A health state detection method, device, equipment and medium

Abstract

A health state detection method, device and equipment and medium are related to the field of power electronics. A plurality of real-time operation signals of a to-be-detected device and an operation condition parameter of the to-be-detected device are acquired; a plurality of real-time operation characteristics are determined according to the plurality of real-time operation signals; a plurality of reference operation characteristics corresponding to the operation condition parameter are determined from a preset health condition reference library; a plurality of health deviation characteristics are determined according to the deviation degree between each real-time operation characteristic and the corresponding reference operation characteristic; a fusion weight of each corresponding health deviation characteristic is determined according to the credibility of each real-time operation signal; the plurality of health deviation characteristics are weighted and spliced according to the corresponding fusion weight to obtain a fusion deviation characteristic; the fusion deviation characteristic is input into a state detection model, and an abnormal state, a fault type, a fault position and a health level of the to-be-detected device are acquired through the state detection model. Thus, the detection accuracy of the health state can be improved.

Description

Technical Field

[0001] This application relates to the field of power electronics technology, and in particular to a method, apparatus, equipment and medium for detecting health status. Background Technology

[0002] In industrial production processes, power electronic devices such as solid-state transformers, converters, inverters, rectifiers, and energy storage converters typically use sensors to collect real-time operating signals during operation to reflect the health status of the equipment under test.

[0003] However, in real-world operating environments, sensors may be affected by factors such as long-term operation, changes in ambient temperature, electromagnetic interference, and communication anomalies, leading to abnormalities such as sensor drift, decreased measurement accuracy, response lag, increased signal noise, and signal distortion. These anomalies can cause the acquired real-time operating signals to fail to accurately reflect the actual operating characteristics of the device under test, resulting in low accuracy in health status detection.

[0004] Furthermore, power electronic equipment typically operates under various conditions, and its operating characteristics are closely related to parameters such as load rate, input voltage, output voltage, output power, switching frequency, and ambient temperature. Even if the equipment under test is in a healthy state, its operating characteristics, such as voltage, current, power, temperature, and control quantities, will change significantly when the operating conditions change. If fixed reference operating characteristics or fixed judgment thresholds are directly used to judge real-time operating characteristics, normal changes in operating conditions can easily be misjudged as equipment abnormalities, resulting in false alarms and consequently, low accuracy in health status detection.

[0005] Therefore, how to reduce the impact of sensor anomalies on health status detection and avoid false alarms caused by changes in operating characteristics due to different operating conditions, thereby improving the accuracy of health status detection, is a technical problem that needs to be solved in this field. Summary of the Invention

[0006] To address the aforementioned issues, this application provides a method, apparatus, device, and medium for detecting health status, which can reduce the impact of sensor malfunctions on health status detection and avoid false alarms caused by changes in characteristics due to different operating conditions, thereby improving the accuracy of health status detection.

[0007] The embodiments of this application disclose the following technical solutions: In a first aspect, this application discloses a method for detecting a health status, the method comprising: Acquire multiple real-time operating signals of the device under test and the operating condition parameters of the device under test; the multiple real-time operating signals include at least two of electrical signals, control drive signals, and temperature status signals. Based on the multiple real-time operation signals, several real-time operation characteristics are determined; From a preset health condition reference library, determine multiple reference operating characteristics corresponding to the operating condition parameters; The multiple real-time operating features are compared with the multiple reference operating features in a one-to-one correspondence manner, and multiple health deviation features are determined based on the degree of deviation between each real-time operating feature and the corresponding reference operating feature; Based on the reliability of each real-time operating signal, the fusion weight of each corresponding health deviation feature is determined; the reliability is determined at least based on the degree to which the physical constraint relationship between the real-time operating signal and the associated operating signal is satisfied, and the associated operating signal is an operating signal that has a physical correlation with the real-time operating signal; The multiple health deviation features are weighted and concatenated according to their corresponding fusion weights to obtain the fused deviation features; The fusion deviation features are input into the state detection model, and the abnormal state, fault type, fault location and health level of the device under test are obtained through the state detection model.

[0008] Optionally, the determination of multiple health deviation features based on the degree of deviation between each real-time operating feature and its corresponding reference operating feature includes: From the preset health condition reference library, determine the reference fluctuation range corresponding to each of the reference operating characteristics; Multiple health deviation features are determined based on the degree of deviation between the multiple real-time operating features and the corresponding reference fluctuation range.

[0009] Optionally, determining multiple reference operating characteristics corresponding to the operating condition parameters from a preset health condition reference library includes: If, according to the health condition reference library, the operating condition parameter is located between the operating condition parameters corresponding to two adjacent health conditions, interpolation is performed according to the multiple reference operating features corresponding to the two adjacent health conditions to determine the multiple reference operating features corresponding to the operating condition parameter.

[0010] Optionally, the reliability of the real-time operation signal is determined as follows: Determine the data integrity score, noise level score, sensor drift score, and physical consistency score of the real-time operating signal; The reliability of the real-time operating signal is determined based on the data integrity score, the noise level score, the sensor drift score, and the physical consistency score. The data integrity score of the real-time operating signal is determined based on the ratio of the number of valid sampling points to the theoretical number of sampling points within a preset time window. The noise level score of the real-time operating signal is determined based on the difference between the current noise level of the real-time operating signal within a preset time window and the normal noise level corresponding to the operating condition parameter. The sensor drift score of the real-time operating signal is determined based on the deviation between the measured value of the real-time operating signal and the estimated value determined based on the associated operating signal. The physical consistency score of the real-time operation signal is determined based on the degree to which the physical constraint relationship between the real-time operation signal and the associated operation signal is satisfied. The preset physical constraint relationships include at least one of the following: input-output power balance relationship, node current balance relationship, DC capacitor current and capacitor voltage change relationship, high-frequency transformer primary and secondary side relationship, and dual active bridge converter shift ratio and transmission power relationship.

[0011] Secondly, this application discloses a health status detection device, the device comprising: The signal acquisition module is used to acquire multiple real-time operating signals of the device under test and the operating condition parameters of the device under test; the multiple real-time operating signals include at least two of electrical signals, control drive signals and temperature status signals. The first determining module is used to determine multiple real-time operating features based on the multiple real-time operating signals; The second determining module is used to determine multiple reference operating features corresponding to the operating condition parameters from a preset health condition reference library; The third determining module is used to compare the multiple real-time operating features with the multiple reference operating features in a one-to-one correspondence manner, and determine multiple health deviation features based on the degree of deviation between each real-time operating feature and the corresponding reference operating feature; The weight determination module is used to determine the fusion weight of each corresponding health deviation feature based on the credibility of each real-time operation signal; the credibility is determined at least based on the degree to which the physical constraint relationship between the real-time operation signal and the associated operation signal is the operation signal that has a physical correlation with the real-time operation signal; The feature acquisition module is used to weight and concatenate the multiple health deviation features according to the corresponding fusion weights to obtain the fused deviation features; The status acquisition module is used to input the fusion deviation features into the status detection model, and obtain the abnormal status, fault type, fault location and health level of the device under test through the status detection model.

[0012] Optionally, the third determining module is specifically used to: determine the reference fluctuation range corresponding to each of the reference operating characteristics from a preset health condition reference library; and determine multiple health deviation characteristics based on the degree of deviation between the multiple real-time operating characteristics and the corresponding reference fluctuation ranges.

[0013] Optionally, the second determining module is specifically used to: if, according to the health condition reference library, the operating condition parameter is located between the operating condition parameters corresponding to two adjacent health conditions, interpolate according to the multiple reference operating features corresponding to the two adjacent health conditions to determine the multiple reference operating features corresponding to the operating condition parameter.

[0014] Optionally, the weight determination module is specifically used to: determine the data integrity score, noise level score, sensor drift score, and physical consistency score of the real-time operating signal; determine the reliability of the real-time operating signal based on the data integrity score, the noise level score, the sensor drift score, and the physical consistency score; the data integrity score of the real-time operating signal is determined based on the ratio of the number of effective sampling points to the theoretical number of sampling points within a preset time window; the noise level score of the real-time operating signal is determined based on the difference between the current noise level of the real-time operating signal within the preset time window and the normal noise level corresponding to the operating condition parameter; the sensor drift score of the real-time operating signal is determined based on the deviation between the measured value of the real-time operating signal and the estimated value determined based on the associated operating signal; the physical consistency score of the real-time operating signal is determined based on the degree to which the physical constraint relationship between the real-time operating signal and the associated operating signal is satisfied; the preset physical constraint relationship includes at least one of the following: input-output power balance relationship, node current balance relationship, DC capacitor current and capacitor voltage change relationship, high-frequency transformer primary-secondary side relationship, and dual active bridge converter shift ratio and transmission power relationship.

[0015] Thirdly, this application discloses a health status detection device, the device comprising: a memory and a processor; The memory is used to store programs; The processor is configured to execute the program to implement the various steps of the health status detection method as described in the first aspect.

[0016] Fourthly, this application discloses a computer-readable medium having a computer program stored thereon, which, when executed by a processor, implements the various steps of the health status detection method as described in the first aspect.

[0017] Compared with the prior art, this application has the following beneficial effects: This application provides a method, apparatus, device, and medium for detecting health status. Firstly, this application acquires multiple real-time operating signals from the device under test, including at least two of electrical signals, control drive signals, and temperature status signals. Utilizing the redundancy and complementarity among these signals, even when a sensor experiences abnormalities such as sensor drift, decreased measurement accuracy, response lag, increased signal noise, or signal distortion, the other real-time operating signals can still provide effective health status judgment. Simultaneously, this application determines multiple health deviation characteristics based on the degree of deviation between multiple real-time operating characteristics and corresponding multiple reference operating characteristics, and determines each corresponding health deviation characteristic based on the reliability of each real-time operating signal. The fusion weights of multiple health deviation features are then weighted and concatenated according to their corresponding fusion weights to obtain fused deviation features. This allows health deviation features corresponding to low-reliability real-time operating signals (i.e., abnormal signals) to be dynamically downweighted during the fusion process, thereby suppressing the contamination of fused deviation features by sensor anomalies and making the fused deviation features more accurately and stably reflect the true health status of the device under test. Finally, the fused deviation features are input into the status detection model, which makes judgments based on the fused deviation features after anomaly suppression and multi-channel information fusion, thereby reducing the impact of sensor anomalies on health status monitoring results and improving the accuracy of health status detection. Secondly, this application determines multiple reference operating characteristics corresponding to the current operating condition from a preset health condition reference library based on the operating condition parameters of the equipment under test. It also determines multiple health deviation characteristics based on the degree of deviation between the multiple real-time operating characteristics and the corresponding multiple reference operating characteristics. This allows the health deviation characteristics to reflect the degree of deviation of the equipment under test from the normal state under the same operating conditions, rather than the absolute degree of deviation from a fixed reference value. This avoids misjudging the equipment as abnormal due to changes in real-time operating characteristics caused by changes in operating conditions such as load rate, input voltage, output voltage, output power, switching frequency, and ambient temperature. It also reduces false alarms caused by changes in operating conditions and further improves the accuracy of health status detection. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the 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.

[0019] Figure 1 A flowchart illustrating a health status detection method provided in an embodiment of this application; Figure 2 A schematic diagram of a health status detection device provided in an embodiment of this application; Figure 3 This is a schematic diagram of a computer-readable medium provided in an embodiment of this application. Detailed Implementation

[0020] As described earlier, in actual operating environments, sensors may be affected by factors such as long-term operation, changes in ambient temperature, electromagnetic interference, and communication anomalies, leading to abnormal conditions such as sensor drift, decreased measurement accuracy, response lag, increased signal noise, and signal distortion. These anomalies can cause the collected real-time operating signals to fail to accurately reflect the actual operating characteristics of the device under test, resulting in low accuracy in health status detection.

[0021] Furthermore, power electronic equipment typically operates under various conditions, and its operating characteristics are closely related to parameters such as load rate, input voltage, output voltage, output power, switching frequency, and ambient temperature. Even if the equipment under test is in a healthy state, its operating characteristics, such as voltage, current, power, temperature, and control quantities, will change significantly when the operating conditions change. If fixed reference operating characteristics or fixed judgment thresholds are directly used to judge real-time operating characteristics, normal changes in operating conditions can easily be misjudged as equipment abnormalities, resulting in false alarms and consequently, low accuracy in health status detection.

[0022] Therefore, how to reduce the impact of sensor anomalies on health status detection and avoid false alarms caused by changes in operating characteristics due to different operating conditions, thereby improving the accuracy of health status detection, is a technical problem that needs to be solved in this field.

[0023] The inventors, through research, have proposed a method, apparatus, device, and medium for detecting health status. Firstly, this application acquires multiple real-time operating signals from the device under test, including at least two of electrical signals, control drive signals, and temperature status signals. Utilizing the redundancy and complementarity among these signals, even when a sensor experiences abnormalities such as sensor drift, decreased measurement accuracy, response lag, increased signal noise, or signal distortion, the other real-time operating signals can still provide effective health status judgment. Simultaneously, this application determines multiple health deviation characteristics based on the degree of deviation between multiple real-time operating characteristics and corresponding multiple reference operating characteristics, and determines each corresponding health deviation characteristic based on the reliability of each real-time operating signal. The fusion weights of multiple health deviation features are then weighted and concatenated according to their corresponding fusion weights to obtain fused deviation features. This allows health deviation features corresponding to low-reliability real-time operating signals (i.e., abnormal signals) to be dynamically downweighted during the fusion process, thereby suppressing the contamination of fused deviation features by sensor anomalies and making the fused deviation features more accurately and stably reflect the true health status of the device under test. Finally, the fused deviation features are input into the status detection model, which makes judgments based on the fused deviation features after anomaly suppression and multi-channel information fusion, thereby reducing the impact of sensor anomalies on health status monitoring results and improving the accuracy of health status detection. Secondly, this application determines multiple reference operating characteristics corresponding to the current operating condition from a preset health condition reference library based on the operating condition parameters of the equipment under test. It also determines multiple health deviation characteristics based on the degree of deviation between the multiple real-time operating characteristics and the corresponding multiple reference operating characteristics. This allows the health deviation characteristics to reflect the degree of deviation of the equipment under test from the normal state under the same operating conditions, rather than the absolute degree of deviation from a fixed reference value. This avoids misjudging the equipment as abnormal due to changes in real-time operating characteristics caused by changes in operating conditions such as load rate, input voltage, output voltage, output power, switching frequency, and ambient temperature. It also reduces false alarms caused by changes in operating conditions and further improves the accuracy of health status detection.

[0024] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0025] See Figure 1The figure is a flowchart of a health status detection method provided in an embodiment of this application. This method is applied to power electronic equipment such as solid-state transformers, converters, inverters, rectifiers, and energy storage converters. The method includes: S101: Acquire multiple real-time operating signals of the device under test and the operating condition parameters of the device under test; the multiple real-time operating signals include at least two of the electrical signals, control drive signals and temperature status signals.

[0026] In this embodiment, the multiple real-time operating signals include at least two of the following: electrical signals, control drive signals, and temperature status signals. Exemplarily, the electrical signals include one or more of the following: input voltage, output voltage, input current, output current, module voltage, module current, input power, output power, harmonic signals, and ripple signals. The control drive signals include one or more of the following: drive pulses, switching states, shift ratios, modulation ratios, control setpoints, control feedback values, and control errors. The temperature status signals include one or more of the following: power device temperatures, magnetic component temperatures, and heat sink temperatures.

[0027] It should be noted that this application acquires at least two of the electrical signals, control drive signals, and temperature status signals, rather than relying solely on a single real-time operating signal for threshold judgment. This is to utilize the redundancy and complementarity among multiple real-time operating signals. When a sensor experiences abnormal conditions such as sensor drift, decreased measurement accuracy, response lag, increased signal noise, or signal distortion, other real-time operating signals can still provide effective health status judgment. When the operating conditions of the device under test change, multiple real-time operating signals can also jointly reflect the true state of the device under test from different dimensions such as electrical, control, and thermal status, thereby reducing false alarms caused by changes in operating conditions affecting a single real-time operating signal.

[0028] In this embodiment of the application, the operating condition parameters are used to determine the current operating conditions of the device under test, including one or more of the following: load rate, input voltage, output voltage, output power, switching frequency, and ambient temperature.

[0029] For example, when the device under test is a three-phase CHB+DAB solid-state transformer, the operating parameters may include: load rate 50%, input voltage 10kV, and output power 500kW.

[0030] S102: Determine multiple real-time operating characteristics based on multiple real-time operating signals.

[0031] In this embodiment, multiple real-time operating signals are time-aligned, preprocessed, and feature-extracted to determine multiple real-time operating features. These multiple real-time operating features are used to characterize at least two of the electrical, control, and thermal states of the device under test.

[0032] First, since the sampling frequencies of multiple real-time operating signals are often different, it is necessary to time-align these signals, mapping electrical signals, control drive signals, and temperature status signals to a unified time base to obtain time-aligned real-time operating signals. Therefore, time alignment allows subsequent calculations of health deviation characteristics to be based on operating information at the same time scale, avoiding erroneous judgments caused by asynchronous signal sampling.

[0033] Subsequently, the time-aligned multi-channel real-time operating signals are preprocessed to obtain preprocessed multi-channel real-time operating signals. Preprocessing includes at least one of the following: filtering, outlier removal, missing data identification, sampling frequency unification, and data normalization. Specifically, filtering is used to suppress high-frequency noise and measurement glitches in the time-aligned multi-channel real-time operating signals; outlier removal is used to remove sampling points that significantly deviate from physical laws; missing data identification is used to mark packet loss, breakpoints, or invalid samples in the time-aligned multi-channel real-time operating signals; sampling frequency unification is used to resample the time-aligned multi-channel real-time operating signals from different sources; and data normalization is used to eliminate the influence of different dimensions and orders of magnitude (such as voltage, current, temperature, and shift ratio) on subsequent calculations in the time-aligned multi-channel real-time operating signals.

[0034] Finally, feature extraction is performed on the preprocessed multi-channel real-time operating signals to obtain multiple real-time operating features. Feature extraction can employ methods such as time-domain statistics, frequency-domain analysis, and time-frequency-domain analysis. For example, for preprocessed electrical signals, one or more of the following can be extracted: RMS voltage, RMS current, peak value, average value, power, harmonic content, ripple amplitude, module voltage deviation, and current imbalance. For preprocessed control drive signals, one or more of the following can be extracted: drive pulse duty cycle, switching state, shift ratio, modulation ratio, control setpoint, control feedback value, and control error. For preprocessed temperature status signals, one or more of the following can be extracted: power device temperature, magnetic component temperature, heat sink temperature, temperature rise rate, temperature gradient, and inter-module temperature difference.

[0035] S103: Determine multiple reference operating characteristics corresponding to the operating condition parameters from the preset health condition reference library.

[0036] The health condition reference library is built based on historical operating data collected under different operating conditions when the device under test is in a healthy state. The library stores reference operating characteristics (e.g., reference values) and reference fluctuation ranges corresponding to different operating conditions.

[0037] For example, a health condition reference library can be shown in Table 1 below: Table 1

[0038] As shown in Table 1, when the operating conditions are 50% load, 10kV input voltage, and 500kW output power, operating condition G2 can be matched from the preset health condition reference library, and several reference operating characteristics corresponding to the operating conditions can be determined, including: silicon carbide device temperature reference value and fluctuation range 75℃±5℃, and input current reference value and fluctuation range 3.0%±0.8%.

[0039] It should be noted that if the operating condition parameters of the device under test are determined to be between the operating condition parameters corresponding to two adjacent healthy operating conditions according to the health condition reference library, interpolation is performed according to the multiple reference operating characteristics corresponding to the two adjacent healthy operating conditions to determine the multiple reference operating characteristics corresponding to the operating condition parameters.

[0040] As shown in Table 1, when the operating parameters are a load rate of 75%, an input voltage of 10kV, and an output power of 750kW, operating conditions G2 and G3 can be matched from the preset health operating condition reference library. At this time, multiple reference operating characteristics corresponding to G2 and G3 can be interpolated to obtain multiple reference operating characteristics corresponding to the current operating condition, such as the silicon carbide device temperature reference value and fluctuation range of 85℃±5℃, and the input current reference value and fluctuation range of 2.75%±0.65%.

[0041] S104: Compare multiple real-time operating features with multiple reference operating features in a one-to-one correspondence manner, and determine multiple health deviation features based on the degree of deviation between each real-time operating feature and its corresponding reference operating feature.

[0042] Health deviation features are used to characterize the degree of deviation of the real-time operating state of the device under test from its healthy state. In one specific implementation, multiple health deviation features can be determined based on the degree of deviation between multiple real-time operating features and corresponding multiple reference operating features (e.g., reference values). In another specific implementation, a reference fluctuation range corresponding to each reference operating feature can first be determined from a preset health condition reference library; subsequently, multiple health deviation features can be determined based on the degree of deviation between multiple real-time operating features and their corresponding reference fluctuation ranges.

[0043] It should be noted that this application matches operating condition parameters to a health condition reference library and determines the reference operating characteristics corresponding to the current operating condition. Then, it compares the real-time operating characteristics with the reference operating characteristics to obtain health deviation characteristics, rather than directly using a fixed threshold to determine whether the device under test is abnormal. Therefore, this application can reduce the interference of changes in operating conditions such as load rate, input / output voltage, output power, and ambient temperature on fault diagnosis, making the health deviation characteristics more reflective of changes in the device's own health status, reducing false alarms caused by changes in normal operating conditions, and facilitating the identification of early minor anomalies.

[0044] S105: Determine the fusion weight of each corresponding health deviation feature based on the reliability of each real-time operation signal; the reliability is determined at least based on the degree to which the physical constraint relationship between the real-time operation signal and the associated operation signal is satisfied, and the associated operation signal is the operation signal that has a physical relationship with the real-time operation signal.

[0045] In this embodiment, a physical consistency score is first determined; the physical consistency score characterizes whether the real-time operation signal and the associated operation signal satisfy a preset physical constraint relationship. Then, based on the physical consistency score, the reliability of the real-time operation signal is determined. Finally, based on the reliability of each real-time operation signal, the fusion weight of each corresponding health deviation feature is determined.

[0046] Specifically, the preset physical constraint relationships include at least one of the following: First, the input-output power balance relationship. Specifically, the power balance residual is determined by the difference between the sum of the output power and the normal losses of the equipment and the input power. If the power balance residual is small, for example, less than the preset difference threshold C3, the physical consistency score of the relevant signal is high; if the power balance residual is large, for example, greater than the preset difference threshold C3, the physical consistency score of the relevant signal is low.

[0047] Second, the node current balance relationship. Specifically, firstly, the current balance residual is determined based on the current flowing into the electrical node and the sum of the currents of each branch flowing out of the electrical node. If the current balance residual is zero, it indicates that the relevant current signals have good consistency and the physical consistency score of the relevant signals is high; if the current balance residual is large, for example, greater than the preset difference threshold C4, the physical consistency score of the relevant signals decreases.

[0048] Third, the relationship between DC capacitor current and capacitor voltage changes. Specifically, when the net current flowing into the DC capacitor is positive, the capacitor voltage usually shows an upward trend; when the net current flowing into the DC capacitor is negative, the capacitor voltage usually shows a downward trend. Therefore, if the current direction is consistent with the trend of capacitor voltage change, the physical consistency score of the related signal is higher; if the current direction is inconsistent with the trend of capacitor voltage change, the physical consistency score of the related signal is lower.

[0049] Fourth, the relationship between the primary and secondary sides of the high-frequency transformer. Specifically, based on the transformer turns ratio, the primary and secondary voltages are cross-checked; simultaneously, power verification is performed based on the relationship between primary-side power transmission, secondary-side power transmission, and transformer losses. If the preset turns ratio or power transmission relationship is met, the physical consistency score of the relevant signals is high; if the preset turns ratio or power transmission relationship is not met, the physical consistency score of the relevant signals is low.

[0050] Fifth, the relationship between the phase shift and the transmitted power of the dual active bridge converter. Specifically, within the preset operating range, the change in the phase shift of the dual active bridge converter should correspond to the actual change in transmitted power, and the phase shift direction should generally match the power transmission direction. If the phase shift signal and the transmitted power signal are consistently consistent, the physical consistency score of the related signals is high; if the phase shift signal and the transmitted power signal are consistently inconsistent, the physical consistency score of the related signals is low.

[0051] In this embodiment of the application, the fusion weight of each corresponding health deviation feature can also be determined through the following steps A1-A3: A1: Determine the data integrity score, noise level score, sensor drift score, and physical consistency score of the real-time operating signal.

[0052] The data integrity score is used to characterize the proportion of valid data in real-time operating signals within a preset time window T1.

[0053] Specifically, firstly, within a preset time window T1, the theoretical number N of sampling points to be sampled at a preset sampling frequency according to the real-time operating signal is determined, and the number M of valid sampling points that are actually acquired and meet the validity requirements is determined. Then, based on the theoretical number of sampling points N and the number of valid sampling points M, the data integrity score S is determined. c =M / N.

[0054] Understandably, if the real-time signal experiences continuous packet loss, communication interruption, abnormal timestamps, or invalid sample values ​​(i.e., invalid sampling), the number of valid sample points M will decrease accordingly, and the data integrity score S will decrease. c reduce.

[0055] It should be noted that if the number of valid sampling points M within the preset time window T1 is lower than the preset lower threshold, the data integrity of the real-time operating signal can be determined to be unqualified, and the data integrity score S will be reduced accordingly. c Set to zero. Alternatively, if the duration of invalid sampling within the preset time window T1 exceeds the preset duration threshold, the data integrity of the real-time operating signal can be determined to be unqualified, and the data integrity score S will be set to zero. c Set to zero. This application does not impose any restrictions on this.

[0056] The noise level score is used to characterize the degree of random fluctuations and high-frequency disturbances in the real-time operating signal.

[0057] Specifically, firstly, the current noise level of the real-time operating signal within a preset time window T2 is determined using any one of the following methods: signal standard deviation, variance, high-frequency energy, spectral characteristics, wavelet coefficients, or the difference between the signal before and after filtering. Then, the normal noise level corresponding to the operating parameters of the device under test is obtained from the health condition reference library. Next, the current noise level is compared with the normal noise level. If the difference between the current noise level and the normal noise level is less than a preset difference threshold C1, the noise of the real-time operating signal is considered to be within the normal range, and the noise level score is high. If the difference between the current noise level and the normal noise level is greater than the preset difference threshold C1, the real-time operating signal is considered to be possibly affected by sensor noise, communication interference, switching interference, or abnormal acquisition circuitry, and the noise level score is lowered.

[0058] Sensor drift score is used to characterize whether there is a persistent offset in the measured values ​​of real-time operating signals.

[0059] Specifically, firstly, the estimated value of the real-time operating signal is determined based on the associated operating signals that are related to it. Then, the measured value of the real-time operating signal is compared with the estimated value to obtain the measurement deviation. If the measurement deviation is small within a preset time window T3, for example, if the deviation between the measured value and the estimated value is less than a preset difference threshold C2, it indicates that the real-time operating signal has not experienced significant drift, and the sensor drift score is high. If the measurement deviation persists for multiple consecutive preset time windows, and the direction of the measurement deviation remains consistent or the measurement deviation value gradually increases, it indicates that the sensor may be experiencing slow drift, and the sensor drift score decreases.

[0060] The physical consistency score is used to characterize whether the real-time operation signal and the associated operation signal meet the preset physical constraint relationship.

[0061] It should be noted that, for ease of calculation, the above data integrity score, noise level score, sensor drift score, and physical consistency score can all be mapped to the range of 0-100.

[0062] A2: Determine the reliability of the real-time operating signal based on the data integrity score, noise level score, sensor drift score, and physical consistency score.

[0063] First, the data integrity score, noise level score, sensor drift score, and physical consistency score are weighted according to preset weights to obtain the credibility of the real-time operating signal.

[0064] It should be noted that the above preset weights can be set according to the signal type, application scenario, and fault mechanism. For example, for electrical signals, the weights of data integrity and physical consistency can be appropriately increased; for temperature signals, the weights of noise level and sensor drift can be appropriately increased.

[0065] Subsequently, to reduce the impact of transient noise, occasional packet loss, or short-term physical imbalances on credibility, multiple credibility scores obtained within the continuous analysis time window are time-smoothed to obtain the credibility score of the real-time running signal used to determine the fusion weights. Time smoothing can employ methods such as moving average, exponential weighted average, median filtering, or low-pass filtering to suppress the impact of transient anomalies on credibility and improve the stability of the credibility evaluation.

[0066] A3: Determine the fusion weight of each corresponding health deviation feature based on the reliability of each real-time operating signal.

[0067] The reliability of each real-time operating signal is normalized to obtain the corresponding fusion weight. The sum of the fusion weights for each real-time operating signal is a preset value, such as 1 or 100%. For example, if the reliability of the electrical signal, control drive signal, and temperature signal are 90, 80, and 50 points respectively, the total reliability is 220 points, and the corresponding fusion weights are approximately 40.9%, 36.4%, and 22.7%. If the health deviation characteristics corresponding to the three types of signals are 1.2, 0.8, and 1.5 respectively, the weighted health deviation characteristics are approximately 0.49, 0.29, and 0.34 respectively.

[0068] It should be noted that this application comprehensively evaluates the reliability of each real-time operating signal from multiple dimensions, including data integrity, noise level, sensor drift, and physical consistency, and determines the fusion weights based on the reliability, rather than assuming that all sensors are in a reliable state. Therefore, when a real-time operating signal experiences increased noise, missing data, measurement saturation, slow drift, or failure, the corresponding health deviation characteristics of that signal will be dynamically downweighted in subsequent fusion processes, thereby avoiding excessive influence of low-quality or abnormal signals on the health status diagnosis results. Even if some sensors malfunction, other real-time operating signals with higher reliability can still participate in health status assessment, improving the stability and accuracy of health status monitoring.

[0069] S106: Weight and concatenate multiple health deviation features according to their corresponding fusion weights to obtain fused deviation features.

[0070] After determining the fusion weight corresponding to each real-time operating signal in step S105, the multiple health deviation features obtained in step S104 are multiplied by the corresponding fusion weight according to their respective signal categories to obtain multiple weighted health deviation features.

[0071] Subsequently, the weighted health deviation features are spliced ​​together in a preset fixed order to form a fused deviation feature. The fused deviation feature includes at least two of the following: electrical health deviation features, control drive health deviation features, and temperature health deviation features spliced ​​together in a preset fixed order.

[0072] It should be noted that when the reliability of any real-time operating signal decreases, the fusion weight of the corresponding health deviation feature in the fused deviation feature is reduced; when the reliability of any real-time operating signal is zero, the weighted health deviation feature corresponding to that real-time operating signal is set to zero. Therefore, this application can dynamically adjust the fusion weights of different types of health deviation features, such as electrical, control drive, and temperature features, based on signal reliability. This ensures that signals with higher reliability have a greater impact on the diagnostic results, while reducing the impact of signals with lower reliability or failure on the diagnostic results. Even in the event of sensor drift, increased noise, or failure, it can still maintain a relatively stable health status monitoring capability and reduce the contamination of the fused deviation feature by sensor anomalies.

[0073] S107: Input the fusion deviation features into the state detection model, and obtain the abnormal state, fault type, fault location and health level of the device under test through the state detection model.

[0074] In this embodiment, the state detection model is a neural network model, comprising an input layer, a normalization layer, a first hidden layer, a second hidden layer, a comprehensive health feature layer, and a diagnostic output layer connected in sequence. The input layer receives fusion deviation features. The normalization layer unifies the scale of different types of fusion deviation features. The first hidden layer learns the combination relationships between different fusion deviation features. Faults in power electronic equipment are often not isolated changes of a single feature, but rather the result of multiple features changing synergistically. For example, an open-circuit fault in a power semiconductor device may simultaneously cause module current distortion, module power reduction, inconsistency between drive commands and actual electrical responses, and abnormal local temperature rise. The first hidden layer uses a nonlinear activation function to combine and map the input features, enabling the extraction of correlations between different health deviation features. The second hidden layer further extracts higher-level features related to the fault state from the fusion deviation features. Based on the output of the first hidden layer, the second hidden layer can form a more abstract and discriminative fault feature representation. The comprehensive health feature layer forms feature information characterizing the overall operating state of the device under test. The diagnostic output layer can adopt a multi-task output structure, outputting the abnormal state, fault type, fault location, and health level respectively.

[0075] For example, the fault types include at least one of the following: power semiconductor device open circuit fault, high frequency transformer malfunction, sensor drift fault, sensor failure fault, and inter-module power imbalance fault.

[0076] For example, the fault location includes at least one of a power module, a cascaded H-bridge unit, a dual active bridge unit, a high-frequency transformer, and a sensor.

[0077] For example, the health level includes four levels: healthy, sub-healthy, warning, and faulty.

[0078] It should be noted that the state detection model is built using offline training and online detection. Offline training includes: B1: Obtain healthy operation samples and samples of different fault states, and label the healthy operation samples and samples of different fault states with at least one of the following: abnormal state label, fault type label, fault location label, and health level label.

[0079] It should be noted that different fault state samples can be obtained through simulation, fault injection on experimental platforms, or actual fault records, covering faults such as open circuit of power semiconductor devices, abnormal high-frequency transformers, sensor drift, sensor failure, and power imbalance between modules.

[0080] B2: Use labeled samples to train the neural network model to obtain the state detection model.

[0081] It should be noted that during training, the labeled fusion bias features can be input into the neural network model. The predicted output is obtained through forward propagation. A loss function is constructed based on the difference between the predicted output and the labels, and the network parameters are updated using the backpropagation algorithm. The loss function can be cross-entropy loss, mean squared error, or joint loss, etc. For multi-task outputs, the losses for anomaly detection, fault type, fault location, and health level can be calculated separately and then summed with weights. After training, the neural network parameters can be fixed, and the state detection model can be deployed on a state monitoring terminal.

[0082] Online detection includes inputting the currently obtained fusion deviation features into a trained neural network model to obtain the abnormal state, fault type, fault location, and health level of the power electronic equipment to be detected.

[0083] In summary, this application provides a method for detecting health status. Firstly, this application acquires multiple real-time operating signals from the device under test, including at least two of electrical signals, control drive signals, and temperature status signals. Utilizing the redundancy and complementarity among these signals, even when a sensor experiences abnormalities such as sensor drift, decreased measurement accuracy, response lag, increased signal noise, or signal distortion, other real-time operating signals can still provide effective health status judgment. Simultaneously, this application determines multiple health deviation characteristics based on the degree of deviation between multiple real-time operating characteristics and corresponding multiple reference operating characteristics, and determines each corresponding health deviation characteristic based on the reliability of each real-time operating signal. The fusion weights of multiple health deviation features are then weighted and concatenated according to their corresponding fusion weights to obtain fused deviation features. This allows health deviation features corresponding to low-reliability real-time operating signals (i.e., abnormal signals) to be dynamically downweighted during the fusion process, thereby suppressing the contamination of fused deviation features by sensor anomalies and making the fused deviation features more accurately and stably reflect the true health status of the device under test. Finally, the fused deviation features are input into the status detection model, which makes judgments based on the fused deviation features after anomaly suppression and multi-channel information fusion, thereby reducing the impact of sensor anomalies on health status monitoring results and improving the accuracy of health status detection. Secondly, this application determines multiple reference operating characteristics corresponding to the current operating condition from a preset health condition reference library based on the operating condition parameters of the equipment under test. It also determines multiple health deviation characteristics based on the degree of deviation between the multiple real-time operating characteristics and the corresponding multiple reference operating characteristics. This allows the health deviation characteristics to reflect the degree of deviation of the equipment under test from the normal state under the same operating conditions, rather than the absolute degree of deviation from a fixed reference value. This avoids misjudging the equipment as abnormal due to changes in real-time operating characteristics caused by changes in operating conditions such as load rate, input voltage, output voltage, output power, switching frequency, and ambient temperature. It also reduces false alarms caused by changes in operating conditions and further improves the accuracy of health status detection.

[0084] See Figure 2 The figure shows a health status detection device provided in an embodiment of this application. The health status detection device 200 includes: The signal acquisition module 201 is used to acquire multiple real-time operating signals of the device under test and the operating condition parameters of the device under test; the multiple real-time operating signals include at least two of electrical signals, control drive signals and temperature status signals. The first determining module 202 is used to determine multiple real-time operating characteristics based on multiple real-time operating signals; The second determining module 203 is used to determine multiple reference operating features corresponding to the operating condition parameters from a preset health condition reference library; The third determining module 204 is used to compare multiple real-time operating features with multiple reference operating features in a one-to-one correspondence manner, and determine multiple health deviation features based on the degree of deviation between each real-time operating feature and the corresponding reference operating feature. The weight determination module 205 is used to determine the fusion weight of each corresponding health deviation feature based on the credibility of each real-time operation signal; the credibility is determined at least based on the degree to which the physical constraint relationship between the real-time operation signal and the associated operation signal is the operation signal that has a physical correlation with the real-time operation signal. The feature acquisition module 206 is used to weight and concatenate multiple health deviation features according to their corresponding fusion weights to obtain fused deviation features. The status acquisition module 207 is used to input the fusion deviation features into the status detection model, and obtain the abnormal status, fault type, fault location and health level of the device under test through the status detection model.

[0085] In one specific implementation, the third determining module 204 is specifically used to: determine the reference fluctuation range corresponding to each reference operating feature from a preset health condition reference library; and determine multiple health deviation features based on the degree of deviation between multiple real-time operating features and their corresponding reference fluctuation ranges.

[0086] In one specific implementation, the second determining module 203 is specifically used to: if, according to the health condition reference library, the operating condition parameter is located between the operating condition parameters corresponding to two adjacent health conditions, interpolate according to the multiple reference operating features corresponding to the two adjacent health conditions to determine the multiple reference operating features corresponding to the operating condition parameter.

[0087] In one specific implementation, the weight determination module 205 is specifically used to: determine the data integrity score, noise level score, sensor drift score, and physical consistency score of the real-time operating signal; determine the reliability of the real-time operating signal based on the data integrity score, noise level score, sensor drift score, and physical consistency score; the data integrity score of the real-time operating signal is determined based on the ratio of the number of effective sampling points to the theoretical number of sampling points within a preset time window; the noise level score of the real-time operating signal is determined based on the difference between the current noise level of the real-time operating signal within the preset time window and the normal noise level corresponding to the operating condition parameters; the sensor drift score of the real-time operating signal is determined based on the deviation between the measured value of the real-time operating signal and the estimated value determined based on the associated operating signal; the physical consistency score of the real-time operating signal is determined based on the degree to which the physical constraint relationship between the real-time operating signal and the associated operating signal is satisfied; the preset physical constraint relationship includes at least one of the following: input-output power balance relationship, node current balance relationship, DC capacitor current and capacitor voltage change relationship, high-frequency transformer primary-secondary side relationship, and shift ratio and transmission power relationship of dual active bridge converter.

[0088] In summary, this application provides a health status detection device. Firstly, this application acquires multiple real-time operating signals from the device under test, including at least two of electrical signals, control drive signals, and temperature status signals. Utilizing the redundancy and complementarity among these signals, even when a sensor experiences abnormalities such as sensor drift, decreased measurement accuracy, response lag, increased signal noise, or signal distortion, other real-time operating signals can still provide effective health status judgment. Simultaneously, this application determines multiple health deviation characteristics based on the degree of deviation between multiple real-time operating characteristics and corresponding multiple reference operating characteristics, and determines each corresponding health deviation characteristic based on the reliability of each real-time operating signal. The fusion weights of multiple health deviation features are then weighted and concatenated according to their corresponding fusion weights to obtain fused deviation features. This allows health deviation features corresponding to low-reliability real-time operating signals (i.e., abnormal signals) to be dynamically downweighted during the fusion process, thereby suppressing the contamination of fused deviation features by sensor anomalies and making the fused deviation features more accurately and stably reflect the true health status of the device under test. Finally, the fused deviation features are input into the status detection model, which makes judgments based on the fused deviation features after anomaly suppression and multi-channel information fusion, thereby reducing the impact of sensor anomalies on health status monitoring results and improving the accuracy of health status detection. Secondly, this application determines multiple reference operating characteristics corresponding to the current operating condition from a preset health condition reference library based on the operating condition parameters of the equipment under test. It also determines multiple health deviation characteristics based on the degree of deviation between the multiple real-time operating characteristics and the corresponding multiple reference operating characteristics. This allows the health deviation characteristics to reflect the degree of deviation of the equipment under test from the normal state under the same operating conditions, rather than the absolute degree of deviation from a fixed reference value. This avoids misjudging the equipment as abnormal due to changes in real-time operating characteristics caused by changes in operating conditions such as load rate, input voltage, output voltage, output power, switching frequency, and ambient temperature. It also reduces false alarms caused by changes in operating conditions and further improves the accuracy of health status detection.

[0089] This application also provides a corresponding health status detection device and a computer-readable medium for implementing the health status detection method provided in this application.

[0090] The health status detection device includes a memory and a processor. The memory is used to store instructions or code, and the processor is used to execute the instructions or code so that the device performs a health status detection method according to any embodiment of this application.

[0091] See Figure 3This figure is a schematic diagram of a computer-readable medium provided in an embodiment of this application. The computer-readable medium 300 stores a computer program 311, which, when executed by a processor, implements the above-described... Figure 1 The steps of the health status detection method.

[0092] It should be noted that, in the context of this application, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0093] It should be noted that the machine-readable medium described above in this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0094] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device.

[0095] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

[0096] While several specific implementation details are included in the foregoing discussion, these should not be construed as limiting the scope of this application. Certain features described in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments.

[0097] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.

Claims

1. A method for detecting a health status, characterized in that, The method includes: Acquire multiple real-time operating signals of the device under test and the operating condition parameters of the device under test; the multiple real-time operating signals include at least two of electrical signals, control drive signals, and temperature status signals. Based on the multiple real-time operation signals, several real-time operation characteristics are determined; From a preset health condition reference library, determine multiple reference operating features corresponding to the operating condition parameters; The multiple real-time operating features are compared with the multiple reference operating features in a one-to-one correspondence manner, and multiple health deviation features are determined based on the degree of deviation between each real-time operating feature and the corresponding reference operating feature; Based on the reliability of each real-time operating signal, the fusion weight of each corresponding health deviation feature is determined; the reliability is determined at least based on the degree to which the physical constraint relationship between the real-time operating signal and the associated operating signal is satisfied, and the associated operating signal is an operating signal that has a physical correlation with the real-time operating signal; The multiple health deviation features are weighted and concatenated according to their corresponding fusion weights to obtain the fused deviation features; The fusion deviation features are input into the state detection model, and the abnormal state, fault type, fault location and health level of the device under test are obtained through the state detection model.

2. The method according to claim 1, characterized in that, The determination of multiple health deviation features based on the degree of deviation between each real-time operating feature and its corresponding reference operating feature includes: From the preset health condition reference library, determine the reference fluctuation range corresponding to each of the reference operating characteristics; Multiple health deviation features are determined based on the degree of deviation between the multiple real-time operating features and the corresponding reference fluctuation range.

3. The method according to claim 1, characterized in that, The step of determining multiple reference operating characteristics corresponding to the operating condition parameters from a preset health condition reference library includes: If, according to the health condition reference library, the operating condition parameter is located between the operating condition parameters corresponding to two adjacent health conditions, interpolation is performed according to the multiple reference operating features corresponding to the two adjacent health conditions to determine the multiple reference operating features corresponding to the operating condition parameter.

4. The method according to claim 1, characterized in that, The reliability of the real-time operation signal is determined as follows: Determine the data integrity score, noise level score, sensor drift score, and physical consistency score of the real-time operating signal; The reliability of the real-time operating signal is determined based on the data integrity score, the noise level score, the sensor drift score, and the physical consistency score. The data integrity score of the real-time operating signal is determined based on the ratio of the number of valid sampling points to the theoretical number of sampling points within a preset time window. The noise level score of the real-time operating signal is determined based on the difference between the current noise level of the real-time operating signal within a preset time window and the normal noise level corresponding to the operating condition parameter. The sensor drift score of the real-time operating signal is determined based on the deviation between the measured value of the real-time operating signal and the estimated value determined based on the associated operating signal. The physical consistency score of the real-time operation signal is determined based on the degree to which the physical constraint relationship between the real-time operation signal and the associated operation signal is satisfied. The preset physical constraint relationships include at least one of the following: input-output power balance relationship, node current balance relationship, DC capacitor current and capacitor voltage change relationship, high-frequency transformer primary and secondary side relationship, and dual active bridge converter shift ratio and transmission power relationship.

5. A health status detection device, characterized in that, The device includes: The signal acquisition module is used to acquire multiple real-time operating signals of the device under test and the operating condition parameters of the device under test; the multiple real-time operating signals include at least two of electrical signals, control drive signals and temperature status signals. The first determining module is used to determine multiple real-time operating characteristics based on the multiple real-time operating signals; The second determining module is used to determine multiple reference operating features corresponding to the operating condition parameters from a preset health condition reference library; The third determining module is used to compare the multiple real-time operating features with the multiple reference operating features in a one-to-one correspondence manner, and determine multiple health deviation features based on the degree of deviation between each real-time operating feature and the corresponding reference operating feature; The weight determination module is used to determine the fusion weight of each corresponding health deviation feature based on the credibility of each real-time operation signal; the credibility is determined at least based on the degree to which the physical constraint relationship between the real-time operation signal and the associated operation signal is the operation signal that has a physical correlation with the real-time operation signal; The feature acquisition module is used to weight and concatenate the multiple health deviation features according to the corresponding fusion weights to obtain the fused deviation features; The status acquisition module is used to input the fusion deviation features into the status detection model, and obtain the abnormal status, fault type, fault location and health level of the device under test through the status detection model.

6. The apparatus according to claim 5, characterized in that, The third determining module is specifically used to: determine the reference fluctuation range corresponding to each of the reference operating features from a preset health condition reference library; and determine multiple health deviation features based on the degree of deviation between the multiple real-time operating features and the corresponding reference fluctuation ranges.

7. The apparatus according to claim 5, characterized in that, The second determining module is specifically used to: if, according to the health condition reference library, the operating condition parameter is located between the operating condition parameters corresponding to two adjacent health conditions, interpolate according to the multiple reference operating features corresponding to the two adjacent health conditions to determine the multiple reference operating features corresponding to the operating condition parameter.

8. The apparatus according to claim 5, characterized in that, The weight determination module is specifically used to: determine the data integrity score, noise level score, sensor drift score, and physical consistency score of the real-time operating signal; and determine the reliability of the real-time operating signal based on the data integrity score, the noise level score, the sensor drift score, and the physical consistency score. The data integrity score of the real-time operating signal is determined based on the ratio of the number of valid sampling points to the theoretical number of sampling points within a preset time window. The noise level score of the real-time operating signal is determined based on the difference between the current noise level of the real-time operating signal within a preset time window and the normal noise level corresponding to the operating condition parameter; the sensor drift score of the real-time operating signal is determined based on the deviation between the measured value of the real-time operating signal and the estimated value determined based on the associated operating signal; the physical consistency score of the real-time operating signal is determined based on the degree to which the physical constraint relationship between the real-time operating signal and the associated operating signal is satisfied; the preset physical constraint relationship includes at least one of the following: input-output power balance relationship, node current balance relationship, DC capacitor current and capacitor voltage change relationship, high-frequency transformer primary and secondary side relationship, and shift ratio and transmission power relationship of dual active bridge converter.

9. A health status detection device, characterized in that, The device includes: a memory and a processor; The memory is used to store programs; The processor is configured to execute the program to implement each step of the health status detection method as described in any one of claims 1 to 4.

10. A computer-readable medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements each step of the health status detection method as described in any one of claims 1 to 4.

Images