A method and system for detecting the state of a load connected to an energy storage power supply

Abstract

The application belongs to the technical field of load access, and provides a kind of energy storage power load access state detection method and system, comprising: real-time monitoring the current after energy storage power load access, by analyzing the current change characteristics in monitoring period, the current change after energy storage power load access is preliminarily evaluated;Current prediction model is constructed, whether the predicted current shows a downward trend in the prediction period is judged;The specific recovery time point of current recovery to rated current is determined by using linear interpolation method, and the abnormal degree of energy storage power load access state is evaluated, the current recovery time point is determined, the comprehensive transient evaluation index is formed by fusing real-time and prediction data, the hidden faults such as battery internal resistance rise and component aging can be identified in advance, and the abnormal identification sensitivity is improved;For low scene of abnormal degree, pulse width modulation is used to dynamically adjust output power, and the current abnormal time is shortened under the premise of uninterrupted power supply.

Description

Technical Field

[0001] This invention belongs to the field of load access technology, specifically a method and system for detecting the load access status of an energy storage power supply. Background Technology

[0002] With the continuous development of energy storage technology, energy storage power supplies are being used more and more widely in various fields. During the process of connecting energy storage power supplies to loads, accurately detecting the load connection status and effectively analyzing current changes is crucial. However, existing technologies have many shortcomings and deficiencies in detecting the load connection status of energy storage power supplies.

[0003] When a load is connected to an energy storage power supply, the current will usually exceed the rated current briefly due to load characteristics and circuit transient processes. However, existing technologies cannot accurately determine whether such fluctuations are within the normal range through quantitative means, leading to misjudgment or omission of abnormal situations. This makes it impossible to detect hidden anomalies such as increased battery internal resistance or component aging in advance, and cannot provide reliable quantitative early warning basis for the safe operation of the system.

[0004] When an abnormal change in current is detected after a load is connected, existing technologies are not sufficiently sophisticated in analyzing the rate of change of predicted current. They cannot accurately determine whether the predicted current is decreasing within the predicted period, thus failing to provide strong support for taking reasonable measures. They only consider real-time monitored current data and do not integrate predicted current data for comprehensive analysis. It is difficult to accurately determine the specific recovery time point when the current recovers to the rated current using linear interpolation, resulting in an inaccurate assessment of the degree of abnormality and failing to provide a reliable basis for subsequent control decisions.

[0005] When the load connection status is of low abnormality, existing technology cannot effectively regulate the current without interrupting the load. It is difficult to reduce the output power in a reasonable way to shorten the current abnormality time, which may lead to unnecessary power outages and affect the normal operation of the system.

[0006] Therefore, the present invention provides a method and system for detecting the load connection status of an energy storage power source. Summary of the Invention

[0007] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0008] In a first aspect, the present invention provides a method for detecting the load connection status of an energy storage power source, comprising:

[0009] Real-time monitoring of the current after the energy storage power supply load is connected; by analyzing the current change characteristics during the monitoring period, a preliminary assessment of the current change after the energy storage power supply load is connected is made.

[0010] If the current change after the energy storage power supply load is connected is assessed as abnormal, a current prediction model is constructed. All current data within the monitoring period are used as input, the prediction period is preset, the predicted current within the prediction period is output, the rate of change of the predicted current is analyzed, and it is determined whether the predicted current shows a downward trend within the prediction period.

[0011] If the predicted current shows a downward trend during the prediction period, the specific recovery time point when the current recovers to the rated current is determined by linear interpolation. The time point when the current first exceeds the rated current during the monitoring period is recorded as the start time point. All current data between the start time point and the specific recovery time point are comprehensively analyzed to assess the degree of abnormality in the load connection status of the energy storage power supply.

[0012] If the load connection status of the energy storage power supply is low-level abnormal, pulse width modulation is activated to adjust the power without interrupting the load, thereby achieving current regulation and shortening the current abnormality time.

[0013] As a further aspect of the present invention, the specific process for the preliminary assessment of current changes after the energy storage power supply load is connected is as follows:

[0014] The monitoring period is divided into several monitoring intervals with equal time intervals. During the monitoring period, the current when the energy storage power supply load is connected is obtained in real time through the Hall current sensor. The current change curve is constructed to determine the excess time ratio and excess degree ratio.

[0015] The transient current excess index is obtained by multiplying the excess duration ratio and the excess degree ratio.

[0016] If the transient current excess index is greater than or equal to the transient current excess index threshold, it indicates that the current change is abnormal after the energy storage power supply load is connected.

[0017] As a further aspect of the present invention: the excess time ratio is obtained as follows:

[0018] Extract the section within the monitoring range that exceeds the rated current and record it as the excess range;

[0019] The duration corresponding to the excess interval within the statistical monitoring period is recorded as the excess duration. The excess duration is then compared with the duration corresponding to the monitoring period to obtain the excess duration ratio.

[0020] As a further aspect of the present invention: the process of obtaining the excess ratio is as follows:

[0021] In the current change curve, the area exceeding the rated current within the measured monitoring interval is summed, and the ratio of the summation of the areas exceeding the rated current in all monitoring intervals during the monitoring period is calculated with the standard area to obtain the excess current ratio.

[0022] As a further aspect of the present invention: the specific process for determining whether the predicted current shows a decreasing trend during the prediction period is as follows:

[0023] The prediction period is divided into several prediction time points with equal time intervals; based on the trained current prediction model, the predicted current at all prediction time points within the prediction period is output; the predicted current at all prediction time points within the prediction period is integrated into a prediction current time series according to the time order.

[0024] The least squares method is used to perform linear fitting on the predicted current time series to determine the rate of change of the predicted current; if the rate of change of the predicted current is less than 0, it indicates that the predicted current is decreasing during the prediction period.

[0025] As a further aspect of the present invention: the specific process for determining the specific recovery time point is as follows:

[0026] Extract the predicted current from the predicted current time series. If the predicted current is greater than the rated current, the predicted current is recorded as the over-limit current; otherwise, the predicted current is recorded as the normal current.

[0027] All normal currents within the prediction period are integrated into a normal current sequence according to time order. The prediction time point corresponding to the first normal current in the normal current sequence is recorded as the approximate recovery time point. Based on the approximate recovery time point, the specific recovery time point is calculated using the linear interpolation method.

[0028] As a further aspect of the present invention: the specific process for assessing the degree of abnormality in the load connection status of the energy storage power source is as follows:

[0029] A comprehensive analysis of all current data between the start time and the specific recovery time is conducted to determine the comprehensive excess duration ratio and the comprehensive excess degree ratio.

[0030] The comprehensive transient assessment index is obtained by multiplying the comprehensive excess duration ratio and the comprehensive excess degree ratio.

[0031] If the comprehensive transient assessment index is less than the comprehensive transient assessment index threshold, it indicates that the degree of abnormal current change after the energy storage power supply load is connected is low.

[0032] As a further aspect of the present invention: the process for obtaining the comprehensive excess time ratio and the comprehensive excess degree ratio is as follows:

[0033] The time between the start time and the specific recovery time is recorded as the comprehensive excess time. The comprehensive excess time is then compared with the time corresponding to the monitoring interval to obtain the comprehensive excess time ratio.

[0034] Construct a predicted current change curve within the prediction period, measure the area exceeding the rated current during the prediction period, sum the area with the corresponding excess area during the monitoring period to obtain the comprehensive excess area, and then compare it with the standard area to obtain the comprehensive excess degree ratio.

[0035] As a further aspect of the present invention: the specific process of adjusting the power is as follows:

[0036] All overcurrents within the predicted period are integrated into an overcurrent sequence in chronological order. The first overcurrent in the overcurrent sequence is extracted and its difference from the rated current is calculated to obtain the excess current. The excess current is then compared with the rated current, and the square of the ratio is used to calculate the power regulation ratio. The power is then adjusted according to the power regulation ratio.

[0037] Secondly, the present invention also provides an energy storage power supply load connection status detection system, the system comprising:

[0038] Preliminary assessment module: Real-time monitoring of the current after the energy storage power supply load is connected; By analyzing the current change characteristics during the monitoring period, a preliminary assessment of the current change after the energy storage power supply load is connected is performed.

[0039] Trend Analysis Module: If the current change after the energy storage power supply load is connected is assessed as abnormal, a current prediction model is constructed. All current data within the monitoring period are used as input, the prediction period is preset, the predicted current within the prediction period is output, the rate of change of the predicted current is analyzed, and it is determined whether the predicted current shows a downward trend within the prediction period.

[0040] Anomaly Assessment Module: If the predicted current shows a downward trend during the prediction period, the linear interpolation method is used to determine the specific recovery time point when the current recovers to the rated current. The time point when the current first exceeds the rated current during the monitoring period is recorded as the start time point. All current data between the start time point and the specific recovery time point are comprehensively analyzed to assess the degree of anomaly in the load connection status of the energy storage power supply.

[0041] Control and Analysis Module: If the load connection status of the energy storage power supply is low-level abnormal, pulse width modulation is activated to adjust the power without interrupting the load, thereby achieving current regulation and shortening the current abnormality time.

[0042] The beneficial effects of this invention are as follows: Real-time monitoring of the current after the energy storage power supply load is connected; analysis of the current change characteristics during the monitoring period to preliminarily assess the current change after the energy storage power supply load is connected; if the current change after the energy storage power supply load is connected is assessed as abnormal, a current prediction model is constructed, using all current data during the monitoring period as input, a preset prediction period, and the output of the predicted current during the prediction period. The rate of change of the predicted current is analyzed to determine whether the predicted current shows a downward trend during the prediction period; if the current shows a downward trend after reaching a certain time point, linear interpolation is used to determine the specific recovery time point when the current recovers to the rated current. The time point when the monitoring period first exceeds the rated current is recorded as the start time point, and the start time point is compared with the specific recovery time... The invention comprehensively analyzes all current data between points to assess the degree of abnormality in the energy storage power supply load connection status. If the abnormality level of the energy storage power supply load connection status is low, pulse width modulation is initiated to adjust the power without interrupting the load, thereby regulating the current and shortening the current abnormality time. This invention distinguishes between normal transient fluctuations and abnormal current changes when the energy storage power supply load is connected by constructing a transient current excess index. By determining the current recovery time point and integrating real-time and predicted data to form a comprehensive transient assessment index, hidden faults such as increased battery internal resistance and component aging can be identified in advance, improving the sensitivity of anomaly identification. For scenarios with low abnormality levels, pulse width modulation is used to dynamically adjust the output power, shortening the current abnormality time without interrupting power supply. Attached Figure Description

[0043] The invention will now be further described with reference to the accompanying drawings.

[0044] Figure 1 This is a flowchart illustrating the steps of a method for detecting the load connection status of an energy storage power supply according to an embodiment of the present invention.

[0045] Figure 2 This is a system block diagram of an energy storage power supply load access status detection system according to an embodiment of the present invention. Detailed Implementation

[0046] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0047] Example 1

[0048] Please see Figure 1 As shown in the figure, the method for detecting the load connection status of an energy storage power source according to an embodiment of the present invention includes the following steps:

[0049] Step 1: Monitor the current after the energy storage power supply load is connected in real time. By analyzing the current change characteristics during the monitoring period, make a preliminary assessment of the current change after the energy storage power supply load is connected.

[0050] When the energy storage power supply is connected to the load, the current usually exceeds the rated current for a short period of time. This is a normal phenomenon caused by the load characteristics and circuit transient process. Therefore, a monitoring period is preset to analyze the current change characteristics during the monitoring period.

[0051] Specifically, the monitoring period is divided into several monitoring intervals with equal time intervals. During the monitoring period, the current when the energy storage power supply load is connected is obtained in real time through the Hall current sensor. The current data is preprocessed, including noise reduction using the moving average filtering method and normalization processing of the current data, mapping the current data to the [0, 1] interval.

[0052] Based on the preprocessed current data, a current change curve is constructed, and the intervals within the monitoring range that exceed the rated current are extracted and recorded as the excess intervals.

[0053] The duration corresponding to the excess interval within the statistical monitoring period is recorded as the excess duration. The excess duration is then compared with the duration corresponding to the monitoring period to obtain the excess duration ratio.

[0054] The area exceeding the rated current within the monitoring interval is measured. The areas exceeding the rated current in all monitoring intervals during the monitoring period are summed to obtain the excess area. The excess area is then compared with the standard area to obtain the excess degree ratio.

[0055] It is understandable that the standard area can be set by professionals in the field based on historical experience;

[0056] The transient current excess index is obtained by multiplying the excess duration ratio and the excess degree ratio.

[0057] In some embodiments, the transient current excess index is compared with the transient current excess index threshold. The specific comparison process is as follows:

[0058] If the transient current excess index is greater than or equal to the transient current excess index threshold, it indicates that the current change is abnormal after the energy storage power supply load is connected.

[0059] If the transient current excess index is less than the transient current excess index threshold, it indicates that the current change is normal after the energy storage power supply load is connected.

[0060] The transient current excess index serves the following purposes: In energy storage power supply load access scenarios, the transient current excess index is obtained by calculating the excess duration ratio and the excess degree ratio, and multiplying the two. On the one hand, it uses a dual-dimensional quantitative model of time and amplitude to distinguish whether the current fluctuation caused by the transient process during load access is a normal current fluctuation; on the other hand, its quantitative characteristics can detect hidden anomalies such as increased battery internal resistance and component aging in advance, and improve the sensitivity of anomaly identification by combining duration and energy accumulation effects, providing a quantitative early warning basis for the safe operation of the system.

[0061] Step 2: If the current change after the energy storage power supply load is connected is assessed as abnormal, a current prediction model is constructed. All current data within the monitoring period are used as input, the prediction period is preset, the predicted current within the prediction period is output, and the rate of change of the predicted current is analyzed to determine whether the predicted current shows a downward trend within the prediction period.

[0062] Based on a machine learning model, the time period in a historical cycle during which the current exceeds the rated current and reaches the peak current after the energy storage power supply load is connected is denoted as the peak period, and the time period in a historical cycle during which the current returns to the rated current after reaching the peak current is denoted as the peak elimination period. All current data between the peak period and the peak elimination period in multiple historical cycles are extracted, and the training set and test set are divided in an 8:2 ratio. The training set is used for learning the parameters of the convolutional neural network model, and the test set is used to verify the generalization ability. The mean squared error is used as the loss function, and Adam is used as the optimizer to build a current prediction model.

[0063] It should be noted that the peak current is the maximum current after the load is connected within the historical period.

[0064] The peak reduction duration is obtained by summing and averaging the durations of peak reduction periods within multiple historical periods.

[0065] The preset prediction period is longer than the peak elimination period. The purpose is to ensure that the current data predicted by the current prediction model includes the current data that recovers to the rated current.

[0066] Extract all current data within the monitoring period, divide the prediction period into several prediction time points with equal time intervals; input all current data within the monitoring period into the trained current prediction model, and output the predicted current at all prediction time points within the prediction period.

[0067] The predicted currents at all predicted time points within the prediction period are integrated into a predicted current time series in chronological order.

[0068] The least squares method is used to linearly fit the predicted current data in the predicted current time series to determine the fitted trend line y = kt + b, where the slope k is the rate of change of the predicted current, t represents time, and b represents the intercept of the fitted linear trend line on the y-axis.

[0069] Compare the predicted rate of change of current k with 0:

[0070] If the rate of change k of the predicted current is less than 0, it indicates that the predicted current shows a downward trend during the prediction period.

[0071] If the rate of change of the predicted current k is greater than or equal to 0, it means that the predicted current does not show a decreasing trend during the prediction period. The load should be disconnected immediately and the energy storage power supply should be inspected.

[0072] Step 3: If the predicted current shows a downward trend during the prediction period, use linear interpolation to determine the specific recovery time point when the current recovers to the rated current. Record the time point when the current first exceeds the rated current during the monitoring period as the start time point. Perform comprehensive analysis on all current data between the start time point and the specific recovery time point to assess the degree of abnormality in the load connection status of the energy storage power supply.

[0073] All current data includes real-time monitored current data and predicted current data. The predicted current in the predicted current time series is compared with the rated current.

[0074] If the predicted current is greater than the rated current, the predicted current is recorded as the over-limit current.

[0075] If the predicted current is less than or equal to the rated current, the predicted current is recorded as the normal current.

[0076] All normal currents within the prediction period are integrated into a normal current sequence according to time order. The prediction time point corresponding to the first normal current in the normal current sequence is recorded as the approximate recovery time point.

[0077] If the previous prediction time point on the timeline The predicted current is the over-limit current and the next prediction time point The predicted current is the normal current, indicating that the current is within the normal range. The predicted time interval is from the over-limit current down to below the rated current. This is the approximate recovery time point. To improve time accuracy, linear interpolation is used to calculate the specific recovery time point. ;

[0078] It is understandable that the time interval between two adjacent prediction time points is a prediction time interval;

[0079] Specifically, the linear relationship of current change within the predicted time interval is calculated: Solve In the formula, This represents the (h-1)th prediction time point. This represents the h-th prediction time point. Indicates the rated current. This represents the over-limit current corresponding to the (h-1)th prediction time point. This represents the normal current corresponding to the h-th prediction time point;

[0080] The time between the start time and the specific recovery time is recorded as the comprehensive excess time. The comprehensive excess time is then compared with the time corresponding to the monitoring interval to obtain the comprehensive excess time ratio.

[0081] Construct a predicted current change curve within the prediction period, measure the area exceeding the rated current during the prediction period to obtain the predicted excess area, sum the predicted excess area with the corresponding excess area within the monitoring period to obtain the comprehensive excess area, and ratio the comprehensive excess area with the standard area to obtain the comprehensive excess degree ratio.

[0082] The comprehensive transient assessment index is obtained by multiplying the comprehensive excess duration ratio and the comprehensive excess degree ratio.

[0083] In some embodiments, the comprehensive transient assessment index is compared with the comprehensive transient assessment index threshold. The specific comparison process is as follows:

[0084] If the comprehensive transient assessment index is greater than or equal to the comprehensive transient assessment index threshold, it indicates that the current change is seriously abnormal after the energy storage power supply load is connected. The load should be disconnected immediately and the energy storage power supply should be inspected.

[0085] If the comprehensive transient assessment index is less than the comprehensive transient assessment index threshold, it indicates that the degree of abnormal current change after the energy storage power supply load is connected is low.

[0086] The comprehensive transient assessment index serves the following purposes: it considers both the duration of current exceeding the rated value and the proportion of monitoring time, and incorporates the energy accumulation effect of excess current (quantified by the ratio of excess area to standard area). It also integrates real-time monitoring data with predicted data, enabling dynamic tracking of the current transient process. Based on normal transient fluctuations and abnormal current changes during load connection, it proactively detects hidden faults such as increased battery internal resistance and component aging through trend analysis during the predicted time period. Combined with precise calculation of recovery time points, it effectively improves the sensitivity and reliability of energy storage power supply load connection status assessment, providing quantitative early warning support for safe equipment operation.

[0087] Step 4: If the load connection status of the energy storage power supply is low-level abnormal, start pulse width modulation to adjust the power without interrupting the load, thereby achieving current regulation and shortening the current abnormality time.

[0088] All overcurrents within the prediction period are integrated into an overcurrent sequence in chronological order. The first overcurrent in the sequence is extracted and its difference from the rated current is calculated to obtain the excess current. According to Ohm's law, the current of a resistive load is proportional to the voltage. When the average voltage is modulated by pulse width modulation, the current also changes linearly. Therefore, under a resistive load, the power is proportional to the square of the current. The excess current is compared with the rated current, and the square of the ratio is used to calculate the power regulation ratio.

[0089] While continuing to monitor the current, the power is adjusted according to the power adjustment ratio, thereby achieving current regulation;

[0090] The rationale for pulse width modulation in scenarios with low current anomaly levels: The current over-limit level is low and has not reached an emergency state that threatens the safety of the equipment; the cause of the over-limit may be transient process fluctuations (such as load start-up impact) or slight parameter deviations (such as a slight increase in battery internal resistance), rather than a serious fault. Pulse width modulation reduces the current to the rated value by adjusting the output power pulse width, thereby shortening the current anomaly time and avoiding power outages caused by directly disconnecting the load.

[0091] The technical solution of this embodiment is as follows: Real-time monitoring of the current after the energy storage power supply load is connected; analysis of the current change characteristics during the monitoring period to preliminarily assess the current change after the energy storage power supply load is connected; if the current change after the energy storage power supply load is connected is assessed as abnormal, a current prediction model is constructed, using all current data during the monitoring period as input, a preset prediction period, and the output predicted current during the prediction period. A rate of change analysis is performed on the predicted current to determine whether the predicted current shows a downward trend during the prediction period; if the current shows a downward trend after reaching the specified time point, linear interpolation is used to determine the specific recovery time point when the current returns to the rated current. The time point when the monitoring period first exceeds the rated current is recorded as the start time point, and the start time point is compared with the specific recovery time... The invention comprehensively analyzes all current data between points to assess the degree of abnormality in the energy storage power supply load connection status. If the abnormality level of the energy storage power supply load connection status is low, pulse width modulation is initiated to adjust the power without interrupting the load, thereby regulating the current and shortening the current abnormality time. This invention distinguishes between normal transient fluctuations and abnormal current changes when the energy storage power supply load is connected by constructing a transient current excess index. By determining the current recovery time point and integrating real-time and predicted data to form a comprehensive transient assessment index, hidden faults such as increased battery internal resistance and component aging can be identified in advance, improving the sensitivity of anomaly identification. For scenarios with low abnormality levels, pulse width modulation is used to dynamically adjust the output power, shortening the current abnormality time without interrupting power supply.

[0092] Example 2

[0093] Please see Figure 2 As shown in the embodiment of the present invention, an energy storage power supply load access status detection system includes the following modules:

[0094] Preliminary assessment module: Real-time monitoring of the current after the energy storage power supply load is connected; By analyzing the current change characteristics during the monitoring period, a preliminary assessment of the current change after the energy storage power supply load is connected is performed.

[0095] Trend Analysis Module: If the current change after the energy storage power supply load is connected is assessed as abnormal, a current prediction model is constructed. All current data within the monitoring period are used as input, the prediction period is preset, the predicted current within the prediction period is output, the rate of change of the predicted current is analyzed, and it is determined whether the predicted current shows a downward trend within the prediction period.

[0096] Anomaly Assessment Module: If the current shows a downward trend after reaching the time point, the specific recovery time point of the current returning to the rated current is determined by linear interpolation. The time point when the current first exceeds the rated current during the monitoring period is recorded as the start time point. All current data between the start time point and the specific recovery time point are comprehensively analyzed to assess the degree of anomaly in the load connection status of the energy storage power supply.

[0097] Control and Analysis Module: If the load connection status of the energy storage power supply is low-level abnormal, pulse width modulation is activated to adjust the power without interrupting the load, thereby achieving current regulation and shortening the current abnormality time.

[0098] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for detecting the state of a load connected to an energy storage power supply, characterized by: The application relates to a method for evaluating the abnormality of current change after a load of an energy storage power supply is connected, and the method comprises the following steps: real-time monitoring of the current after the load of the energy storage power supply is connected, preliminary evaluation of the current change after the load of the energy storage power supply is connected by analyzing the current change characteristics in the monitoring period; if the evaluation of the current change after the load of the energy storage power supply is connected is abnormal, a current prediction model is constructed, all current data in the monitoring period is taken as input, a preset prediction period is output, the change trend of the predicted current in the prediction period is analyzed, and whether the predicted current in the prediction period shows a downward trend is judged; the process of constructing the current prediction model is as follows: based on a machine learning model, a time period in which the current exceeds the rated current to reach a current peak value after the load of the energy storage power supply is connected in a historical period is recorded as a peak reaching time period, a time period in which the current returns to the rated current after reaching the current peak value in the historical period is recorded as a peak eliminating time period, all current data between the peak reaching time period and the peak eliminating time period in multiple historical periods is extracted, a training set and a test set are divided according to a 8:2 ratio, the training set is used for convolutional neural network model parameter learning, the test set verifies the generalization ability, a mean square error is taken as a loss function, Adam is taken as an optimizer, and the current prediction model is constructed; if the predicted current in the prediction period shows a downward trend, the specific recovery time point of the current returning to the rated current is determined by using a linear interpolation method, a time point when the current first exceeds the rated current in the monitoring period is recorded as a starting time point, and all current data between the starting time point and the specific recovery time point is comprehensively analyzed to evaluate the abnormality degree of the load connection state of the energy storage power supply; the specific process of evaluating the abnormality degree of the load connection state of the energy storage power supply is as follows: all current data between the starting time point and the specific recovery time point is comprehensively analyzed to determine a comprehensive excess time length ratio and a comprehensive excess degree ratio; the process of obtaining the comprehensive excess time length ratio and the comprehensive excess degree ratio is as follows: the time length between the starting time point and the specific recovery time point is recorded as a comprehensive excess time length, the comprehensive excess time length is subjected to ratio processing with the time length corresponding to the monitoring interval, and a comprehensive excess time length ratio is obtained; a prediction current change curve in the prediction period is constructed, the area exceeding the rated current in the prediction period is measured, and the area exceeding the rated current in the monitoring period is summed to obtain a comprehensive excess area, and then the comprehensive excess area is subjected to ratio processing with a standard area to obtain a comprehensive excess degree ratio; the comprehensive excess time length ratio and the comprehensive excess degree ratio are subjected to product processing to obtain a comprehensive transient state evaluation index; if the comprehensive transient state evaluation index is smaller than a comprehensive transient state evaluation index threshold value, the abnormality degree of the current change after the load of the energy storage power supply is connected is low; if the load connection state of the energy storage power supply is low in abnormality degree, pulse width modulation is started, the power is adjusted without interrupting the load, current adjustment is realized, and the current abnormality time is shortened.

2. The method of claim 1, wherein: the specific process of preliminarily evaluating the current change after the load of the energy storage power supply is connected is as follows: the monitoring period is divided into a plurality of time intervals with equal monitoring intervals, the current when the load of the energy storage power supply is connected is acquired in real time in the monitoring period by a Hall current sensor, a current change curve is constructed, and an excess time length ratio and an excess degree ratio are determined; the excess time length ratio is obtained in the following manner: Extract the interval segment exceeding the rated current in the monitoring interval, denoted as an excessive interval; Statistical monitoring duration corresponding to the excessive interval, denoted as an excessive duration, and the excessive duration is processed by ratio with the duration corresponding to the monitoring period to obtain an excessive duration ratio; The process of obtaining the excessive degree ratio is: In the current change curve, measure the area exceeding the rated current in the monitoring interval, sum the areas exceeding the rated current in all monitoring intervals in the monitoring period, and process by ratio with the standard area to obtain the excessive degree ratio; The excessive duration ratio and the excessive degree ratio are processed by product to obtain the transient current excessive index; If the transient current excessive index is greater than or equal to the transient current excessive index threshold, it indicates that the current change after the energy storage power load is connected is abnormal.

3. The method of claim 1, wherein: The specific process of judging whether the predicted current presents a downward trend in the prediction period is: Divide the prediction period into several prediction time points with equal time intervals; according to the trained current prediction model, output the predicted current at all prediction time points in the prediction period; integrate the predicted current at all prediction time points in the prediction period in time sequence as the predicted current time sequence; Use the least square method to linearly fit the predicted current time sequence to determine the change rate of the predicted current; if the change rate of the predicted current is less than 0, it indicates that the predicted current presents a downward trend in the prediction period.

4. The method of claim 3, wherein: The specific determination process of the specific recovery time point is: Extract the predicted current in the predicted current time sequence, if the predicted current is greater than the rated current, the predicted current is recorded as an over-limit current; otherwise, the predicted current is recorded as a normal current; Integrate all normal currents in the prediction period in time sequence as a normal current sequence, record the prediction time point corresponding to the normal current located in the first sequence in the normal current sequence as an approximate recovery time point, and calculate the specific recovery time point by linear interpolation method according to the approximate recovery time point.

5. The method of claim 4, wherein: The specific process of adjusting the power is: Integrate all over-limit currents in the prediction period in time sequence as an over-limit current sequence, extract the over-limit current located in the first sequence in the over-limit current sequence, and process by difference with the rated current to obtain an excessive current, process the excessive current by ratio with the rated current, and then take the square of the ratio processing result to calculate the power adjustment ratio, and adjust the power according to the power adjustment ratio.

6. An energy storage power supply load access status detection system, characterized by, The system is used to execute the method of any one of claims 1-5, and the system comprises: A preliminary evaluation module: real-time monitoring of the current after the energy storage power load is connected, and preliminary evaluation of the current change after the energy storage power load is connected by analyzing the current change characteristics in the monitoring period; A trend analysis module: if the current change after the energy storage power load is connected is evaluated as abnormal, a current prediction model is constructed, all current data in the monitoring period are taken as input, a prediction period is preset, the predicted current in the prediction period is output, the change rate of the predicted current is analyzed, and whether the predicted current presents a downward trend in the prediction period is judged; The process of constructing the current prediction model is: Based on the machine learning model, the time period in which the load of the energy storage power supply exceeds the rated current to reach the current peak value after the load is connected in the historical period is recorded as the peak reaching time period, the time period in which the current returns to the rated current after reaching the current peak value in the historical period is recorded as the peak eliminating time period, all current data between the peak reaching time period and the peak eliminating time period in multiple historical periods are extracted, a training set and a test set are divided according to a ratio of 8:2, the training set is used for parameter learning of a convolutional neural network model, the test set is used for verifying the generalization ability, a mean square error is used as a loss function, Adam is used as an optimizer, and a current prediction model is constructed; The abnormality evaluation module: if the predicted current shows a downward trend in the prediction period, the linear interpolation method is used to determine the specific recovery time point at which the current returns to the rated current, the time point at which the current first exceeds the rated current in the monitoring period is recorded as the starting time point, and all current data between the starting time point and the specific recovery time point are comprehensively analyzed to evaluate the abnormality degree of the load connection state of the energy storage power supply; The specific process of evaluating the abnormality degree of the load connection state of the energy storage power supply is as follows: The comprehensive excess duration ratio and the comprehensive excess degree ratio are determined by comprehensively analyzing all current data between the starting time point and the specific recovery time point. The comprehensive excess duration ratio and the comprehensive excess degree ratio are obtained as follows: The time duration between the starting time point and the specific recovery time point is recorded as the comprehensive excess duration, the comprehensive excess duration is processed by ratio with the duration corresponding to the monitoring interval to obtain the comprehensive excess duration ratio; A predicted current change curve in the prediction period is constructed, the area exceeding the rated current in the prediction period is measured, and the area is summed with the corresponding excess area in the monitoring period to obtain the comprehensive excess area, and then the comprehensive excess area is processed by ratio with the standard area to obtain the comprehensive excess degree ratio; The comprehensive transient state evaluation index is obtained by multiplying the comprehensive excess duration ratio and the comprehensive excess degree ratio; If the comprehensive transient state evaluation index is less than the comprehensive transient state evaluation index threshold, it indicates that the abnormality degree of the current change after the load of the energy storage power supply is connected is low. The control analysis module: if the load connection state of the energy storage power supply is low in abnormality degree, pulse width modulation is started, the power is adjusted without interrupting the load, current adjustment is realized, and the current abnormality time is shortened.

Images