Day-ahead and real-time optimization bidding strategy method of optical storage direct current integrated system in electricity spot market

Abstract

The present application relates to the technical field of power spot market quotation, in particular to a day-ahead and real-time optimization quotation strategy method of a photovoltaic and energy storage direct current integrated system in a power spot market, comprising: generating a day-ahead market quotation strategy according to photovoltaic power generation prediction values, energy storage system charge prediction values and market electricity price prediction information; in the real-time market, generating a real-time quotation strategy according to real-time environmental data, the day-ahead market quotation strategy and the real-time state of the energy storage charge. The present application improves the flexibility and robustness of the quotation strategy by formulating a quotation strategy based on multi-source prediction information in the day-ahead stage and dynamically adjusting according to the latest environmental data and charge state in the real-time stage, realizes the coordination and unification of day-ahead prediction and real-time response, and improves the flexibility of the quotation strategy and the economy of system operation. By fully exerting the synergistic effect of photovoltaic and energy storage, the present application improves the new energy utilization efficiency and the income level in the spot market, and has good popularization and application prospect.

Description

Technical Field

[0001] This invention relates to the field of electricity spot market pricing technology, specifically to a day-ahead and real-time optimized pricing strategy method for a photovoltaic-storage-DC integrated system in the electricity spot market. Background Technology

[0002] With the large-scale integration of new energy sources, especially photovoltaic power generation, the power system faces challenges such as high output uncertainty, widening peak-valley differences, and insufficient regulation capacity. To improve the utilization rate of new energy resources and the stability of the system, integrated photovoltaic-storage-DC power systems are gradually becoming an important component of the new power system. This system integrates photovoltaic power generation with energy storage devices and combines them with DC power distribution technology, which can not only improve energy utilization efficiency but also achieve flexible power dispatch and multi-market participation.

[0003] In the electricity spot market, market participants typically need to formulate bidding strategies based on day-ahead forecasts and adjust them according to actual conditions during real-time operation. Because photovoltaic power generation is highly dependent on weather conditions, actual power generation often deviates from day-ahead forecasts. Simultaneously, the state of charge of energy storage systems may also differ from expectations due to dispatch changes, leading to deviations in the execution of original bidding strategies, affecting market returns and even system stability.

[0004] Existing pricing strategies often rely on static forecasts or adjustments based on a single factor, making it difficult to take into account the impact of multiple uncertainties on system output and market response, resulting in scheduling schemes lacking flexibility and economy. Summary of the Invention

[0005] (a) Purpose of the invention The purpose of this invention is to provide a method for optimizing the day-ahead and real-time bidding strategies of a photovoltaic-storage-DC integrated system in the electricity spot market by formulating a bidding strategy based on multi-source forecast information in the day-ahead stage and dynamically adjusting it in the real-time stage according to the latest environmental data and state of charge.

[0006] (II) Technical Solution To address the above problems, this invention provides a method for optimizing day-ahead and real-time pricing strategies for integrated photovoltaic-storage DC systems in the electricity spot market, comprising: Based on photovoltaic module parameters, environmental data, and historical power generation data, short-term forecasts of photovoltaic output are made to obtain photovoltaic power generation forecasts. Based on the preset energy storage charge prediction model, the predicted charge value of the energy storage system is obtained; Based on historical electricity prices and load demand data in the electricity spot market, market electricity price forecasting information is obtained; Based on the photovoltaic power generation forecast, the energy storage system charge forecast, and the market electricity price forecast, a day-ahead market pricing strategy is generated. In the real-time market, a real-time pricing strategy is generated based on real-time environmental data, day-ahead market pricing strategies, and the real-time status of energy storage charge.

[0007] In another aspect of the present invention, preferably, the step of making short-term predictions of photovoltaic output based on photovoltaic module parameters, environmental data, and historical power generation data to obtain predicted photovoltaic power generation values ​​includes: The parameters of the photovoltaic module include module area, conversion efficiency and installation angle; the environmental data include solar irradiance and ambient temperature. A photovoltaic power generation prediction model is constructed. The photovoltaic power generation prediction model is based on a deep neural network. The input of the photovoltaic power generation prediction model is the component area, conversion efficiency, installation angle, solar irradiance and ambient temperature. The output of the photovoltaic power generation prediction model is the photovoltaic power generation prediction value. The photovoltaic power generation prediction model is trained using the historical power generation data, and the predicted value of photovoltaic power generation is obtained using the trained photovoltaic power generation prediction model.

[0008] In another aspect of the present invention, preferably, obtaining the predicted charge value of the energy storage system based on a preset energy storage charge prediction model includes: Obtain the initial state of charge, rated capacity, photovoltaic power generation forecast, and load-side electricity consumption forecast of the energy storage system; A state of charge (SOC) prediction model is constructed. The SOC prediction model is based on a deep neural network. The inputs of the SOC prediction model are the initial SOC, rated capacity, predicted photovoltaic power generation, and predicted load-side electricity consumption. The output of the SOC prediction model is the predicted SOC of the energy storage system.

[0009] In another aspect of the present invention, preferably, obtaining market electricity price forecast information based on historical electricity prices and load demand data from the electricity spot market includes: Obtain historical electricity price data and corresponding load demand data for the electricity spot market within the target area; Extract time series features from the historical electricity price data and load demand data, including periodic features, trend features, and abrupt change features; Based on the periodicity, trend, and abrupt change characteristics, market electricity price forecast information is obtained using a pre-defined electricity price forecasting model.

[0010] In another aspect of the present invention, preferably, obtaining market electricity price forecast information based on a preset electricity price forecasting model according to the periodic characteristics, trend characteristics, and abrupt change characteristics includes: The pre-defined electricity price prediction model is structurally initialized and periodic factors are embedded using the periodic features to obtain the modeled electricity price prediction model. The modeled electricity price prediction model is dynamically corrected and its output is calibrated using trend and abrupt change characteristics.

[0011] In another aspect of the present invention, preferably, the modeled electricity price prediction model is dynamically corrected and its output calibrated using trend features and abrupt change features, including: The long-term trend offset is constructed using the aforementioned trend characteristics, and the output of the modeled electricity price prediction model is adjusted linearly based on this trend. By utilizing the aforementioned mutation characteristics, abnormal electricity price events are identified, and the prediction results are adjusted through an anomaly-sensitive weighting mechanism.

[0012] In another aspect of the present invention, preferably, a day-ahead market pricing strategy is generated based on the photovoltaic power generation forecast, the energy storage system charge forecast, and the market electricity price forecast information, including: Based on the photovoltaic power generation forecast and the energy storage system charge forecast, an available power forecast curve is constructed. The constraints of the available power forecast curve include the photovoltaic output capacity and the charging and discharging capacity boundary of the energy storage system. The available electricity forecast curve and market electricity price forecast information are input into the pricing strategy generation model to generate the day-ahead market pricing strategy.

[0013] In another aspect of the present invention, preferably, the available electricity forecast curve and market electricity price forecast information are input into a pricing strategy generation model to generate a day-ahead market pricing strategy, including: The forecast period is divided into multiple pricing decision cycles; Based on the available electricity forecast curve, market electricity price forecast information and the charging and discharging capacity boundary of the energy storage system, a clearing electricity and bidding price pair is generated in each bidding decision cycle. Based on the goal of maximizing profits, the cleared electricity volume and quoted price are optimized to generate a day-ahead market pricing strategy.

[0014] In another aspect of the present invention, preferably, In the real-time market, a real-time pricing strategy is generated based on real-time environmental data, day-ahead market pricing strategies, and the real-time status of energy storage charge, including: The deviation value of photovoltaic power generation is obtained based on real-time environmental data; Based on the real-time charge status of the energy storage system, the charge deviation value of the energy storage system is obtained; Based on the photovoltaic power generation deviation value and the energy storage system charge deviation value, the day-ahead market pricing strategy is optimized to generate a real-time pricing strategy.

[0015] In another aspect of the present invention, preferably, the step of optimizing the day-ahead market pricing strategy based on the photovoltaic power generation deviation value and the energy storage system charge deviation value to generate a real-time pricing strategy includes: A set of real-time pricing adjustment factors is established based on the deviation values ​​of photovoltaic power generation and the charge deviation values ​​of energy storage systems; Based on the set of adjustment factors, calculate the expected return change of the pricing strategy for each time period in the day-ahead market. Based on the principle of maximizing profits and the operational constraints of the energy storage system, the cleared electricity volume and bid price pair of the day-ahead bidding strategy are adjusted to generate a real-time bidding strategy.

[0016] (III) Beneficial Effects The above-described technical solution of the present invention has the following beneficial technical effects: This invention integrates photovoltaic module parameters, environmental data, and historical power generation data to improve the accuracy of photovoltaic output prediction and effectively reduce market risks caused by fluctuations in new energy output. Simultaneously, by constructing a state-of-charge (SOC) prediction model for energy storage, the adjustability of the energy storage system can be assessed in advance, providing a reliable basis for strategy formulation. Combining historical market electricity prices and load demand data, a forward-looking and economical day-ahead pricing strategy is generated. Furthermore, the day-ahead strategy is dynamically adjusted in real-time based on environmental changes and the real-time status of the energy storage system, making real-time pricing more closely aligned with actual operating conditions. The overall solution achieves coordinated unity between day-ahead prediction and real-time response, improving the flexibility of the pricing strategy and the economic efficiency of system operation. By fully leveraging the synergistic effect of photovoltaics and energy storage, this invention can significantly improve the efficiency of new energy utilization and the profitability in the spot market, demonstrating promising prospects for widespread application. Attached Figure Description

[0017] Figure 1 This is an overall flowchart of one embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments and the accompanying drawings. It should be understood that these descriptions are merely exemplary and not intended to limit the scope of the invention. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0019] Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0020] Furthermore, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0021] The invention will now be described in more detail with reference to the accompanying drawings. In the various drawings, the same elements are indicated by similar reference numerals. For clarity, the various parts in the drawings are not drawn to scale.

[0022] Example 1 A method for optimizing day-ahead and real-time pricing strategies for integrated photovoltaic-storage DC systems in the electricity spot market. Figure 1 An overall flowchart of one embodiment of the present invention is shown, as follows: Figure 1 As shown, it includes: Based on photovoltaic module parameters, environmental data, and historical power generation data, short-term photovoltaic output is predicted to obtain the predicted photovoltaic power generation value; in this embodiment, it includes: The parameters of the photovoltaic module include module area, conversion efficiency, and installation angle. Module area reflects the total effective light-receiving area of ​​the photovoltaic system to receive solar radiation; photoelectric conversion efficiency represents the ability to convert solar energy into electrical energy per unit area; installation angle, including tilt angle and azimuth angle, affects incident light intensity and effective irradiation time. The environmental data includes solar irradiance and ambient temperature; solar irradiance is a direct factor determining instantaneous power generation capacity; ambient temperature affects module temperature, thus having a certain impact on module efficiency.

[0023] A photovoltaic (PV) power generation prediction model is constructed based on a deep neural network. The inputs to the model are module area, conversion efficiency, installation angle, solar irradiance, and ambient temperature. The output is the predicted PV power generation value. The constructed PV power generation prediction model employs a deep neural network, such as a multilayer feedforward neural network or an LSTM network. The model's inputs include the five parameters mentioned above: module area, conversion efficiency, installation angle, solar irradiance, and ambient temperature. To improve prediction accuracy, the input data can be organized in a time-series manner according to the actual collection frequency, forming a time-window input structure.

[0024] The photovoltaic power generation prediction model is trained using the historical power generation data, and the predicted photovoltaic power generation value is obtained using the trained model. To construct and optimize this prediction model, the system trains the model using historical power generation data, which includes measured photovoltaic output values ​​recorded synchronously with the input parameters. During training, a supervised learning approach is adopted, aiming to minimize the error between the predicted and measured values ​​(e.g., mean squared error, MSE), and parameter optimization is performed using the backpropagation algorithm. To prevent overfitting, regularization strategies, early stopping mechanisms, or cross-validation can be introduced for parameter tuning.

[0025] Based on a pre-defined energy storage charge prediction model, the predicted charge values ​​of the energy storage system are obtained, including: Obtain the initial state of charge, rated capacity, photovoltaic power generation forecast, and load-side electricity consumption forecast of the energy storage system; the initial state of charge of the energy storage system is used to represent the electricity basis at the starting point of the model prediction; the rated capacity of the energy storage system is the maximum electrical energy that the energy storage device can store. A state-of-charge (POC) prediction model is constructed based on a deep neural network. The inputs to the POC model are the initial POC, rated capacity, predicted photovoltaic (PV) power generation, and predicted load-side electricity consumption. The output of the POC model is the predicted POC value of the energy storage system. Depending on the specific implementation, the POC model may employ a feedforward neural network, a recurrent neural network, or a long short-term memory (LSTM) network to enhance its modeling capability for time-series inputs. Through supervised training on historical operating data, the model learns the nonlinear mapping relationship between input variables and POC changes. This allows for dynamic prediction of the energy storage system's charging and discharging capabilities, taking into account fluctuations in PV power generation and electricity demand, and proactively identifying charging capacity or discharging limitations. This provides crucial input data for market strategy formulation and system operation optimization.

[0026] Based on historical electricity price and load demand data from the electricity spot market, market electricity price forecasting information is obtained, including: Obtain historical electricity price data and corresponding load demand data for the electricity spot market within the target area; The system extracts time-series features from the historical electricity price data and load demand data. These time-series features include periodicity, trend, and abrupt change characteristics. The system performs time-series analysis on the historical electricity price and load data, extracting the main time-series features, including: periodicity, such as intraday load fluctuation patterns and weekday / weekend electricity price cycle changes; trend, such as seasonal load growth trends and medium- to long-term electricity price trends; and abrupt change characteristics, such as drastic price fluctuations caused by abnormal weather, grid failures, or policy adjustments.

[0027] Based on the aforementioned periodic characteristics, trend characteristics, and abrupt change characteristics, and using a pre-defined electricity price prediction model, market electricity price prediction information is obtained, including: The system utilizes the aforementioned periodic features to initialize the structure of a pre-defined electricity price prediction model and embed periodic factors, resulting in a modeled electricity price prediction system. The periodic features primarily include intraday cycles, such as peak-valley price changes; weekly cycles, such as price patterns between weekdays and weekends; and seasonal cycles, such as peak air conditioning loads in summer. When constructing deep prediction models, such as LSTM, TCN, or Transformer, the system introduces a periodic embedding module to embed these periodic factors into the model's input layer or intermediate hidden layers in numerical or vector form. This allows the model structure to adapt to periodic fluctuations during the initialization phase, helping to capture the basic fluctuation patterns of electricity prices.

[0028] The modeled electricity price prediction model is dynamically corrected and its output calibrated using trend and abrupt change characteristics, including: The system utilizes the aforementioned trend characteristics to construct a long-term trend offset, which is then used to linearly adjust the output of the modeled electricity price prediction system. The trend characteristics primarily refer to the long-term upward or downward trend of electricity prices over time, which may be caused by seasonal load growth, changes in power supply structure, or adjustments in market policies. To eliminate systematic errors caused by the trend offset, the system constructs a long-term trend offset function. By fitting a sliding window to the historical average electricity price series, linear or non-linear trend factors are extracted and superimposed or applied to the model's output layer to adjust the trend of the prediction results, making the output more closely reflect long-term price trends.

[0029] By utilizing the aforementioned mutation features, abnormal electricity price events are identified, and the prediction results are adjusted through an anomaly-sensitive weighting mechanism. Mutation features reflect outliers, drastic fluctuations, or unforeseen events in the price curve, such as extreme weather, power grid accidents, and sudden adjustments to electricity price policies. To enhance the model's ability to perceive and respond to sudden events, an anomaly-sensitive weighting mechanism is designed. Specifically, anomaly detection algorithms, such as the Z-score method and Isolation Forest, are used to mark mutation points in historical data. During the training phase, these samples are assigned higher weights, or during the prediction output phase, the confidence interval of the output value or the predicted value itself is dynamically adjusted based on the confidence coefficient of the mutation signal, thereby reducing the error fluctuation of the model during abnormal periods.

[0030] The electricity price forecasting model, which has undergone periodic modeling, trend correction, and abrupt change calibration, can output more stable, accurate market electricity price forecast information with multi-scale dynamic response capabilities. It covers day-ahead market prices and real-time market prices and can be used to guide application scenarios such as pricing strategy optimization, energy storage charging and discharging scheduling, and economic dispatch of power assets.

[0031] Based on the aforementioned photovoltaic power generation forecast, energy storage system charge forecast, and market electricity price forecast information, a day-ahead market pricing strategy is generated, including: Based on the predicted photovoltaic power generation and the predicted charge of the energy storage system, a forecast curve for available electricity is constructed. The constraints of this forecast curve include the photovoltaic output capacity and the charging / discharging capacity boundaries of the energy storage system. This forecast curve represents the maximum electricity capacity that the system can sell to the market in each future time period, serving as the capacity boundary for pricing.

[0032] The available electricity forecast curve and market electricity price forecast information are input into the pricing strategy generation model to generate a day-ahead market pricing strategy, including: The forecast period is divided into multiple pricing decision cycles; the entire day-ahead market forecast period is divided into several pricing decision cycles, for example, in units of 1 hour or 15 minutes. Each cycle is an independent pricing point, corresponding to a set of decisions regarding cleared electricity volume and pricing price.

[0033] Within each bidding decision cycle, based on the available electricity forecast curve, market electricity price forecast information, and the charging and discharging capacity boundary of the energy storage system, a clearing electricity volume and bidding price pair are generated. Within each bidding cycle, the model makes strategy judgments based on the upper limit of available electricity volume, the market electricity price forecast, and the energy storage charging and discharging capacity for that cycle. If the electricity price is higher than a preset threshold or higher than the average trend, the system prioritizes discharging or sells photovoltaic electricity to the maximum extent; if the electricity price is lower, the strategy may favor energy storage charging or retaining electricity. Considering the electricity volume boundary and electricity price forecast, the expected clearing electricity volume and corresponding bidding price for each bidding cycle are output, forming a preliminary bidding pair.

[0034] Based on the profit maximization objective, the cleared electricity volume and quoted price pair are optimized to generate a day-ahead market pricing strategy. The profit maximization objective is used to optimize the price pair across all pricing periods. The objective can be set to maximize total revenue, i.e., maximizing the sum of total electricity volume multiplied by the price, without considering other factors. The final output is a complete day-ahead market pricing strategy, including electricity volume and quoted price matching information for multiple time periods, which can be directly used as a basis for participating in day-ahead market pricing decisions. This not only enhances the system's competitiveness in the spot market but also strengthens the pricing strategy's adaptability to renewable energy fluctuations and market price uncertainties, contributing to maximizing system revenue and long-term operational stability.

[0035] In the real-time market, a real-time pricing strategy is generated based on real-time environmental data, day-ahead market pricing strategies, and the real-time status of energy storage charge, including: Based on real-time environmental data, the photovoltaic power generation deviation value is obtained; the photovoltaic power generation deviation value is the difference between the current actual power generation value and the day-ahead photovoltaic power generation prediction value, reflecting the direction and magnitude of the prediction error.

[0036] Based on the real-time charge status of the energy storage system, the charge deviation value of the energy storage system is obtained; Based on the photovoltaic power generation deviation value and the energy storage system charge deviation value, the day-ahead market pricing strategy is optimized to generate a real-time pricing strategy, including: Based on the deviation values ​​of photovoltaic power generation and the charge deviation values ​​of the energy storage system, a set of real-time pricing adjustment factors is established. The set of real-time pricing adjustment factors includes a photovoltaic output offset factor, which increases or decreases the available clearing power according to the direction (positive / negative) and magnitude of the deviation; and an energy storage charge state correction factor, which assesses whether the current energy storage can support the original discharge plan or whether it has the ability to absorb excess photovoltaic power.

[0037] Based on the set of adjustment factors, the expected change in revenue for each time period of the day-ahead market pricing strategy is calculated; by comparing the difference between the current executable output and the original bid-cleared power, and in conjunction with the revised market electricity price forecast, the revenue increase or decrease trend for each time period is assessed.

[0038] Based on the principle of maximizing profits and the operational constraints of the energy storage system, the clearing volume and bid price pair of the day-ahead bidding strategy are adjusted to generate a real-time bidding strategy, including: appropriately reducing the bid volume or increasing the bid price to avoid losses when photovoltaic output is insufficient; increasing the discharge power to obtain higher profits when energy storage capacity is sufficient and electricity price increases are expected; and proactively reducing the bid volume or participating in the ancillary services market when electricity prices are predicted to be low or the risk of system overload is high.

[0039] The final output real-time pricing strategy is an adjusted time-series pricing dataset, including the electricity bid value and corresponding bid price for each decision period. This dataset can be submitted to the electricity spot market system for execution in real time or used to adjust internal energy management plans. The method in this embodiment enables flexible adjustments based on the actual system state without altering the day-ahead strategy structure, giving the photovoltaic-storage system stronger market adaptability and adjustment capabilities, effectively improving revenue stability and operational economy.

[0040] This invention integrates photovoltaic module parameters, environmental data, and historical power generation data to improve the accuracy of photovoltaic output prediction and effectively reduce market risks caused by fluctuations in new energy output. Simultaneously, by constructing a state-of-charge (SOC) prediction model for energy storage, the adjustability of the energy storage system can be assessed in advance, providing a reliable basis for strategy formulation. Combining historical market electricity prices and load demand data, a forward-looking and economical day-ahead pricing strategy is generated. Furthermore, the day-ahead strategy is dynamically adjusted in real-time based on environmental changes and the real-time status of the energy storage system, making real-time pricing more closely aligned with actual operating conditions. The overall solution achieves coordinated unity between day-ahead prediction and real-time response, improving the flexibility of the pricing strategy and the economic efficiency of system operation. By fully leveraging the synergistic effect of photovoltaics and energy storage, this invention can significantly improve the efficiency of new energy utilization and the profitability in the spot market, demonstrating promising prospects for widespread application.

[0041] It should be understood that the specific embodiments described above are merely illustrative or explanatory of the principles of the invention and do not constitute a limitation thereof. Therefore, any modifications, equivalent substitutions, improvements, etc., made without departing from the spirit and scope of the invention should be included within the protection scope of the invention. Furthermore, the appended claims are intended to cover all variations and modifications falling within the scope and boundaries of the appended claims, or equivalent forms of such scope and boundaries.

[0042] The present invention has been described above with reference to embodiments thereof. However, these embodiments are merely illustrative and not intended to limit the scope of the invention. The scope of the invention is defined by the appended claims and their equivalents. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of the invention, and all such substitutions and modifications should fall within the scope of the invention.

[0043] Although embodiments of the present invention have been described in detail, it should be understood that various changes, substitutions, and modifications can be made to the embodiments of the present invention without departing from the spirit and scope of the invention.

[0044] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for optimizing day-ahead and real-time pricing strategies for a photovoltaic-storage DC integrated system in the electricity spot market, characterized in that, include: Based on photovoltaic module parameters, environmental data, and historical power generation data, short-term forecasts of photovoltaic output are made to obtain photovoltaic power generation forecasts. Based on the preset energy storage charge prediction model, the predicted charge value of the energy storage system is obtained; Based on historical electricity prices and load demand data in the electricity spot market, market electricity price forecasting information is obtained; Based on the photovoltaic power generation forecast, the energy storage system charge forecast, and the market electricity price forecast, a day-ahead market pricing strategy is generated. In the real-time market, a real-time pricing strategy is generated based on real-time environmental data, day-ahead market pricing strategies, and the real-time status of energy storage charge.

2. The method for optimizing day-ahead and real-time pricing strategies of the photovoltaic-storage DC integrated system in the electricity spot market according to claim 1, characterized in that, The method of making short-term predictions of photovoltaic output based on photovoltaic module parameters, environmental data, and historical power generation data to obtain predicted photovoltaic power generation values ​​includes: The parameters of the photovoltaic module include module area, conversion efficiency and installation angle; the environmental data include solar irradiance and ambient temperature. A photovoltaic power generation prediction model is constructed. The photovoltaic power generation prediction model is based on a deep neural network. The input of the photovoltaic power generation prediction model is the component area, conversion efficiency, installation angle, solar irradiance and ambient temperature. The output of the photovoltaic power generation prediction model is the photovoltaic power generation prediction value. The photovoltaic power generation prediction model is trained using the historical power generation data, and the predicted value of photovoltaic power generation is obtained using the trained photovoltaic power generation prediction model.

3. The method for optimizing day-ahead and real-time pricing strategies of the photovoltaic-storage DC integrated system in the electricity spot market according to claim 1, characterized in that, The process of obtaining the predicted charge value of the energy storage system based on the preset energy storage charge prediction model includes: Obtain the initial state of charge, rated capacity, photovoltaic power generation forecast, and load-side electricity consumption forecast of the energy storage system; A state of charge (SOC) prediction model is constructed. The SOC prediction model is based on a deep neural network. The inputs of the SOC prediction model are the initial SOC, rated capacity, predicted photovoltaic power generation, and predicted load-side electricity consumption. The output of the SOC prediction model is the predicted SOC of the energy storage system.

4. The method for optimizing day-ahead and real-time pricing strategies of the photovoltaic-storage DC integrated system in the electricity spot market according to claim 1, characterized in that, The market electricity price forecast information obtained based on historical electricity price and load demand data from the electricity spot market includes: Obtain historical electricity price data and corresponding load demand data for the electricity spot market within the target area; Extract time series features from the historical electricity price data and load demand data, including periodic features, trend features, and abrupt change features; Based on the periodicity, trend, and abrupt change characteristics, market electricity price forecast information is obtained using a pre-defined electricity price forecasting model.

5. The method for optimizing day-ahead and real-time pricing strategies of the photovoltaic-storage DC integrated system in the electricity spot market according to claim 4, characterized in that, The step of obtaining market electricity price forecast information based on the periodic characteristics, trend characteristics, and abrupt change characteristics, using a preset electricity price forecasting model, includes: The pre-defined electricity price prediction model is structurally initialized and periodic factors are embedded using the periodic features to obtain the modeled electricity price prediction model. The modeled electricity price prediction model is dynamically corrected and its output is calibrated using trend and abrupt change characteristics.

6. The method for optimizing day-ahead and real-time pricing strategies of the photovoltaic-storage DC integrated system in the electricity spot market according to claim 5, characterized in that, The modeled electricity price prediction model is dynamically corrected and its output calibrated using trend and abrupt change characteristics, including: The long-term trend offset is constructed using the aforementioned trend characteristics, and the output of the modeled electricity price prediction model is adjusted linearly based on this trend. By utilizing the aforementioned mutation characteristics, abnormal electricity price events are identified, and the prediction results are adjusted through an anomaly-sensitive weighting mechanism.

7. The method for optimizing day-ahead and real-time pricing strategies of the photovoltaic-storage DC integrated system in the electricity spot market according to claim 1, characterized in that, Based on the aforementioned photovoltaic power generation forecast, energy storage system charge forecast, and market electricity price forecast information, a day-ahead market pricing strategy is generated, including: Based on the photovoltaic power generation forecast and the energy storage system charge forecast, an available power forecast curve is constructed. The constraints of the available power forecast curve include the photovoltaic output capacity and the charging and discharging capacity boundary of the energy storage system. The available electricity forecast curve and market electricity price forecast information are input into the pricing strategy generation model to generate the day-ahead market pricing strategy.

8. The method for optimizing day-ahead and real-time pricing strategies of the photovoltaic-storage DC integrated system in the electricity spot market according to claim 7, characterized in that, The available electricity forecast curve and market electricity price forecast information are input into the pricing strategy generation model to generate a day-ahead market pricing strategy, including: The forecast period is divided into multiple pricing decision cycles; Based on the available electricity forecast curve, market electricity price forecast information and the charging and discharging capacity boundary of the energy storage system, a clearing electricity volume and bid price pair are generated in each bidding decision cycle. Based on the goal of maximizing profits, the cleared electricity volume and bid price pair are optimized to generate a day-ahead market bidding strategy.

9. The method for optimizing day-ahead and real-time pricing strategies of the photovoltaic-storage DC integrated system in the electricity spot market according to claim 8, characterized in that, In the real-time market, a real-time pricing strategy is generated based on real-time environmental data, day-ahead market pricing strategies, and the real-time status of energy storage charge, including: The deviation value of photovoltaic power generation is obtained based on real-time environmental data; Based on the real-time charge status of the energy storage system, the charge deviation value of the energy storage system is obtained; Based on the photovoltaic power generation deviation value and the energy storage system charge deviation value, the day-ahead market pricing strategy is optimized to generate a real-time pricing strategy.

10. The method for optimizing day-ahead and real-time pricing strategies of the photovoltaic-storage DC integrated system in the electricity spot market according to claim 9, characterized in that... The step of optimizing the day-ahead market pricing strategy based on the photovoltaic power generation deviation value and the energy storage system charge deviation value to generate a real-time pricing strategy includes: A set of real-time pricing adjustment factors is established based on the deviation values ​​of photovoltaic power generation and the charge deviation values ​​of energy storage systems; Based on the set of adjustment factors, calculate the expected change in the price return of the daytime market pricing strategy for each time period; Based on the principle of maximizing profits and the operational constraints of the energy storage system, the cleared electricity volume and bid price pair of the day-ahead bidding strategy are adjusted to generate a real-time bidding strategy.

Images