A battery-powered collaborative control method and system

Abstract

This invention provides a battery-powered coordinated control method and system, comprising: collecting and preprocessing battery voltage, current, and temperature data; supplementing the collection of microgrid photovoltaic, load, and grid parameters; calculating prediction error coefficients to form a corresponding dataset; distinguishing charging and discharging states based on battery operating data; calculating a coordinated proportional factor and extracting state features; dynamically adjusting the proportional factor in conjunction with prediction errors; generating initial control commands and then optimizing them through amplitude-limited double-exponential smoothing; adjusting command parameters in conjunction with multi-objective constraints; executing power supply switching; optimizing controller parameters in reverse through a reward function; filling power supply gaps through energy storage power correction and controllable load adjustment; and monitoring battery and grid states in all dimensions to achieve automatic power supply circuit switching and dynamic adjustment of charging and discharging strategies, thereby improving the system's adaptability to complex operating conditions, ensuring stable power supply, improving energy utilization, and reducing operating costs.

Description

Technical Field

[0001] This invention belongs to the field of power supply coordination technology, and more specifically, relates to a battery power supply coordination control method and system. Background Technology

[0002] With the continuous advancement of rural energy structure transformation, the penetration rate of new energy sources such as distributed photovoltaics in rural microgrids is constantly increasing. Batteries, as the core energy storage equipment, have been widely used in power supply systems to alleviate the contradiction between the intermittency of new energy output and load fluctuation.

[0003] Rural microgrids are characterized by complex load types, photovoltaic output being significantly affected by the natural environment, and relatively weak grid operating conditions. Traditional battery power supply control methods mainly focus on the basic parameters of the battery itself, such as voltage and current, without fully integrating multi-dimensional data on the actual output of photovoltaic power and load forecasts of rural microgrids. This makes it difficult to adapt to complex operating scenarios, resulting in the power supply system struggling to achieve a balance between stability, economy, and energy utilization. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a battery-powered collaborative control method and system, which solves the technical problem that traditional power supply control methods are difficult to adapt to complex operating scenarios, resulting in the difficulty of achieving a balance between stability, economy, and energy utilization in the power supply system.

[0005] The purpose and effectiveness of the battery-powered coordinated control method and system of the present invention are achieved by the following specific technical means: A battery-powered collaborative control method includes the following steps: S1: Obtain the original battery operation dataset, obtain the original power grid supply dataset based on the original battery operation dataset, and obtain the power supply prediction error dataset based on the prediction error of the original power grid supply dataset. S2: Obtain the battery state feature dataset based on the original battery operation dataset, and obtain the collaborative proportional factor dataset based on the power supply prediction error dataset and the battery state feature dataset; S3: Generate control commands based on the collaborative proportional factor dataset and battery state feature dataset, and perform multi-objective adjustment of the control commands according to the original power grid data set to obtain a multi-objective control command dataset; S4: Based on the multi-objective control command dataset, power supply switching is performed, controller parameters are optimized, power supply gaps are filled, and power compensation dataset is obtained; S5: Monitor the power supply battery based on the power compensation dataset to obtain the dimensional monitoring dataset, and automatically switch back the power supply loop and adjust the strategy based on the dimensional monitoring dataset.

[0006] As a further aspect of the present invention, obtaining the original battery operating dataset and obtaining the original power grid data set based on the original battery operating dataset includes: Connect the data acquisition device to the voltage, current and temperature detection terminals of the main battery, backup battery or multiple power supply devices respectively, and continuously collect the battery's operating data under charging and discharging, steady state or transient conditions to obtain the battery's original operating data set. Based on the original battery operating dataset, the state of charge (SOC) value of each battery is estimated using the ampere-hour integration method. At the same time, abnormal interference components in the data are identified and removed, and the voltage, current, and SOC information of the batteries are integrated to obtain the battery preprocessing dataset. Based on the battery preprocessing dataset, additional data were collected on photovoltaic parameters, load parameters, and grid parameters of the rural microgrid to obtain the original power grid power supply dataset.

[0007] As a further aspect of the present invention, the power supply prediction error dataset based on the original power supply dataset includes: Based on the original power grid data set, the deviations between the actual and predicted output of photovoltaic power and the deviations between the actual and predicted load values ​​are quantified to obtain the photovoltaic prediction error coefficient and the load prediction error coefficient, thus acquiring the power supply prediction error dataset.

[0008] As a further aspect of the present invention, the step of obtaining a battery state feature dataset based on the original battery operating dataset and obtaining a collaborative scaling factor dataset based on the power supply prediction error dataset and the battery state feature dataset includes: Based on the original battery operation dataset, the charging and discharging states of each battery are distinguished according to their state of charge and output current. The cooperative proportional factor under different states is calculated to quantify the power allocation weight of each battery and obtain the battery cooperative proportional factor dataset. Based on the battery cooperative scaling factor dataset and the battery original operation dataset, the battery output voltage and cooperative scaling factor input state encoder are normalized in terms of feature range, and key features that can characterize the battery operation state are extracted to obtain the battery state feature dataset. Error triggering and scaling factor adjustment are performed based on power supply prediction error dataset and battery status feature dataset. It is determined whether the photovoltaic or load prediction error coefficient exceeds the preset threshold. If it does, the collaborative scaling factor is dynamically adjusted according to the error magnitude to strengthen the power compensation weight of energy storage batteries and obtain the adjusted collaborative scaling factor dataset.

[0009] As a further aspect of the present invention, the step of generating control commands based on the collaborative proportional factor dataset and the battery state feature dataset, and adjusting the control commands according to the original power grid data set to obtain a multi-objective control command dataset includes: Based on the collaborative proportional factor dataset and the battery state feature dataset, the feature data is input into the device-level controller. Combined with the power allocation weight, the initial charging, discharging, and switching commands for each battery are generated. At the same time, photovoltaic consumption and cost control targets are incorporated to obtain a multi-objective initial control command dataset. Based on the multi-objective initial control command dataset, a limiting double exponential smoothing strategy is adopted. According to the horizontal and trend components of the command at the previous time step, the initial control command is optimized and adjusted to constrain the command fluctuation range and obtain the control command dataset. Based on the control command dataset and the original power grid dataset, with the goals of maximizing photovoltaic absorption rate, minimizing operating cost, and ensuring voltage stability, and combined with power balance, battery SOC, and transformer load rate constraints, the control command parameters are adjusted to obtain a multi-objective control command dataset.

[0010] As a further aspect of the present invention, the power supply switching based on the multi-objective control command dataset includes: Based on the multi-target control command dataset, the final control command is transmitted to the dual power supply switching module to detect the voltage surge magnitude and duration of the backup battery interface. When the voltage meets the preset conditions and is greater than the relay action voltage, the relay coil is energized, the power supply circuit is switched to the target battery, and the power supply switching execution dataset is obtained. As a further aspect of the present invention, the optimization of controller parameters and filling of power supply gaps to obtain power compensation dataset includes: Based on the power supply switching execution dataset, collaborative control feedback optimization is performed. After switching, voltage stability data and power distribution data of each battery are collected. A reward function including dynamic proportional factor error and voltage error is established. The loss function and gradient are calculated. The device-level controller parameters are updated in reverse optimization, and the controller parameter optimization dataset is obtained. Energy storage power correction is performed based on the controller parameter optimization dataset and the power supply prediction error dataset. Specifically, if the actual output of photovoltaic power is lower than the predicted value or the load suddenly increases, the additional energy storage discharge power is calculated and the original discharge command is corrected. If the photovoltaic output is excessive, the additional charging power is calculated and the charging command is corrected to obtain the energy storage correction power command dataset. Controllable load coordinated regulation is performed based on the energy storage correction power command dataset; if the energy storage correction power still cannot fully compensate for the error, the controllable load power to be reduced is calculated, and load regulation commands are generated to fill the power supply gap; otherwise, the load is not regulated; and the power compensation dataset is obtained.

[0011] As a further aspect of the present invention, the step of monitoring the power supply battery based on the power compensation dataset to obtain a dimensional monitoring dataset, and performing automatic power supply circuit switching and strategy adjustment based on the dimensional monitoring dataset, includes: Based on the power compensation dataset, the current power supply battery's charge, voltage, and temperature changes are monitored in real time. Simultaneously, the voltage deviation of the distribution area, line loss, transformer load rate, and photovoltaic output fluctuations are tracked to assess the battery's remaining power supply capacity and grid operation safety, and to obtain a full-dimensional monitoring dataset. Based on the full-dimensional monitoring dataset, when the current power supply battery is detected to be depleted, the dual power supply switching module is triggered to disconnect the current circuit and automatically switch back to the original main battery or other backup power supply equipment circuit; at the same time, the charging and discharging strategy is dynamically adjusted according to the monitored voltage deviation and load rate.

[0012] A battery-powered collaborative control system includes: The data acquisition module is used to obtain the battery's original operating dataset; The feature extraction module, connected to the data acquisition module, is used to obtain a battery state feature dataset based on the original battery operating dataset. The control command generation module is connected to the feature extraction module and the data acquisition module, and is used to generate control commands based on the collaborative proportional factor dataset and the battery state feature dataset. A power supply switching module, connected to a control command generation module and a data acquisition module, is used to switch power supplies based on the multi-target control command dataset. The loop control module, connected to the power supply switching module, is used to monitor the power supply battery based on the power compensation dataset, obtain a full-dimensional monitoring dataset, and then automatically switch back the power supply loop and adjust the strategy based on the full-dimensional monitoring dataset.

[0013] Compared with the prior art, the present invention has the following beneficial effects: 1. Collect basic operating data such as battery voltage, current, and temperature, and supplement them with photovoltaic parameters, load parameters, and grid parameters of rural microgrids to build comprehensive basic data support; at the same time, by quantifying the prediction deviation between photovoltaic and load, a power supply prediction error dataset is formed, and the coordination ratio factor is dynamically adjusted according to the error magnitude to strengthen the power compensation weight of energy storage batteries, so that the control strategy can respond in real time to the fluctuation of new energy output and load changes, avoiding the adaptability problems caused by insufficient data support and delayed error response in traditional methods.

[0014] 2. In the process of generating control commands, the core objectives of photovoltaic consumption and cost control are integrated, and the range of command fluctuations is constrained by a limit-amplitude double exponential smoothing strategy to ensure stable power supply. At the same time, the command parameters are adjusted in combination with constraints such as power balance, battery state of charge, and transformer load rate to avoid the loss of one aspect due to optimization of a single objective. This enables the power supply system to reduce operating costs and improve the utilization efficiency of new energy sources while ensuring voltage stability and safe operation of equipment.

[0015] 3. In response to photovoltaic power output gaps or sudden load surges, the charging and discharging commands are first adjusted by correcting the energy storage power. If the error still cannot be fully compensated, controllable load collaborative regulation is initiated to form a multi-complementary power compensation mechanism, effectively avoiding power outages. At the same time, by monitoring the battery status and grid operating parameters in all dimensions, the power supply circuit is automatically switched back, and the charging and discharging strategy is dynamically adjusted according to voltage deviation, load rate, and other conditions, so that the system can adapt to different operating conditions and continuously maintain efficient and stable operation. Attached Figure Description

[0016] Figure 1 This is a flowchart of the steps of a battery-powered collaborative control method according to the present invention.

[0017] Figure 2 This is a schematic diagram of a battery-powered collaborative control system according to the present invention. Detailed Implementation

[0018] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate the technical solutions of the present invention, but should not be used to limit the scope of protection of the present invention.

[0019] Example:

[0020] As attached Figure 1 As shown: This invention provides a battery-powered coordinated control method, comprising the following steps: S1: Obtain the original battery running dataset, obtain the original power grid power supply dataset based on the original battery running dataset, and obtain the power supply prediction error dataset based on the prediction error of the original power grid power supply dataset.

[0021] In this embodiment, the signal acquisition port of the data acquisition device is connected one-to-one with the voltage detection terminal, current detection terminal, and temperature detection terminal of the main battery, backup battery, or multiple power supply devices. A reasonable acquisition frequency is set according to the operating characteristics of the power supply devices to ensure that complete operating data of the battery during charging and discharging, steady-state operation, or transient fluctuation conditions can be captured. The data is continuously collected for a period of time to cover different operating states, obtaining the original battery operating dataset containing information such as battery voltage changes, current output, and temperature fluctuations. This provides basic data support for subsequent battery state analysis and control strategy formulation. Based on the acquired original battery operating dataset, the ampere-hour integration method is used, combined with the battery's rated capacity and discharge efficiency, to gradually accumulate and calculate the state of charge (SOC) value of each battery. At the same time, a data filtering algorithm is used to identify and remove abnormal interference components in the data. These abnormal components include data points that deviate from the normal range due to detection errors and transient electromagnetic interference. The filtered battery voltage and current data are then integrated with the calculated SOC information and sorted by time series to form a battery preprocessing dataset, reducing the interference of invalid data on subsequent analysis and improving data quality.

[0022] Based on the battery operating status reflected in the battery preprocessing dataset, additional photovoltaic parameters, load parameters, and grid parameters of the rural microgrid are collected. The photovoltaic parameters include the actual output, predicted output, irradiance, and panel temperature of the photovoltaic system; the load parameters include the actual load, predicted load, and controllable load capacity of the microgrid; and the grid parameters include transformer substation voltage, line loss, and transformer load rate. These additionally collected parameters are integrated with the battery preprocessing data and organized according to a unified time dimension to form the original power grid power supply dataset, which comprehensively covers the key parameters related to microgrid power supply and provides complete data support for subsequent prediction error calculation and multi-objective control.

[0023] Based on the actual and predicted photovoltaic power output data and the actual and predicted load data in the original power grid dataset, the deviation between the two is calculated using a deviation quantification algorithm to obtain the photovoltaic prediction error coefficient and the load prediction error coefficient. The photovoltaic prediction error coefficient is used to characterize the difference between the predicted and actual photovoltaic power output, while the load prediction error coefficient is used to characterize the difference between the predicted and actual load demand. These two types of error coefficients are integrated to form a power supply prediction error dataset, which provides a quantitative basis for dynamically adjusting the control strategy according to the magnitude of the error.

[0024] S2: Obtain the battery state feature dataset based on the original battery operation dataset, and obtain the collaborative scaling factor dataset based on the power supply prediction error dataset and the battery state feature dataset.

[0025] In this embodiment, based on the original battery operating dataset, the state of charge (SOC) value and real-time output current data of each battery are extracted. The battery charging / discharging state is distinguished by the sign of the output current: a positive current indicates a discharging state, and a negative current indicates a charging state. For batteries in the discharging state, a coordination ratio factor is calculated using the ratio of output current to SOC. For batteries in the charging state, a coordination ratio factor is calculated using the ratio of output current to the missing SOC. This method quantifies the power allocation weight of each battery in the power supply system, clarifies the output priority of different batteries, integrates the ratio factor data of all batteries, and obtains a battery coordination ratio factor dataset, providing a weighting basis for subsequent power allocation and control command generation. Based on the battery coordination ratio... Example factor dataset and original battery operation dataset are used to extract the output voltage data and corresponding proportional factors of each battery. The two sets of data are aligned by time series and then input into the state encoder. The encoder's built-in normalization algorithm performs feature range normalization processing on the data. Specifically, it subtracts the mean of the corresponding features in the dataset and divides by the standard deviation to bring feature data of different magnitudes into a uniform numerical range. Then, the feature extraction algorithm is used to select key features that can characterize the battery's operating state, including voltage fluctuation trends, proportional factor stability, and charge / discharge state transition frequency. These key features are integrated to form a battery state feature dataset, which simplifies the data dimensions while retaining core operating information, providing efficient data support for the controller to identify the battery state.

[0026] Based on the power supply prediction error dataset and battery status characteristic dataset, a preset error threshold is first retrieved. This threshold is determined according to the microgrid's operational stability requirements and the equipment's tolerance range. The photovoltaic prediction error coefficient and the load prediction error coefficient are checked one by one to see if they exceed the preset threshold. If any error coefficient exceeds the threshold, the adjustment range is determined based on the difference between the error coefficient and the threshold. The larger the error difference, the larger the adjustment range of the collaborative proportional factor. The power compensation weight of the energy storage battery is strengthened, so that the energy storage battery can undertake more power regulation tasks when the prediction deviation is large. After the adjustment is completed, the new proportional factors of all batteries are integrated to obtain the adjusted collaborative proportional factor dataset, so that the power allocation weight is adapted to the current prediction error situation, thereby improving the system's adaptability to fluctuations in photovoltaic output and load demand.

[0027] S3: Generate control commands based on the collaborative proportional factor dataset and battery state feature dataset, and perform multi-objective adjustments to the control commands based on the original power grid data set to obtain a multi-objective control command dataset.

[0028] In this embodiment, based on the power allocation weight information of each battery contained in the collaborative proportional factor dataset and the key battery operation features covered by the battery state feature dataset, the two types of data are organized according to the input format of the device-level controller and then transmitted to the controller. The controller combines the power allocation ratio of each battery determined by the collaborative proportional factor to generate charging and discharging intensity commands and initial power supply circuit switching commands for the main battery, backup battery and each group of power supply equipment. At the same time, the photovoltaic consumption target and cost control target are incorporated into the command generation process, that is, the command is made to prioritize adapting to the actual output of photovoltaic power to reduce curtailment. Meanwhile, the command parameters are optimized with reference to the charging and discharging cost of energy storage unit and the demand response compensation cost, forming a multi-objective initial control command dataset, which provides a basis for subsequent command optimization.

[0029] Based on a multi-objective initial control command dataset, a limited-amplitude double-exponential smoothing strategy is adopted to extract the horizontal component (i.e., the basic value of the previous control command) and the trend component (i.e., the trend value of the previous command relative to the previous period) of the control command at the previous moment. The current initial command is then weighted and adjusted according to a preset safety factor to constrain the command fluctuation amplitude within a preset safety range, preventing sudden command changes from impacting power supply equipment and grid operation, thus obtaining a control command dataset with smooth fluctuations. Based on the photovoltaic parameters, load parameters, grid parameters, and battery parameters in the control command dataset and the original power supply dataset, with the core objectives of maximizing photovoltaic absorption rate, minimizing operating costs, and ensuring voltage stability, and strictly adhering to power balance constraints, battery power compensation dataset SOC power compensation dataset constraints, and transformer load rate constraints, the charging and discharging power parameters and switching timing parameters in the control commands are refined and adjusted. This ensures that the commands can adapt to changes in photovoltaic output and load while controlling operating costs and voltage deviations, ultimately obtaining a multi-objective control command dataset that meets the requirements of multiple objectives.

[0030] S4: Based on the multi-objective control command dataset, power supply switching is performed, controller parameters are optimized, power supply gaps are filled, and power compensation dataset is obtained.

[0031] In this embodiment, based on the power supply circuit switching instructions and parameters contained in the multi-target control instruction dataset, the final control instruction is transmitted to the dual power supply switching module through the communication module. The dual power supply switching module monitors the voltage change of the backup battery interface in real time, records the magnitude and duration of voltage surges. When the monitored voltage surge reaches the preset standard and the duration meets the set requirements, and the control voltage output by the backup battery is greater than the relay action voltage, the relay coil is energized, and the main contacts in the power supply system switch from the current power supply circuit to the circuit corresponding to the target battery, so that the target battery, controller, and electrical load form a complete power supply circuit. The circuit switching time, voltage and current changes before and after the switching, target battery activation status, and other information are recorded synchronously during the switching process to obtain the power supply switching execution dataset, providing actual operating data support for subsequent controller parameter optimization.

[0032] Based on the power supply switching execution dataset, voltage stability data such as voltage stabilization time and voltage fluctuation range of each battery after switching, as well as power allocation data such as the actual power share of each battery and power allocation response speed, are collected. A reward function incorporating dynamic scaling factor error and voltage error is established based on this data. The execution effect of the current control strategy is quantified through the reward function. Then, based on the quantification result, the loss function and gradient are calculated. Using the gradient descent principle, the internal parameters of the device-level controller are adjusted in reverse to optimize the controller's response logic to power supply switching and power allocation, obtaining a controller parameter optimization dataset to improve the adaptability of subsequent control commands. The adjusted controller parameter dataset is then used to further refine the control strategy. The control logic, combined with the photovoltaic and load deviation reflected in the power supply prediction error dataset, performs energy storage power correction. If the actual photovoltaic output is lower than the predicted value or a sudden increase in load causes a power supply gap, the required supplementary energy storage discharge power is calculated based on the magnitude of the deviation and the current charging and discharging capacity of the energy storage battery. The power parameters in the original discharge command are corrected accordingly to ensure that the energy storage battery outputs sufficient power to fill the gap. If the actual photovoltaic output is higher than the predicted value and there is excess output, the charging power that the energy storage battery can increase is calculated, and the original charging command is corrected to fully absorb the excess photovoltaic power and reduce power curtailment. The energy storage correction power command dataset is obtained, and the corrected charging and discharging parameters and command execution requirements are recorded.

[0033] Based on the energy storage correction power command dataset, the output power after energy storage correction is compared with the actual power supply gap to determine whether the energy storage correction power can fully compensate for the error. If the energy storage correction power still cannot cover the power supply gap, the controllable load power to be reduced is calculated according to the upper limit of the controllable load capacity of the rural microgrid, and the corresponding load adjustment command is generated. The command is then sent to the controllable load equipment through the controller to reduce the load and fill the remaining power supply gap. If the energy storage correction power can fully compensate for the error, the controllable load adjustment is not initiated. Data such as the energy storage correction process, load adjustment status, and power supply gap filling effect are recorded simultaneously and integrated to form a power compensation dataset, providing a basis for subsequent full-dimensional monitoring and strategy adjustment.

[0034] S5: Monitor the power supply battery based on the power compensation dataset to obtain the dimensional monitoring dataset, and automatically switch back the power supply loop and adjust the strategy based on the dimensional monitoring dataset.

[0035] In this embodiment, based on data such as energy storage correction power parameters, load regulation execution, and power supply gap filling effect in the power compensation dataset, a full-dimensional operation monitoring program is initiated. The battery monitoring module collects real-time data on changes in the charge, voltage, and temperature of the current power supply battery. Simultaneously, the grid monitoring module tracks the voltage deviation of the rural microgrid, line loss, transformer load rate, and photovoltaic output fluctuations. Combining the battery rated capacity parameters, current grid load demand, and transformer safety operation standards, the remaining power supply capacity of the current power supply battery and the safety of grid operation are comprehensively evaluated. All monitoring data and evaluation results are organized and summarized according to a unified time dimension to obtain a full-dimensional monitoring dataset.

[0036] Based on a comprehensive monitoring dataset, when the current power supply battery's charge level drops to a preset lower threshold and is determined to be depleted, a circuit switching trigger signal is immediately sent to the dual power supply switching module. This triggers the module's internal control logic, disconnecting the current power supply battery from the load and automatically connecting the original main battery or other backup power supply equipment, achieving seamless power supply circuit switching and preventing power outages. Simultaneously, based on the voltage deviation and transformer load rate recorded in the comprehensive monitoring dataset, the charging and discharging strategy of the energy storage battery is dynamically adjusted. When the voltage in the distribution area is high and photovoltaic output is sufficient, the charging power of the energy storage battery is appropriately increased to absorb excess energy in the grid. When the voltage in the distribution area is low or the transformer load rate is close to the safe upper limit, the discharging power of the energy storage battery is appropriately increased to assist in grid voltage regulation and load reduction. When both voltage deviation and transformer load rate are within the normal range, the current charging and discharging strategy remains unchanged to ensure stable grid operation.

[0037] Please see as follows Figure 2 As shown, the present invention also provides a battery-powered collaborative control system, comprising: The data acquisition module is used to obtain the battery's original operating dataset; The feature extraction module, connected to the data acquisition module, is used to obtain a battery state feature dataset based on the original battery operating dataset. The control command generation module is connected to the feature extraction module and the data acquisition module, and is used to generate control commands based on the collaborative proportional factor dataset and the battery state feature dataset. A power supply switching module, connected to a control command generation module and a data acquisition module, is used to switch power supplies based on the multi-target control command dataset. The loop control module, connected to the power supply switching module, is used to monitor the power supply battery based on the power compensation dataset, obtain a full-dimensional monitoring dataset, and then automatically switch back the power supply loop and adjust the strategy based on the full-dimensional monitoring dataset.

[0038] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A battery-powered coordinated control method, characterized in that, It includes the following steps: S1: Obtain the original battery operation dataset, obtain the original power grid supply dataset based on the original battery operation dataset, and obtain the power supply prediction error dataset based on the prediction error of the original power grid supply dataset. S2: Obtain the battery state feature dataset based on the original battery operation dataset, and obtain the collaborative scaling factor dataset based on the power supply prediction error dataset and the battery state feature dataset; S3: Generate control commands based on the collaborative proportional factor dataset and battery state feature dataset, and perform multi-objective adjustment of the control commands according to the original power grid data set to obtain a multi-objective control command dataset; S4: Based on the multi-objective control command dataset, power supply switching is performed, controller parameters are optimized, power supply gaps are filled, and power compensation dataset is obtained; S5: Based on the power compensation dataset, monitor the power supply battery to obtain the dimensional monitoring dataset, and perform automatic power supply circuit switching and strategy adjustment based on the dimensional monitoring dataset.

2. The battery-powered coordinated control method according to claim 1, characterized in that, The step of obtaining the original battery operating dataset and obtaining the original power grid data based on the original battery operating dataset includes: Connect the data acquisition device to the voltage, current and temperature detection terminals of the main battery, backup battery or multiple power supply devices respectively, and continuously collect the battery's operating data under charging and discharging, steady state or transient conditions to obtain the battery's original operating data set. Based on the original battery operating dataset, the state of charge (SOC) value of each battery is estimated using the ampere-hour integration method. At the same time, abnormal interference components in the data are identified and removed, and the voltage, current, and SOC information of the batteries are integrated to obtain the battery preprocessing dataset. Based on the battery preprocessing dataset, additional data were collected on photovoltaic parameters, load parameters, and grid parameters of the rural microgrid to obtain the original power grid power supply dataset.

3. The battery-powered coordinated control method according to claim 2, characterized in that, The prediction error dataset based on the original power grid power supply dataset includes: Based on the original power grid data set, the deviations between the actual and predicted output of photovoltaic power and the deviations between the actual and predicted load values ​​are quantified to obtain the photovoltaic prediction error coefficient and the load prediction error coefficient, thus acquiring the power supply prediction error dataset.

4. The battery-powered coordinated control method according to claim 1, characterized in that, The process of obtaining the battery state feature dataset based on the original battery operating dataset and the collaborative scaling factor dataset based on the power supply prediction error dataset and the battery state feature dataset includes: Based on the original battery operation dataset, the charging and discharging states of each battery are distinguished according to their state of charge and output current. The cooperative proportional factor under different states is calculated to quantify the power allocation weight of each battery and obtain the battery cooperative proportional factor dataset. Based on the battery cooperative scaling factor dataset and the battery original operation dataset, the battery output voltage and cooperative scaling factor input state encoder are normalized in terms of feature range, and key features that can characterize the battery operation state are extracted to obtain the battery state feature dataset. Error triggering and scaling factor adjustment are performed based on power supply prediction error dataset and battery status feature dataset. It is determined whether the photovoltaic or load prediction error coefficient exceeds the preset threshold. If it does, the collaborative scaling factor is dynamically adjusted according to the error magnitude to strengthen the power compensation weight of energy storage batteries and obtain the adjusted collaborative scaling factor dataset.

5. The battery-powered coordinated control method according to claim 1, characterized in that, The control commands generated based on the cooperative proportional factor dataset and the battery state feature dataset are then adjusted according to the original power grid data to achieve multiple objectives. The resulting multi-objective control command dataset includes: Based on the collaborative proportional factor dataset and the battery state feature dataset, the feature data is input into the device-level controller. Combined with the power allocation weight, the initial charging, discharging, and switching commands for each battery are generated. At the same time, photovoltaic consumption and cost control targets are incorporated to obtain a multi-objective initial control command dataset. Based on the multi-objective initial control command dataset, a limiting double exponential smoothing strategy is adopted. According to the horizontal and trend components of the command at the previous time step, the initial control command is optimized and adjusted to constrain the command fluctuation range and obtain the control command dataset. Based on the control command dataset and the original power grid dataset, with the goals of maximizing photovoltaic absorption rate, minimizing operating cost, and ensuring voltage stability, and combined with power balance, battery SOC, and transformer load rate constraints, the control command parameters are adjusted to obtain a multi-objective control command dataset.

6. The battery-powered coordinated control method according to claim 1, characterized in that, The power supply switching based on the multi-objective control command dataset includes: Based on the multi-target control command dataset, the final control command is transmitted to the dual power supply switching module to detect the voltage surge magnitude and duration of the backup battery interface. When the voltage meets the preset conditions and is greater than the relay operating voltage, the relay coil is energized, the power supply circuit is switched to the target battery, and the power supply switching execution dataset is obtained.

7. The battery-powered coordinated control method according to claim 6, characterized in that, The process of optimizing controller parameters and filling power supply gaps to obtain a power compensation dataset includes: Based on the power supply switching execution dataset, collaborative control feedback optimization is performed. After switching, voltage stability data and power distribution data of each battery are collected. A reward function including dynamic proportional factor error and voltage error is established. The loss function and gradient are calculated. The device-level controller parameters are updated in reverse optimization, and the controller parameter optimization dataset is obtained. Energy storage power correction is performed based on the controller parameter optimization dataset and the power supply prediction error dataset. Specifically, if the actual output of photovoltaic power is lower than the predicted value or the load suddenly increases, the additional energy storage discharge power is calculated and the original discharge command is corrected. If the photovoltaic output is excessive, the additional charging power is calculated and the charging command is corrected to obtain the energy storage correction power command dataset. Controllable load coordinated regulation is performed based on the energy storage correction power command dataset; if the energy storage correction power still cannot fully compensate for the error, the controllable load power to be reduced is calculated, and load regulation commands are generated to fill the power supply gap; otherwise, the load is not regulated; and the power compensation dataset is obtained.

8. The battery-powered coordinated control method according to claim 1, characterized in that, The process of monitoring the power supply battery based on the power compensation dataset to obtain a dimensional monitoring dataset, and then automatically switching back to the power supply circuit and adjusting the strategy based on the dimensional monitoring dataset, includes: Based on the power compensation dataset, the current power supply battery's charge, voltage, and temperature changes are monitored in real time. Simultaneously, the voltage deviation of the distribution area, line loss, transformer load rate, and photovoltaic output fluctuations are tracked to assess the battery's remaining power supply capacity and grid operation safety, and to obtain a full-dimensional monitoring dataset. Based on the full-dimensional monitoring dataset, when the current power supply battery is detected to be depleted, the dual power supply switching module is triggered to disconnect the current circuit and automatically switch back to the original main battery or other backup power supply equipment circuit; at the same time, the charging and discharging strategy is dynamically adjusted according to the monitored voltage deviation and load rate.

9. A battery-powered collaborative control system, characterized in that, include: The data acquisition module is used to obtain the battery's original operating dataset; The feature extraction module, connected to the data acquisition module, is used to obtain a battery state feature dataset based on the original battery operating dataset. The control command generation module is connected to the feature extraction module and the data acquisition module, and is used to generate control commands based on the collaborative proportional factor dataset and the battery state feature dataset. A power supply switching module, connected to a control command generation module and a data acquisition module, is used to switch power supplies based on the multi-target control command dataset. The loop control module, connected to the power supply switching module, is used to monitor the power supply battery based on the power compensation dataset, obtain a full-dimensional monitoring dataset, and then automatically switch back the power supply loop and adjust the strategy based on the full-dimensional monitoring dataset.

Images