Current sensor fault detection method and device, computer equipment and storage medium

Abstract

The invention relates to a current sensor fault detection method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring a current value obtained by detecting a to-be-detected line by at least one current sensor at a current moment; for each current sensor, determining the confidence of the current sensor at the current moment according to the current value of the current sensor at the current moment; performing fusion processing on the current values of the different current sensors according to the magnitude relationship between the confidence coefficients of the different current sensors at the current moment and a preset confidence coefficient threshold value to obtain a target fusion current; and determining the fault condition of each current sensor according to the confidence coefficient fluctuation condition of each current sensor at the current moment and the magnitude relationship between the target fusion current and a preset current threshold. The method can improve the fault detection accuracy of the current sensor.

Description

Technical Field

[0001] This application relates to the field of fault detection technology, and in particular to a method, apparatus, computer equipment, and storage medium for fault detection of a current sensor. Background Technology

[0002] With the rapid development of smart grids and industrial automation technologies, current sensors, as key measurement devices, are playing an increasingly important role in system status monitoring, fault protection, and energy efficiency management. Currently, most systems still use a single current sensor or a simple multi-sensor redundancy scheme for current measurement. In the single-sensor scheme, the system relies on a single sensor to provide current data. This method is simple in structure and low in cost, but it is susceptible to factors such as temperature drift, electromagnetic interference, and sensor aging, which can lead to decreased measurement accuracy or even complete failure, making it difficult to meet the requirements of high-reliability applications.

[0003] To improve system reliability, some high-end systems have begun to adopt multi-sensor redundancy designs, fusing data from the outputs of multiple sensors to obtain more stable and accurate measurement results. Traditional data fusion methods often employ weighted averaging with fixed weights or selective fusion based on simple rules. However, these methods fail to fully consider the dynamic performance changes of each sensor under different operating conditions and cannot adaptively adjust the weight of a sensor in the fusion process when a sensor experiences sudden interference or performance degradation. This "static" fusion strategy not only limits the improvement of overall system performance but may even lead to fusion results inferior to those from a single sensor output due to contamination from faulty sensor data.

[0004] Furthermore, most existing fault diagnosis functions focus on detecting anomalies in the measured lines or equipment, lacking the ability to specifically assess the health status of the current sensor itself. Systems often struggle to effectively distinguish between "real line faults" and "false alarms caused by sensor malfunctions," leading to confusion in maintenance decisions and hindering the effective implementation of predictive maintenance strategies.

[0005] Therefore, existing current sensor measurement and fault diagnosis methods have significant shortcomings in terms of reliability, adaptability, and fault differentiation accuracy. There is an urgent need for an integrated solution that can dynamically assess sensor reliability, achieve intelligent adaptive data fusion, and simultaneously and accurately diagnose line faults and sensor faults. Summary of the Invention

[0006] Therefore, it is necessary to provide a current sensor fault detection method, device, computer equipment, and storage medium that can accurately measure current sensor faults, addressing the aforementioned technical problems.

[0007] Firstly, this application provides a method for detecting faults in a current sensor, including:

[0008] Obtain the current value detected by at least one current sensor on the circuit under test at the current moment;

[0009] For each current sensor, determine the confidence level of the current sensor at the current moment based on the current value of the current sensor at the current moment;

[0010] Based on the relationship between the confidence levels of different current sensors at the current moment and the preset confidence threshold, the current values ​​of different current sensors are fused to obtain the target fused current.

[0011] Based on the confidence fluctuation of each current sensor at the current moment, and the relationship between the target fused current and the preset current threshold, the fault condition of each current sensor is determined.

[0012] In one embodiment, determining the confidence level of the current sensor at the current moment based on the current value of the current sensor at the current moment includes:

[0013] Based on the current value of the current sensor at the current moment, determine the evaluation value of the current sensor under at least one evaluation index; the evaluation index includes current consistency index, timing consistency index and current quality index.

[0014] The evaluation values ​​under different evaluation indicators are spliced ​​together to obtain the target evaluation data;

[0015] Based on the confidence assessment model, the confidence level of the current sensor at the current moment is determined according to the target assessment data; the confidence assessment model is obtained by training a long short-term memory network model.

[0016] In one embodiment, determining the evaluation value of the current sensor under at least one evaluation metric based on the current value of the current sensor at the current moment includes:

[0017] Based on the current sensor's current value at the current moment, the median and standard deviation of the corresponding current values ​​from different current sensors, determine the evaluation value of the current sensor under the current consistency index; and,

[0018] Based on the current sensor's current value at the current moment and its predicted value at future moments, the evaluation value of the current sensor under the timing consistency index is determined; the predicted value is obtained based on the current prediction model's prediction of the current sensor's current values ​​at historical moments; and...

[0019] Based on the current value of the current sensor at the current moment, determine the current signal-to-noise ratio of the current sensor at the current moment, and based on the current signal-to-noise ratio and the historical signal-to-noise ratio of the current sensor at historical moments, determine the evaluation value of the current sensor under the current quality index.

[0020] In one embodiment, the current values ​​of different current sensors are fused based on the relationship between the confidence levels of different current sensors at the current moment and a preset confidence threshold to obtain a target fused current, including:

[0021] A current sensor with a confidence level greater than a preset confidence threshold is selected as the target sensor.

[0022] For each target sensor, a confidence weight is determined based on the confidence level of the target sensor.

[0023] Based on the current values ​​of different target sensors and their corresponding confidence weights, the current values ​​of each target sensor are fused to obtain the target fused current.

[0024] In one embodiment, the method further includes:

[0025] If the confidence levels of all current sensors are not greater than the preset confidence threshold, an alarm for sensor abnormality will be output.

[0026] In one embodiment, the fault status of each current sensor is determined based on the confidence fluctuation of each current sensor at the current moment and the relationship between the target fused current and a preset current threshold, including:

[0027] If the target fusion current is not greater than the preset current threshold, it is determined that the current sensor with drastic fluctuation in confidence at the current moment has failed.

[0028] If the target fusion current is greater than the preset current threshold, and the number of current sensors with drastic confidence fluctuations at the current moment is greater than the preset number threshold, then the circuit under test is determined to be faulty.

[0029] Secondly, this application also provides a current sensor fault detection device, comprising:

[0030] The acquisition module is used to acquire the current value detected by at least one current sensor on the circuit under test at the current moment;

[0031] The confidence module is used to determine the confidence level of each current sensor at the current moment based on the current value of the current sensor at the current moment.

[0032] The fusion module is used to fuse the current values ​​of different current sensors based on the relationship between the confidence level of different current sensors at the current moment and the preset confidence threshold, so as to obtain the target fused current.

[0033] The detection module is used to determine the fault status of each current sensor based on the confidence fluctuation of each current sensor at the current moment and the relationship between the target fused current and the preset current threshold.

[0034] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0035] Obtain the current value detected by at least one current sensor on the circuit under test at the current moment;

[0036] For each current sensor, determine the confidence level of the current sensor at the current moment based on the current value of the current sensor at the current moment;

[0037] Based on the relationship between the confidence levels of different current sensors at the current moment and the preset confidence threshold, the current values ​​of different current sensors are fused to obtain the target fused current.

[0038] Based on the confidence fluctuation of each current sensor at the current moment, and the relationship between the target fused current and the preset current threshold, the fault condition of each current sensor is determined.

[0039] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0040] Obtain the current value detected by at least one current sensor on the circuit under test at the current moment;

[0041] For each current sensor, determine the confidence level of the current sensor at the current moment based on the current value of the current sensor at the current moment;

[0042] Based on the relationship between the confidence levels of different current sensors at the current moment and the preset confidence threshold, the current values ​​of different current sensors are fused to obtain the target fused current.

[0043] Based on the confidence fluctuation of each current sensor at the current moment, and the relationship between the target fused current and the preset current threshold, the fault condition of each current sensor is determined.

[0044] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0045] Obtain the current value detected by at least one current sensor on the circuit under test at the current moment;

[0046] For each current sensor, determine the confidence level of the current sensor at the current moment based on the current value of the current sensor at the current moment;

[0047] Based on the relationship between the confidence levels of different current sensors at the current moment and the preset confidence threshold, the current values ​​of different current sensors are fused to obtain the target fused current.

[0048] Based on the confidence fluctuation of each current sensor at the current moment, and the relationship between the target fused current and the preset current threshold, the fault condition of each current sensor is determined.

[0049] The aforementioned current sensor fault detection method, device, computer equipment, and storage medium effectively improve the system's measurement robustness and reliability under sensor performance degradation or sudden failure conditions by dynamically evaluating the real-time confidence level of each current sensor and performing adaptive weighted fusion based on this confidence level. Specifically, confidence level evaluation quantifies the data credibility of each sensor; the confidence level-based fusion strategy automatically reduces the weight of low-confidence sensors, suppressing the interference of abnormal data on the fusion result, thereby ensuring the accuracy of the target fused current. Furthermore, by analyzing the correlation characteristics between confidence level fluctuations and fused current values, the system can accurately distinguish between sensor faults and actual line faults: if only a few sensors have abnormal confidence levels while the fused current is normal, it is determined to be a sensor fault; if most sensors have synchronously abnormal confidence levels and the fused current exceeds the limit, it is determined to be a line fault. This method improves the system's fault tolerance while effectively identifying fault types, providing key technical support for safe system operation and predictive maintenance. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 This is an application environment diagram of a current sensor fault detection method provided in this embodiment;

[0052] Figure 2 This is a flowchart illustrating the first current sensor fault detection method provided in this embodiment;

[0053] Figure 3A This is a flowchart illustrating a step for determining confidence level, as provided in this embodiment.

[0054] Figure 3B This is a model structure diagram provided in this embodiment;

[0055] Figure 4 This is a flowchart illustrating a step for obtaining the target fusion current in this embodiment;

[0056] Figure 5 This is a structural block diagram of a current sensor fault detection device provided in this embodiment;

[0057] Figure 6 This is an internal structural diagram of a computer device provided in this embodiment. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0059] The current sensor fault detection method provided in this application embodiment can be applied to, for example, Figure 1 The application environment is illustrated. The terminal communicates with the server via a network. The data storage system stores the data the server needs to process. The data storage system can be integrated on the server or located in the cloud or on other network servers. The computer device acquires the current value detected by at least one current sensor at the current moment for the circuit under test. For each current sensor, the confidence level of the current sensor at the current moment is determined based on its current value. Based on the relationship between the confidence levels of different current sensors at the current moment and a preset confidence threshold, the current values ​​of different current sensors are fused to obtain a target fused current. Based on the fluctuation of the confidence levels of each current sensor at the current moment, and the relationship between the target fused current and the preset current threshold, the fault status of each current sensor is determined. The computer device can be a terminal or a server. The terminal can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. The server can be implemented using a standalone server or a server cluster composed of multiple servers.

[0060] In one exemplary embodiment, such as Figure 2 As shown, a fault detection method for a current sensor is provided, which can be applied to... Figure 1 The following steps, from S201 to S204, are used as an example of computer equipment.

[0061] S201 acquires the current value detected by at least one current sensor at the current moment for the circuit under test.

[0062] A current sensor is a measuring device used to detect the magnitude of current in a circuit. It converts current signals into standard electrical or digital signals that can be acquired, processed, and identified. The circuit under test (DUT) refers to the electrical loop or power supply path being monitored; it is the specific carrier and target object for the current sensor to perform its measurement. The current value is the physical quantity acquired by the current sensor that characterizes the magnitude of the current in the DUT, usually measured in amperes. It serves as the fundamental data for condition monitoring, fault diagnosis, and fusion calculations.

[0063] In some embodiments, the initial current value detected by at least one current sensor at the current moment is obtained; the initial current value is preprocessed to obtain a preprocessed current value. The preprocessing may be at least one of filtering and noise reduction.

[0064] For example, the filtering of the initial current value is implemented as shown in the following formula (1):

[0065] (1)

[0066] in, Let be the current value of the i-th current sensor after bandpass filtering at time t. , Here are the coefficients of the bandpass filter, and M is the order of the filter.

[0067] For example, the method for denoising the initial current value is shown in the following formula (2):

[0068] (2)

[0069] Among them, z i (n) represents the current value of the i-th current sensor after noise reduction by moving average at time t, and L is the length of the moving average window.

[0070] It should be noted that the raw measurements output from the multi-source current sensors undergo signal preprocessing (filtering / denoising), combined with historical data, signal quality analysis, and optional machine learning algorithms to construct a dynamic confidence assessment model, generating the confidence score for each measurement in real time. This real-time confidence score is then fed into an adaptive weighted fusion center. At the fusion strategy selection node, three fusion paths are automatically selected based on the confidence score interval and consistency results: weighted fusion of high-confidence intervals, selective fusion of medium- and low-confidence intervals, and evidence-based fusion of conflicting confidence score intervals. The fusion output forms the optimal estimated current value, and the change in confidence score over time is sent to the confidence score trajectory analysis module; both work together to drive the fault diagnosis engine for anomaly detection. Subsequently, at the fault type discrimination node, the fault is attributed to either the line side or the sensor side: when it is determined to be on the line side, the fault enters the line fault branch, outputting evidence signals such as "abnormal current / synchronous fluctuation of confidence level," triggering line protection actions, and providing high-precision current measurement output to the display and control system; when it is determined to be on the sensor side, the fault enters the sensor fault branch, outputting evidence signals such as "sudden drop / abrupt change in confidence level," triggering sensor health warnings and predictive maintenance suggestions. The blue path connecting back from the right side in the diagram represents the feedback loop: statistical and trend information from "confidence level trajectory analysis" is fed back to the "dynamic confidence level assessment model" to update assessment parameters and thresholds, thereby forming a closed-loop adaptive adjustment.

[0071] S202 determines the confidence level of each current sensor at the current moment based on the current value of the current sensor at the current moment.

[0072] Among them, confidence level refers to a dynamic parameter that is calculated online using machine learning algorithms based on multi-source information such as the consistency of current sensor readings, the continuity of historical data, and signal quality characteristics. It is used to evaluate the reliability of the current sensor's measurement data under the current operating conditions.

[0073] In some embodiments, based on a confidence assessment model, for each current sensor, the confidence level of the current sensor at the current moment is determined according to the current value of the current sensor at the current moment. The confidence assessment model uses the consistency of current sensor readings, the continuity of historical data, and signal quality characteristics as input parameters, and calculates the real-time confidence level of each sensor online through a built-in machine learning algorithm.

[0074] For example, as shown in the following formula (3):

[0075] (3)

[0076] Where, x i (n) represents the current value of the i-th current sensor at time t, s i (n) represents the continuous-time signal of the i-th current sensor, Ts To synchronize the sampling period, Let be the measurement noise of the i-th current sensor.

[0077] S203 performs fusion processing on the current values ​​of different current sensors based on the relationship between the confidence level of different current sensors at the current moment and the preset confidence threshold, to obtain the target fused current.

[0078] The confidence threshold is a pre-set critical value used to judge the reliability of current sensor data. When the real-time confidence of the sensor is higher than the threshold, its measurement data is considered reliable; when it is lower than the threshold, the data is considered to be abnormal or unreliable. The target fused current is the optimal current estimate calculated by a data fusion algorithm after dynamically adjusting the weights based on the real-time confidence of each current sensor. It is used to reflect the true current state of the measured circuit.

[0079] In some embodiments, the relationship between the confidence levels of different current sensors at the current moment and a preset confidence threshold is determined; current sensors with a confidence level greater than the preset confidence threshold at the current moment are designated as normal sensors; current sensors with a confidence level not greater than the preset confidence threshold at the current moment are designated as abnormal sensors; when the number of abnormal sensors is less than a preset number threshold, the current values ​​of different current sensors are fused to obtain a target fused current; when the number of abnormal sensors is not less than the preset number threshold, only the current values ​​of different normal sensors are fused to obtain a target fused current.

[0080] S204 determines the fault status of each current sensor based on the confidence fluctuation of each current sensor at the current moment and the relationship between the target fused current and the preset current threshold.

[0081] The confidence level fluctuation refers to the stability of the real-time confidence level of the current sensor over time during continuous monitoring, including abnormal trends such as short-term sharp fluctuations or long-term monotonic declines. This reflects the dynamic degradation or sudden failure of the sensor's performance. The preset current threshold is a current value boundary pre-set according to the normal operating range of the circuit under test. When the fused current exceeds this threshold, it may trigger an overload or fault alarm. The fault condition refers to the sensor's own performance degradation (such as gradual deviation) or sudden hard fault (such as signal interruption), as well as actual overload or short-circuit faults in the circuit under test, distinguished by analyzing the confidence level fluctuation and the fused current value.

[0082] In some embodiments, faults in the current sensors are diagnosed by analyzing the long-term evolution trajectory and abrupt change characteristics of the confidence levels of each current sensor. If the confidence level of a current sensor shows a long-term monotonically decreasing trend, it is determined that its performance has deteriorated; if its confidence level drops sharply, it is determined to be a sudden hard fault. If the target fused current is abnormal, and the confidence levels of all or most current sensors fluctuate synchronously and drastically, it is determined to be a line fault. If the target fused current is normal, and only the confidence levels of one or a few current sensors are significantly reduced, it is determined to be a sensor fault.

[0083] For example, if the target fusion current is not greater than a preset current threshold, it is determined that the current sensor with a drastic fluctuation in confidence level at the current moment has failed; if the target fusion current is greater than the preset current threshold, and the number of current sensors with a drastic fluctuation in confidence level at the current moment is greater than a preset number threshold, then it is determined that the circuit under test has failed.

[0084] It should be noted that the measurement data output from the multi-source current sensors undergoes data acquisition and preprocessing, dynamic confidence assessment, adaptive weighted fusion, and dual fault diagnosis before entering the fault type determination stage. If a sensor-related fault is determined, a sensor fault warning is triggered; if a line / system-side fault is determined, line fault protection is triggered. Subsequently, a real-time update mechanism is implemented to perform online correction of the fusion weights, diagnostic thresholds, and model parameters, and the update results are written to the status database and alarm records during the system status update step. After the status update is completed, the fault determination results, along with the updated weights and thresholds, are sent back to the "dynamic confidence assessment" and "adaptive weighted fusion" stages via the system feedback channel, forming a closed-loop adaptive adjustment.

[0085] The aforementioned current sensor fault detection method effectively improves the system's measurement robustness and reliability under sensor performance degradation or sudden failure conditions by dynamically evaluating the real-time confidence level of each current sensor and performing adaptive weighted fusion accordingly. Specifically, confidence level assessment quantifies the reliability of data from each sensor; the confidence level-based fusion strategy automatically reduces the weight of low-confidence sensors, suppressing the interference of abnormal data on the fusion result, thereby ensuring the accuracy of the target fused current. Furthermore, by analyzing the correlation characteristics between confidence level fluctuations and fused current values, the system can accurately distinguish between sensor faults and actual line faults: if only a few sensors have abnormal confidence levels while the fused current is normal, it is determined to be a sensor fault; if most sensors have synchronously abnormal confidence levels and the fused current exceeds the limit, it is determined to be a line fault. This method improves the system's fault tolerance while effectively identifying fault types, providing key technical support for safe system operation and predictive maintenance.

[0086] Figure 3AThis is a flowchart illustrating the process of determining the confidence level in one embodiment. This embodiment refines the steps in the above embodiment for determining the confidence level of the current sensor at the current moment based on the current value of the current sensor at the current moment, including the following steps:

[0087] S301 determines the evaluation value of the current sensor under at least one evaluation index based on the current value of the current sensor at the current moment.

[0088] Evaluation metrics refer to multiple dimensions of parameters used to quantify the performance of current sensors, including current consistency metrics (reflecting the deviation of sensor readings from group or historical data), time series consistency metrics (reflecting the time series prediction accuracy of sensor readings), and current quality metrics (reflecting the signal-to-noise ratio or waveform distortion of the sensor signal). The evaluation value is a quantitative value used to characterize the sensor's performance under a given metric, derived from the current value of the current sensor at the current moment, combined with the calculation methods of each evaluation metric (such as deviation calculation, prediction error analysis, signal-to-noise ratio estimation, etc.).

[0089] In some embodiments, the evaluation value of the current sensor under the current consistency index is determined based on the current value of the current sensor at the current moment, the median and standard deviation of the corresponding current values ​​of different current sensors.

[0090] For example, the evaluation value of the current sensor under the current consistency index can be determined based on the following formula (4).

[0091] (4)

[0092] in, Let be the evaluation value of the i-th current sensor under the current consistency index at time t; The median value for all current sensors at time t; Let be the standard deviation of the readings of all current sensors at time t. is the consistency sensitivity coefficient. Wherein, .

[0093] In some embodiments, the evaluation value of the current sensor under the timing consistency index is determined based on the current value of the current sensor at the current time and the predicted value of the current sensor at a future time; the predicted value is obtained by predicting the current value of the current sensor at historical times based on the current prediction model.

[0094] For example, the evaluation value of the current sensor under the timing consistency index can be determined based on the following formula (5).

[0095] (5)

[0096] in, Let be the historical continuity factor of the i-th current sensor at time t; This is a predicted value for a given moment based on historical data. This is the sensitivity coefficient of the historical continuity factor. For the forgetting factor in the prediction model; This is the predicted value for time -1 based on historical data.

[0097] In some embodiments, the current signal-to-noise ratio of the current sensor is determined based on the current value of the current sensor at the current moment, and the evaluation value of the current sensor under the current quality index is determined based on the current signal-to-noise ratio and the historical signal-to-noise ratio of the current sensor at historical moments.

[0098] For example, the evaluation value of the current sensor under the current quality index can be determined based on the following formula (6).

[0099] (6)

[0100] in, Let be the signal quality factor of the i-th current sensor at time t; Let be the signal-to-noise ratio of the i-th current sensor at time t; These are the adjustment parameters for the signal quality factor; Signal power; This represents noise power.

[0101] It should be noted that, in addition to determining the evaluation value of the current sensor under the current quality index based on the signal-to-noise ratio, this embodiment can also evaluate the signal quality based on specific frequency band energy or waveform distortion.

[0102] S302 splices the evaluation values ​​under different evaluation indicators to obtain the target evaluation data.

[0103] In some embodiments, the evaluation values ​​under different evaluation indicators are spliced ​​together using the following formula (7) to obtain the target evaluation data.

[0104] (7)

[0105] Among them, F i (n) represents the target evaluation data of the i-th current sensor at time t; Let be the gradient of the reading of the i-th current sensor at time t.

[0106] Based on the confidence assessment model, S303 determines the confidence level of the current sensor at the current moment according to the target assessment data.

[0107] The confidence assessment model is trained based on a Long Short-Term Memory (LSTM) network model. For example, ... Figure 3B The model structure diagram is shown below. The feature inputs from the multi-source channels (including consistency features, historical stability features, signal quality features, and observations and their gradients) are vectorized and combined in the feature concatenation module, then fed into three Sigmoid fully connected gated branches and one tanh candidate state branch. The three gated branches sequentially generate the gating coefficients for the forget gate, input gate, and output gate; the candidate branch generates new candidate memory content. The forget gate coefficient is multiplied by the long-term memory information from the previous time step through a multiplication gate, achieving selective retention of historical memory; the input gate coefficient is multiplied by the candidate memory content through a multiplication gate and then added to the aforementioned retained information to obtain the updated long-term memory for the current time step. Subsequently, the updated long-term memory is activated and multiplied by the output gate coefficient at the multiplication gate to form the short-term output for the current time step (shown by the red dashed line), which serves as one of the recursive inputs for the next time step. This structure, while ensuring long-term dependency modeling capabilities, achieves gating suppression of sudden noise and anomalies, thereby outputting robust confidence estimates and intermediate states.

[0108] In some embodiments, the confidence level of the current sensor at the current moment is determined based on the target evaluation data using the confidence evaluation model, according to the following formula (8).

[0109] (8)

[0110] in, These are the first and second hidden states of the neural network, respectively; W 1 W 2 W 3 is the weight of the three factors in the weighted fusion; b is the offset.

[0111] It should be noted that during the data fusion process, a real-time update mechanism is used to update the parameters as shown in the following formula (9):

[0112] (9)

[0113] in, The consistency sensitivity coefficient is adjusted over time. This is the initial consistency sensitivity coefficient; Adaptive learning rate; For the past The variance of the confidence level at each time point; This represents the length of the time window.

[0114] It should be noted that the loss function for online learning of the model is shown in the following formula (10):

[0115] (10)

[0116] Where L(n) is the loss function; I ground_truth Let be the true value of the current at time t. For regularization parameters; The model parameter vector is used; gradient descent is employed for parameter updates. This is the learning rate.

[0117] The aforementioned technical solution achieves dynamic and accurate calculation of the confidence level of current sensors by comprehensively evaluating data from multiple dimensions, including current consistency, timing consistency, and current quality indicators, and by using a Long Short-Term Memory (LSTM) network model to train a confidence assessment model. Its advantages lie in the fact that the LSTM model can effectively capture long-term dependencies and timing characteristics in sensor data, thus more accurately reflecting the real-time performance status of the sensor under different operating conditions. Simultaneously, the multi-indicator fusion assessment avoids the limitations of a single indicator, improving the comprehensiveness and robustness of the confidence assessment, and providing a reliable basis for subsequent intelligent adaptive data fusion and accurate fault diagnosis.

[0118] Figure 4 This is a flowchart illustrating the process of obtaining the target fused current in one embodiment. This embodiment refines the steps in the above embodiment where the current values ​​of different current sensors are fused based on the relationship between their confidence levels at the current moment and a preset confidence threshold to obtain the target fused current. The steps include the following:

[0119] S401 selects a current sensor with a confidence level greater than a preset confidence threshold as the target sensor.

[0120] S402 determines the confidence weight of each target sensor based on its confidence level.

[0121] In some embodiments, when the confidence (n) of all current sensors is higher than a preset confidence threshold, the weights are normalized using the following formula (11) to determine the confidence weight of the target sensor.

[0122] (11)

[0123] Among them, w i (n) represents the fusion weight of the i-th sensor at time t;

[0124] In some embodiments, when the confidence level of some current sensors decreases significantly, the system automatically switches to adopting only high-confidence sensor data or employs a median fusion strategy to eliminate interference from faulty sensors. A significant decrease in the confidence level (n) of some sensors means that the confidence level of at least one sensor falls below a certain low-confidence threshold. A selective fusion strategy is employed to construct a high-confidence sensor ensemble. , ,like If the set is not empty, use the sensor data within the set for weighted averaging.

[0125] S403 performs fusion processing on the current values ​​of each target sensor according to the current values ​​of different target sensors and the corresponding confidence weights of the target sensors to obtain the target fused current.

[0126] In some embodiments, when the confidence (n) of all current sensors is higher than a preset confidence threshold, the target fused current is obtained by the following formula (12).

[0127] (12)

[0128] in, The fusion current value at time t;

[0129] In some embodiments, if the confidence level (n) of any current sensor is not higher than a preset confidence threshold, the target fused current is obtained by the following formula (13).

[0130] (13)

[0131] if If the value is empty, meaning all sensor confidence levels are at a medium or low level, the system outputs an alarm signal to remind maintenance. When performing adaptive weighted fusion of sensor data, fusion uncertainty estimation is considered. The variance of the fusion current value; Let be the measurement method of the i-th sensor at time t.

[0132] In some embodiments, a sensor malfunction alarm is output when the confidence levels of all current sensors are not greater than a preset confidence threshold.

[0133] It should be noted that, for signals in a fault detection system, a discrete state space is constructed for multi-source signals:

[0134] (14)

[0135] Where X(n) is the system state variable, which is the state vector containing all key variables; Corresponding to multi-source data acquisition, For dynamic confidence assessment, For adaptive weighted fusion, establish the state transition equation:

[0136] (15)

[0137] Where f() is the state transition function, describing the real-time state evolution of the system; X(n-1) is the historical input data; U(n) is the adaptive environmental parameter; V(n) is the process noise, representing the uncertainty of the system model; the system's observation equation is:

[0138] (16)

[0139] Where Y(n) is the observation vector, For the fusion current value, The result is the fault diagnosis result, H is the observation matrix, and W(n) is the observation noise.

[0140] In the above embodiments, threshold screening ensures the quality of data participating in the fusion, avoiding distortion of the fusion results due to faulty or abnormal sensors. The confidence-based weight allocation mechanism allows higher-performing sensors to contribute more to the final result, thus maintaining high accuracy and stability of the target fused current even under complex operating conditions. This logic not only improves the system's anti-interference capability but also fully utilizes the complementarity of multi-source data, significantly enhancing the reliability and adaptability of current measurement.

[0141] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0142] Based on the same inventive concept, this application also provides a current sensor fault detection device for implementing the current sensor fault detection method described above. The solution provided by this device is similar to the implementation described in the above method; therefore, the specific limitations in one or more current sensor fault detection device embodiments provided below can be found in the limitations of the current sensor fault detection method described above, and will not be repeated here.

[0143] In one exemplary embodiment, such as Figure 5 As shown, a current sensor fault detection device is provided, comprising: an acquisition module, a confidence module, a fusion module, and a detection module, wherein:

[0144] The acquisition module is used to acquire the current value detected by at least one current sensor on the circuit under test at the current moment;

[0145] The confidence module is used to determine the confidence level of each current sensor at the current moment based on the current value of the current sensor at the current moment.

[0146] The fusion module is used to fuse the current values ​​of different current sensors based on the relationship between the confidence level of different current sensors at the current moment and the preset confidence threshold, so as to obtain the target fused current.

[0147] The detection module is used to determine the fault status of each current sensor based on the confidence fluctuation of each current sensor at the current moment and the relationship between the target fused current and the preset current threshold.

[0148] In some embodiments, the confidence module is further configured to determine the evaluation value of the current sensor under at least one evaluation index based on the current value of the current sensor at the current moment; the evaluation index includes a current consistency index, a timing consistency index, and a current quality index; the evaluation values ​​under different evaluation indexes are spliced ​​together to obtain target evaluation data; based on the confidence evaluation model, the confidence level of the current sensor at the current moment is determined according to the target evaluation data; the confidence evaluation model is obtained by training a long short-term memory network model.

[0149] In some embodiments, the confidence module is further configured to: determine the evaluation value of the current sensor under the current consistency index based on the current value of the current sensor at the current moment, the median and standard deviation of the corresponding current values ​​of different current sensors; and determine the evaluation value of the current sensor under the time consistency index based on the current value of the current sensor at the current moment and the predicted value of the current sensor at a future moment; wherein the predicted value is obtained by predicting the current value of the current sensor at a historical moment based on the current prediction model; and determine the current signal-to-noise ratio of the current sensor at the current moment based on the current value of the current sensor at the current moment, and determine the evaluation value of the current sensor under the current quality index based on the current signal-to-noise ratio and the historical signal-to-noise ratio of the current sensor at a historical moment.

[0150] In some embodiments, the fusion module is used to select current sensors with a confidence level greater than a preset confidence threshold as target sensors; for each target sensor, determine the confidence weight of the target sensor based on the confidence level of the target sensor; and perform fusion processing on the current values ​​of each target sensor according to the current values ​​of different target sensors and the corresponding confidence weights of the target sensors to obtain the target fused current.

[0151] In some embodiments, the fusion module is configured to output a sensor anomaly alarm when the confidence levels of all current sensors are not greater than a preset confidence threshold.

[0152] In some embodiments, the detection module is used to determine that a current sensor with a drastic fluctuation in confidence level at the current moment has failed if the target fusion current is not greater than a preset current threshold; and to determine that the circuit under test has failed if the target fusion current is greater than the preset current threshold and the number of current sensors with drastic fluctuation in confidence level at the current moment is greater than a preset number threshold.

[0153] Each module in the aforementioned current sensor fault detection device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0154] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 6As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network. When executed by the processor, the computer program implements a current sensor fault detection method.

[0155] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0156] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0157] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0158] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0159] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0160] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0161] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0162] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for detecting faults in a current sensor, characterized in that, The method includes: Obtain the current value detected by at least one current sensor on the circuit under test at the current moment; For each current sensor, the confidence level of the current sensor at the current moment is determined based on the current value of the current sensor at the current moment. Based on the relationship between the confidence levels of different current sensors at the current moment and the preset confidence threshold, the current values ​​of different current sensors are fused to obtain the target fused current. Based on the confidence fluctuation of each current sensor at the current moment, and the relationship between the target fused current and the preset current threshold, the fault condition of each current sensor is determined.

2. The method according to claim 1, characterized in that, The step of determining the confidence level of the current sensor at the current moment based on the current value of the current sensor at the current moment includes: Based on the current value of the current sensor at the current moment, determine the evaluation value of the current sensor under at least one evaluation index; the evaluation index includes current consistency index, timing consistency index, and current quality index. The evaluation values ​​under different evaluation indicators are spliced ​​together to obtain the target evaluation data; Based on the confidence assessment model, the confidence level of the current sensor at the current moment is determined according to the target assessment data; the confidence assessment model is obtained by training a long short-term memory network model.

3. The method according to claim 2, characterized in that, The step of determining the evaluation value of the current sensor under at least one evaluation index based on the current value of the current sensor at the current moment includes: Based on the current value of the current sensor at the current moment, the median and standard deviation of the corresponding current values ​​of different current sensors, determine the evaluation value of the current sensor under the current consistency index; and, Based on the current value of the current sensor at the current moment and the predicted value of the current sensor at future moments, an evaluation value of the current sensor under a timing consistency index is determined; the predicted value is obtained based on the current prediction model predicting the current value of the current sensor at historical moments; and... Based on the current value of the current sensor at the current moment, the current signal-to-noise ratio of the current sensor at the current moment is determined, and based on the current signal-to-noise ratio and the historical signal-to-noise ratio of the current sensor at historical moments, the evaluation value of the current sensor under the current quality index is determined.

4. The method according to claim 1, characterized in that, The step of fusing the current values ​​of different current sensors based on the relationship between their confidence levels at the current moment and a preset confidence threshold to obtain the target fused current includes: A current sensor with a confidence level greater than a preset confidence threshold is selected as the target sensor. For each target sensor, a confidence weight is determined based on the confidence level of the target sensor. Based on the current values ​​of different target sensors and their corresponding confidence weights, the current values ​​of each target sensor are fused to obtain the target fused current.

5. The method according to claim 4, characterized in that, The method further includes: If the confidence levels of all current sensors are not greater than the preset confidence threshold, an alarm for sensor abnormality will be output.

6. The method according to any one of claims 1-4, characterized in that, The step of determining the fault status of each current sensor based on the confidence fluctuation of each current sensor at the current moment and the relationship between the target fused current and the preset current threshold includes: If the target fusion current is not greater than the preset current threshold, it is determined that the current sensor with drastic fluctuation in confidence at the current moment has failed. If the target fusion current is greater than the preset current threshold, and the number of current sensors with drastic confidence fluctuations at the current moment is greater than the preset number threshold, then the circuit under test is determined to be faulty.

7. A current sensor fault detection device, characterized in that, The device includes: The acquisition module is used to acquire the current value detected by at least one current sensor on the circuit under test at the current moment; The confidence module is used to determine the confidence level of each current sensor at the current moment based on the current value of the current sensor at the current moment. The fusion module is used to fuse the current values ​​of different current sensors based on the relationship between the confidence level of different current sensors at the current moment and the preset confidence threshold, so as to obtain the target fused current. The detection module is used to determine the fault status of each current sensor based on the confidence fluctuation of each current sensor at the current moment and the relationship between the target fused current and the preset current threshold.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images