand range of the photovoltaic energy storage device, and selling the remaining electricity of the photovoltaic energy storage device through a wall; and a sales module for selling all the electricity of the photovoltaic energy storage device through a wall if the highest electricity price period and the electricity consumption period are not within a preset time interval.

[0016] This application also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the computer program to implement the AI ​​model-based electricity price tracking and electricity sales method described above.

[0017] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when run by a processor, causes the processor to perform the AI-based electricity price tracking and electricity sales method as described in any of the above-mentioned methods.

[0018] Compared with existing technologies, this application has the following advantages: high flexibility and strong economic benefits. Through intelligent electricity price forecasting and user electricity demand analysis, it achieves optimal power allocation and flexible electricity sales for photovoltaic energy storage devices. When peak electricity price periods are predicted, priority is given to ensuring user electricity demand, while the remaining power is sold during off-peak hours, maximizing the economic benefits of the energy storage devices. By introducing an electricity price fluctuation forecasting model and user electricity consumption habit analysis, it achieves the function of real-time adjustment of power allocation based on electricity price fluctuations and changes in user demand. This not only improves the economic efficiency of user electricity use but also promotes the efficient utilization of photovoltaic energy storage devices, mitigating the problems of untimely response and missed energy storage discharge opportunities in situations with highly frequent electricity price fluctuations, complex grid response mechanisms, or uncertain user load behavior. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] The structures, proportions, sizes, etc., shown in the accompanying drawings of this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed in the specification, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0021] Figure 1 This is a flowchart illustrating the electricity price tracking and electricity sales method based on an AI model provided in an embodiment of the present invention.

[0022] Figure 2 This is a schematic block diagram of the structure of the AI-based electricity price tracking and sales system provided in this embodiment of the invention;

[0023] Figure 3 This is a schematic block diagram of the structure of the electronic device provided in the embodiment of the present invention.

[0024] Explanation of reference numerals in the attached figures:

[0025] 10. Electricity price tracking and sales system based on AI model; 11. Acquisition module; 12. Prediction module; 13. Comparison module; 14. Sales module; 20. Electronic equipment; 21. Memory; 22. Processor. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the described order. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.

[0028] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0029] It should also be further understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0030] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0031] Example 1:

[0032] like Figure 1 As shown, this application provides an AI-based method for electricity price tracking and sales, applied to photovoltaic energy storage devices, including the following steps:

[0033] Step S100: Obtain user electricity consumption habits, predict the solar radiation value of the next day and the electricity price fluctuation curve of the day, and calculate the photovoltaic power generation of the next day based on the solar radiation value of the next day.

[0034] In this step, users' electricity consumption habits are first obtained through historical electricity consumption data. Specifically, time-series predictive networks (such as Long Short-Term Memory networks, LSTM) are used to analyze past electricity consumption data, extracting the frequency, magnitude of electricity consumption changes, and periodic characteristics of users' electricity consumption in different time periods. These characteristics help to build a model of users' electricity consumption habits and predict future electricity consumption trends. Simultaneously, based on historical weather data and combined with meteorological prediction models (e.g., using Support Vector Machines (SVR) for weather data regression), key meteorological parameters for the next day, including temperature, cloud cover, and humidity, are predicted. This data will be input into the solar radiation prediction sub-model.

[0035] For example, when acquiring user electricity consumption habits, analyzing data from the past 30 days identifies periods of higher electricity demand during the morning peak (7:00-9:00) and evening peak (17:00-19:00). This information will influence future electricity demand forecasts. Regarding solar radiation forecasting, combining historical radiation data from the local meteorological bureau with current weather conditions, the model predicts the solar radiation intensity for the following day, ensuring the accuracy of photovoltaic power generation calculations.

[0036] Step S200: If the photovoltaic power generation on the next day is higher than the capacity of the photovoltaic energy storage device, then predict the period of highest electricity price based on the electricity price fluctuation curve of the day, and analyze the user's actual electricity demand range and actual electricity consumption period based on the user's electricity consumption habits.

[0037] In this step, the photovoltaic power generation calculation model calculates the photovoltaic power generation for the next day based on the predicted solar radiation value and the physical parameters of the photovoltaic system (such as the area of ​​the photovoltaic panel and the conversion efficiency). If this power generation exceeds the capacity of the energy storage device, the period of highest electricity price is predicted based on the electricity price fluctuation curve (obtained by combining historical electricity price data from the power grid with user electricity load analysis). To further meet user needs, the actual range of electricity demand is analyzed based on user electricity consumption habits, and the actual electricity consumption time periods are identified.

[0038] For example, when calculating photovoltaic (PV) power generation, the solar radiation value is combined with parameters such as the tilt angle and conversion efficiency of the PV panels to obtain the PV power generation curve for the following day. By analyzing the daily electricity price fluctuation curve, peak electricity price periods can be predicted, such as 10:00 AM to 12:00 PM and 6:00 PM to 8:00 PM, which are the periods with the highest electricity prices. At the same time, based on users' electricity consumption habits (e.g., users use air conditioning from 8:00 AM to 10:00 AM and kitchen appliances during midday), the actual electricity demand periods can be calculated, and the electricity consumption during those periods can be predicted.

[0039] Step S300: Compare the time period with the electricity consumption period. If the time period with the highest electricity price is within the preset time interval, the photovoltaic energy storage device reserves enough electricity to meet the electricity demand range, and sells the remaining electricity of the photovoltaic energy storage device through the wall.

[0040] In this step, the predicted peak electricity price period is compared with the user's actual electricity consumption period. If there is an overlap, and the overlap is within a preset time interval range (e.g., the set time interval is ±30 minutes), the photovoltaic energy storage device reserves enough electricity to meet the user's needs, and the remaining electricity is used for selling electricity through the wall.

[0041] For example, if the predicted peak electricity price period is from 10:00 AM to 12:00 PM, while the user's electricity demand period is from 8:00 AM to 10:00 AM, then the stored energy will be reserved for the user between 10:00 AM and 10:30 AM, and the remaining energy will be sold after 10:30 AM. This method ensures that sufficient electricity is provided to users during the peak electricity price period, while efficiently selling off the remaining energy.

[0042] Step S400: If the time period with the highest electricity price and the time period for electricity consumption are not within the preset time interval, then all the electricity generated by the photovoltaic energy storage device will be sold off through the wall.

[0043] In this step, when the peak electricity price period and the user's electricity consumption period are not within the preset time interval, all the electricity generated by the photovoltaic energy storage device will be sold off through the wall to ensure that the electricity in the energy storage device receives timely economic returns.

[0044] For example, if the predicted peak electricity price period is from 6 PM to 8 PM, while the user's electricity demand period is from 8 AM to 10 AM, the two periods do not overlap. In this case, all the electricity stored in the photovoltaic energy storage device can be sold to obtain the highest market electricity price return.

[0045] In this embodiment, by acquiring users' electricity consumption habits and combining historical weather data with a time-series forecasting network, the system predicts the user's electricity consumption trend for the next 24 hours and the key meteorological parameters for the following day, while also predicting the solar radiation value for the next day. Next, based on the solar radiation value for the following day, the system calculates the photovoltaic power generation for the next day using a photovoltaic power generation calculation model and compares it with the capacity of the photovoltaic energy storage device. If the photovoltaic power generation for the next day exceeds the capacity of the photovoltaic energy storage device, the system predicts the highest electricity price period based on the electricity price fluctuation curve for that day. Subsequently, based on the user's electricity consumption habits, the system analyzes the user's actual electricity demand range and actual electricity consumption time period, and compares the highest electricity price period with the electricity consumption time period. If the highest electricity price period and the electricity consumption time period are within a preset time interval, the photovoltaic energy storage device reserves enough electricity to meet the electricity demand range and sells the remaining electricity through a barrier; if the highest electricity price period and the electricity consumption time period are not within the preset time interval, all electricity from the photovoltaic energy storage device is sold through a barrier, realizing intelligent power allocation and optimized electricity sales for the photovoltaic energy storage device.

[0046] By using AI-based models to predict electricity price fluctuations and analyze user electricity demand, priority can be given to meeting user needs during periods of high electricity prices, avoiding over-discharge during periods of low prices, and maximizing the economic benefits of photovoltaic energy storage equipment. Simultaneously, the power output of photovoltaic energy storage equipment can be flexibly allocated according to electricity price fluctuations, ensuring maximum utilization of the storage equipment during periods of highest electricity prices. This allocation of power resources not only reduces unnecessary electricity waste but also effectively lowers electricity procurement costs, improves the economic efficiency of users' electricity consumption, promotes the efficient use of green energy, and addresses the problems of untimely response and missed energy storage discharge opportunities in situations with highly frequent electricity price fluctuations, complex grid response mechanisms, or uncertain user load behavior.

[0047] Example 2:

[0048] Step S100, which involves acquiring users' electricity consumption habits and predicting the solar radiation value for the next day and the electricity price fluctuation curve for the current day, specifically includes:

[0049] Acquire users' historical electricity consumption data, perform time series analysis on the data, and extract the frequency of electricity consumption, the magnitude of changes in electricity consumption, and periodic characteristics in different time periods to construct users' electricity consumption habits.

[0050] Historical electricity consumption data is analyzed using time-series prediction networks. Specifically, historical electricity consumption data is first obtained from users' smart meters or power management systems. This data includes users' electricity consumption at different time periods (e.g., hourly or minute-by-minute). Then, time series analysis methods (such as Long Short-Term Memory (LSTM) networks or ARIMA models) are used to process this data. This method can effectively capture electricity consumption patterns, such as peak hours, off-peak hours, and periodic changes during holidays. Through these analysis results, characteristics such as users' electricity consumption frequency, electricity consumption fluctuation amplitude, and seasonal electricity consumption changes can be extracted to construct a model of users' electricity consumption habits.

[0051] For example, when acquiring user electricity consumption habits, suppose a user's historical data shows that their peak electricity consumption occurs between 7:00 AM and 9:00 AM and between 5:00 PM and 7:00 PM daily, with generally higher electricity consumption on weekends. By analyzing this data using an LSTM model, these peak periods can be identified, and the user's future electricity demand can be predicted. Based on this analysis, targeted scheduling strategies can be provided.

[0052] Users' electricity consumption habits and historical weather data are input into a preset time-series prediction network to predict the user's electricity consumption trend in the next 24 hours and the key meteorological parameters for the next day, including temperature, cloud cover and humidity. The key meteorological parameters are input into a preset solar radiation prediction sub-model, and combined with historical sunshine data of similar types and current geographical location parameters, the solar radiation value for the next day is output.

[0053] This method combines meteorological data and user electricity consumption data for joint forecasting. Specifically, users' electricity consumption habits, along with historical meteorological data (including temperature, humidity, cloud cover, etc.), are first input into a time-series forecasting network, such as an LSTM model or a GRU (Gated Recurrent Unit) model. This network can predict users' electricity consumption trends for the next 24 hours and, combined with weather changes, predict key meteorological parameters (such as temperature, cloud cover, humidity, etc.) for the next 24 hours. This data is then further fed into a solar radiation forecasting sub-model, which uses historical sunshine data of similar periods and the user's current geographical location parameters to predict the solar radiation value for the following day. In this way, it is possible to comprehensively predict users' electricity demand and photovoltaic power generation potential based on multiple factors.

[0054] For example, in practical applications, assuming the weather in a user's region is cold in winter and meteorological data predicts high cloud cover the following day, this information will be input into the solar radiation prediction model. The model's prediction shows low radiation levels the next day, which will affect the expected amount of photovoltaic power generation. Simultaneously, based on the user's electricity consumption patterns, the model will predict the user's electricity demand during cold weather, further optimizing power dispatch.

[0055] Feature extraction networks are used to extract features from historical electricity price records, holiday information, and demand response policies to construct electricity price fluctuation-related features. These features are then input into a pre-defined short-term price prediction model, which outputs the daily electricity price fluctuation curve.

[0056] Predictions are made by constructing a model of electricity price fluctuation characteristics. Specifically, historical electricity price data is obtained from the power grid management platform or a third-party electricity market data provider. To accurately capture the trend of electricity price fluctuations, features are extracted from electricity price records, holiday information (such as price fluctuations before and after holidays), and demand response policies. Feature extraction methods can employ convolutional neural networks (CNNs) or feature selection methods based on regression analysis to ensure that key factors of electricity price fluctuations are captured. The extracted electricity price fluctuation features are then used as input to a pre-defined short-term price prediction model (such as an LSTM or random forest model). This model outputs the electricity price fluctuation curve for the day based on historical electricity prices and demand response policies, helping to predict future electricity price trends.

[0057] For example, suppose extracted holiday information indicates that electricity demand increases during holidays, and electricity prices fluctuate accordingly. Therefore, by combining historical data with holiday information, it can be predicted that the electricity price the following day will rise significantly in the lead-up to the holiday. Based on this prediction, the discharge schedule of energy storage devices can be adjusted in advance, allowing electricity to be sold when price fluctuations are most severe.

[0058] The step of inputting electricity price fluctuation-related characteristics into a preset short-term price prediction model and outputting the electricity price fluctuation curve for the day specifically includes:

[0059] By inputting the relevant characteristics of electricity price fluctuations into a pre-defined short-term price prediction model, the predicted hourly electricity price sequence, the electricity price volatility sequence, and the set of abnormal factors for high-frequency price fluctuations are obtained.

[0060] The calculation is performed using feature selection and a short-term price forecasting model. Specifically, firstly, key data from historical electricity price records are extracted using feature selection algorithms (such as information gain in decision trees or principal component analysis (PCA) based on correlation). This key data includes hourly price trends, peak-valley differences, and seasonal fluctuations. After feature selection, this data, along with holiday information and demand response policy-related features, are input into a pre-defined short-term price forecasting model (such as an LSTM or random forest model). This model then outputs three results through time series regression analysis:

[0061] Hourly electricity price series: reflects the predicted fluctuation of electricity prices per hour on the same day;

[0062] Electricity price volatility series: quantifying the rate of electricity price fluctuations;

[0063] A set of anomalous factors in high-frequency price fluctuations: key influencing indicators that identify areas of drastic electricity price changes.

[0064] For example, in electricity price forecasting, historical data indicates that electricity prices in a certain region are low before 8:00 AM, and then rapidly peak between 3:00 PM and 5:00 PM. The short-term price forecasting model analyzes this historical data and demand response policies to output the hourly electricity price sequence for the day, showing that the price jump will peak between 4:00 PM and 4:30 PM. Simultaneously, the electricity price volatility sequence reveals a significant increase in the rate of volatility between 3:00 PM and 5:00 PM, while the set of outliers identifies "proximate holidays" and "seasonal changes" as contributing factors to the price jump.

[0065] A sliding window analysis of fluctuations is performed on the hourly electricity price series, and local maximum points and slope information are extracted based on the analysis results. The price fluctuation intensity and price trend change rate of each time period are constructed using the local maximum points and slope information.

[0066] Price fluctuation characteristics are constructed using time window sliding analysis. Specifically, a fixed time window (e.g., 3 hours or 5 hours) is selected, and hourly electricity price sequences are processed using a sliding window. During the sliding window analysis, local maxima and time slopes are extracted from the sequence using local function fitting methods (e.g., calculating the fitting slope using the least squares method). Local maxima represent the peak of price fluctuations, while the slope represents the rate of change in electricity prices. Combining these two features, the following characteristics are constructed for each time period:

[0067] Price volatility intensity: the peak amplitude during a quantitative volatility period;

[0068] Price trend change rate: Indicates the direction and rate of change in electricity price trends.

[0069] For example, in analyzing the predicted electricity price from 3 PM to 6 PM, a local maximum was found at 4 PM, with the slope rising significantly between 2:30 PM and 3:30 PM. By fitting this window, the price volatility intensity was generated, indicating that the electricity price was expected to peak at 4 PM, while the rate of change of the trend showed that the price gradually decreased after 4:30 PM.

[0070] The price volatility intensity is weighted and fused with the electricity price volatility series to generate a price volatility weight curve.

[0071] A weighted fusion method is used to generate a price volatility weight curve; specifically, the intensity of price volatility is weighted by the electricity price volatility series. The weighting method can utilize the Poisson distribution or a dynamic weight adjustment mechanism based on the standard deviation of electricity price changes to ensure accurate quantification of price changes across different time periods. The resulting price volatility weight curve reflects the overall trend of electricity price volatility and the weight values ​​for each time period, which can be used for subsequent analysis of sensitive periods.

[0072] For example, the electricity price weighting during the evening period from 4 PM to 5 PM is assigned a higher weight value (e.g., 0.8 to 0.9) due to its steeper slope, while the weighting for the morning period is only 0.2 to 0.3 due to lower price volatility. The resulting price volatility weighting curve accurately indicates the importance of each volatility period.

[0073] By identifying the cross-time points between the rate of change in price trends and the set of abnormal factors, a set of sensitive periods for price jumps is generated.

[0074] A set of sensitive time periods is generated through cross-identification. Specifically, by combining cross-analysis of the price trend change rate and the set of abnormal factors, time points are filtered and labeled using the Dynamic Time Warping (DTW) algorithm or graph-based clustering methods to identify sensitive time periods of price jumps. The set of sensitive time periods identifies important time points where electricity prices change significantly, which is used for operational scheduling strategies.

[0075] For example, in the daily electricity price forecast, the period from 16:00 to 16:30 was identified as a sensitive period due to the intensity of price fluctuations and the effectiveness of anomalous factors. This period, combined with holiday factors and seasonal variations, has a significant impact on discharge operations, and it is recommended that energy storage devices release electricity during this period to obtain the highest returns.

[0076] By performing joint cluster analysis on the price fluctuation weight curve and the sensitive period set, the electricity price fluctuation curve for the day can be obtained.

[0077] Electricity price fluctuation curves are generated through joint clustering analysis. Specifically, the price fluctuation weight curve and the set of sensitive periods are input into a joint clustering algorithm (such as density-based DBSCAN algorithm or K-Means clustering) to classify and analyze the characteristics of electricity price fluctuations and mine relationships, thereby generating a complete electricity price fluctuation curve for the day. This curve not only reflects the overall trend of electricity prices but also marks important sensitive periods.

[0078] For example, analyzing the daily electricity price weight curve and the set of sensitive time periods yields a fluctuation curve that shows prices are relatively stable from morning to noon, while significant fluctuations occur from 3 PM to 5 PM, with 4:00 PM to 4:30 PM being the critical discharge window. This fluctuation curve can be applied to photovoltaic energy storage scheduling decisions to ensure maximum returns.

[0079] In step S100, the step of calculating the photovoltaic power generation for the next day based on the solar radiation value of the next day specifically includes:

[0080] The solar radiation value of the next day is combined with the area, tilt angle parameters and conversion efficiency parameters of the photovoltaic panel equipment using a preset combined calculation formula to obtain the photovoltaic power generation for each time period of the next day.

[0081] The power generation for each time period is calculated using photovoltaic power generation formulas. Specifically, based on the predicted solar radiation value for the next day and combined with the physical model of photovoltaic power generation, the power generation P of the photovoltaic panel equipment for each time period on the next day is calculated using the following combined formula. t :

[0082] The combined calculation formula is as follows:

[0083] P t =I t ×A×η×cos(θ);

[0084] Among them, P t I represents the power generation of the photovoltaic device at time t (unit: watts (W)). t η represents the solar radiation value at time t (unit: watts per square meter W / m²), output by the solar radiation prediction sub-model; A represents the effective area of ​​the photovoltaic panel (unit: square meters m²), directly obtained from the equipment installation parameters. For example, a typical standard photovoltaic panel has an area of ​​1.6 m²; η represents the conversion efficiency of the photovoltaic panel (dimensionless), directly read from the photovoltaic panel's equipment parameter table. A typical value is 0.18, or 18%; θ represents the incident angle of solar radiation (unit: radians rad). It is obtained analytically by calculating the angle between solar radiation and the normal to the photovoltaic panel surface through geometric angle calculation, combined with the fixed tilt angle of the photovoltaic panel.

[0085] Within each time period, the above formula is used to iteratively calculate all time points t=1,2,…,n to generate the time-of-use power output curve for the entire next day.

[0086] For example, if a photovoltaic panel has an effective area A of 1.6 m², a conversion efficiency η of 0.18, and the solar radiation value It at time t is 500 W / m², and the tilt angle of the photovoltaic panel is 30 degrees, then the angle (θ) between the solar radiation direction and the normal is 20 degrees. The photovoltaic power generation at time t is calculated as follows:

[0087] Substitute the relevant values ​​into the formula: Pt=500 1.6 0.18 cos(20°);

[0088] According to trigonometric functions: cos(20°)≈0.9397;

[0089] Calculation result: Pt=500 1.6 0.18 0.9397 = 135.11 W;

[0090] Therefore, at time t, the output power of the photovoltaic panel is 135.11 watts. Applying the above calculations iteratively to a full-day period (e.g., 24 hours) yields the power generation per hour or at a finer time granularity.

[0091] In step S200, the step of predicting the period of highest electricity price based on the daily electricity price fluctuation curve specifically includes:

[0092] The daily electricity price fluctuation curve is standardized to construct a time gradient curve reflecting the magnitude of electricity price changes. Peak detection is performed on the time gradient curve to identify candidate time periods of rapid electricity price increases.

[0093] A time gradient curve is constructed and candidate time periods are identified using normalization transformation and local peak detection methods. Specifically, firstly, the electricity value at each time point in the obtained daily predicted electricity price fluctuation curve is subjected to Min-Max standardization, compressing the electricity value to the 0,1 range, thereby unifying the scale of electricity price fluctuations across different time periods. Then, based on the standardized electricity price sequence, the rate of change of electricity prices between adjacent time points (i.e., the first-order difference) is calculated to form the time gradient curve. The sliding window detection method is used to scan this gradient curve, and by comparing each time point with its adjacent values, local peak points are extracted as candidate time points for rapid electricity price increases.

[0094] For example, if the standardized electricity price sequence for a certain time period is:

[0095] The sequence of rates of change, calculated using the gradient curve, is [0.25, 0.32, 0.50, 0.78, 0.91, 0.95, 0.96, 0.85].

[0096] [+0.07,+0.18,+0.28,+0.13,+0.04,+0.01, [0.11]. When the sliding window is 3, the period from the third time point (i.e., 0.50) to the sixth time point (0.95) is identified as the segment with significant electricity price growth, and therefore it is identified as a candidate period for rapid electricity price increases.

[0097] Extract the price growth rate and duration corresponding to the candidate time period, and perform normalized aggregation analysis on the price growth rate and duration to obtain a time period-level volatility intensity score matrix. Input the volatility intensity score matrix into the preset price risk identification model, and combine it with holiday factors, power grid load factors and meteorological conditions to output a confidence-weighted price anomaly risk factor and a price rise rate change spectrum.

[0098] Risk indicators and the structure of electricity price increases are obtained through multi-factor modeling and scoring matrix analysis. Specifically, for each candidate time period, the price growth rate (i.e., the difference between the end and beginning of the electricity price in that period) and duration (i.e., the number of time points in that period) are calculated. After normalizing these two indicators, a two-dimensional volatility intensity scoring matrix is ​​constructed. This scoring matrix is ​​then input into a price risk identification model, which can employ a multilayer perceptron (MLP) or XGBoost classification model, and incorporates holiday factors (e.g., whether it is a working day before a holiday), grid load factors (e.g., load forecast curve), and meteorological conditions (e.g., high temperature, low temperature, thunderstorms, etc.) as additional input features. The model outputs two key results: first, a price anomaly risk factor (risk level label or score); second, a price increase rate change graph, used for subsequent trend analysis.

[0099] For example, during a candidate time period, the electricity price rises from 0.40 to 0.90, lasting for 3 hours, with a growth rate of +0.50. After normalization, the growth rate score is 0.83, the duration score is 0.60, and the combined volatility intensity score is 0.72. Combining factors such as the day before the holiday (assuming it is the day before National Day), high-load weather (e.g., high temperature of 35℃), the model predicts that the price anomaly risk factor for this period is "high," and the continuously rising slope of the electricity price during this period is marked on the rate change graph.

[0100] A tiered threshold analysis of price anomaly risk factors was conducted to identify candidate periods of price surges. The candidate periods of price surges were then reconstructed with the price increase rate change graph to extract the peak periods of rapid electricity price increases and those with discharge value.

[0101] High-value peak periods are extracted through tiered threshold discrimination and gradient trend backtracking. Specifically, price anomaly risk factors are tiered (e.g., low, medium, and high), and thresholds are set (e.g., a threshold of 0.7 for high-risk factors). Only time periods exceeding this threshold are retained as candidate periods for price surges. Then, first- and second-order gradient trend analysis is performed on the price increase rate change spectrum to reconstruct the price growth curve and identify the peak regions—the time intervals during which electricity prices show a significant upward trend and reach maximum discharge revenue within a short period.

[0102] For example, a certain time period was assigned a price anomaly risk factor of 0.85, exceeding the high-risk threshold of 0.7; this was reflected in the price rise rate graph as a continuous price increase from 15:30 to 16:30, peaking at 16:15. After gradient reconstruction, 16:00 to 16:30 was identified as the main peak period with discharge value.

[0103] The peak period is compared with the preset peak electricity price threshold to select the period with the highest electricity price when the price is higher than the average peak level.

[0104] The highest electricity price periods are selected by comparing them with a peak electricity price threshold. Specifically, a peak electricity price threshold is set as a reference benchmark (e.g., 1.1 times the average daily highest electricity price over the past month), and the predicted electricity price for all peak periods is compared with this threshold one by one. Only periods with predicted electricity prices higher than this threshold are retained as the highest electricity price periods. The final selected periods are used to guide the power dispatch and priority electricity sales strategies for photovoltaic energy storage devices.

[0105] For example, if the peak electricity price threshold is set at 0.85, and the predicted electricity price reaches 0.92 between 16:00 and 16:30, exceeding the threshold, then this time period is ultimately confirmed as the period with the highest electricity price and is used as the priority period for the photovoltaic energy storage equipment to discharge.

[0106] Step S200, which involves analyzing the user's actual electricity demand range and actual electricity consumption time period based on the user's electricity usage habits, specifically includes:

[0107] Clustering modeling is performed on users' electricity consumption habits to identify different types of daily electricity consumption behavior patterns, resulting in typical daily electricity consumption templates for users. These templates are then input into a pre-set user behavior prediction model, which, combined with weather changes and holiday characteristics, predicts the user's electricity consumption at different times in the next 24 hours.

[0108] Clustering algorithms are used to identify and predict user electricity consumption patterns. Specifically, clustering algorithms (such as K-means or DBSCAN) are first used to analyze users' historical electricity consumption data. Electricity consumption behavior is categorized based on its variation characteristics (such as morning peak, evening peak, and differences between weekdays and weekends). Each category corresponds to a typical "daily electricity consumption template." For example, the weekday template differs from the weekend template in terms of peak electricity consumption time and duration.

[0109] After clustering is completed, the output typical daily electricity consumption template is input into a preset user behavior prediction model. This model can be a neural network-based regression model (such as an LSTM network), which can predict the electricity consumption trend for the next 24 hours based on external factors such as user electricity consumption patterns, weather changes, and holidays, and output the electricity consumption for different time periods.

[0110] For example, suppose that after cluster analysis, a user's historical electricity consumption data identifies peak weekday electricity consumption between 7:00 AM and 9:00 AM, and peak evening consumption between 5:00 PM and 7:00 PM, while weekend electricity consumption shows a relatively stable load distribution throughout the day. Based on these consumption patterns, a typical daily electricity consumption template for the user will reflect electricity consumption at different times (e.g., 300W at 7:00 AM and 200W at 4:00 PM). Then, combining meteorological data (such as tomorrow's temperature and humidity) and holiday information, the predictive model calculates the hourly electricity consumption for the next 24 hours and adjusts the expected electricity demand.

[0111] By segmenting and statistically analyzing electricity consumption at different times, and performing sliding window regression analysis, the actual range of users' electricity demand and the actual time periods of electricity consumption are obtained.

[0112] Sliding window regression analysis is used to determine the range and time period of electricity demand. Specifically, the predicted electricity consumption for the next 24 hours is divided into multiple time periods by hour, and the electricity consumption for each time period is statistically analyzed. Then, sliding window regression analysis is used to smooth the electricity consumption data for each time period, eliminating sudden fluctuations and noise, thereby extracting the range of stable electricity demand for users (such as basic electricity demand and high-intensity electricity demand). Furthermore, sliding window regression can also be used to identify the actual time periods of electricity consumption for users. For example, if a user's electricity demand fluctuates significantly during certain periods, these fluctuations need to be specifically marked and incorporated into scheduling decisions.

[0113] For example, if a user's electricity consumption is 300W between 8 AM and 9 AM, but under certain circumstances (such as when the air conditioner is turned on or many people are using appliances), the electricity consumption during this period suddenly increases to 500W. Sliding window regression analysis will smooth out this fluctuation and determine that the normal electricity consumption for this period is 300W. It will also identify 8 AM to 9 AM as the "basic electricity demand" period and 9 AM to 11 AM as the "high-intensity electricity demand" period. In this case, power dispatch will be implemented based on the high-intensity electricity demand period to ensure that photovoltaic energy storage can provide sufficient power during these time periods.

[0114] The steps involved in segmenting and statistically analyzing electricity consumption at different times to obtain the user's actual electricity demand range and actual electricity consumption time periods include:

[0115] Electricity consumption is segmented and statistically analyzed for different time periods. The mean and deviation are then extracted from the statistical results, and the range of basic electricity demand and high-intensity electricity demand are identified based on the mean and deviation.

[0116] This method utilizes time-segmented statistics to extract the basic electricity demand range and the high-intensity electricity demand range. Specifically, electricity consumption data for different time periods within the next 24 hours is segmented, for example, by hourly or more finely by minute, forming an independent electricity consumption statistics dataset for each time period. For each dataset, the mean and deviation of electricity consumption are calculated. The mean represents the average load demand for that time period, and the deviation represents the fluctuation range of electricity consumption. Based on preset thresholds, loads within the low deviation range are classified as the "basic electricity demand range," while loads within the high deviation range are classified as the "high-intensity electricity demand range."

[0117] For example, when collecting electricity consumption data from 7:00 AM to 9:00 AM, the average hourly electricity consumption of users was recorded as 400W, with a deviation of 30W. This period was marked as the "basic electricity demand range." However, in the electricity consumption data from 6:00 PM to 8:00 PM, the records showed an average electricity consumption of 600W, with a deviation of 100W. This period was identified as having significant load fluctuations and was marked as the "high-intensity electricity demand range."

[0118] The basic electricity demand range and the high-intensity electricity demand range are merged to eliminate time overlap intervals and generate continuous non-overlapping electricity intensity curves. The electricity intensity curves are time-series fitted, and the peak periods, durations, and surge points are extracted based on the fitting results. The actual electricity demand range of users is constructed based on the peak periods, durations, and surge points.

[0119] The actual electricity demand range is constructed through non-overlap processing and time-series fitting. Specifically, the first step involves cross-merging the basic electricity demand range and the high-intensity electricity demand range. If there is overlap between basic and high-intensity demand in a certain time period, it is preferentially identified as high-intensity demand, and all overlapping intervals are eliminated, thereby generating a continuous non-overlapping electricity intensity curve. The second step involves time-series fitting of the generated electricity intensity curve, using polynomial fitting or nonlinear regression methods to extract data trends, obtaining the peak periods of electricity consumption changes (peak load periods), durations (time range of load fluctuations), and spike points (points of rapid load changes). Finally, the actual electricity demand range for users is constructed based on the fitting results, and combined with key demand parameters, this provides a basis for the power generation scheduling of photovoltaic energy storage equipment.

[0120] For example, through fusion processing, it was found that the electricity demand from 7:00 to 8:30 AM is basic demand, while the demand from 8:30 to 9:00 AM is classified as high-intensity demand due to the startup of high-load equipment (such as kitchen appliances). The demand from 9:00 to 9:30 AM is then attributed to basic load. Through time-series fitting analysis, the peak period for users was extracted from 8:00 to 8:50 AM, lasting 50 minutes, and 8:30 AM was identified as a sudden surge point. Based on these results, the actual electricity demand range was constructed as follows: the peak period is from 8:00 to 8:50 AM, and the reserved load capacity is 550W.

[0121] A sliding window regression analysis was performed on the electricity consumption at different times to extract stable load areas and sudden load areas, and load change trend curves were constructed based on the stable load areas and sudden load areas.

[0122] Load patterns are extracted and trends are constructed using sliding window regression. Specifically, future user electricity demand is processed using a sliding window method, and multi-window regression analysis (such as linear sliding window regression and regression methods based on smoothing kernel functions) is applied to extract two key regions: a stable load region and a sudden load region. The stable load region refers to the segment where user load changes at a relatively low rate, while the sudden load region refers to the segment where electricity demand rises or falls rapidly as detected by sliding window regression. Based on these two types of regions, load change trend curves are generated to further identify the characteristics of user electricity consumption patterns.

[0123] For example, when processing data from 7:00 to 9:00 AM, it was identified that the load fluctuation range from 7:00 to 8:30 AM did not exceed 20W (stable load area), while from 8:30 to 9:00 AM the load rapidly increased from 450W to 650W (sudden load area). Trend curves were generated based on these two areas to show the overall trend of electricity demand changes in the morning and to mark the fluctuation areas.

[0124] By performing regression fitting and cross-analysis on the load change trend curve based on the actual electricity demand range, the user's actual electricity consumption time period can be obtained.

[0125] The actual electricity consumption periods of users are identified through regression fitting and cross-analysis. Specifically, the load change trend curve is cross-analyzed with the user's actual electricity demand range. This analysis, based on a Bayesian decision model or support vector regression (SVR) method, further confirms the user's actual electricity consumption periods based on the intersection of the user's peak periods, high-intensity demand, and stable load areas. For example, if a certain time period belongs to high-intensity demand and also appears in a rapidly increasing area of ​​the load change trend, it is automatically designated as a priority electricity consumption period.

[0126] For example, cross-analysis confirmed that the period from 8:30 AM to 9:00 AM falls both within the range of users' actual electricity demand and within a rapidly fluctuating load trend. This period was ultimately designated as the actual electricity consumption time. Combined with other analysis results, a user's electricity usage plan was generated to ensure that photovoltaic energy storage equipment can prioritize power supply during this phase.

[0127] In this embodiment, by analyzing user electricity consumption habits, predicting solar radiation, and analyzing electricity price fluctuations using time series analysis and data mining methods, a comprehensive prediction of the next day's photovoltaic power generation and the period of highest electricity price for the current day is achieved. First, historical user electricity consumption data is classified through cluster modeling to identify user electricity consumption patterns for different types of days. Then, a preset user behavior prediction model is used, combined with weather changes and holiday characteristics, to accurately predict the user's electricity consumption at different times within the next 24 hours. Next, by combining the physical parameters of the photovoltaic panel equipment and the solar radiation value of the next day, the photovoltaic power generation for each time period of the next day is calculated using a preset formula. A more accurate power generation prediction curve is obtained after corrections for temperature coefficients and equipment attenuation factors.

[0128] In electricity price forecasting, historical electricity price records and real-time characteristic information are first input into a short-term price forecasting model, outputting hourly electricity price sequences, electricity price volatility sequences, and a set of abnormal factors. Temporal gradient analysis and peak detection are then used to identify candidate periods of rapid price increases. Subsequently, parameters such as price growth rate and duration are normalized and aggregated, and combined with holiday factors, grid load factors, and meteorological conditions, price anomaly risk factors and price increase rate change maps are generated. Through hierarchical threshold analysis and gradient reconstruction, peak periods with rapid price increases and discharge value are extracted, and finally, the highest electricity price period meeting the peak price threshold is selected.

[0129] Meanwhile, to analyze the actual electricity demand range and time period of users, segmented statistics are performed based on future electricity consumption at different times. Sliding window regression analysis is then used to identify the basic electricity demand range and the high-intensity electricity demand range. Simultaneously, an electricity intensity curve is generated by eliminating overlapping intervals, and time-series fitting is performed on the curve to extract peak periods, durations, and spike points, used to construct the actual electricity demand range of users. Finally, through cross-analysis of the load change trend curve and actual electricity demand, the actual electricity consumption time period of users is determined, achieving an effective combination of photovoltaic energy storage power dispatch and electricity price tracking. This embodiment can significantly improve the economic efficiency and intelligence level of photovoltaic energy storage, providing strong support for efficient energy operation and maximizing user benefits.

[0130] Example 3:

[0131] like Figure 2 As shown, this application also provides an AI model-based electricity price tracking and sales system 10, which is applied to photovoltaic energy storage equipment, including an acquisition module 11, a prediction module 12, a comparison module 13, and a sales module 14.

[0132] The acquisition module 11 is mainly used to acquire users' electricity consumption habits, predict the solar radiation value of the next day and the electricity price fluctuation curve of the day, and calculate the photovoltaic power generation of the next day based on the solar radiation value of the next day.

[0133] The prediction module 12 is mainly used to predict the period of highest electricity price based on the electricity price fluctuation curve of the day if the photovoltaic power generation of the next day is higher than the capacity of the photovoltaic energy storage device, and to analyze the user's actual electricity demand range and actual electricity consumption period based on the user's electricity consumption habits.

[0134] The comparison module 13 is mainly used to compare the time period with the time period of electricity consumption. If the time period with the highest electricity price is within the preset time interval, the photovoltaic energy storage device reserves enough electricity to meet the electricity demand range, and sells the remaining electricity of the photovoltaic energy storage device through the wall.

[0135] The electricity sales module 14 is mainly used to sell all the electricity generated by the photovoltaic energy storage device through the wall if the time period with the highest electricity price and the time period of electricity consumption are not within the preset time interval.

[0136] In this embodiment, an AI-based electricity price tracking and sales system 10 is designed to comprehensively realize the intelligent integration of electricity price tracking, user demand prediction, and sales strategy formulation and execution for photovoltaic energy storage devices. The system consists of an acquisition module 11, a prediction module 12, a comparison module 13, and a sales module 14 working collaboratively, each responsible for different functional tasks. The acquisition module 11 uses time series analysis and machine learning techniques to acquire user electricity consumption habits, combines weather prediction models to predict key meteorological parameters for the next day, and calculates the photovoltaic power generation for the next day using the physical parameters of the photovoltaic panels and solar radiation values, providing basic data for subsequent modules. The prediction module 12 uses an AI model to accurately predict the peak electricity price periods in the electricity price fluctuation curve, and simultaneously outputs the user's actual electricity demand range and actual electricity consumption time periods based on a clustering model, dynamically updating user behavior trends to optimize matching efficiency. The comparison module 13 compares the predicted peak electricity price period with the user's actual electricity consumption time periods, uses cross-analysis to identify the overlap between peak electricity prices and user demand, and rationally allocates the electricity of the photovoltaic energy storage devices. The electricity sales module 14 enables efficient off-grid electricity sales of photovoltaic power generation when user demand and electricity price periods do not match. It uses a target optimization algorithm to determine the electricity sales period, ensuring maximum revenue. This embodiment integrates forecasting, energy storage scheduling, and dynamic electricity sales management to systematically maximize the economic benefits of photovoltaic energy storage equipment while reducing user electricity costs, providing an efficient and intelligent solution for new energy systems.

[0137] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system and each module described above can be referred to the corresponding process in the aforementioned Embodiment 1, and will not be repeated here.

[0138] Example 4:

[0139] like Figure 3 As shown, this application also provides an electronic device 20, including a memory 21 and a processor 22. The memory 21 stores a computer program that can run on the processor 22. When the processor 22 executes the computer program, it implements the AI ​​model-based electricity price tracking and electricity sales method of Embodiment 1.

[0140] In this embodiment, the memory 21 and processor 22 of the electronic device 20 are used to run a computer program to implement a planning method for electricity price tracking and sales based on an AI model, providing a convenient computing platform and efficient execution mechanism for photovoltaic energy storage device scheduling and electricity sales solutions. The memory 21 stores the computer program designed for Embodiment 1, which includes core modules such as time series analysis, price prediction models, user behavior prediction models, and equipment optimization algorithms. When the processor 22 executes the computer program, it can accurately extract and predict user electricity consumption habits, calculate the solar radiation value for the next day, estimate photovoltaic power generation, and analyze electricity price fluctuation trends. It identifies the highest electricity price period through time series fitting and graded thresholds. The processor 22 can also dynamically allocate the power of the photovoltaic energy storage device based on cross-comparison results to meet users' priority electricity needs and execute a wall-mounted electricity sales strategy during non-demand periods. During operation, the electronic device 20 can efficiently coordinate solar energy resources, user electricity consumption behavior, and grid price fluctuations, making the allocation of energy storage devices more precise. It can also respond in real time to weather changes and policy requirements, improving user benefits and promoting the widespread application of green energy.

[0141] Example 5:

[0142] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when run by a processor, causes the processor to execute an AI-based electricity price tracking and electricity sales method as described in any of Embodiment 1.

[0143] In this embodiment, a computer-readable storage medium is used to implement an AI-based electricity price tracking and sales method for photovoltaic energy storage devices. The storage medium stores a processor-executable computer program, an optimized implementation of the method in Embodiment 1, integrating complex time-series forecasting algorithms, photovoltaic power generation calculation formulas, electricity price fluctuation prediction models, and user demand analysis methods. When the computer program runs on the processor, it can efficiently perform comprehensive analysis of user electricity consumption habits and environmental data, accurately predict user electricity consumption trends and next-day solar radiation values ​​through a time-series forecasting network, and dynamically plan the time-segmented power generation of photovoltaic energy storage devices. The program in the storage medium further utilizes a short-term electricity price forecasting model to extract price anomaly risk factors and periods of highest electricity prices. Simultaneously, it combines the range of electricity demand with high-intensity demand areas to generate a cross-barrier electricity sales strategy through cross-analysis. This computer-readable storage medium can accurately execute energy storage device allocation and intelligent electricity sales operations, providing reliable algorithmic support for the photovoltaic energy storage scheduling system, significantly improving the utilization efficiency of energy storage power, optimizing electricity price revenue, and promoting the efficient use of new energy sources.

[0144] The structures, proportions, sizes, etc., shown in the accompanying drawings of this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed in the specification, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0145] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An AI model-based electricity price tracking and selling method applied to a photovoltaic energy storage device, characterized in that, The method comprises the following steps: Obtaining user electricity usage habits and predicting the next day's solar radiation value and the day's electricity price fluctuation related features; Specifically, inputting the electricity price fluctuation related features into a preset short-term price prediction model to obtain a predicted hourly electricity price sequence, an electricity price fluctuation rate sequence, and an abnormal factor set of high-frequency price fluctuations; performing fluctuation window sliding analysis on the hourly electricity price sequence, and extracting local maximum points and slope information based on the analysis results; using the local maximum points and the slope information to construct price fluctuation intensity and price trend change rate of each time period; weighting and fusing the price fluctuation intensity and the electricity price fluctuation rate sequence to generate a price fluctuation weight curve; Performing cross-time-point identification on the price trend change rate and the abnormal factor set to generate a sensitive time period set of price jumps; jointly clustering the price fluctuation weight curve and the sensitive time period set to obtain the day's electricity price fluctuation curve; calculating the next day's photovoltaic power generation based on the next day's solar radiation value; If the next day's photovoltaic power generation is higher than the capacity of the photovoltaic energy storage device, predicting the highest electricity price time period based on the day's electricity price fluctuation curve; specifically, performing standardization processing on the day's electricity price fluctuation curve to construct a time gradient curve reflecting the electricity price fluctuation range, performing peak value detection on the time gradient curve to identify candidate time periods of rapid electricity price rise; extracting the price growth rate and duration corresponding to the candidate time periods, and performing normalization and aggregation analysis on the price growth rate and the duration to obtain a fluctuation intensity score matrix at the time period level; inputting the fluctuation intensity score matrix into a preset price risk identification model to output a confidence-weighted price abnormal risk factor and a price rise rate change map; performing hierarchical threshold analysis on the price abnormal risk factor to identify a price high-rise candidate time period, and performing gradient reconstruction on the price high-rise candidate time period and the price rise rate change map to extract a main peak time period; Comparing the main peak time period with a preset electricity price peak threshold to filter out the highest electricity price time period with a price higher than the average peak value; and analyzing the user's actual electricity demand range and actual electricity time period according to the user electricity usage habits; Comparing the highest electricity price time period with the electricity time period in terms of time period, if the highest electricity price time period and the electricity time period are within a preset time interval range, the photovoltaic energy storage device reserves electricity to meet the electricity demand range, and the remaining electricity of the photovoltaic energy storage device is sold to a wall; If the highest electricity price time period and the electricity time period are not within the preset time interval range, all the electricity of the photovoltaic energy storage device is sold to a wall. 2.The AI model-based electricity price tracking and selling method of claim 1, wherein, The step of obtaining user electricity usage habits and predicting the next day's solar radiation value and the day's electricity price fluctuation related features comprises: Obtaining user historical electricity data, performing time series analysis on the user historical electricity data, and extracting user electricity frequency, electricity quantity change amplitude and periodicity characteristics in different time periods to construct user electricity usage habits; inputting the user electricity habit and weather history data into a preset time sequence prediction network to predict user electricity trend in the next 24 hours and key meteorological parameters of the next day, inputting the key meteorological parameters into a preset solar radiation prediction sub-model to output solar radiation value of the next day; characteristic extraction is performed on historical power grid price records, holiday information and demand response policies by using a feature extraction network to construct electricity price fluctuation related features. 3.The AI model-based electricity price tracking and selling method of claim 1, wherein, The step of calculating the photovoltaic power generation of the next day based on the solar radiation value of the next day comprises: The solar radiation value of the next day is combined with the area, inclination parameter and conversion efficiency parameter of the photovoltaic panel device by using a preset combination calculation formula to obtain the photovoltaic power generation of each time period of the next day. The combination calculation formula is as follows: P t =I t ×A×η×cos(θ); Wherein, the P t is the photovoltaic power generation at time t; I t is the solar radiation value at time t; A is the effective area of the photovoltaic panel; η is the conversion efficiency of the photovoltaic panel; θ is the inclination angle of the photovoltaic panel. 4.The AI model-based electricity price tracking and selling method of claim 1, wherein, The step of analyzing the actual electricity demand range and actual electricity time period of the user according to the user electricity habit comprises: The user typical daily electricity template is input into a preset user behavior prediction model to predict the electricity consumption of the user in different time periods in the next 24 hours. The actual electricity demand range and actual electricity time period of the user are obtained by performing segmented statistical and sliding window regression analysis on the electricity consumption in different time periods. 5.The AI model-based electricity price tracking and selling method of claim 4, wherein, The step of obtaining the actual electricity demand range and actual electricity time period of the user by performing segmented statistical and sliding window regression analysis on the electricity consumption in different time periods comprises: The electricity consumption in different time periods is subjected to segmented statistical analysis to obtain statistical results, the mean value and deviation are extracted according to the statistical results, and the basic electricity demand range and high-intensity electricity demand range are identified according to the mean value and the deviation. The basic electricity demand range and the high-intensity electricity demand range are fused to generate a continuous non-overlapping electricity intensity curve, the electricity intensity curve is subjected to time sequence fitting, and the peak time period, duration and sudden increase point are extracted based on the fitting results, and the actual electricity demand range of the user is constructed according to the peak time period, the duration and the sudden increase point. The stable load area and the sudden load area are extracted by performing sliding window regression analysis on the electricity consumption in different time periods, and a load change trend curve is constructed according to the stable load area and the sudden load area. The actual electricity time period of the user is obtained by performing regression fitting and cross analysis on the load change trend curve through the actual electricity demand range.

6. An AI model-based electricity price tracking and selling system applied to a photovoltaic energy storage device, characterized in that, It comprises: An acquisition module is configured to acquire a user electricity habit, and predict a solar radiation value of the next day and electricity price fluctuation related features of the day. Specifically, the electricity price fluctuation related features are input into a preset short-term price prediction model to obtain a predicted hourly electricity price sequence, an electricity price fluctuation rate sequence and an abnormal factor set of high-frequency price fluctuation, the hourly electricity price sequence is subjected to fluctuation window sliding analysis, and local maximum points and slope information are extracted based on the analysis results, the price fluctuation intensity and price trend change rate of each time period are constructed by using the local maximum points and the slope information, and the price fluctuation intensity and the electricity price fluctuation rate sequence are subjected to weighted fusion to generate a price fluctuation weight curve. Crossing time point identification is performed on the price trend change rate and the abnormal factor set to generate a sensitive time period set of price jump; joint clustering analysis is performed on the price fluctuation weight curve and the sensitive time period set to obtain a daily electricity price fluctuation curve; and the next day's photovoltaic power generation capacity is calculated based on the next day's solar radiation value; The prediction module is configured to, if the next day's photovoltaic power generation capacity is higher than the capacity of the photovoltaic energy storage device, predict a highest electricity price time period of the electricity price based on the daily electricity price fluctuation curve. Specifically, the daily electricity price fluctuation curve is subjected to standardization processing to construct a time gradient curve reflecting the electricity price variation amplitude, peak value detection is performed on the time gradient curve to identify a candidate time period of rapid electricity price rise, the price growth rate and the duration corresponding to the candidate time period are extracted, and the price growth rate and the duration are subjected to normalization and aggregation analysis to obtain a fluctuation intensity score matrix at the time period level. The fluctuation intensity score matrix is input into a preset price risk identification model to output a confidence-weighted price abnormal risk factor and a price rise rate change map. The price abnormal risk factor is subjected to hierarchical threshold analysis to identify a price high-rise candidate time period, and the price high-rise candidate time period and the price rise rate change map are subjected to gradient reconstruction to extract a main peak time period. The main peak time period is compared with a preset electricity price peak value threshold to filter out a highest electricity price time period in which the price is higher than the average peak value. The actual electricity demand range and the actual electricity time period of the user are analyzed according to the user electricity consumption habit. The comparison module is configured to compare the highest electricity price time period with the electricity time period, and if the highest electricity price time period and the electricity time period are within a preset time interval range, the photovoltaic energy storage device reserves an electricity quantity meeting the electricity demand range, and the remaining electricity quantity of the photovoltaic energy storage device is sold to a neighbor. The electricity selling module is configured to, if the highest electricity price time period and the electricity time period are not within the preset time interval range, sell all the electricity quantity of the photovoltaic energy storage device to a neighbor.

7. An electronic device, comprising: A memory and a processor are included, the memory stores a computer program capable of running on the processor, and the processor implements the AI model-based electricity price tracking and electricity selling method of any one of claims 1 to 5 when executing the computer program.

8. A computer-readable storage medium, characterized in that, A computer program is stored thereon, and the computer program causes the processor to execute the AI model-based electricity price tracking and electricity selling method of any one of claims 1 to 5 when the processor runs the computer program.

Images