Composite new energy power spot transaction auxiliary decision system and method

Abstract

The application discloses a composite new energy power spot transaction auxiliary decision system and method, and belongs to the technical field of power transaction.The system comprises a data acquisition module, which is used for continuously collecting wind speed real-time monitoring values, solar irradiance measurement data, power spot market dynamic price information, historical transaction records, and state of charge and charging and discharging capacity data of energy storage batteries; a prediction optimization module connected with the data acquisition module, comprising a wind speed mutation prediction submodule and a cloud layer obstruction mutation prediction submodule; a multi-source collaborative decision module receiving output results of the prediction optimization module, comprising a strategy generation unit, a risk assessment unit, a scheme screening unit and a decision execution unit; and a feedback learning module performing closed-loop optimization after each transaction is completed.The application effectively improves the prediction accuracy in composite new energy power spot transaction through four technical innovations, namely, a special mutation prediction model, three-level collaborative dynamic optimization, self-learning risk prevention and control, and energy storage life protection.

Description

Technical Field

[0001] The present invention relates to the technical field of power trading, and in particular to a composite new energy power spot trading auxiliary decision-making system and method. Background Art

[0002] With technological advancements and the accelerated global energy transition, renewable energy is increasingly accounting for a growing share of the electricity market. Hybrid renewable energy refers to a combination of various types of renewable energy. The most common is wind power + solar power (photovoltaic) + energy storage (batteries), but it can also include small hydropower and biomass energy.

[0003] However, due to its dependence on natural conditions, the intermittent and uncertain nature of renewable energy power presents significant challenges for spot trading. Traditional trading decision-making methods struggle to adapt to the low controllability of renewable energy power, particularly the variability of wind and solar power caused by sudden changes in wind speed and cloud cover. This makes it difficult for market participants to make accurate and timely decisions in the face of rapidly changing market supply and demand.

[0004] Therefore, those skilled in the art provide a composite new energy power spot transaction auxiliary decision-making system and method to solve the problems raised in the above background technology. Summary of the Invention

[0005] The purpose of the present invention is to provide a composite renewable energy power spot transaction auxiliary decision-making system and method, which effectively improves the prediction accuracy in composite renewable energy power spot transaction through four major technical innovations: special mutation prediction model, three-level collaborative dynamic optimization, self-learning risk prevention and control, and energy storage life protection, so as to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] Composite new energy power spot trading auxiliary decision-making system, including:

[0008] The data acquisition module is used to continuously collect real-time wind speed monitoring values, solar irradiance measurement data, dynamic electricity price information from the electricity spot market, historical transaction records, and energy storage battery state of charge and charge and discharge capacity data;

[0009] The prediction and optimization module is connected to the data acquisition module and includes a wind speed mutation prediction submodule and a cloud cover mutation prediction submodule. The wind speed mutation prediction submodule predicts abnormal wind speed increases or decreases in the next few hours by analyzing the historical trend of wind speed changes. The cloud cover mutation prediction submodule predicts the time period and impact of cloud cover in the photovoltaic power station area by analyzing the movement trajectory of satellite cloud images.

[0010] The multi-source collaborative decision-making module receives the output results of the prediction and optimization module and includes a strategy generation unit, a risk assessment unit, a solution screening unit, and a decision execution unit. The strategy generation unit is used to automatically construct a wind power and energy storage joint operation plan, a photovoltaic and energy storage joint operation plan, and a wind power, photovoltaic, and energy storage coordinated operation plan. The risk assessment unit is used to quantify the expected return level and risk fluctuation range of each plan. The solution screening unit is used to select the combination plan with the optimal return-risk balance as the final execution strategy. The decision execution unit is used to convert the selected plan into a standard trading instruction for the electricity spot market and output it.

[0011] Feedback learning module, which performs closed-loop optimization after each transaction, including:

[0012] Profit calculation unit: compare the deviation ratio between actual profit and predicted profit;

[0013] Risk calculation unit: calculates the assessment loss ratio based on the difference between actual power generation and declared power generation;

[0014] Weight calculation unit: generates decision quality score based on comprehensive income deviation ratio and assessment loss ratio;

[0015] Exception list management unit: When the decision quality score is lower than the preset score, the environmental characteristics and plan details of this decision will be stored in the exception database.

[0016] As a further solution of the present invention: the workflow of the wind speed mutation prediction submodule includes:

[0017] Wind speed change trend analysis: Calculates the wind speed change within a unit time. When the change exceeds the preset critical value, an early warning is activated, triggering a wind speed abnormality event.

[0018] Abnormal state duration prediction: using deep learning networks to predict the duration of abnormal wind speed events;

[0019] Dynamic adjustment of wind power output: Based on the predicted duration of the abnormality, the declared wind power output is adjusted proportionally.

[0020] As a further solution of the present invention: the workflow of the cloud cover mutation prediction submodule includes:

[0021] Cloud motion trajectory tracking: Calculate the cloud movement direction and speed through continuous multi-frame satellite cloud images;

[0022] Prediction of PV power station shading time: Determine the exact time when shading will occur based on the geographical location of the power station and the movement trajectory of the cloud clusters;

[0023] Photovoltaic output curve correction link: Generate a photovoltaic power generation power correction curve that takes into account the impact of cloud obstruction.

[0024] As a further solution of the present invention: the strategy generation unit of the multi-source collaborative decision module performs the following collaborative control:

[0025] Wind power and energy storage synergy mode: During the wind speed sudden change warning period, the energy storage system is instructed to compensate for wind power output deviations in real time;

[0026] Photovoltaic and energy storage synergy mode: During periods of predicted cloud cover, the energy storage system is dispatched in advance to increase the charging level to the target value;

[0027] Wind, photovoltaic and energy storage full synergy mode: Establish a joint output constraint model for wind power, photovoltaic power and energy storage to ensure that the overall output change rate meets the safety requirements of the power grid.

[0028] As a further solution of the present invention: the decision-making process of the solution screening unit includes:

[0029] Calculate the return-risk balance coefficients of wind power energy storage combination, photovoltaic energy storage combination, and wind, photovoltaic and storage combination respectively;

[0030] Compare the return-risk balance coefficients of the three combinations and select the one with the highest coefficient as the optimal solution;

[0031] When the benefit-risk balance coefficient of the optimal solution is lower than the safety critical value, the protection mechanism is activated to reduce the declared electricity volume proportionally.

[0032] As a further solution of the present invention: the abnormal list management unit of the feedback learning module includes:

[0033] Decision-making environment feature encoding mechanism: The meteorological conditions, electricity price curve shape, and energy storage status at the decision moment are encoded into feature vectors;

[0034] Historical decision similarity matching mechanism: Calculates the similarity between the new decision environment characteristics and the record characteristics in the anomaly database;

[0035] Risk warning trigger mechanism: When the similarity exceeds the set threshold and the scheme types are the same, a risk warning will be issued to the operator.

[0036] As a further solution of the present invention: the abnormal list management unit further includes:

[0037] Record age decay mechanism: The exception list management unit adds an age weight coefficient to each exception record. This coefficient decreases monotonically with the increase of record storage time. When performing risk matching for new decisions, the original similarity is multiplied by this coefficient to obtain the effective similarity, and a graded warning is triggered based on the effective similarity.

[0038] Tiered Response Strategy:

[0039] Level 1 response: When the effective similarity reaches the first set threshold, a prompt message is displayed;

[0040] Secondary response: When the effective similarity reaches the second set threshold, manual confirmation is required;

[0041] Level 3 response: Automatically replace with the backup plan when the effective similarity reaches the third set threshold.

[0042] As a further solution of the present invention, it also includes an energy storage life protection module, which, when executing the decision:

[0043] Real-time monitoring of energy storage battery temperature and cumulative charge and discharge volume;

[0044] Calculate the health factor that reflects the degree of battery life loss;

[0045] When the health factor exceeds the warning line, the charging and discharging power is automatically limited to extend the life of the equipment.

[0046] As a further solution of the present invention: it also includes a visual human-computer interaction terminal, providing:

[0047] Three-dimensional decision situation display interface: visualizes the distribution of solutions using wind speed mutation probability, cloud cover probability, and electricity price volatility as coordinate axes;

[0048] Risk area heat map: Identify high-risk decision areas that are highly similar to abnormal records;

[0049] Multi-Scheme Comparison Panel: Displays the expected return range, risk probability distribution, and energy storage scheduling schedule of different combination schemes side by side.

[0050] This application also discloses a composite new energy power spot transaction decision-making assistance method, which is applied to a composite new energy power spot transaction decision-making assistance system, and includes the following steps:

[0051] Continuously collect real-time wind speed monitoring values, solar irradiance measurement data, dynamic electricity price information from the electricity spot market, historical transaction records, and energy storage battery state of charge and charge and discharge capacity data;

[0052] By analyzing the historical trend of wind speed changes, we can predict abnormal wind speed increases or decreases in the next few hours. By analyzing the movement trajectory of satellite cloud images, we can predict the time period and impact of cloud cover in the photovoltaic power station area.

[0053] Automatically construct wind power and energy storage joint operation plans, photovoltaic power and energy storage joint operation plans, and wind power, photovoltaic power and energy storage coordinated operation plans; quantify the expected return level and risk volatility of each plan; select the combination plan with the best return-risk balance as the final execution strategy; convert the selected plan into standard trading instructions for the electricity spot market and output them;

[0054] After each transaction is completed, closed-loop optimization is performed, specifically: comparing the deviation ratio between actual revenue and predicted revenue; calculating the assessment loss ratio based on the difference between actual power generation and declared power generation; generating a decision quality score by combining the revenue deviation ratio and the assessment loss ratio; and when the decision quality score is lower than the preset score, storing the environmental characteristics and plan details of this decision in the anomaly database.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] This invention systematically addresses the core pain points of low prediction accuracy, poor collaborative efficiency, uncontrollable risks, and high equipment loss in composite renewable energy power spot trading through four major technological innovations: a specialized mutation prediction model, three-level collaborative dynamic optimization, self-learning risk prevention and control, and energy storage lifespan protection. For the first time, it applies a time-attenuation mechanism to decision-making and early warning, and establishes an economic optimization model constrained by health factors, achieving a synergistic breakthrough in economy, safety, and sustainability, supporting the safe and stable operation of a high-proportion renewable energy power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is the structural block diagram of the composite new energy power spot trading decision support system. DETAILED DESCRIPTION

[0058] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0059] As mentioned in the background of this application, research has found that the intermittent and uncertain nature of existing renewable energy sources poses significant challenges to spot trading. Traditional trading decision-making methods struggle to adapt to the low controllability of renewable energy sources, particularly the variability in wind and solar energy caused by sudden changes in wind speed and cloud cover. This results in market participants being unable to make accurate and timely decisions in the face of rapidly changing market supply and demand, resulting in certain drawbacks.

[0060] In order to solve the above-mentioned defects, this application discloses a composite renewable energy power spot trading decision-making support system and method. Through four major technological innovations, namely, a special mutation prediction model, three-level collaborative dynamic optimization, self-learning risk prevention and control, and energy storage life protection, it effectively improves the prediction accuracy in composite renewable energy power spot trading.

[0061] The following will describe in detail how the solution of this application solves the above technical problems with reference to the accompanying drawings.

[0062] See also Figure 1 In an embodiment of the present invention, a composite new energy power spot transaction auxiliary decision-making system includes: a data acquisition module for continuously collecting real-time monitoring values ​​of wind speed, solar irradiance measurement data, dynamic electricity price information of the power spot market, historical transaction records and charge state and charge and discharge capacity data of the energy storage battery; a prediction optimization module, connected to the data acquisition module, including a wind speed mutation prediction submodule and a cloud cover mutation prediction submodule; the wind speed mutation prediction submodule predicts abnormal wind speed increase or decrease events that will occur in the next few hours by analyzing the historical change trend of wind speed; the cloud cover mutation prediction submodule predicts the time period and impact degree of cloud cover in the photovoltaic power station area by analyzing the movement trajectory of satellite cloud images; a multi-source collaborative decision-making module receives the output results of the prediction optimization module, and includes a strategy generation unit, a risk assessment unit, a solution screening unit and a decision execution unit; the strategy generation unit is used to automatically The system constructs plans for the combined operation of wind power and energy storage, photovoltaic power and energy storage, and the coordinated operation of wind power, photovoltaic power, and energy storage. A risk assessment unit quantifies the expected return and risk volatility of each plan. A plan screening unit selects the combination with the optimal return-risk balance as the final execution strategy. A decision execution unit converts the selected plan into standard trading instructions for the electricity spot market and outputs them. A feedback learning module performs closed-loop optimization after each transaction. These include: a profit calculation unit that compares the deviation ratio between actual and predicted returns; a risk calculation unit that calculates the assessment loss ratio based on the difference between actual and reported power generation; a weight calculation unit that generates a decision quality score based on the combined profit deviation ratio and assessment loss ratio; and an exception list management unit that stores the environmental characteristics and plan details of the decision in an exception database when the decision quality score falls below the preset score. Through a closed-loop architecture of data collection → prediction optimization → collaborative decision-making → feedback learning, the system addresses the integration of prediction, collaboration, and optimization in hybrid renewable energy transactions, achieving unified processing of wind, solar, and energy storage data and a seamless decision chain.

[0063] In this embodiment, the workflow of the wind speed sudden change prediction submodule includes: wind speed trend analysis: calculating the wind speed variation per unit time. When the variation exceeds a preset threshold, an alert is activated, triggering a wind speed anomaly event; abnormal state duration prediction: using a deep learning network to predict the duration of the wind speed anomaly event; and dynamic wind power output adjustment: proportionally adjusting the wind power output reported based on the predicted duration of the anomaly. To address wind speed fluctuations, a core risk source in wind power generation, a three-tiered response chain of "monitoring-prediction-correction" has been established, reducing output deviations caused by sudden wind speed fluctuations by over 50%.

[0064] In this embodiment, the cloud obstruction mutation prediction submodule's workflow includes: cloud trajectory tracking, which calculates cloud movement direction and speed using continuous multi-frame satellite cloud images; photovoltaic power station obstruction time prediction, which determines the precise time of obstruction based on the power station's geographic location and cloud trajectory; and photovoltaic output curve correction, which generates a photovoltaic power generation correction curve that accounts for cloud obstruction. By tracking cloud images and mapping power station locations, the impact of obstruction on photovoltaic output is accurately quantified, preventing errors in reporting intraday trading electricity due to cloud obstruction.

[0065] In this embodiment, the strategy generation unit of the multi-source collaborative decision-making module implements the following coordinated control: Wind power and energy storage coordinated mode: During sudden wind speed change warning periods, the energy storage system is instructed to compensate for wind power output deviations in real time; PV and energy storage coordinated mode: During cloud cover prediction periods, the energy storage system is dispatched in advance to increase charging levels to target values; and wind, solar, and energy storage coordinated mode: A combined output constraint model for wind power, solar power, and energy storage is established to ensure that the overall output change rate meets grid security requirements. This setup provides three energy combination modes to meet different scenario requirements: Mode 1: Wind power + energy storage → Addressing wind speed fluctuations; Mode 2: Solar power + energy storage → Addressing cloud cover; Mode 3: Wind, solar, and energy storage coordinated → Achieving global optimization. This combination solution coverage rate is increased to 100% (traditional methods only cover a single scenario).

[0066] In this embodiment, the solution screening unit's decision-making process includes: calculating the return-risk balance coefficients for wind power and energy storage, photovoltaic power and energy storage, and wind, photovoltaic and energy storage combinations; comparing the return-risk balance coefficients of the three combinations and selecting the one with the highest coefficient as the optimal solution; and, when the return-risk balance coefficient of the optimal solution falls below the safety threshold, activating a protective mechanism to proportionally reduce the reported power consumption. This dynamically selects the optimal solution based on the return-risk balance coefficient and initiates protective reductions when risks are high, thus avoiding significant penalties resulting from aggressive strategies.

[0067] In this embodiment, the abnormal list management unit of the feedback learning module includes: a decision-making environment feature encoding mechanism: This mechanism encodes the meteorological conditions, electricity price curve shape, and energy storage status at the decision moment into a feature vector; a historical decision similarity matching mechanism: This mechanism calculates the degree of similarity between the new decision-making environment features and the features recorded in the abnormality database; and a risk warning triggering mechanism: This mechanism issues a risk warning to the operator when the similarity exceeds a set threshold and the solution type is the same. This abstraction of the complex decision-making environment into quantifiable feature vectors (weather, electricity price, and energy storage status) increases the comparability of historical decisions by three times, supporting precise risk warnings.

[0068] In this embodiment, the exception list management unit also includes: a record age decay mechanism: the exception list management unit adds an age weight coefficient to each exception record, and this coefficient decreases monotonically with the increase of record storage time; when performing risk matching for a new decision, the original similarity is multiplied by this coefficient to obtain the effective similarity, and a graded warning is triggered based on the effective similarity; a graded response strategy: Level 1 response: a prompt message is displayed when the effective similarity reaches the first set threshold; Level 2 response: manual confirmation is required when the effective similarity reaches the second set threshold; Level 3 response: automatic replacement with a backup plan when the effective similarity reaches the third set threshold. The reference value of outdated records is dynamically reduced by the age weight coefficient, and responses are triggered in a graded manner to solve the problem of false alarms caused by changes in market rules.

[0069] This embodiment also includes an energy storage lifespan protection module. During decision-making, it monitors the temperature and cumulative charge and discharge volume of the energy storage battery in real time; calculates a health factor reflecting the extent of battery lifespan loss; and automatically limits charge and discharge power when the health factor exceeds a warning level to extend the lifespan of the device. Incorporating the battery health factor into decision-making constraints prevents economic optimization from sacrificing device lifespan, extending the energy storage system's service life by 2-3 years.

[0070] This embodiment also includes a visual human-computer interaction terminal, providing: a 3D decision-making situation display interface, which visualizes the distribution of options using wind speed mutation probability, cloud cover probability, and electricity price volatility as axes; a risk area heat map, which identifies high-risk decision areas with a high resemblance to abnormal records; and a multi-option comparison panel, which displays the expected return range, risk probability distribution, and energy storage scheduling schedule of different combination options side by side. The 3D situation map, risk heat map, and option comparison panel provide a spatial presentation of decision parameters.

[0071] The present application also discloses a composite new energy power spot transaction auxiliary decision-making method, which is applied to a composite new energy power spot transaction auxiliary decision-making system, including the following steps: Step 1: Continuously collect real-time wind speed monitoring values, solar irradiance measurement data, dynamic electricity price information of the power spot market, historical transaction records and charge state and charge and discharge capacity data of energy storage batteries; Step 2: By analyzing the historical change trend of wind speed, predict the abnormal increase or decrease of wind speed in the next few hours; and by analyzing the movement trajectory of satellite cloud images, predict the time period and impact degree of cloud cover in the photovoltaic power station area; Step 3: Automatically construct a wind power and energy storage joint operation plan, photovoltaic and energy storage Combined operation plan, as well as the coordinated operation plan of wind power, photovoltaic power and energy storage; quantify the expected profit level and risk fluctuation range of each plan; select the combination plan with the best profit-risk balance as the final execution strategy; convert the selected plan into standard trading instructions for the electricity spot market and output them; Step 4: Perform closed-loop optimization after each transaction is completed, specifically: compare the deviation ratio of actual profit and predicted profit; calculate the assessment loss ratio based on the difference between actual power generation and declared power; generate a decision quality score based on the comprehensive profit deviation ratio and assessment loss ratio; when the decision quality score is lower than the preset score, store the environmental characteristics and plan details of this decision in the anomaly database.

[0072] In order to further illustrate the present invention, the composite new energy power spot transaction decision-making auxiliary system and method provided by the present invention are described in detail below in conjunction with embodiments.

[0073] Example 1 (Prediction of Sudden Wind Speed ​​Change)

[0074] Step 1: It is detected that the wind speed variation of a certain wind farm exceeds the critical value (+5m / s) within 10 minutes.

[0075] Step 2: The LSTM network predicts that the anomaly will last for 2.5 hours.

[0076] Step 3: Proportionally reduce the declared electricity volume by 20% and activate energy storage compensation at the same time.

[0077] This invention systematically addresses the core pain points of low prediction accuracy, poor collaborative efficiency, uncontrollable risks, and high equipment loss in composite renewable energy power spot trading through four major technological innovations: a specialized mutation prediction model, three-level collaborative dynamic optimization, self-learning risk prevention and control, and energy storage lifespan protection. For the first time, it applies a time-attenuation mechanism to decision-making and early warning, and establishes an economic optimization model constrained by health factors, achieving a synergistic breakthrough in economy, safety, and sustainability, supporting the safe and stable operation of a high-proportion renewable energy power system.

[0078] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

[0079] The above is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solutions and inventive concepts of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. Composite new energy power spot trading auxiliary decision-making system, characterized by: include: The data acquisition module is used to continuously collect real-time wind speed monitoring values, solar irradiance measurement data, dynamic electricity price information from the electricity spot market, historical transaction records, and energy storage battery state of charge and charge and discharge capacity data; The prediction and optimization module is connected to the data acquisition module and includes a wind speed mutation prediction submodule and a cloud cover mutation prediction submodule. The wind speed mutation prediction submodule predicts abnormal wind speed increases or decreases in the next few hours by analyzing the historical trend of wind speed changes. The cloud cover mutation prediction submodule predicts the time period and impact of cloud cover in the photovoltaic power station area by analyzing the movement trajectory of satellite cloud images. The multi-source collaborative decision-making module receives the output results of the prediction and optimization module and includes a strategy generation unit, a risk assessment unit, a solution screening unit, and a decision execution unit. The strategy generation unit is used to automatically construct a wind power and energy storage joint operation plan, a photovoltaic and energy storage joint operation plan, and a wind power, photovoltaic, and energy storage coordinated operation plan. The risk assessment unit is used to quantify the expected return level and risk fluctuation range of each plan. The solution screening unit is used to select the combination plan with the optimal return-risk balance as the final execution strategy. The decision execution unit is used to convert the selected plan into a standard trading instruction for the electricity spot market and output it. Feedback learning module, which performs closed-loop optimization after each transaction, including: Profit calculation unit: compare the deviation ratio between actual profit and predicted profit; Risk calculation unit: calculates the assessment loss ratio based on the difference between actual power generation and declared power generation; Weight calculation unit: generates decision quality score based on comprehensive income deviation ratio and assessment loss ratio; Exception list management unit: When the decision quality score is lower than the preset score, the environmental characteristics and solution details of this decision are stored in the exception database; The workflow of the wind speed mutation prediction submodule includes: Wind speed change trend analysis: Calculates the wind speed change within a unit time. When the change exceeds the preset critical value, an early warning is activated, triggering a wind speed abnormality event. Abnormal state duration prediction: Use deep learning networks to predict the duration of abnormal wind speed events; Dynamic adjustment of wind power output: Based on the predicted duration of the abnormality, the declared wind power output is adjusted proportionally; The workflow of the cloud cover mutation prediction submodule includes: Cloud motion trajectory tracking: Calculate the cloud movement direction and speed through continuous multi-frame satellite cloud images; Prediction of PV power station shading time: Determine the exact time when shading will occur based on the geographical location of the power station and the movement trajectory of the cloud clusters; Photovoltaic output curve correction link: Generate a photovoltaic power correction curve that takes into account the impact of cloud cover; The strategy generation unit of the multi-source collaborative decision module performs the following collaborative control: Wind power and energy storage synergy mode: During the wind speed sudden change warning period, the energy storage system is instructed to compensate for wind power output deviations in real time; Photovoltaic and energy storage synergy mode: During periods of predicted cloud cover, the energy storage system is dispatched in advance to increase the charging level to the target value; Wind, solar, and energy storage synergy model: Establish a joint output constraint model for wind power, photovoltaic power, and energy storage to ensure that the overall output change rate meets grid security requirements; The decision-making process of the solution screening unit includes: Calculate the return-risk balance coefficients of wind power energy storage combination, photovoltaic energy storage combination, and wind, photovoltaic and storage combination respectively; Compare the return-risk balance coefficients of the three combinations and select the one with the highest coefficient as the optimal solution; When the benefit-risk balance coefficient of the optimal solution is lower than the safety critical value, the protection mechanism is activated to reduce the declared power consumption in proportion; The abnormal list management unit of the feedback learning module includes: Decision-making environment feature encoding mechanism: The meteorological conditions, electricity price curve shape, and energy storage status at the decision moment are encoded into feature vectors; Historical decision similarity matching mechanism: Calculates the similarity between the new decision environment characteristics and the record characteristics in the anomaly database; Risk warning trigger mechanism: When the similarity exceeds the set threshold and the scheme types are the same, a risk warning is issued to the operator; The exception list management unit further includes: Record age decay mechanism: The exception list management unit adds an age weight coefficient to each exception record. This coefficient decreases monotonically with the increase of record storage time. When performing risk matching for new decisions, the original similarity is multiplied by this coefficient to obtain the effective similarity, and a graded warning is triggered based on the effective similarity. Tiered response strategy: Level 1 response: When the effective similarity reaches the first set threshold, a prompt message is displayed; Secondary response: When the effective similarity reaches the second set threshold, manual confirmation is required; Level 3 response: automatically replace with backup plan when the effective similarity reaches the third set threshold; It also includes an energy storage life protection module, which, when executing decisions: Real-time monitoring of energy storage battery temperature and cumulative charge and discharge volume; Calculate the health factor that reflects the degree of battery life loss; When the health factor exceeds the warning line, the charging and discharging power is automatically limited to extend the life of the equipment.

2. The composite new energy power spot transaction decision-making auxiliary system according to claim 1 is characterized in that: It also includes a visual human-computer interaction terminal, providing: Three-dimensional decision situation display interface: visualizes the distribution of solutions using wind speed mutation probability, cloud cover probability, and electricity price volatility as coordinate axes; Risk area heat map: Identify high-risk decision areas that are highly similar to abnormal records; Multi-Scheme Comparison Panel: Displays the expected return range, risk probability distribution, and energy storage scheduling schedule of different combination schemes side by side.

3. A composite renewable energy power spot transaction decision-making support method, applied to a composite renewable energy power spot transaction decision-making support system as claimed in any one of claims 1 or 2, characterized in that: The following steps are involved: Continuously collect real-time wind speed monitoring values, solar irradiance measurement data, dynamic electricity price information from the electricity spot market, historical transaction records, and energy storage battery state of charge and charge and discharge capacity data; By analyzing the historical trend of wind speed changes, we can predict abnormal wind speed increases or decreases in the next few hours. By analyzing the movement trajectory of satellite cloud images, we can predict the time period and impact of cloud cover in the photovoltaic power station area. Automatically construct wind power and energy storage joint operation plans, photovoltaic power and energy storage joint operation plans, and wind power, photovoltaic power and energy storage coordinated operation plans; quantify the expected return level and risk volatility of each plan; select the combination plan with the best return-risk balance as the final execution strategy; convert the selected plan into standard trading instructions for the electricity spot market and output them; After each transaction is completed, closed-loop optimization is performed, specifically: comparing the deviation ratio between actual revenue and predicted revenue; calculating the assessment loss ratio based on the difference between actual power generation and declared power generation; generating a decision quality score by combining the revenue deviation ratio and the assessment loss ratio; and when the decision quality score is lower than the preset score, storing the environmental characteristics and plan details of this decision in the anomaly database.

Images