A fault recording method for a direct current power supply and distribution circuit

Abstract

The present application relates to the technical field of direct current power supply and distribution system monitoring, and particularly relates to a fault recording method for a direct current power supply and distribution circuit, comprising: S1: collecting a current waveform of a load of the direct current power supply and distribution circuit, and identifying a load type by using a current waveform feature; S2: detecting a predictive fault recording trigger condition, and performing S5 if the condition is met; otherwise, performing S3; S3: detecting a working condition joint fault recording trigger condition, and performing S5 if the condition is met; otherwise, performing S4; S4: detecting a multi-parameter associated fault recording trigger condition, and performing S5 if the condition is met; otherwise, repeatedly performing S1-S4; S5: performing adaptive fault recording in combination with the load type identified in S1. The present application solves the technical problems that the existing fault recording technology for the direct current power supply and distribution circuit cannot adapt to the characteristics of the direct current power supply and distribution system, the diversity of loads, and provides differentiated recording strategies and early warning, etc.

Description

Technical Field

[0001] This invention relates to the field of DC power supply and distribution system monitoring technology, and more specifically, to a fault recording method for DC power supply and distribution circuits. Background Technology

[0002] DC power supply and distribution systems are widely used in data centers, communication base stations, industrial automation, new energy power generation, electric vehicle charging, and other fields, providing stable and reliable power for various DC loads such as servers, PLC controllers, sensors, DC motors, battery packs, and LED lighting. With the rapid development of DC microgrids and distributed energy, the scale of DC power supply and distribution systems is expanding and the types of loads are becoming increasingly diverse. At the same time, the requirements for system safety and reliability are also becoming increasingly stringent.

[0003] Fault recording technology, as an important means of power system fault analysis, recording, and diagnosis, provides data support for fault location, cause analysis, and protection action evaluation by accurately recording electrical parameters before and after a fault occurs. However, existing fault recording technologies are mainly designed for AC distribution networks and high-voltage direct current (HVDC) transmission systems, and are difficult to directly apply to low-voltage DC power supply and distribution scenarios. Specifically, existing technologies have the following shortcomings: First, existing fault recording methods fail to fully consider the diversity of DC loads. The load types in AC distribution networks are relatively simple, mainly lines, transformers, and motors, and their fault characteristics are relatively uniform. However, DC power supply and distribution systems have extremely diverse load types, including motors, batteries, electronic equipment, lighting, etc., and the electrical characteristics, fault modes, and characteristic signals of different loads vary greatly. If a uniform fault recording strategy is adopted, such as a fixed sampling rate, a uniform trigger threshold, and the same recording duration, it is difficult to effectively capture the characteristic fault signals of various loads, resulting in poor specificity of the recorded data and low value for fault analysis. Second, existing fault recording methods cannot adapt to the rapid development of faults in DC systems. AC system currents have a natural zero-crossing point every half cycle, providing conditions for arc extinction and fault detection. However, DC system fault currents lack a natural zero-crossing point; once a fault occurs, the arc continues to burn, easily causing fires and equipment damage. Furthermore, the time window from fault occurrence to loss of control in DC systems is much shorter than in AC systems, necessitating the capture of early signs before a fault occurs. Existing fault recording technologies mostly employ threshold triggering, meaning recording only begins after parameters exceed an abnormal threshold. By this time, the critical early stages of fault development are often missed, making it impossible to fully record the entire fault evolution process and failing to meet the active safety protection requirements of DC systems.

[0004] In summary, existing technologies lack a fault recording method for DC power supply and distribution systems that can adapt to diverse loads, provide early warnings, and offer differentiated recording strategies. Summary of the Invention

[0005] The purpose of this application is to provide a fault recording method for DC power supply and distribution circuits, which solves the technical problems existing in the fault recording technology of DC power supply and distribution circuits, such as the inability to adapt to the characteristics of DC power supply and distribution systems and the diversity of loads, and provides differentiated recording strategies and early warning.

[0006] To solve the above-mentioned technical problems, the solution adopted in this application is as follows:

[0007] A fault recording method for a DC power supply and distribution circuit, characterized by the following steps:

[0008] S1: Collect the current waveform of the load in the DC power supply and distribution circuit, analyze the characteristics of the current waveform, and use the characteristics of the current waveform to identify the load type;

[0009] S2: Perform a predictive fault recording trigger condition check. If the predictive fault recording trigger condition is met, proceed to S5; otherwise, proceed to S3.

[0010] S3: Perform a joint fault recording trigger condition detection. If the joint fault recording trigger condition is met, proceed to S5; otherwise, proceed to S4.

[0011] S4: Perform multi-parameter associated fault recording trigger condition detection. If the multi-parameter associated fault recording trigger condition is met, execute S5; if the multi-parameter associated fault recording trigger condition is not met, repeat S1-S4.

[0012] S5: Combine the load type identified in S1 to perform adaptive fault recording.

[0013] Preferably, the specific implementation method of S1 includes the following steps:

[0014] S1.1: Collect current signal data of each load in the DC power supply and distribution circuit, and store the collected current signal data to form continuous current waveform data;

[0015] S1.2: Preprocessing of current waveform data;

[0016] S1.3: Extract multidimensional feature parameters from the preprocessed current waveform data;

[0017] S1.4: Load type identification based on multi-dimensional feature parameters.

[0018] Preferably, the preprocessing of the current waveform in S1.2 includes the following steps:

[0019] S1.2.1: Filter and denoise the current waveform data;

[0020] S1.2.2: Perform zero-point calibration on the current waveform data after filtering and denoising;

[0021] S1.2.3: Perform waveform segmentation processing on continuous current waveform data.

[0022] Preferably, the predictive fault recording trigger condition detection is performed in step S2, and the specific implementation method includes the following steps:

[0023] S2.1: Establish a mapping table between each load type and the predictive fault recording trigger condition, and set the threshold for the predictive fault recording trigger condition;

[0024] S2.2: Monitor the rate of change exceeding the trigger condition based on the set predictive fault recording trigger condition threshold;

[0025] S2.3: Monitor long-term drift trigger conditions based on the set predictive fault recording trigger condition threshold.

[0026] Preferably, the monitoring of the rate of change exceeding the trigger condition based on the set predictive fault recording trigger condition threshold in S2.2 includes the following steps:

[0027] S2.2.1: The instantaneous current change rate is calculated using the backward difference method. When the instantaneous current change rate exceeds the instantaneous current change rate threshold, it is determined that the trigger condition of the change rate exceeding the limit is met.

[0028] S2.2.2: The instantaneous voltage change rate is calculated using backward differential distribution. When the instantaneous voltage change rate exceeds the instantaneous voltage change rate threshold, it is determined that the trigger condition of the change rate exceeding the limit is met.

[0029] Preferably, the long-term drift trigger condition monitoring based on the set predictive fault recording trigger condition threshold in S2.3 includes the following steps:

[0030] S2.3.1: The trend of electrical parameters within a time window is fitted using linear regression to obtain the drift rate and total drift of the electrical parameters;

[0031] S2.3.2: Determine whether the drift rate and total drift of the electrical parameters exceed the drift rate threshold and the total drift threshold, respectively. If they do, the long-term drift trigger condition is met.

[0032] Preferably, the specific implementation method of S3 includes the following steps:

[0033] S3.1: Detect the start-up event of the load, collect the start-up electrical parameters of the start-up event, and determine whether the start-up process meets the abnormal triggering conditions based on the characteristics of the start-up electrical parameters.

[0034] S3.2: Detect shutdown events for the load and determine whether the abnormal triggering conditions of the shutdown process are met;

[0035] S3.3: Detect load mutation events and determine whether they meet the abnormal triggering conditions for load mutation.

[0036] Preferably, the specific implementation method of S3.1 includes the following steps:

[0037] S3.1.1: Obtain the normal startup characteristics of each load type and establish a standard library of normal startup characteristics;

[0038] S3.1.2: Extract the electrical parameter features of the startup event and compare them with the normal startup feature standard library. If a startup electrical parameter feature deviates from the electrical parameter feature in the normal startup feature standard library, it is determined that it meets the abnormal triggering conditions of the startup process.

[0039] Preferably, the multi-parameter correlation fault recording trigger condition detection in S4 is specifically implemented by the following steps:

[0040] S4.1: Collect voltage and current signals of the load in the DC power supply and distribution circuit, establish a voltage-current physical correlation model, and perform voltage-current correlation anomaly trigger detection on the load based on the voltage-current physical correlation model. When the preset voltage-current correlation trigger condition is met, it is determined that the multi-parameter correlation fault recording trigger condition is met.

[0041] S4.2: Collect the current and temperature signals of the load circuit, establish a current-temperature thermal effect correlation model, and perform current-temperature correlation anomaly trigger detection on the load based on the current-temperature thermal effect correlation model. When the preset current-temperature correlation trigger conditions are met, it is determined that the multi-parameter correlation fault recording trigger conditions are met.

[0042] Preferably, in step S5, adaptive fault recording is performed in conjunction with the load type identified in step S1. This is specifically implemented by the following steps:

[0043] S5.1: Establish a mapping table between load type and waveform recording parameter set;

[0044] S5.2: Based on the specific type of triggering condition, dynamically optimize the waveform recording parameter set in S5.1 to generate the final waveform recording parameter set;

[0045] S5.3: Perform fault recording based on the adjusted recording parameter set.

[0046] The technical solution of this application has at least the following advantages and beneficial effects:

[0047] 1. This invention discloses a fault recording method for DC power supply and distribution circuits. The method identifies the load type by acquiring the current waveform of the load in the DC power supply and distribution circuit and analyzing its waveform characteristics. It then performs a three-level progressive fault recording trigger condition detection for the power supply and distribution circuit: predictive fault recording trigger condition, combined operating condition fault recording trigger condition, and multi-parameter correlated fault recording trigger condition. If any fault recording trigger condition is met, adaptive fault recording is performed according to the load type. This method provides different fault recording methods for different load types. Compared to using the same fault recording method for all load types, the fault recording method of this invention obtains more targeted data, better reflects the fault characteristics of different load types, and is more helpful for subsequent fault analysis. Effects: This invention also sets up a three-level progressive fault recording trigger condition, which has a wider detection range for possible fault situations. Furthermore, through predictive fault recording condition detection, it can analyze the rate of change of load parameters in advance and determine whether the load is faulty and trigger fault recording based on the pattern of the load parameter change rate. Compared with the traditional method of recording faults only after all parameters have deviated from normal values, the method of this invention can record data before and after the fault more completely, ensuring the reliability of data in subsequent fault analysis. In addition, DC power distribution systems have diverse load types, typically including motors, batteries, and electronic equipment. The electrical characteristics and fault modules of different loads vary greatly. If a uniform fault recording strategy is used, it is difficult to effectively capture the characteristic fault signals of various loads. DC power distribution system fault currents do not have a natural zero-crossing point, and the time window from fault occurrence to loss of control in DC power distribution systems is much shorter than that of AC systems. Once a fault occurs, the arc continues to burn, which can easily cause fires and equipment damage. Therefore, it is necessary to perform predictive fault recording trigger condition detection before the fault occurs, identifying abnormal trends before load parameters exceed abnormal thresholds, and realizing predictive fault recording. Attached Figure Description

[0048] Figure 1 This is a flowchart of the fault recording method of the present invention. Detailed Implementation

[0049] 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, and 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.

[0050] This invention discloses a fault recording method for DC power supply and distribution circuits. The method identifies the load type by acquiring and analyzing the current waveforms of the load in the DC power supply and distribution circuit. It then performs a three-level progressive fault recording trigger condition detection: predictive fault recording trigger condition, combined operating condition fault recording trigger condition, and multi-parameter correlated fault recording trigger condition. If any fault recording trigger condition is met, adaptive fault recording is performed based on the load type. This method provides different fault recording methods for different load types. Compared to using the same fault recording method for all load types, the fault recording method of this invention yields more targeted data that better reflects the fault characteristics of different load types, making subsequent fault analysis more effective. Furthermore, this invention also establishes a three-level progressive fault recording trigger condition, which has a wider detection range for possible fault situations. Through predictive fault recording condition detection, it can analyze the rate of change of load parameters in advance and determine whether the load is faulty and trigger fault recording based on the pattern of the load parameter change rate. Compared to the traditional method of recording faults only after all parameters have deviated from normal values, this invention can record data before and after the fault more completely, ensuring the reliability of data in subsequent fault analysis. In addition, DC power distribution systems have diverse load types, typically including motors, batteries, and electronic equipment. The electrical characteristics and fault modules of different loads vary greatly. If a uniform fault recording strategy is used, it is difficult to effectively capture the characteristic fault signals of various loads. DC power distribution system fault currents do not have a natural zero-crossing point, and the time window from fault occurrence to loss of control in DC power distribution systems is much shorter than that of AC systems. Once a fault occurs, the arc continues to burn, which can easily cause fires and equipment damage. Therefore, it is necessary to perform predictive fault recording trigger condition detection before the fault occurs, identifying abnormal trends before load parameters exceed abnormal thresholds, and realizing predictive fault recording.

[0051] Referring to the figure, which is a flowchart of a fault recording method for a DC power supply and distribution circuit according to the present invention, its specific implementation includes the following steps:

[0052] S1: Collect the current waveform of the load in the DC power supply and distribution circuit, analyze the characteristics of the current waveform, and use the characteristics of the current waveform to identify the load type;

[0053] S2: Perform a predictive fault recording trigger condition check. If the predictive fault recording trigger condition is met, proceed to S5; otherwise, proceed to S3.

[0054] S3: Perform a joint fault recording trigger condition detection. If the joint fault recording trigger condition is met, proceed to S5; otherwise, proceed to S4.

[0055] S4: Perform multi-parameter associated fault recording trigger condition detection. If the multi-parameter associated fault recording trigger condition is met, execute S5; if the multi-parameter associated fault recording trigger condition is not met, repeat S1-S4.

[0056] S5: Combine the load type identified in S1 to perform adaptive fault recording.

[0057] In this embodiment, the specific implementation method for acquiring the current waveform of the DC power supply and distribution circuit load in S1, analyzing the current waveform characteristics, and identifying the load type using the current waveform characteristics includes the following steps:

[0058] S1.1: Collect current signal data of each load in the DC power supply and distribution circuit, and store the collected current signal data to form continuous current waveform data;

[0059] Specifically, the current signal can be acquired through a current sampling circuit in the DC power supply and distribution circuit. The power sampling circuit may include a current sampling resistor or a Hall sensor. The sampling frequency of the current signal can be set between 1kHz and 10kHz, and this invention does not limit it.

[0060] S1.2: Preprocessing of current waveform data;

[0061] Furthermore, S1.2 performs preprocessing on the current waveform, including the following steps:

[0062] S1.2.1: Filter and denoise the current waveform data;

[0063] Specifically, a digital low-pass filter can be used to filter the current waveform data to remove high-frequency noise interference. The cutoff frequency of the filter can be set to 100Hz-500Hz depending on the load type.

[0064] S1.2.2: Perform zero-point calibration on the current waveform data after filtering and denoising;

[0065] Specifically, the following implementation method can be used for zero-point calibration: when the load is detected to be in a power-off state, record the actual current signal value as the zero-point offset, and subtract the offset from the current signal data value obtained by subsequent sampling to obtain the actual current signal value; the influence of DC bias when collecting current signal data is eliminated through zero-point calibration processing.

[0066] S1.2.3: Perform waveform segmentation processing on continuous current waveform data;

[0067] Specifically, when performing waveform segmentation processing on continuous current waveform data, the length of each time window can be set to 1s-10s, and the overlap rate between windows can be set to 50%. The above parameters can be adjusted according to different loads, and this invention does not limit them.

[0068] S1.3: Extract multidimensional feature parameters from the preprocessed current waveform data;

[0069] Specifically, the extracted multidimensional feature parameters include the following categories: time-domain feature parameters, frequency-domain feature parameters, statistical feature parameters, and waveform morphology feature parameters;

[0070] S1.4: Load type identification based on multi-dimensional feature parameters;

[0071] Specifically, in this step, the load type can be identified using a rule-based threshold discrimination method, which involves comparing the extracted multi-dimensional feature parameters with preset multi-dimensional feature parameter thresholds and determining the load type based on the comparison results.

[0072] Alternatively, a template matching method can be used. This involves first establishing a template library for each load type, collecting standard current waveform data of each load device before it leaves the factory, extracting standard multi-dimensional feature parameters from the standard current waveform data, comparing the currently collected multi-dimensional feature parameters with the standard multi-dimensional feature parameters, calculating a similarity index, and taking the load type corresponding to the template with the highest similarity as the identification result. The Euclidean distance method can be used to calculate the similarity index, and this invention does not limit it.

[0073] Here, the identification of load type can also be achieved by establishing a classification model, and this invention does not limit it;

[0074] It should be noted that, as shown in Table 1, load types can be divided into the following main categories and subcategories;

[0075]

[0076] In this embodiment, predictive fault recording trigger condition detection is performed in S2, and the specific implementation method includes the following steps:

[0077] In this step, predictive fault recording trigger condition detection is performed on the DC power supply and distribution circuit to capture the characteristics before the fault occurs. That is, by monitoring the changing trend, fluctuation characteristics or long-term drift characteristics of electrical parameters, abnormal states are identified before the electrical parameters exceed the traditional threshold, so as to start fault recording before the fault occurs and keep complete parameters before, during and after the fault occurs.

[0078] S2.1: Establish a mapping table between each load type and the predictive fault recording trigger condition, and set the threshold for the predictive fault recording trigger condition;

[0079] For example, for a brushed DC motor, the current rate of change threshold can be set to 50A / s, the voltage rate of change threshold can be set to 10V / s, and the drift threshold can be set to 0.5A; for a brushless DC motor, the current rate of change threshold can be set to 30A / s, the voltage rate of change threshold can be set to 8V / s, and the drift threshold can be set to 0.3A.

[0080] S2.2: Monitor the rate of change exceeding the trigger condition based on the set predictive fault recording trigger condition threshold;

[0081] S2.3: Monitor long-term drift trigger conditions based on the set predictive fault recording trigger condition threshold;

[0082] Long-term drift refers to the slow change in electrical parameters over a relatively long time window.

[0083] Furthermore, in S2.2, the monitoring of rate of change exceeding the trigger condition is based on the set predictive fault recording trigger condition threshold. The specific implementation method includes the following steps:

[0084] S2.2.1: The instantaneous current change rate is calculated using the backward difference method. When the instantaneous current change rate exceeds the instantaneous current change rate threshold, it is determined that the trigger condition of the change rate exceeding the limit is met.

[0085] The formula for calculating the instantaneous rate of change of current is:

[0086] ;

[0087] in, This represents the current value at the current sampling time. The current value at the previous sampling time. The sampling interval is denoted as .

[0088] S2.2.2: The instantaneous voltage change rate is calculated using backward differential distribution. When the instantaneous voltage change rate exceeds the instantaneous voltage change rate threshold, it is determined that the trigger condition of the change rate exceeding the limit is met.

[0089] Furthermore, in S2.3, long-term drift triggering condition monitoring is performed based on the set predictive fault recording triggering condition threshold. The specific implementation method includes the following steps:

[0090] S2.3.1: The trend of electrical parameters within a time window is fitted using linear regression to obtain the drift rate and total drift of the electrical parameters;

[0091] Specifically, in this step, the electrical parameters include current and voltage; we will use current as an example for explanation.

[0092] The electrical parameters of all sampling points within the acquisition time window are used to form a time series. And the least squares method was used to analyze the time series. By fitting the lines, a straight line is obtained. , where the slope The current drift rate is the velocity through a straight line. The total current drift can be calculated, and the total current drift is expressed as... express.

[0093] S2.3.2: Determine whether the drift rate and total drift of the electrical parameters exceed the drift rate threshold and the total drift threshold, respectively. If they do, the long-term drift trigger condition is met.

[0094] Specifically, the present invention can adopt a sliding window mechanism to update the electrical parameter trend fitting results every 1 second to avoid repeated triggering. The window length can be dynamically adjusted according to the load type. For example, a longer time window can be used for loads operating in steady state, and a shorter time window can be used for loads with changing operating conditions.

[0095] It should be noted that the fault recording operation in step S5 will be executed immediately upon the fulfillment of any of the above predictive fault recording trigger conditions.

[0096] In this embodiment, in step S3, the joint fault recording trigger condition detection is performed. In this step, the joint fault recording trigger uses a specific operating condition stage in the load operation process as a detection window. Under these operating conditions, the electrical parameters of the load have typical dynamic characteristics. Abnormal operating condition characteristics that deviate from normal operating condition characteristics indicate that there may be a fault in the DC power supply and distribution circuit. By comparing the real-time operating condition waveform with the preset normal operating condition template and threshold, the fault is detected in time and the fault signal that is only displayed when the operating condition changes is captured.

[0097] In this invention, the combined fault recording triggering conditions under operating conditions include the following three categories: abnormal triggering conditions during startup, abnormal triggering conditions during shutdown, and abnormal triggering conditions due to sudden load changes.

[0098] Monitoring electrical parameter waveforms during the load startup phase under abnormal trigger conditions; monitoring electrical parameter waveforms during the load shutdown phase under abnormal trigger conditions; monitoring electrical parameter waveforms during the load operation phase under abnormal trigger conditions for sudden load changes.

[0099] The above three types of combined fault recording trigger conditions are selected and enabled based on the load type, and the detection is triggered based on the load condition. If any condition is met, it is considered that the combined fault recording trigger conditions are met, and step S5 is executed; if none of them are met, the process proceeds to step S4.

[0100] The specific implementation method of step S3 includes the following steps:

[0101] S3.1: Detect the start-up event of the load, collect the start-up electrical parameters of the start-up event, and determine whether the start-up process meets the abnormal triggering conditions based on the characteristics of the start-up electrical parameters.

[0102] Specifically, a load start-up event is determined by monitoring the load's start-up electrical parameters. If the parameters rise from less than 5% of the load's rated electrical parameters to more than 10% of the rated electrical parameters and remain stable for more than a preset minimum start-up time, the start-up event is determined to have ended.

[0103] Furthermore, determining whether the startup process meets the abnormal triggering conditions involves the following steps:

[0104] S3.1.1: Obtain the normal startup characteristics of each load type and establish a standard library of normal startup characteristics;

[0105] The normal startup characteristics of each load type can be determined by conducting electrical tests on the load; the normal startup characteristics include the range of peak electrical parameters during startup, the range of startup time, and the waveform envelope of the electrical parameters during startup.

[0106] S3.1.2: Extract the electrical parameter features of the startup event and compare them with the normal startup feature standard library. If a startup electrical parameter feature deviates from the electrical parameter feature in the normal startup feature standard library, it is determined that it meets the abnormal triggering conditions of the startup process.

[0107] For example, if the peak current of a load deviates from the standard peak current range when the load starts normally, it is determined to meet the abnormal triggering conditions of the startup process.

[0108] S3.2: Detect shutdown events for the load and determine whether the abnormal triggering conditions of the shutdown process are met;

[0109] Specifically, a load shutdown event is determined by monitoring the load's shutdown electrical parameters. If the load's shutdown electrical parameters drop from normal operation to below a shutdown threshold and remain below that threshold for a preset minimum shutdown time, the shutdown event is considered to have started. When the load's shutdown electrical parameters drop to near zero and remain stable for more than 1 second, the shutdown event is considered to have ended.

[0110] S3.3: Detect load change events and determine whether they meet the abnormal triggering conditions for load change.

[0111] Specifically, during steady-state operation of the load, the rate of change of the load's electrical parameters is monitored. When the rate of change of the load's electrical parameters exceeds the sudden change detection threshold and lasts for more than 5 ms, it is determined that a load sudden change event has begun. When the rate of change of the electrical parameters re-enters a new steady-state range and remains stable for more than 0.5 seconds, it is determined that the sudden change event has ended.

[0112] It should be noted that the specific implementation method for determining whether the shutdown process abnormal triggering condition and the load change abnormal triggering condition are met is similar to that for determining whether the startup process abnormal triggering condition is met. Both methods first establish a standard feature library and then compare the electrical parameter characteristics of the shutdown event and the load change event with the standard feature library. If the electrical parameter characteristics of the shutdown event and the load change event deviate from the standard electrical parameters in the standard feature library, then it is determined that the shutdown process abnormal triggering condition and the load change abnormal triggering condition are met. The implementation steps will not be elaborated here.

[0113] In this embodiment, the multi-parameter associated fault recording trigger condition detection is performed in S4, and its specific implementation includes the following steps:

[0114] S4.1: Collect voltage and current signals of the load in the DC power supply and distribution circuit, establish a voltage-current physical correlation model, and perform voltage-current correlation anomaly trigger detection on the load based on the voltage-current physical correlation model. When the preset voltage-current correlation trigger condition is met, it is determined that the multi-parameter correlation fault recording trigger condition is met.

[0115] Furthermore, the preset voltage-current associated triggering conditions include at least one of the following:

[0116] Triggering conditions for abnormal static impedance, abnormal dynamic impedance, abnormal power consistency, and abnormal voltage-current correlation coefficient.

[0117] For example, static impedance anomaly triggering occurs during the steady-state operation of the load, by calculating the instantaneous impedance. ,in for Voltage sample value at time 10:00 for The current sample value at time t; the average impedance is obtained by smoothing the instantaneous impedance. When the average impedance When the load exceeds the preset normal impedance range for this type of load, it is determined that the static impedance abnormality triggering condition is met.

[0118] S4.2: Collect the current and temperature signals of the load circuit, establish a current-temperature thermal effect correlation model, and perform current-temperature correlation abnormal trigger detection on the load based on the current-temperature thermal effect correlation model. When the preset current-temperature correlation trigger conditions are met, it is determined that the multi-parameter correlation fault recording trigger conditions are met.

[0119] Furthermore, the preset current-temperature correlated triggering conditions include at least one of the following:

[0120] Anomaly triggering conditions for steady-state temperature rise, anomaly triggering conditions for thermal time constant, and anomaly triggering conditions for thermal runaway trend;

[0121] For example, the steady-state temperature rise anomaly triggering condition is to calculate the expected temperature rise based on the preset thermal resistance parameters, and at the same time calculate the actual temperature based on the real-time temperature and the ambient temperature; when the deviation between the actual temperature rise and the expected temperature rise exceeds the preset temperature rise deviation threshold, it is determined that the steady-state temperature rise anomaly triggering condition is met.

[0122] It should be noted that step S5 will be executed if any of the above conditions are met; otherwise, steps S1-S4 will be executed.

[0123] In this example, S5 combines the load type identified in S1 to perform adaptive fault recording, which specifically includes the following steps:

[0124] S5.1: Establish a mapping table between load type and waveform recording parameter set;

[0125] The waveform recording parameter set includes at least one or more of the following: sampling rate, trigger sensitivity, and waveform recording duration;

[0126] The mapping table can be customized according to the electrical characteristics and fault modes of different types of loads. Some waveform recording parameters are shown in Table 2:

[0127]

[0128] S5.2: Based on the specific type of triggering condition, dynamically optimize the waveform recording parameter set in S5.1 to generate the final waveform recording parameter set;

[0129] Dynamic optimization includes at least one of the following adjustment methods:

[0130] Sampling rate adaptive adjustment, recording duration adaptive adjustment, trigger sensitivity adaptive adjustment;

[0131] Specifically, the sampling rate is adaptively adjusted as follows: when the fault recording trigger condition is a predictive fault recording trigger condition (step S2), the sampling rate is doubled to improve the ability to capture early weak signals; when the fault recording trigger condition is a combined operating condition fault recording trigger condition (step S3), the sampling rate remains unchanged; when the trigger condition is a multi-parameter associated fault recording trigger condition (step S4), the sampling rate is adjusted according to the type of associated parameters. For example, for voltage-current associated parameter anomalies, the sampling rate remains unchanged.

[0132] The waveform recording duration is adaptively adjusted by dynamically extending or shortening the recording duration based on the severity of the fault. The severity of the fault is quantified by the over-limit multiple or deviation of the triggering condition. When the over-limit multiple is n, the recording duration is extended to n times the original recording duration.

[0133] The trigger sensitivity is adaptively adjusted as follows: if the load electrical parameters are continuously abnormal during fault recording, the trigger sensitivity will be increased by one level.

[0134] S5.3: Perform fault recording based on the adjusted recording parameter set.

[0135] The various embodiments of the present invention have now been described in detail. To avoid obscuring the concept of the invention, some details known in the art have not been described. Those skilled in the art will fully understand how to implement the technical solutions of this invention based on the above description, and the scope of the invention is defined by the appended claims.

Claims

1. A fault recording method for a DC power supply and distribution circuit, characterized in that: Includes the following steps: S1: Collect the current waveform of the load in the DC power supply and distribution circuit, analyze the characteristics of the current waveform, and use the characteristics of the current waveform to identify the load type; S2: Perform a predictive fault recording trigger condition check. If the predictive fault recording trigger condition is met, proceed to S5; otherwise, proceed to S3. S3: Perform a joint fault recording trigger condition detection. If the joint fault recording trigger condition is met, proceed to S5; otherwise, proceed to S4. S4: Perform multi-parameter associated fault recording trigger condition detection. If the multi-parameter associated fault recording trigger condition is met, execute S5; if the multi-parameter associated fault recording trigger condition is not met, repeat S1-S4. S5: Combine the load type identified in S1 to perform adaptive fault recording.

2. The fault recording method for a DC power supply and distribution circuit according to claim 1, characterized in that, The specific implementation method of S1 includes the following steps: S1.1: Collect current signal data of each load in the DC power supply and distribution circuit, and store the collected current signal data to form continuous current waveform data; S1.2: Preprocessing of current waveform data; S1.3: Extract multidimensional feature parameters from the preprocessed current waveform data; S1.4: Load type identification based on multi-dimensional feature parameters.

3. The fault recording method for a DC power supply and distribution circuit according to claim 2, characterized in that, The preprocessing of the current waveform in S1.2 includes the following steps: S1.2.1: Filter and denoise the current waveform data; S1.2.2: Perform zero-point calibration on the current waveform data after filtering and denoising; S1.2.3: Perform waveform segmentation processing on continuous current waveform data.

4. The fault recording method for a DC power supply and distribution circuit according to claim 1, characterized in that, The predictive fault recording trigger condition detection is performed in S2, and its specific implementation method includes the following steps: S2.1: Establish a mapping table between each load type and the predictive fault recording trigger condition, and set the threshold for the predictive fault recording trigger condition; S2.2: Monitor the rate of change exceeding the trigger condition based on the set predictive fault recording trigger condition threshold; S2.3: Monitor long-term drift trigger conditions based on the set predictive fault recording trigger condition threshold.

5. The fault recording method for a DC power supply and distribution circuit according to claim 4, characterized in that, The monitoring of rate of change exceeding the trigger condition in S2.2 based on the set predictive fault recording trigger condition threshold includes the following steps: S2.2.1: The instantaneous current change rate is calculated using the backward difference method. When the instantaneous current change rate exceeds the instantaneous current change rate threshold, it is determined that the trigger condition of the change rate exceeding the limit is met. S2.2.2: The instantaneous voltage change rate is calculated using backward differential distribution. When the instantaneous voltage change rate exceeds the instantaneous voltage change rate threshold, it is determined that the trigger condition of the change rate exceeding the limit is met.

6. The fault recording method for a DC power supply and distribution circuit according to claim 4, characterized in that, The long-term drift trigger condition monitoring in S2.3 based on the set predictive fault recording trigger condition threshold includes the following steps: S2.3.1: The trend of electrical parameters within a time window is fitted using linear regression to obtain the drift rate and total drift of the electrical parameters; S2.3.2: Determine whether the drift rate and total drift of the electrical parameters exceed the drift rate threshold and the total drift threshold, respectively. If they do, the long-term drift trigger condition is met.

7. The fault recording method for a DC power supply and distribution circuit according to claim 1, characterized in that, The specific implementation method of S3 includes the following steps: S3.1: Detect the start-up event of the load, collect the start-up electrical parameters of the start-up event, and determine whether the start-up process meets the abnormal triggering conditions based on the characteristics of the start-up electrical parameters. S3.2: Detect shutdown events for the load and determine whether the abnormal triggering conditions of the shutdown process are met; S3.3: Detect load mutation events and determine whether they meet the abnormal triggering conditions for load mutation.

8. The fault recording method for a DC power supply and distribution circuit according to claim 7, characterized in that, The specific implementation method of S3.1 includes the following steps: S3.1.1: Obtain the normal startup characteristics of each load type and establish a standard library of normal startup characteristics; S3.1.2: Extract the electrical parameter features of the startup event and compare them with the normal startup feature standard library. If a startup electrical parameter feature deviates from the electrical parameter feature in the normal startup feature standard library, it is determined that it meets the abnormal triggering conditions of the startup process.

9. The fault recording method for a DC power supply and distribution circuit according to claim 1, characterized in that, The multi-parameter correlation fault recording trigger condition detection in S4 includes the following steps: S4.1: Collect voltage and current signals of the load in the DC power supply and distribution circuit, establish a voltage-current physical correlation model, and perform voltage-current correlation anomaly trigger detection on the load based on the voltage-current physical correlation model. When the preset voltage-current correlation trigger condition is met, it is determined that the multi-parameter correlation fault recording trigger condition is met. S4.2: Collect the current and temperature signals of the load circuit, establish a current-temperature thermal effect correlation model, and perform current-temperature correlation anomaly trigger detection on the load based on the current-temperature thermal effect correlation model. When the preset current-temperature correlation trigger conditions are met, it is determined that the multi-parameter correlation fault recording trigger conditions are met.

10. A fault recording method for a DC power supply and distribution circuit according to claim 1, characterized in that, In step S5, adaptive fault recording is performed by combining the load type identified in step S1. The specific implementation of this method is as follows: Includes the following steps: S5.1: Establish a mapping table between load type and waveform recording parameter set; S5.2: Based on the specific type of triggering condition, dynamically optimize the waveform recording parameter set in S5.1 to generate the final waveform recording parameter set; S5.3: Perform fault recording based on the adjusted recording parameter set.

Images