EMI disturbance source automatic identification and classification method and system based on artificial intelligence

Abstract

The invention provides an EMI disturbance source automatic identification and classification method and system based on artificial intelligence, and belongs to the technical field of data processing, and the method specifically comprises the steps: carrying out the determination of a combined identification and classification processing strategy of interval interference signals of a disturbance source based on the change data of the time sequence characteristics of different interval interference signals, and carrying out the classification of the disturbance source; according to a joint identification classification processing strategy of different interval interference signals in a disturbance source, and in combination with distribution data of the interval interference signals of the disturbance source in electromagnetic interference signals of the disturbance source, an updating processing method of the joint identification processing strategy of the disturbance source when the interval interference signals change is determined. And the reliability degree of identification processing of the electromagnetic interference signal is improved.

Description

Technical Field

[0001] This invention belongs to the field of data processing technology, and in particular relates to an automatic identification and classification method and system for EMI interference sources based on artificial intelligence. Background Technology

[0002] During equipment testing in a darkroom, some sources of interference are inevitable. Currently, the identification of these interference sources is often done manually, which is not only time-consuming and labor-intensive, but also often fails to effectively detect them.

[0003] Existing technical solutions utilize artificial intelligence models to identify and process electromagnetic interference signals from different sources, thereby improving the reliability of source identification and processing. However, these solutions suffer from the following technical problems: Due to the influence of the external environment, the electromagnetic interference signals of interference sources may appear intermittently, that is, at regular intervals. Therefore, they will inevitably overlap with other intermittently appearing electromagnetic interference signals, making it difficult to identify and process the interference source. Thus, determining how to jointly identify and classify all electromagnetic interference signals in the interference source based on the intermittently appearing electromagnetic interference signal data, and thus determining whether the electromagnetic interference signals of the interference source can be effectively identified, has become an urgent technical problem to be solved.

[0004] Therefore, there is an urgent need for an artificial intelligence-based method and system for automatic identification and classification of EMI interference sources. Summary of the Invention

[0005] To achieve the objectives of this invention, the following technical solution is adopted: Specifically, this application provides an artificial intelligence-based method for automatic identification and classification of EMI interference sources, which includes: S1 uses the identification data of the interference source in the dark room to determine the interval interference signal of the interference source. Based on the monitoring data of the interval interference signal of the interference source, it determines the interference source that is jointly identified and classified in the artificial intelligence model with the interval interference signal of all the interference sources, and uses it as a reliable interference source for analysis. S2, based on the reliable analysis of the composition of the interference sources in the anechoic chamber and combined with the monitoring data of the variation of the interval interference signals of different interference sources, determines the joint identification and classification processing strategy of the interval interference signals of the interference sources when the temporal characteristics of the interval interference signals of the interference sources change, and when it is not necessary to use a preset strategy for joint identification and classification processing, the joint identification and classification processing strategy of the interval interference signals of the interference sources is determined based on the data of the variation of the temporal characteristics of different interval interference signals. S3 determines the update processing method of the joint identification and classification processing strategy when the interval interference signal of the interference source changes, based on the joint identification and classification processing strategy of different interval interference signals in the interference source and the distribution data of the interval interference signal of the interference source in the electromagnetic interference signal of the interference source.

[0006] The beneficial effects of this invention are as follows: Based on the variation data of the temporal characteristics of different interval interference signals, a joint identification and classification processing strategy for the interval interference signals of the interference source is determined. Thus, when the temporal characteristics of the interval interference signals change, i.e. when the interval duration of the electromagnetic interference signals changes, the reliability of the joint identification and analysis processing of interval interference signals with different varying frequencies and the consistency with the temporal characteristics of other interval interference signals are fully considered. The determination of the interval interference signals for joint identification and analysis processing is carried out, thereby ensuring the reliability of the classification and identification processing of interval interference signals.

[0007] Based on the joint identification and classification processing strategy of different interval interference signals in the interference source and the distribution data of the interval interference signals in the electromagnetic interference signals of the interference source, an update processing method for the joint identification and classification strategy of the interference source when the interval interference signals change is determined. This method takes into account both the frequency of changes in the temporal characteristics of the electromagnetic interference signals in the interference source and the consistency of the temporal characteristics of the interval interference signals in the interference source with those of other interval interference signals. This enables timely updates to the joint identification and classification strategy of all electromagnetic interference signals in the interference source, ensuring the reliability of the classification and identification processing of some interference sources.

[0008] Furthermore, the identification data of the interference source includes the electromagnetic interference signal of the interference source and the timing characteristics of the electromagnetic interference signal.

[0009] Furthermore, the interval interference signal of the disturbance source is an electromagnetic interference signal with an interval distribution in terms of time sequence characteristics, that is, there are interval periods between the time periods in which electromagnetic interference signals exist.

[0010] Furthermore, the method for reliably analyzing and determining the source of interference is as follows: Based on the monitoring data of the interval interference signals of the interference source, the time periods in which the interval interference signals exist in the interference source are determined and are taken as the interference periods; Based on the distribution data of the interference periods, the dates on which the interference periods exist in the interference sources are determined and used as the interference dates; Based on the composition data of the interference dates in different interference sources, it is determined whether the interference source is a reliable source for analysis.

[0011] Furthermore, the method for determining the update processing method of the joint identification and processing strategy for the harassment source is as follows: Using a joint identification and classification processing strategy for different interval interference signals in the aforementioned interference source, different interval interference signals are identified, and the interval interference signals that are jointly identified and classified are used as matching interference signals. Based on the distribution data of the interval interference signal of the interference source in the electromagnetic interference signal of the interference source, the proportion of the interval interference signal of the interference source in all electromagnetic interference signals is determined, and it is used as the interval influence coefficient. Based on the interval influence coefficient of the disturbance source and the matching interference signal of different interval interference signals, an update processing method for the joint identification and processing strategy of the disturbance source when the interval interference signal changes is determined.

[0012] Secondly, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the aforementioned method for automatic identification and classification of EMI interference sources based on artificial intelligence when running the computer program.

[0013] Other features and advantages will be set forth in the following description, and the objects and other advantages of the invention are realized and obtained through the structures particularly pointed out in the description and the drawings.

[0014] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0015] The above and other features and advantages of the present invention will become more apparent from a detailed description of exemplary embodiments thereof with reference to the accompanying drawings.

[0016] Figure 1 This is a flowchart of an artificial intelligence-based method for automatic identification and classification of EMI interference sources; Figure 2 This is a flowchart of a reliable method for identifying sources of interference. Figure 3 This is a flowchart illustrating a method for determining that joint identification and classification processing does not require the use of a pre-defined strategy; Figure 4 This is a flowchart illustrating the method for determining a joint identification and classification processing strategy for spaced interference signals. Detailed Implementation

[0017] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0018] Example 1 like Figure 1 As shown, this application provides an artificial intelligence-based method for automatic identification and classification of EMI interference sources, specifically including: S1 uses the identification data of the interference source in the dark room to determine the interval interference signal of the interference source. Based on the monitoring data of the interval interference signal of the interference source, it determines the interference source that is jointly identified and classified in the artificial intelligence model with the interval interference signal of all the interference sources, and uses it as a reliable interference source for analysis. Furthermore, the identification data of the interference source includes the electromagnetic interference signal of the interference source and the timing characteristics of the electromagnetic interference signal.

[0019] Furthermore, the interval interference signal of the disturbance source is an electromagnetic interference signal with an interval distribution in terms of time sequence characteristics, that is, there are interval periods between the time periods in which electromagnetic interference signals exist.

[0020] Specifically, such as Figure 2 As shown, the method for reliably analyzing and determining the source of interference is as follows: The purpose of this embodiment is to screen out those electromagnetic interference sources whose interference behavior is regular and persistent from among numerous sources, and classify them as "reliable analysis interference sources." These types of interference sources are the primary targets for in-depth analysis and targeted suppression measures.

[0021] The core criterion for judgment is the frequency of the intermittent interference from the source. The more frequent the interference, the more stable its behavioral pattern, and the more worthy it is of in-depth analysis.

[0022] S11 Based on the monitoring data of the interval interference signal of the interference source, determine the time period in which the interval interference signal exists in the interference source, and take it as the interference period; In the above steps, the interference period is identified. The interference period is defined as the period during which the interference source emits an intermittent interference signal.

[0023] Intermittent interference signals: Electromagnetic interference signals that are discontinuous in time and have a "transmit-pause-transmit" characteristic. For example, periodic pulse noise.

[0024] S12 determines the dates in the interference source where interference occurs based on the distribution data of the interference periods, and uses these dates as the interference dates; Specifically, interference dates are identified. A date is marked as an interference date if at least one interference period occurs within a given date range.

[0025] S13 determines whether the harassment source is a reliable harassment source based on the composition data of the interference dates in different harassment sources.

[0026] It is understandable that, based on the composition data of the interference dates in different interference sources, determining whether the interference source is a reliable source for analysis specifically includes: The interference impact coefficient is determined based on the proportion of the number of interference days from the aforementioned interference source within the most recent preset time period; Case 1: If the interference influence coefficient of the interference source is greater than the preset influence coefficient threshold, then the interference source is determined as a reliable interference source for analysis. Case 1: If the interference impact coefficient is greater than the preset impact coefficient threshold, then the interference source is determined to be a reliable interference source for analysis.

[0027] Case 2: If the interference influence coefficient of the interference source is not greater than the preset influence coefficient threshold, then the interference source is determined not to be a reliable interference source.

[0028] It should be noted that when performing joint analysis and processing of interval interference signals, a certain interval interference signal is superimposed with the interval interference signal to generate a new electromagnetic interference signal. Based on the new electromagnetic interference signal, it is determined whether the interference source can be identified.

[0029] In one possible embodiment, the target of the analysis is an air conditioning unit in a darkroom area, suspected to be a source of interference causing abnormal data from nearby monitoring stations.

[0030] Preset time period: the last 30 days; preset impact coefficient threshold: 0.5 (i.e., 50%). Interference Period Identification: Monitoring revealed that the frequency converter emits periodic electromagnetic noise, each lasting approximately 10 minutes, followed by a 50-minute pause, and this cycle repeats. Each 10-minute emission period constitutes an interference period.

[0031] Interference Dates: Interference was detected on 20 out of the 30 days, meaning the number of interference dates was 20.

[0032] Calculate the interference impact coefficient: Interference impact coefficient = 20 / 30 ≈ 0.67 (i.e. 67%), interference impact coefficient (0.67) > preset impact coefficient threshold (0.5).

[0033] Conclusion: Based on situation 1, the frequency converter drive device is identified as a "reliable source of interference".

[0034] Specifically, such as Figure 3 As shown, it is determined that a preset strategy is not required for joint identification and classification processing, specifically including: The purpose of this invention is to determine whether the electromagnetic environment of the current anechoic chamber is so complex and variable that a more advanced "preset strategy" (i.e., using an AI model for joint signal analysis) must be activated in order to effectively identify the source of interference.

[0035] The judgment logic revolves around two core issues: when there are many "unknown" and changing sources of disturbance, and their proportion is large, joint analysis and processing are required.

[0036] S21 Determine the percentage of reliable analytical interference sources in the anechoic chamber based on the composition of reliable analytical interference sources in the anechoic chamber; S22 determines the interference source whose temporal characteristics of the interval interference signal change based on the monitoring data of the variation of the interval interference signal of different interference sources, and regards it as the variable interference source. The maintenance team installed multiple fixed and portable electromagnetic field probes in and around the anechoic chamber for 24-hour continuous monitoring. Using a spectrum analyzer and time-domain recording, they identified several potential sources of interference and their spaced interference signals.

[0037] Interference Source A: The "lighting system ballast" inside the darkroom. Its interference signal is synchronized with the light switching time, occurs daily, and is stable. Interference Source B: The "periodic environmental stress screening test chamber" in the adjacent laboratory. Its compressor operates for one hour and then stops for half an hour, generating regular, intermittent interference. Interference Source C: The "fresh air system variable frequency motor" in the building. Its interference signal is intermittent and its frequency fluctuates, classifying it as a variable interference source. Interference Source D: An unknown, very weak periodic pulse signal, difficult to detect and locate.

[0038] S23 determines whether a preset strategy is needed for joint identification and classification processing based on the proportion of reliable analysis interference sources in the darkroom and the data on changing interference sources.

[0039] Interference source A (ballast): Interference occurred on 28 out of 30 days. Interference impact coefficient = 28 / 30 ≈ 0.93. Interference source B (test chamber): Operational and interference occurred on 25 out of 30 days. Interference impact coefficient = 25 / 30 ≈ 0.83. Interference source C (variable frequency motor): Obvious interference was detected on only 8 out of 30 days. Interference impact coefficient = 8 / 30 ≈ 0.27. Interference source D (unknown pulse): The signal was too weak to be directly statistically analyzed.

[0040] Determining "Reliable Analysis Sources of Interference": Since the interference impact coefficients of A (0.93) and B (0.83) are both greater than the threshold (0.5), they are determined to be "Reliable Analysis Sources of Interference". C and D do not belong to this category.

[0041] It is understood that, based on the proportion of reliable analytical interference sources in the anechoic chamber and the data on changing interference sources, it is determined whether a preset strategy is needed for joint identification and classification processing, specifically including: S231 Based on the proportion of reliable analysis interference sources in the anechoic chamber, determine the joint analysis reliability coefficient of the interval interference signals of the interference sources, and determine whether the joint analysis reliability coefficient of the interval interference signals of the interference sources is greater than the preset reliability coefficient threshold. If yes, proceed to the next step; if no, determine that a preset strategy needs to be used for joint identification and classification processing. In the above steps, the joint analysis reliability coefficient is calculated. The proportion of reliable analysis interference sources = 2 / 4 identified in total = 0.5. The judgment is: Reliability coefficient (0.5) < preset reliability coefficient threshold (0.7). Conclusion: The condition is not met. According to logic, it should be determined that "a preset strategy needs to be adopted". At this time, the number of reliable analysis interference sources is relatively small. The joint analysis strategy between interval interference signals is to superimpose the interval interference signals of other interference sources on the interval interference signals whose temporal characteristics change, and then feed them into the artificial intelligence model to determine whether the interference source can be accurately classified and identified.

[0042] S232 removes the interference sources that have been removed from reliable analysis as other interference sources, and determines whether the number of other interference sources is greater than a preset interference source number threshold. If yes, proceed to the next step; if no, determine that a preset strategy needs to be used for joint identification and classification processing. If the preset reliability coefficient threshold is 0.4, then other interference sources = total interference sources (4) - reliable analysis interference sources (2) = 2 (i.e., C and D). Therefore, the number of other interference sources (2) < the preset threshold (3). Conclusion: The condition is not met. Logically, the number of other interference sources is relatively small, and it should be determined that "the preset strategy needs to be adopted".

[0043] S233 Based on the identification result of the changing disturbance source, determine whether there is a changing disturbance source among other disturbance sources. If yes, proceed to the next step; otherwise, determine that no preset strategy is needed for joint identification and classification processing. S234 determines the joint analysis requirement coefficient of the interval interference signal of the interference source based on the proportion of the number of the variable interference source among the other interference sources, and in combination with the proportion of the number of reliable analysis interference sources in the anechoic chamber. Based on the joint analysis requirement coefficient of the interval interference signal, it determines whether a preset strategy is needed for joint identification and classification processing.

[0044] In another embodiment, if the number of other interference sources is 4, the number of other interference sources (4) > the threshold (3). Conclusion: The condition is met, proceed to S233.

[0045] S233 (Check for the presence of a variable source): Among the four other sources of disturbance, source C (variable frequency motor) is clearly identified as a variable source of disturbance. Conclusion: The condition is met (a variable source exists), proceed to S234.

[0046] S234 (Comprehensive Demand Assessment): Calculate the joint analysis demand coefficient for interval interference signals: Factor 1: The proportion of variable interference sources among other interference sources = 1 / 4 = 0.25, Factor 2: In this example, it is 1 - the proportion of reliable analysis interference sources = 1 - 0.33 = 0.67. (Because the lower the proportion of reliable sources, the higher the analysis demand for joint identification and classification processing of interval interference signals), Joint analysis demand coefficient = (0.25 + 0.75) / 2 = 0.5, Judgment: Demand coefficient (0.375) < preset demand coefficient threshold (0.5), Final decision: It is determined that the preset strategy is not required for joint identification and classification processing.

[0047] It is understood that the joint analysis demand coefficient of the interval interference signal is determined based on the proportion of the variable interference source among the other interference sources, and in combination with the proportion of the reliable analysis interference sources in the anechoic chamber. The higher the proportion of the variable interference source among the other interference sources and the lower the proportion of the reliable analysis interference sources in the anechoic chamber, the larger the joint analysis demand coefficient of the interval interference signal.

[0048] Specifically, when the joint analysis demand coefficient of the interval interference signal is greater than the preset demand coefficient threshold, it is determined that a preset strategy needs to be used for joint identification and classification processing.

[0049] It should be noted that when a preset strategy is required for joint identification and classification, for interval interference signals whose temporal characteristics change, the interval interference signals of other interference sources are superimposed and fed into the artificial intelligence model to determine whether the interference source can be accurately classified and identified.

[0050] In one possible embodiment, when a preset strategy is required for joint identification and classification, if the interval influence coefficient of the interference source is greater than 0.7, all electromagnetic interference signals in the interference source are superimposed with the interval interference signal to determine whether reliable identification is possible.

[0051] S2, based on the reliable analysis of the composition of the interference sources in the anechoic chamber and combined with the monitoring data of the variation of the interval interference signals of different interference sources, determines the joint identification and classification processing strategy of the interval interference signals of the interference sources when the temporal characteristics of the interval interference signals of the interference sources change, and when it is not necessary to use a preset strategy for joint identification and classification processing, the joint identification and classification processing strategy of the interval interference signals of the interference sources is determined based on the data of the variation of the temporal characteristics of different interval interference signals. Specifically, such as Figure 4 As shown, the method for determining the joint identification and classification processing strategy for the interval interference signal is as follows: Scenario: A 3-meter semi-anechoic chamber at the same testing center. Background: In previous analysis, the system ultimately determined that "no pre-set strategy is needed for joint identification and classification." Now, we need to develop specific AI analysis strategies for the multiple "difficult-to-identify" interference sources detected in the anechoic chamber.

[0052] Key sources of interference: Source C: A variable interference source, the building's "fresh air system inverter motor". Its interference signal timing characteristics (such as pulse repetition frequency) change frequently, with a high frequency of change. Source D: An unknown weak pulse, the signal is weak and hidden, difficult to detect. Source E: A newly discovered "automatic door control motor", its interference only occurs when people enter or exit, its timing characteristics are somewhat regular, and its frequency of change is moderate. Source F: "Server room cooling pump", its interference signal timing is stable, and its frequency of change is low.

[0053] S31 uses the variation data of the timing characteristics of the interval interference signal to determine the number of times the timing characteristics change, and determines the variation frequency coefficient of the interval interference signal based on the number of times the timing characteristics change. The frequency of change coefficient is the product of the number of times the time-series feature changes and a preset scaling factor. The more times the feature changes, the larger the frequency of change coefficient will be. The preset scaling factor can be 0.1.

[0054] C (Variable Frequency Motor): Frequency of Change Coefficient V = 0.9 (Working mode frequently changes with load), D (Unknown Pulse): V = 0.7 (Although weak, the captured patterns differ), E (Automatic Door): V = 0.5, F (Cooling Pump): V = 0.2, and the Frequency of Change Coefficient of G is 0.

[0055] S32, based on the overlap between the timing characteristics of the interval interference signal and the timing characteristics of other interval interference signals, determines other interval interference signals that have the same timing characteristics as the interval interference signal during the monitoring process, and treats them as consistent interference signals. In the above steps, the consistent interference signal refers to other strong signals that appear completely synchronously with the target interference signal (such as the weak pulse of D) on the time axis (timing characteristics). At this time, the risk of them superimposing with the target interference signal is higher.

[0056] S33 determines the joint identification and classification processing strategy for the interval interference signal based on the interval process data where the timing characteristics of the consistent interference signal and the interval interference signal are consistent, and the variation frequency coefficients of the different interval interference signals.

[0057] Furthermore, the frequency coefficient of the interval interference signal is determined based on the number of times the timing characteristics of the interval interference signal change, wherein the more times the timing characteristics of the interval interference signal change, the greater the frequency coefficient of the interval interference signal.

[0058] It is understood that, based on the time-series data of the consistent interference signal and the interval interference signal, which are consistent with each other, and the frequency of change coefficients of the different interval interference signals, a joint identification and classification processing strategy for the interval interference signal is determined, specifically including: S331 determines whether the frequency coefficient of the interval interference signal is greater than the preset frequency coefficient threshold. If so, it is determined that the interval interference signal is superimposed with the interval interference signals of other interference sources, and is sent to the artificial intelligence model to determine whether it can be accurately classified and the interference source is identified. If not, proceed to the next step. In the above steps, (high variability screening): Determine: The variability frequency coefficient (0.9) of harassment source C > the threshold (0.8). Conclusion: The condition is met. Therefore, it is immediately determined that for harassment source C, signals from other harassment sources must be superimposed and fed into the AI ​​model for classification and identification.

[0059] S332 determines whether the frequency coefficient of the interval interference signal is within the preset frequency coefficient range. If yes, proceed to the next step. If no, only the consistent interference signal and the interval interference signal with a frequency coefficient greater than the preset frequency coefficient threshold are superimposed and sent to the artificial intelligence model to determine whether the interference source can be accurately classified and identified. If no, proceed to the next step. In the above steps (low variability exclusion and medium variability screening): Determination: The coefficients of interference sources D (0.7) and E (0.5) are within the preset frequency coefficient range [0.3, 0.8]. Interference source F (0.2) is not in this range. Conclusion: The condition is met (signals exist within the range), proceed to the next step. Due to its low variability, interference source F is only superimposed on consistent interference signals and interval interference signals with a variability frequency coefficient greater than the preset frequency coefficient threshold.

[0060] S333 uses the interval interference signal with a frequency coefficient within a preset frequency coefficient range as the filtering interference signal, obtains the number of the filtering interference signals, and determines whether the number of the filtering interference signals is less than a preset filtering number threshold. If so, only the consistent interference signal, the interval interference signal with a frequency coefficient greater than the preset frequency coefficient threshold, and the filtering interference signal are superimposed and sent to the artificial intelligence model to determine whether the interference source can be accurately classified and identified. If not, proceed to the next step. In the above steps, interference signals D and E are filtered. The judgment is: the number of interference signals filtered (2) = the preset number of filtering threshold (2). The conclusion is: the condition is not met (only when it is "less than" the threshold is met). At this time, the number of interference signals is large, so the risk of interference signals with inconsistent timing characteristics superimposed is high. Therefore, proceed to the next step.

[0061] S334 determines the classification reliability coefficient of different selected interference signals based on the number of consistent interference signals and interval interference signals with a frequency coefficient greater than the preset frequency coefficient threshold. It then determines whether there are selected interference signals with a classification reliability coefficient less than the preset reliability coefficient threshold. If so, it proceeds to the next step; otherwise, it determines that only consistent interference signals and interval interference signals with a frequency coefficient greater than the preset frequency coefficient threshold are superimposed and fed into the artificial intelligence model to determine whether the interference source can be accurately classified and identified.

[0062] In the above steps, the classification reliability coefficient is calculated: this coefficient is based on the number of consistent interference signals. Assumption: D (unknown pulse): the number of times its changed timing characteristics are consistent with the timing characteristics of other spaced interference signals is 4, and the classification reliability coefficient 4 / 5 = 0.8.

[0063] E (Automatic Door): It was found that the number of times its changed timing characteristics are consistent with the timing characteristics of other interval interference signals is 2. The classification reliability coefficient is 2 / 5 = 0.4. Determine: Is there a screening signal with a classification reliability coefficient < 0.6? E (0.4 < 0.6) meets the condition. Conclusion: The condition is met. Proceed to the next step. At this time, there is an interval interference signal with poor reliability in classification and identification processing.

[0064] S335 determines the comparison deviation coefficient based on the proportion of screening signals whose classification reliability coefficient is less than a preset reliability coefficient threshold. It then determines whether the comparison deviation coefficient is greater than a preset deviation coefficient threshold. If so, for all screening signals, it performs superposition processing on consistent interference signals, interval interference signals with a change frequency coefficient greater than a preset frequency coefficient threshold, and screening interference signals, and sends the results to the artificial intelligence model to determine whether the interference source can be accurately classified and identified. If not, for screening signals whose classification reliability coefficient is less than the preset reliability coefficient threshold, it performs superposition processing on consistent interference signals, interval interference signals with a change frequency coefficient greater than a preset frequency coefficient threshold, and screening interference signals, and sends the results to the artificial intelligence model to determine whether the interference source can be accurately classified and identified.

[0065] For the filtered signals whose classification reliability coefficient is not less than the preset reliability coefficient threshold, the signals are superimposed on the consistent interference signals, the interval interference signals whose change frequency coefficient is greater than the preset frequency coefficient threshold, and the filtered signals whose classification reliability coefficient is less than the preset reliability threshold. The superposition is then fed into the artificial intelligence model to determine whether the interference source can be accurately classified and identified.

[0066] In this step (final strategy formulation - tiered processing): calculate the comparison deviation coefficient: the proportion of the number of screening signals (i.e., E) with a classification reliability coefficient less than the threshold in all screening signals (D and E) = 1 / 2 = 0.5, judgment: assuming the preset deviation coefficient threshold is 0.6, then 0.5 < 0.6; Final decision: Adopt a tiered processing strategy. Strategy A (Targeted Analysis): For the screening signal (i.e., E) whose classification reliability coefficient is less than the preset reliability coefficient threshold, its signal is superimposed with the signals of consistent interference signal, high variability signal (C) and other low reliability signals, and then fed into the AI ​​model. That is, each time the signal is fed into the AI ​​model with one of the signals to determine whether it can be reliably identified.

[0067] Strategy B (Basic Analysis): For the screening signal (i.e., D) whose classification reliability coefficient is not less than the preset reliability coefficient threshold, the superposition result of its signal with the consistent interference signal, the high variability signal (C) and the signal (E) selected according to strategy A is fed into the AI ​​model. That is, each time the signal is fed into the AI ​​model with one of the signals to determine whether it can be reliably identified.

[0068] S3 determines the update processing method of the joint identification and classification processing strategy when the interval interference signal of the interference source changes, based on the joint identification and classification processing strategy of different interval interference signals in the interference source and the distribution data of the interval interference signal of the interference source in the electromagnetic interference signal of the interference source.

[0069] Scene: A 3-meter semi-anechoic chamber at the same testing center.

[0070] Background: A preliminary joint identification strategy has been developed for interference sources C, D, and E. Now, a dynamic update mechanism needs to be established: how should the system automatically adjust (update) the previous analysis strategy when the interference signal characteristics of these interference sources change?

[0071] Focus of interference source: Pay special attention to interference source C (inverter motor of fresh air system) because it is a known, highly variable interference source.

[0072] Core concept: Spacing Influence Coefficient: This measures the proportion of spacing interference signals among all electromagnetic interference signals emitted by the interference source. A high proportion indicates that the device's operating mode is inherently intermittent and susceptible to external scheduling influences.

[0073] Matching coefficient analysis: For a specific interval interference signal from this interference source, the proportion of the number of matching interference signals that are consistent with it in time to the total number of interval interference signal types. A high proportion indicates that this signal is frequently subjected to joint classification and identification processing with other interval interference signals, thus its identification reliability is relatively high.

[0074] Furthermore, the method for determining the update processing method of the joint identification and processing strategy for the harassment source is as follows: Using a joint identification and classification processing strategy for different interval interference signals in the aforementioned interference source, different interval interference signals are identified, and the interval interference signals that are jointly identified and classified are used as matching interference signals. In the steps above, definition 1: interference source, defined as: a physical device or system that generates electromagnetic interference. It is the main emitter of all interference signals; examples include: frequency converters, switching power supplies, and microprocessor units.

[0075] Definition 2: Intermittent Interference Signal. Definition: An electromagnetic interference signal belonging to a certain interference source that is discontinuous in time and exhibits a "transmission-pause-transmission" characteristic. A single interference source can generate multiple interval interference signals of different modes. Example: For the interference source of a "variable frequency motor," its "start pulse," "PWM carrier signal," and "fan start-stop noise" are three different interval interference signals.

[0076] Definition 3: Electromagnetic interference signal. Definition: The general term for all electromagnetic emissions generated by interference sources. It is a higher-level concept than "intermittent interference signal". Interpretation: It includes continuous, transient and intermittent interference signals.

[0077] Definition 4: Matching interference signal. Definition: In the existing joint identification and classification processing strategy, the interval interference signal that has been selected to be superimposed with signals from other interference sources and sent to the AI ​​model for analysis. Interpretation: This is the result of historical decisions and is the input condition for this process.

[0078] Based on the distribution data of the interval interference signal of the interference source in the electromagnetic interference signal of the interference source, the proportion of the interval interference signal of the interference source in all electromagnetic interference signals is determined, and it is used as the interval influence coefficient. Definition 5: Spacing Influence Coefficient. Definition: An indicator that measures the instability and complexity of the operating mode of an interference source. Calculation formula: Spacing Influence Coefficient = Number of types of interference signals at intervals of the interference source / Total number of types of electromagnetic interference signals identified by the interference source. Interpretation: The higher the coefficient, the more fragmented the operating mode of the device is, and the more easily it is affected by external factors.

[0079] Based on the interval influence coefficient of the disturbance source and the matching interference signal of different interval interference signals, an update processing method for the joint identification and processing strategy of the disturbance source when the interval interference signal changes is determined.

[0080] Furthermore, based on the interval influence coefficient of the interference source and the matching interference signals of different interval interference signals, an update processing method for the joint identification and processing strategy of the interference source when the interval interference signal changes is determined, specifically including: If the interval influence coefficient of the interference source is greater than the preset influence coefficient threshold, it is determined that the number of interval interference signals of the interference source is large, which may indicate that the interference source is affected by external factors, causing the interference signal to be in an interval state. Therefore, the update processing method of the joint identification processing strategy when the interval interference signal of the interference source changes is determined as follows: whenever a new interval interference signal appears or the temporal characteristics of the interval interference signal change, the interference source, whether it is an interval interference signal or an electromagnetic interference signal, needs to be superimposed with all the interval interference signals and sent to the artificial intelligence model to determine whether it can be accurately classified and identified as the interference source. If not, proceed to the next step. In the above steps, a preset impact coefficient threshold (used for the interval impact coefficient) is a preset critical value used to determine whether the interval interference behavior of the harassment source is so prevalent that the most aggressive update strategy needs to be triggered. Its function is to compare it with the interval impact coefficient in the first step.

[0081] Analysis of interference source C (variable frequency motor): Calculate the key coefficient: Interval Influence Coefficient: Analysis shows that among the 10 different electromagnetic interference signals emitted by interference source C, 8 are intermittent (e.g., startup surge, speed regulation pulse, shutdown transient, etc.). Interval Influence Coefficient = 8 / 10 = 0.8.

[0082] Judgment: The interval influence coefficient (0.8) > the threshold (0.7). Conclusion: The condition is met! Therefore, the strategy update method for immediately determining the interference source C is as follows: whenever C has a new interval interference signal or the original signal changes, all interference signals of C (regardless of whether they are intervals) need to be superimposed with all other interval interference signals in the anechoic chamber and sent to the AI ​​model for re-analysis. That is, each time, one type of interference signal is superimposed with one other interval interference signal for analysis.

[0083] The analysis matching coefficient of the interval interference signal is determined by the proportion of the number of matching interference signals of the interval interference signal of the interference source to all interval interference signals. It is then determined whether there is an interval interference signal with an analysis matching coefficient greater than a preset analysis matching coefficient threshold. If so, proceed to the next step. If not, it indicates that the temporal characteristics of the interval interference signal of the interference source are not very consistent with those of other interval interference signals. Therefore, the update processing method of the joint identification processing strategy of the interference source when the interval interference signal changes is determined to be that no update processing is required, and the joint identification processing strategy of the interval interference signal is still determined in the original way. The analysis matching coefficient is defined as an index that measures the correlation between a specific interference signal from a given interval and interference signals from other devices. The calculation formula is: Analysis matching coefficient of a specific interval interference signal = Number of matching interference signals for this signal / Total number of interval interference signal types in the system. This coefficient applies to a single signal. A higher coefficient indicates that the signal more frequently appears in conjunction with other devices.

[0084] The preset analysis matching coefficient threshold is defined as a preset critical value used to determine whether the correlation of a single interval interference signal is strong enough. Its function is to compare it with the analysis matching coefficient in the second step to filter out "strong interference signals".

[0085] Interval interference signals with an analysis matching coefficient greater than a preset analysis matching coefficient threshold are considered strong interference signals. It is determined whether the number of strong interference signals of the interference source is greater than a preset strong interference signal number threshold. If so, the update processing method of the joint identification processing strategy when the interval interference signal changes is determined as follows: whenever a new interval interference signal appears or the temporal characteristics of the interval interference signal change, the interference source, whether it is an interval interference signal or an electromagnetic interference signal, needs to be superimposed with all interval interference signals and sent to the artificial intelligence model to determine whether the interference source can be accurately classified and identified. If not, proceed to the next step. Strong interference signals refer to interval interference signals whose analysis matching coefficient is greater than the preset analysis matching coefficient threshold. These signals are key signals that are closely related to other devices and may cause systemic interference.

[0086] A preset threshold for the number of strong interference signals is defined as a preset critical value used to determine whether there are enough key, highly correlated signals in the interference source. Its function is to be compared with the number of strong interference signals in the third step.

[0087] Based on the average value of the interval influence coefficient of the interference source and the analysis matching coefficient of different interval interference signals, the comprehensive influence coefficient of the interference source is determined, and based on the comprehensive influence coefficient, the update processing method of the joint identification and processing strategy of the interference source when the interval interference signal changes is determined.

[0088] The comprehensive influence coefficient is defined as a composite index that integrates the overall instability of the interference source (interval influence coefficient) and its average signal correlation (mean of the analytical matching coefficient). The calculation formula is: Comprehensive influence coefficient = (interval influence coefficient + average of the analytical matching coefficients of all interval interference signals) / 2. This is the ultimate quantification of the potential influence of the interference source.

[0089] , Preset impact coefficient threshold (used for comprehensive impact coefficient), defined as: a preset critical value used to make the final update strategy decision based on the comprehensive impact coefficient in the final step, and compared with the comprehensive impact coefficient in the fourth step.

[0090] It is understandable that the update processing method for determining the joint identification and processing strategy of the interference source when the interval interference signal changes based on the comprehensive influence coefficient specifically includes: When the comprehensive influence coefficient is greater than the preset influence coefficient threshold, the updated processing method of the joint identification and processing strategy for the interference source when the interval interference signal changes is determined as follows: whenever a new interval interference signal appears or the temporal characteristics of the interval interference signal change, the interference source, whether it is an interval interference signal or an electromagnetic interference signal, needs to be superimposed with all the interval interference signals and fed into the artificial intelligence model to determine whether the interference source can be accurately classified and identified. If not, the updated processing method of the joint identification and processing strategy for the interference source when the interval interference signal changes is determined as follows: when the number of electromagnetic interference signals that were not originally interval interference signals but have become interval interference signals is greater than the preset interference signal number threshold, the interference source, whether it is an interval interference signal or an electromagnetic interference signal, needs to be superimposed with all the interval interference signals and fed into the artificial intelligence model to determine whether the interference source can be accurately classified and identified.

[0091] In another possible embodiment, assuming a scenario variant: the interval influence coefficient of harassment source C is 0.6, step one: not satisfied (0.6 < 0.7), proceed to step two. Step 2 (Analysis and Matching Coefficient Check): Determine whether there are interval interference signals with an analysis matching coefficient > 0.6, such as signal C1 (0.9 > 0.6) and signal C2 (0.7 > 0.6). Conclusion: The condition is met (a high matching degree signal exists). At this time, the synchronization degree with other interval interference signals is relatively high, and proceed to step 3.

[0092] Step 3 (Checking the number of strong interference signals): Strong interference signals: C1 and C2. Judgment: Number of strong interference signals (2) = preset threshold (2), conclusion: condition is not met (2 is not greater than 2), proceed to the fourth step, at this time the number of strong interference signals is small.

[0093] Step 4 (Comprehensive Impact Coefficient Assessment): Calculate the comprehensive impact coefficient: Factor 1: Interval impact coefficient = 0.6, Factor 2: Mean of analysis matching coefficient = (0.9 + 0.7 + 0.1 ...) / 10 ≈ 0.4, Comprehensive impact coefficient = (0.6 + 0.4) / 2 = 0.5, Judgment: Comprehensive impact coefficient (0.5) < threshold (0.65), Final decision: Therefore, a conservative update strategy is adopted. The strategy update method for harassment source C is as follows: When the number of electromagnetic interference signals that were not originally interval interference signals becomes interval interference signals exceeds a preset threshold for the number of interference signals (e.g., 3), then all interference signals need to be superimposed with all interval interference signals and fed into the AI ​​model. That is, each time, one type of interference signal is superimposed with one of the other interval interference signals for analysis.

[0094] Example 2 Secondly, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the aforementioned method for automatic identification and classification of EMI interference sources based on artificial intelligence when running the computer program.

[0095] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, devices, and non-volatile computer storage media are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0096] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0097] The above description is merely one or more embodiments of this specification and is not intended to limit this specification. Various modifications and variations can be made to the one or more embodiments of this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of one or more embodiments of this specification should be included within the scope of the claims of this specification.

Claims

1. A method for automatic identification and classification of EMI interference sources based on artificial intelligence, characterized in that, Specifically, it includes: Using the identification data of the interference sources in the dark room, the interval interference signals of the interference sources are determined. Based on the monitoring data of the interval interference signals of the interference sources, the interference sources that are jointly identified and classified in the artificial intelligence model are identified and used as reliable interference sources for analysis. Based on the reliable analysis of the composition of the interference sources in the anechoic chamber, and combined with the monitoring data of the variation of the interval interference signals of different interference sources, it is determined that when the temporal characteristics of the interval interference signals of the interference sources change, and it is not necessary to use a preset strategy for joint identification and classification processing, the joint identification and classification processing strategy of the interval interference signals of the interference sources is determined based on the data of the variation of the temporal characteristics of different interval interference signals. Based on the joint identification and classification processing strategy of different interval interference signals in the interference source, and combined with the distribution data of the interval interference signals of the interference source in the electromagnetic interference signal of the interference source, the update processing method of the joint identification and processing strategy of the interference source when the interval interference signal changes is determined.

2. The method for automatic identification and classification of EMI interference sources based on artificial intelligence as described in claim 1, characterized in that, The identification data of the interference source includes the electromagnetic interference signal of the interference source and the timing characteristics of the electromagnetic interference signal.

3. The method for automatic identification and classification of EMI interference sources based on artificial intelligence as described in claim 1, characterized in that, The interval interference signal of the disturbance source is an electromagnetic interference signal with an interval distribution in terms of time sequence characteristics, that is, there are interval periods between the time periods in which electromagnetic interference signals exist.

4. The method for automatic identification and classification of EMI interference sources based on artificial intelligence as described in claim 1, characterized in that, The method for reliably analyzing and determining the source of interference is as follows: Based on the monitoring data of the interval interference signals of the interference source, the time periods in which the interval interference signals exist in the interference source are determined and are taken as the interference periods; Based on the distribution data of the interference periods, the dates on which the interference periods exist in the interference sources are determined and used as the interference dates; Based on the composition data of the interference dates in different interference sources, it is determined whether the interference source is a reliable source for analysis.

5. The method for automatic identification and classification of EMI interference sources based on artificial intelligence as described in claim 1, characterized in that, When performing joint analysis and processing of interval interference signals, a certain interval interference signal is superimposed with the interval interference signal to generate a new electromagnetic interference signal. Based on the new electromagnetic interference signal, it is determined whether the interference source can be identified.

6. The method for automatic identification and classification of EMI interference sources based on artificial intelligence as described in claim 1, characterized in that, It has been determined that a pre-defined strategy is not required for joint identification and classification processing, specifically including: Based on the composition of reliable analytical interference sources in the anechoic chamber, determine the percentage of reliable analytical interference sources in the anechoic chamber. Based on the monitoring data of the variation of the interval interference signal from different interference sources, the interference sources whose temporal characteristics of the interval interference signal change are identified and identified as the variable interference sources. Based on the proportion of reliable harassment sources in the darkroom and the data on changing harassment sources, it is determined whether a preset strategy is needed for joint identification and classification processing.

7. The method for automatic identification and classification of EMI interference sources based on artificial intelligence as described in claim 6, characterized in that, Based on the proportion of reliable analytical interference sources in the anechoic chamber and the data on changing interference sources, it is determined whether a preset strategy is needed for joint identification and classification processing, specifically including: Based on the proportion of reliable interference sources in the anechoic chamber, the joint analysis reliability coefficient of the interval interference signals of the interference sources is determined. If the joint analysis reliability coefficient of the interval interference signals of the interference sources is not greater than the preset reliability coefficient threshold, then it is determined that a preset strategy needs to be used for joint identification and classification processing.

8. The method for automatic identification and classification of EMI interference sources based on artificial intelligence as described in claim 1, characterized in that, The method for determining the update processing method of the joint identification and processing strategy for the harassment sources is as follows: Using a joint identification and classification processing strategy for different interval interference signals in the aforementioned interference source, different interval interference signals are identified, and the interval interference signals that are jointly identified and classified are used as matching interference signals. Based on the distribution data of the interval interference signal of the interference source in the electromagnetic interference signal of the interference source, the proportion of the interval interference signal of the interference source in all electromagnetic interference signals is determined, and it is used as the interval influence coefficient. Based on the interval influence coefficient of the disturbance source and the matching interference signal of different interval interference signals, an update processing method for the joint identification and processing strategy of the disturbance source when the interval interference signal changes is determined.

9. A computer system, comprising: A memory and processor connected by communication, and a computer program stored in the memory and capable of running on the processor, characterized in that, when the processor runs the computer program, it executes an AI-based automatic identification and classification method for EMI interference sources as described in any one of claims 1-8.

Images