Parallel operation data interaction system and method of hybrid energy storage inverter

Abstract

The application discloses a kind of hybrid energy storage inverters and methods of parallel machine data interaction, it is related to power system technical field, including parallel machine data interaction center, data acquisition monitoring module, for collecting load data in distribution network, and monitoring the operating state of each hybrid energy storage inverter, carry out feature analysis to extract load operating characteristics, obtain load operating characteristic sequence table;Load prediction module is used to utilize historical data and load operating characteristic in sequence table, in combination with pre-trained load prediction model, obtains load trend index, analyzes the change trend of load.The application can fully master system operation by collecting load data and monitoring inverter operating state, timely find and handle exception, in combination with load prediction model, adjust inverter output and switch operating mode in advance, effectively cope with load fluctuation and power grid fault, significantly improve the stability and reliability of power system.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and specifically to a parallel data interaction system and method for hybrid energy storage inverters. Background Technology

[0002] With the rapid development of renewable energy sources such as solar and wind power, the stability and reliability of power systems face new challenges. In order to improve the utilization rate of renewable energy and the stability of power systems, the introduction of energy storage systems is particularly important. Hybrid energy storage inverters can combine battery energy storage systems with other types of energy storage devices to optimize the energy storage and conversion process.

[0003] For example, the parallel data interaction method for hybrid energy storage inverters, as disclosed in Chinese Patent Publication No. CN119765648A, has good real-time performance, smooth and efficient interaction, and can realize distributed communication and control between inverters, making it convenient for expansion and configuration.

[0004] In existing technologies, interactive data communication is performed according to standard frame classification to address the issues of high system bandwidth requirements and the susceptibility to communication delays and even failures. However, in parallel operation, multiple inverters share the load and jointly perform power dispatch. Under conditions of large load fluctuations or grid faults, the system's load regulation capability is insufficient, leading to inverter instability and affecting the security of remote control of switching devices in the power distribution network. Therefore, how to predict load changes in advance, adjust the inverter output, and switch operating modes to ensure the normal operation of switching devices in the power distribution network is the problem that this invention aims to solve. To this end, a parallel data interaction system and method for hybrid energy storage inverters is proposed. Summary of the Invention

[0005] The purpose of this invention is to provide a parallel data interaction system and method for hybrid energy storage inverters to solve the problems mentioned in the background art.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0007] In a first aspect, a parallel data interaction system for a hybrid energy storage inverter includes a parallel data interaction center, wherein the parallel data interaction center is communicatively connected to the following modules, wherein:

[0008] The data acquisition and monitoring module is used to collect load data in the power distribution network and monitor the operating status of each hybrid energy storage inverter to ensure a comprehensive understanding of the inverter's operating status, detect abnormalities in a timely manner, perform feature analysis to extract load operating characteristics, and obtain a load operating characteristic sequence table.

[0009] The load prediction module is used to obtain the load trend index by combining historical data and load operation characteristics in the load operation characteristic sequence table with a pre-trained load prediction model, and to analyze the load change trend.

[0010] The control and scheduling module is used to adjust the output power of each hybrid energy storage inverter based on the load forecast results and the operating status of each hybrid energy storage inverter, and to output the strategy for switching the inverter between different operating modes.

[0011] The remote control module is used to combine the switching strategies of the operating modes of each hybrid energy storage inverter and send remote control commands to the switching devices in the power distribution network through the parallel data interaction center to adjust the on / off state of the switches.

[0012] A further improvement to the technical solution of the present invention is that the data acquisition and monitoring module specifically includes:

[0013] Data acquisition equipment is deployed in the power distribution network and installed at the load equipment. It collects load data in the power distribution network at preset time intervals and transmits the collected raw load data to the central processing unit of the data acquisition and monitoring module via wireless communication.

[0014] While collecting load data, the monitoring sensors built into each hybrid energy storage inverter are used to monitor the operating status data of each hybrid energy storage inverter, and the monitored inverter operating status data is synchronized with the collected load data in time.

[0015] The collected load data and the monitored inverter operating status data are integrated, and the integrated data is cleaned and standardized to remove abnormal data and noise caused by equipment failure or communication interference, forming a load-status dataset.

[0016] Feature analysis is performed on the load data and operating status data in the load-state dataset to extract load operating features, including load power, load power fluctuation rate, load power factor, load harmonic content, and load average daily power. These load operating features are then integrated to obtain a load operating feature sequence table.

[0017] A further improvement to the technical solution of the present invention is that the load prediction module specifically includes:

[0018] Historical load data related to the current prediction period is retrieved from the historical database. At the same time, the corresponding load operation features are extracted from the load operation feature sequence table. The load operation feature data is normalized to unify data of different dimensions into a specific range. The data is then aggregated according to the prediction time scale to form a complete feature dataset.

[0019] The feature dataset is input into a pre-trained load prediction model based on a long short-term memory network as the model architecture. The load prediction model runs the feature data according to the input load and outputs a load change prediction value of the same length as the current prediction period.

[0020] Based on the load change prediction values ​​output by the load prediction model, the load change trend during the current prediction period is analyzed, and the prediction results are compared with the actual load data to evaluate the accuracy of the prediction. If the prediction error is large, the data acquisition strategy is adjusted or the model is retrained. At the same time, the prediction results and their evaluation information are stored in the system database.

[0021] A further improvement to the technical solution of the present invention lies in that: the analysis process of the load change trend during the current forecast period specifically includes:

[0022] The preprocessed feature dataset is input into a pre-trained load prediction model based on a long short-term memory network. The feature dataset includes normalized load power, load power fluctuation rate, load power factor, load harmonic content, and daily average load power. Through the transmission and activation of neurons layer by layer, the load change is predicted, and the predicted load change value with the same time length as the current prediction period is output.

[0023] Based on the load change prediction values ​​output by the load forecasting model, the degree of deviation between the load change prediction values ​​of each load operating characteristic and their corresponding standard values ​​is analyzed. Then, the load trend index is calculated by comprehensively considering the degree of deviation between the prediction values ​​of all load operating characteristics and the standard values.

[0024] By comparing the load trend index calculated for the current forecast period with the load trend index for the previous period, the load change trend for the current forecast period can be analyzed to determine whether the load in the current forecast period shows an upward, downward, or stable fluctuation trend.

[0025] A further improvement to the technical solution of the present invention is that the calculation process of the load tendency index is as follows:

[0026] For each load operating characteristic, calculate the relative deviation between its predicted value and the standard value. After calculating the relative deviation of each load operating characteristic, a set containing the relative deviations of all characteristics is obtained, where the relative deviation is obtained by subtracting the predicted value from the standard value and dividing by the standard value.

[0027] The relative deviation of each load operating characteristic is squared, and all squared values ​​are summed to obtain the sum of squares of the load operating characteristics. The sum of squares of the load operating characteristics is divided by the number of load operating characteristics to obtain the mean squared deviation. Then, the square root of the calculated mean squared deviation is taken to obtain the deviation in the form of standard deviation.

[0028] Calculate the absolute value of the relative deviation for each load operating characteristic, and then calculate the average of the absolute values ​​of the relative deviations to obtain the absolute average of the relative deviations. Finally, take the natural logarithm of the absolute average of the relative deviations to obtain the absolute deviation value.

[0029] The final load tendency index is obtained by combining the deviation in the form of standard deviation and the absolute deviation value.

[0030] A further improvement of the technical solution of the present invention is that the control and scheduling module includes a power scheduling unit and an operating mode switching unit;

[0031] The power dispatching unit is used to formulate a power allocation strategy for each inverter based on the output of the load prediction model, combined with the performance curves of each hybrid energy storage inverter and the grid dispatching requirements, and to dispatch the output power of each inverter to ensure load balance.

[0032] The operating mode switching unit is used to control the inverter to switch between different operating modes according to the system operating status and load requirements.

[0033] A further improvement to the technical solution of the present invention is that the power scheduling unit specifically includes:

[0034] The load forecast model obtains the predicted load change value and load trend index for the current forecast period. At the same time, it collects the performance curve data of each hybrid energy storage inverter and obtains grid dispatch requirements, including the power balance requirements and voltage stability requirements of the grid. The system integrates and analyzes these data to calculate the total power required by the system and the power range that each inverter should bear under the current load forecast and grid requirements.

[0035] Based on the output of the load forecasting model and the performance curves of each hybrid energy storage inverter, a power allocation strategy for each inverter is formulated. At the same time, according to the formulated power allocation strategy, the total power demand is allocated to each hybrid energy storage inverter, the output power target value of each inverter is defined, and a detailed power allocation scheme is formed.

[0036] According to the established power allocation strategy, the output power of each inverter is scheduled, and power adjustment commands are sent to each inverter through the communication interface to ensure that it operates in accordance with the predetermined power allocation strategy. At the same time, during the power scheduling process, the operating status and output power of each inverter are monitored in real time and compared with the allocated target value. If there is a deviation between the actual output power and the target value, the cause is analyzed, the power allocation strategy is adjusted or the inverter is troubleshooted to ensure that the output power of each inverter always conforms to the allocation scheme.

[0037] A further improvement to the technical solution of the present invention is that the operating mode switching unit specifically includes:

[0038] The system continuously monitors the operating status and load demand in real time. At the same time, it combines the load change trend output by the load prediction model with the current actual load value to evaluate the size and rate of change of the load demand. Based on the monitored data and evaluation results, it calculates the system stability index and load adaptability index to determine whether the current system operation is in a stable state and whether the load demand is within the adaptability range of the inverter's current operating mode.

[0039] Based on the monitoring and evaluation results of system operating status and load demand, make decisions on inverter operating mode switching and clarify the new operating mode for each inverter;

[0040] According to the established operating mode switching strategy, the mode switching command is sent to each inverter through the communication interface to control the inverter to switch from the current operating mode to the new operating mode. During the switching process, the operating status of each inverter is monitored in real time. After the switching is completed, the operating status of each inverter is monitored in real time to ensure stable operation in the new operating mode.

[0041] The calculation process for the system stability index is as follows:

[0042] Extract the current actual load power from real-time monitoring data, obtain the load power prediction value for the current prediction period from the load prediction model, calculate the difference between the current actual load power and the load power prediction value, and then divide it by the load power prediction value and take the absolute value to obtain the load deviation value.

[0043] The power balance demand of the power grid is obtained from the power grid dispatching system, and the total power required by the current system is extracted. The power balance impact value of the power grid is obtained by subtracting the ratio of the power grid's power balance demand to the total power required by the system from 1.

[0044] The current grid voltage deviation is obtained from real-time monitoring data, and the grid rated voltage is obtained from grid parameters. Then, the grid voltage deviation impact value is obtained by subtracting the ratio of grid voltage deviation to grid rated voltage from 1.

[0045] By adding 1 to the reciprocal of the load deviation value and multiplying it by the power grid power balance influence value and the power grid voltage deviation influence value, the system stability index is obtained, and the stability of the current system operation is analyzed.

[0046] The calculation process for the load adaptability index is as follows:

[0047] The minimum and maximum output power of the j-th inverter are obtained from the inverter performance curve data. The power range value is obtained by subtracting the minimum output power of the j-th inverter from the current actual load power and dividing by the difference between the maximum and minimum output power of the j-th inverter.

[0048] By subtracting the calculated power range value from 1, the power range adaptability value is obtained. Then, the product of 1 plus the reciprocal of the load deviation value and the power range adaptability value is calculated to obtain the load adaptability index, which determines whether the load demand is within the adaptability range of the inverter's current operating mode.

[0049] A further improvement to the technical solution of the present invention is that the remote control module specifically includes:

[0050] The system receives the operating mode switching strategies of each hybrid energy storage inverter from the control and scheduling module, including the new operating mode of each inverter, the expected output power adjustment, and related control commands. At the same time, it obtains the power allocation scheme.

[0051] Based on the received operating mode switching strategy and power allocation scheme, the remote control module generates specific remote control commands, including adjusting the on / off status of switching devices in the distribution network to ensure the dynamic balance of the power grid load. Then, the generated remote control commands are safely and stably transmitted to the switching devices in the distribution network through the parallel data interaction center.

[0052] After receiving a remote control command, the switching device immediately parses and verifies it to confirm its validity. If the remote control command is correct, the switching device adjusts its own on / off state according to the requirements of the remote control command to realize the on / off control of the power grid line. At the same time, the switching device feeds back the operation result to the parallel data interaction center. The remote control module receives the feedback information in real time to confirm whether the command has been successfully executed, ensuring that each switching device can accurately respond to the control command.

[0053] After the switching devices complete their status adjustment, the remote control module continuously collects the operating data of the distribution network through the parallel data interaction center. It compares and analyzes the actual operating data with the expected targets to determine whether the power grid has reached a safe and stable operating state. If it finds that the load is still unbalanced or the power grid parameters exceed the safe range, it immediately activates the dynamic adjustment mechanism. Based on the new operating data and evaluation results, it regenerates remote control commands to readjust the relevant switching devices until the power grid load achieves dynamic balance, and the safety and stability of the distribution network are effectively guaranteed. At the same time, the entire control process and results are recorded and stored.

[0054] Secondly, a parallel data interaction method for hybrid energy storage inverters, implemented based on the aforementioned parallel data interaction system for hybrid energy storage inverters, includes the following steps:

[0055] Data acquisition equipment is deployed in the power distribution network to collect load data and monitor inverter operating status data in real time, forming a load-status dataset.

[0056] The collected load data and inverter operating status data are integrated, feature analysis is performed to extract load operating features, and a load operating feature sequence table is obtained.

[0057] By utilizing historical data and feature sequence lists, combined with a load prediction model pre-trained based on a long short-term memory network model, a load trend index is obtained to assess the load change trend.

[0058] Based on the load forecast results and inverter performance curves, a power allocation strategy is formulated, and the target output power value and operating mode switching strategy of each inverter are defined.

[0059] Based on the operation mode switching strategy and power distribution scheme, remote control commands are generated and transmitted to the power distribution network switchgear through the parallel data interaction center;

[0060] The switching device executes remote control commands, adjusts the on / off state, monitors the execution status in real time, and dynamically adjusts based on feedback information until the power grid load is balanced.

[0061] Due to the adoption of the above technical solution, the technical progress achieved by this invention compared to the prior art is as follows:

[0062] 1. This invention provides a parallel data interaction system and method for hybrid energy storage inverters. By collecting load data and monitoring the inverter's operating status, it can comprehensively grasp the system's operating status, promptly detect and handle anomalies, and, combined with a load prediction model, adjust the inverter output and switch operating modes in advance to effectively cope with load fluctuations and grid faults, significantly improving the stability and reliability of the power system.

[0063] 2. This invention provides a parallel data interaction system and method for hybrid energy storage inverters. By utilizing historical data and load operation characteristics, combined with a pre-trained load prediction model, it can accurately predict load change trends, thereby formulating a reasonable power allocation strategy to ensure that each inverter operates at optimal power, avoid energy waste, improve energy utilization efficiency, and optimize energy allocation.

[0064] 3. This invention provides a parallel data interaction system and method for hybrid energy storage inverters, which supports flexible switching of inverters between different operating modes and automatic adjustment according to system operating status and load demand, enabling the system to adapt to various complex operating conditions, improving the system's adaptability and flexibility, and ensuring stable operation under different load conditions. Attached Figure Description

[0065] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0066] Figure 1 This is a schematic diagram of the workflow of the present invention;

[0067] Figure 2 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0068] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0069] Example 1, as Figure 1 As shown, this invention provides a parallel data interaction system for hybrid energy storage inverters, including a parallel data interaction center. The parallel data interaction center is communicatively connected to the following modules, wherein:

[0070] The data acquisition and monitoring module is used to collect load data in the power distribution network and monitor the operating status of each hybrid energy storage inverter. This ensures a comprehensive understanding of the inverter's operating status, timely detection of anomalies, and feature analysis to extract load operating characteristics, resulting in a load operating characteristic sequence table. Data acquisition devices are deployed in the power distribution network and installed at the load devices. These devices collect load data at preset time intervals. The data acquisition devices include high-precision current transformers and voltage sensors, which collect load data such as current, voltage, and power in real time, ensuring data accuracy and real-time performance. The collected raw load data is transmitted wirelessly to the central processing unit of the data acquisition and monitoring module. Simultaneously, the module utilizes the built-in monitoring sensors of each hybrid energy storage inverter to monitor its operating status data. This operating status data includes, but is not limited to, the inverter's input and output voltage, current, frequency, temperature, power factor, and the operating status of key internal components. The module synchronizes the monitored inverter operating status data with the collected load data. The monitored inverter operating status data is integrated, and the integrated data is cleaned and standardized to remove abnormal data and noise caused by equipment failure or communication interference, forming a load-state dataset. Feature analysis is performed on the load data and operating status data in the load-state dataset to extract load operating characteristics, including load power, load power fluctuation rate, load power factor, load harmonic content, and load average daily power. The load operating characteristics are then integrated to obtain a load operating characteristic sequence table. Load power represents the power demand of the load and is used to monitor load demand in real time and guide the power allocation of the inverter. Load power fluctuation rate characterizes the rate of change of load power and reflects the dynamic characteristics of the load. Load power factor represents the ratio of active power consumed by the load to apparent power and reflects the electrical characteristics of the load. It is used to optimize grid efficiency and reduce reactive power loss. Load harmonic content characterizes the degree of harmonic distortion of the load current or voltage waveform and reflects the impact of the load on grid quality. It is used to assess the impact of the load on grid quality and optimize power system operation. Load average daily power reflects the average power demand of the load over a day and is used to assess long-term load change trends.

[0071] The load prediction module utilizes historical data and load operation characteristics from a load operation characteristic sequence table, combined with a pre-trained load prediction model, to obtain a load trend index and analyze load change trends. It retrieves historical load data relevant to the current prediction period from a historical database and extracts corresponding load operation characteristics from the load operation characteristic sequence table. The load operation characteristic data is normalized to unify data of different dimensions to a specific range, and aggregated according to the prediction time scale to form a complete feature dataset. This feature dataset is input into a pre-trained load prediction model based on a Long Short-Term Memory (LSTM) network. The load prediction model outputs a predicted load change value for the same time period as the current prediction period based on the input load operation characteristic data. This involves collecting a large amount of historical load data and corresponding load-state comprehensive characteristic data, including data from different time periods and different load types. The collected data is divided into training, validation, and test sets in a 7:2:1 ratio. The training set is used for learning and adjusting model parameters, the validation set is used to evaluate model performance during training to prevent overfitting, and the test set is used for... Finally, the generalization ability of the model is evaluated. A deep learning model architecture based on a long short-term memory network is selected. Based on data characteristics and prediction requirements, the number of network layers, the number of neurons in each layer, and other hyperparameters are determined. The long short-term memory network model is trained using a training set. During training, backpropagation and gradient descent optimization methods are used to adjust the model parameters to minimize the loss function value between the model's predicted output and the actual label. At the same time, the performance of the model on the validation set is monitored in real time. When the performance on the validation set no longer improves, training is stopped early to prevent overfitting. The trained long short-term memory network model is evaluated using a test set. Various evaluation indicators are calculated, and the model structure and parameters are continuously adjusted to optimize model performance until the model achieves satisfactory prediction results on the test set, thus obtaining the trained load prediction model. Based on the load change prediction values ​​output by the load prediction model, the load change trend of the current prediction period is analyzed, and the prediction results are compared with the actual load data to evaluate the accuracy of the prediction. If the prediction error is large, the data acquisition strategy is adjusted or the model is retrained. Meanwhile, the prediction results and evaluation information are stored in the system database.

[0072] Furthermore, the analysis process for the load change trend during the current forecast period specifically includes:

[0073] The preprocessed feature dataset is input into a pre-trained load prediction model based on a long short-term memory network. The feature dataset includes normalized load power, load power volatility, load power factor, load harmonic content, and daily average load power. Through the transmission and activation of layers of neurons, load changes are predicted, and the predicted load change value with the same time length as the current prediction period is output. Based on the predicted load change value output by the load prediction model, the deviation of the predicted load change value of each load operation feature from its corresponding standard value is analyzed. Then, the deviation of all predicted load operation features from the standard value is combined to calculate the load trend index. The load trend index calculated in the current prediction period is compared with the load trend index in the previous period to analyze the load change trend in the current prediction period and determine whether the load in the current prediction period shows an upward, downward, or stable fluctuation trend.

[0074] Furthermore, the calculation process for the load tendency index is as follows:

[0075] For each load operating characteristic, the relative deviation between its predicted value and the standard value is calculated. After calculating the relative deviation of each load operating characteristic, a set containing the relative deviations of all characteristics is obtained. The relative deviation is obtained by subtracting the predicted value from the standard value and dividing by the standard value. The relative deviation of each load operating characteristic is squared, and all squared values ​​are added together to obtain the sum of squares of the load operating characteristics. The influence of the deviation is amplified so that larger deviations occupy a more important position in the sum. The sum of squares of the load operating characteristics is divided by the number of load operating characteristics to obtain the mean squared deviation. By calculating the mean deviation of all load operating characteristics, it is ensured that each characteristic contributes equally to the final result. Then, the square root of the calculated mean squared deviation is taken to obtain the deviation in the form of standard deviation. The deviation value is restored to the same dimension as the original data. The absolute value of the relative deviation of each load operating characteristic is calculated, and the average of the absolute values ​​of the relative deviations is calculated to obtain the absolute mean of the relative deviation. Then, the natural logarithm of the absolute mean of the relative deviation is taken to obtain the absolute deviation value. The deviation range is scaled by the logarithmic operation to make the influence of the deviation more in line with actual needs. The deviation in the form of standard deviation and the absolute deviation value are combined to obtain the final load tendency index.

[0076] The expression for calculating the load tendency index is as follows:

[0077] ;

[0078] In the formula, For load directional index, For the first Predicted values ​​for each load operating characteristic For the first Standard values ​​for each load operating characteristic For the number of load operating characteristics, when all predicted values ​​are equal to their corresponding standard values, Taking the minimum value of 0, as the deviation between the predicted value and the standard value increases, The value also increases, indicating that the drastic change in load is increasing and the trend of load change is obvious, requiring corresponding control measures;

[0079] The control and scheduling module is used to adjust the output power of each hybrid energy storage inverter based on the load forecast results and the operating status of each hybrid energy storage inverter, and to output the strategy for switching the inverter between different operating modes.

[0080] The remote control module is used to combine the switching strategies of the operating modes of each hybrid energy storage inverter and send remote control commands to the switching devices in the power distribution network through the parallel data interaction center to adjust the on / off state of the switches, ensure the dynamic balance of the power grid load, and guarantee the safety and stability of the power distribution network.

[0081] Example 2, as Figure 1 As shown, based on Embodiment 1, the present invention provides a technical solution: preferably, the control scheduling module includes a power scheduling unit and an operating mode switching unit;

[0082] The power dispatch unit is used to formulate power allocation strategies for each inverter based on the output of the load forecasting model, combined with the performance curves of each hybrid energy storage inverter and grid dispatching requirements. It dispatches the output power of each inverter to ensure load balance. It obtains the load change forecast and load trend index for the current forecast period from the load forecasting model. Simultaneously, it collects performance curve data of each hybrid energy storage inverter, including key parameters such as the inverter's output power range, efficiency curve, and maximum output power. It also obtains grid dispatching requirements, including power balance requirements and voltage stability requirements, and performs integrated analysis to calculate the total power required by the system and the power range that each inverter should handle under the current load forecast and grid requirements. Based on the output of the load forecasting model and the performance curves of each hybrid energy storage inverter, it formulates power allocation strategies for each inverter. Specifically, for performance... High-performance, fast-response, and efficient inverters are appropriately allocated more power tasks, while less powerful inverters are allocated a suitable amount of power to avoid overload operation. Simultaneously, according to the established power allocation strategy, the total power demand is distributed to each hybrid energy storage inverter, clearly defining the output power target value for each inverter and forming a detailed power allocation scheme. Based on the established power allocation strategy, the output power of each inverter is scheduled, and power adjustment commands are sent to each inverter through the communication interface to ensure operation according to the predetermined power allocation strategy. During power scheduling, the operating status and output power of each inverter are monitored in real time and compared with the allocated target value. If there is a deviation between the actual output power and the target value, the cause is analyzed, the power allocation strategy is adjusted, or the inverter is troubleshooted to ensure that the output power of each inverter always conforms to the allocation scheme.

[0083] The formula for calculating the total power required by the system is as follows:

[0084] ;

[0085] ;

[0086] In the formula, The total power required by the system, For the first Predicted values ​​for each load operating characteristic The power balance requirement of the power grid is represented by a positive value, indicating that additional power needs to be supplied to the grid, and a negative value, indicating that additional power needs to be absorbed from the grid. This represents the total load power demand for the current forecast period. This represents the total power generation for the current forecast period. This represents the total output power of the energy storage system for the current forecast period. Power loss during power transmission and distribution in the power grid;

[0087] The calculation formula for the power range that each inverter should handle is as follows:

[0088] ;

[0089] In the formula, For the first The power range that an inverter should be able to handle. For the first The minimum output power of each inverter For the first The maximum output power of each inverter;

[0090] The calculation expression for the target output power value of each inverter is as follows:

[0091] ;

[0092] In the formula, For the first Target output power value of each inverter This is the sum of the maximum output power of all inverters. The number of inverters, Used to indicate a specific inverter Used to iterate through all inverters during summation. Indicates all The sum of the maximum output power of each inverter, each inverter Target output power Based on its maximum output power The power distribution is calculated as a proportion of the sum of the maximum output power of all inverters to ensure its rationality and fairness.

[0093] The operating mode switching unit controls the inverter to switch between different operating modes based on the system operating status and load demand. This ensures that the inverter operates in the optimal mode under different conditions, improving system adaptability and stability. It continuously monitors the system operating status and load demand in real time. Simultaneously, it assesses the magnitude and rate of change of load demand by combining the load change trend output from the load prediction model with the current actual load value. Based on the monitored data and assessment results, it calculates the system stability index and load adaptability index to determine whether the current system operation is stable and whether the load demand is within the adaptability range of the inverter's current operating mode. Based on the monitoring and assessment results of the system operating status and load demand, it formulates inverter operating mode switching decisions, clarifying the new operating mode for each inverter. Specifically, when the load demand suddenly increases and the current inverter operating mode cannot meet the demand, if the detection... If the conditions for switching the inverter to high-power operation mode are detected (temperature not exceeding limits, voltage stable, etc.), the inverter will be switched to high-power operation mode to increase output power and meet load demand. Conversely, when the load demand decreases, in order to improve system efficiency and reduce energy consumption, if the inverter's current operation mode is inefficient and meets the conditions for switching to low-power or standby mode, the corresponding switch will be performed. According to the established operation mode switching strategy, mode switching commands are sent to each inverter through the communication interface to control the inverter to switch from the current operation mode to the new operation mode. During the switching process, the operating status of each inverter is monitored in real time to ensure that the switching process is smooth and without abnormalities. If any problems are detected during the switching process, measures are taken immediately to handle the fault, including resending the command or switching to standby mode. After the switching is completed, the operating status of each inverter continues to be monitored in real time to ensure stable operation in the new operation mode.

[0094] Furthermore, the calculation process for the system stability index is as follows:

[0095] The system extracts the current actual load power from real-time monitoring data and obtains the predicted load power value for the current forecast period from the load forecast model. It calculates the difference between the current actual load power and the predicted load power value, divides it by the predicted load power value, and takes the absolute value to obtain the load deviation value. It obtains the power balance demand of the power grid from the power grid dispatch system and extracts the total power required by the current system. It subtracts the ratio of the power balance demand of the power grid to the total power required by the system from 1 to obtain the power grid power balance impact value. It obtains the current power grid voltage deviation from real-time monitoring data and obtains the rated voltage of the power grid from the power grid parameters. It subtracts the ratio of the power grid voltage deviation to the rated voltage of the power grid from 1 to obtain the power grid voltage deviation impact value. It adds 1 to the load deviation value, takes the reciprocal, and multiplies it with the power grid power balance impact value and the power grid voltage deviation impact value to obtain the system stability index. It then analyzes the stability of the current system operation.

[0096] The formula for calculating the system stability index is as follows:

[0097] ;

[0098] In the formula, This is the system stability index, with values ​​ranging from 0 to 1. The closer the value is to 1, the more stable the system. This represents the current actual load power. This is the predicted load power value. To meet the power balance requirements of the power grid, The total power required by the system, For grid voltage deviation, The rated voltage of the power grid, when near , Smaller and When smaller, A value close to 1 indicates that the system is running stably. and Large deviation Larger or When it is large, A value close to 0 indicates that the system is unstable.

[0099] Furthermore, the calculation process for the load adaptability index is as follows:

[0100] The minimum and maximum output power of the j-th inverter are obtained from the inverter performance curve data. The minimum output power of the j-th inverter is subtracted from the current actual load power, and then divided by the difference between the maximum and minimum output power of the j-th inverter to obtain the power range value. The calculated power range value is obtained by subtracting the calculated power range value from 1. Then, the product of 1 plus the reciprocal of the load deviation value and the power range value is calculated to obtain the load adaptability index, which determines whether the load demand is within the adaptability range of the inverter's current operating mode.

[0101] The formula for calculating the load adaptability index is as follows:

[0102] ;

[0103] In the formula, This is the load adaptability index, with a value ranging from 0 to 1. The closer the value is to 1, the better the load demand adapts to the current operating mode. For the first The minimum output power of each inverter For the first The maximum output power of the inverter, when near And the current actual load power Within the power range of the inverter A value close to 1 indicates that the load demand is adapted to the current operating mode. and Large deviation or When the power range of the inverter is exceeded. A value close to 0 indicates that the load demand is not suitable for the current operating mode;

[0104] The remote control module specifically includes:

[0105] The system receives operating mode switching strategies from each hybrid energy storage inverter from the control and scheduling module. This includes the new operating mode for each inverter, the expected output power adjustment, and related control commands. Simultaneously, it acquires a power allocation scheme to ensure that the inverter's output power adjustment matches the grid load demand. Using stability and load adaptability indices, it ensures that the system is in a stable state and load demand is met when executing remote control commands. Based on the received operating mode switching strategies and power allocation scheme, the remote control module generates specific remote control commands, including adjusting the on / off states of switchgear in the distribution network to ensure dynamic grid load balance. These remote control commands are then securely and stably transmitted to the switchgear in the distribution network via the parallel data interaction center. Upon receiving the remote control commands, the switchgear immediately parses and verifies them to confirm their validity. If the remote control commands are correct, the switchgear... According to the remote control commands, the switching devices adjust their own on / off states to control the on / off state of the power grid lines. At the same time, the switching devices feed back the operation results to the parallel data interaction center. The remote control module receives the feedback information in real time, confirms whether the command was successfully executed, and ensures that each switching device can accurately respond to the control command. After the switching devices complete the state adjustment, the remote control module continuously collects the operation data of the distribution network through the parallel data interaction center, compares and analyzes the actual operation data with the expected target, and judges whether the power grid has reached a safe and stable operating state. If it is found that the load is still unbalanced or the power grid parameters exceed the safe range, the dynamic adjustment mechanism is immediately activated. Based on the new operation data and evaluation results, the remote control command is regenerated to readjust the relevant switching devices until the power grid load achieves dynamic balance. The safety and stability of the distribution network are effectively guaranteed. At the same time, the entire control process and results are recorded and stored.

[0106] Example 3, as Figure 1 , Figure 2 As shown, based on Embodiments 1-2, the present invention also provides a parallel data interaction method for hybrid energy storage inverters, implemented based on the above-mentioned parallel data interaction system for hybrid energy storage inverters, including the following steps:

[0107] Data acquisition equipment is deployed in the power distribution network to collect load data and monitor inverter operating status data in real time, forming a load-status dataset.

[0108] The collected load data and inverter operating status data are integrated, feature analysis is performed to extract load operating features, and a load operating feature sequence table is obtained.

[0109] By utilizing historical data and feature sequence lists, combined with a load prediction model pre-trained based on a long short-term memory network model, a load trend index is obtained to assess the load change trend.

[0110] Based on the load forecast results and inverter performance curves, a power allocation strategy is formulated, and the target output power value and operating mode switching strategy of each inverter are defined.

[0111] Based on the operation mode switching strategy and power distribution scheme, remote control commands are generated and transmitted to the power distribution network switchgear through the parallel data interaction center;

[0112] The switching device executes remote control commands, adjusts the on / off state, monitors the execution status in real time, and dynamically adjusts based on feedback information until the power grid load is balanced.

[0113] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A parallel data interaction system for a hybrid energy storage inverter, comprising a parallel data interaction center, characterized in that, The parallel data interaction center has the following communication connections, including: The data acquisition and monitoring module is used to collect load data in the power distribution network, monitor the operating status of each hybrid energy storage inverter, perform feature analysis to extract load operating characteristics, and obtain a load operating characteristic sequence table. The load forecasting module utilizes historical data and load operation characteristics from a load operation characteristic sequence table, combined with a pre-trained load forecasting model, to obtain a load trend index and analyze load change trends. Specifically, it includes: Historical load data related to the current prediction period is retrieved from the historical database. At the same time, the corresponding load operation features are extracted from the load operation feature sequence table. The load operation feature data is normalized and aggregated according to the prediction time scale to form a complete feature dataset. The feature dataset is input into a pre-trained load prediction model based on a long short-term memory network as the model architecture. The load prediction model runs the feature data according to the input load and outputs a load change prediction value of the same length as the current prediction period. Based on the load change prediction values ​​output by the load prediction model, analyze the load change trend during the current prediction period, compare the prediction results with the actual load data, evaluate the accuracy of the prediction, and if the prediction error is large, adjust the data acquisition strategy or retrain the model. At the same time, store the prediction results and evaluation information in the system database. The analysis process of the load change trend during the current forecast period specifically includes: The preprocessed feature dataset is input into a pre-trained load prediction model based on a long short-term memory network. The feature dataset includes normalized load power, load power fluctuation rate, load power factor, load harmonic content, and daily average load power. Through the transmission and activation of neurons layer by layer, the load change is predicted, and the predicted load change value with the same time length as the current prediction period is output. Based on the load change prediction values ​​output by the load forecasting model, the degree of deviation between the load change prediction values ​​of each load operating characteristic and their corresponding standard values ​​is analyzed. Then, the load trend index is calculated by comprehensively considering the degree of deviation between the prediction values ​​of all load operating characteristics and the standard values. By comparing the load trend index calculated for the current forecast period with the load trend index for the previous period, the load change trend for the current forecast period can be analyzed to determine whether the load in the current forecast period shows an upward, downward, or stable fluctuation trend. The calculation process for the load tendency index is as follows: For each load operating characteristic, calculate the relative deviation between its predicted value and the standard value. After calculating the relative deviation of each load operating characteristic, a set containing the relative deviations of all characteristics is obtained, where the relative deviation is obtained by subtracting the predicted value from the standard value and dividing by the standard value. The relative deviation of each load operating characteristic is squared, and all squared values ​​are summed to obtain the sum of squares of the load operating characteristics. The sum of squares of the load operating characteristics is divided by the number of load operating characteristics to obtain the mean squared deviation. Then, the square root of the calculated mean squared deviation is taken to obtain the deviation in the form of standard deviation. Calculate the absolute value of the relative deviation for each load operating characteristic, and then calculate the average of the absolute values ​​of the relative deviations to obtain the absolute average of the relative deviations. Finally, take the natural logarithm of the absolute average of the relative deviations to obtain the absolute deviation value. The load tendency index is obtained by combining the deviation in the form of standard deviation and the absolute deviation value. The control and scheduling module is used to adjust the output power of each hybrid energy storage inverter based on the load forecast results and the operating status of each hybrid energy storage inverter, and to output the strategy for switching the inverter between different operating modes. The remote control module is used to send remote control commands to the switching devices in the power distribution network and adjust the switching status by combining the switching strategies of the operating modes of each hybrid energy storage inverter.

2. The parallel data interaction system for a hybrid energy storage inverter according to claim 1, characterized in that: The data acquisition and monitoring module specifically includes: Data acquisition equipment is deployed in the power distribution network and installed at the load equipment. It collects load data in the power distribution network at preset time intervals and transmits the collected raw load data to the central processing unit of the data acquisition and monitoring module via wireless communication. While collecting load data, the monitoring sensors built into each hybrid energy storage inverter are used to monitor the operating status data of each hybrid energy storage inverter, and the monitored inverter operating status data is synchronized with the collected load data in time. The collected load data and the monitored inverter operating status data are integrated, and the integrated data is cleaned and standardized to form a load-status dataset. Feature analysis is performed on the load data and operating status data in the load-state dataset to extract load operating features, including load power, load power fluctuation rate, load power factor, load harmonic content, and load average daily power. These load operating features are then integrated to obtain a load operating feature sequence table.

3. The parallel data interaction system for a hybrid energy storage inverter according to claim 1, characterized in that: The control and scheduling module includes a power scheduling unit and an operating mode switching unit; The power dispatching unit is used to formulate a power allocation strategy for each inverter and dispatch the output power of each inverter based on the output of the load prediction model, combined with the performance curves of each hybrid energy storage inverter and the grid dispatching requirements. The operating mode switching unit is used to control the inverter to switch between different operating modes according to the system operating status and load requirements.

4. The parallel data interaction system for a hybrid energy storage inverter according to claim 3, characterized in that: The power scheduling unit specifically includes: The load forecast model obtains the predicted load change value and load trend index for the current forecast period. At the same time, it collects the performance curve data of each hybrid energy storage inverter and obtains the grid dispatch requirements. The data is then integrated and analyzed to calculate the total power required by the system and the power range that each inverter should bear under the current load forecast and grid requirements. Based on the output of the load forecasting model and the performance curves of each hybrid energy storage inverter, a power allocation strategy for each inverter is formulated. At the same time, according to the formulated power allocation strategy, the total power demand is allocated to each hybrid energy storage inverter, the output power target value of each inverter is defined, and a detailed power allocation scheme is formed. According to the established power allocation strategy, the output power of each inverter is scheduled, and power adjustment commands are sent to each inverter through the communication interface. At the same time, the operating status and output power of each inverter are monitored in real time and compared with the allocated target value. If there is a deviation between the actual output power and the target value, the cause is analyzed, the power allocation strategy is adjusted, or the inverter is troubleshooted.

5. The parallel data interaction system for a hybrid energy storage inverter according to claim 4, characterized in that: The operating mode switching unit specifically includes: The system continuously monitors the operating status and load demand in real time. At the same time, it combines the load change trend output by the load prediction model with the current actual load value to evaluate the size and rate of change of the load demand. Based on the monitored data and evaluation results, it calculates the system stability index and load adaptability index to determine whether the current system operation is in a stable state and whether the load demand is within the adaptability range of the inverter's current operating mode. Based on the monitoring and evaluation results of system operating status and load demand, make decisions on inverter operating mode switching and clarify the new operating mode for each inverter; According to the established operating mode switching strategy, the mode switching command is sent to each inverter through the communication interface to control the inverter to switch from the current operating mode to the new operating mode. During the switching process, the operating status of each inverter is monitored in real time. After the switching is completed, the operating status of each inverter continues to be monitored in real time. The calculation process for the system stability index is as follows: Extract the current actual load power from real-time monitoring data, obtain the load power prediction value for the current prediction period from the load prediction model, calculate the difference between the current actual load power and the load power prediction value, and then divide it by the load power prediction value and take the absolute value to obtain the load deviation value. The power balance demand of the power grid is obtained from the power grid dispatching system, and the total power required by the current system is extracted. The power balance impact value of the power grid is obtained by subtracting the ratio of the power grid's power balance demand to the total power required by the system from 1. The current grid voltage deviation is obtained from real-time monitoring data, and the grid rated voltage is obtained from grid parameters. Then, the grid voltage deviation impact value is obtained by subtracting the ratio of grid voltage deviation to grid rated voltage from 1. By adding 1 to the reciprocal of the load deviation value and multiplying it by the power grid power balance influence value and the power grid voltage deviation influence value, the system stability index is obtained, and the stability of the current system operation is analyzed. The calculation process for the load adaptability index is as follows: The minimum and maximum output power of the j-th inverter are obtained from the inverter performance curve data. The power range value is obtained by subtracting the minimum output power of the j-th inverter from the current actual load power and dividing by the difference between the maximum and minimum output power of the j-th inverter. By subtracting the calculated power range value from 1, the power range adaptability value is obtained. Then, the product of 1 plus the reciprocal of the load deviation value and the power range adaptability value is calculated to obtain the load adaptability index, which determines whether the load demand is within the adaptability range of the inverter's current operating mode.

6. The parallel data interaction system for a hybrid energy storage inverter according to claim 5, characterized in that: The remote control module specifically includes: The system receives the operating mode switching strategies of each hybrid energy storage inverter from the control and scheduling module, including the new operating mode of each inverter, the expected output power adjustment, and related control commands. At the same time, it obtains the power allocation scheme. Based on the received operating mode switching strategy and power allocation scheme, the remote control module generates specific remote control commands, including adjusting the on / off state of the switching devices in the power distribution network, and then transmits the generated remote control commands to the switching devices in the power distribution network through the parallel data interaction center. After receiving the remote control command, the switch device immediately parses and verifies it to confirm the validity of the remote control command. If the remote control command is correct, the switch device adjusts its own on / off state according to the requirements of the remote control command. At the same time, the switch device feeds back the operation result to the parallel data interaction center. The remote control module receives the feedback information in real time to confirm whether the command has been successfully executed. After the switching devices complete their status adjustment, the remote control module continuously collects the operating data of the distribution network through the parallel data interaction center. It compares and analyzes the actual operating data with the expected targets to determine whether the power grid has reached a safe and stable operating state. If it finds that the load is still unbalanced or the power grid parameters exceed the safe range, it immediately activates the dynamic adjustment mechanism. Based on the new operating data and evaluation results, it regenerates remote control commands to readjust the relevant switching devices until the power grid load achieves dynamic balance. At the same time, the entire control process and results are recorded and stored.

7. A parallel data interaction method for a hybrid energy storage inverter, implemented based on the parallel data interaction system for the hybrid energy storage inverter as described in any one of claims 1-6, characterized in that, Includes the following steps: Data acquisition equipment is deployed in the power distribution network to collect load data and monitor inverter operating status data in real time, forming a load-status dataset. The collected load data and inverter operating status data are integrated, feature analysis is performed to extract load operating features, and a load operating feature sequence table is obtained. By utilizing historical data and feature sequence lists, combined with a load prediction model pre-trained based on a long short-term memory network model, a load trend index is obtained to assess the load change trend. Based on the load forecast results and inverter performance curves, a power allocation strategy is formulated, and the target output power value and operating mode switching strategy of each inverter are defined. Based on the operation mode switching strategy and power distribution scheme, remote control commands are generated and transmitted to the power distribution network switchgear through the parallel data interaction center; The switching device executes remote control commands, adjusts the on / off state, monitors the execution status in real time, and dynamically adjusts based on feedback information until the power grid load is balanced.

Images