A wind-solar integrated power generation automatic bin charging battery module control system

Abstract

The application relates to the field of new energy and artificial intelligence, and discloses a wind-solar integrated power generation automatic bin adjusting and charging battery module control system, which comprises a power generation bus connected to a wind-solar integrated power generation system at a charging input end, and a charging battery in a charging groove connected to a charging output end, which is used for converting the power supply input by the power generation bus into output power supply with adjustable voltage; the charging groove, a pop-up groove, a plurality of insertion grooves and a battery bin adjusting mechanical structure are included, the insertion grooves are used for storing the charging battery, the bin adjusting mechanical structure sequentially sends the charging battery into the charging groove, and the battery is transferred to the pop-up groove after being fully charged; according to the real-time data of the charging battery obtained by a voltage sampling module, a current sampling module and a temperature monitoring module, the output voltage and current of a direct current conversion module are adjusted, and the charging process is controlled. The application has the advantages of reducing the complexity and volume of the charging device.

Description

Technical Field

[0001] The present invention relates to the fields of new energy and artificial intelligence, and specifically to an automatic storage adjustment rechargeable battery module control system for wind-solar integrated power generation. Background Art

[0002] As global strategic competition in the new energy industry intensifies, lithium rechargeable batteries with multiple advantages have been widely used in the field of new energy batteries. During the charging process of lithium batteries, multiple charging stages are required to cause a series of chemical reactions to occur inside the lithium batteries. The charging voltage and current in different stages are different, and the charging voltage and current need to be finely adjusted and controlled. In order to use rechargeable batteries for a long time and safely, it is also necessary to monitor the battery voltage and temperature, and estimate the remaining battery capacity, etc., so the battery management device is very complex. Traditional parallel or series charging modules charge multiple batteries synchronously, and the current and voltage of each lithium battery cannot be guaranteed to be completely consistent. In addition, it is difficult to ensure the consistency of each lithium battery due to the influence of the output accuracy of the charging equipment, which increases the complexity of the charging operation exponentially. Therefore, it is very necessary to design a rechargeable battery module control system for wind and solar integrated power generation with automatic adjustment to reduce the complexity and volume of the charging device. Summary of the Invention

[0003] In response to the shortcomings of the existing technology, the present invention provides an automatic adjustment and charging battery module control system for wind and solar integrated power generation, which has the advantages of reducing the complexity and volume of the charging device and solves the problems in the above-mentioned background technology.

[0004] To achieve the above-mentioned purpose of reducing the complexity and volume of the charging device, the present invention provides the following technical solution: a control system for automatically adjusting the storage of charging battery modules for wind-solar integrated power generation, comprising:

[0005] DC conversion module: The charging input end is connected to the generating bus of the wind power integrated power generation system, and the charging output end is connected to the rechargeable battery in the charging tank. It is used to convert the power input from the generating bus into output power with adjustable voltage;

[0006] Automatic battery adjustment mechanism: includes a charging slot, an ejection slot, multiple slots, and a battery adjustment mechanism. The slots are used to store batteries to be charged. The adjustment mechanism sequentially feeds the batteries to be charged into the charging slots and transfers the batteries to the ejection slots after they are fully charged.

[0007] The control module adjusts the output voltage and current of the DC conversion module and controls the charging process according to the real-time data of the rechargeable battery obtained by the voltage sampling module, the current sampling module and the temperature monitoring module.

[0008] Preferably, the control module executes the control process including:

[0009] After starting the system, check the status of each submodule;

[0010] Detect whether the charging slot has a battery inserted. If there is no battery, the battery in the slot is called into the charging slot through the automatic adjustment structure;

[0011] During the charging process, the battery voltage, current and temperature data are continuously collected;

[0012] Dynamically adjust the charging voltage and current according to the target parameters corresponding to the charging stage to achieve multi-stage fine charging;

[0013] After the battery is fully charged, the automatic adjustment mechanism is controlled to transfer the battery to the ejection slot and introduce the next battery to continue charging.

[0014] Preferably, the automatic adjustment structure has a cyclic charging function, which supports the automatic introduction of a new battery for the next round of charging after a fully charged battery is ejected from the pop-up slot. It has a standby monitoring mechanism. If there is no battery to be charged in the slot, it automatically enters a low-power standby state, continuously detects the slot status, and automatically wakes up and starts the charging process after a new battery is inserted.

[0015] A control method for automatically adjusting the storage of rechargeable battery modules for wind-solar integrated power generation includes the following steps:

[0016] The input power parameters of the wind-solar integrated power generation system under different operating scenarios are obtained. Combined with the operating parameters of the battery module, a feature aggregation algorithm is used to construct a charging state feature dataset.

[0017] Based on the charging state feature dataset, a hierarchical clustering algorithm and a charging stage identification model are applied to segment the charging process, distinguish typical charging states, and automatically match the target charging parameters for each stage according to the battery type.

[0018] Based on the identified charging stage, a reinforcement learning control strategy is combined to build a real-time voltage and current adjustment mechanism, which dynamically adjusts the output power according to the battery response feedback during the charging process.

[0019] During dynamic charging control, the battery module's response stability indicators are evaluated in real time to monitor whether the charging parameters have reached a stable range. If so, an automatic repositioning strategy is triggered, prioritizing battery packs with similar charging status to join the charging process.

[0020] Apply multi-source scheduling optimization algorithm to dynamically allocate wind and solar power generation resources, make real-time corrections to charging scheduling plans, and generate energy efficiency optimization reports.

[0021] Preferably, the process of constructing a charge state feature dataset using a feature aggregation algorithm is as follows:

[0022] The wind and light integrated power generation system collects photovoltaic power output, grid feeding ratio, wind energy fluctuation parameters and environmental meteorological factors in different operation periods, and synchronously obtains the operation characteristics of the battery module;

[0023] The collected data is subjected to data standardization processing, the scale range of each characteristic data is unified, and the time series sample of the charging state is constructed through the time window mechanism;

[0024] The moving average and weighted dynamic smoothing algorithm are introduced to extract representative statistical characteristics reflecting the actual response behavior of the battery module;

[0025] Based on the multi-dimensional feature fusion mechanism, the wind and light power generation input features and the battery module response features are combined and aggregated to form a multi-source heterogeneous fusion feature vector;

[0026] The multi-dimensional feature vector after aggregation is used to construct the charging state characteristic data set.

[0027] Preferably, the hierarchical clustering algorithm and the charging stage recognition model are applied to the segmented recognition process of the charging process, which is:

[0028] The charging state characteristic data set constructed is input into the hierarchical clustering model, and the Euclidean distance and Mahalanobis distance joint similarity measurement are combined to perform clustering analysis on different charging states;

[0029] The hierarchical depth control parameter of the clustering tree is set, and the charging stage clusters are retained through the pruning strategy;

[0030] For each cluster, the charging stage recognition model is used for label classification to determine whether the label belongs to the typical stage;

[0031] In the recognition process, the historical annotation data and the model supervision verification mechanism are combined;

[0032] Finally, the multi-stage segmented results of the charging process are formed.

[0033] Preferably, the process of automatically matching the target charging parameters according to the battery type is:

[0034] According to the segmented recognition results, the built-in battery type recognition module of the system is called to obtain the type of the battery module currently connected to the charging process;

[0035] According to the identified battery type, the corresponding stage voltage, current and temperature rise safety threshold parameters in the preset charging parameter database are matched;

[0036] In the target parameter matching process, the influence factors are considered to dynamically adjust the adaptive range of the target parameters;

[0037] Combined with the charging stage classification labels, the target parameter templates of each stage are automatically bound to form a complete charging stage target parameter configuration file.

[0038] Preferably, the process of building a voltage and current real-time adjustment mechanism in combination with the reinforcement learning control strategy is as follows:

[0039] Based on the phased target charging parameter profile, the reward function of the reinforcement learning model is set to minimize charging time and maximize energy conversion efficiency;

[0040] The state space is defined as the battery's current SOC, voltage, current, and ambient temperature parameters, and the action space is defined as the amplitude values ​​for adjusting the output power and the charging voltage and current.

[0041] The proximal policy optimization algorithm is used to train the control strategy, and the reinforcement learning model continuously updates the strategy through online feedback;

[0042] During the actual charging process, the control model dynamically generates control actions based on the real-time feedback signals from the battery module.

[0043] Preferably, the process of monitoring whether the charging parameters have reached a stable range is:

[0044] During the implementation of the reinforcement learning-based real-time voltage and current adjustment mechanism, the key response parameters of the currently connected battery module are continuously collected and recorded at multiple points within the set time sliding window;

[0045] For each key parameter, a sliding window statistical algorithm is applied to calculate the mean, standard deviation and volatility within the window period and compared with the preset stability threshold;

[0046] Set up the criteria for judging the stability interval and build a multi-factor stability evaluation index system;

[0047] If the fluctuation rate of the key parameters is less than the set stability threshold in multiple consecutive sliding window cycles, it is determined that the current battery module charging process has entered a convergent and stable state;

[0048] If the fluctuation rate of the key parameter is greater than or equal to the set stability threshold in multiple consecutive sliding window cycles, it is determined that the current battery module charging process has not entered a converged stable state.

[0049] Preferably, the process of generating an energy efficiency optimization report is:

[0050] After completing the charging stage control and automatic position adjustment process, the utilization efficiency, average conversion efficiency and position adjustment response delay of wind and solar power generation resources in the entire charging process are calculated;

[0051] Combined with the energy received by each module during the charging process and the achievement rate of the target charging parameters, the comprehensive energy efficiency index is calculated, including the energy input density per unit time and the number of dispatches per unit power;

[0052] A multi-source scheduling optimization algorithm is introduced to analyze the resource scheduling paths and battery pack switching strategies during the completed charging process, identifying redundant paths and energy waste points.

[0053] Combine system operation logs with charging behavior records to automatically generate energy efficiency optimization reports.

[0054] Compared with the existing technology, the present invention provides a control system for automatic storage adjustment of rechargeable battery modules for wind-solar integrated power generation, which has the following beneficial effects:

[0055] By integrating the multi-source input features of the wind-solar integrated power generation system with the operating status of the battery modules, the present invention uses a feature aggregation algorithm to construct a high-dimensional charging state feature dataset, significantly improving the system's ability to accurately perceive charging behavior under complex operating conditions. A hierarchical clustering algorithm and a stage recognition model are introduced to achieve fine-grained segmentation and classification of the charging process into typical stages, thereby matching the optimal stage target parameters based on the battery type, effectively enhancing the system's adaptability and intelligence. Combined with a reinforcement learning control strategy, a real-time voltage and current adjustment mechanism is constructed to achieve dynamic and precise control of the charging process, optimizing charging efficiency while ensuring battery safety. Furthermore, through real-time monitoring of battery response stability and determination of stable intervals, combined with an automatic repositioning strategy, the balanced access of battery packs and the continuity of system operation are ensured. Furthermore, a multi-source scheduling optimization algorithm is used to dynamically adjust the wind and solar power generation resource allocation path, improving the real-time resource scheduling and energy conversion efficiency. An optimization report is generated containing multi-dimensional energy efficiency indicators, scheduling strategy recommendations, and performance evolution trends, comprehensively improving the intelligent management level of system operation, charging energy efficiency, and operation and maintenance decision-making capabilities. The system has excellent engineering practical value and promotion prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a schematic diagram of the structure of the present invention;

[0057] Figure 2 Schematic diagram of the method of the present invention;

[0058] Figure 3 This is a structural diagram of the automatic storage adjustment module of the present invention;

[0059] Figure 4 It is the workflow diagram of the present invention. DETAILED DESCRIPTION

[0060] 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. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0061] Example 1: Please refer to Figure 1 As shown, an automatic storage adjustment charging battery module control system for wind-solar integrated power generation according to an embodiment of the present invention includes:

[0062] DC conversion module: The charging input end is connected to the generating bus of the wind power integrated power generation system, and the charging output end is connected to the rechargeable battery in the charging tank. It is used to convert the power input from the generating bus into output power with adjustable voltage;

[0063] Automatic battery adjustment mechanism: includes a charging slot, an ejection slot, multiple slots, and a battery adjustment mechanism. The slots are used to store batteries to be charged. The adjustment mechanism sequentially feeds the batteries to be charged into the charging slots and transfers the batteries to the ejection slots after they are fully charged.

[0064] The control module adjusts the output voltage and current of the DC conversion module and controls the charging process according to the real-time data of the rechargeable battery obtained by the voltage sampling module, the current sampling module and the temperature monitoring module.

[0065] The control module executes the control process including:

[0066] After starting the system, check the status of each submodule;

[0067] Detect whether the charging slot has a battery inserted. If no battery is present, the battery in the slot is automatically transferred to the charging slot through the automatic adjustment mechanism.

[0068] During the charging process, the battery voltage, current and temperature data are continuously collected;

[0069] Dynamically adjust the charging voltage and current according to the target parameters corresponding to the charging stage to achieve multi-stage fine charging;

[0070] After the battery is fully charged, the automatic adjustment mechanism is controlled to transfer the battery to the ejection slot and introduce the next battery to continue charging.

[0071] The automatic adjustment structure has a cyclic charging function, which supports automatically introducing a new battery for the next round of charging after a fully charged battery is ejected from the ejection slot. It has a standby monitoring mechanism. If there is no battery to be charged in the slot, it automatically enters a low-power standby state, continuously detects the slot status, and automatically wakes up and starts the charging process after a new battery is inserted.

[0072] Example 2: Figure 2 As shown, a control method for automatically adjusting the storage of charging battery modules for wind-solar integrated power generation includes the following steps:

[0073] S1: Obtain the input power parameters of the wind-solar integrated power generation system under different operating scenarios, combine them with the operating parameters of the battery module, and use the feature aggregation algorithm to construct a charging state feature dataset.

[0074] The process of constructing the charging state feature data set using the feature aggregation algorithm in S1 is as follows:

[0075] Collect the photovoltaic power output, grid feed ratio, wind energy volatility parameters, and environmental meteorological factors of the wind-solar integrated power generation system during different operating periods, and simultaneously obtain the operating characteristics of the battery module, such as voltage, current, temperature, and state of charge; obtain the power output value of the photovoltaic module, wind energy volatility parameters, grid feed ratio, and environmental meteorological factors to fully reflect the input status of renewable energy; and simultaneously record the operating status parameters of the battery module, including voltage, current, temperature, and state of charge, to ensure the corresponding relationship between energy input and storage unit response. All collected data must be time-stamped and continuously recorded with a unified sampling period.

[0076] The collected data is standardized to unify the scale range of each feature data, and a time series sample of the charging state is constructed through a time window mechanism. For the various physical quantity data collected, a standardization method is used to unify the numerical range of features of different dimensions. For example, through Z-score standardization or Min-Max normalization, various features are mapped to the same scale interval to avoid the influence of unit or order of magnitude differences on subsequent analysis. Based on the set time window mechanism, continuous data is sliced ​​to construct a time series sample of the charging state with temporal continuity.

[0077] A sliding average and weighted dynamic smoothing algorithm are introduced to extract representative statistical features that reflect the actual response behavior of the battery module. The sliding average algorithm is used to denoise the data within each time window to eliminate the interference of high-frequency fluctuations. A weighted dynamic smoothing algorithm is introduced to extract representative statistical features within each time window, such as average, maximum, minimum, standard deviation, and rate of change, to quantify the dynamic behavior characteristics of the charging process.

[0078] Based on the multi-dimensional feature fusion mechanism, the wind and solar power generation input features and the battery module response features are jointly aggregated to form a multi-source heterogeneous fusion feature vector; after the wind and solar power generation-related input features and the battery module response features are aligned according to the timestamp, they are aggregated through the multi-dimensional feature fusion mechanism, and the environmental input features and battery response features are spliced ​​in the same time window. The most representative information is extracted by combining the weighting strategy or fusion algorithm to construct a fusion feature vector that can simultaneously reflect the input disturbance and response behavior. The feature vector has the ability to integrate multi-source heterogeneous features and can comprehensively characterize the charging status at that moment; the charging status feature dataset is constructed using the aggregated multi-dimensional feature vector.

[0079] S2: Based on the charging state feature dataset, a hierarchical clustering algorithm and a charging stage identification model are applied to segment the charging process, distinguish typical charging states, and automatically match the stage target charging parameters according to the battery type.

[0080] The process of segmenting the charging process by applying the hierarchical clustering algorithm and the charging stage identification model in S2 is as follows:

[0081] The constructed charging state feature dataset is input into the hierarchical clustering model, and the different charging states are clustered and analyzed by combining the Euclidean distance and Mahalanobis distance joint similarity measurement. The previously constructed charging state feature dataset is used as input data and input into the hierarchical clustering model for preliminary analysis. The joint similarity measurement method combining the Euclidean distance and Mahalanobis distance is adopted. The covariance structure between each feature is introduced on the basis of considering the numerical difference, which enhances the ability to characterize the intrinsic correlation between different charging states. The clustering tree structure is constructed by layer-by-layer merging through the distance matrix.

[0082] The hierarchical depth control parameters of the clustering tree are set, and a pruning strategy is used to retain clusters with significant representative charging stages. The pruning strategy is introduced to prune the clustering tree structure, filtering out redundant subclusters with too deep hierarchies or unclear boundaries, and only retaining significant representative clusters. Each retained cluster corresponds to a potential stage in the charging process and has consistent characteristic behavior patterns and physical meaning.

[0083] For each cluster, the charging stage recognition model is used to classify its labels to determine whether it belongs to a typical stage such as constant current charging, constant voltage charging, trickle charging, or maintenance charging. For each cluster, the cluster is input into the trained charging stage recognition model to perform discriminant analysis on its overall feature distribution. Based on the model's judgment results, each cluster is classified into a typical charging stage label, such as constant current charging stage, constant voltage charging stage, trickle charging stage, maintenance charging stage, etc., completing the transition from unsupervised clustering to supervised recognition.

[0084] During the recognition process, historical annotation data and model supervision verification mechanisms are combined to improve the accuracy and generalization ability of segmented recognition;

[0085] Finally, a multi-stage segmented result of the charging process is formed.

[0086] The process of automatically matching the staged target charging parameters according to the battery type in S2 is as follows:

[0087] Based on the segmented recognition results, the system's built-in battery type recognition module is called to obtain the type of battery module currently connected to the charging process, including lithium-ion battery, lithium iron phosphate battery, nickel-metal hydride battery, etc.

[0088] Based on the identified battery type, the system matches the corresponding stage-by-stage voltage, current, and temperature rise safety threshold parameters in the preset charging parameter database. Based on the identified battery type, the system accesses and matches the locally preset charging parameter database, which stores standard target parameters for various battery types in different charging stages, including: target current and voltage range in the constant current stage; target voltage and termination current threshold in the constant voltage stage; target power and lower time limit in the trickle charge stage; and safety control parameters such as upper temperature rise rate limit and internal resistance increase critical value. The system then extracts the stage-by-stage target voltage, current, temperature rise, and other parameter templates that match the current battery type.

[0089] During the target parameter matching process, factors such as ambient temperature, battery cycle life status, and health are taken into account to dynamically adjust the adaptation range of the target parameters. On the basis of performing standard parameter matching, a dynamic adaptation mechanism is further introduced to adjust the stage target parameters to adapt to the actual working conditions. This mainly includes:

[0090] Current ambient temperature: If the ambient temperature is too high, appropriately lower the upper limit of the constant current or shorten the constant voltage time;

[0091] Battery cycle life status: If it is already in a high aging stage, increase the safety margin and reduce the peak current;

[0092] Battery health: Dynamically adjust termination conditions based on SOH value. If SOH < 80%, trickle charging will be terminated early.

[0093] Combined with the charging stage classification labels, the target parameter templates of each stage are automatically bound to form a complete charging stage target parameter configuration file; the adapted target parameters are structured and bound according to the stage classification labels to form a complete target configuration file: the target current, voltage and termination conditions are bound for the constant current stage; the target voltage and current thresholds are bound for the constant voltage stage; and the target power and maintenance time are bound for the trickle charge stage. The configuration file serves as the input constraint condition of the reinforcement learning controller and the voltage and current adjustment mechanism, and is dynamically referenced in the actual control link.

[0094] S3: According to the identified charging stage, combined with the reinforcement learning control strategy, a real-time voltage and current adjustment mechanism is constructed, and the output power is dynamically adjusted according to the battery response feedback during the charging process.

[0095] The process of constructing the real-time voltage and current adjustment mechanism in S3 combined with the reinforcement learning control strategy is:

[0096] Based on the stage target charging parameter configuration file, the reward function of the reinforcement learning model is set, aiming to minimize the charging time and maximize the energy conversion efficiency, while ensuring that the safety indicators in the charging process are not triggered; According to the control requirements of the charging system in different stages, a stage target charging parameter configuration file is constructed, including target charging time, energy conversion efficiency, safety operation constraints and other indicators; Based on this, the reward function of the reinforcement learning model is designed: when the charging time is shortened and the energy conversion efficiency is improved, a positive reward is given; If the charging process triggers overheating, overvoltage or current overshoot and other safety warnings, a penalty term is applied; At the same time, a penalty smoothing coefficient is introduced to balance the relationship between policy exploration and stability. The reward function drives the model to produce control actions that take into account efficiency and safety during the control strategy training process;

[0097] The state space is defined as the current SOC, voltage, current and environmental temperature parameters of the battery, and the action space is the adjustment of the output power and the amplitude value of the charging voltage and current;

[0098] The proximal policy optimization algorithm is used for control strategy training, and the reinforcement learning model updates the policy through online feedback; Through simulation of multiple charging cycles, the reinforcement learning agent generates control actions based on the policy network, and obtains feedback from the environment, iteratively updating the policy network and value function network; The clipping target function and experience replay mechanism are used to improve the stability of the policy and the sample utilization rate in the training process; Disturbance factors can be introduced during training to simulate abnormal scenarios;

[0099] In the actual charging process, the control model dynamically generates control actions according to real-time feedback signals of the battery module, such as voltage mutation rate and current fluctuation rate.

[0100] S4: In the dynamic charging control process, the response stability indicators of the battery module are evaluated in real time, and it is monitored whether the charging parameters reach the stable interval. If so, the automatic portfolio strategy is triggered, and the battery group with similar charging state is preferentially selected to access the charging process.

[0101] The process of monitoring whether the charging parameters reach the stable interval in S4 is:

[0102] During the implementation of the reinforcement learning-based real-time voltage and current adjustment mechanism, the key response parameters of the currently connected battery module are continuously collected, including but not limited to voltage, current, temperature rise rate, internal resistance change rate, etc., and these parameters are recorded at multiple points within the set time sliding window;

[0103] For each key parameter, a sliding window statistical algorithm is applied to calculate the mean, standard deviation and volatility within the window period and compared with the preset stability threshold;

[0104] Set up stability interval judgment criteria, including the maximum allowable fluctuation rate, the minimum mean change rate, and the stability factor of the coupling relationship between parameters, and construct a multi-factor stability evaluation index system; preset the stability judgment threshold of each key parameter, including but not limited to: maximum allowable fluctuation rate; minimum mean change rate; parameter coupling stability factor: for example, the temperature rise rate and current change should be negatively correlated. If the correlation coefficient |r| is <0.3, it is judged to be unstable; these indicators constitute a multi-factor stability evaluation index system, which is used to determine whether the current charging process is in a convergent and stable state.

[0105] If the fluctuation rate of the key parameters is less than the set stability threshold in multiple consecutive sliding window cycles, it is determined that the current battery module charging process has entered a convergent and stable state;

[0106] If the fluctuation rate of the key parameter is greater than or equal to the set stability threshold in multiple consecutive sliding window cycles, it is determined that the current battery module charging process has not entered a converged stable state.

[0107] S5: Apply multi-source scheduling optimization algorithm to dynamically allocate wind and solar power generation resources, make real-time corrections to charging scheduling plans, and generate energy efficiency optimization reports.

[0108] The process of generating the energy efficiency optimization report in S5 is as follows:

[0109] After completing the charging stage control and automatic position adjustment process, the utilization efficiency, average conversion efficiency and position adjustment response delay of wind and solar power generation resources in the entire charging process are calculated;

[0110] Wind and solar power resource utilization: Compare the actual output power of photovoltaic modules and wind units with the theoretical maximum output power and calculate the average utilization rate;

[0111] Average energy conversion efficiency: This measures the ratio of the original wind and solar power input to the total amount of energy effectively absorbed by the battery module, reflecting the end-to-end energy transmission efficiency.

[0112] Position adjustment response delay: Record the interval between the time the automatic position adjustment instruction is issued and the time the battery pack is actually switched. Extract the average and maximum values ​​of multiple responses as an evaluation indicator of the system control reaction speed.

[0113] Combined with the energy received by each module during the charging process and the achievement rate of the target charging parameters, the comprehensive energy efficiency index is calculated, including the energy input density per unit time and the number of unit power dispatches. The following energy efficiency-related parameters are calculated:

[0114] Energy reception statistics: Statistics on the energy absorbed by each module per unit time, forming a module-energy mapping table;

[0115] Target parameter achievement rate: Determines whether each module has achieved the preset charging voltage / current target at each stage, and calculates the achievement rate;

[0116] Energy input density per unit time: reflects the energy charging efficiency of the system as a whole within a certain time window;

[0117] Unit power dispatch times: This evaluates the number of adjustments / dispatching actions required for each 1kWh of energy charged by the system, used to determine the complexity and energy consumption of the control strategy;

[0118] A multi-source scheduling optimization algorithm is introduced to analyze the resource scheduling paths and battery pack switching strategies during the completed charging process, identifying redundant paths and energy waste points. The actual resource scheduling path diagram is reconstructed using the wind and solar power output direction, battery pack switching sequence, power regulation curve, and other data recorded in the system. Path nodes with high frequency but low returns are identified and marked as scheduling redundancy points. Energy loss areas caused by time delays, path overlaps, and other factors are analyzed in the path and identified as energy waste points, providing a basis for subsequent optimization.

[0119] Combining system operation logs and charging behavior records, an energy efficiency optimization report is automatically generated, which includes energy utilization efficiency curves, scheduling strategy optimization suggestions, battery performance evolution trend charts, etc.

[0120] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0121] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A control method for automatic adjustment of charging battery modules for wind-solar integrated power generation, characterized in that: include: DC conversion module: The charging input end is connected to the generating bus of the wind power integrated power generation system, and the charging output end is connected to the rechargeable battery in the charging tank. It is used to convert the power input from the generating bus into output power with adjustable voltage; Automatic battery adjustment mechanism: includes a charging slot, an ejection slot, multiple slots, and a battery adjustment mechanism. The slots are used to store batteries to be charged. The adjustment mechanism sequentially feeds the batteries to be charged into the charging slots and transfers the batteries to the ejection slots after they are fully charged. Control module: adjusts the output voltage and current of the DC conversion module and controls the charging process according to the real-time data of the rechargeable battery obtained by the voltage sampling module, current sampling module and temperature monitoring module; The control module includes: The input power parameters of the wind-solar integrated power generation system under different operating scenarios are obtained. Combined with the operating parameters of the battery module, a feature aggregation algorithm is used to construct a charging state feature dataset. Based on the charging state feature dataset, a hierarchical clustering algorithm and a charging stage identification model are applied to segment the charging process, distinguish typical charging states, and automatically match the target charging parameters for each stage according to the battery type. Based on the identified charging stage, a reinforcement learning control strategy is combined to build a real-time voltage and current adjustment mechanism, which dynamically adjusts the output power according to the battery response feedback during the charging process. During dynamic charging control, the battery module's response stability indicators are evaluated in real time to monitor whether the charging parameters have reached a stable range. If so, an automatic repositioning strategy is triggered, prioritizing battery packs with similar charging status to join the charging process. Apply multi-source scheduling optimization algorithms to dynamically allocate wind and solar power generation resources, make real-time corrections to charging scheduling plans, and generate energy efficiency optimization reports; The process of generating an energy efficiency optimization report is as follows: After completing the charging stage control and automatic position adjustment process, the utilization efficiency, average conversion efficiency and position adjustment response delay of wind and solar power generation resources in the entire charging process are calculated; Combined with the energy received by each module during the charging process and the achievement rate of the target charging parameters, the comprehensive energy efficiency index is calculated, including the energy input density per unit time and the number of dispatches per unit power; A multi-source scheduling optimization algorithm is introduced to analyze the resource scheduling paths and battery pack switching strategies during the completed charging process, identifying redundant paths and energy waste points. Combine system operation logs with charging behavior records to automatically generate energy efficiency optimization reports.

2. The method for controlling a rechargeable battery module for wind-solar integrated power generation automatic adjustment according to claim 1, characterized in that: The control module executes the control process including: After starting the system, check the status of each submodule; Detect whether the charging slot has a battery inserted. If no battery is present, the battery in the slot is automatically transferred to the charging slot through the automatic adjustment mechanism. During the charging process, the battery voltage, current and temperature data are continuously collected; Dynamically adjust the charging voltage and current according to the target parameters corresponding to the charging stage to achieve multi-stage fine charging; After the battery is fully charged, the automatic adjustment mechanism is controlled to transfer the battery to the ejection slot and introduce the next battery to continue charging.

3. The method for controlling a rechargeable battery module for automatic adjustment of wind-solar integrated power generation according to claim 1, characterized in that: The automatic adjustment structure has a cyclic charging function. It supports automatically introducing a new battery for the next round of charging after a fully charged battery is ejected from the ejection slot. It has a standby monitoring mechanism. If there is no battery to be charged in the slot, it will automatically enter a low-power standby state, continuously detect the slot status, and automatically wake up and start the charging process when a new battery is inserted.

4. The method for controlling a rechargeable battery module for automatic adjustment of wind-solar integrated power generation according to claim 1, characterized in that: The process of constructing the state of charge feature dataset using the feature aggregation algorithm is as follows: Collect the photovoltaic power output, grid feed-in ratio, wind energy volatility parameters and environmental meteorological factors of the wind-solar integrated power generation system in different operating periods, and simultaneously obtain the operating characteristics of the battery module; Perform data standardization on the collected data, unify the scale range of each feature data, and construct a time series sample of the charging state through the time window mechanism; The sliding average and weighted dynamic smoothing algorithms are introduced to extract representative statistical features; Based on the multi-dimensional feature fusion mechanism, the wind and solar power input features and the battery module response features are jointly aggregated to form a multi-source heterogeneous fusion feature vector; The charge state feature dataset is constructed using the aggregated multi-dimensional feature vectors.

5. The method for controlling a rechargeable battery module for automatic adjustment of wind-solar integrated power generation according to claim 1, characterized in that: The process of segmenting and identifying the charging process using the hierarchical clustering algorithm and the charging stage identification model is as follows: The constructed charging state feature dataset is input into the hierarchical clustering model, and the Euclidean distance and Mahalanobis distance are combined to measure similarity and perform cluster analysis on different charging states. Set the hierarchical depth control parameters of the clustering tree and retain the charging stage clusters through pruning strategies; For each cluster, the charging stage recognition model is used to classify the labels and determine whether the labels belong to the typical stage; In the recognition process, historical annotation data is combined with the model supervision verification mechanism; Finally, a multi-stage segmented result of the charging process is formed.

6. The method for controlling a rechargeable battery module for automatic adjustment of wind-solar integrated power generation according to claim 1, characterized in that: The process of automatically matching the staged target charging parameters according to the battery type is as follows: Based on the segmented recognition results, the system's built-in battery type recognition module is called to obtain the type of battery module currently connected to the charging process; According to the identified battery type, the corresponding stage voltage, current and temperature rise safety threshold parameters are matched in the preset charging parameter database; In the process of target parameter matching, the influencing factors are taken into consideration and the adaptation range of the target parameters is dynamically adjusted; Combined with the charging stage classification labels, the target parameter templates of each stage are automatically bound to form a complete charging stage target parameter configuration file.

7. The method for controlling a rechargeable battery module for automatic adjustment of wind-solar integrated power generation according to claim 1, characterized in that: The process of building a real-time voltage and current adjustment mechanism by combining reinforcement learning control strategy is as follows: Based on the phased target charging parameter profile, the reward function of the reinforcement learning model is set to minimize charging time and maximize energy conversion efficiency; The state space is defined as the battery's current SOC, voltage, current, and ambient temperature parameters, and the action space is defined as the amplitude values ​​for adjusting the output power and the charging voltage and current. The proximal policy optimization algorithm is used to train the control strategy, and the reinforcement learning model continuously updates the strategy through online feedback; During the actual charging process, the control model dynamically generates control actions based on the real-time feedback signals from the battery module.

8. The method for controlling a rechargeable battery module for automatic adjustment of wind-solar integrated power generation according to claim 1, characterized in that: The process of monitoring whether the charging parameters have reached the stable range is as follows: During the implementation of the reinforcement learning-based real-time voltage and current adjustment mechanism, the key response parameters of the currently connected battery module are continuously collected and recorded at multiple points within the set time sliding window; For each key parameter, a sliding window statistical algorithm is applied to calculate the mean, standard deviation and volatility within the window period and compared with the preset stability threshold; Set up the criteria for judging the stability interval and build a multi-factor stability evaluation index system; If the fluctuation rate of the key parameters is less than the set stability threshold in multiple consecutive sliding window cycles, it is determined that the current battery module charging process has entered a convergent and stable state; If the fluctuation rate of the key parameter is greater than or equal to the set stability threshold in multiple consecutive sliding window cycles, it is determined that the current battery module charging process has not entered a converged stable state.

Images