Current sensor error compensation method and system

Abstract

The present invention discloses a current sensor error compensation method and system, comprising: first, obtaining the current sensor operating environment and dividing it into time windows; extracting the environmental coding features of each window; dynamically evaluating the contribution coefficients and fusing them into environmental features; simultaneously obtaining electrical signal time series segments and fusing them into joint features; and reconstructing the circuit state features through a feature fusion network. The environmental and circuit state features are aggregated, and a compensation strategy module is matched accordingly. Compensation parameter fusion operations are performed in the two modules to obtain first and second compensation features, respectively. Finally, the current error compensation result is determined by combining the error type identifier, thereby achieving high-precision current sensor error compensation.

Description

Technical Field

[0001] The present invention relates to the field of automated detection, and in particular to a current sensor error compensation method and system. Background Art

[0002] Current sensors are widely used in numerous fields, including industrial production and power systems. Their measurement accuracy directly impacts system stability and reliability. However, operating environment factors (such as temperature and electromagnetic interference) and the circuit's inherent state can cause errors in current sensors. Existing error compensation methods often fail to fully account for the dynamic changes and mutual influence of the environment and circuit state, resulting in limited compensation accuracy. Therefore, a current sensor error compensation method that comprehensively considers the dynamic changes of multiple factors is urgently needed to improve measurement accuracy. Summary of the Invention

[0003] The object of the present invention is to provide a current sensor error compensation method and system.

[0004] In a first aspect, an embodiment of the present invention provides a current sensor error compensation method, comprising:

[0005] Acquire an operating environment corresponding to the current sensor, and divide the operating environment into multiple time windows;

[0006] Performing feature extraction on the multiple time windows to obtain environmental coding features corresponding to the multiple time windows;

[0007] Dynamically evaluating the environmental coding features corresponding to the multiple time windows to obtain contribution coefficients corresponding to the multiple time windows;

[0008] Dynamically fusing the environmental coding features corresponding to the multiple time windows and the contribution coefficients corresponding to the multiple time windows to obtain environmental features corresponding to the operating environment;

[0009] Acquire multiple electrical signal time series segments corresponding to the current sensor, and fuse the multiple electrical signal time series segments into an electrical signal joint feature;

[0010] Inputting the electrical signal joint features into a feature fusion network, performing a feature reconstruction operation on the electrical signal joint features based on the feature fusion network, and obtaining circuit state features corresponding to the current sensor;

[0011] An error type identifier is obtained, and a current error compensation result corresponding to the current sensor is determined based on the error type identifier, the environmental characteristics, and the circuit state characteristics.

[0012] Furthermore, the obtaining of the error type identifier and determining the current error compensation result corresponding to the current sensor based on the error type identifier, the environmental characteristics, and the circuit state characteristics include:

[0013] Aggregating the environmental features and the circuit state features within the same feature domain to obtain a first aggregated feature, and matching a first compensation strategy module for the environmental features in a plurality of compensation strategy modules based on the first aggregated feature; the plurality of compensation strategy modules configuring differentiated compensation strategies to adjust features of a plurality of data sources;

[0014] In the first compensation strategy module, a compensation parameter fusion operation is performed on the environmental feature based on the circuit state feature to obtain a first compensation feature;

[0015] Aggregating the circuit state feature and the first compensation feature in the same feature domain to obtain a second aggregate feature, and matching a second compensation strategy module for the circuit state feature among the multiple compensation strategy modules based on the second aggregate feature;

[0016] In the second compensation strategy module, a compensation parameter fusion operation is performed on the circuit state feature based on the first compensation feature to obtain a second compensation feature;

[0017] An error type identifier is obtained, and a current error compensation result corresponding to the current sensor is determined based on the error type identifier, the first compensation characteristic, and the second compensation characteristic.

[0018] Furthermore, the aggregating the environmental features and the circuit state features within the same feature domain to obtain a first aggregated feature includes:

[0019] Inputting the environmental features and the circuit state features into a dynamic fusion module, and performing feature alignment on the environmental features based on a first feature projection unit in the dynamic fusion module to obtain an environmental compensation feature;

[0020] Performing feature alignment on the circuit state feature based on the second feature projection unit in the dynamic fusion module to obtain a circuit compensation feature; the environment compensation feature and the circuit compensation feature belong to the same feature domain;

[0021] Acquiring a first effective value corresponding to the environmental compensation feature, and performing calibration processing on the environmental compensation feature based on the first effective value to obtain an environmental calibration feature;

[0022] Dynamically adjust the gain of the environmental calibration feature to obtain an environmental gain feature, and smooth the environmental gain feature to obtain an environmental smoothing feature;

[0023] Acquiring a second effective value corresponding to the circuit compensation feature, and performing calibration processing on the circuit compensation feature based on the second effective value to obtain a circuit calibration feature;

[0024] Dynamically gain-adjusting the circuit calibration feature to obtain a circuit gain feature, and smoothing the circuit gain feature to obtain a circuit smoothing feature;

[0025] The environmental smoothing feature and the circuit smoothing feature are combined to obtain a comprehensive compensation feature, and a feature reconstruction operation is performed on the comprehensive compensation feature based on a compensation parameter matrix corresponding to a parameter fusion module in the dynamic fusion module to obtain the first aggregated feature.

[0026] Furthermore, matching a first compensation strategy module for the environmental feature among multiple compensation strategy modules based on the first aggregated feature includes:

[0027] Performing a strategy evaluation on the first aggregated feature based on a strategy decision module in the dynamic fusion module to obtain fitness coefficients corresponding to multiple compensation strategy modules;

[0028] The compensation strategy module corresponding to the largest fitness coefficient is determined as the first compensation strategy module matching the environmental features.

[0029] Furthermore, the first compensation strategy module includes a dynamic weight fuser;

[0030] The first compensation strategy module performs a compensation parameter fusion operation on the environmental feature based on the circuit state feature to obtain a first compensation feature, including:

[0031] Perform channel combination on the environment feature and the circuit state feature to obtain a first combined feature;

[0032] Mapping the first joint feature into a first environment correlation matrix, a first circuit correlation matrix, and a first compensation parameter matrix based on a transformation compensation parameter matrix corresponding to the dynamic weight fuser;

[0033] constructing a first coupling parameter matrix based on a multiplication result between the first environment correlation matrix and the environment feature, and a multiplication result between the first environment correlation matrix and the circuit state feature;

[0034] constructing a second coupling parameter matrix based on a dimensionality conversion operation result of a multiplication result between the first circuit association matrix and the environmental feature, and a dimensionality conversion operation result of a multiplication result between the first circuit association matrix and the circuit state feature;

[0035] constructing a third coupling parameter matrix based on a multiplication result between the first compensation parameter matrix and the environmental characteristics, and a multiplication result between the first compensation parameter matrix and the circuit state characteristics;

[0036] Determine the product of the first coupling parameter matrix and the second coupling parameter matrix as a fourth coupling parameter matrix, and obtain the total number of channels of the environmental feature;

[0037] performing a normalized proportional calculation on the fourth coupling parameter matrix and the normalized scaling factor of the total number of channels to obtain a cross-source dynamic weight matrix, and determining a multiplication result of the cross-source dynamic weight matrix and the third coupling parameter matrix as a first dynamic correlation feature;

[0038] The first X groups of channel features in the first dynamic association feature are determined as first compensation features, where X is the number of main channels corresponding to the environmental feature.

[0039] Furthermore, the first compensation strategy module includes a dynamic weight fuser;

[0040] The first compensation strategy module performs a compensation parameter fusion operation on the environmental feature based on the circuit state feature to obtain a first compensation feature, including:

[0041] Based on the dynamic weight fuser, the environmental characteristics are mapped into a second environmental correlation matrix, and the circuit state characteristics are mapped into a second circuit correlation matrix and a second compensation parameter matrix;

[0042] Performing a matrix multiplication operation on a dimension conversion operation result of a multiplication result between the second environment association matrix and the environment feature, and a dimension conversion operation result between the second circuit association matrix and the circuit state feature to obtain a fifth coupling parameter matrix, and acquiring a total number of channels of the environment feature;

[0043] performing a normalized proportional calculation on the fifth coupling parameter matrix and a normalized scaling factor of the total number of channels to obtain a multi-source contribution coefficient matrix, and performing a matrix multiplication operation on the multi-source contribution coefficient matrix and a multiplication result of the second compensation parameter matrix and the circuit state characteristic to obtain a multi-source compensation characteristic;

[0044] The first X groups of channel features in the multi-source compensation features are determined as first compensation features; X is the number of main channels corresponding to the environmental features.

[0045] Furthermore, the first compensation strategy module includes a multi-source dynamic fuser;

[0046] The first compensation strategy module performs a compensation parameter fusion operation on the environmental feature based on the circuit state feature to obtain a first compensation feature, including:

[0047] Obtaining a fused compression feature, performing channel union on the fused compression feature and the circuit state feature to obtain a second joint feature; the fused compression feature is used to establish a connection between multiple source features;

[0048] Based on the first dynamic weight allocation branch in the multi-source dynamic fuser, a cross-source dynamic correlation analysis is performed on the second joint feature to obtain a second dynamic correlation feature, and the first Y groups of main channel features in the second dynamic correlation feature are determined as target compensation features; Y is the number of main channels corresponding to the fused compression feature;

[0049] Performing channel combination on the environmental feature and the target compensation feature to obtain a third combined feature, and performing cross-source dynamic correlation analysis on the third combined feature based on the second dynamic weight allocation branch in the multi-source dynamic fuser to obtain a third dynamic correlation feature;

[0050] Determine the first X groups of channel features in the third dynamic association feature as the first compensation feature; X is the number of main channels corresponding to the environmental feature;

[0051] The first compensation strategy module further includes a main channel fusion unit;

[0052] The first compensation strategy module performs a compensation parameter fusion operation on the environmental feature based on the circuit state feature to obtain a first compensation feature, including:

[0053] Acquire an environmental calibration feature corresponding to the environmental feature, and a circuit calibration feature corresponding to the circuit state feature;

[0054] Based on the first dynamic time series processor in the main channel fusion unit, the environmental feature is modulated by a time series signal to obtain a first time series response feature;

[0055] Based on the second dynamic timing processor in the main channel fusion unit, the circuit state feature is modulated by a timing signal to obtain a second timing response feature;

[0056] The second time series response characteristic is smoothed to obtain a time series smoothing characteristic, and the first time series response characteristic and the time series smoothing characteristic are combined to obtain a first compensation characteristic.

[0057] Further, the first compensation strategy module includes a reference signal suppressor;

[0058] The first compensation strategy module performs a compensation parameter fusion operation on the environmental feature based on the circuit state feature to obtain a first compensation feature, including:

[0059] The circuit state feature is cleared, and the environmental feature is used as a first compensation feature output by the reference signal suppressor.

[0060] Furthermore, determining a current error compensation result corresponding to the current sensor based on the error type identifier, the first compensation feature, and the second compensation feature includes:

[0061] Performing channel combination on the first compensation feature and the second compensation feature to obtain a fourth joint feature, and adding the error type identifier to the front end of the fourth joint feature to obtain a joint compensation feature;

[0062] The joint compensation feature is dynamically evaluated based on a cross-source dynamic weight allocation branch to obtain a dynamic weight fusion parameter, and a feature reconstruction operation is performed on the dynamic weight fusion parameter to obtain a current error compensation result corresponding to the current sensor.

[0063] In a second aspect, an embodiment of the present invention provides a server system, including a server, wherein the server is configured to execute the method described in the first aspect.

[0064] Compared to existing technologies, the present invention offers the following advantages: A current sensor error compensation method and system disclosed herein obtains the current sensor's operating environment and divides it into time windows, extracts environmental coding features from each window, dynamically evaluates the contribution coefficient, and fuses them into environmental features. Simultaneously, electrical signal time series segments are obtained and fused into joint features, which are then reconstructed into circuit state features via a feature fusion network. The environmental and circuit state features are aggregated and matched to compensation strategy modules. Compensation parameter fusion operations are performed in the two modules to obtain first and second compensation features, respectively. Finally, the current error compensation result is determined based on the error type identifier, achieving high-precision current sensor error compensation. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly describes the drawings required for use in the embodiments. It should be understood that the following drawings illustrate only certain embodiments of the present invention and should not be construed as limiting the scope of the present invention. Those skilled in the art can, without inventive effort, derive other relevant drawings from these drawings.

[0066] Figure 1 A schematic flow chart of the steps of a current sensor error compensation method provided by an embodiment of the present invention;

[0067] Figure 2 A schematic block diagram of the structure of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0068] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings of the embodiments of the present invention. It should be understood that the described embodiments are only a portion of the embodiments of the present invention, not all of them. Generally, the components of the embodiments of the present invention described and illustrated in the drawings herein may be arranged and designed in a variety of different configurations.

[0069] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0070] In order to solve the technical problems in the above background technology, Figure 1 This is a flow chart of a current sensor error compensation method provided by an embodiment of the present disclosure. The current sensor error compensation method is introduced in detail below.

[0071] Step S201, obtaining an operating environment corresponding to the current sensor, and dividing the operating environment into multiple time windows;

[0072] Step S202: performing feature extraction on the multiple time windows to obtain environment coding features corresponding to the multiple time windows;

[0073] Step S203, dynamically evaluating the environmental coding features corresponding to the multiple time windows to obtain contribution coefficients corresponding to the multiple time windows;

[0074] Step S204: dynamically fusing the environmental coding features corresponding to the multiple time windows and the contribution coefficients corresponding to the multiple time windows to obtain environmental features corresponding to the operating environment;

[0075] Step S205 , obtaining a plurality of electrical signal time series segments corresponding to the current sensor, and fusing the plurality of electrical signal time series segments into an electrical signal joint feature;

[0076] Step S206: inputting the electrical signal joint feature into a feature fusion network, performing a feature reconstruction operation on the electrical signal joint feature based on the feature fusion network, and obtaining a circuit state feature corresponding to the current sensor;

[0077] Step S207: Aggregate the environmental features and the circuit state features within the same feature domain to obtain a first aggregated feature. Based on the first aggregated feature, match a first compensation strategy module to the environmental features in multiple compensation strategy modules; the multiple compensation strategy modules configure differentiated compensation strategies to adjust features of multiple data sources.

[0078] Step S208: In the first compensation strategy module, a compensation parameter fusion operation is performed on the environmental feature based on the circuit state feature to obtain a first compensation feature;

[0079] Step S209: Aggregate the circuit state feature and the first compensation feature in the same feature domain to obtain a second aggregate feature, and match a second compensation strategy module for the circuit state feature among the multiple compensation strategy modules based on the second aggregate feature;

[0080] Step S210: In the second compensation strategy module, a compensation parameter fusion operation is performed on the circuit state feature based on the first compensation feature to obtain a second compensation feature.

[0081] Step S211 : obtaining an error type identifier, and determining a current error compensation result corresponding to the current sensor based on the error type identifier, the first compensation characteristic, and the second compensation characteristic.

[0082] In an embodiment of the present invention, for example, within this industrial production park, different production equipment may experience different operating environments during different time periods. For example, during peak daytime production periods, the ambient temperature of some equipment may rise due to the intensive operation of the equipment, and electromagnetic interference may also increase due to the simultaneous operation of many devices. Meanwhile, during low-temperature periods at night, the temperature may drop, and electromagnetic interference may also decrease accordingly. The server obtains operating environment information corresponding to a current sensor. This information may include various environmental parameters such as temperature, humidity, and electromagnetic intensity. The server divides this operating environment into time windows, for example, every 10 minutes. This is because, based on the experience of this production park, a 10-minute interval can capture dynamic changes in environmental parameters without increasing the processing burden by consuming excessive data. For example, from 8:00 AM to 9:00 AM, the server divides the operating environment into six 10-minute time windows, each containing complete environmental parameter information for that time period. For each 10-minute time window, the server uses specific algorithms and models to extract features. For example, for temperature parameters, the server may extract features such as the mean, maximum, minimum, and slope of temperature variation. For electromagnetic intensity parameters, it may extract features such as peak and valley values, as well as the frequency of fluctuations within the time period. For example, within a time window, the mean temperature is 30°C, the maximum is 32°C, the minimum is 28°C, and the temperature slope increases by 1°C every 10 minutes. The peak electromagnetic intensity is 50 microteslas, the valley is 30 microteslas, and the frequency of fluctuation is 5 times per minute. The server encodes these extracted features to form a unique environmental signature. This signature acts as a "fingerprint" of the operating environment within that time window, containing key information about the environment during that time period for further analysis and processing. The impact of the environment in different time windows on current sensor error may vary. The server dynamically evaluates the environmental signatures for each time window to determine their contribution to the overall operating environment characteristics. For example, during peak production periods, temperature and electromagnetic interference have a greater impact on current sensor error, while during off-peak periods, humidity may have a greater impact. The server analyzes historical data and real-time current sensor error monitoring to determine the importance of environmental coding features in each time window. For example, during the 9:00 AM to 10:00 AM period, analysis reveals that the electromagnetic interference from a nearby large motor starts up, significantly impacting the current sensor error. Therefore, the contribution coefficient of the environmental coding features in this time window is relatively high, assuming it is 0.8. In contrast, during the 9:20 AM to 9:30 AM period, environmental parameters are relatively stable, with a smaller impact on the current sensor error, resulting in a contribution coefficient of perhaps 0.3.The server then determines a contribution coefficient for each time window, reflecting its importance to the overall operating environment. The server then fuses the environmental coding features of each time window with their corresponding contribution coefficients. For example, the environmental coding features for the 9:00-9:10 time window are [30, 32, 28, 1, 50, 30, 5], corresponding to the mean temperature, maximum temperature, minimum temperature, temperature slope, peak electromagnetic intensity, valley electromagnetic intensity, and fluctuation frequency, respectively, with a contribution coefficient of 0.8. The environmental coding features for the 9:20-9:30 time window are [29, 30, 28, 0.5, 40, 35, 3], with a contribution coefficient of 0.3. The server then fuses this information using a specific fusion algorithm, such as weighted averaging. The fusion formula is: ,in, After fusion Environmental characteristic values, such as mean temperature, peak electromagnetic intensity, etc.; The sequence number of the time window, ranging from 1 to , is the total number of time windows, For the The time window The original eigenvalues, the features extracted in each time window include the temperature mean, maximum value, minimum value, electromagnetic intensity peak value, etc. For the The contribution coefficient of each time window is dynamically evaluated, reflecting the importance of that time window to the overall environmental characteristics. Based on this, the fused mean temperature characteristic is: [(30 × 0.8 + 29 × 0.3) / (0.8 + 0.3)] ≈ 29.73°C. By fusing all features in this way, the server obtains an environmental characteristic corresponding to the entire operating environment. This environmental characteristic comprehensively considers the impact of different time window environments on current sensor errors. Current sensors collect electrical signals in real time, which exist in the form of time series segments. In industrial production, different production processes and equipment operating states can cause variations in electrical signals. For example, the electrical signal of a motor exhibits different characteristics during startup, stable operation, and shutdown. The server obtains multiple time series segments of the electrical signal from a current sensor over a period of time, each of which may last from a few milliseconds to a few seconds. For example, at the moment of motor startup, the electrical signal experiences a large peak, which then gradually stabilizes. The server fuses these different electrical signal time series segments to form a joint electrical signal feature. Fusion can be accomplished by splicing these time series segments in chronological order, or by integrating frequency, phase, and other features from different segments through signal processing algorithms, such as Fourier transforms, to generate a joint electrical signal feature containing comprehensive information about the electrical signal within that time period. The server inputs the joint electrical signal feature into a pre-trained feature fusion network. This feature fusion network, acting as an intelligent information processing center, can identify and reconstruct various patterns and features within the joint electrical signal feature. For example, the feature fusion network might identify periodic variations or abnormal fluctuations in the electrical signal and, based on these features, reconstruct a feature that reflects the current circuit state of the current sensor. For example, if the electrical signal exhibits abnormal high-frequency fluctuations, the feature fusion network might reconstruct this fluctuation into a circuit state feature indicating potential interference. By reconstructing the joint electrical signal feature, the server obtains a circuit state feature that accurately reflects the current circuit state of the current sensor, providing a critical basis for subsequent error compensation. The server must first ensure that the environmental features and circuit state features are within the same feature domain for aggregation. For example, the server might use feature mapping and normalization methods to adjust environmental and circuit state features to the same numerical range and representation. For example, consider the temperature feature, which might have a range of 20-40°C, and the voltage fluctuation feature, which might have a range of 100-200V. The server normalizes the temperature feature to a range of 0-1, placing it in the same feature domain as the similarly processed voltage fluctuation feature. The server then inputs the environmental and circuit state features into the dynamic fusion module.In the dynamic fusion module, the first feature projection unit performs feature alignment on the environmental features to obtain an environmental compensation feature. The second feature projection unit performs feature alignment on the circuit state features to obtain a circuit compensation feature. These two compensation features reside in the same feature domain. Next, the server obtains a first effective value corresponding to the environmental compensation feature. For example, by calculating the effective value of the environmental compensation feature in each dimension, assuming the first effective value is 0.8. Based on this first effective value, the environmental compensation feature is calibrated by multiplying each dimension value of the environmental compensation feature by 0.8 to obtain the calibrated environmental feature. The calibrated environmental feature is then dynamically gain-adjusted. For example, according to a pre-defined gain rule, certain dimensions of the environmental calibration feature are amplified or reduced to obtain the environmental gain feature. The environmental gain feature is then smoothed to remove any noise and sudden changes to obtain the smoothed environmental feature. Similarly, the server obtains a second effective value corresponding to the circuit compensation feature, assuming it is 0.6. Based on this value, the circuit compensation feature is calibrated, dynamically gain-adjusted, and smoothed to obtain the smoothed circuit feature. Finally, the server combines the smoothed environmental feature and the smoothed circuit feature to obtain the comprehensive compensation feature. Based on the compensation parameter matrix corresponding to the parameter fusion module in the dynamic fusion module, a feature reconstruction operation is performed on the comprehensive compensation feature to obtain a first aggregated feature. Based on this first aggregated feature, the server uses the policy decision module in the dynamic fusion module to perform policy evaluation. The policy decision module analyzes the dimensions and feature patterns of the first aggregated feature and compares them with the fitness of multiple compensation policy modules to obtain fitness coefficients corresponding to the multiple compensation policy modules. For example, for a compensation policy module with a dynamic weight fuser, the policy decision module finds that its fitness coefficient with the first aggregated feature is 0.7; for another compensation policy module with a main channel fusion unit, the fitness coefficient is 0.5. The server identifies the compensation policy module corresponding to the largest fitness coefficient as the first compensation policy module that matches the environmental feature. Assume that the first compensation policy module includes a dynamic weight fuser. The server first performs a channel union on the environmental feature and the circuit state feature, concatenating the two features along the channel dimension to obtain the first joint feature. Based on the transformed compensation parameter matrix corresponding to the dynamic weight fuser, the server maps the first joint feature into a first environmental correlation matrix, a first circuit correlation matrix, and a first compensation parameter matrix. For example, the first joint signature is converted into these three matrices through matrix multiplication and specific mapping rules. Then, based on the multiplication results between the first environmental correlation matrix and the environmental signature, and the multiplication results between the first environmental correlation matrix and the circuit state signature, a first coupling parameter matrix is constructed. Specifically, each row of the first environmental correlation matrix is multiplied by the corresponding element of the environmental signature and the circuit state signature, respectively. These product results are then combined according to specific rules to form the first coupling parameter matrix.In a similar manner, a second coupling parameter matrix is constructed based on the dimensionality conversion results of the multiplication results between the first circuit correlation matrix and the environmental features, and the dimensionality conversion results of the multiplication results between the first circuit correlation matrix and the circuit state features. The dimensionality conversion operation here may adjust the matrix dimensions of the multiplication results to accommodate subsequent calculations. A third coupling parameter matrix is constructed based on the multiplication results between the first compensation parameter matrix and the environmental features, and the multiplication results between the first compensation parameter matrix and the circuit state features. The multiplication results between the first coupling parameter matrix and the second coupling parameter matrix are determined as a fourth coupling parameter matrix. The server obtains the total number of channels of the environmental features, assuming it is 10. A normalized scaling factor, such as 1 / 10, is applied to the fourth coupling parameter matrix and the total number of channels to obtain a cross-source dynamic weight matrix. The multiplication result between the cross-source dynamic weight matrix and the third coupling parameter matrix is determined as the first dynamic correlation feature. Finally, the first X groups of channel features in the first dynamic correlation feature are determined as the first compensation feature. Assuming X is the number of primary channels corresponding to the environmental features, such as 5, the first five groups of channel features of the first dynamic correlation feature are selected as the first compensation feature. Similar to the previous aggregation of environmental features and circuit state features, the server again ensures that the circuit state features and the first compensation features are within the same feature domain. Through operations such as feature alignment, calibration, gain adjustment, and smoothing, the circuit state features and the first compensation features are aggregated to obtain a second aggregated feature. For example, feature alignment is also performed to obtain the compensation features of the circuit and the first compensation features, which are then calibrated, gain adjusted, and smoothed separately. After merging to obtain a comprehensive feature, feature reconstruction is performed based on the compensation parameter matrix of the parameter fusion module to obtain a second aggregated feature. The server then uses the policy decision module to evaluate the fitness of multiple compensation strategy modules based on the second aggregated feature, obtaining fitness coefficients corresponding to the multiple compensation strategy modules. For example, for a compensation strategy module with a multi-source dynamic fuser, the fitness coefficient is 0.6; for another compensation strategy module with a reference signal suppressor, the fitness coefficient is 0.4. The server determines the compensation strategy module corresponding to the largest fitness coefficient as the second compensation strategy module that matches the circuit state features. Assume that the second compensation strategy module is a multi-source dynamic fuser. The server obtains a fused compressed feature, which may be obtained by compressing and fusing some intermediate features in the previous processing process, and is used to establish a connection between multi-source features. The fused compressed feature and the circuit state feature are channel-joined to obtain a second joint feature. Based on the first dynamic weight allocation branch in the multi-source dynamic fuser, a cross-source dynamic correlation analysis is performed on the second joint feature to obtain a second dynamic correlation feature. For example, the first dynamic weight allocation branch will assign different weights based on the correlation and importance between different channel features in the second joint feature, thereby obtaining a second dynamic correlation feature.The first Y groups of main channel features in the second dynamic correlation feature are determined as the target compensation features. Assuming Y is the number of main channels corresponding to the fused compression feature, for example, 3, the first three groups of main channel features are selected as the target compensation features. Channel-wise union of the environmental features and the target compensation features is performed to obtain a third joint feature. Based on the second dynamic weight allocation branch in the multi-source dynamic fusion module, cross-source dynamic correlation analysis is performed on the third joint feature to obtain a third dynamic correlation feature. Finally, the first X groups of channel features in the third dynamic correlation feature are determined as the second compensation feature. Assuming X is the number of main channels corresponding to the circuit state feature, for example, 4, the first four groups of channel features in the third dynamic correlation feature are selected as the second compensation feature. The server obtains an error type identifier from the current sensor or related monitoring system. For example, error types may include temperature drift error, electromagnetic interference error, circuit component aging error, etc. Each error type has a specific identifier. Assuming the currently obtained error type identifier is "temperature drift error," the server performs channel-wise union of the first and second compensation features to obtain a fourth joint feature. The error type identifier is added to the leading tag of the fourth joint feature to obtain the joint compensation feature. For example, the identifier for "temperature drift error" is encoded as [1,0,0]. Assuming there are three error types, this is added to the front end of the fourth joint feature to form a joint compensation feature. The joint compensation feature is dynamically evaluated based on the cross-source dynamic weight allocation branch. Different weights are assigned to each feature in the joint compensation feature based on its importance and correlation with the error type, resulting in a dynamic weight fusion parameter. Feature reconstruction operations are performed on the dynamic weight fusion parameter. For example, through specific matrix operations and transformations, the dynamic weight fusion parameter is converted into a numerical value that can be directly used to compensate for the current sensor error. The current error compensation result corresponding to the current sensor is ultimately obtained. This compensation result can be used to adjust the output data of the current sensor to more accurately reflect the actual current value, thereby ensuring the stable operation of industrial production equipment. Through the detailed steps and scenario examples above, the server can effectively compensate for the errors of the current sensor, improve the accuracy of the current data, and provide strong support for the stable operation of industrial production and other fields.

[0083] In the embodiment of the present invention, the aggregation processing of the environmental feature and the circuit state feature in the same feature domain to obtain the first aggregated feature can be implemented through the following examples.

[0084] Inputting the environmental features and the circuit state features into a dynamic fusion module, and performing feature alignment on the environmental features based on a first feature projection unit in the dynamic fusion module to obtain an environmental compensation feature;

[0085] Performing feature alignment on the circuit state feature based on the second feature projection unit in the dynamic fusion module to obtain a circuit compensation feature; the environment compensation feature and the circuit compensation feature belong to the same feature domain;

[0086] The environment compensation feature and the circuit compensation feature are aggregated to obtain a first aggregate feature.

[0087] In an embodiment of the present invention, the server illustratively inputs previously obtained environmental features and circuit state features into the dynamic fusion module. Environmental features include processed, comprehensive features such as temperature, humidity, and electromagnetic intensity, while circuit state features reflect the current circuit condition of the current sensor, including characteristics such as electrical signal fluctuation and frequency. The first feature projection unit in the dynamic fusion module begins to align the environmental features. For example, the temperature in the environmental feature originally ranges from 20-40°C, while the module's uniform feature range is set to 0-1. The first feature projection unit maps the temperature feature to the range of 0-1 using methods such as linear transformation. It also processes other environmental features such as humidity and electromagnetic intensity according to corresponding rules to obtain an environmental compensation feature. Next, the second feature projection unit aligns the circuit state features. For example, the electrical signal fluctuation amplitude in the circuit state feature may originally be 10-100mV. The second feature projection unit adjusts it to the same range as the environmental compensation feature, such as 0-1, based on predefined rules. Similar processing is performed on other circuit state features such as electrical signal frequency to obtain the circuit compensation feature. At this point, the environmental compensation feature and the circuit compensation feature belong to the same feature domain. Finally, the server aggregates the environmental compensation features and circuit compensation features. Assume that the environmental compensation feature is represented by a vector [0.3, 0.5, 0.4], representing the processed temperature, humidity, and electromagnetic intensity eigenvalues, respectively; the circuit compensation feature is represented by a vector [0.6, 0.2, 0.7], representing the processed electrical signal fluctuation, frequency, and phase eigenvalues. The server uses a specific algorithm, such as adding and normalizing the corresponding elements: [(0.3 + 0.6) / 2, (0.5 + 0.2) / 2, (0.4 + 0.7) / 2] = [0.45, 0.35, 0.55]. This new vector is the first aggregated feature. This first aggregated feature integrates environmental and circuit status information, providing a basis for the subsequent matching compensation strategy module.

[0088] In the embodiment of the present invention, the aggregating the environment compensation feature and the circuit compensation feature to obtain a first aggregate feature may be implemented through the following example.

[0089] Acquiring a first effective value corresponding to the environmental compensation feature, and performing calibration processing on the environmental compensation feature based on the first effective value to obtain an environmental calibration feature;

[0090] Dynamically adjust the gain of the environmental calibration feature to obtain an environmental gain feature, and smooth the environmental gain feature to obtain an environmental smoothing feature;

[0091] Acquiring a second effective value corresponding to the circuit compensation feature, and performing calibration processing on the circuit compensation feature based on the second effective value to obtain a circuit calibration feature;

[0092] Dynamically gain-adjusting the circuit calibration feature to obtain a circuit gain feature, and smoothing the circuit gain feature to obtain a circuit smoothing feature;

[0093] The environmental smoothing feature and the circuit smoothing feature are combined to obtain a comprehensive compensation feature, and a feature reconstruction operation is performed on the comprehensive compensation feature based on a compensation parameter matrix corresponding to a parameter fusion module in the dynamic fusion module to obtain the first aggregated feature.

[0094] In an embodiment of the present invention, the server first obtains the first effective value corresponding to the environmental compensation feature. For example, the environmental compensation feature is composed of multiple features such as temperature, humidity, and electromagnetic intensity, and its first effective value is calculated to be 0.75. Based on this first effective value, the server calibrates the environmental compensation feature. Assuming that the temperature feature value in the environmental compensation feature is 0.6, the temperature feature value becomes 0.6 × 0.75 = 0.45 after calibration. Other features such as humidity and electromagnetic intensity are processed in the same manner, thereby obtaining the calibrated environmental feature. Next, the server dynamically adjusts the gain of the calibrated environmental feature. Based on an empirical model of the impact of different regional environments on current sensors within the campus, for the temperature feature, if the equipment in that area is temperature-sensitive, the server may set the gain coefficient of the temperature feature to 1.2. After adjustment, the temperature feature value becomes 0.45 × 1.2 = 0.54. Other features are adjusted according to the corresponding gain rules to obtain the environmental gain feature. The server then smoothes the environmental gain feature to remove minor fluctuations that may be introduced by the gain adjustment, thereby obtaining the smoothed environmental feature. Afterwards, the server obtains the second effective value corresponding to the circuit compensation feature, assuming it is 0.8. Based on this, the circuit compensation feature is calibrated. For example, if the electrical signal fluctuation characteristic value in the circuit compensation feature is 0.5, it becomes 0.5×0.8=0.4 after calibration. Other electrical signal frequency, phase and other characteristics are processed in the same way to obtain the circuit calibration feature. The server then performs dynamic gain adjustment on the circuit calibration feature. For example, according to the circuit characteristics, the gain coefficient of the electrical signal frequency feature is set to 0.9. After adjustment, the electrical signal frequency characteristic value becomes 0.4×0.9=0.36. Other features are adjusted according to the rules to obtain the circuit gain feature. The circuit gain feature is smoothed to obtain the circuit smoothing feature. Finally, the server merges the environmental smoothing feature and the circuit smoothing feature to obtain the comprehensive compensation feature. Assuming the environmental smoothing feature is [0.54, 0.3, 0.4] and the circuit smoothing feature is [0.36, 0.4, 0.5], the combined comprehensive compensation feature is [0.54, 0.3, 0.4, 0.36, 0.4, 0.5]. A feature reconstruction operation is performed on the comprehensive compensation feature based on the compensation parameter matrix corresponding to the parameter fusion module in the dynamic fusion module. A matrix operation is performed on the compensation parameter matrix and the comprehensive compensation feature, and the features are recombined to obtain the first aggregated feature, which lays the foundation for the subsequent matching compensation strategy module.

[0095] In the embodiment of the present invention, matching the first compensation strategy module for the environmental feature among multiple compensation strategy modules based on the first aggregate feature can be implemented through the following examples.

[0096] Performing a strategy evaluation on the first aggregated feature based on a strategy decision module in the dynamic fusion module to obtain fitness coefficients corresponding to multiple compensation strategy modules;

[0097] The compensation strategy module corresponding to the largest fitness coefficient is determined as the first compensation strategy module matching the environmental features.

[0098] In an embodiment of the present invention, the server, using the policy decision module within the dynamic fusion module, conducts a policy evaluation on the previously obtained first aggregated feature. Within the campus, multiple compensation policy modules each have their own unique characteristics. For example, the dynamic weight fusion module excels at weighting multi-source data to achieve compensation; the main channel fusion unit module focuses on fusion compensation of key channel features; and the reference signal suppressor module suppresses and compensates for specific interfering reference signals. The policy decision module begins its work, deeply analyzing the various dimensions and characteristic patterns of the first aggregated feature. For example, the dynamic weight fusion module examines the correlation between environmental features and circuit state features in the first aggregated feature, as well as the trend of feature changes. If the first aggregated feature indicates a close correlation between temperature and electrical signal fluctuations in the environment, and this correlation exhibits a certain nonlinear pattern, the dynamic weight fusion module may be better suited to this situation through its unique weight allocation mechanism. The policy decision module, based on its own algorithm, assigns this module a fitness coefficient, such as 0.7. For the main channel fusion unit module, the policy decision module examines the importance and interrelationships of the main channel features within the first aggregated feature. If the primary channel features in the first aggregated feature, such as temperature and electrical signal frequency, significantly affect the error and their variations exhibit certain temporal characteristics, and if the primary channel fusion unit module can effectively capture and fuse these features, the policy decision module will assign it a fitness coefficient, assuming it is 0.5. Similarly, the policy decision module evaluates the reference signal suppressor module. If the first aggregated feature indicates that electromagnetic interference in the environment has an impact on the current sensor similar to the reference signal, and if the reference signal suppressor module can effectively suppress this interference, the policy decision module will assign a corresponding fitness coefficient, such as 0.4. After evaluating all compensation strategy modules, the server obtains the fitness coefficients corresponding to multiple compensation strategy modules. The server then identifies the compensation strategy module corresponding to the largest fitness coefficient as the first compensation strategy module matching the environmental features. In the above example, the dynamic weight fusion module has the largest fitness coefficient of 0.7, so the server identifies the dynamic weight fusion module as the first compensation strategy module matching the environmental features, allowing for more accurate error compensation based on this module.

[0099] In an embodiment of the present invention, the first compensation strategy module includes a dynamic weight fuser;

[0100] In the first compensation strategy module, a compensation parameter fusion operation is performed on the environmental feature based on the circuit state feature to obtain a first compensation feature, which can be implemented through the following example.

[0101] Perform channel combination on the environment feature and the circuit state feature to obtain a first combined feature;

[0102] Performing cross-source dynamic correlation analysis on the first joint feature based on the dynamic weight fuser to obtain a first dynamic correlation feature;

[0103] The first X groups of channel features in the first dynamic association feature are determined as first compensation features, where X is the number of main channels corresponding to the environmental feature.

[0104] In an embodiment of the present invention, for example, in the scenario of a large industrial production park, the server uses the dynamic weight fusion module in the first compensation strategy module to process environmental features and circuit state features to obtain a first compensation feature. The server first performs a channel-wise union of the environmental and circuit state features. Environmental features include processed feature data such as temperature, humidity, and electromagnetic intensity, represented as the vector [0.6, 0.4, 0.5], corresponding to temperature, humidity, and electromagnetic intensity, respectively. Circuit state features include feature data such as electrical signal fluctuations, frequency, and phase, represented as the vector [0.7, 0.3, 0.8]. The server concatenates these two vectors along the channel dimension to obtain the first joint feature [0.6, 0.4, 0.5, 0.7, 0.3, 0.8]. Next, the server performs cross-source dynamic correlation analysis on the first joint feature using the dynamic weight fusion module. The dynamic weight fusion module analyzes the correlation between the different sources of the first joint feature—environmental and circuit features. For example, it finds a correlation between temperature and electrical signal fluctuations, and a potential link between humidity and electrical signal frequency. Based on these correlations, the dynamic weight fusion assigns different weights to each feature. For the temperature feature, due to its close correlation with electrical signal fluctuations, a weight of 0.8 might be assigned; humidity, with its relatively weak correlation with electrical signal frequency, might be assigned a weight of 0.6, and so on. By assigning and calculating weights to all features, a first dynamic correlation feature is obtained. Assume that the calculated first dynamic correlation feature is [0.48, 0.24, 0.3, 0.56, 0.18, 0.64]. Finally, the number X of primary channels corresponding to the environmental features is determined. Assume that previous analysis of the current sensor data within the campus determined that temperature and humidity are the primary channels corresponding to the environmental features, i.e., X = 2. The server identifies the first two sets of channel features [0.48, 0.24] in the first dynamic correlation feature as the first compensation feature. This first compensation feature comprehensively considers the dynamic correlation between environmental features and circuit state features, providing an important data foundation for subsequent processing to compensate for current sensor errors.

[0105] In the embodiment of the present invention, the cross-source dynamic correlation analysis of the first joint feature based on the dynamic weight fuser to obtain the first dynamic correlation feature can be implemented through the following examples.

[0106] Based on the transformation compensation parameter matrix corresponding to the dynamic weight fuser, mapping the first joint feature into a first environment correlation matrix, a first circuit correlation matrix and a first compensation parameter matrix;

[0107] constructing a first coupling parameter matrix based on a multiplication result between the first environment correlation matrix and the environment feature, and a multiplication result between the first environment correlation matrix and the circuit state feature;

[0108] constructing a second coupling parameter matrix based on a dimensionality conversion operation result of a multiplication result between the first circuit association matrix and the environmental feature, and a dimensionality conversion operation result of a multiplication result between the first circuit association matrix and the circuit state feature;

[0109] constructing a third coupling parameter matrix based on a multiplication result between the first compensation parameter matrix and the environmental characteristics, and a multiplication result between the first compensation parameter matrix and the circuit state characteristics;

[0110] Determine the product of the first coupling parameter matrix and the second coupling parameter matrix as a fourth coupling parameter matrix, and obtain the total number of channels of the environmental feature;

[0111] The fourth coupling parameter matrix and the normalized scaling factor of the total number of channels are normalized to obtain a cross-source dynamic weight matrix, and the multiplication result between the cross-source dynamic weight matrix and the third coupling parameter matrix is determined as the first dynamic association feature.

[0112] In an embodiment of the present invention, for example, in a large industrial production park, the server continues to use the dynamic weight fusion device to perform a more detailed cross-source dynamic correlation analysis on the first joint feature to obtain the first dynamic correlation feature. The server processes the first joint feature according to the transformation compensation parameter matrix corresponding to the dynamic weight fusion device. Assuming that the first joint feature is [0.6, 0.4, 0.5, 0.7, 0.3, 0.8], it is mapped into a first environment correlation matrix, a first circuit correlation matrix, and a first compensation parameter matrix through specific matrix operation rules. For example, after complex linear transformation and other operations, the first environment correlation matrix is obtained as [[0.2, 0.3], [0.4, 0.1]], the first circuit correlation matrix is [[0.5, 0.2], [0.3, 0.4]], and the first compensation parameter matrix is [[0.6, 0.1], [0.2, 0.5]]. Next, the first coupling parameter matrix is constructed. It is known that the environmental feature is [0.6, 0.4] and the circuit state feature is [0.7, 0.3]. The first environmental correlation matrix is multiplied by the environmental characteristics, resulting in [0.2×0.6+0.3×0.4,0.4×0.6+0.1×0.4]=[0.24,0.28]. The first environmental correlation matrix is multiplied by the circuit state characteristics, resulting in [0.2×0.7+0.3×0.3,0.4×0.7+0.1×0.3]=[0.23,0.31]. These two results are combined to construct the first coupling parameter matrix: [[0.24,0.23],[0.28,0.31]]. To construct the second coupling parameter matrix, first perform a dimension conversion on the product of the first circuit correlation matrix with the environmental characteristics and circuit state characteristics. The first circuit correlation matrix multiplied by the environmental characteristics is [0.5×0.6+0.2×0.4,0.3×0.6+0.4×0.4]=[0.38,0.34]. After dimensionality conversion, it is assumed to become [[0.38],[0.34]]. The first circuit correlation matrix multiplied by the environmental characteristics is [0.5×0.7+0.2×0.3,0.3×0.7+0.4×0.3]=[0.41,0.33]. After dimensionality conversion, it is assumed to become [[0.41],[0.33]]. The second coupling parameter matrix is constructed from the results of these two dimensional conversions as [[0.38,0.41],[0.34,0.33]]. To construct the third coupling parameter matrix, the first compensation parameter matrix multiplied by the environmental characteristics is [0.6×0.6+0.1×0.4,0.2×0.6+0.5×0.4]=[0.4,0.32]; and multiplied by the circuit state characteristics is [0.6×0.7+0.1×0.3,0.2×0.7+0.5×0.3]=[0.45,0.29]. Thus, the third coupling parameter matrix is constructed as [[0.4,0.45],[0.32,0.29]]. Then, the first coupling parameter matrix is multiplied by the second coupling parameter matrix to obtain the fourth coupling parameter matrix.That is, [[0.24×0.38+0.23×0.34,0.24×0.41+0.23×0.33],[0.28×0.38+0.31×0.34,0.28×0.41+0.31×0.33]] = [[0.17,0.18],[0.20,0.21]]. The server obtains the total number of channels for the environmental feature, which is [0.6,0.4], and the total number of channels is 2. The fourth coupling parameter matrix is normalized by the scaling factor (1 / 2) of the total number of channels to obtain the cross-source dynamic weight matrix. For example, [[0.17×0.5,0.18×0.5],[0.20×0.5,0.21×0.5]] = [[0.085,0.09],[0.1,0.105]]. Finally, the cross-source dynamic weight matrix is multiplied by the third coupling parameter matrix to obtain the first dynamic correlation feature. That is, [[0.085×0.4+0.09×0.32,0.085×0.45+0.09×0.29],[0.1×0.4+0.105×0.32,0.1×0.45+0.105×0.29]]=[[0.0628,0.0621],[0.0736,0.0755]], which is the first dynamic correlation feature.

[0113] In an embodiment of the present invention, the first compensation strategy module includes a dynamic weight fuser;

[0114] In the first compensation strategy module, a compensation parameter fusion operation is performed on the environmental feature based on the circuit state feature to obtain a first compensation feature, which can be implemented through the following example.

[0115] Based on the dynamic weight fuser, the environmental characteristics are mapped into a second environmental correlation matrix, and the circuit state characteristics are mapped into a second circuit correlation matrix and a second compensation parameter matrix;

[0116] Performing a matrix multiplication operation on a dimension conversion operation result of a multiplication result between the second environment association matrix and the environment feature, and a dimension conversion operation result between the second circuit association matrix and the circuit state feature to obtain a fifth coupling parameter matrix, and acquiring a total number of channels of the environment feature;

[0117] performing a normalized proportional calculation on the fifth coupling parameter matrix and a normalized scaling factor of the total number of channels to obtain a multi-source contribution coefficient matrix, and performing a matrix multiplication operation on the multi-source contribution coefficient matrix and a multiplication result of the second compensation parameter matrix and the circuit state characteristic to obtain a multi-source compensation characteristic;

[0118] The first X groups of channel features in the multi-source compensation features are determined as first compensation features; X is the number of main channels corresponding to the environmental features.

[0119] In an embodiment of the present invention, for example, in a large industrial production park, the server uses the dynamic weight fusion device in the first compensation strategy module to perform a compensation parameter fusion operation on the environmental characteristics based on the circuit state characteristics to obtain a first compensation characteristic. The server uses the dynamic weight fusion device to map the environmental characteristics and the circuit state characteristics. Assume that the environmental characteristics are represented by the vector [0.7, 0.6, 0.5], which correspond to temperature, humidity, and electromagnetic intensity respectively; the circuit state characteristics are represented by the vector [0.8, 0.4, 0.9], which correspond to electrical signal fluctuations, frequency, and phase respectively. Through the specific operation rules of the dynamic weight fusion device, the environmental characteristics are mapped into the second environmental correlation matrix [[0.3, 0.2, 0.1], [0.4, 0.3, 0.2]]; the circuit state characteristics are mapped into the second circuit correlation matrix [[0.5, 0.3, 0.2], [0.4, 0.3, 0.3]] and the second compensation parameter matrix [[0.6, 0.4, 0.5], [0.3, 0.5, 0.4]]. Next, the server performs operations on the relevant results to obtain the fifth coupling parameter matrix. The second environment correlation matrix is multiplied by the environment characteristics, resulting in [0.3*0.7+0.2×0.6+0.1×0.5, 0.4×0.7+0.3×0.6+0.2×0.5]=[0.38, 0.52]. The second circuit correlation matrix is multiplied by the circuit state characteristics to obtain [0.5×0.8+0.3×0.4+0.2×0.9, 0.4×0.8+0.3×0.4+0.3×0.9]=[0.7, 0.67]. After dimensionality conversion, this matrix becomes [[0.7], [0.67]]. Matrix multiplication of these two results yields the fifth coupling parameter matrix: [[0.38×0.7], [0.52×0.67]]=[[0.266], [0.3484]]. The server obtains a total of 3 channels for environmental characteristics. The server then normalizes the fifth coupling parameter matrix by the normalized scaling factor of 1 / 3 for the total number of channels to obtain the multi-source contribution coefficient matrix. Specifically, [[0.266 × 1 / 3], [0.3484 × 1 / 3]] = [[0.0887], [0.1161]]. The second compensation parameter matrix multiplied by the circuit state characteristics yields [0.6 × 0.8 + 0.4 × 0.4 + 0.5 × 0.9, 0.3 × 0.8 + 0.5 × 0.4 + 0.4 × 0.9] = [1.21, 1.0]. This result is matrix multiplied by the multi-source contribution coefficient matrix to obtain the multi-source compensation characteristics. Specifically, [[0.0887 × 1.21], [0.1161 × 1.0]] = [[0.1073], [0.1161]]. Finally, assume that long-term analysis of campus current sensor data indicates that temperature and humidity are the primary channels corresponding to environmental features, that is, X = 2. The server then determines the first two sets of channel features [0.1073, 0.1161] in the multi-source compensation feature set as the first compensation feature.This first compensation feature comprehensively considers the relationship between environmental characteristics and circuit state characteristics, providing key data for subsequent current sensor error compensation.

[0120] In an embodiment of the present invention, the first compensation strategy module includes a multi-source dynamic fuser;

[0121] In the first compensation strategy module, a compensation parameter fusion operation is performed on the environmental feature based on the circuit state feature to obtain a first compensation feature, which can be implemented through the following example.

[0122] Obtaining a fused compression feature, performing channel union on the fused compression feature and the circuit state feature to obtain a second joint feature; the fused compression feature is used to establish a connection between multiple source features;

[0123] Based on the first dynamic weight allocation branch in the multi-source dynamic fuser, a cross-source dynamic correlation analysis is performed on the second joint feature to obtain a second dynamic correlation feature, and the first Y groups of main channel features in the second dynamic correlation feature are determined as target compensation features; Y is the number of main channels corresponding to the fused compression feature;

[0124] Performing channel combination on the environmental feature and the target compensation feature to obtain a third combined feature, and performing cross-source dynamic correlation analysis on the third combined feature based on the second dynamic weight allocation branch in the multi-source dynamic fuser to obtain a third dynamic correlation feature;

[0125] The first X groups of channel features in the third dynamic association feature are determined as the first compensation feature, where X is the number of main channels corresponding to the environmental feature.

[0126] In an exemplary embodiment of the present invention, in a large industrial production park, a server utilizes a first compensation strategy module comprising a multi-source dynamic fusion module to perform compensation parameter fusion operations on environmental features based on circuit state features, generating a first compensation feature. The server first obtains a fused compressed feature. This feature is obtained by deeply analyzing and fusing various data related to current sensors within the park, such as historical environmental data and circuit performance indicators, and applying a specific compression algorithm. For example, by analyzing data such as temperature, humidity, and electromagnetic intensity at different times over the past week, as well as the electrical signal fluctuations and frequency of the current sensors at corresponding moments, algorithms such as principal component analysis are used to extract a set of key information as a fused compressed feature. Assume that this feature is represented by the vector [0.4, 0.3, 0.2]. The server then performs a channel-wise union of the fused compressed feature and the circuit state feature. Assume that the circuit state feature is [0.7, 0.5, 0.6, 0.8], representing the electrical signal fluctuation, frequency, phase, and amplitude, respectively. This union yields a second joint feature [0.4, 0.3, 0.2, 0.7, 0.5, 0.6, 0.8]. Next, the server uses the first dynamic weight assignment branch in the multi-source dynamic fusion engine to perform cross-source dynamic correlation analysis on the second joint feature. This first dynamic weight assignment branch considers potential connections between different features, such as the correlation between electrical signal fluctuations and temperature and humidity, and the correlation between frequency and electromagnetic intensity. Through complex algorithmic analysis, it assigns weights to each feature. For example, if electrical signal fluctuations are closely correlated with temperature, the corresponding feature is assigned a higher weight of 0.8, and other features are weighted accordingly. After weighted calculation and processing, the second dynamic correlation feature is obtained, assuming it is [0.32, 0.18, 0.12, 0.56, 0.25, 0.36, 0.64]. Assuming the number of main channels Y corresponding to the fused compressed feature is 2, the server identifies the first two main channel features [0.32, 0.18] in the second dynamic correlation feature as the target compensation features. The server then performs channel-wise combination of the environmental features and the target compensation features. Assuming the environmental feature is [0.6, 0.5, 0.4], the third joint feature [0.6, 0.5, 0.4, 0.32, 0.18] is obtained after combination. The server then performs cross-source dynamic correlation analysis on the third joint feature based on the second dynamic weight allocation branch in the multi-source dynamic fusion module. The second dynamic weight allocation branch re-examines the relationship between the features, reallocates weights, and performs weighted calculation and processing to obtain the third dynamic correlation feature, for example, [0.48, 0.35, 0.24, 0.16, 0.09]. Assuming the number of main channels X corresponding to the environmental feature is 3, the server determines the first three groups of channel features [0.48, 0.35, 0.24] in the third dynamic correlation feature as the first compensation feature. This first compensation feature fully integrates the relevant information of the environmental features and circuit state features, laying the foundation for subsequent accurate compensation of current sensor errors.

[0127] In an embodiment of the present invention, the first compensation strategy module includes a main channel fusion unit;

[0128] In the first compensation strategy module, a compensation parameter fusion operation is performed on the environmental feature based on the circuit state feature to obtain a first compensation feature, which can be implemented through the following example.

[0129] Acquire an environmental calibration feature corresponding to the environmental feature, and a circuit calibration feature corresponding to the circuit state feature;

[0130] Based on the first dynamic time series processor in the main channel fusion unit, the environmental feature is modulated by a time series signal to obtain a first time series response feature;

[0131] Based on the second dynamic timing processor in the main channel fusion unit, the circuit state feature is modulated by a timing signal to obtain a second timing response feature;

[0132] The second time series response characteristic is smoothed to obtain a time series smoothing characteristic, and the first time series response characteristic and the time series smoothing characteristic are combined to obtain a first compensation characteristic.

[0133] In an embodiment of the present invention, for example, in a large industrial production park, the server uses a first compensation strategy module including a main channel fusion unit to process environmental features and circuit state features to obtain a first compensation feature. The server first obtains the environmental calibration features corresponding to the environmental features and the circuit calibration features corresponding to the circuit state features. Assuming that the environmental features are [0.7, 0.6, 0.5], representing temperature, humidity, and electromagnetic intensity, respectively, according to a pre-set calibration rule, for example, each feature value is multiplied by a calibration coefficient of 0.8 to obtain the environmental calibration features [0.56, 0.48, 0.4]. For the circuit state features, assuming that they are [0.8, 0.4, 0.9], representing electrical signal fluctuations, frequency, and phase, similarly according to the corresponding calibration rule, such as multiplying by 0.9, the circuit calibration features [0.72, 0.36, 0.81] are obtained. Then, the server uses the first dynamic timing processor in the main channel fusion unit to perform timing signal modulation on the environmental features. In industrial production scenarios, environmental factors have certain changing patterns over time. For example, temperature gradually rises during daytime production hours, and humidity fluctuates in areas with dense equipment operation. The first dynamic time series processor modulates environmental features based on these time-series variations. It may add a time-dependent weight to the current environmental feature value based on historical data showing trends in temperature, humidity, and electromagnetic intensity at different times. Assuming the current time is during peak morning production, the temperature feature weight increases, resulting in a first time series response feature after modulation, such as [0.6, 0.5, 0.45]. Simultaneously, the server uses the second dynamic time series processor in the main channel fusion unit to perform time series signal modulation on the circuit state feature. Circuit states also change over time, such as when electrical signals exhibit different behavior during equipment startup, operation, and shutdown. The second dynamic time series processor modulates the circuit calibration features, taking into account changes in electrical signal fluctuations, frequency, and phase across different production stages. For example, when equipment is initially started, electrical signal fluctuations are significant, resulting in a corresponding feature weight increase, resulting in a second time series response feature after modulation, such as [0.85, 0.4, 0.9]. The server then smoothes this second time series response feature. During device operation, electrical signals may experience transient interference, causing characteristic value fluctuations. Smoothing can remove these unnecessary fluctuations. Using a specific smoothing algorithm, such as the moving average method, [0.85, 0.4, 0.9] is processed to obtain a time-series smoothing feature, assuming it is [0.83, 0.4, 0.88]. Finally, the server merges the first time-series response feature [0.6, 0.5, 0.45] with the time-series smoothing feature [0.83, 0.4, 0.88]. By sequentially concatenating or using a specific merging algorithm, the first compensation feature is obtained, for example, [0.6, 0.5, 0.45, 0.83, 0.4, 0.88].This first compensation feature integrates the time-series change information of environmental features and circuit state features, providing more comprehensive data support for subsequent current sensor error compensation.

[0134] In an embodiment of the present invention, the first compensation strategy module includes a reference signal suppressor;

[0135] In the first compensation strategy module, a compensation parameter fusion operation is performed on the environmental feature based on the circuit state feature to obtain a first compensation feature, which can be implemented through the following example.

[0136] The circuit state feature is cleared, and the environmental feature is used as a first compensation feature output by the reference signal suppressor.

[0137] In an exemplary embodiment of the present invention, in the complex electromagnetic environment of a large industrial production park, the server utilizes a first compensation strategy module, including a reference signal suppressor, to process the environmental characteristics and circuit state characteristics of the current sensor to obtain a first compensation characteristic. Within this park, the current sensor is surrounded by various interference sources, which can generate interference similar to the reference signal and affect measurement accuracy. When the server invokes this module, it first analyzes the currently acquired environmental characteristics and circuit state characteristics. Assume that the environmental characteristics are represented by the vector [0.6, 0.5, 0.4], representing temperature, humidity, and electromagnetic intensity, respectively; and the circuit state characteristics are represented by [0.7, 0.3, 0.8], representing electrical signal fluctuations, frequency, and phase. Based on its design principles, the reference signal suppressor assumes that, given the current compensation requirements, the signal contained in the circuit state characteristics may be the primary source of interference, similar to the reference signal, interfering with the accurate analysis of the environmental characteristics, thereby affecting the accuracy of the current sensor error compensation. Therefore, the server, in accordance with the functional settings of the reference signal suppressor, directly removes the circuit state characteristics [0.7, 0.3, 0.8]. The server then uses the environmental signature [0.6, 0.5, 0.4] as the first compensation signature output by the reference signal suppressor. This is because, in this case, after the reference signal suppressor determines that the circuit state signature has been removed, the environmental signature itself more accurately reflects information related to current sensor errors and can be used for subsequent error compensation calculations. For example, in a specific area of the campus, strong electromagnetic interference generated by the startup of large motors causes abnormal changes in parameters such as electrical signal fluctuations and frequency in the circuit state signature. These abnormal changes act as reference signals, interfering with the determination of actual current conditions. Environmental factors such as temperature, humidity, and electromagnetic intensity are relatively stable and can more directly correlate to potential current sensor errors. By removing the interfered circuit state signature and retaining the environmental signature as the first compensation signature, the server takes a critical step toward accurately compensating for current sensor errors. This ensures that current sensor data more accurately reflects actual current conditions in complex electromagnetic environments, providing strong support for the stable operation of industrial production equipment.

[0138] In the embodiment of the present invention, determining the current error compensation result corresponding to the current sensor based on the error type identifier, the first compensation feature, and the second compensation feature can be implemented through the following example.

[0139] Performing channel combination on the first compensation feature and the second compensation feature to obtain a fourth joint feature, and adding the error type identifier to the front end of the fourth joint feature to obtain a joint compensation feature;

[0140] The joint compensation feature is dynamically evaluated based on a cross-source dynamic weight allocation branch to obtain a dynamic weight fusion parameter, and a feature reconstruction operation is performed on the dynamic weight fusion parameter to obtain a current error compensation result corresponding to the current sensor.

[0141] In an exemplary embodiment of the present invention, in a large industrial production park, a server determines the current error compensation result corresponding to the current sensor based on the error type identifier, the first compensation feature, and the second compensation feature. Assume that the first compensation feature is represented by the vector [0.5, 0.3, 0.4]. This is obtained by processing environmental features using a specific compensation strategy module and reflects the impact of environmental factors on the current error. The second compensation feature is represented by the vector [0.6, 0.2, 0.7]. This is obtained by performing a compensation parameter fusion operation on the circuit state features and reflects the effect of the circuit state on the current error. The server first performs a channel union on the first and second compensation features. The two vectors are concatenated in sequence to obtain the fourth joint feature [0.5, 0.3, 0.4, 0.6, 0.2, 0.7]. At this point, the server obtains the error type identifier. For example, after in-depth analysis of the current sensor data, the current error type is determined to be "electromagnetic interference error." The system encodes this error type, assuming it is encoded as [1, 0, 0]. If multiple error types exist, they are represented using a one-hot encoding. The server adds the error type identifier to the front of the fourth joint feature, resulting in the joint compensation feature [1, 0, 0, 0.5, 0.3, 0.4, 0.6, 0.2, 0.7]. Next, the server dynamically evaluates the joint compensation feature based on the cross-source dynamic weight allocation branch. This branch analyzes the correlation between each element in the joint compensation feature and the error type, as well as the relationships between different feature sources. For example, it finds that the "1" in the error type identifier, representing electromagnetic interference error, is closely correlated with the electromagnetic intensity-related element in the first compensation feature (assumed to be 0.4 here) and the electrical signal fluctuation-related element in the second compensation feature (assumed to be 0.6 here). Based on this analysis, the cross-source dynamic weight allocation branch assigns different weights to each element of the joint compensation feature. After a series of complex calculations, the dynamic weight fusion parameter is obtained, assuming it is [0.2, 0.1, 0.1, 0.3, 0.1, 0.2]. Finally, the server performs feature reconstruction on the dynamic weight fusion parameter. Using a specific algorithm and a pre-trained model, the dynamic weight fusion parameter is converted into a numerical value that can be directly used to compensate for current sensor errors. For example, through matrix operations, nonlinear transformations, and other operations, the dynamic weight fusion parameters are mapped to the numerical range of current error compensation, ultimately obtaining the current error compensation result corresponding to the current sensor. This result may be a specific numerical value, which is used to adjust the current value output by the current sensor to correct errors caused by factors such as electromagnetic interference, ensuring that the data provided by the current sensor more accurately reflects the actual current situation and guarantees the stable operation of industrial production equipment.

[0142] An embodiment of the present invention provides a computer device 100, which includes a processor and a non-volatile memory storing computer instructions. When the computer instructions are executed by the processor, the computer device 100 executes the above-mentioned current sensor error compensation method. Figure 2 As shown, Figure 2 This is a block diagram of the structure of a computer device 100 provided in an embodiment of the present invention. Computer device 100 includes a memory 111, a processor 112, and a communication unit 113. To enable data transmission or exchange, memory 111, processor 112, and communication unit 113 are electrically connected to each other, directly or indirectly. For example, these components can be electrically connected via one or more communication buses or signal lines.

[0143] For illustrative purposes, the foregoing description has been made with reference to specific embodiments. However, the above illustrative discussion is not intended to be exhaustive or to limit the present disclosure to the precise forms disclosed. Numerous modifications and variations are possible in light of the above teachings. These embodiments have been selected and described in order to best illustrate the principles of the present disclosure and its practical application, thereby enabling those skilled in the art to best utilize the present disclosure and to utilize various embodiments with various modifications as appropriate for the specific application contemplated.

Claims

1. A current sensor error compensation method, characterized in that: include: Acquire an operating environment corresponding to the current sensor, and divide the operating environment into multiple time windows; Performing feature extraction on the multiple time windows to obtain environmental coding features corresponding to the multiple time windows; Dynamically evaluating the environmental coding features corresponding to the multiple time windows to obtain contribution coefficients corresponding to the multiple time windows; Dynamically fusing the environmental coding features corresponding to the multiple time windows and the contribution coefficients corresponding to the multiple time windows to obtain environmental features corresponding to the operating environment; Acquire multiple electrical signal time series segments corresponding to the current sensor, and fuse the multiple electrical signal time series segments into an electrical signal joint feature; Inputting the electrical signal joint features into a feature fusion network, performing a feature reconstruction operation on the electrical signal joint features based on the feature fusion network, and obtaining circuit state features corresponding to the current sensor; Obtaining an error type identifier, and determining a current error compensation result corresponding to the current sensor based on the error type identifier, the environmental characteristics, and the circuit state characteristics; The obtaining of the error type identifier and determining the current error compensation result corresponding to the current sensor based on the error type identifier, the environmental characteristics, and the circuit state characteristics includes: Aggregating the environmental features and the circuit state features within the same feature domain to obtain a first aggregated feature, and matching a first compensation strategy module for the environmental features in a plurality of compensation strategy modules based on the first aggregated feature; the plurality of compensation strategy modules configuring differentiated compensation strategies to adjust features of a plurality of data sources; In the first compensation strategy module, a compensation parameter fusion operation is performed on the environmental feature based on the circuit state feature to obtain a first compensation feature; Aggregating the circuit state feature and the first compensation feature in the same feature domain to obtain a second aggregate feature, and matching a second compensation strategy module for the circuit state feature among the multiple compensation strategy modules based on the second aggregate feature; In the second compensation strategy module, a compensation parameter fusion operation is performed on the circuit state feature based on the first compensation feature to obtain a second compensation feature; An error type identifier is obtained, and a current error compensation result corresponding to the current sensor is determined based on the error type identifier, the first compensation characteristic, and the second compensation characteristic.

2. A current sensor error compensation method according to claim 1, characterized in that: The aggregating the environmental feature and the circuit state feature in the same feature domain to obtain a first aggregated feature includes: Inputting the environmental features and the circuit state features into a dynamic fusion module, and performing feature alignment on the environmental features based on a first feature projection unit in the dynamic fusion module to obtain an environmental compensation feature; Performing feature alignment on the circuit state feature based on the second feature projection unit in the dynamic fusion module to obtain a circuit compensation feature; the environment compensation feature and the circuit compensation feature belong to the same feature domain; Acquiring a first effective value corresponding to the environmental compensation feature, and performing calibration processing on the environmental compensation feature based on the first effective value to obtain an environmental calibration feature; Dynamically adjust the gain of the environmental calibration feature to obtain an environmental gain feature, and smooth the environmental gain feature to obtain an environmental smoothing feature; Acquiring a second effective value corresponding to the circuit compensation feature, and performing calibration processing on the circuit compensation feature based on the second effective value to obtain a circuit calibration feature; Dynamically gain-adjusting the circuit calibration feature to obtain a circuit gain feature, and smoothing the circuit gain feature to obtain a circuit smoothing feature; The environmental smoothing feature and the circuit smoothing feature are combined to obtain a comprehensive compensation feature, and a feature reconstruction operation is performed on the comprehensive compensation feature based on a compensation parameter matrix corresponding to a parameter fusion module in the dynamic fusion module to obtain the first aggregated feature.

3. The current sensor error compensation method according to claim 2, characterized in that: The matching of a first compensation strategy module for the environmental feature among multiple compensation strategy modules based on the first aggregated feature includes: Performing a strategy evaluation on the first aggregated feature based on a strategy decision module in the dynamic fusion module to obtain fitness coefficients corresponding to multiple compensation strategy modules; The compensation strategy module corresponding to the largest fitness coefficient is determined as the first compensation strategy module matching the environmental features.

4. The current sensor error compensation method according to claim 1, characterized in that: The first compensation strategy module includes a dynamic weight fuser; The first compensation strategy module performs a compensation parameter fusion operation on the environmental feature based on the circuit state feature to obtain a first compensation feature, including: Perform channel combination on the environment feature and the circuit state feature to obtain a first combined feature; Based on the transformation compensation parameter matrix corresponding to the dynamic weight fuser, mapping the first joint feature into a first environment correlation matrix, a first circuit correlation matrix and a first compensation parameter matrix; constructing a first coupling parameter matrix based on a multiplication result between the first environment correlation matrix and the environment feature, and a multiplication result between the first environment correlation matrix and the circuit state feature; constructing a second coupling parameter matrix based on a dimensionality conversion operation result of a multiplication result between the first circuit association matrix and the environmental feature, and a dimensionality conversion operation result of a multiplication result between the first circuit association matrix and the circuit state feature; constructing a third coupling parameter matrix based on a multiplication result between the first compensation parameter matrix and the environmental characteristics, and a multiplication result between the first compensation parameter matrix and the circuit state characteristics; Determine the product of the first coupling parameter matrix and the second coupling parameter matrix as a fourth coupling parameter matrix, and obtain the total number of channels of the environmental feature; performing a normalized proportional calculation on the fourth coupling parameter matrix and the normalized scaling factor of the total number of channels to obtain a cross-source dynamic weight matrix, and determining a multiplication result of the cross-source dynamic weight matrix and the third coupling parameter matrix as a first dynamic correlation feature; The first X groups of channel features in the first dynamic association feature are determined as first compensation features, where X is the number of main channels corresponding to the environmental feature.

5. The current sensor error compensation method according to claim 1, characterized in that: The first compensation strategy module includes a dynamic weight fuser; The first compensation strategy module performs a compensation parameter fusion operation on the environmental feature based on the circuit state feature to obtain a first compensation feature, including: Based on the dynamic weight fuser, the environmental characteristics are mapped into a second environmental correlation matrix, and the circuit state characteristics are mapped into a second circuit correlation matrix and a second compensation parameter matrix; Performing a matrix multiplication operation on a dimension conversion operation result of a multiplication result between the second environment association matrix and the environment feature, and a dimension conversion operation result between the second circuit association matrix and the circuit state feature to obtain a fifth coupling parameter matrix, and acquiring a total number of channels of the environment feature; performing a normalized proportional calculation on the fifth coupling parameter matrix and a normalized scaling factor of the total number of channels to obtain a multi-source contribution coefficient matrix, and performing a matrix multiplication operation on the multi-source contribution coefficient matrix and a multiplication result of the second compensation parameter matrix and the circuit state characteristic to obtain a multi-source compensation characteristic; The first X groups of channel features in the multi-source compensation features are determined as first compensation features, where X is the number of main channels corresponding to the environmental features.

6. The current sensor error compensation method according to claim 1, characterized in that: The first compensation strategy module includes a multi-source dynamic fuser; The first compensation strategy module performs a compensation parameter fusion operation on the environmental feature based on the circuit state feature to obtain a first compensation feature, including: Obtaining a fused compression feature, performing channel union on the fused compression feature and the circuit state feature to obtain a second joint feature; the fused compression feature is used to establish a connection between multiple source features; Based on the first dynamic weight allocation branch in the multi-source dynamic fuser, a cross-source dynamic correlation analysis is performed on the second joint feature to obtain a second dynamic correlation feature, and the first Y groups of main channel features in the second dynamic correlation feature are determined as target compensation features; Y is the number of main channels corresponding to the fused compression feature; Performing channel combination on the environmental feature and the target compensation feature to obtain a third combined feature, and performing cross-source dynamic correlation analysis on the third combined feature based on the second dynamic weight allocation branch in the multi-source dynamic fuser to obtain a third dynamic correlation feature; Determine the first X groups of channel features in the third dynamic association feature as the first compensation feature; X is the number of main channels corresponding to the environmental feature; The first compensation strategy module further includes a main channel fusion unit; The first compensation strategy module performs a compensation parameter fusion operation on the environmental feature based on the circuit state feature to obtain a first compensation feature, including: Acquire an environmental calibration feature corresponding to the environmental feature, and a circuit calibration feature corresponding to the circuit state feature; Based on the first dynamic time series processor in the main channel fusion unit, the environmental feature is modulated by a time series signal to obtain a first time series response feature; Based on the second dynamic timing processor in the main channel fusion unit, the circuit state feature is modulated by a timing signal to obtain a second timing response feature; The second time series response characteristic is smoothed to obtain a time series smoothing characteristic, and the first time series response characteristic and the time series smoothing characteristic are combined to obtain a first compensation characteristic.

7. The current sensor error compensation method according to claim 1, characterized in that: The first compensation strategy module includes a reference signal suppressor; The first compensation strategy module performs a compensation parameter fusion operation on the environmental feature based on the circuit state feature to obtain a first compensation feature, including: The circuit state feature is cleared, and the environmental feature is used as a first compensation feature output by the reference signal suppressor.

8. The current sensor error compensation method according to claim 1, characterized in that: The determining, based on the error type identifier, the first compensation feature, and the second compensation feature, a current error compensation result corresponding to the current sensor includes: Performing channel combination on the first compensation feature and the second compensation feature to obtain a fourth joint feature, and adding the error type identifier to the front end of the fourth joint feature to obtain a joint compensation feature; The joint compensation feature is dynamically evaluated based on a cross-source dynamic weight allocation branch to obtain a dynamic weight fusion parameter, and a feature reconstruction operation is performed on the dynamic weight fusion parameter to obtain a current error compensation result corresponding to the current sensor.

9. A server system, characterized in that: The system comprises a server, wherein the server is used to execute the current sensor error compensation method according to any one of claims 1 to 8.

Images