A DC bus current control method connected to a backup battery

Abstract

The present invention provides a method for controlling the flow of a DC bus connected to a backup battery, which relates to the field of power system control technology. The method includes: defining the topological structure of a DC microgrid and constructing a mathematical model framework including state parameters, operating status, and constraints; filtering, interpolating and sparsifying, time series matching, combined calculation, dimensional transformation, and normalization processing the measured parameters obtained by the microgrid to obtain processed data; parsing the current system state parameters based on the processed data; and predicting future state changes through a genetic algorithm to obtain a predicted result of the future state. The intelligent control scheme of the present invention predicts and optimizes the DC bus flow, effectively improving system stability while meeting the operating requirements of the backup battery directly connected to the quality bus, and supporting energy cost-optimized scheduling.

Description

Technical Field

[0001] The present invention relates to the technical field of power system control, and in particular to a method for controlling a DC bus current connected to a backup battery. Background Art

[0002] The monitoring and control units of traditional uninterruptible DC power supply systems are primarily designed to manage the charging and discharging processes of backup batteries, and this is achieved solely by controlling the output of the AC rectifier. When the DC bus is connected to renewable energy power generation and energy storage devices, this single control logic cannot effectively manage the coordinated operation of multiple power sources and energy storage devices, potentially negatively impacting the safety and service life of the backup battery. Existing monitoring and control units are unable to comprehensively monitor the operating modes of each device connected to the DC bus, including the status and requirements of renewable energy power generation, energy storage devices, and loads. Therefore, flexible scheduling and utilization of energy sources based on different costs or priorities cannot be achieved, limiting the system's energy efficiency and economic benefits.

[0003] Traditional systems lack the ability to predict future trends in renewable energy sources, energy storage devices, and electricity loads. This means the system cannot make advance scheduling plans to optimize energy distribution and reduce operating costs. This lack of forward-looking data makes the system vulnerable to emergencies or energy price fluctuations. The power flow control algorithms for DC microgrids are typically designed for a constant-voltage DC bus. However, the DC bus of a traditional uninterruptible DC power supply system must be connected to a backup battery, whose voltage varies with the battery's status. This makes existing power flow control algorithms incapable of directly applying to this new type of DC power supply system and unable to meet its unique control requirements. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a DC bus current control method connected to a backup battery, which effectively solves the problem of operability of the microgrid.

[0005] In order to solve the above technical problems, the technical solutions of the present invention are as follows:

[0006] In a first aspect, a method for controlling a DC bus current connected to a backup battery is provided, the method comprising:

[0007] Define the topology of the DC microgrid and build a mathematical model framework including state parameters, operating status and constraints;

[0008] The measurement parameters obtained by the microgrid are filtered, interpolated and sparse, time series matched, combined and calculated, dimensionally transformed, and normalized to obtain processed data;

[0009] Analyze the current system status parameters based on the processed data;

[0010] By randomly generating an initial population and iteratively performing mutation, crossover, and selection operations, the final solution is finally determined to predict future state changes and obtain the prediction results of the future state;

[0011] Based on the predicted results of future states, power flow, and energy costs, a control plan is developed and distributed to each controller;

[0012] The control scheme is continuously optimized based on the execution feedback process to respond to changes in system status and environment.

[0013] Furthermore, the topology of the DC microgrid is defined, and a mathematical model framework including state parameters, operating status, and constraints is constructed, including:

[0014] The DC microgrid is divided into a primary circuit topology and a secondary circuit topology with the DC bus as the center. The primary circuit includes the DC bus, distribution branches, intermediate control layer, and branch equipment for power flow. The secondary circuit includes a virtual converged data bus, branch virtual measurement devices, and equipment communication ports for data collection and transmission. The branch equipment includes power generation, energy storage, and load equipment, and at least one backup battery pack directly connected to the bus. The final DC microgrid topology is obtained.

[0015] According to the topological structure of the DC microgrid, a mathematical model framework of the microgrid is constructed, including state parameters, defined operating states and defined constraint sets. Among them, the state parameters include bus voltage, branch current and direction, branch power, controller efficiency, device connection port voltage and current, inflow and outflow electric power, inflow and outflow electric energy price, energy storage margin and value, and energy storage loop loss rate per unit time; the operating state includes various states of the bus, distribution branch, intermediate controller and branch equipment; the constraint set includes various operating restrictions of the bus, distribution branch, intermediate controller and branch equipment.

[0016] Furthermore, the measurement parameters obtained by the microgrid are filtered, interpolated and sparse, time series matched, combined and calculated, dimensionally transformed, and normalized to obtain processed data, including:

[0017] Obtain raw measurement parameter data from sensors and monitoring systems in the microgrid;

[0018] Process the original measurement parameter data to remove noise and outliers to obtain filtered data;

[0019] Interpolate the missing values ​​in the filtered data and perform sparse processing on the overly dense data points to obtain the interpolated and sparse processed data;

[0020] Align the interpolated and sparsely processed data to a unified time series to obtain time series matching data;

[0021] Perform combined calculations based on the matched data of the time series to obtain new derivative data;

[0022] Convert the measurement data and new derived data of different dimensions into a unified dimension to obtain dimension-transformed data;

[0023] The data after dimension transformation is normalized to obtain the processed data.

[0024] Furthermore, based on the processed data, the current system status parameters are analyzed, including:

[0025] Using a preset neural network model, the extracted system status-related features are used as input data to identify the current state of the system, including normal operation, overload, undervoltage, overvoltage, and frequency deviation;

[0026] Based on the identified current state of the system, state parameters are calculated, including the system's stability index, load level, and failure probability.

[0027] Furthermore, a genetic algorithm is used to randomly generate an initial population and iteratively perform mutation, crossover, and selection operations to ultimately determine the final solution to predict future state changes. The prediction results of the future state are obtained, including:

[0028] Randomly generate an initial population, each individual represents a set of state parameter prediction values;

[0029] For each individual in the population, three different individuals are identified to generate a new variant individual;

[0030] Cross the new mutant individuals with the original individuals to generate test individuals;

[0031] Use the fitness function to evaluate each individual in the test individual set and the original individual set, calculate the fitness value of each individual, and determine whether the individual will enter the next generation population based on the size of the fitness value;

[0032] Repeat the mutation, crossover, and selection operations until the preset number of iterations is reached, and determine the final solution from the final population based on the individual fitness values;

[0033] Based on the final solution, the future state of the DC microgrid is analyzed, including state change trends and anomaly detection.

[0034] Based on the final solution and the analysis of the future state of the DC microgrid, the prediction results of the future state are generated, including the predicted value, prediction error range, and confidence key information.

[0035] Furthermore, the fitness value calculation formula for each individual is:

[0036]

[0037] Among them, F represents the fitness value; w e represents the prediction error weight; n represents the number of observations; y i represents the actual observed value; represents the predicted value; w s represents the stability weight; Represents the mean of the predicted values.

[0038] Furthermore, based on the predicted results of future states, power flow, and energy costs, a control plan is developed and distributed to each controller, including:

[0039] Collect relevant forecast data on future status, mainly including the power load demand forecast of each branch, and obtain current power flow information, including the power supply status of the power grid, the voltage and current of each node; collect energy cost data, including electricity prices in different time periods and energy consumption characteristics of equipment;

[0040] Construct a prediction model to predict the future state, and use the collected prediction data of the future state, mainly the power load prediction of each branch, as input data, input it into the prediction model to obtain the prediction result of the future state;

[0041] Develop control plans based on future state predictions, power flow, and energy cost information, including adjusting equipment operating power, optimizing power usage hours, and starting or shutting down equipment;

[0042] The developed control scheme is encoded into instructions recognized by the controller, and the control instructions are sent to each controller.

[0043] In a second aspect, a DC bus power flow control system connected to a backup battery includes:

[0044] The acquisition module is used to define the topology of the DC microgrid and build a mathematical model framework that includes state parameters, operating status, and constraints. It performs filtering, interpolation and sparseness, time series matching, combined calculation, dimension transformation, and normalization on the measured parameters obtained from the microgrid to obtain processed data. Based on the processed data, the current system state parameters are analyzed.

[0045] The processing module is used to randomly generate an initial population and iteratively perform mutation, crossover, and selection operations to ultimately determine the final solution, thereby predicting future state changes and obtaining prediction results of the future state; based on the prediction results of the future state and the power flow and energy cost, a control plan is formulated and issued to each controller; and the control plan is continuously optimized according to the execution feedback program to cope with changes in the system state and environment.

[0046] According to a third aspect, a computing device includes:

[0047] one or more processors;

[0048] The storage device is used to store one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors implement the method.

[0049] In a fourth aspect, a computer-readable storage medium stores a program, which implements the method when executed by a processor.

[0050] The above solution of the present invention includes at least the following beneficial effects:

[0051] By defining the DC microgrid topology and constructing a mathematical model framework encompassing state parameters, operating states, and constraints, a solid theoretical foundation is provided for the system's stable operation. This helps accurately describe the system's dynamic behavior and promptly identify and address potential instabilities. The measured parameters acquired from the microgrid are filtered, interpolated and sparsified, time series matched, combined, calculated, dimensionally transformed, and normalized, effectively improving data accuracy and reliability. Furthermore, by randomly generating an initial population and iteratively performing mutation, crossover, and selection operations, the final solution is determined to predict future state changes. This data processing and prediction method, based on intelligent algorithms, more accurately grasps future system trends and provides strong support for the development of control plans. Based on the predicted future state, power flow, and energy costs, a control plan is developed and distributed to each controller. This control strategy, which comprehensively considers power flow and energy costs, helps achieve efficient energy utilization, reduce system operating costs, and improve economic benefits.

[0052] By continuously optimizing the control scheme based on feedback loops to respond to changes in system status and environment, the system's adaptability and flexibility are enhanced. This enables the system to better cope with various emergencies and uncertainties, ensuring stable operation and power supply quality. The entire control process, based on intelligent algorithms and data processing technologies, enables intelligent management of the microgrid. This helps reduce manual intervention, improve management efficiency, and lower operation and maintenance costs, providing a strong guarantee for the long-term, stable operation of the microgrid. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 The present invention provides a flow chart of a method for controlling a DC bus current connected to a backup battery.

[0054] Figure 2 The figure is a schematic diagram of a DC bus power flow control system connected to a backup battery provided by an embodiment of the present invention.

[0055] Figure 3 This is a topology diagram of the primary circuit of a DC microgrid provided by an embodiment of the present invention.

[0056] Figure 4 This is an example diagram of a primary circuit topology diagram of a DC microgrid provided by an embodiment of the present invention.

[0057] Figure 5 This is a topology diagram of a secondary circuit of a DC microgrid provided by an embodiment of the present invention.

[0058] Figure 6 This is an example diagram based on the primary loop topology diagram of a DC microgrid provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0059] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0060] like Figure 1 As shown, an embodiment of the present invention provides a method for controlling a DC bus current connected to a backup battery, the method comprising the following steps:

[0061] Step 11: define the topology of the DC microgrid and construct a mathematical model framework including state parameters, operating status and constraints;

[0062] Step 12: filtering, interpolation and sparseness, time series matching, combined calculation, dimension transformation, and normalization are performed on the measurement parameters obtained by the microgrid to obtain processed data;

[0063] Step 13: Analyze the current system status parameters based on the processed data;

[0064] Step 14: Using a genetic algorithm, by randomly generating an initial population and iteratively performing mutation, crossover, and selection operations, a final solution is finally determined to predict future state changes and obtain a prediction result of the future state;

[0065] Step 15: Based on the predicted results of the future state, power flow, and energy cost, a control plan is formulated and distributed to each controller;

[0066] Step 16: Continuously optimize the control scheme according to the execution feedback procedure to cope with changes in system status and environment.

[0067] In an embodiment of the present invention, by clarifying the topological structure of the microgrid, a clear foundation is provided for power flow calculation and control strategy formulation. A mathematical model framework including state parameters, operating status and constraints is constructed, so that the system can more accurately describe and simulate the actual operating conditions of the microgrid, providing a solid theoretical basis for optimization control. By performing a series of preprocessing on the measurement parameters, noise and outliers are effectively removed, and the accuracy and reliability of the data are improved. At the same time, through dimensional transformation and normalization processing, parameters of different dimensions and value ranges can be compared and calculated, providing a unified data basis for state analysis and prediction. By analyzing the processed data, the current state parameters of the microgrid, such as voltage, current, power, etc., can be obtained in real time, providing real-time and accurate information support for the formulation of the control plan.

[0068] Using intelligent optimization algorithms to predict future states fully accounts for system uncertainty and nonlinear characteristics, resulting in more accurate and reliable predictions. This helps identify potential problems and risks in advance, providing forward-looking guidance for control plan development. By combining future state predictions with power flow and energy costs, the resulting control plan better meets the system's actual operational needs and economic efficiency targets. By distributing these predictions to each controller for execution, precise control of the microgrid is achieved, improving system efficiency and stability. By collecting real-time execution feedback and continuously optimizing and adjusting the control plan, the system can better adapt to changing conditions and environmental fluctuations. This closed-loop control strategy enhances the system's adaptability and robustness, ensuring the long-term stable operation of the microgrid.

[0069] In a preferred embodiment of the present invention, the above step 11, defining the topology of the DC microgrid and constructing a mathematical model framework including state parameters, operating status and constraints, may include:

[0070] Step 112: The DC microgrid is divided into a primary circuit topology and a secondary circuit topology with the DC bus as the center. The primary circuit includes the DC bus, distribution branches, an intermediate control layer, and branch devices for power flow; the secondary circuit includes a virtual converged data bus, branch virtual measurement devices, and device communication ports for data collection and transmission; the branch devices include power generation, energy storage, and load devices, and include at least one backup battery pack directly connected to the bus, ultimately obtaining the DC microgrid topology.

[0071] Step 113: Based on the topological structure of the DC microgrid, a microgrid mathematical model framework is constructed, including state parameters, defined operating states, and defined constraint sets. The state parameters include bus voltage, branch current and direction, branch power, controller efficiency, device connection port voltage and current, inflow and outflow power, inflow and outflow electricity price, energy storage margin and value, and energy storage loop loss rate per unit time. The operating states include various states of the bus, distribution branches, intermediate controllers, and branch devices. The constraint set includes various operational restrictions of the bus, distribution branches, intermediate controllers, and branch devices.

[0072] In the embodiment of the present invention, the primary circuit and the secondary circuit are divided:

[0073] Primary circuit: centered on the DC bus, it connects the distribution branches, intermediate control layers, and branch equipment, and is responsible for the flow and distribution of electrical energy.

[0074] Secondary circuit: Build a virtual converged data bus to achieve data collection, transmission and monitoring through branch virtual measurement devices and equipment communication ports.

[0075] Branch circuit equipment includes power generation equipment (such as solar panels, wind turbines, and diesel generator sets), energy storage equipment (such as smart lithium batteries), and load equipment (such as transmission, 4G / 5G equipment, and lighting). Ensure that at least one backup battery branch circuit is included to provide backup power when necessary. Based on these elements, draw a DC microgrid topology diagram, clearly showing the connections between the primary and secondary circuits, as well as the distribution of branch circuit equipment.

[0076] Define the status parameters:

[0077] Bus voltage: Indicates the voltage level of the DC bus.

[0078] Branch current and direction: Describes the magnitude and direction of the current in each branch.

[0079] Branch Power: Calculates the power flow on a branch.

[0080] Controller efficiency: reflects the energy conversion efficiency of the intermediate controller.

[0081] Device connection port voltage and current: Records the voltage and current values ​​at the device access point.

[0082] Inflow and outflow electrical power: represent the input and output power of the equipment or branch respectively.

[0083] Inflow and outflow electricity prices: price factors considered when electricity is traded.

[0084] Energy storage margin and value: represents the remaining energy of the energy storage equipment and its economic value.

[0085] Energy storage circuit loss rate per unit time: describes the loss of energy storage equipment during the process of storing and releasing energy.

[0086] Define various possible states for busbars, distribution branches, intermediate controllers, and branch devices, such as normal operation, fault, maintenance, and standby. Establish a series of operational limits and constraints based on device characteristics, system requirements, and safety regulations, such as busbar voltage fluctuation range, branch current limits, and energy storage device charge and discharge rate limits.

[0087] Consider a simple DC microgrid consisting of the following elements:

[0088] DC bus: voltage level is 48V.

[0089] DC distribution branch: more than 6 branches, respectively connected to the solar controller backup battery, energy storage lithium battery and load equipment.

[0090] Intermediate control equipment: includes a solar controller bidirectional DC / DC, AC rectifier, etc., used for power conversion and regulation of DC output voltage and current.

[0091] Power generation equipment: One solar panel with a rated power of 2kW.

[0092] Backup battery: 2 sets of lead-acid batteries, rated voltage 48V, capacity 500Ah.

[0093] Energy storage equipment: 2 smart lithium batteries, each with a capacity of 5kWh.

[0094] Load equipment: A group of 4G / 5G communication equipment occupies two distribution branches with a rated power of 5kW.

[0095] DC bus voltage: 54V (under normal operation).

[0096] Branch current and direction: solar controller branch current I1 (direction is positive), load branch current I2 (direction is negative).

[0097] Branch power: solar controller output power P1, load input power P2.

[0098] Controller efficiency: Solar controller converter efficiency η.

[0099] Device connection port voltage and current: solar controller port voltage V1 and current I1, load port voltage V2 and current I2.

[0100] Inflow and outflow of electric power: Solar inflow power P i , load outflow power P o .

[0101] Energy storage margin and value: the remaining power Qs of the smart lithium battery, the economic value V of the battery b .

[0102] Backup battery energy reserve: The remaining power of the backup battery is QB, and the economic value of the battery is also V b .

[0103] Energy storage and backup power circuit loss rate per unit time: battery loss rate r.

[0104] Busbar: Normal operation / fault. Distribution branch: On / off. Controller: Operating / Standby / Fault. Backup power and energy storage battery: Charging / Discharging / Standby / Protection. Load equipment: Running / Stop (power-off protection).

[0105] The constraints set are: bus voltage fluctuation range: 42V-58V, voltage follows the backup lead-acid battery port voltage. Branch current limit: I1 ≤ 15A, I2 ≤ 10A. Energy storage device charge and discharge rate limit: charge rate ≤ 0.1C, discharge rate ≤ 0.15. Battery discharge protection: low voltage protection mode.

[0106] Dividing the DC microgrid into primary and secondary circuits clarifies the paths for power flow, data collection, and transmission, providing a clear foundation for system design and optimization. By defining state parameters such as bus voltage, branch current and direction, and branch power, the system's operating status can be comprehensively reflected, providing accurate data support for system monitoring, analysis, and optimization. Defining the various states of the bus, distribution branches, intermediate controllers, and branch equipment helps accurately determine the system's operating status and promptly detect and address abnormal conditions. By defining a set of constraints, including various operational limits for the bus, distribution branches, intermediate controllers, and branch equipment, the system can operate within a safe and reasonable range, avoiding system instability or failures caused by improper operation. Based on the mathematical model framework, more precise and optimized control strategies can be developed to achieve accurate control of system status and improve system stability and reliability.

[0107] Incorporating the prices of incoming and outgoing electricity into state parameters helps fully consider economic factors when formulating control strategies, achieving efficient use of electricity and minimizing costs. By defining state parameters such as the energy storage margin and value, and the energy storage circuit loss rate per unit time, energy storage equipment can be managed more scientifically, energy storage strategies can be optimized, and the utilization efficiency of energy storage equipment and the overall energy efficiency of the system can be improved. The setting of secondary circuits and the application of virtual measurement devices and equipment communication ports enable automatic data collection and transmission, providing a foundation for the intelligence and automation of the system. Based on the mathematical model framework, an intelligent solution support system can be developed to achieve real-time monitoring, analysis, and solutions for system status, improving the system's intelligence level and operational efficiency. The topology of the DC microgrid can be flexibly adjusted according to actual needs to adapt to microgrid systems of different scales and complexities.

[0108] In a preferred embodiment of the present invention, the above step 12, filtering, interpolation and sparseness, time series matching, combined calculation, dimension conversion, and normalization of the measurement parameters obtained by the microgrid to obtain processed data, may include:

[0109] Step 123, obtaining raw measurement parameter data from sensors and monitoring systems of the microgrid;

[0110] Step 124, processing the original measurement parameter data to remove noise and outliers to obtain filtered data;

[0111] Step 125, interpolating missing values ​​in the filtered data and performing sparse processing on overly dense data points to obtain interpolated and sparse processed data;

[0112] Step 126: align the interpolated and sparsely processed data to a unified time series to obtain time series matched data;

[0113] Step 137, performing combined calculations based on the time series matched data to obtain new derivative data;

[0114] Step 128, converting the measurement data of different dimensions and the new derived data into a unified dimension to obtain dimension-transformed data;

[0115] Step 129: normalize the data after dimension transformation to obtain processed data.

[0116] In an embodiment of the present invention, raw measurement parameter data is obtained from sensors and monitoring systems of the microgrid in real time or periodically. These data include various physical quantities such as voltage, current, power, frequency, and temperature. A filtering algorithm, such as median filtering, is applied to the raw measurement parameter data to remove noise and outliers in the data. The filtered data should be smoother and more accurately reflect the actual operating status of the microgrid. Median filtering is a nonlinear filtering method that replaces the value of the data point with the median of the data point and its neighborhood. For any data point x j (can be voltage, current, power, etc.), median filtering can be expressed as: filtered data = Median{x j -k,…,x j ,…,x j +k}; where k is the size of the filter window, Median represents the median operation, and j represents the index of the data point. For missing values ​​in the filtered data, interpolation methods (such as linear interpolation) are used to fill them. For data points that are too dense, sparse processing is performed according to actual needs to reduce data redundancy and improve data processing efficiency. Assume d c is a missing value, and its adjacent non-missing value is d c -1 and d c +1, then the linear interpolation can be expressed as: j =d j -1+(j+1)-(j-1)(d j +1-d j -1)×(j-(j-1))=2d j -1+d j +1. Aligning interpolated and sparsely processed data to a unified time series ensures consistent timestamps for all data points. This facilitates data analysis and processing, especially when comparing data from different time points. Based on the aligned time series data, combined calculations can be performed to generate new derivative data.

[0117] For example, the product of voltage and current is calculated as power p = V × I, where V is voltage and I is current. Convert measured data and newly derived data of different dimensions to a uniform dimension. For example, convert power from watts to kilowatts to ensure that all data has the same dimension, facilitating comparison and analysis. Normalize the transformed data to scale it to a specific range (e.g., between 0 and 1). Normalization helps eliminate dimensional differences between data and improves the accuracy and efficiency of data processing.

[0118] Assume that the following raw measurement parameter data are obtained from the sensors of the microgrid:

[0119] Voltage (V): [53.5, 53.6, 53.5, ..., 53.7] (unit: volt).

[0120] Current (I): [5, 5.1, 4.9, ..., 5.2] (unit: ampere). Apply the mean filter algorithm to smooth the voltage and current data to remove noise and outliers. Assume that at a certain point in time, the voltage data is missing, and use linear interpolation to fill the missing value. At the same time, if the current data is too dense, we choose to retain some data points according to actual needs and perform sparse processing. Align the filtered and sparse voltage and current data to a unified time series to ensure that each data point has a corresponding timestamp. Calculate the product of voltage and current to obtain power data: [p1, p2, p3..., p m ]. Convert power from watts to kilowatts (P / 1000). Scale the voltage and power data after dimension transformation to between 0 and 1. After the above processing steps, the processed data is obtained. This data is smoother, more accurate, and has a unified dimension and normalized range, which is convenient for data analysis and processing.

[0121] Filtering removes noise and outliers from the raw measurement parameter data, making the data smoother and more accurate, reflecting the actual operating status of the microgrid more realistically. Missing values ​​in the filtered data are interpolated to ensure data integrity and continuity, preventing the adverse effects of missing data on subsequent analysis. Overly dense data points are thinned to reduce data redundancy, improve data processing and storage efficiency, and reduce computational costs. The interpolated and thinned data are aligned to a unified time series to ensure temporal consistency between different measurement parameters, facilitating subsequent data analysis and comparison. Combined calculations are performed based on the matched time series data to generate new derived data (such as power, energy, impedance, and price difference), providing more information support for microgrid monitoring, control, and optimization. Measurement data and new derived data of different dimensions are converted to a unified dimension, eliminating comparison barriers caused by dimensional differences and enabling direct comparison and analysis. The transformed data are normalized to scale the data to a specific range, enhancing comparability and interpretability, and facilitating the discovery of patterns and trends within the data.

[0122] In a preferred embodiment of the present invention, the above step 13, parsing the current system state parameters based on the processed data, may include:

[0123] Step 131: Using a preset neural network model, the extracted system state-related features are used as input data to identify the current state of the system, including normal operation, overload, undervoltage, and overvoltage;

[0124] Step 132 : Calculate state parameters based on the identified current state of the system, including the stability index, load level, and failure probability of the system.

[0125] Assume that the voltage time series data obtained after processing is [V1, V2, ..., V r ], the current time series data is [I1, I2, ..., I b Using a pre-set neural network model, the extracted features are used as input data to identify the current state of the system. A trained neural network model, such as a long short-term memory (LSTM) network, is constructed. The extracted features (voltage level, current, power factor, and energy consumption) are fed into the neural network as input data. The neural network outputs the current state of the system, such as normal operation, overload, undervoltage, or overvoltage.

[0126] Assume that a trained LSTM model is used. The extracted features [voltage level, current size, power factor, energy consumption] are input as input vectors to the LSTM model. The model outputs a probability distribution, which represents the probability of the system being in various states. For example, [0.9, 0.05, 0.02, 0.02, 0.01] means that there is a 90% probability that the system is in normal operation, a 5% probability that it is in an overload state, and so on. Based on the identified system state, the state parameters are calculated, including the system load level and the failure probability. The load level of the system Among them, I c Indicates the current; I r Indicates rated current; V f The fault probability FP = 1-P n Among them, P n Indicates the normal operation status of the power system.

[0127] Different stability index thresholds are set based on the system status. The load level is calculated as the ratio of the current load to the rated load. The failure probability is estimated based on historical data and the current status. If the system is in normal operation, the stability index can be set to 1; if it is in an overload or undervoltage state, the stability index can be reduced; if it is in an overvoltage or frequency offset state, the stability index can be further reduced. Assume that the current load is L and the rated load is L r , the load level is Based on the duration of various states and the frequency of failures in historical data, the probability of failure in the current state is estimated. For example, if the system is in an overloaded state for a long time, the probability of failure will increase.

[0128] Based on the identified system status, the calculated stability index quantifies the system's stability level. This helps operators promptly identify potential instability factors and take appropriate measures to prevent system crashes or failures, thereby improving overall system stability. The load level calculation provides the ratio of the current system load to the rated load or maximum load capacity. This helps operators understand the system's load situation and appropriately adjust load distribution to avoid overload operation, thereby extending equipment life and improving system efficiency. The failure probability calculation, based on historical system data and current status, predicts the likelihood of system failure within a given period of time. This helps operators plan maintenance in advance, prevent failures, and reduce the risk of power outages and equipment damage caused by failures. By comprehensively considering the system's stability index, load level, and failure probability, operators gain a more comprehensive understanding of the system's operating status and risks. This provides strong data support for their plan formulation, helping to optimize system operation strategies and improve system reliability and cost-effectiveness.

[0129] In a preferred embodiment of the present invention, step 14 utilizes a genetic algorithm to randomly generate an initial population and iteratively perform mutation, crossover, and selection operations to ultimately determine a final solution to predict future state changes. The predicted results of the future state may include:

[0130] Step 145, randomly generate an initial population, each individual representing a set of state parameter prediction values;

[0131] Step 146, for each individual in the population, determine three different individuals to generate a new variant individual;

[0132] Step 147, cross the new variant individual with the original individual to generate a test individual;

[0133] Step 148, using the fitness function to evaluate each individual in the test individual set and the original individual set, calculating the fitness value of each individual, and determining whether the individual will enter the next generation population based on the size of the fitness value;

[0134] Step 149, repeating the mutation, crossover, and selection operations until a preset number of iterations is reached, and determining a final solution from the final population based on individual fitness values;

[0135] Step 150 , analyzing the future state of the DC microgrid based on the final solution, including state change trends and anomaly detection;

[0136] Step 151 : Based on the final solution and the analysis of the future state of the DC microgrid, a prediction result of the future state is generated, including a prediction value, a prediction error range, and confidence key information.

[0137] In an embodiment of the present invention, a population size is set (e.g., 100 individuals). Each individual represents a set of predicted values ​​for the state parameters of the DC microgrid (e.g., voltage, current, power, etc.). These parameters can be randomly generated or initialized based on historical data. For example, each individual in the initial population is a vector containing predicted values ​​for voltage, current, and power. For each individual in the population, three different individuals are identified as parents. A new mutant individual is generated using a mutation strategy (e.g., randomly changing the value of a parameter). The mutation operation aims to increase the diversity of the population and explore new solution spaces. The new mutant individual is cross-pollinated with the original individual to generate a test individual. The crossover operation can be implemented by exchanging some parameter values ​​or adopting other crossover strategies. The test individual inherits some characteristics of the parent individual while introducing new mutations. A fitness function is used to evaluate each individual in the test individual set and the original individual set. The fitness function can be defined based on the actual operating conditions of the DC microgrid and the prediction target, such as the sum of squares of the prediction error, energy efficiency index, etc. Based on the size of the fitness value, individuals with higher fitness are selected to enter the next generation of the population.

[0138] Mutation, crossover, and selection operations are repeated until the preset number of iterations is reached or other stopping conditions are met. With each iteration, the population evolves toward a more optimal solution, gradually approaching the final solution. The final solution—the set of predicted state parameter values ​​with the best prediction performance—is determined from the final population based on individual fitness values. Based on the final solution, the future state of the DC microgrid is analyzed, including state change trends and anomaly detection. Data analysis tools or visualization methods can be used to present the prediction and analysis results. Based on the final solution and the analysis of the future state of the DC microgrid, a prediction of the future state is generated. The prediction results can include key information such as the predicted value, prediction error range, and confidence level. This information helps operators understand the future operating state of the DC microgrid and formulate appropriate plans and plans.

[0139] Suppose you want to predict the voltage variation of a DC microgrid for the next day. An initial population is randomly generated: 100 individuals are generated, each containing a vector of 24-hour voltage forecasts. For each individual, three different individuals are randomly selected as parents, and the voltage forecast for one hour is randomly changed to generate a variant individual. The variant individual is crossed with the original individual to generate a test individual. For example, the voltage forecasts for certain hours can be swapped between two individuals. The test individuals and the original individuals are evaluated using a fitness function (such as the sum of squared prediction errors), and the individual with the lower fitness is selected for the next generation of the population. This process is repeated 100 times (or until other stopping conditions are met). The individual with the lowest fitness is selected from the final population as the final solution, which is the set of voltage forecasts with the best prediction performance. Based on this final solution, the voltage variation trend and possible anomalies for the DC microgrid for the next day are analyzed. A prediction result report is generated, including the predicted voltage value, prediction error range, and confidence level. For example, the voltage for a certain hour can be predicted to be ±0.1V of the backup battery voltage (with a 95% confidence level).

[0140] By simulating the biological evolution process, the algorithm can search for near-final solutions within a complex solution space. By iteratively performing mutation, crossover, and selection operations, the algorithm continuously optimizes individuals (i.e., predicted state parameter values) so that the final solution is closer to the actual value. This improves the accuracy of DC microgrid future state predictions. In DC microgrid predictions, the algorithm adaptively adjusts predicted state parameter values ​​to account for various uncertainties and changes in grid operation. During the analysis of the DC microgrid's future state, it can identify changing trends and anomalies in state parameters. This helps operators promptly identify potential grid problems, such as equipment failures and overloads, and take appropriate preventive measures. Early warning and intervention can avoid or minimize the losses and impacts of grid failures. Prediction results include key information such as predicted values, prediction error ranges, and confidence levels, providing comprehensive solution support for operators. Based on the prediction results, operators can formulate more effective grid operation, maintenance, and emergency response plans. This helps improve grid efficiency and reliability, while reducing operating costs and risks.

[0141] In a preferred embodiment of the present invention, the fitness value of each individual is calculated as follows:

[0142]

[0143] Among them, F represents the fitness value; w e represents the prediction error weight; n represents the number of observations; y i represents the actual observed value; represents the predicted value; w s represents the stability weight; Represents the mean of the predicted values.

[0144] In the embodiment of the present invention, w e is the prediction error weight, which is used to adjust the importance of the prediction error in the fitness value calculation. s is the stability weight, which is used to adjust the importance of prediction stability (i.e., the degree of change in the predicted value) in the fitness value calculation. n is the number of observations, i.e., how many actual observations are compared with the predicted values. i (for i=1 to n) is the array of actual observations. (For i = 1 to n) is the array of predicted values. Calculate the square of the difference between the actual observation value and the predicted value for each observation point and sum these square values: Calculate the square of the error for each point (For i=1 to n). Sum all squared errors The sum of squared errors is divided by the number of observations n to get the average prediction error. Compute the square of the difference between each predicted value and the mean: (for i=1 to n). Sum the squares of all differences and divide by n-1 to get Take the square root of the obtained variance to get the standard deviation of the predicted value. Finally, the fitness value F is obtained.

[0145] The mean square error (MSE) between the predicted value and the actual observed value is calculated in part, which is a common indicator of forecast accuracy. e (Prediction Error Weight) adjustment can emphasize or weaken the importance of prediction accuracy in fitness calculation. The standard deviation of the predicted value is calculated in part to reflect the fluctuation or stability of the predicted results. s (Stability Weight) adjustment can emphasize or weaken the importance of predicted stability in fitness calculation. e and w s As a weight parameter, it allows users to adjust the relative importance of prediction accuracy and stability in the fitness calculation according to specific application scenarios and needs. This flexibility makes the formula applicable to different optimization objectives and problem contexts.

[0146] In a preferred embodiment of the present invention, the above step 15, formulating a control plan based on the prediction results of the future state and the power flow and energy cost and sending it to each controller, may include:

[0147] Step 151: Collect relevant forecast data for future states, including weather conditions and user demand forecasts, and obtain current power flow information, including the power supply status of the power grid, the voltage and current of each node; collect energy cost data, including electricity prices in different time periods and energy consumption characteristics of equipment;

[0148] Step 152: construct a prediction model to predict the future state, and use the collected prediction data of the future state, including weather conditions and user demand prediction, as input data into the prediction model to obtain the prediction result of the future state;

[0149] Step 153: Develop a control plan based on the future state prediction results and the power flow and energy cost information, including adjusting the operating power of the equipment, optimizing the power consumption period, and starting or shutting down the equipment;

[0150] In step 154 ​​, the developed control scheme is encoded into instructions recognized by the controller, and the control instructions are sent to each controller.

[0151] In an embodiment of the present invention, weather data for a future period (e.g., 24 or 48 hours) is obtained from a meteorological agency or weather forecast API, including temperature, humidity, wind speed, and irradiance. This data is crucial for predicting the power generation of renewable energy sources (e.g., solar and wind power). Based on historical electricity usage data, user behavior patterns, holiday effects, and other factors, a prediction model (e.g., time series analysis or machine learning algorithms) is used to predict user electricity demand for each future time period. The total power supply of the power grid, the load of each power line, and the presence of faults or maintenance plans are monitored. Sensors or smart meters in the power grid collect real-time voltage and current data at each node to assess the stability and security of the grid. Time-of-use electricity price information is obtained from the power market, typically including peak, normal, and off-peak prices, as well as possible special pricing policies (e.g., holiday prices). Parameters such as energy consumption curves, power factor, and efficiency of each electrical device are collected to calculate energy costs under different operating modes. An appropriate prediction method (e.g., a neural network) is selected and the model is trained based on historical data and the collected future state prediction data. The model is validated and tuned to ensure its prediction accuracy. The collected data, such as weather conditions and user demand forecasts, is fed into the forecasting model. The model then outputs forecasts for various time periods in the future, including renewable energy generation, user power demand, and grid load.

[0152] Combining future state predictions with current power flow information, the system analyzes the grid's supply and demand balance, fluctuations in renewable energy generation, and peak and valley periods of user demand. Based on electricity prices and the energy consumption characteristics of each device during different time periods, the system calculates the energy consumption cost of each device during different operating periods. Adjusting device operating power: Adjust the operating power of adjustable devices based on grid load and renewable energy generation to balance supply and demand. High-energy-consuming devices are scheduled to operate during periods with lower electricity prices to reduce electricity costs. Based on the predictions and actual demand, it determines whether to activate backup power supplies, energy storage devices, or shut down non-essential equipment. The developed control plan is converted into a command format that the controller can recognize, including the device number, control action (such as start, stop, power adjustment), and execution time. The control commands are sent to each controller via a communication network (such as a wired network, wireless network, or the Internet of Things). Upon receiving the command, the controller executes the corresponding control action to adjust the device's operating status.

[0153] Consider a microgrid system consisting of a solar power station, a wind power station, a battery storage system, and multiple user loads. A forecast predicts cloudy skies and low wind speeds for the coming day, resulting in a reduction in renewable energy generation. Furthermore, user electricity demand is predicted to increase significantly in the evening. Based on this information, the following control plan is developed:

[0154] Reduce the operating power of non-essential equipment to reduce the load on the grid. Encourage users to use electricity during the day when electricity prices are lower and reduce their use at night when prices are higher. During periods of insufficient renewable energy generation, discharge energy storage batteries to supplement power; during periods of sufficient renewable energy generation, charge energy storage batteries for future use.

[0155] By collecting real-time grid power supply information, including voltage and current information at each node, and combining it with forecasts of future conditions (such as load fluctuations caused by weather changes), grid overloads or faults can be identified and prevented in advance, ensuring the stable operation of the microgrid system. Adjusting device operating power and power usage time based on these forecasts allows for more efficient energy allocation. By collecting electricity price information for different time periods and combining it with load demand forecasts, strategies can be developed to optimize power usage time, scheduling high-energy-consuming tasks during periods with lower prices, significantly reducing user electricity costs. Adjusting device operating power based on their energy consumption characteristics allows them to operate at a more efficient level, reducing energy waste and further lowering energy costs. By forecasting weather conditions (such as sunlight intensity and wind speed), the power generation of renewable energy sources (such as solar and wind) can be more accurately predicted, enabling more efficient renewable energy utilization plans and improving their grid integration and utilization rates.

[0156] By optimizing power usage periods and equipment operating power, we can reduce dependence on traditional fossil fuels (such as coal and oil), lower carbon emissions, and promote sustainable energy use and environmental protection. Control plans can be dynamically adjusted based on actual conditions, such as rapid responses to weather changes, changes in user demand, or grid failures, enhancing system flexibility and adaptability. By building predictive models, collecting and analyzing large amounts of data to develop control plans, and encoding them into controller-recognizable instructions, these plans are then distributed to each controller, enabling intelligent management of the power system and improving both efficiency and accuracy.

[0157] In a preferred embodiment of the present invention, the above step 16, continuously optimizing the control scheme according to the execution feedback program to cope with changes in system status and environment, may include:

[0158] In this embodiment of the present invention, a series of key performance indicators (KPIs), such as microgrid stability, equipment energy efficiency, electricity costs, and renewable energy utilization, are set to evaluate the effectiveness of control decision execution. Monitoring devices, such as sensors and smart meters, are deployed within the microgrid system to collect real-time information on grid status, equipment operating status, and energy consumption. Data communication between the monitoring devices and the central control system is ensured to facilitate timely feedback on execution. Based on the previously developed control plan, control instructions are encoded and issued to each controller. These instructions include equipment adjustments and optimized power usage periods. Monitoring devices collect real-time data during the execution process, such as actual equipment operating power, grid voltage and current changes, and actual user electricity usage. Based on this collected data, the actual values ​​of each feedback indicator are calculated and compared with the expected values ​​to evaluate the effectiveness of the control decision execution. This collected feedback data is then analyzed in depth to identify issues, deviations, or deficiencies in the execution process, such as delayed equipment response or energy consumption exceeding expectations. Further diagnosis is performed on identified issues to identify their root causes, such as equipment aging or inaccurate prediction models. Based on the problem diagnosis results, the original control decisions are adjusted and optimized, such as revising the prediction model, adjusting equipment control parameters, and optimizing power usage time. The adjusted control decisions are integrated into the new control scheme to ensure that the scheme can more accurately reflect system status and environmental changes.

[0159] The updated control plan is then distributed to each controller for execution, and the feedback collection, analysis, and adjustment process is repeated to form a closed-loop control system. Long-term tracking and monitoring of the microgrid system's operating status accumulates data and experience, providing a more accurate basis for future control decisions. With the continuous advancement of technology and the development of power systems, new monitoring technologies, control algorithms, and intelligent methods are being introduced in a timely manner to improve the overall performance and efficiency of the control system.

[0160] like Figure 2As shown, an embodiment of the present invention further provides a DC bus current control system 20 connected to a backup battery, comprising:

[0161] The acquisition module 21 is used to define the topology of the DC microgrid and construct a mathematical model framework including state parameters, operating status, and constraints; filter, interpolate and sparse, time series match, combine calculations, dimension conversion, and normalize the measured parameters obtained from the microgrid to obtain processed data; and analyze the current system state parameters based on the processed data;

[0162] The processing module 22 is used to randomly generate an initial population and iteratively perform mutation, crossover, and selection operations to ultimately determine the final solution, thereby predicting future state changes and obtaining prediction results of the future state; based on the prediction results of the future state and the power flow and energy cost, a control plan is formulated and issued to each controller; and the control plan is continuously optimized according to the execution feedback program to cope with changes in the system state and environment.

[0163] like Figure 3 As shown, the embodiment of the present invention also provides a primary circuit topology including:

[0164] The DC bus, located at the top of the entire topology, serves as the system's central aggregation node and is responsible for providing DC power. It maintains a single voltage state and provides stable DC power to all connected branches and devices.

[0165] Multiple distribution branches extend from the DC bus, distributing DC power to various devices or loads. These branches primarily consist of switches (such as B1, B2, B3, ..., Bn) and relays to ensure safe power distribution and circuit isolation. Branch status parameters primarily focus on branch currents, which can be monitored to understand branch load conditions and operating status.

[0166] The intermediate control layer, located in the middle of the distribution branch, contains power controllers, such as power electronic converters. These controllers monitor and control parameters such as current and voltage in the branch, ensuring stable system operation and efficient energy conversion. Converters are typically two- or multi-port devices. A key parameter is conversion efficiency, which determines energy loss during the power conversion process.

[0167] Branch equipment is connected to the end of the distribution branch, including power generation equipment, energy storage equipment, and load devices (such as C1, C2, C3, ..., Cn). These devices are connected to the system based on actual needs to meet different power and energy storage requirements. Each device has its own specific parameters, including cost parameters.

[0168] like Figure 4 As shown in the figure, an example of the primary circuit topology diagram of a DC microgrid is shown:

[0169] The DC bus is the core of the entire system. As the main power supply, it provides stable DC power to each distribution branch. It represents a single voltage state and serves as the hub for energy aggregation and distribution in the system.

[0170] The distribution branch extends from the DC bus and is used to distribute DC power to different devices or loads. The branch mainly contains air switches (smart circuit breakers can also be used) and contactors to control the on and off of the circuit and protect the safety of the equipment. In addition, the branch also contains a DC / DC converter to ensure voltage conversion and maintain the stability of the DC bus voltage. The energy storage battery is also directly connected to the bus, connected through switches and protective contactors, without an intermediate controller, to provide energy storage for the system and enhance the stability of the power supply.

[0171] The middle control layer is the "brain" of the system, responsible for monitoring and managing the operating status of the entire DC microgrid. Controllers include an MPPT (photovoltaic controller) to maximize the power generation efficiency of PV panels; an AC / DC (AC / DC converter, rectifier) ​​to convert AC power to DC; a DC / DC (bidirectional DC converter, DC energy storage controller) for bidirectional conversion and control of DC power; and a DC / AC (DC / AC converter, inverter) to convert DC power to AC to meet the needs of AC loads. These controllers work together to ensure stable system operation and efficient energy conversion.

[0172] Branch circuit equipment includes various electrical devices and protection devices connected to the distribution branch. These devices are connected to the system based on actual needs to meet different power demands and protection requirements. Electrical devices may include DC loads and AC loads, which are connected to the DC bus and the converted AC power supply, respectively.

[0173] like Figure 5 As shown, the secondary circuit topology includes:

[0174] The virtual busbar acts as a central node, serving as the convergence point for various signal lines. This virtual busbar does not physically exist, but rather serves as a logical concept for signal aggregation and processing, centralizing and managing signals from various branches. It is connected to the smart edge gateway (microgrid controller) via a dotted line, indicating that busbar status information is aggregated and processed by the controller.

[0175] The branch virtual measurement device is a concept proposed to facilitate branch modeling and analysis. It centralizes key branch status parameters to form a unified, standardized state parameter feature vector for each branch. The branch virtual measurement is also connected to the smart edge gateway via a dotted line, indicating that this measurement data is aggregated and processed by the controller.

[0176] Device communication ports are physical ports that connect devices to the gateway, such as RS-232, RS-485, or CAN bus ports. The seven modules (m1 to m7) are each connected to the smart edge gateway via communication ports (n1 to n7). These connections, represented by solid or dashed lines, symbolize the communication and control relationships between them. Through these communication ports, devices can transmit and interact with the smart edge gateway, achieving system coordination and control.

[0177] like Figure 6 Figure 2 shows an example of a microgrid secondary circuit: the smart edge gateway (microgrid controller) serves as the control center of the entire microgrid. It is responsible for coordinating and managing various devices and components in the microgrid, enabling centralized data processing and intelligent control.

[0178] Key equipment and components include:

[0179] Virtual electricity meter: used to monitor and record electricity data in the microgrid, providing basic information for energy management.

[0180] AC / DC converter: realizes the conversion of AC into DC to meet the power requirements of different devices in the microgrid.

[0181] EMU (Microgrid Control Unit): Used to analyze, predict, allocate and manage the electric energy in the microgrid to ensure the stable operation of the system.

[0182] DC / DC: It is a DC potential transformer used to monitor and control DC power.

[0183] Smart switches and contactors: used to control the on and off of circuits in microgrids and protect the safety of equipment and systems.

[0184] Lithium battery packs and smart lithium batteries: As energy storage devices, they provide backup power for microgrids and enhance the system's power supply stability and reliability.

[0185] The virtual data bus serves as a key point for virtual data aggregation, gathering energy and power parameter information from each branch to ensure information flow within the microgrid. The smart edge gateway connects to key devices and components via dotted lines, enabling data transmission and issuing control commands. Devices and components may also be interconnected, forming the microgrid's core control network.

[0186] It should be noted that this system is a system corresponding to the above method, and all implementation methods in the above method embodiment are applicable to this embodiment and can achieve the same technical effects.

[0187] An embodiment of the present invention further provides a computing device comprising: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the above-described method. All implementations in the above-described method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0188] The embodiment of the present invention further provides a computer-readable storage medium storing instructions, which, when executed on a computer, causes the computer to execute the above-described method. All implementations in the above-described method embodiment are applicable to this embodiment and can achieve the same technical effects.

[0189] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A method for controlling the flow of a DC bus connected to a backup battery, characterized in that: The method comprises: Define the topology of the DC microgrid and build a mathematical model framework including state parameters, operating status and constraints; The measurement parameters obtained by the microgrid are filtered, interpolated and sparse, time series matched, combined and calculated, dimensionally transformed, and normalized to obtain processed data; Analyze the current system status parameters based on the processed data; Use genetic algorithms to predict future state changes and obtain prediction results of future states; Based on the predicted results of future states, power flow, and energy costs, a control plan is developed and distributed to each controller; Continuously optimize control solutions based on execution feedback procedures to respond to changes in system state and environment; Define the topology of the DC microgrid and build a mathematical model framework that includes state parameters, operating states, and constraints, including: The DC microgrid is divided into a primary circuit topology and a secondary circuit topology with the DC bus as the center. The primary circuit includes the DC bus, distribution branches, intermediate control layer, and branch equipment for power flow. The secondary circuit includes a virtual converged data bus, branch virtual measurement devices, and equipment communication ports for data collection and transmission. The branch equipment includes power generation, energy storage, and load equipment, and at least one backup battery pack directly connected to the bus. The final DC microgrid topology is obtained. Based on the topological structure of the DC microgrid, a microgrid mathematical model framework is constructed, including state parameters, defined operating states, and defined constraint sets. The state parameters include bus voltage, branch current and direction, branch power, controller efficiency, device connection port voltage and current, inflow and outflow power, inflow and outflow electricity price, energy storage margin and value, and energy storage loop loss rate per unit time. The operating state includes various states of the bus, distribution branch, intermediate controller, and branch equipment. The constraint set includes various operational restrictions of the bus, distribution branch, intermediate controller, and branch equipment. The measurement parameters obtained by the microgrid are filtered, interpolated and sparse, time series matched, combined and calculated, dimensionally transformed, and normalized to obtain processed data, including: Obtain raw measurement parameter data from sensors and monitoring systems in the microgrid; Process the original measurement parameter data to remove noise and outliers to obtain filtered data; Interpolate the missing values ​​in the filtered data and perform sparse processing on the overly dense data points to obtain the interpolated and sparse processed data; Align the interpolated and sparsely processed data to a unified time series to obtain time series matching data; Perform combined calculations based on the matched data of the time series to obtain new derivative data; Convert the measurement data and new derived data of different dimensions into a unified dimension to obtain dimension-transformed data; Normalize the data after dimension transformation to obtain processed data; According to the processed data, analyze the current system status parameters, including; Using a preset genetic algorithm neural network model, the extracted system status-related features are used as input data to identify the current state of the system, including normal operation, overload, undervoltage, and overvoltage; Based on the identified current state of the system, calculate state parameters, including system stability indicators, load levels, and failure probabilities; Through the genetic algorithm, the initial population is randomly generated and mutation, crossover, and selection operations are iteratively performed to finally determine the final solution and predict future state changes. The prediction results of the future state are obtained, including: Randomly generate an initial population, each individual represents a set of state parameter prediction values; For each individual in the population, three different individuals are identified to generate a new variant individual; Cross the new mutant individuals with the original individuals to generate test individuals; Use the fitness function to evaluate each individual in the test individual set and the original individual set, calculate the fitness value of each individual, and determine whether the individual will enter the next generation population based on the size of the fitness value; Repeat the mutation, crossover, and selection operations until the preset number of iterations is reached, and determine the final solution from the final population based on the individual fitness values; Based on the final solution, the future state of the DC microgrid is analyzed, including state change trends and anomaly detection. Based on the final solution and analysis of the future state of the DC microgrid, a prediction result of the future state is generated, including the predicted value, prediction error range, and confidence level key information; The fitness value calculation formula for each individual is: ； Among them, F represents the fitness value; represents the prediction error weight; n represents the number of observations; represents the actual observed value; represents the predicted value; represents the stability weight; Represents the mean of the predicted values.

2. The DC bus current control method connected to a backup battery according to claim 1, characterized in that: Based on the predicted results of future states, power flow, and energy costs, a control plan is developed and distributed to each controller, including: Collect relevant forecast data on future status and obtain current power flow information, including the power supply status of the power grid, the voltage and current of each node; collect energy cost data, including electricity prices in different time periods and energy consumption characteristics of equipment; Construct a prediction model to predict the future state, and use the collected relevant prediction data of the future state, mainly the power load demand forecast, as input data, input it into the prediction model to obtain the prediction result of the future state; Develop control plans based on future state predictions and information on power flow and energy costs, including adjusting equipment operating power, optimizing power usage periods, and starting or shutting down equipment; The developed control scheme is encoded into instructions recognized by the controller, and the control instructions are sent to each controller.

3. A DC bus current control system connected to a backup battery, the system implementing the method according to claim 1 or 2, characterized in that: include: The acquisition module is used to define the topology of the DC microgrid and build a mathematical model framework including state parameters, operating status and constraints; The measurement parameters obtained by the microgrid are filtered, interpolated and sparse, time series matched, combined and calculated, dimensionally transformed, and normalized to obtain processed data; Analyze the current system status parameters based on the processed data; The processing module is used to randomly generate an initial population and iteratively perform mutation, crossover, and selection operations to ultimately determine the final solution, thereby predicting future state changes and obtaining prediction results of the future state; based on the prediction results of the future state and the power flow and energy cost, a control plan is formulated and issued to each controller; and the control plan is continuously optimized according to the execution feedback program to cope with changes in the system state and environment.

4. A computing device, characterized in that include: one or more processors; A storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 3.

5. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a program, which implements the method according to any one of claims 1 to 3 when executed by a processor.

Images