Electromagnetic detection data recovery method and system of intelligent sensing equipment

Abstract

The invention provides an electromagnetic detection data recovery method and system for intelligent sensing equipment, and the method comprises the steps: firstly obtaining an electromagnetic detection original data set comprising a plurality of groups of detection data units, carrying out the missing and damage detection of the electromagnetic detection original data set, obtaining a data abnormal unit and the abnormal type information of the data abnormal unit, extracting associated redundant data of the data exception unit, generating a redundant data set, executing data recovery processing according to exception type information based on the redundant data set and an electromagnetic signal feature analysis result, generating a recovered data unit, and finally performing consistency verification processing on the recovered data unit. And generating a final valid data set containing the verified recovered data units and the unabnormal original data units. According to the method, data relevance and electromagnetic signal characteristics can be fully utilized, and the accuracy and reliability of electromagnetic detection data recovery are improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent sensing technology, and more specifically, to a method and system for recovering electromagnetic detection data from an intelligent sensing device. Background Technology

[0002] In the field of intelligent sensing devices, electromagnetic detection technology is widely used in environmental monitoring, equipment condition detection, and other scenarios. Intelligent sensing devices continuously collect electromagnetic detection data to obtain real-time information about the environment or equipment. However, in practical applications, the quality of electromagnetic detection data is often affected by a variety of factors.

[0003] On the one hand, due to hardware failures, communication interference, or unstable power supply in the sensing devices themselves, the collected electromagnetic detection data may be missing or corrupted. Missing data leads to incomplete information, affecting the accurate assessment of the environment or equipment status; corrupted data may introduce erroneous interference information, misleading subsequent data analysis and decision-making. On the other hand, existing data recovery methods are mostly quite simple, typically relying on simple interpolation or averaging algorithms to fill in the data, without fully considering the characteristics of the electromagnetic signals themselves or the correlation between data points. For example, the characteristics of electromagnetic signals may differ under different environmental parameters, and data from adjacent time periods under the same identification information may have continuity. However, existing methods often ignore these important factors, resulting in low accuracy and reliability of data recovery, failing to meet the high-quality electromagnetic detection data requirements of intelligent sensing devices. Summary of the Invention

[0004] In view of the aforementioned problems, and in conjunction with the first aspect of the present invention, embodiments of the present invention provide a method for recovering electromagnetic detection data from an intelligent sensing device, the method comprising: Acquire a raw electromagnetic detection data set, which includes multiple sets of continuously collected detection data units with identification information. Each set of detection data units consists of electromagnetic signal sampling values ​​and corresponding environmental parameter records. The original electromagnetic detection data set is subjected to missing or damaged detection processing to obtain data anomaly units and their anomaly type information, wherein the anomaly type information includes data missing type and data damaged type; Extract the associated redundant data of the data anomaly unit and generate a redundant data group corresponding to the data anomaly unit. The redundant data group includes detection data units in adjacent time periods under the same identification information and parallel detection data units under the same environmental parameter records. Based on the redundant data set and the electromagnetic signal feature analysis results, and according to the anomaly type information of the data anomaly unit, data recovery processing is performed to generate the recovered data unit. The electromagnetic signal feature analysis results include signal waveform continuity features and signal intensity distribution features. The recovered data units are subjected to consistency verification processing to generate a final valid data set, which includes the verified recovered data units and the original probe data units that are not abnormal.

[0005] In another aspect, embodiments of the present invention also provide an electromagnetic detection data recovery system for an intelligent sensing device, including a processor and a machine-readable storage medium connected to the processor. The machine-readable storage medium is used to store programs, instructions, or code, and the processor is used to execute the programs, instructions, or code in the machine-readable storage medium to implement the above-described method.

[0006] Based on the above, this embodiment of the invention acquires a raw electromagnetic detection data set containing multiple sets of continuously acquired detection data units with identification information. This allows for accurate identification of abnormal data units and their anomaly types during the missing / damaged detection processing of the raw electromagnetic detection data set. It extracts associated redundant data from these abnormal data units, generating a redundant data group containing detection data units from adjacent time periods with the same identification information and parallel detection data units recorded with the same environmental parameters. This fully utilizes the correlation between data, providing rich reference information for data recovery. Based on the redundant data group and electromagnetic signal feature analysis results, data recovery processing is performed according to the anomaly type information of the abnormal data units. This combines the waveform continuity characteristics and signal intensity distribution characteristics of the electromagnetic signals, improving the accuracy of data recovery. Consistency verification processing is performed on the recovered data units to ensure they meet the requirements of waveform consistency, intensity consistency, and environmental parameter correlation consistency, further improving data quality and reliability. Finally, a final valid data set is generated, containing verified recovered data units and non-abnormal raw detection data units, which helps improve the performance and application effect of intelligent sensing devices. Attached Figure Description

[0007] Figure 1 This is a schematic diagram of the execution flow of the electromagnetic detection data recovery method for intelligent sensing devices provided in this embodiment of the invention.

[0008] Figure 2 This is a schematic diagram of exemplary hardware and software components of the electromagnetic detection data recovery system for an intelligent sensing device provided in an embodiment of the present invention. Detailed Implementation

[0009] The present invention will now be described in detail with reference to the accompanying drawings. Figure 1 This is a flowchart illustrating an electromagnetic detection data recovery method for an intelligent sensing device according to an embodiment of the present invention. The electromagnetic detection data recovery method for the intelligent sensing device will be described in detail below.

[0010] Step S110: Obtain the electromagnetic detection raw data set, which contains multiple sets of detection data units with identification information collected continuously. Each set of detection data units consists of electromagnetic signal sampling values ​​and corresponding environmental parameter records.

[0011] In electromagnetic detection scenarios, intelligent sensing devices can continuously collect electromagnetic signals at set time intervals and simultaneously record environmental parameters at the time of collection. This collected data is systematically stored by the system, forming a raw electromagnetic detection data set. Each set of detection data units is equipped with identification information, which can take various forms, such as a timestamp to clearly identify the collection time of the data unit, or a device number to distinguish data collected by different sensing devices. Through these methods, each set of detection data units can be accurately identified and differentiated.

[0012] Each data unit contains two important parts: electromagnetic signal sample values ​​and corresponding environmental parameter records. The electromagnetic signal sample values ​​are the results of collecting electromagnetic signal values ​​at different times; they are a multi-dimensional sequence reflecting how the electromagnetic signal changes over time. The environmental parameter records contain various environmental factors during electromagnetic signal acquisition, such as temperature, humidity, and electromagnetic interference intensity. These environmental parameters affect the propagation and acquisition of the electromagnetic signal.

[0013] For example, in a complex electromagnetic environment, a smart sensing device might collect electromagnetic signals at specific time intervals, while simultaneously recording environmental parameters such as temperature, humidity, and the intensity of surrounding electromagnetic interference. This collected data is stored sequentially, forming a continuous set of raw electromagnetic detection data, where each data unit has unique identification information.

[0014] Step S120: Perform missing or damaged detection processing on the original electromagnetic detection data set to obtain data anomaly units and their anomaly type information, wherein the anomaly type information includes data missing type and data damaged type.

[0015] After obtaining the raw electromagnetic detection data set, various factors may affect the data during acquisition, transmission, or storage, leading to missing or corrupted data. Therefore, it is necessary to perform missing / corruption detection processing on the data set to identify abnormal data units and determine their anomaly types. The specific steps are as follows: Step S121: Perform integrity verification processing on the electromagnetic signal sampling values ​​of each group of detection data units in the original electromagnetic detection data set, and identify units with interruptions or length mismatches in the sampling value sequence as candidate units for missing data.

[0016] To verify the integrity of electromagnetic signal sampled values, a detailed analysis of the characteristics of the sampled value sequence is required. Specifically, it is necessary to check for interruptions or length mismatches in the sampled value sequence. If the sampled value sequence is abruptly interrupted at a certain point, or if the length of the sampled value sequence is inconsistent with the expected standard length, then that unit may have a data missing problem and can be identified as a candidate unit for data missing. To achieve this goal, the following specific operations are required: Step S1211: Extract the sequence length information of the electromagnetic signal sampling value of each group of detection data units, and obtain the standard sequence length information corresponding to the detection data unit.

[0017] In this step, the electromagnetic signal sample values ​​of each group of detection data units can be analyzed to statistically determine the length information of their sequences. Simultaneously, based on the sampling rules and preset parameters of the intelligent sensing device, the standard sequence length information corresponding to that detection data unit is obtained. The standard sequence length information is the ideal length that the electromagnetic signal sample value sequence should have. For example, when the intelligent sensing device is working normally, it samples according to a fixed sampling frequency and time period, so the electromagnetic signal sample value sequence of each data unit should have a definite length. By comparing the actual sequence length with the standard sequence length, a preliminary determination can be made as to whether there is a length mismatch.

[0018] Step S1212: Calculate the difference between the sequence length information of the electromagnetic signal sample value and the standard sequence length information, and identify units with a difference greater than a preset threshold as candidate units for length mismatch.

[0019] After obtaining the actual sequence length information and the standard sequence length information, the difference between them can be calculated. A preset threshold is a critical value set based on actual conditions and experience, used to determine whether the length difference exceeds an acceptable range. If the difference is greater than the preset threshold, it indicates that the electromagnetic signal sample value sequence length of the unit differs significantly from the standard length, potentially indicating missing data. This unit will be identified as a length mismatch candidate unit. For example, if the preset threshold is set to a specific length difference, and the difference between the electromagnetic signal sample value sequence length and the standard length of a set of probe data units exceeds this threshold, then that unit will be marked as a length mismatch candidate unit.

[0020] Step S1213: Perform interruption point detection processing on the electromagnetic signal sample values ​​of the candidate units with mismatched lengths. By analyzing the continuity of the timestamps of the sample values, identify units with timestamp jumps as interruption candidate units.

[0021] For probe data units identified as candidate units of length mismatch, further examination is needed to check for interruptions in their electromagnetic signal sampling values. This can be achieved by analyzing the timestamps of the sampling values. Timestamps record the acquisition time of each sample value; normally, timestamps should be continuous. If a jump is found in the timestamp sequence—that is, the interval between a timestamp and the previous timestamp is significantly greater than the normal sampling interval—it indicates a potential data interruption at that location. This unit will be identified as an interruption candidate unit. For example, in a continuous timestamp sequence, if a timestamp suddenly appears with an interval much larger than the normal sampling interval, it can be determined that the electromagnetic signal sampling value of this unit may be interrupted at that location.

[0022] Step S1214: Merge the length mismatch candidate units and the interruption candidate units to obtain the data missing candidate units.

[0023] After identifying candidate cells with length mismatch and candidate cells with interruptions, these two types of candidate cells are merged to obtain the final candidate cells with missing data. These candidate cells with missing data are cells that may have data missing issues and require further analysis and processing.

[0024] Step S122: Perform outlier detection processing on the electromagnetic signal sampling values ​​of each group of detection data units in the original electromagnetic detection data set, and identify units whose sampling values ​​deviate from the standard waveform range as candidate units of data corruption by using the signal waveform morphology matching method.

[0025] While verifying the integrity of electromagnetic signal sample values, outlier detection and processing are also required. Outliers may be caused by electromagnetic interference, equipment malfunctions, or other reasons, which can affect the accuracy and reliability of the data. To identify outliers, a signal waveform shape matching method is used.

[0026] First, a standard waveform range needs to be established. The standard waveform range represents the waveform characteristics that electromagnetic signal sample values ​​should possess under normal conditions. This can be achieved by analyzing and statistically processing a large amount of normally acquired data to obtain characteristics such as the rising edge shape, falling edge shape, and periodicity of the standard waveform. Then, the waveform of the electromagnetic signal sample values ​​for each group of probe data units is matched against the standard waveform range. If the waveform of the sample values ​​deviates from the standard waveform range, it indicates that the electromagnetic signal sample values ​​of that probe data unit may be corrupted, and it can be identified as a candidate unit for data corruption.

[0027] For example, the rising edge of a standard waveform should have a certain slope and shape. If the rising edge slope of the electromagnetic signal sample value of a certain set of probe data units is significantly different from the slope of the standard rising edge, or the shape of the rising edge changes significantly, then the unit may be identified as a candidate unit for data corruption.

[0028] Step S123: Perform identification information verification processing on the candidate units for missing data and candidate units for corrupted data, eliminate misjudged units caused by incorrect identification information, and obtain the abnormal data units.

[0029] After identifying candidate cells for missing or corrupted data, their identification information needs to be verified. Incorrect identification information may lead to misjudgment, identifying normal data cells as abnormal ones. Therefore, the accuracy of the identification information must be carefully checked.

[0030] For example, check whether the timestamp in the identification information matches the actual data collection time, and whether the device number matches the smart sensing device used. If an error is found in the identification information, the unit may be a misjudgment due to the incorrect identification information and needs to be excluded. After the identification information verification process, the remaining units are the data anomaly units.

[0031] Step S124: Perform type classification processing on the abnormal features of the data abnormal unit, determine the data missing type based on the interruption position feature of the missing candidate unit, determine the data damage type based on the deviation degree feature of the damage candidate unit, and generate abnormal type information.

[0032] For data anomaly units, it is necessary to further determine their anomaly type. For candidate units of missing data, their interruption location characteristics can be analyzed. Interruption location characteristics can reflect the specific circumstances of the data missing, such as whether the interruption occurs at the beginning, middle, or end of the sampled value sequence, and the length of the interruption. Based on these interruption location characteristics, the data missing type can be classified, such as beginning missing, middle missing, and end missing.

[0033] For candidate data corruption units, the degree of deviation from the standard waveform range can be analyzed. This deviation reflects the severity of the data corruption, such as the magnitude of the difference between the sampled value and the standard waveform, and the degree of change in the waveform shape. Based on these deviation characteristics, data corruption types can be classified, such as minor corruption, moderate corruption, and severe corruption. By classifying the anomaly characteristics of abnormal data units, anomaly type information containing both data missing type and data corruption type is generated.

[0034] Step S130: Extract the associated redundant data of the data anomaly unit and generate a redundant data group corresponding to the data anomaly unit. The redundant data group includes detection data units in adjacent time periods under the same identification information and parallel detection data units under the same environmental parameter records.

[0035] After identifying the data anomaly units and their anomaly types, it is necessary to extract their associated redundant data in order to recover this abnormal data. Associated redundant data refers to data that has a certain correlation with the data anomaly units; this associated redundant data can provide reference and basis for data recovery. The specific steps are as follows: Step S131: Obtain the identification information of the abnormal data unit, and filter the adjacent time period detection data units that are the same as the identification information in the electromagnetic detection raw data set. The adjacent time period detection data units include the detection data units of the time period before and after the abnormal unit.

[0036] In this step, the identification information of the data anomaly unit is first obtained. Then, the adjacent time period detection data units with the same identification information are selected from the original electromagnetic detection data set. Adjacent time period detection data units refer to detection data units that are temporally adjacent to the data anomaly unit, including detection data units from the time period before and after the anomaly unit. Because the electromagnetic signals of adjacent time periods have a certain continuity and correlation, these adjacent time period detection data units can provide important reference information for the recovery of the data anomaly unit.

[0037] For example, if the identification information of the abnormal data unit is a specific timestamp, then the detection data units of the time period before and after that timestamp will be selected from the original electromagnetic detection data set as adjacent time period detection data units.

[0038] Step S132: Obtain the environmental parameter record of the data anomaly unit, and filter the parallel detection data unit that is the same as the environmental parameter record in the electromagnetic detection raw data set. The parallel detection data unit includes detection data units of different sensing channels at the same acquisition time.

[0039] In addition to adjacent time-period data units, parallel data units also need to be acquired. Parallel data units refer to data units acquired by different sensor channels at the same acquisition time. Because data acquired by different sensor channels at the same time are affected by the same environmental factors, they exhibit a certain degree of correlation.

[0040] In this step, environmental parameter records of the data anomaly unit can be obtained, and then parallel detection data units with the same environmental parameter records can be selected from the original electromagnetic detection data set. For example, if the environmental parameter records of the data anomaly unit show parameters such as temperature, humidity, and electromagnetic interference intensity at the time of acquisition, then detection data units from different sensor channels with the same environmental parameter records at the same acquisition time will be selected from the original electromagnetic detection data set as parallel detection data units.

[0041] Step S133: Perform validity screening on the adjacent time period detection data units and parallel detection data units, exclude units with abnormal characteristics, and retain valid redundant data.

[0042] After acquiring adjacent time-period probe data units and parallel probe data units, they need to be filtered for validity. This is because these data units may contain anomalous characteristics, and directly using this anomalous data for data recovery can affect the accuracy and reliability of the recovery. To exclude anomalous units, the following steps are required: Step S1331: Perform missing or damaged detection processing on the adjacent time period detection data units to identify abnormal units.

[0043] In this step, missing or corrupted detection processing can be performed on the data units detected in adjacent time periods. The specific method is the same as the method for missing or corrupted detection processing of the original electromagnetic detection data set in step S120. That is, the integrity verification and outlier detection are performed on the electromagnetic signal sampling values ​​of the data units detected in adjacent time periods, and units with missing or corrupted data are identified as abnormal units.

[0044] Step S1332: Perform missing or corrupted detection processing on the parallel probe data unit to identify abnormal units.

[0045] Similarly, missing or damaged data units also need to be detected. Using the same method as in step S120, the electromagnetic signal sampling values ​​of the parallel probe data units are checked for integrity and outlier detection to identify any abnormal units.

[0046] Step S1333: The portion of the adjacent time period detection data unit that is not identified as an abnormal unit is taken as the effective redundant data of the adjacent time period.

[0047] After missing or damaged data detection, the portions of the data units detected in adjacent time periods that were not identified as abnormal units constitute the valid redundant data for those adjacent time periods. This data can be used for subsequent data recovery processing.

[0048] Step S1334: The portion of the parallel detection data unit that is not identified as an abnormal unit is used as effective redundant data of the parallel channel.

[0049] Similarly, the portion of the parallel probe data unit that is not identified as an abnormal unit is the effective redundant data of the parallel channel.

[0050] Step S1335: Merge the effective redundant data of adjacent time periods and the effective redundant data of parallel channels to obtain effective redundant data.

[0051] Finally, the effective redundant data from adjacent time periods and the effective redundant data from parallel channels are merged to obtain the final effective redundant data. This effective redundant data will be used for the recovery processing of data anomaly units.

[0052] Step S134: Sort the effective redundant data according to the degree of correlation with the data anomaly unit to generate a redundant data group containing effective redundant data of adjacent time periods and effective redundant data of parallel channels.

[0053] After obtaining the effective redundant data, it needs to be sorted according to its correlation with the data anomaly units. The correlation can be determined based on factors such as temporal correlation and environmental correlation. For example, effective redundant data in adjacent time periods are temporally adjacent to data anomaly units, and their temporal correlation is strong; effective redundant data in parallel channels are identical to data anomaly units in terms of environmental parameters, and their environmental correlation is strong.

[0054] After sorting the effective redundant data according to their correlation, the effective redundant data from adjacent time periods and the effective redundant data from parallel channels are combined to generate a redundant data group. The redundant data group contains effective redundant data related to the data anomaly unit.

[0055] Step S140: Based on the redundant data group and the electromagnetic signal feature analysis results, and according to the anomaly type information of the data anomaly unit, perform data recovery processing to generate the recovered data unit. The electromagnetic signal feature analysis results include signal waveform continuity features and signal intensity distribution features.

[0056] After obtaining the redundant data set and electromagnetic signal characteristic analysis results, data recovery processing needs to be performed on the data anomaly units based on the anomaly type information. The specific steps are as follows: Step S141: Perform signal waveform continuity analysis on the adjacent time period detection data units in the redundant data group, and extract the waveform change trend features of the electromagnetic signal sampling values ​​in adjacent time periods as the first continuity feature.

[0057] Electromagnetic signals in adjacent time periods exhibit a certain continuity. Therefore, by analyzing the waveform variation trend of the electromagnetic signal sample values ​​of the detection data unit in adjacent time periods, the waveform variation trend feature can be extracted as the first continuity feature. The first continuity feature can reflect the variation law of the electromagnetic signal waveform in adjacent time periods.

[0058] For example, analyzing the rising edge slope, falling edge slope, and periodic stability of electromagnetic signal sample values ​​in adjacent time periods can reflect the changing trend of the waveform. By analyzing and statistically processing the detection data units from multiple adjacent time periods, a more accurate first continuity characteristic can be obtained.

[0059] Step S142: Perform signal intensity distribution analysis on the parallel detection data units in the redundant data group, and extract the intensity distribution pattern features of the electromagnetic signal sampling values ​​of the parallel channel as the first distribution feature.

[0060] Parallel detection data units are affected by the same environmental factors at the same acquisition time, and the intensity distribution of their electromagnetic signals exhibits certain regularities. Therefore, signal intensity distribution analysis is performed on the parallel detection data units, and the intensity distribution pattern characteristics of the electromagnetic signal sampling values ​​of the parallel channels are extracted as the first distribution feature.

[0061] The primary distribution characteristic reflects the intensity distribution pattern of electromagnetic signals acquired by different sensing channels under the same environmental conditions. For example, analyzing the intensity mean, intensity variance, and intensity distribution curve of electromagnetic signal samples from parallel channels can describe the intensity distribution pattern. A more accurate primary distribution characteristic can be obtained by analyzing and statistically processing multiple parallel detection data units.

[0062] Step S143: Perform residual feature extraction processing on the original electromagnetic signal sampling value of the data anomaly unit, extract the local waveform features of the unmissing or undamaged parts as the second continuity features, and extract the local intensity features of the undamaged parts as the second distribution features.

[0063] Although the original electromagnetic signal sample values ​​of the data anomaly unit may contain missing or corrupted data, the unmissing or uncorrupted portions still contain some useful information. Therefore, feature extraction processing is required for these residual portions.

[0064] The local waveform features of the unmissing or undamaged portions are extracted as the second continuity feature. The second continuity feature can reflect the waveform features of the unaffected parts of the data anomaly unit. Combined with the first continuity feature, the waveform of the data anomaly unit can be recovered more accurately.

[0065] Simultaneously, the local intensity features of the undamaged portions are extracted as a second distribution feature. This second distribution feature reflects the intensity characteristics of the unaffected portions within the data anomaly units. Combined with the first distribution feature, it allows for a more accurate adjustment of the intensity distribution of the recovered sampled values.

[0066] Step S144: Construct a waveform recovery model by combining the first continuity feature and the second continuity feature, and use waveform trend fitting method to complete the electromagnetic signal sampling values ​​of the missing data unit or correct the abnormal sampling values ​​of the corrupted data unit.

[0067] To recover the waveform of an anomaly data unit, a waveform recovery model needs to be constructed by combining the first and second continuity features. The waveform recovery model can predict the waveform of the missing data portion or correct the abnormal waveform of the corrupted data portion based on the waveform characteristics of adjacent time periods and the anomaly data unit itself.

[0068] Specifically, a waveform trend fitting method is employed. This method uses the waveform's changing trend to fill in missing data units with sampled values, or to correct abnormal sampled values ​​in corrupted data units. To achieve this goal, the following specific operations are required: Step S1441: Extract the rising edge slope, falling edge slope and periodic stability parameters of the electromagnetic signal sampling values ​​in adjacent time periods based on the first continuity feature.

[0069] In this step, the rising edge slope, falling edge slope, and periodic stability parameter of the electromagnetic signal sample values ​​from adjacent time periods are extracted from the first continuity feature. These parameters reflect the changing patterns of the electromagnetic signal waveforms from adjacent time periods. For example, the rising edge slope reflects the rate of change of the electromagnetic signal during the rising phase, the falling edge slope reflects the rate of change of the electromagnetic signal during the falling phase, and the periodic stability parameter reflects the stability of the electromagnetic signal period. Accurate values ​​of these parameters are obtained by analyzing and statistically processing multiple electromagnetic signal sample values ​​from adjacent time periods.

[0070] Step S1442: Based on the second continuity feature, extract the local rising edge slope, local falling edge slope, and local periodic stability parameters of the data anomaly unit that is not missing or damaged.

[0071] Similarly, the local rising edge slope, local falling edge slope, and local periodic stability parameters of the unaffected or undamaged portions of the data anomaly unit are extracted from the second continuity feature. These local parameters reflect the waveform characteristics of the unaffected portion of the data anomaly unit itself. For example, the local rising edge slope reflects the rate of change of the rising edge of the unaffected or undamaged portion of the data anomaly unit, the local falling edge slope reflects the rate of change of the falling edge of that portion, and the local periodic stability parameter reflects the stability of the period of that portion. These local parameters are obtained by analyzing the electromagnetic signal sampling values ​​of the unaffected or undamaged portions of the data anomaly unit.

[0072] Step S1443: Perform a weighted average of the rising edge slopes of the adjacent time periods and the local rising edge slopes to generate the recovered rising edge slope.

[0073] After obtaining the rising edge slope and local rising edge slope of adjacent time periods, a weighted average is needed to comprehensively consider the characteristics of adjacent time periods and the data anomaly unit itself. The weighted average process assigns different weights to the two slopes based on the importance of adjacent time periods and the data anomaly unit, and then calculates the weighted average. For example, if the characteristics of adjacent time periods are considered more representative, a larger weight can be assigned to the rising edge slope of adjacent time periods; if the local characteristics of the data anomaly unit itself are considered more important, a larger weight can be assigned to the local rising edge slope. Through the weighted average process, the recovered rising edge slope is obtained, which can more accurately reflect the rising edge change trend after the data anomaly unit is recovered.

[0074] Step S1444: Perform a weighted average of the falling edge slopes of the adjacent time periods and the local falling edge slopes to generate the recovered falling edge slope.

[0075] Similar to the method for generating the recovery rising edge slope, a weighted average is applied to the falling edge slopes of adjacent time periods and the local falling edge slopes. Weights are assigned based on the importance of adjacent time periods and the data anomaly units, and the weighted average is calculated to obtain the recovery falling edge slope. The recovery falling edge slope can more accurately reflect the changing trend of the falling edge after the data anomaly unit is recovered.

[0076] Step S1445: Perform a weighted average of the periodic stability parameters of the adjacent time periods and the local periodic stability parameters to generate the restored periodic stability parameters.

[0077] Similarly, a weighted average is applied to the periodic stability parameters of adjacent time periods and the local periodic stability parameters. Weights are assigned based on their importance, and the weighted average is calculated to obtain the restored periodic stability parameter. The restored periodic stability parameter can more accurately reflect the degree of periodic stability after the recovery of data anomaly units.

[0078] Step S1446: Construct a waveform trend fitting function based on the recovery rising edge slope, recovery falling edge slope, and recovery period stability parameters.

[0079] After obtaining the recovery rising edge slope, recovery falling edge slope, and recovery period stability parameters, a waveform trend fitting function is constructed using these parameters. The waveform trend fitting function is a mathematical function that describes the waveform's changing trend based on the recovery rising edge slope, recovery falling edge slope, and recovery period stability parameters. For example, the waveform trend fitting function can determine the rate of change during the rising phase of the waveform based on the recovery rising edge slope, the rate of change during the falling phase based on the recovery falling edge slope, and the stability of the waveform period based on the recovery period stability parameters. By constructing the waveform trend fitting function, missing sample values ​​can be filled in for data gaps, or abnormal sample values ​​from corrupted data units can be corrected.

[0080] Step S1447: Use the waveform trend fitting function to fill in the missing sample values ​​of the missing data unit, or to correct the abnormal sample values ​​of the corrupted data unit, to obtain the initially recovered electromagnetic signal sample values.

[0081] After constructing the waveform trend fitting function, it is applied to the data anomaly units. For units with missing data, the missing sample values ​​are predicted based on the waveform trend fitting function and then filled in. For units with corrupted data, the abnormal sample values ​​are corrected based on the waveform trend fitting function to make them more consistent with normal waveform characteristics. After processing, the preliminary recovered electromagnetic signal sample values ​​are obtained.

[0082] Step S145: Construct an intensity recovery model by combining the first distribution feature and the second distribution feature, adjust the intensity distribution of the recovered sampled values ​​by using the intensity pattern matching method, and generate the recovered data unit.

[0083] After obtaining the initially recovered electromagnetic signal sample values, it is necessary to adjust their intensity distribution to better reflect the actual situation. To achieve this goal, an intensity recovery model is constructed by combining the first and second distribution characteristics.

[0084] The intensity recovery model can predict the intensity distribution of missing data or adjust the abnormal intensity distribution of damaged data based on the intensity characteristics of adjacent time periods and the data anomaly units themselves. An intensity pattern matching method is used to match the intensity distribution of the initially recovered electromagnetic signal sample values ​​with the intensity recovery model. If the intensity distribution of the initially recovered sample values ​​does not match the intensity recovery model, adjustments are necessary.

[0085] For example, the intensity recovery model can determine the mean intensity, variance intensity, and intensity distribution curve of an electromagnetic signal under the same environmental conditions based on the first and second distribution characteristics. The mean intensity, variance intensity, and intensity distribution curve of the initially recovered sampled values ​​are compared with the intensity recovery model. If differences exist, the magnitude or proportion of the sampled values ​​can be adjusted to make the intensity distribution more consistent with the intensity recovery model. After intensity pattern matching and adjustment, the recovered data units are generated.

[0086] Step S150: Perform consistency verification processing on the recovered data units to generate a final valid data set, which includes the recovered data units that have passed verification and the original probe data units that are not abnormal.

[0087] After obtaining the recovered data units, consistency verification is required to ensure their accuracy and reliability. Consistency verification mainly includes waveform consistency verification, intensity consistency verification, and environmental parameter correlation consistency verification. The specific steps are as follows: Step S151: Perform waveform consistency verification processing on the electromagnetic signal sample values ​​of the recovered data unit. By calculating the matching degree parameter between the recovered sample values ​​and the standard waveform template, determine whether the waveform recovery effect meets the requirements.

[0088] In this step, the waveform consistency verification process needs to be performed on the electromagnetic signal sample values ​​of the recovered data units. Specifically: Step S1511: Obtain the standard waveform template corresponding to the recovered data unit identification information. The standard waveform template includes a standard rising edge pattern, a standard falling edge pattern, and a standard periodic pattern. First, based on the identification information of the recovered data unit, the corresponding standard waveform template is searched from a pre-stored standard waveform template library. This standard waveform template is a representative waveform pattern obtained through statistical analysis based on a large amount of normally acquired data. The standard rising edge pattern describes the characteristics of the rising phase of the waveform under normal conditions, the standard falling edge pattern describes the characteristics of the falling phase, and the standard periodic pattern describes the characteristics of the waveform period.

[0089] Step S1512: Extract the actual rising edge shape, actual falling edge shape, and actual period shape of the electromagnetic signal sampled values ​​of the recovered data unit. Perform a detailed analysis of the electromagnetic signal sampled values ​​of the recovered data unit, and identify the morphological characteristics of the actual rising edge, actual falling edge, and actual period using a specific signal processing algorithm. For example, for the actual rising edge shape, analyze its rising rate, the start and end positions of the rise, etc.; for the actual falling edge shape, analyze its falling rate and position, etc.; for the actual period shape, analyze its period duration and stability, etc.

[0090] Step S1513: Calculate the shape similarity parameter between the actual rising edge pattern and the standard rising edge pattern. Compare the features of the actual rising edge pattern with the corresponding features of the standard rising edge pattern, and obtain the shape similarity parameter between the two using a specific similarity calculation method. This parameter reflects the degree of similarity between the actual rising edge pattern and the standard rising edge pattern; the closer the parameter value is to 1, the more similar the two are.

[0091] Step S1514: Calculate the shape similarity parameters between the actual falling edge shape and the standard falling edge shape. Similarly, compare the features of the actual falling edge shape with the corresponding features of the standard falling edge shape, and use the same or similar similarity calculation method to obtain the shape similarity parameters between the two.

[0092] Step S1515: Calculate the time matching parameter between the actual cycle pattern and the standard cycle pattern. Analyze the corresponding characteristics of the actual cycle pattern, such as cycle length and stability, with the standard cycle pattern. Calculate the degree of time matching between the two to obtain the time matching parameter. This parameter reflects the consistency between the actual cycle pattern and the standard cycle pattern in terms of time.

[0093] Step S1516: The shape similarity parameter and the time matching parameter are weighted and summed to generate a waveform matching parameter. Based on the importance of the shape similarity parameter and the time matching parameter in judging the overall waveform matching degree, different weights are assigned to them, and then a weighted summation is performed to obtain the waveform matching parameter. This parameter integrates the matching of the rising edge, falling edge, and period, and can more comprehensively reflect the matching degree between the recovered sampled value waveform and the standard waveform.

[0094] Step S1517: Determine whether the waveform matching degree parameter is greater than the preset matching threshold. If so, the waveform recovery effect is deemed to meet the requirements; otherwise, it is deemed not to meet the requirements. Compare the calculated waveform matching degree parameter with the preset matching threshold. If the waveform matching degree parameter is greater than the preset matching threshold, it indicates that the waveform of the recovered sampled value matches the standard waveform to a high degree, and the waveform recovery effect meets the requirements; otherwise, it indicates that the waveform recovery effect does not meet the requirements.

[0095] Step S152: Perform intensity consistency verification processing on the electromagnetic signal sampling values ​​of the recovered data unit. By calculating the overlap parameter between the recovered sampling values ​​and the standard intensity range, determine whether the intensity recovery effect meets the requirements.

[0096] In addition to waveform consistency verification, the electromagnetic signal sample values ​​of the recovered data units also need to undergo intensity consistency verification. First, the standard intensity range needs to be determined. The standard intensity range is the range of intensity that electromagnetic signal sample values ​​should have under normal conditions.

[0097] Then, the electromagnetic signal sample values ​​of the recovered data units are compared with the standard intensity range to calculate their overlap parameter. The overlap parameter reflects the degree of matching between the intensity of the recovered sample values ​​and the standard intensity range. For example, if the intensity of the recovered sample values ​​mostly falls within the standard intensity range, the overlap parameter will be high; conversely, if the intensity of the recovered sample values ​​mostly falls outside the standard intensity range, the overlap parameter will be low.

[0098] Finally, by comparing the overlap parameter with the preset intensity matching threshold, it is determined whether the intensity recovery effect meets the requirements. If the overlap parameter is greater than the preset intensity matching threshold, the intensity recovery effect meets the requirements; otherwise, the intensity recovery effect does not meet the requirements.

[0099] Step S153: Perform an association consistency verification process on the environmental parameter records of the recovered data unit. By checking the logical correspondence between the environmental parameter records and the electromagnetic signal sampling values, determine whether the environmental parameter records match the recovered sampling values.

[0100] For example, step S1531: Extract the environmental parameter records of the recovered data unit. The environmental parameter records include temperature parameters, humidity parameters, and electromagnetic interference intensity parameters. These environmental parameters, which are recorded synchronously during electromagnetic signal acquisition, are extracted from the recovered data unit and affect the propagation and acquisition of electromagnetic signals.

[0101] Step S1532: Extract the characteristic parameters of the electromagnetic signal sample values ​​of the recovered data unit. These characteristic parameters include the average signal amplitude, average signal frequency, and average signal-to-noise ratio. Statistical analysis is performed on the electromagnetic signal sample values ​​of the recovered data unit to calculate characteristic parameters such as the average signal amplitude, average signal frequency, and average signal-to-noise ratio. These characteristic parameters reflect the basic characteristics of the electromagnetic signal.

[0102] Step S1533: Obtain preset correlation rules between environmental parameters and electromagnetic signal characteristics. These preset correlation rules include a positive correlation between temperature parameters and the average signal amplitude, a negative correlation between humidity parameters and the average signal frequency, and a positive correlation between electromagnetic interference intensity parameters and the average signal-to-noise ratio. These preset correlation rules are summarized based on extensive experiments and data analysis, describing the inherent logical relationship between environmental parameters and electromagnetic signal characteristics.

[0103] Step S1534: Check whether the environmental parameter records and feature parameters of the recovered data unit satisfy the preset association rules. Substitute the extracted environmental parameter records and feature parameters into the preset association rules for checking. For example, check whether the temperature parameter and the average signal amplitude are positively correlated, that is, whether the average signal amplitude increases accordingly as the temperature rises; check whether the humidity parameter and the average signal frequency are negatively correlated; check whether the electromagnetic interference intensity parameter and the average signal-to-noise ratio are positively correlated.

[0104] Step S1535: If all preset association rules are met, the environmental parameter record and the recovered sampled value are determined to match; otherwise, they are determined not to match. If the environmental parameter record and the feature parameter meet all preset association rules, it means that the environmental parameter record and the recovered sampled value are logically corresponding and they match; if there are cases where the preset association rules are not met, the environmental parameter record and the recovered sampled value are determined not to match.

[0105] Step S154: Select the recovered data units that simultaneously meet the waveform consistency requirements, intensity consistency requirements, and environmental parameter correlation consistency requirements as valid recovery units.

[0106] After waveform consistency verification, intensity consistency verification, and environmental parameter correlation consistency verification, the recovered data units that simultaneously meet all three requirements are selected as valid recovery units. The electromagnetic signal sample values ​​of valid recovery units meet the requirements in terms of waveform, intensity, and correlation with environmental parameters, exhibiting high accuracy and reliability.

[0107] Step S155: Merge the effective recovery unit with the original detection data units that are not abnormal in the original electromagnetic detection data set to generate the final effective data set.

[0108] After identifying the valid recovered data units, they are merged with the original electromagnetic detection data units that are not abnormal from the raw data set. This merging process combines the valid recovered data units and the original, non-abnormal data units to form the final valid data set. The final valid data set contains both verified recovered data units and original, non-abnormal data units, which can be used for subsequent data analysis and applications.

[0109] In summary, the electromagnetic detection data recovery method for intelligent sensing devices provided in this embodiment acquires the original electromagnetic detection data set, performs missing or damaged detection processing, extracts associated redundant data, performs data recovery processing based on the redundant data and electromagnetic signal feature analysis results, and finally performs consistency verification processing on the recovered data units to generate a final valid data set. This method can effectively recover missing and damaged parts of electromagnetic detection data, improving data integrity and accuracy.

[0110] Figure 2 The illustration shows exemplary hardware and software components of an electromagnetic detection data recovery system 100 for an intelligent sensing device that can implement the ideas of this application, according to some embodiments of this application. For example, a processor 120 can be used in the electromagnetic detection data recovery system 100 for an intelligent sensing device and to perform the functions described in this application.

[0111] The electromagnetic detection data recovery system 100 for intelligent sensing devices can be a general-purpose server or a special-purpose server; both can be used to implement the electromagnetic detection data recovery method for intelligent sensing devices described in this application. Although only one server is shown in this application, for convenience, the functions described in this application can be implemented in a distributed manner on multiple similar platforms to balance the load.

[0112] For example, the electromagnetic detection data recovery system 100 of an intelligent sensing device may include a network port 110 connected to a network, one or more processors 120 for executing program instructions, a communication bus 130, and various forms of storage media 140, such as a disk, ROM, or RAM, or any combination thereof. Exemplarily, the electromagnetic detection data recovery system 100 of an intelligent sensing device may also include program instructions stored in ROM, RAM, or other types of non-transitory storage media, or any combination thereof. The methods of this application can be implemented according to these program instructions. The electromagnetic detection data recovery system 100 of an intelligent sensing device also includes an I / O interface 150 between a computer and other input / output devices.

[0113] For ease of explanation, only one processor is described in the electromagnetic detection data recovery system 100 for intelligent sensing devices. However, it should be noted that the electromagnetic detection data recovery system 100 for intelligent sensing devices in this application may also include multiple processors. Therefore, the steps performed by one processor as described in this application may also be performed jointly or individually by multiple processors. For example, if the processor of the electromagnetic detection data recovery system 100 for intelligent sensing devices performs steps A and B, it should be understood that steps A and B may also be performed jointly by two different processors or individually by one processor. For example, the first processor performs step A, the second processor performs step B, or the first processor and the second processor jointly perform steps A and B.

[0114] Furthermore, this embodiment of the invention also provides a readable storage medium, wherein computer-executable instructions are preset in the readable storage medium, and when the processor executes the computer-executable instructions, the electromagnetic detection data recovery method of the above-mentioned intelligent sensing device is implemented.

[0115] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. A method for recovering electromagnetic detection data from an intelligent sensing device, characterized in that, The method includes: Acquire a raw electromagnetic detection data set, which includes multiple sets of continuously collected detection data units with identification information. Each set of detection data units consists of electromagnetic signal sampling values ​​and corresponding environmental parameter records. The original electromagnetic detection data set is subjected to missing or damaged detection processing to obtain data anomaly units and their anomaly type information, wherein the anomaly type information includes data missing type and data damaged type; Extract the associated redundant data of the data anomaly unit and generate a redundant data group corresponding to the data anomaly unit. The redundant data group includes detection data units in adjacent time periods under the same identification information and parallel detection data units under the same environmental parameter records. Based on the redundant data set and the electromagnetic signal feature analysis results, and according to the anomaly type information of the data anomaly unit, data recovery processing is performed to generate the recovered data unit. The electromagnetic signal feature analysis results include signal waveform continuity features and signal intensity distribution features. The recovered data units are subjected to consistency verification processing to generate a final valid data set, which includes the verified recovered data units and the original probe data units that are not abnormal.

2. The electromagnetic detection data recovery method of the intelligent sensing device according to claim 1, characterized in that, The process of performing missing or corrupted detection on the original electromagnetic detection data set to obtain data anomaly units and their anomaly type information includes: The electromagnetic signal sampling values ​​of each group of detection data units in the original electromagnetic detection data set are subjected to integrity verification processing, and units with interruptions or length mismatches in the sampling value sequence are identified as candidate units for missing data. Anomaly detection processing is performed on the electromagnetic signal sample values ​​of each group of detection data units in the original electromagnetic detection data set. Units whose sample values ​​deviate from the standard waveform range are identified as candidate units for data corruption by the signal waveform morphology matching method. The candidate units for missing data and candidate units for corrupt data are checked for identification information. Units that are misjudged due to incorrect identification information are excluded, and the abnormal data units are obtained. The abnormal features of the data anomaly units are classified, the data missing type is determined based on the interruption position feature of the missing candidate unit, the data corruption type is determined based on the deviation degree feature of the corruption candidate unit, and anomaly type information is generated.

3. The electromagnetic detection data recovery method of the intelligent sensing device according to claim 2, characterized in that, The process of performing integrity verification on the electromagnetic signal sample values ​​of each group of detection data units in the original electromagnetic detection data set, and identifying units with interruptions or length mismatches in the sample value sequence as candidate units for missing data, includes: Extract the sequence length information of the electromagnetic signal sample values ​​of each group of detection data units, and obtain the standard sequence length information corresponding to that detection data unit; Calculate the difference between the sequence length information of the electromagnetic signal sampled value and the standard sequence length information, and identify units with a difference greater than a preset threshold as candidate units of length mismatch; The electromagnetic signal sampling values ​​of the candidate units with mismatched lengths are subjected to interruption point detection processing. By analyzing the continuity of the sampling value timestamps, units with timestamp jumps are identified as interruption candidate units. The length mismatch candidate unit and the interrupt candidate unit are merged to obtain the data missing candidate unit.

4. The electromagnetic detection data recovery method of the intelligent sensing device according to claim 1, characterized in that, The step of extracting the associated redundant data of the data anomaly unit and generating a redundant data group corresponding to the data anomaly unit includes: Obtain the identification information of the abnormal data unit, and filter the adjacent time period detection data units that are the same as the identification information in the original electromagnetic detection data set. The adjacent time period detection data units include the detection data units of the time period before and after the abnormal unit. Obtain the environmental parameter record of the data anomaly unit, and filter the parallel detection data unit that is the same as the environmental parameter record in the electromagnetic detection raw data set. The parallel detection data unit includes detection data units of different sensing channels at the same acquisition time. The adjacent time period detection data units and parallel detection data units are subjected to validity screening to exclude units with abnormal characteristics and retain valid redundant data. The effective redundant data are sorted according to their correlation with the data anomaly units to generate a redundant data group containing effective redundant data from adjacent time periods and effective redundant data from parallel channels.

5. The electromagnetic detection data recovery method of the intelligent sensing device according to claim 4, characterized in that, The validity screening process for the adjacent time period detection data units and parallel detection data units, excluding units with abnormal characteristics and retaining valid redundant data, includes: Perform missing or corrupted detection processing on the adjacent time period detection data units to identify abnormal units; Perform missing or corrupted detection processing on the parallel probe data units to identify abnormal units; The portion of the adjacent time period detection data unit that is not identified as an abnormal unit is taken as the effective redundant data of the adjacent time period. The portion of the parallel detection data unit that is not identified as an abnormal unit is used as effective redundant data for the parallel channel. The effective redundant data of adjacent time periods and the effective redundant data of parallel channels are merged to obtain the effective redundant data.

6. The electromagnetic detection data recovery method of the intelligent sensing device according to claim 1, characterized in that, Based on the analysis results of the redundant data group and electromagnetic signal characteristics, and according to the anomaly type information of the data anomaly unit, data recovery processing is performed to generate the recovered data unit, including: The signal waveform continuity analysis is performed on the adjacent time period detection data units in the redundant data group, and the waveform change trend characteristics of the electromagnetic signal sampling values ​​in adjacent time periods are extracted as the first continuity feature. The parallel detection data units in the redundant data group are subjected to signal intensity distribution analysis and processing, and the intensity distribution pattern features of the electromagnetic signal sampling values ​​of the parallel channel are extracted as the first distribution feature. The original electromagnetic signal sampling values ​​of the data anomaly unit are subjected to residual feature extraction processing. The local waveform features of the unmissing or undamaged parts are extracted as the second continuity features, and the local intensity features of the undamaged parts are extracted as the second distribution features. A waveform recovery model is constructed by combining the first and second continuity features, and the electromagnetic signal sampling values ​​of missing data units or the abnormal sampling values ​​of corrupted data units are supplemented by waveform trend fitting methods. An intensity recovery model is constructed by combining the first and second distribution features. The intensity distribution of the recovered sampled values ​​is adjusted by the intensity pattern matching method to generate the recovered data units.

7. The electromagnetic detection data recovery method of the intelligent sensing device according to claim 6, characterized in that, The step of constructing a waveform recovery model by combining the first and second continuity features, and supplementing the electromagnetic signal sampling values ​​of missing data units or correcting the abnormal sampling values ​​of corrupted data units using a waveform trend fitting method, includes: Based on the first continuity feature, the rising edge slope, falling edge slope, and periodic stability parameters of the electromagnetic signal sampling values ​​in adjacent time periods are extracted; Based on the second continuity feature, extract the local rising edge slope, local falling edge slope, and local periodic stability parameters of the non-missing or undamaged parts of the abnormal data units. The rising edge slope of the adjacent time periods is weighted and averaged with the local rising edge slope to generate the recovered rising edge slope. The falling edge slope of the adjacent time periods is weighted and averaged with the local falling edge slope to generate the recovered falling edge slope; The periodic stability parameters of adjacent time periods are weighted and averaged with the local periodic stability parameters to generate the restored periodic stability parameters. A waveform trend fitting function is constructed based on the recovery rising edge slope, recovery falling edge slope, and recovery period stability parameters; The waveform trend fitting function is used to fill in the missing sample values ​​of the missing data units or to correct the abnormal sample values ​​of the corrupted data units, so as to obtain the preliminary recovered electromagnetic signal sample values.

8. The electromagnetic detection data recovery method of the intelligent sensing device according to claim 1, characterized in that, The process of performing consistency verification on the recovered data units to generate a final valid data set includes: The electromagnetic signal sample values ​​of the recovered data unit are subjected to waveform consistency verification processing. By calculating the matching degree parameter between the recovered sample values ​​and the standard waveform template, it is determined whether the waveform recovery effect meets the requirements. The electromagnetic signal sampling values ​​of the recovered data unit are subjected to intensity consistency verification processing. By calculating the overlap parameter between the recovered sampling values ​​and the standard intensity range, it is determined whether the intensity recovery effect meets the requirements. The environmental parameter records of the recovered data unit are subjected to association consistency verification processing. By checking the logical correspondence between the environmental parameter records and the electromagnetic signal sampling values, it is determined whether the environmental parameter records match the recovered sampling values. Data units that simultaneously meet waveform consistency requirements, intensity consistency requirements, and environmental parameter correlation consistency requirements are selected as valid recovery units. The effective recovery unit is merged with the original detection data units that are not abnormal in the original electromagnetic detection data set to generate the final effective data set.

9. The electromagnetic detection data recovery method of the intelligent sensing device according to claim 8, characterized in that, The step of performing waveform consistency verification processing on the electromagnetic signal sample values ​​of the recovered data unit, and determining whether the waveform recovery effect meets the requirements by calculating the matching degree parameter between the recovered sample values ​​and the standard waveform template, includes: Obtain a standard waveform template corresponding to the recovered data unit identification information. The standard waveform template includes a standard rising edge pattern, a standard falling edge pattern, and a standard periodic pattern. Extract the actual rising edge shape, actual falling edge shape, and actual period shape of the electromagnetic signal sampling value of the recovered data unit; Calculate the shape similarity parameter between the actual rising edge shape and the standard rising edge shape; Calculate the shape similarity parameter between the actual falling edge shape and the standard falling edge shape; Calculate the time matching parameter between the actual cycle pattern and the standard cycle pattern; The shape similarity parameter and the time matching parameter are weighted and summed to generate the waveform matching parameter; Determine whether the waveform matching degree parameter is greater than the preset matching threshold. If it is, the waveform recovery effect is determined to meet the requirements; otherwise, it is determined to not meet the requirements.

10. An electromagnetic detection data recovery system for an intelligent sensing device, characterized in that, The device includes a processor and a memory, the memory and the processor being connected. The memory is used to store programs, instructions, or code, and the processor is used to execute the programs, instructions, or code in the memory to implement the electromagnetic detection data recovery method of the intelligent sensing device according to any one of claims 1-9.

Images