Energy storage charging and discharging control method and system under micro-grid weather risk

Abstract

The application discloses a kind of energy storage charge-discharge control method and system under microgrid meteorological risk, it is related to microgrid energy storage regulation and control technical field, the method includes: the meteorological forecast data calling of preset time window is carried out;Local operating state acquisition is carried out to target microgrid, and real-time operating state data is obtained;New energy power generation output prediction under meteorological risk condition is carried out;The load demand prediction of target microgrid is carried out, and the output load demand prediction value;Energy storage charge-discharge strategy is constructed;Optimize energy storage charge-discharge strategy.The application solves the technical problems that existing technology does not combine meteorological risk prediction to roll optimization energy storage strategy, and extreme weather new energy output fluctuation easily causes microgrid power imbalance, poor operating stability under extreme weather, achieves to realize the dynamic self-adapting charge-discharge regulation and control of microgrid energy storage under extreme weather condition, improves the technical effects of microgrid power supply and demand balance stability and operating reliability under meteorological risk scenario.

Description

Technical Field

[0001] This invention relates to the field of microgrid energy storage regulation technology, specifically to a method and system for controlling the charging and discharging of energy storage under meteorological risks in microgrids. Background Technology

[0002] Wind and solar renewable energy are core power supply units of microgrids, and their output is highly dependent on meteorological conditions. Extreme weather conditions such as strong winds, intense radiation, and sudden temperature drops can cause drastic changes in wind and solar output, directly leading to power imbalance between the microgrid's source and load. Current traditional microgrid energy storage charging and discharging control technologies have multiple shortcomings: Firstly, existing wind and solar output forecasts only use basic meteorological parameters for conversion, without collecting real-time local operating data of the microgrid to construct an operating deviation vector, and lack a multi-dimensional deviation hierarchical correction mechanism. Under extreme weather conditions, the output forecast error increases significantly, and there is no specific extraction of extreme values ​​of meteorological elements for extreme weather model matching and risk quantification assessment, making it impossible to predict weather conditions in advance. On the one hand, there is the power impact caused by disturbances; on the other hand, traditional energy storage charging and discharging strategies are mostly static day-ahead fixed scheduling schemes, which only rely on the single forecast result to generate charging and discharging commands. They do not take the corrected wind and solar power output and load forecast values ​​as dual constraints of power generation and load to jointly build energy storage control strategies. At the same time, there is a lack of a rolling optimization mechanism driven by incremental updates of meteorological forecast data. It is impossible to adaptively adjust the energy storage charging and discharging power with dynamic changes in the weather. Under extreme weather risk scenarios, energy storage control is lagging and the ability to smooth fluctuations is insufficient. It is very easy to have problems such as wind and solar curtailment, short-term power shortage, and voltage and frequency deviation, which significantly reduces the power supply stability of microgrids and the level of new energy consumption.

[0003] Existing technologies have technical problems such as failing to combine meteorological risk prediction with rolling optimization of energy storage strategies, and the drastic fluctuations in the output of new energy sources under extreme weather conditions can easily cause power imbalance and poor operational stability of microgrids. Summary of the Invention

[0004] This application provides a method and system for controlling the charging and discharging of energy storage under meteorological risks in microgrids. It is used to address the technical problems of existing technologies, such as the failure to combine meteorological risk prediction with rolling optimization of energy storage strategies, and the resulting drastic fluctuations in the output of new energy sources under extreme weather conditions, which can easily lead to power imbalance and poor operational stability in microgrids.

[0005] In view of the above problems, this application provides a method and system for controlling the charging and discharging of energy storage under meteorological risks in microgrids.

[0006] The first aspect of this application provides a method for controlling the charging and discharging of energy storage under meteorological risks in microgrids, the method comprising: The system retrieves weather forecast data for a preset time window through the data interface of an external weather forecast service module; it collects local operating status data of the target microgrid to obtain real-time operating status data; using the weather forecast data and real-time operating status data, it predicts the output of new energy power generation under meteorological risk conditions and outputs a corrected predicted output; using the weather forecast data as input variables, it predicts the load demand of the target microgrid and outputs a predicted load demand value; using the corrected predicted output as a power generation-side constraint and the predicted load demand value as a load-side constraint, it constructs an energy storage charging and discharging strategy to drive the target microgrid to perform energy storage charging and discharging operations; based on the incremental updates of the weather forecast data, it continuously optimizes the energy storage charging and discharging strategy to adjust the energy storage charging and discharging operations, thereby achieving dynamic power control of the energy storage system of the target microgrid.

[0007] Among possible implementation methods, the method further includes: predicting the new energy power generation output of the target microgrid based on the meteorological forecast data to obtain an initial predicted output; calculating the operating deviation vector of the real-time operating status data relative to the standard operating status data; correcting the error of the initial predicted output based on the operating deviation vector; and outputting the corrected predicted output.

[0008] Among possible implementation methods, the method further includes: constructing a meteorological element spatiotemporal mapping engine based on the geographic information of the new energy power stations of the target microgrid, performing spatiotemporal reconstruction of the meteorological elements in the meteorological forecast data to obtain a photovoltaic meteorological element sequence and a wind power meteorological element sequence; performing power generation output conversion on the photovoltaic meteorological element sequence and the wind power meteorological element sequence to output the initial photovoltaic power output prediction value and the initial wind power power output prediction value; performing environmental additional correction on the initial photovoltaic power output prediction value and the initial wind power power output prediction value based on the site environmental characteristics of the target microgrid, and outputting the corrected photovoltaic power output and the corrected wind power output; and superimposing the corrected photovoltaic power output and the corrected wind power output based on the temporal synchronization relationship of the photovoltaic meteorological element sequence and the wind power meteorological element sequence to obtain the initial predicted power output.

[0009] Among possible implementation methods, the method further includes: aligning the corrected predicted output and load demand forecast values ​​to perform time-step calculation of the net load sequence; assessing the meteorological risk level based on the meteorological forecast data and outputting the meteorological risk level; retrieving the real-time state of charge from the energy storage system of the target microgrid, and determining the energy storage operation mode by combining the meteorological risk level and the net load sequence to obtain the target operation mode; and coordinating the allocation of charging and discharging power and reserve capacity based on the target operation mode to obtain the charging and discharging power command sequence and reserve capacity reservation command, which constitute the energy storage charging and discharging strategy and are sent to the energy storage converter of the energy storage system for energy storage charging and discharging operations.

[0010] Among possible implementation methods, the following steps are also included: Analyzing the horizontal irradiance sequence, horizontal diffuse irradiance sequence, ambient temperature sequence, and wind speed sequence from the photovoltaic meteorological element sequence; performing a slant projection conversion of the horizontal irradiance sequence based on the installation angle information of the photovoltaic array in the target microgrid, and then combining the horizontal diffuse irradiance sequence with the sky scattering synthesis to obtain the tilted total irradiance sequence; estimating the cell operating temperature based on the tilted total irradiance sequence, ambient temperature sequence, and wind speed sequence to obtain the cell temperature sequence; inputting the tilted total irradiance sequence and cell temperature sequence into the photovoltaic module equivalent circuit model to calculate the DC power of the photovoltaic module, obtaining the module-level maximum power sequence; and, based on the string structure of the photovoltaic array, performing a collector-collector accumulation of the module-level maximum power sequence, and then combining it with the inverter configuration of the photovoltaic array to perform inverter efficiency conversion, outputting the predicted initial photovoltaic output value.

[0011] Among possible implementation methods, the method further includes: using the hub height of the wind turbines in the target microgrid to perform meteorological element profile transformation of the wind power meteorological element sequence, obtaining hub height wind speed, hub height wind direction, hub height temperature, and hub height air pressure; using the hub height temperature and hub height air pressure to calculate the actual air density sequence; based on the overall power characteristics of the wind turbines, calculating the field wake attenuation coefficient sequence for the hub height wind direction, and reducing the wake wind speed of the hub height wind speed to obtain the wake correction wind speed sequence; performing power curve density correction calculation based on the wake correction wind speed sequence and the actual air density sequence, outputting the single-unit initial output prediction value, and then summarizing the overall output of the wind turbines to output the initial wind power output prediction value.

[0012] Among possible implementation methods, the method further includes: inputting the operating deviation vector into multiple single-dimensional deviation component mapping channels in the deviation-output error mapping model, performing deviation pattern matching and determination to obtain multivariate error correction parameters; using the correction component priority of the multivariate error correction parameters to perform step-by-step composite correction of the initial predicted output, and outputting the corrected predicted output.

[0013] In some possible implementations, the method further includes: calculating multiple operational parameter deviation components of the real-time operational status data relative to the standard operational status data; using a preset deviation judgment threshold, filtering out P valid deviation components from the multiple operational parameter deviation components, and then performing vectorized concatenation to obtain the operational deviation vector.

[0014] Among the possible implementation methods, it also includes: extracting extreme values ​​of meteorological elements from the meteorological forecast data, performing extreme weather model matching, and quantitatively outputting the meteorological risk level.

[0015] A second aspect of this application provides an energy storage charging and discharging control system for microgrids under meteorological risks, the system comprising: The system comprises the following modules: a data retrieval module for retrieving weather forecast data for a preset time window via the data interface of an external weather forecast service module; a real-time operating status data acquisition module for collecting local operating status data of the target microgrid; a corrected forecast output module for predicting the output of renewable energy generation under meteorological risk conditions using the weather forecast data and real-time operating status data, and outputting a corrected forecast output; a load demand forecast output module for predicting the load demand of the target microgrid using the weather forecast data as input variables, and outputting a load demand forecast value; a charge / discharge execution module for constructing an energy storage charge / discharge strategy using the corrected forecast output as a generation-side constraint and the load demand forecast value as a load-side constraint, and driving the target microgrid to perform energy storage charge / discharge operations; and a dynamic power control module for continuously optimizing the energy storage charge / discharge strategy based on incremental updates of the weather forecast data, adjusting the energy storage charge / discharge operations to achieve dynamic power control of the energy storage system of the target microgrid.

[0016] One or more technical solutions provided in this application have at least the following technical effects or advantages: The system retrieves weather forecast data for a preset time window through the data interface of an external weather forecast service module; it collects local operating status data of the target microgrid to obtain real-time operating status data; using the weather forecast data and real-time operating status data, it predicts the output of new energy power generation under meteorological risk conditions and outputs a corrected predicted output; using the weather forecast data as input variables, it predicts the load demand of the target microgrid and outputs the predicted load demand value; it constructs an energy storage charging and discharging strategy to drive the target microgrid to perform energy storage charging and discharging operations; and it continuously optimizes the energy storage charging and discharging strategy to adjust the energy storage charging and discharging operations, thereby achieving dynamic power control of the energy storage system. This achieves the technical effect of realizing dynamic adaptive charging and discharging regulation of microgrid energy storage under extreme weather conditions, improving the stability and operational reliability of microgrid power supply and demand balance under meteorological risk scenarios. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A schematic flowchart of an energy storage charging and discharging control method under meteorological risks in a microgrid, provided in an embodiment of this application; Figure 2 This is a schematic diagram of a microgrid energy storage charging and discharging control system under meteorological risks, provided as an embodiment of this application.

[0019] Explanation of reference numerals in the attached diagram: Data retrieval module 10, real-time operating status data acquisition module 20, corrected predicted output module 30, load demand predicted value output module 40, charging and discharging execution module 50, dynamic power control module 60. Detailed Implementation

[0020] This application provides a method and system for controlling the charging and discharging of energy storage under meteorological risks in microgrids. It addresses the technical problems of existing technologies that fail to incorporate meteorological risk prediction and rolling optimization of energy storage strategies, leading to severe fluctuations in renewable energy output under extreme weather conditions, which can cause power imbalances and poor operational stability in microgrids.

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0022] Example 1, as Figure 1 As shown, this application provides a method for controlling the charging and discharging of energy storage under meteorological risks in microgrids, the method comprising: Step S100: Call the weather forecast data for the preset time window through the data interface of the external weather forecast service module.

[0023] Specifically, the system uses the data interface of an external weather forecast service module to retrieve weather forecast data for a preset time window. This involves building a bidirectional interactive data interface to connect with the external weather forecast service module. The system initiates timed data retrieval requests according to a pre-configured preset time window, and batches up conventional meteorological time-series data including irradiance, ambient temperature, wind speed, wind direction, and air pressure within that time window. Simultaneously, it collects risk-related forecast parameters such as extreme weather indicators (typhoons, blizzards, continuous rain) and extreme values ​​of meteorological elements. The system then compiles a standardized time-series weather forecast dataset, which serves as the basic input data source for subsequent new energy output forecasting, load demand forecasting, and meteorological risk level assessment.

[0024] Step S200: Collect local operating status data of the target microgrid to obtain real-time operating status data.

[0025] Specifically, relying on the photovoltaic inverter acquisition units, wind turbine status monitoring terminals, energy storage BMS acquisition devices, bus power sensors, and load metering equipment of each branch deployed within the microgrid, the operating parameters of the power station are collected in real time. These parameters include the current actual power generation output of photovoltaic and wind power, real-time state of charge of energy storage, AC and DC bus voltage and current, real-time power of various loads, equipment operating temperature, and inherent loss coefficient of the power station, among other multi-dimensional parameters. The collected raw data is then processed by noise reduction, alignment, and standardization to form complete and standardized real-time operating status data.

[0026] Step S300: Using the meteorological forecast data and real-time operating status data, predict the output of new energy power generation under meteorological risk conditions, and output the corrected predicted output.

[0027] Specifically, a spatiotemporal mapping engine for meteorological elements is first built based on the geographic information of the target microgrid renewable energy power stations. The retrieved meteorological forecast data is then used for spatiotemporal reconstruction of meteorological elements, generating separate sequences for photovoltaic (PV) and wind power (WHP) meteorological elements. For the PV side, sequences of horizontal irradiance, diffuse irradiance, temperature, and wind speed are extracted. Combined with the PV installation angle, slope irradiance conversion and cell temperature estimation are performed, and then input into the PV equivalent circuit model to calculate the module-level power. The initial PV output prediction value is then obtained through string bus and inverter efficiency conversion. For the WHP side, meteorological profile transformation, air density calculation, and wake wind speed reduction are performed based on the wind turbine hub height, combined with the wind turbine power curve. After obtaining the output of a single unit, the initial predicted output value of wind power is superimposed across the entire field. The output of photovoltaic and wind power is integrated, and environmental additional corrections are made based on the site environmental characteristics to obtain the overall initial predicted output. Then, the collected real-time operating status data is compared with the preset standard operating status data to calculate the multi-dimensional operating parameter deviation components. The effective deviation components are selected and spliced ​​to form an operating deviation vector. This operating deviation vector is sent to the deviation-output error mapping model to complete the deviation pattern matching to obtain multi-dimensional error correction parameters. The initial predicted output is then subjected to step-by-step composite error correction according to the priority of the correction components. Finally, the corrected predicted output is output to adapt to the current meteorological risk conditions and the actual operating conditions of the microgrid.

[0028] Step S400: Using the meteorological forecast data as input variables, perform load demand forecasting for the target microgrid and output the load demand forecast value.

[0029] Specifically, a multi-input temporal LSTM long short-term memory neural network is adopted as the core algorithm for load forecasting. The algorithm consists of four layers: an input layer, a feature fusion layer, an LSTM temporal memory layer, and a fully connected output layer. During the training phase, historical meteorological time-series samples from the past three consecutive months are collected. These samples include irradiance, temperature, wind speed, extreme weather labels, and the actual power consumption of industrial, residential, and critical loads in the microgrid during the corresponding time periods, forming a labeled training dataset. This dataset is then divided into training, validation, and test sets in a 7:2:1 ratio. Meteorological features are normalized and standardized, and load data undergoes temporal difference smoothing preprocessing. The multi-dimensional meteorological forecast features are concatenated into a multi-dimensional input vector and fed into the input layer. The feature fusion layer completes the cross-coupling of meteorological features through a fully connected weight matrix, capturing... To capture the nonlinear impact of extreme weather events such as high temperatures, blizzards, and continuous rain on electricity load, the LSTM time-series memory layer relies on cell states and gating structures to remember the day-ahead and intraday load time-series changes, solving the problem of gradient vanishing in long-term forecasts. During training, the Adam optimizer is used to minimize the mean squared error loss function between the predicted load and the actual load, and the optimal model weights are saved after the validation set loss converges. In the online forecast stage, the real-time retrieved time-series weather forecast data is input into the trained LSTM model after undergoing standardized preprocessing consistent with the training stage. The model outputs the predicted power of various loads in different time periods step by step, and the summaries are used to obtain the overall load demand forecast value of the target microgrid, which serves as the load-side constraint basis for the construction of energy storage regulation strategies.

[0030] Step S500: Using the modified predicted output as the generation-side constraint and the predicted load demand as the load-side constraint, construct an energy storage charging and discharging strategy to drive the target microgrid to perform energy storage charging and discharging operations.

[0031] Specifically, the modified forecast output and load demand forecast values ​​for the same time interval are first aligned by time step, and the net load sequence of the microgrid is calculated for each time period. Then, extreme values ​​of meteorological elements at each time step are extracted from meteorological forecast data to carry out extreme weather model matching, and the meteorological risk levels are quantitatively divided into high, medium and low levels. The real-time charge status of the energy storage system is retrieved simultaneously. The target operation mode of energy storage is determined by combining the meteorological risk level and the net load sequence. Under the high-risk mode, a large capacity of backup power is reserved to ensure uninterrupted power supply to critical loads. Under the medium and low-risk modes, peak shaving and valley filling of new energy sources and power fluctuation smoothing are taken into account simultaneously. Then, based on the target operation mode determined, the energy storage charging and discharging power and the reserve capacity are allocated in a coordinated manner for each time period, and a complete energy storage charging and discharging strategy including the time-step charging and discharging power command sequence and the reserve capacity reserve command is generated. The strategy is sent to the energy storage converter of the energy storage system, and the energy storage converter receives the command and executes the corresponding charging, discharging or standby backup energy storage charging and discharging operations.

[0032] Step S600: Based on the incremental update of the meteorological forecast data, the energy storage charging and discharging strategy is continuously optimized to adjust the energy storage charging and discharging operation, so as to realize dynamic power control of the energy storage system of the target microgrid.

[0033] Specifically, the system continuously monitors incremental updates of meteorological data pushed by external weather forecasting service modules. Once meteorological elements such as irradiance, wind speed, temperature, and extreme weather warning levels change, the system re-executes the new energy power output forecast, microgrid load demand forecast, and meteorological risk level assessment processes based on the updated weather forecast data. Simultaneously, the system refreshes the net load sequence and re-determines the target energy storage operation mode. It then re-solves and optimizes the energy storage charging and discharging power allocation and reserve capacity reservation scheme, generating a new energy storage charging and discharging strategy adapted to the latest meteorological conditions. The updated power adjustment command is then sent to the energy storage converter in real time to dynamically adjust the energy storage charging and discharging power, charging and discharging duration, and reserve capacity reservation ratio. This forms a closed-loop rolling control logic of meteorological data update - full-process re-forecast - strategy iteration - real-time energy storage power adjustment, continuously completing the all-time adaptive dynamic power control of the target microgrid energy storage system.

[0034] In one possible implementation, step S300 further includes: Step S310: Based on the meteorological forecast data, predict the new energy power generation output of the target microgrid to obtain the initial predicted output.

[0035] Step S320: Calculate the operating deviation vector of the real-time operating status data relative to the standard operating status data.

[0036] Step S330: Based on the running deviation vector, perform error correction on the initial predicted output and output the corrected predicted output.

[0037] Specifically, to retrieve complete geographic information such as latitude and longitude, altitude, photovoltaic array tilt angle and azimuth, and wind turbine hub height corresponding to the target microgrid renewable energy power station, a spatiotemporal mapping engine for meteorological elements is built. Based on this engine, the input discrete time-series meteorological forecast data is interpolated and spatiotemporally reconstructed to generate photovoltaic meteorological element sequences and wind power meteorological element sequences with completely unified time steps. Based on the two types of meteorological sequences, power generation output conversion is carried out according to type. Combining the equivalent circuit model of photovoltaic modules and the power curve of wind turbines, the initial power output prediction values ​​of photovoltaic and wind power are calculated and output sequentially. Then, environmental correction coefficients are introduced to make environmental additional corrections to the initial power output of photovoltaic and wind power based on the site environmental characteristics such as dust obstruction, altitude attenuation, and surrounding shadows, so as to obtain the corrected power output of photovoltaic and wind power. Finally, based on the one-to-one time-series synchronization relationship between the photovoltaic meteorological element sequences and the wind power meteorological element sequences, the corrected power output values ​​of photovoltaic and wind power at the same time step are superimposed and summed. After integrating the total power output data of wind and solar, the overall initial predicted power output of renewable energy in the target microgrid is obtained.

[0038] First, real-time operating status data of microgrid wind and solar turbines, converters, and meteorological monitoring equipment are collected simultaneously with standard operating status data under matching conditions. Then, the difference calculations are performed on parameters such as irradiance deviation, wind speed deviation, component temperature deviation, inverter loss deviation, wind turbine wake loss deviation, and equipment aging attenuation deviation to obtain the deviation components corresponding to each operating parameter. Preset independent deviation judgment thresholds for each parameter are retrieved, and the absolute value of each deviation component is compared with the corresponding threshold. Only P-term valid deviation components with absolute values ​​exceeding the threshold and exhibiting significant prediction disturbances are retained, while invalid weak deviations caused by small fluctuations are eliminated. Finally, the selected P-term valid deviation components are vectorized and concatenated sequentially according to a fixed dimension to generate an operating deviation vector with uniform dimensions that contains only significant disturbance information, providing standardized input features for the subsequent error mapping correction model.

[0039] The operational deviation vector, which includes multi-dimensional operational deviation indicators, is input into a pre-trained deviation-output error mapping model. Multiple single-dimensional deviation component mapping channels corresponding to various deviation indicators are set up in parallel within the model. Each channel independently extracts deviation features for its corresponding dimension and performs deviation pattern matching with a historical deviation sample library. The matching results from all channels are then fused to obtain a multivariate error correction parameter containing multi-dimensional compensation coefficients. Following a preset priority order for correction components of meteorological deviation, equipment operating condition deviation, and site loss deviation, the corresponding correction coefficients within the multivariate error correction parameter are sequentially retrieved to perform a step-by-step composite correction operation of superposition and scaling on the initial predicted output. The output value after the correction of the previous dimension is used as the input for the correction of the next dimension. After all components are corrected step-by-step, the corrected predicted output, which eliminates multi-source prediction errors and conforms to the actual power generation conditions on site, is output.

[0040] In one possible implementation, step S310 further includes: Step S311: Based on the geographic information of the new energy power stations of the target microgrid, construct a meteorological element spatiotemporal mapping engine, perform spatiotemporal reconstruction of the meteorological elements of the meteorological forecast data, and obtain photovoltaic meteorological element sequences and wind power meteorological element sequences.

[0041] Step S312: Perform power generation output conversion on the photovoltaic meteorological element sequence and the wind power meteorological element sequence, and output the initial power output prediction value of photovoltaic power and the initial power output prediction value of wind power.

[0042] Step S313: Based on the site environment characteristics of the target microgrid, perform environmental additional corrections on the initial photovoltaic power output prediction value and the initial wind power output prediction value, and output the corrected photovoltaic power output and the corrected wind power output.

[0043] Step S314: Based on the temporal synchronization relationship between the photovoltaic meteorological element sequence and the wind power meteorological element sequence, the photovoltaic corrected output and the wind power corrected output are superimposed to obtain the initial predicted output.

[0044] Specifically, the process begins by collecting complete geographic information of the target microgrid renewable energy power station, including the station's latitude and longitude, altitude, tilt / azimuth angle of each photovoltaic array, planar coordinates of each wind turbine location, design height of the wind turbine hub, and elevation data of surrounding terrain shading. Based on this geographic dataset, a meteorological element spatiotemporal mapping engine with built-in Kriging spatial interpolation and linear temporal interpolation algorithms is built. The engine first connects to external gridded weather forecast raw data, and uses Kriging interpolation to reconstruct the spatial dimensions based on the precise coordinates of the power station. This interpolates and converts large-scale gridded meteorological values ​​into site-specific irradiance, temperature, wind speed, air pressure, and wind direction parameters, resolving the issue of the relationship between meteorological grids and weather forecasts. Spatial deviations due to mismatched site locations; then, inconsistent sampling timestamps from multi-source weather forecasts are read, and missing time data are filled in through linear time-series interpolation and standardized to a fixed time step of 15 minutes, completing the temporal dimension reconstruction; subsequently, the reconstructed meteorological dataset is split according to the parameter requirements for photovoltaic and wind power output calculations, and horizontal irradiance, diffuse irradiance, near-surface temperature, and near-surface wind speed are extracted and organized into photovoltaic meteorological element sequences, and hub height is extracted to convert wind speed, wind direction, and air temperature and pressure parameters and organized into wind power meteorological element sequences. The timestamps of the two types of sequences are completely aligned and the temporal sequence is continuous, which can be directly used for subsequent step-by-step calculations of photovoltaic and wind power output.

[0045] The horizontal irradiance sequence, horizontal diffuse irradiance sequence, ambient temperature sequence, and wind speed sequence are extracted from the photovoltaic meteorological element sequence. The horizontal direct irradiance is converted to a projected horizontal irradiance using the photovoltaic array's installation tilt angle and azimuth angle. Combined with the horizontal diffuse irradiance sequence, the total tilted irradiance sequence is synthesized using a sky scattering model. Simultaneously, the cell temperature sequence is estimated hourly using the total tilted irradiance, ambient temperature, and wind speed in a thermal balance model. The total tilted irradiance sequence and cell temperature sequence are input into a single-diode photovoltaic equivalent circuit model. A full-domain voltage and current scan is used to solve for the maximum power sequence at each module level. Then, based on the series-parallel string structure of the photovoltaic array, the maximum power of each module is accumulated stage by stage to obtain the total DC power of the array. Finally, the inverter power is calculated using the inverter efficiency curve. The process involves rate conversion, ultimately outputting a time-series continuous prediction of the initial photovoltaic power output. Next, wind power output conversion is performed: the hub height parameters of the wind turbines are read to perform meteorological profile transformation on the wind power meteorological element sequence, generating time-series data of hub height wind speed, wind direction, temperature, and air pressure. The hub height temperature and air pressure are substituted into the ideal gas law to calculate the real-time air density sequence for the entire field. Combined with the overall wind turbine layout and hub height wind direction, the corresponding wake attenuation coefficient sequence is solved. The attenuation coefficient is used to reduce the original wind speed at hub height to obtain the wake correction wind speed sequence. Based on the original power characteristic curve of the wind turbine and real-time air density, density correction is performed, and the instantaneous output of a single wind turbine is obtained by looking up a table. The output of all wind turbines in the field is then summed up hourly to output a complete time-series prediction of the initial wind power output.

[0046] First, the site environmental characteristic parameters obtained from long-term monitoring of the target microgrid plant area are collected, including the proportion of photovoltaic array dust accumulation attenuation, the proportion of shadow shading by surrounding buildings and vegetation, the altitude irradiance attenuation coefficient, the high temperature power reduction coefficient, the terrain obstruction coefficient of the wind turbine area, the near-surface airflow turbulence loss coefficient, and the additional surface roughness coefficient. The photovoltaic comprehensive environmental correction coefficient and the wind power comprehensive environmental correction coefficient are calculated separately. The photovoltaic initial output prediction value at each time node is multiplied by the photovoltaic comprehensive environmental correction coefficient to offset the power loss caused by dust accumulation, shadow, altitude, and high temperature to obtain the photovoltaic corrected output. At the same time, the wind power initial output prediction value at the corresponding time is multiplied by the wind power comprehensive environmental correction coefficient to compensate for the wind energy capture loss caused by terrain obstruction, airflow turbulence, and surface roughness to obtain the wind power corrected output. After completing the independent environmental compensation calculation of wind and solar output at each time step, the photovoltaic corrected output and wind power corrected output that are time-aligned and closely match the actual environmental loss level of the plant area are output.

[0047] Using the timestamps of the photovoltaic meteorological element sequence and the wind power meteorological element sequence as the time series benchmark, the sampling interval, time series nodes, and prediction duration of the two types of meteorological sequences are completely one-to-one correspondences, eliminating time series misalignment deviations. Under each same time series step, the photovoltaic power output value after environmental additional correction and the wind power output value are precisely linearly superimposed and summed at each moment. The wind and solar power output data of the whole time period are accumulated and merged point by point according to the time series order, and finally a total new energy initial prediction output that is time-continuous, dimension-unified, and synchronously matched with the meteorological time series change law of the microgrid is formed.

[0048] In one possible implementation, step S500 further includes: Step S510: Align the corrected predicted output and load demand prediction values, and perform time-step calculation of the net load sequence.

[0049] Step S520: Based on the meteorological forecast data, conduct a meteorological risk level assessment and output the meteorological risk level.

[0050] Step S530: Retrieve the real-time state of charge from the energy storage system of the target microgrid, and determine the energy storage operation mode by combining the meteorological risk level and net load sequence to obtain the target operation mode.

[0051] Step S540: Based on the target operating mode, perform coordinated allocation of charging and discharging power and reserve capacity to obtain charging and discharging power instruction sequence and reserve capacity reservation instruction, which constitute the energy storage charging and discharging strategy and send it to the energy storage converter of the energy storage system for energy storage charging and discharging operation.

[0052] Specifically, the process begins by extracting the corrected predicted power output time series array and the predicted load demand time series array. The start and end timestamps and single-step durations of both sets of data are then uniformly verified. Missing or misaligned time series nodes are corrected using linear interpolation, ensuring one-to-one alignment of all time steps across the entire time period. After alignment, the net load calculation formula is applied to each synchronized time step. In the formula For the current time step net load, Forecast value of load demand at this moment To correct the predicted output at that moment, if the calculated output is... A value greater than 0 indicates that local power generation is insufficient to cover the load, requiring energy storage to discharge and make up the power gap. A value less than 0 indicates that there is excess output from new energy sources. Excess energy can be absorbed and recharged using energy storage. The above difference calculation is repeated for each time step. The calculation results of all time steps are arranged in chronological order to generate a complete and continuous net load sequence, which serves as the basic input data for subsequent energy storage operation mode determination.

[0053] The system reads complete time-series meteorological forecast data and extracts the maximum and minimum values ​​of key meteorological elements such as wind speed, ambient temperature, irradiance, air pressure, and precipitation intensity within the forecast period as meteorological element extremes. It retrieves pre-stored thresholds and standard feature models for various extreme weather events and matches the extracted multi-dimensional meteorological extremes with extreme weather model intervals such as high temperature, strong wind, extreme irradiance, and heavy rainfall. It statistically analyzes the types of extreme weather events, the deviation of extreme values ​​from the normal interval, and the duration of extreme weather. It calculates the comprehensive risk score using a preset quantitative scoring formula and classifies the corresponding meteorological risk level based on the interval to which the score belongs. Finally, it outputs the meteorological risk level result with a quantitative score.

[0054] The current State of Charge (SOC) value of the energy storage battery pack is read in real time through communication interaction. This real-time SOC, hourly net load sequence, and all-time meteorological risk level are simultaneously input into a preset multi-dimensional mode judgment logic library. When the meteorological risk level is high, the priority is to ensure stable power supply. The emergency supply mode is determined by combining the positive and negative trends of net load and the remaining SOC, and sufficient energy storage capacity is reserved to cope with drastic fluctuations in wind and solar power output. When the meteorological risk level is medium, the balance between renewable energy consumption and stable load supply is taken into account. The mode of smoothing fluctuations or storing surplus energy is switched according to the net load gap / surplus and SOC high and low. When the meteorological risk level is low, the goal is to maximize the consumption of renewable energy and reduce grid interaction power. The conventional charge and discharge arbitrage mode is determined based on the time sequence changes of net load. After completing the all-time hourly logic judgment by comprehensively considering the three types of parameters, the target operation mode of energy storage corresponding to each time sequence node is output.

[0055] Based on the target operating mode of energy storage corresponding to each time step, preset capacity constraints and upper and lower power safety boundaries are matched. Under the high-risk supply guarantee mode, a fixed safe reserve capacity threshold is first defined, and then the corresponding discharge compensation power or charging absorption power is allocated based on the net load value. Under the medium-risk mitigation mode, the reserve capacity size and charging and discharging power amplitude are dynamically adjusted in a way that takes into account both fluctuation suppression and capacity reserve requirements. Under the low-risk optimized absorption mode, redundant reserve capacity is reduced to the maximum extent, and the optimal charging and discharging power is matched based on the output surplus / load gap. The coordinated optimization allocation of charging and discharging power and reserved reserve capacity for each time period is completed sequentially. The power values ​​of all time steps are combined in an orderly manner to form a charging and discharging power command sequence, and the reserve capacity reservation command corresponding to each time period is generated simultaneously. The two types of commands are integrated to form a complete energy storage charging and discharging strategy. The entire strategy is sent to the energy storage converter inside the energy storage system through the communication bus. The energy storage converter analyzes the received power and capacity commands in real time and performs charging and discharging operations of energy storage charging, discharging, or capacity static retention accordingly.

[0056] In one possible implementation, step S310 further includes: The horizontal irradiance sequence, horizontal diffuse irradiance sequence, ambient temperature sequence, and wind speed sequence were analyzed from the photovoltaic meteorological element sequence.

[0057] Based on the installation angle information of the photovoltaic array in the target microgrid, the horizontal irradiance sequence is converted by oblique projection, and then combined with the horizontal scattering irradiance sequence to perform sky scattering synthesis to obtain the total irradiance sequence of the oblique surface.

[0058] Based on the total irradiance sequence of the inclined surface, the ambient temperature sequence, and the wind speed sequence, the operating temperature of the solar cell is estimated to obtain the solar cell temperature sequence.

[0059] The total irradiance sequence of the tilted surface and the cell temperature sequence are input into the equivalent circuit model of the photovoltaic module to calculate the DC power of the photovoltaic module and obtain the module-level maximum power sequence.

[0060] Based on the string structure of the photovoltaic array, after accumulating the maximum power sequence at the module level, the inverter efficiency is calculated in conjunction with the inverter configuration of the photovoltaic array, and the predicted initial output value of the photovoltaic is output.

[0061] Specifically, the system reads the structured storage of photovoltaic meteorological element sequences, extracts four independent time-series data by splitting them step-by-step according to preset data field identifiers, extracts all time node values ​​by matching the field identified as total horizontal irradiance, arranges them in chronological order to generate a continuous horizontal irradiance sequence, extracts corresponding time-series values ​​by matching the diffuse irradiance-specific field to form a horizontal diffuse irradiance sequence, retrieves the time-by-time monitoring values ​​of the corresponding field of ambient temperature and organizes them into an ambient temperature sequence, and reads the time-step data under the near-surface wind speed field and combines them in an orderly manner to obtain a wind speed sequence. The timestamps of the four parsed sequences are completely aligned and the sampling step size is uniform, which can be directly used for iterative calculations of parameters such as photovoltaic tilt surface irradiance and battery operating temperature.

[0062] Adjust the installation tilt angle of the photovoltaic array Array azimuth angle Combined with real-time solar altitude angle Sun azimuth Constructing a direct radiation projection conversion formula In the formula For single-time-step horizontal direct irradiance, For the direct irradiance of the inclined surface, Given the incident angle between the solar incident ray and the normal to the photovoltaic panel, the hourly stepwise direct component of the horizontal plane irradiance sequence is converted by oblique projection; then, the scattered components are converted using an isotropic sky scattering model, and the calculation formula is as follows: In the formula For the horizontal irradiance at the synchronization time, To calculate the irradiance scattered by the inclined surface, the direct irradiance and the scattered irradiance of the inclined surface, converted at the same time stamp, are added together using the following formula: , Representing the total irradiance of the tilted surface, projection conversion, scattering reduction, and irradiance synthesis operations are sequentially performed on all time-series nodes of the horizontal surface irradiance sequence and the horizontal surface scattered irradiance sequence. Finally, a time-synchronized and time-continuous total irradiance sequence of the tilted surface is generated, providing core irradiance input parameters for subsequent calculations of photovoltaic cell operating temperature and power generation.

[0063] The cell temperature is calculated hourly using a photovoltaic module thermal balance estimation model. The estimation formula is as follows: In the formula The operating temperature of the solar cells. This represents the time-time values ​​corresponding to the ambient temperature sequence. The total irradiance sequence of the inclined surface is the irradiance value at the synchronous moment. The inherent equivalent thermal resistance of photovoltaic modules, The air convection heat transfer coefficient varies with wind speed v, which is taken from the time-series data corresponding to the wind speed sequence. The convection heat transfer coefficient increases linearly with increasing wind speed. The higher the wind speed, the stronger the heat dissipation effect and the lower the battery temperature rise. The total irradiance of the inclined surface, ambient temperature, and wind speed at the same time stamp are read synchronously and substituted into the formula to solve the instantaneous battery temperature. All calculation results are arranged in chronological order to generate a battery temperature sequence that is time-continuous and perfectly matches the meteorological sequence, providing temperature parameters for subsequent iterative solutions of photovoltaic module output power.

[0064] A time-step iterative solution is performed using a single-diode photovoltaic equivalent circuit model. The core circuit equation is: In the formula: I is the component output current, and V is the component terminal voltage. Total irradiance of the inclined surface Cell temperature The photocurrent is jointly determined by the irradiance; the higher the irradiance, the greater the photocurrent, and the higher the temperature, the slightly inhibits the photocurrent. The reverse saturation leakage current of the temperature-dependent diode increases significantly with the rise of the cell temperature. For the series resistance of the component, Here, n is the bypass parallel resistor, and n is the diode ideality factor. The thermal voltage is related to the cell temperature. For each synchronization time step, the total irradiance of the tilted surface and the cell temperature at that moment are read, substituted into the model, and the voltage value range is traversed to scan the corresponding output current. The DC output power under each voltage-current pairing is calculated. The maximum power value at the specified time point is selected as the instantaneous maximum power of the component. The maximum power values ​​at all times are summarized in chronological order to generate a continuous and complete component-level maximum power sequence with the same step size as the time-series meteorological data.

[0065] First, read the series and parallel arrangement parameters of the photovoltaic array in the field to clarify the number of modules in each string and the parallel bus topology of each string. Then, sum the maximum power sequence values ​​of all modules at the same time node according to the string structure to obtain the total DC power time sequence data of the array. Next, retrieve the rated capacity and segmented inverter efficiency curves of the corresponding inverters, match the inverter efficiency coefficient under the corresponding operating condition based on the current DC input power, and use the conversion formula. Calculate the AC output, where This represents the total DC power of the array after the current is accumulated. This represents the inverter conversion efficiency corresponding to the current power range. This is the initial photovoltaic power output prediction value at a single moment. The DC current accumulation and inverter efficiency conversion are completed step by step to output the complete time-series photovoltaic initial power output prediction value.

[0066] In one possible implementation, step S310 further includes: The wind power meteorological element sequence profile is transformed using the hub height of the wind turbine in the target microgrid to obtain the hub height wind speed, hub height wind direction, hub height temperature and hub height air pressure.

[0067] The actual air density sequence is calculated using the air temperature and air pressure at the wheel hub height.

[0068] Based on the overall power characteristics of the wind turbine, the wake attenuation coefficient sequence at the hub height and wind direction is calculated, and the wake wind speed at the hub height is reduced to obtain the wake corrected wind speed sequence.

[0069] Based on the wake-corrected wind speed sequence and the actual air density sequence, the power curve density correction calculation is performed, and the initial output prediction value of a single unit is output. Then, the total output of the wind turbine is summed and aggregated, and the initial output prediction value of the wind power is output.

[0070] Specifically, the original wind speed, wind direction, temperature, and air pressure time-series data at the reference wind measurement height are extracted from the wind power meteorological element sequence. Using the actual hub installation height of the wind turbine as the target height, the wind speed-to-height conversion is completed using the wind speed power-law profile formula. Simultaneously, the temperature and air pressure are corrected for vertical height based on the atmospheric vertical temperature and pressure gradient model. The wind direction does not change with the height gradient and directly uses the original time-series values. The height profile transformation calculation of all meteorological parameters is completed step by step, generating four independent time-series sequences with the same step size and timestamp as the original time-series data: hub height wind speed, hub height wind direction, hub height temperature, and hub height air pressure. This provides accurate meteorological parameters with height matching the wind turbine impeller plane for subsequent air density calculation and wake wind speed correction.

[0071] Extracting the wheel hub height and air pressure at synchronous timing Hub height thermodynamic temperature It is derived from Celsius temperature and is based on the ideal gas law. Calculate air density at each time step In the formula, M represents the molar mass of dry air and R is a universal gas constant. First, the temperature at the hub height in Celsius is converted to Kelvin thermodynamic temperature. Then, the hub height air pressure at the corresponding time is substituted into the synchronous calculation. The air density calculated at each time node is arranged in chronological order to generate an actual air density sequence that matches the step size of the time-series meteorological data and corresponds one-to-one with the timestamp.

[0072] The coordinates of all wind turbines, rotor diameter, turbine spacing, and parameters of the standard wake attenuation model are pre-entered. For each time step, based on the hub height and wind direction, the shading coverage of upstream turbines on downstream turbines is determined, and the wake attenuation coefficients of each turbine under the corresponding wind direction are calculated using the wake loss formula. The wake attenuation coefficients are arranged in chronological order to form a continuous sequence; then the reduction formula is applied. Hourly step-by-hour wheel hub height and original wind speed Perform wake loss calculation. To determine the effective wake correction wind speed after deducting the airflow obstruction loss from the upstream wind turbine, a wake correction wind speed sequence with consistent meteorological time series step size and corresponding timestamps is generated after all time series nodes have completed wind speed reduction calculations.

[0073] Retrieve the baseline power curve of the fan under the standard rated air density, read the wake correction velocity and actual air density at the same time step, and apply the density correction formula. Complete power correction, where To correct the wind speed for the wake, This represents the baseline output at the corresponding wind speed for the baseline density. For real-time air density, Using the standard rated air density, the initial output prediction value of a single wind turbine is obtained by solving time-by-time. All wind turbines in the field are traversed, and the output values ​​of all single turbines at the same time node are linearly accumulated and summarized. The total output data of all time periods are sorted out in time sequence to generate the initial output prediction value of wind power that is completely matched with the time sequence meteorological parameter step size and corresponds one-to-one with the timestamp.

[0074] In one possible implementation, step S330 further includes: Step S331: Input the running deviation vector into the multiple one-dimensional deviation component mapping channels of the deviation-output error mapping model, perform deviation pattern matching and determination, and obtain multivariate error correction parameters.

[0075] Step S332: Using the priority of the correction components of the multivariate error correction parameters, perform stepwise composite correction of the initial predicted output, and output the corrected predicted output.

[0076] Specifically, the operational deviation vector is input into multiple single-dimensional deviation component mapping channels in the deviation-output error mapping model, and deviation pattern matching is performed to obtain multivariate error correction parameters. The deviation-output error mapping model is pre-constructed offline based on historical microgrid operation datasets. The construction process involves first splitting the historical time period's wind and solar output prediction deviation, irradiance deviation, wind turbine wind speed profile deviation, energy storage operating condition loss deviation, inverter converter loss deviation, and site environment attenuation deviation into dimensions, and dividing them into independent single-dimensional deviation component mapping channels according to the deviation type. Each mapping channel is equipped with an independent feature extraction layer, a historical deviation sample library, and a similarity matching discrimination unit. The sample library of each channel stores different deviation amplitudes and deviation durations under the corresponding dimension. The standard error correction coefficient sample set matched with the slope of the deviation change is used. During the online calculation stage, the complete running deviation vector is split into dimensions and sent to the corresponding single-dimensional deviation component mapping channel. Each channel first performs normalized feature extraction on the input single-dimensional deviation component, extracting three types of features: deviation amplitude, time-series fluctuation trend, and offset direction. Then, the Euclidean distance algorithm is used to calculate the similarity between the extracted features and the historical standard deviation samples in the channel one by one. The standard deviation sample with the highest similarity is selected as the matching deviation pattern, and the single-dimensional basic correction coefficient bound to the sample is extracted. After all channels complete independent pattern matching, the basic correction coefficients of each dimension are summarized and combined to form a multivariate error correction parameter containing multi-dimensional compensation coefficients, which is used for subsequent step-by-step error compensation calculation of the initial predicted output.

[0077] Based on the degree of influence of errors on the power output prediction results, the priority of the correction components is determined as follows: meteorological forecast deviation component, unit equipment loss deviation component, site environment attenuation deviation component, and time-series dynamic lag deviation component. The proportional correction coefficient and offset correction constant corresponding to each priority are extracted from the multivariate error correction parameters. Let the initial predicted power output be... First, the highest priority weather deviation correction is performed, and the correction formula is as follows: In the formula This is the meteorological deviation proportionality coefficient. As a meteorological offset constant, the first-order corrected output is obtained. Then with As input, the secondary unit equipment loss deviation correction is carried out, and the calculation formula is as follows: , , These parameters correspond to the proportion of equipment loss and the offset correction parameters, respectively, and output the secondary corrected output power. Then Substitute it into the site environment attenuation deviation correction stage, through Complete the three-level correction to obtain Finally, for Perform the lowest priority timing dynamic lag bias correction, the calculation formula is as follows: , , The time lag correction coefficients are calculated in four stages of step-by-step composite calculations according to priority, ultimately outputting a corrected prediction output that integrates multi-dimensional error compensation and offsets multi-level prediction offsets. .

[0078] In one possible implementation, step S320 further includes: Step S321: Calculate the deviation components of multiple operating parameters of the real-time operating status data relative to the standard operating status data.

[0079] Step S322: Using a preset deviation judgment threshold, after filtering out P effective deviation components from the multiple operating parameter deviation components, vectorize and concatenate them to obtain the operating deviation vector.

[0080] Specifically, the real-time operating status datasets of microgrid photovoltaic, wind power, and power distribution equipment at the same time point are first matched with standard operating status datasets under the same operating conditions. These two datasets contain various corresponding parameters such as tilt surface irradiance, cell operating temperature, effective wind speed at hub height, inverter conversion loss, wind turbine air density, module aging degradation coefficient, and in-situ wake loss coefficient. A deviation calculation formula is then used for each pair of parameters. Solve the individual deviation components independently, where For the i-th running parameter deviation component, This represents the real-time value of the i-th parameter. This represents the standard reference value of the parameter under the corresponding working condition; the difference calculation is performed on all parameters in sequence to obtain the original deviation components corresponding to meteorological, equipment loss, and site attenuation categories, respectively, and the positive and negative offset directions and deviation amplitudes of each parameter deviation are completely preserved to form a set of original deviation components containing deviation information of all dimensions.

[0081] Retrieve the pre-configured deviation judgment threshold for each type of operating parameter Take each original deviation component one by one. absolute value With the corresponding threshold In comparison, if If the deviation is determined to be an effective deviation component with actual error disturbance effect, then the weak invalid deviation is eliminated. After traversing all the original deviation components, a total of P effective deviation components are selected. According to the fixed parameter dimension sorting rules predefined by the system, the selected P effective deviation components are sequentially and orderly concatenated into a one-dimensional ordered array with the array dimension fixed at P. This ordered array is the standardized running deviation vector, which fully preserves the amplitude and positive and negative offset characteristics of each effective deviation, and serves as the input feature quantity of the subsequent deviation-output error mapping model.

[0082] In one possible implementation, step S520 further includes: Extreme values ​​of meteorological elements are extracted from the meteorological forecast data, extreme weather models are matched, and the meteorological risk level is quantitatively output.

[0083] Specifically, the system first iterates through the full-time time-series meteorological forecast data, selecting the maximum and minimum values ​​for each time period for hub height wind speed, ambient temperature, tilt surface irradiance, precipitation intensity, and atmospheric pressure as extreme values ​​for meteorological elements. It then retrieves multiple built-in extreme weather standard models, including high temperature, super gusts, extreme strong light, heavy rain, and low-pressure strong winds, binding upper and lower thresholds to each model. The extracted extreme values ​​for each element are then compared with the threshold ranges of each extreme model to complete model matching. For each matched extreme weather category, a single-element risk score is calculated. The single-element scoring formula is as follows: In the formula For the i-th type of meteorological element, the risk score is... Preset a weighting coefficient for this element. The extracted feature extreme values, This is the standard baseline value for this element. To correspond to the critical threshold for extreme weather events; the overall risk score is obtained by summing the individual scores of all successfully matched elements. Then, pre-define four score intervals: Low risk Medium risk High risk, For extremely high risk, the corresponding meteorological risk level, total risk score, and extreme weather type that triggers the risk are matched and quantified based on the interval where the comprehensive total score is located.

[0084] To intuitively verify the advantages of the control method of the present invention for microgrid regulation under extreme weather conditions, this embodiment selects a 24-hour extreme weather simulation period and compares the meteorological risk rolling optimization energy storage control scheme of the present invention with the existing traditional static energy storage regulation scheme for all time periods. The quantitative performance data are shown in Table 1: As can be seen from the data in the table, this invention significantly reduces the prediction error of new energy output by predicting meteorological risks, correcting multi-dimensional deviations in layered output, and implementing a rolling iterative energy storage charging and discharging strategy. It effectively suppresses the drastic fluctuations in bus power caused by extreme weather, shortens the duration of microgrid power deficit, significantly improves the local consumption level of wind and solar new energy, and shortens the energy storage power regulation response delay. It solves the defects of existing technologies in power imbalance and poor operational stability under extreme weather conditions, and realizes dynamic adaptive and precise control of energy storage under meteorological risk scenarios.

[0085] Example 2, based on the same inventive concept as the microgrid meteorological risk control method in the previous examples, such as... Figure 2 As shown, this application provides an energy storage charging and discharging control system for microgrids under meteorological risks. The system and method embodiments in this application are based on the same inventive concept. The system includes: The data retrieval module 10 is used to retrieve meteorological forecast data for a preset time window through the data interface of the external meteorological forecast service module.

[0086] The real-time operating status data acquisition module 20 is used to collect the local operating status of the target microgrid and obtain real-time operating status data.

[0087] The modified forecast output module 30 is used to use the meteorological forecast data and real-time operating status data to predict the output of new energy power generation under meteorological risk conditions and output the modified forecast output.

[0088] The load demand forecast output module 40 is used to forecast the load demand of the target microgrid using the meteorological forecast data as input variables and output the load demand forecast value.

[0089] The charge / discharge execution module 50 is used to construct an energy storage charge / discharge strategy with the modified predicted output as the generation-side constraint and the predicted load demand as the load-side constraint, and drive the target microgrid to perform energy storage charge / discharge operations.

[0090] The dynamic power control module 60 is used to adjust the energy storage charging and discharging operation by incrementally updating the weather forecast data and continuously optimizing the energy storage charging and discharging strategy, so as to realize dynamic power control of the energy storage system of the target microgrid.

[0091] Furthermore, the system is also used to implement the following functions: Based on the meteorological forecast data, the power output of the target microgrid is predicted by new energy sources to obtain the initial predicted power output; the operating deviation vector of the real-time operating status data relative to the standard operating status data is calculated; based on the operating deviation vector, the error correction of the initial predicted power output is performed, and the corrected predicted power output is output.

[0092] Furthermore, the system is also used to implement the following functions: Based on the geographic information of the new energy power stations of the target microgrid, a spatiotemporal mapping engine for meteorological elements is constructed to reconstruct the meteorological elements of the meteorological forecast data in a spatiotemporal manner, obtaining a photovoltaic meteorological element sequence and a wind power meteorological element sequence. The power output of the photovoltaic and wind power meteorological element sequences is converted, and the initial predicted output values ​​of photovoltaic and wind power are output. Based on the site environmental characteristics of the target microgrid, environmental additional corrections are made to the initial predicted output values ​​of photovoltaic and wind power, outputting the corrected output values ​​of photovoltaic and wind power. Based on the temporal synchronization relationship between the photovoltaic and wind power meteorological element sequences, the corrected output values ​​of photovoltaic and wind power are superimposed to obtain the initial predicted output.

[0093] Furthermore, the system is also used to implement the following functions: Align the corrected predicted output and load demand forecast values, and perform time-step calculation of the net load sequence; assess the meteorological risk level based on the meteorological forecast data, and output the meteorological risk level; retrieve the real-time state of charge from the energy storage system of the target microgrid, and determine the energy storage operation mode by combining the meteorological risk level and the net load sequence to obtain the target operation mode; based on the target operation mode, coordinate the allocation of charging and discharging power and reserve capacity to obtain the charging and discharging power command sequence and reserve capacity reservation command, which constitute the energy storage charging and discharging strategy, and send it to the energy storage converter of the energy storage system for energy storage charging and discharging operation.

[0094] Furthermore, the system is also used to implement the following functions: The system analyzes the horizontal irradiance sequence, horizontal diffuse irradiance sequence, ambient temperature sequence, and wind speed sequence from the photovoltaic meteorological element sequence. Based on the installation angle information of the photovoltaic array in the target microgrid, it performs oblique projection conversion of the horizontal irradiance sequence and then combines it with the horizontal diffuse irradiance sequence to perform sky scattering synthesis, obtaining the total irradiance sequence of the oblique surface. Based on the total irradiance sequence of the oblique surface, ambient temperature sequence, and wind speed sequence, it estimates the cell operating temperature, obtaining the cell temperature sequence. It inputs the total irradiance sequence of the oblique surface and the cell temperature sequence into the equivalent circuit model of the photovoltaic module to calculate the DC power of the photovoltaic module, obtaining the module-level maximum power sequence. Based on the string structure of the photovoltaic array, it performs the current accumulation of the module-level maximum power sequence and combines it with the inverter configuration of the photovoltaic array to perform inverter efficiency conversion, outputting the predicted initial output value of the photovoltaic system.

[0095] Furthermore, the system is also used to implement the following functions: The wind turbine hub height of the wind turbines in the target microgrid is used to transform the meteorological element profile of the wind power meteorological element sequence to obtain hub height wind speed, hub height wind direction, hub height air temperature, and hub height air pressure. The actual air density sequence is calculated using the hub height air temperature and hub height air pressure. Based on the overall power characteristics of the wind turbines, the wake attenuation coefficient sequence of the hub height wind direction is calculated, and the wake wind speed of the hub height wind speed is reduced to obtain the wake corrected wind speed sequence. Based on the wake corrected wind speed sequence and the actual air density sequence, the power curve density correction calculation is performed, and the initial power output prediction value of a single unit is output. Then, the overall power output of the wind turbines is summed and aggregated to output the initial wind power output prediction value.

[0096] Furthermore, the system is also used to implement the following functions: The operating deviation vector is input into multiple single-dimensional deviation component mapping channels in the deviation-output error mapping model, and deviation pattern matching is performed to obtain multivariate error correction parameters. The correction component priority of the multivariate error correction parameters is used to perform step-by-step composite correction of the initial predicted output, and the corrected predicted output is output.

[0097] Furthermore, the system is also used to implement the following functions: Calculate the deviation components of multiple operating parameters of the real-time operating status data relative to the standard operating status data; after filtering out P valid deviation components from the multiple operating parameter deviation components using a preset deviation judgment threshold, perform vectorization and concatenation to obtain the operating deviation vector.

[0098] Furthermore, the system is also used to implement the following functions: Extreme values ​​of meteorological elements are extracted from the meteorological forecast data, extreme weather models are matched, and the meteorological risk level is quantitatively output.

[0099] It should be noted that the order of the embodiments described above is for descriptive purposes only and does not represent the superiority or inferiority of the embodiments. Specific embodiments of this specification have been described above. Furthermore, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0100] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0101] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A method for controlling the charging and discharging of energy storage under meteorological risks in microgrids, characterized in that, The method includes: Meteorological forecast data for preset time windows can be retrieved through the data interface of the external meteorological forecast service module; The target microgrid's local operating status is collected to obtain real-time operating status data; Using the aforementioned meteorological forecast data and real-time operational status data, the power output of new energy generation under meteorological risk conditions is predicted, and the corrected predicted power output is output. Using the meteorological forecast data as input variables, the load demand of the target microgrid is predicted, and the predicted load demand value is output. Using the modified predicted output as the generation-side constraint and the predicted load demand as the load-side constraint, an energy storage charging and discharging strategy is constructed to drive the target microgrid to perform energy storage charging and discharging operations. Based on the incremental updates of the meteorological forecast data, the energy storage charging and discharging strategy is continuously optimized to adjust the energy storage charging and discharging operation, so as to achieve dynamic power control of the energy storage system of the target microgrid.

2. The energy storage charging and discharging control method for microgrids under meteorological risks as described in claim 1, characterized in that, Using the aforementioned meteorological forecast data and real-time operational status data, the method for predicting the output of new energy power generation under meteorological risk conditions and outputting a corrected predicted output includes: Based on the meteorological forecast data, the power output of the new energy generation of the target microgrid is predicted to obtain the initial predicted power output; Calculate the operating deviation vector of the real-time operating status data relative to the standard operating status data; Based on the operating deviation vector, the error correction of the initial predicted output is performed, and the corrected predicted output is output.

3. The energy storage charging and discharging control method for microgrids under meteorological risks as described in claim 2, characterized in that, The method for predicting the renewable energy power output of the target microgrid based on the meteorological forecast data to obtain an initial predicted output includes: Based on the geographic information of the new energy power stations of the target microgrid, a meteorological element spatiotemporal mapping engine is constructed to reconstruct the meteorological elements of the meteorological forecast data in spatiotemporal form, thereby obtaining photovoltaic meteorological element sequences and wind power meteorological element sequences. The photovoltaic meteorological element sequence and the wind power meteorological element sequence are converted into power generation output, and the initial power output prediction values ​​of photovoltaic and wind power are output. Based on the site environment characteristics of the target microgrid, environmental additional corrections are made to the initial photovoltaic power output prediction and the initial wind power output prediction, and the corrected photovoltaic power output and the corrected wind power output are output. Based on the temporal synchronization relationship between the photovoltaic meteorological element sequence and the wind power meteorological element sequence, the photovoltaic corrected output and the wind power corrected output are superimposed to obtain the initial predicted output.

4. The energy storage charging and discharging control method for microgrids under meteorological risks as described in claim 1, characterized in that, Using the revised predicted output as the generation-side constraint and the predicted load demand as the load-side constraint, an energy storage charging and discharging strategy is constructed to drive the target microgrid to perform energy storage charging and discharging operations. The method includes: Align the revised forecast output and load demand forecast values, and perform time-step calculation of the net load sequence; Based on the aforementioned meteorological forecast data, a meteorological risk level assessment is performed, and the meteorological risk level is output. The real-time state of charge is retrieved from the energy storage system of the target microgrid, and the energy storage operation mode is determined by combining the meteorological risk level and net load sequence to obtain the target operation mode; Based on the target operating mode, the charging and discharging power and reserve capacity are coordinated and allocated to obtain the charging and discharging power command sequence and the reserve capacity reservation command, which constitute the energy storage charging and discharging strategy and are sent to the energy storage converter of the energy storage system for energy storage charging and discharging operation.

5. The energy storage charging and discharging control method for microgrids under meteorological risks as described in claim 3, characterized in that, The method further includes: The horizontal surface irradiance sequence, horizontal surface diffuse irradiance sequence, ambient temperature sequence, and wind speed sequence were analyzed from the photovoltaic meteorological element sequence. Based on the installation angle information of the photovoltaic array in the target microgrid, the oblique projection conversion of the horizontal plane irradiance sequence is performed, and then the sky scattering synthesis is performed in conjunction with the horizontal plane scattered irradiance sequence to obtain the total irradiance sequence of the oblique plane. Based on the total irradiance sequence of the inclined surface, the ambient temperature sequence, and the wind speed sequence, the operating temperature of the solar cell is estimated to obtain the solar cell temperature sequence. The total irradiance sequence of the inclined surface and the cell temperature sequence are input into the equivalent circuit model of the photovoltaic module to calculate the DC power of the photovoltaic module and obtain the module-level maximum power sequence. Based on the string structure of the photovoltaic array, after accumulating the maximum power sequence at the module level, the inverter efficiency is calculated in conjunction with the inverter configuration of the photovoltaic array, and the predicted initial output value of the photovoltaic is output.

6. The energy storage charging and discharging control method for microgrids under meteorological risks as described in claim 3, characterized in that, The method further includes: The wind power meteorological element sequence meteorological element profile transformation is performed using the hub height of the wind turbine in the target microgrid to obtain hub height wind speed, hub height wind direction, hub height air temperature and hub height air pressure. The actual air density sequence is calculated using the air temperature and air pressure at the wheel hub height. Based on the overall power characteristics of the wind turbine, the wake attenuation coefficient sequence of the wind direction at the hub height is calculated, and the wake wind speed at the hub height is reduced to obtain the wake corrected wind speed sequence. Based on the wake-corrected wind speed sequence and the actual air density sequence, the power curve density correction calculation is performed, and the initial output prediction value of a single unit is output. Then, the total output of the wind turbine is summed and aggregated, and the initial output prediction value of the wind power is output.

7. The energy storage charging and discharging control method for microgrids under meteorological risks as described in claim 2, characterized in that, Based on the operating deviation vector, the initial predicted output is corrected for error, and the corrected predicted output is output. The method includes: The operational deviation vector is input into multiple single-dimensional deviation component mapping channels in the deviation-output error mapping model, and deviation pattern matching is performed to determine the multivariate error correction parameters. The initial predicted output is corrected stepwise by using the priority of the correction components of the multivariate error correction parameters, and the corrected predicted output is output.

8. The energy storage charging and discharging control method for microgrids under meteorological risks as described in claim 2, characterized in that, The method for calculating the operational deviation vector of the real-time operational status data relative to the standard operational status data includes: Calculate the deviation components of multiple operating parameters of the real-time operating status data relative to the standard operating status data; Using a preset deviation judgment threshold, P valid deviation components are selected from the multiple operational parameter deviation components, and then vectorized and concatenated to obtain the operational deviation vector.

9. The energy storage charging and discharging control method for microgrids under meteorological risks as described in claim 4, characterized in that, Extreme values ​​of meteorological elements are extracted from the meteorological forecast data, extreme weather models are matched, and the meteorological risk level is quantitatively output.

10. A microgrid energy storage charging and discharging control system under meteorological risks, characterized in that, The system is used to implement the energy storage charging and discharging control method for microgrids under meteorological risks as described in any one of claims 1-9, and the system includes: The data retrieval module is used to retrieve weather forecast data for a preset time window through the data interface of the external weather forecast service module; The real-time operation status data acquisition module is used to collect the local operation status of the target microgrid and obtain real-time operation status data. The modified forecast output module is used to use the meteorological forecast data and real-time operating status data to predict the output of new energy power generation under meteorological risk conditions and output the modified forecast output. The load demand forecast output module is used to forecast the load demand of the target microgrid using the meteorological forecast data as input variables and output the load demand forecast value. The charge / discharge execution module is used to construct an energy storage charge / discharge strategy with the corrected predicted output as the generation-side constraint and the predicted load demand value as the load-side constraint, and drive the target microgrid to perform energy storage charge / discharge operations. The dynamic power control module is used to adjust the energy storage charging and discharging operation by incrementally updating the weather forecast data and continuously optimizing the energy storage charging and discharging strategy, so as to achieve dynamic power control of the energy storage system of the target microgrid.

Images