An Electromagnetic Interference Prediction Method Based on Statistical Analysis of Electromagnetic Environment Monitoring Data

Abstract

This invention discloses an electromagnetic interference prediction method based on statistical analysis of electromagnetic environment monitoring data, belonging to the field of equipment electronics technology. The method collects spectral data from various monitoring points in the electromagnetic environment; the spectrum, time, and operating condition information form a spectral sample. Through spectral stability calculation, the monitoring point X with the lowest stability is determined. Based on the spectral sample, statistics reflecting electromagnetic environment fluctuations are calculated from data at different times and the same frequency, generating a fluctuation characteristic spectrum of the statistics as a function of frequency. Based on the correlation between the fluctuation characteristic spectra of monitoring point X and other monitoring points, and combined with the type of equipment at monitoring point X, potential interference source equipment is identified. This invention can predict electromagnetic interference based on statistical analysis of spectral data.

Description

Technical Field

[0001] This invention belongs to the field of equipment electronics technology, and specifically relates to an electromagnetic interference prediction method based on statistical analysis of electromagnetic environment monitoring data. Background Technology

[0002] With the development of science and technology, electronic and electrical equipment and systems are being used more and more widely. Numerous new technologies and products are being applied to various weapon platforms, and high-power, high-voltage, and high-current equipment is frequently used. While the technological level of equipment has significantly improved, electromagnetic interference problems have also increased. How to solve electromagnetic interference and how to improve the efficiency of electromagnetic interference troubleshooting have always been key issues requiring attention. In the past, during the troubleshooting of electromagnetic interference problems, there was often a lack of electromagnetic environment level data at the time of the interference incident, which affected the localization and diagnosis of electromagnetic interference incidents. Therefore, an electromagnetic environment monitoring system is needed. This system would allow operators to monitor the electromagnetic environment situation in real time, and in the event of sudden electromagnetic interference, it would enable systematic analysis and processing of a large amount of electromagnetic environment monitoring data, allowing for retrospective analysis of the monitoring data. This would greatly improve the efficiency of electromagnetic interference troubleshooting and better ensure overall electromagnetic compatibility.

[0003] Currently, electromagnetic environment monitoring systems monitor a wide variety of items, cover a broad range, and operate over long periods, generating a large amount of overall electromagnetic environment monitoring data. However, this makes it difficult to intuitively grasp the overall electromagnetic environment status, such as the trends in power grid quality changes and the fluctuations of large amounts of spectrum data over time. Therefore, how to systematically analyze and process electromagnetic environment monitoring data to intuitively present the comprehensive electromagnetic environment situation, extract overall electromagnetic environment characteristics, and achieve early prediction of electromagnetic interference has become a crucial problem that electromagnetic environment monitoring data analysis must solve. Summary of the Invention

[0004] In view of this, the present invention provides an electromagnetic interference prediction method based on statistical analysis of electromagnetic environment monitoring data, which can predict electromagnetic interference based on statistical analysis of spectrum data.

[0005] To solve the above-mentioned technical problems, the present invention is implemented as follows.

[0006] An electromagnetic interference prediction method based on statistical analysis of electromagnetic environment monitoring data includes:

[0007] Step 1: Set up multiple monitoring points in the electromagnetic environment and acquire electromagnetic environment spectrum data of each monitoring point over a period of time; the electromagnetic environment spectrum data, time and operating condition information constitute a spectrum sample;

[0008] Step 2: Calculate the spectral stability of each monitoring point based on the spectral samples; determine the monitoring point X with the lowest spectral stability, and the corresponding device is x;

[0009] Step 3: Based on the spectrum samples, calculate the statistics reflecting electromagnetic environment fluctuations by analyzing data from different times and the same frequency points, generate the fluctuation characteristic spectrum of the statistics as a function of frequency, and determine the potential interference source devices based on the correlation between the fluctuation characteristic spectra of monitoring point X and other monitoring points, combined with the type of device x.

[0010] Specifically, step 3 involves the following steps: If device x is a high-power device, it indicates that device x is unstable and is a potential source of interference. If device x is a sensitive electronic device, based on the spectrum sample of monitoring point X, a statistical measure reflecting electromagnetic environment fluctuations is calculated from data at the same frequency at different times, generating a fluctuation characteristic spectrum of the statistical measure as a function of frequency. It is then determined whether the fluctuation characteristic spectrum of other monitoring points is positively correlated with that of monitoring point X. If the positively correlated monitoring points correspond to high-power devices, the positively correlated monitoring point devices are identified as potential sources of interference. If the positively correlated monitoring points correspond to sensitive electronic devices, or if there are no positively correlated monitoring points, the process proceeds to step 4.

[0011] Step 4: Based on the spectrum sample of monitoring point X, construct a three-dimensional spectrum diagram showing the spectrum change over time, and search for abrupt spectrum changes on the three-dimensional spectrum diagram; if the frequency range of the abrupt spectrum changes matches the frequency range of the over-threshold fluctuations in the fluctuation characteristic spectrum of monitoring point X, then use the abrupt spectrum changes as the analysis sample, trace the operating condition information before and after the abrupt change, and further locate potential interference source devices based on the operating condition information; if there is no abrupt change or the frequency range does not match, proceed to step 5.

[0012] Step 5: Conduct typical characteristic spectrum analysis based on the spectrum samples of monitoring point X.

[0013] Preferably, the mode and standard deviation are selected as statistical measures to reflect electromagnetic environment fluctuations.

[0014] Preferably, step 2 is: calculating the mode spectral stability and standard deviation spectral stability of each monitoring point based on the spectral samples;

[0015] The method for calculating the mode spectrum stability is as follows: based on spectrum samples over a period of time, the amplitude mode is calculated for data at the same frequency point at different times to generate the mode spectrum; the proportion of frequency points in the mode spectrum where the mode value is within a set mode interval is calculated as the mode spectrum stability.

[0016] The standard deviation spectral stability is calculated as follows: based on spectral samples over a period of time, the amplitude standard deviation is calculated for data at the same frequency point at different times, generating a standard deviation spectrum; the proportion of frequency points in the standard deviation spectrum where the standard deviation value is within a set standard deviation interval is calculated as the standard deviation spectral stability.

[0017] By comprehensively comparing the mode spectral stability and standard deviation spectral stability of each monitoring point, the monitoring point with the lowest spectral stability is determined.

[0018] Preferably, in step 3, determining whether the fluctuation characteristic spectrum of other monitoring points is positively correlated with that of monitoring point X is as follows:

[0019] The frequency bands whose amplitude exceeds the fluctuation threshold in the fluctuation characteristic spectrum are obtained as the fluctuation frequency bands;

[0020] Determine whether the fluctuation characteristic spectrum of other monitoring points also exceeds the fluctuation threshold in the fluctuation frequency band; if so, determine that the fluctuation trend of the corresponding monitoring point is positively correlated with that of monitoring point X.

[0021] Preferably, the fluctuation characteristic spectrum includes fluctuation characteristic spectra of multiple statistics;

[0022] The method for determining whether the fluctuation characteristic spectrum of other monitoring points also exceeds the fluctuation threshold in the fluctuation frequency band is as follows: if the fluctuation characteristic spectrum of any statistic exceeds the fluctuation threshold in any fluctuation frequency band, then the corresponding monitoring point is determined to exceed the fluctuation threshold.

[0023] Preferably, in step 4, the method for determining whether the frequency range of the mutation spectrum matches the frequency range of the over-threshold fluctuation in the fluctuation characteristic spectrum of monitoring point X is as follows:

[0024] Determine whether the frequency range of the sudden change spectrum is within the fluctuation frequency band, or whether the overlap with the fluctuation frequency band exceeds a certain proportion; if so, it is determined that the frequency range is matched.

[0025] Preferably, in step 5, the typical feature spectrum analysis is as follows:

[0026] Typical characteristic spectra include the maximum typical value spectrum and the minimum limit margin spectrum;

[0027] The spectrum with the maximum typical value is determined as follows: Important frequency bands are obtained by dividing the bandwidth according to GJB151; for each spectrum sample, the maximum amplitude of the spectrum sample within the important frequency band is determined. Find all spectral samples The largest sample is used as the spectrum of the largest typical value.

[0028] The minimum limit margin spectrum is determined as follows: For each spectrum sample, determine the closest amplitude within the important frequency band that is closest to the given spectrum limit. Find all spectral samples The smallest sample is used as the spectrum of the minimum limit margin.

[0029] The maximum typical value spectrum and the minimum limit margin spectrum are used as the analysis samples. and minimum Typical characteristics

[0030] If the frequency range of typical features appearing in the analysis sample matches the frequency range of the over-threshold fluctuation, then the operating condition information for a period of time before and after time t2 is traced through the time t2 when the typical features appear; the potential interference source equipment is further located based on the operating condition information; if the frequency ranges do not match, then a typical interference source correlation spectrum analysis is performed on the analysis sample to retrieve the potential interference source equipment.

[0031] Preferably, the step of further locating potential interference source devices based on operating condition information is as follows:

[0032] Based on the operating condition information traced over a period of time, determine whether there has been a change in operating conditions; if so, identify the equipment that caused the change in operating conditions as a potential source of interference; if the operating conditions are stable, conduct a typical interference source correlation spectrum analysis on the analysis sample to search for potential interference source equipment.

[0033] Preferably, the correlation spectrum analysis of the typical interference source is as follows:

[0034] The analyzed samples were placed into a typical interference source spectrum feature database for retrieval and matching, and frequency point correlation analysis was carried out: peaks exceeding a set amplitude threshold were selected from both the analyzed samples and typical interference sources; the frequency points corresponding to the peaks in the analyzed samples were selected as the frequency points to be matched, and the frequency points corresponding to the peaks in the typical interference sources were selected as the interference frequency points; it was determined whether the frequency points to be matched and the interference frequency points matched and their correlation; if they matched, the device corresponding to the analyzed sample was a potential interference source device, and the most relevant typical interference source was the interference type.

[0035] Preferably, the determination of the correlation between the frequency point to be matched and the interference frequency point is as follows:

[0036] If all the frequency points to be matched in the analysis sample have matching interference frequency points within the same typical interference source, then the analysis sample matches the typical interference source and is strongly correlated.

[0037] If some of the frequency points to be matched in the analysis sample have matching interference frequency points within the same typical interference source, then the analysis sample matches the typical interference source, but is weakly correlated.

[0038] If both strong and weak correlations exist, then the typical source of strong correlation is the type of interference from potential interference source devices.

[0039] Beneficial effects:

[0040] (1) This invention collects spectrum samples over a period of time, calculates statistics for amplitude data at the same frequency point at different times, and constructs an array of statistics that change with frequency, called the fluctuation characteristic spectrum. This fluctuation characteristic spectrum reflects the change of spectrum fluctuation over time. Based on the fluctuation characteristic spectrum, the spectrum stability can be determined, and the lowest stability monitoring point is used as the basis for analysis and comparison. At the same time, when collecting spectrum data, operating condition information is also recorded, which facilitates electromagnetic interference analysis and prediction based on the operating condition information. Based on these designs, this invention achieves electromagnetic interference prediction through improvements in fluctuation characteristic spectrum calculation, spectrum stability calculation, and the addition of operating condition information. It predicts the operating conditions and frequency range in which electromagnetic interference may occur, and releases electromagnetic interference risks in advance. It can also quickly find and extract the spectrum characteristics of interference signals when electromagnetic interference occurs, improve the efficiency of electromagnetic interference investigation, and provide data support for overall electromagnetic compatibility assessment.

[0041] (2) The present invention designs an electromagnetic interference prediction process, which proceeds from strong to weak in terms of the certainty of the interference source determination, and combines spectrum stability calculation, spectrum correlation analysis, three-dimensional spectrum analysis, typical feature spectrum analysis, and typical interference source correlation spectrum analysis to achieve electromagnetic interference prediction.

[0042] (3) When determining potential interference sources based on the maximum typical value spectrum, the minimum limit margin spectrum, or the abrupt change spectrum, the changes in operating conditions are further considered, and the equipment that causes the changes in operating conditions is considered a potential interference source. Although the analyzed samples or data are the maximum typical value spectrum, the minimum limit margin spectrum, or the abrupt change spectrum, the reason for these spectral characteristics is that the operating conditions of some equipment have changed. Therefore, this invention adopts a reverse derivation process, extracts time from the characteristic spectrum, obtains the changes in operating conditions through time, and thus finds potential suspected interference sources.

[0043] (4) The data analysis accuracy of this invention is high. Experimental verification shows that the data analysis accuracy is higher than 90%.

[0044] (5) The present invention has strong applicability. It can meet the electromagnetic environment monitoring data analysis in different scenarios such as shielded rooms / anechoic chambers, land-based joint commissioning systems, and cabin environments. It can also analyze electromagnetic compatibility test data, and it does not depend on a hardware platform, making it highly applicable.

[0045] (6) High degree of modularity. This data analysis method divides the statistical analysis of data into several parts, including spectrum stability calculation, spectrum correlation analysis, three-dimensional spectrum analysis, typical feature spectrum analysis, and typical interference source correlation spectrum analysis. It can adopt a modular design, has strong independence, is flexible in portability, and is easy to extend and maintain the algorithm. Attached Figure Description

[0046] Figure 1This is a diagram illustrating the composition of electromagnetic environment monitoring data statistical analysis involved in the embodiments of the present invention.

[0047] Figure 2 This is a flowchart of an electromagnetic interference prediction method based on statistical analysis of electromagnetic environment monitoring data in an embodiment of the present invention. Detailed Implementation

[0048] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0049] The electromagnetic interference prediction method based on statistical analysis of electromagnetic environment monitoring data provided by this invention mainly targets spectrum data, such as power grid conducted interference (low frequency, radio frequency spectrum) and electromagnetic field radiated emission spectrum.

[0050] The electromagnetic interference prediction method provided in this invention mainly includes three aspects: historical retrieval analysis, statistical analysis, and spectral stability calculation. Figure 1 As shown. Historical retrieval analysis refers to retrieving historical electromagnetic environment monitoring data, replaying the data's changes over a period of time, and analyzing typical characteristics and three-dimensional spectra. Statistical calculations mainly include mode and standard deviation calculations. Based on the concepts of standard deviation and mode in statistics, the mode and standard deviation of electromagnetic environment monitoring data samples are statistically analyzed to generate the mode spectrum and standard deviation spectrum of electromagnetic environment monitoring. These can reflect the distribution of the spectrum over time, thus reflecting the state of spectrum fluctuation. Spectrum stability calculation, based on the mode and standard deviation statistics, analyzes the fluctuation of the interference emission spectrum within the specified standard deviation range and the mode spectrum interval, quantitatively reflecting the stability of the interference emission spectrum within that time period.

[0051] The method of this invention can perform statistical analysis on electromagnetic environment monitoring data from multiple monitoring points under the same environment. By utilizing the aforementioned historical retrieval analysis, statistical analysis, and spectral stability calculation, it can intuitively obtain the comprehensive trend of electromagnetic environment monitoring data, monitor electromagnetic environment fluctuations and spectral stability in real time, and thus predict the operating conditions and frequency ranges where electromagnetic interference may occur, releasing electromagnetic interference risks in advance. It can also quickly find and extract the spectral characteristics of interference signals when electromagnetic interference occurs, improving the efficiency of electromagnetic interference investigation and providing data support for overall electromagnetic compatibility assessment.

[0052] First, the three main analytical techniques used in this invention will be described in detail.

[0053] (I) Historical Search Analysis

[0054] Multiple monitoring points are set up in the electromagnetic environment to continuously collect or replay electromagnetic environment spectrum data over a period of time, and the operating condition information at each time point is recorded; the electromagnetic environment spectrum data and operating condition information at a time point constitute a spectrum sample.

[0055] Here, the operating condition information can be the operating parameters of the object under test, the current position of the movable object, the power on / off status, or any other information about the object under test that affects the electromagnetic data.

[0056] Three-dimensional spectrum data (time-frequency-amplitude) can be constructed based on spectrum samples over a period of time. This three-dimensional spectrum data uses the spectrum obtained from an electromagnetic environment monitoring system at each moment to represent the relationship between the electromagnetic interference emission spectrum and time. (Continuous signal spectrum) The mathematical expression is:

[0057]

[0058] Where A is the amplitude. Angular frequency, Let t be the initial phase and t be the time variable.

[0059] Suppose there are K spectral samples, where the frequency axis is the same discrete frequency point. The discrete signal of the k-th spectral sample at time t. The expression is:

[0060]

[0061] In the formula, For spectral amplitude, This refers to the spectral phase.

[0062] Based on three-dimensional spectral data, three-dimensional spectral analysis, typical feature spectral analysis, and correlation spectral analysis of typical interference sources can be performed. The specific details are as follows:

[0063] (1) Three-dimensional spectrum analysis

[0064] Three-dimensional spectrum data can intuitively reflect the changes in spectrum data over system operation time and the amplitude level of interference spectrum. Adding a time dimension allows for the visual display of abrupt spectrum changes in the three-dimensional spectrum data plot. It also enables the identification and extraction of two-dimensional spectrum samples where abrupt changes occur, and the extraction of operating condition information before and after the abrupt change. This allows for the identification of suspicious operating condition changes from an operating condition perspective, thereby predicting potential electromagnetic interference conditions and proactively mitigating electromagnetic interference risks. Furthermore, it allows for the rapid identification and extraction of interference signal spectrum characteristics when electromagnetic interference occurs.

[0065] (2) Typical characteristic spectrum analysis

[0066] Typical characteristic spectrum types mainly include: maximum typical value spectrum and minimum limit margin spectrum.

[0067] ① Maximum typical value spectrum

[0068] Important frequency bands are designated according to the standard bandwidth specified in GJB151.

[0069] For each spectrum sample, determine the maximum amplitude of the spectrum sample within the important frequency band. Find all spectral samples The largest sample is used as the spectrum of the maximum typical value. This is a typical characteristic.

[0070] Based on the occurrence time of typical characteristics and the changes in operating conditions over a period of time before and after, potential interference can be identified. This allows for the rapid identification and extraction of the spectral characteristics of interference signals when electromagnetic interference occurs, as well as the prediction of possible operating conditions for electromagnetic interference, thus mitigating the risk of electromagnetic interference in advance.

[0071] The formula is expressed as: for each spectrum sample Let the important frequency band be bandwidth B, calculate its maximum amplitude within bandwidth B:

[0072]

[0073] Find all spectral samples The largest spectral sample, used as the maximum typical value spectrum.

[0074]

[0075] ② Spectrum of Minimum Limit Margin

[0076] Important frequency bands are designated according to the standard bandwidth specified in GJB151.

[0077] For each spectrum sample, determine the nearest amplitude within the important frequency band that is closest to the given spectral limit. Find all spectral samples The smallest sample is used as the spectrum of the minimum limiting margin. This is a typical characteristic.

[0078] Based on the occurrence time of typical characteristics and the changes in operating conditions over a period of time before and after, potential interference can be identified. This allows for the rapid identification and extraction of the spectral characteristics of interference signals when electromagnetic interference occurs, as well as the prediction of possible operating conditions for electromagnetic interference, thus mitigating the risk of electromagnetic interference in advance.

[0079] The formula is expressed as: for each spectrum sample Let the important frequency band be bandwidth B. Calculate the typical amplitude with the minimum standard limiting margin within bandwidth B:

[0080]

[0081] In the formula, It is the spectral limit function;

[0082] Find all spectral samples The smallest spectral sample is used as the minimum limit margin spectrum.

[0083]

[0084] (3) Correlation spectrum analysis of typical interference sources

[0085] Electromagnetic environment monitoring spectrum data is entered into a typical interference source spectrum feature library (equipment electromagnetic compatibility test spectrum data, including typical harmonic components, characteristic frequency points and phase, etc.) for retrieval and matching, and correlation analysis is carried out to extract the typical interference source spectrum most relevant to the electromagnetic environment monitoring spectrum data. This spectrum is then compared with the system operating conditions. This allows for the rapid location of interference sources or the narrowing of the interference source location range when electromagnetic interference occurs, and also enables the early prediction of electromagnetic interference risks.

[0086] (II) Statistical Analysis

[0087] In this embodiment, the statistical measures selected are the mode and standard deviation.

[0088] For any given time period (randomly selected), the mode and standard deviation of a spectral sample are statistically analyzed to generate the mode spectrum and standard deviation spectrum. This visually presents the dispersion of the spectral data, reflecting its distribution over time and indicating the fluctuation of the spectral data during that period. By comparing the mode and standard deviation spectra of different monitoring points, the correlation of spectral fluctuations at different monitoring points can be observed, i.e., the degree of mutual influence. If a monitoring point device is a sensitive electronic device, and a device with very similar spectral fluctuations is a high-power device, it can be predicted that the high-power device is the source of interference. If a device with very similar spectral fluctuations is also a sensitive electronic device, then both electronic devices may be interfered with by other devices.

[0089] (1) The standard deviation spectrum is generated as follows: based on spectrum samples over a period of time, the amplitude standard deviation is calculated for data at the same frequency point at different times, and the standard deviation spectrum is generated. Specifically, suppose there are K spectrum samples in total, each spectrum sample is a complex spectrum, and the kth spectrum data is represented as: (k=1,2,3…K), frequency axis N represents N frequency points.

[0090] Standard deviation spectrum Each point represents all frequency samples at frequency , where each point represents the frequency . Standard deviation of amplitude at:

[0091]

[0092] In the formula, It is frequency Mean of all sample amplitudes at different times:

[0093]

[0094] (2) The mode spectrum is generated as follows: based on spectrum samples over a period of time, the amplitude mode is calculated for data at the same frequency point at different times, thus generating the mode spectrum. Specifically, the mode spectrum is denoted as... Each point represents all spectral samples at a frequency of Mode of amplitude:

[0095]

[0096] In the formula, Mode represents the mode.

[0097] (III) Calculation of Spectral Stability

[0098] Based on the aforementioned statistics of mode and standard deviation, the proportion of the spectrum within the specified standard deviation range and the mode spectrum interval (the probability of the spectrum occurring within the range) is calculated, thereby obtaining the standard deviation spectrum stability and mode spectrum stability of each monitoring point.

[0099] The standard deviation range and mode spectrum interval vary depending on the type of environment. For example, the standard deviation range for a workshop is less than 6 and the mode spectrum interval is less than 10, while other environments may choose a standard deviation range of less than 3 and a mode spectrum interval of less than 8.

[0100] Based on the above-mentioned main analytical techniques, the flowchart of the electromagnetic interference prediction method based on statistical analysis of electromagnetic environment monitoring data provided in this embodiment is as follows: Figure 2 As shown, it includes the following steps:

[0101] Step 1: Set up multiple monitoring points in the electromagnetic environment, acquire electromagnetic environment spectrum data of each monitoring point over a period of time, and record the operating condition information at each time point; the electromagnetic environment spectrum data and operating condition information at a time point constitute a spectrum sample.

[0102] Assume that a large-scale system test site has four typical electromagnetic environment monitoring points (A, B, C, and D). The equipment at these monitoring points can be categorized into high-power power generation and transformation equipment and sensitive electronic equipment. High-power power generation and transformation equipment includes inverters, generators, electric pumps, and chillers, while sensitive electronic equipment includes monitoring equipment, displays, and communication equipment.

[0103] Before the experiment, the electromagnetic environment monitoring system was connected to four monitoring points A, B, C, and D. At the start of the experiment, the system was activated to collect electromagnetic environment data in real time at each monitoring point. For conducted interference from the power grid (low frequency and radio frequency), the power grid voltage was collected; for radiated electromagnetic emissions, the environmental electromagnetic field signal was collected. During real-time acquisition, the operating status information of relevant equipment at various times was recorded, including start-up, shutdown, movement, and changes in operating parameters. After the system had been running for a period of time, the electromagnetic environment monitoring data from each monitoring point was processed and analyzed, and the spectrum data was calculated. The spectrum data, corresponding time, and operating status information constituted a spectrum sample. Now, either the power grid conducted spectrum or the electromagnetic field radiated emission spectrum is selected for analysis.

[0104] Step 2: Calculate the spectral stability of each monitoring point based on the spectral sample, and determine the monitoring point with the lowest spectral stability and its corresponding equipment.

[0105] In this step, spectral stability is calculated based on statistics reflecting electromagnetic environment fluctuations. In this embodiment, the mode and standard deviation are used as statistics reflecting electromagnetic environment fluctuations.

[0106] Spectral stability was calculated for the spectral data from four monitoring points: A, B, C, and D. Spectral stability includes mode spectral stability and standard deviation spectral stability.

[0107] The method for obtaining mode spectrum stability is as follows: based on spectrum samples over a period of time, the amplitude mode is calculated for data of the same frequency point at different times to generate the mode spectrum; the proportion of the mode value in the mode spectrum within the set mode interval is calculated as the mode spectrum stability.

[0108] The standard deviation spectral stability is obtained as follows: based on spectral samples over a period of time, the amplitude standard deviation is calculated for data at the same frequency point at different times, and a standard deviation spectrum is generated; the proportion of frequency points in the standard deviation spectrum where the standard deviation value is within the set standard deviation interval is calculated as the standard deviation spectral stability.

[0109] By comprehensively comparing the mode spectral stability and standard deviation spectral stability of each monitoring point, the monitoring point with the lowest spectral stability is determined, which is assumed to be monitoring point D. Notably, the mode spectral stability and standard deviation spectral stability often show a consistent trend; a monitoring point with low mode spectral stability also has low standard deviation spectral stability.

[0110] Step 3: If the equipment at monitoring point D is a high-power power generation and transformation equipment, it indicates that the equipment at monitoring point D is operating in a non-continuous and stable manner or in a state of continuous switching of operating conditions. This indicates that the equipment is operating unstable and may be a potential source of interference, or it may pose a significant risk as a source of interference.

[0111] If the equipment at monitoring point D is a sensitive electronic device and cannot be a source of interference, then by judging the correlation of the statistical spectrum with other monitoring points, we can find monitoring points with the same fluctuations as monitoring point D. If the equipment at that monitoring point is a high-power power generation and transformation equipment, it indicates that the operating state is not continuous and stable or in a state of continuous switching, which may be a potential source of interference, or the risk of it being a source of interference is relatively high. Therefore, it is predicted to be a potential source of interference. If the monitoring points with the same fluctuations are all sensitive electronic devices, or if no monitoring points with the same fluctuations can be found, then we perform the non-correlation analysis in step 4.

[0112] This step specifically involves obtaining the standard deviation spectrum and the mode spectrum. You can choose either the standard deviation spectrum or the mode spectrum, or both.

[0113] Taking the selection of standard deviation spectrum as an example, we determine whether the standard deviation spectrum of other monitoring points is positively correlated with that of monitoring point D. The method for determining positive correlation is as follows: the frequency band where the standard deviation spectrum exceeds a set fluctuation threshold th1 is designated as the fluctuation band. This fluctuation band may consist of multiple segments. We then determine whether the standard deviation spectrum of other monitoring points also exceeds the fluctuation threshold th1 within the specified fluctuation band. If so, the standard deviation spectrum of the corresponding monitoring point is determined to be positively correlated with that of monitoring point D, meaning the two devices exhibit the same spectral fluctuation trend.

[0114] In practice, if multiple statistical fluctuation characteristic spectra are selected, such as the standard deviation spectrum and the mode spectrum, for positive correlation judgment, then if the spectral amplitude in one fluctuation band of any one of the standard deviation spectrum and the mode spectrum exceeds the fluctuation threshold th1, it is determined to exceed the fluctuation threshold.

[0115] Assume that the standard deviation spectrum of monitoring points A, B, and C is positively correlated with the standard deviation spectrum of monitoring point D:

[0116] If the positively correlated monitoring point A device is a high-power power generation and transformation equipment, then monitoring point A device can be identified as the potential source of interference that causes the lowest spectral stability of monitoring point D device within the standard deviation fluctuation frequency range.

[0117] If the positively correlated monitoring point A is a sensitive electronic device, or if no positively correlated monitoring point is found, then all other monitoring points are considered to be interfered with, and proceed to step 4 to perform irrelevance analysis.

[0118] Step 4: Irrelevant analysis.

[0119] If the standard deviation spectra of monitoring points A, B, and C are determined to be uncorrelated with the standard deviation spectrum of monitoring point D, then a three-dimensional spectrum analysis is performed on monitoring point D. At this point, a three-dimensional spectrum map of monitoring point D is generated. Abrupt spectrum changes are searched on the three-dimensional spectrum map; the frequency range of these changes is matched with the fluctuation band of the standard deviation spectrum of monitoring point D. If the frequency range matches, the abrupt spectrum change can be extracted from the three-dimensional spectrum map as an analysis sample. Based on the analysis sample, the time of occurrence of the abrupt spectrum change is found, and the operating conditions before and after the abrupt change are traced back to further locate potential interfering devices based on the operating conditions information. If no abrupt change or inconsistent frequency range matching occurs in the three-dimensional spectrum map, step 5 is executed.

[0120] The specific determination method for "frequency range matching" can be as follows: if the frequency range of the sudden change spectrum is within the fluctuation frequency band, or if the overlap with the fluctuation frequency band exceeds a set value, then it is determined that the frequency range is matching.

[0121] The specific implementation of "further locating potential interference devices based on operating condition information" can be as follows: Determine whether the operating conditions have changed based on operating condition information over a period of time. If the operating conditions change frequently or significantly, it indicates that the monitoring point equipment corresponding to the analysis sample (in this step, the abrupt spectrum sample) is very likely to experience electromagnetic interference due to the change in operating conditions. This suggests a necessary connection between the equipment causing the operating condition change and the abrupt spectrum, and the equipment causing the operating condition change may be a potential interference source. If the operating conditions are stable and the changes are minor, it is still impossible to determine the potential interference source. Further typical interference source correlation spectrum analysis can be conducted: the abrupt spectrum sample is used as the analysis sample and placed into the typical interference source spectrum feature database for retrieval and matching to search for potential interference source equipment. See step 5 for the implementation of "typical interference source correlation spectrum analysis".

[0122] Step 5: If, in Step 4, it was determined that the 3D spectrogram does not contain any abrupt frequency changes, or that the abrupt frequency range does not match the standard deviation fluctuation frequency range, then proceed to this step. This step performs typical characteristic spectrum analysis on the spectrum sample of monitoring point D.

[0123] The specific process of typical feature spectrum analysis in this step is as follows:

[0124] Determine the spectrum of maximum typical value: Based on the standard bandwidth specified in GJB151, designate important frequency bands; for each spectrum sample at monitoring point D, determine the maximum amplitude of the spectrum sample within the important frequency band. Find all spectral samples The largest sample is used as the spectrum of the largest typical value.

[0125] The minimum limit margin spectrum is determined as follows: For each spectrum sample at monitoring point D, the closest amplitude of the spectrum sample to the given spectrum limit within the important frequency band is determined. Find all spectral samples The smallest sample is used as the spectrum of the minimum limit margin.

[0126] If the typical characteristics obtained from the analysis [maximum typical value (largest)] ) or the minimum limit margin (the smallest) If the frequency range of the above-mentioned typical feature matches the aforementioned fluctuation frequency band, then the operating condition information for a period of time before and after time t2 is traced based on the occurrence time t2 of the above-mentioned typical feature. Then, the potential interference source device is further located based on the operating condition information. Here, the specific method of "further locating the potential interference source device based on the operating condition information" is the same as step 4.

[0127] If the operating conditions change significantly, it indicates that the equipment causing the change is necessarily related to the spectrum of the maximum typical value and the spectrum of the minimum limit margin, thus identifying the equipment causing the change as a potential source of interference. If the operating conditions are stable, further correlation spectrum analysis of typical interference sources is needed: the spectrum of the maximum typical value and the spectrum of the minimum limit margin are used as analysis samples and matched in the spectrum feature library of typical interference sources to retrieve potential source of interference equipment.

[0128] If the frequency range of typical features does not match the frequency range of the fluctuation band, then conduct a typical interference source correlation spectrum analysis directly on the spectrum of the maximum typical value and the spectrum of the minimum limit margin to search for potential interference source devices.

[0129] This invention involves correlation spectrum analysis of typical interference sources multiple times, and the specific implementation can be as follows:

[0130] The analysis samples are placed into a typical interference source spectrum feature database for retrieval and matching, and frequency point correlation analysis is carried out to find the most relevant typical interference sources, so as to locate the interference source or narrow down the location range of the interference source and predict the electromagnetic interference risk in advance.

[0131] One approach to frequency correlation analysis involves selecting peaks exceeding a set amplitude threshold from both the analysis sample and typical interference sources. The frequency points corresponding to the selected peaks in the analysis sample are called the target frequency points, and the frequency points corresponding to the selected peaks in the typical interference sources are called the interference frequency points. The matching frequency points and their correlation are then determined. If they match, the device corresponding to the analysis sample is considered a potential interference source device, and the most relevant typical interference source is identified as the interference type.

[0132] In a preferred embodiment, the correlation between the frequency point to be matched and the interfering frequency point is determined as follows:

[0133] If all the frequency points to be matched in the analysis sample have matching interference frequency points within the same typical interference source, then the analysis sample matches the typical interference source and is strongly correlated, thus the credibility is high.

[0134] If some of the frequency points to be matched in the analysis sample have matching interference frequency points in the same typical interference source, then the analysis sample matches the typical interference source, but the correlation is weak and the confidence level is low.

[0135] If both strong and weak correlations exist, then the typical source of strong correlation is the type of interference from potential interference source devices.

[0136] If either a strong correlation or a weak correlation exists, the typical interference source matched is the interference type of the potential interference source device.

[0137] When locating potential interference source devices using the above methods, the time corresponding to the relevant samples can also be extracted as the predicted time of interference occurrence.

[0138] This concludes the process of this embodiment.

[0139] In summary, this embodiment performs mode and standard deviation statistics on spectrum samples over a period of time, calculating the proportion of spectra within the specified standard deviation range and the mode spectral deviation interval. This allows for the determination of the standard deviation spectral stability and mode spectral stability at each monitoring point, and provides a clear visual representation of stability differences. Based on these stability differences, the standard deviation spectrum is retrieved, the amplitude disturbance of the standard deviation spectrum is identified, and the correlation of the standard deviation spectra at each monitoring point in the system is compared and analyzed. This predicts and locates the interference frequency range, enabling the rapid identification and extraction of the spectral characteristics of interference signals when electromagnetic interference occurs. It also allows for the prediction of potential electromagnetic interference conditions, thus mitigating the risk of electromagnetic interference in advance.

[0140] By applying this invention, electromagnetic environment monitoring data under different scenarios can be analyzed and processed, the comprehensive trend of electromagnetic environment monitoring data can be obtained intuitively, the electromagnetic environment fluctuation and spectrum stability can be grasped in real time, the efficiency of electromagnetic interference investigation can be improved, and data support can be provided for overall electromagnetic compatibility assessment.

[0141] Based on the disclosure of this invention, those skilled in the art can employ other similar methods to perform specific electromagnetic environment monitoring data analysis, and are not limited to this approach. Figure 1 and 2 As shown in the diagram.

[0142] The specific embodiments described above only illustrate the design principles of the present invention. The shapes and names of the components in this description may differ and are not limited. Therefore, those skilled in the art can modify or make equivalent substitutions to the technical solutions described in the foregoing embodiments; and these modifications and substitutions do not depart from the inventive spirit and technical solutions of the present invention, and should all fall within the protection scope of the present invention.

Claims

1. A method for predicting electromagnetic interference based on statistical analysis of electromagnetic environment monitoring data, characterized in that, include: Step 1: Set up multiple monitoring points in the electromagnetic environment and acquire electromagnetic environment spectrum data of each monitoring point over a period of time; the electromagnetic environment spectrum data, time and operating condition information constitute a spectrum sample; Step 2: Calculate the spectral stability of each monitoring point based on the spectral samples; determine the monitoring point X with the lowest spectral stability, and the corresponding device is x; Step 3: Based on the spectrum sample, calculate the statistics reflecting electromagnetic environment fluctuations by analyzing data from different times and the same frequency points, generate the fluctuation characteristic spectrum of the statistics as a function of frequency, and determine the potential interference source devices based on the correlation between the fluctuation characteristic spectra of monitoring point X and other monitoring points, combined with the type of device x. Step 3 specifically involves: If device x is a high-power device, it indicates that device x is unstable and is a potential source of interference. If device x is a sensitive electronic device, based on the spectrum sample of monitoring point X, a statistical measure reflecting electromagnetic environment fluctuations is calculated by analyzing data from different times and the same frequency points, generating a fluctuation characteristic spectrum of the statistical measure as a function of frequency. It is then determined whether the fluctuation characteristic spectrum of other monitoring points is positively correlated with that of monitoring point X. If the positively correlated monitoring points correspond to high-power devices, the positively correlated monitoring point devices are identified as potential sources of interference. If the positively correlated monitoring points correspond to sensitive electronic devices, or if there are no positively correlated monitoring points, proceed to step 4. Step 4: Based on the spectrum sample of monitoring point X, construct a three-dimensional spectrum diagram showing the spectrum change over time, and search for abrupt spectrum changes on the three-dimensional spectrum diagram; if the frequency range of the abrupt spectrum changes matches the frequency range of the over-threshold fluctuations in the fluctuation characteristic spectrum of monitoring point X, then use the abrupt spectrum changes as the analysis sample, trace the operating condition information before and after the abrupt change, and further locate potential interference source devices based on the operating condition information; if there is no abrupt change or the frequency range does not match, proceed to step 5. Step 5: Conduct typical feature spectrum analysis based on the spectrum samples of monitoring point X; the typical feature spectrum analysis is as follows: Typical characteristic spectra include the maximum typical value spectrum and the minimum limit margin spectrum; The spectrum with the maximum typical value is determined as follows: Important frequency bands are obtained by dividing the bandwidth according to GJB151; for each spectrum sample, the maximum amplitude of the spectrum sample within the important frequency band is determined. Find all spectral samples The largest sample is used as the spectrum of the largest typical value. The minimum limit margin spectrum is determined as follows: For each spectrum sample, determine the closest amplitude within the important frequency band that is closest to the given spectrum limit. Find all spectral samples The smallest sample is used as the spectrum of the minimum limit margin. The maximum typical value spectrum and the minimum limit margin spectrum are used as the analysis samples. and minimum Typical characteristics; If the frequency range of typical features appearing in the analysis sample matches the frequency range of the over-threshold fluctuation, then the operating condition information for a period of time before and after time t2 is traced through the time t2 when the typical features appear; the potential interference source equipment is further located based on the operating condition information; if the frequency ranges do not match, then a typical interference source correlation spectrum analysis is performed on the analysis sample to retrieve the potential interference source equipment.

2. The electromagnetic interference prediction method based on statistical analysis of electromagnetic environment monitoring data as described in claim 1, characterized in that, The mode and standard deviation were chosen as the statistical measures to reflect the fluctuations in the electromagnetic environment.

3. The electromagnetic interference prediction method based on statistical analysis of electromagnetic environment monitoring data as described in claim 2, characterized in that, Step 2 is as follows: The mode spectral stability and standard deviation spectral stability of each monitoring point are calculated based on the spectral samples. The method for calculating the mode spectrum stability is as follows: based on spectrum samples over a period of time, the amplitude mode is calculated for data at the same frequency point at different times to generate the mode spectrum; the proportion of frequency points in the mode spectrum where the mode value is within a set mode interval is calculated as the mode spectrum stability. The standard deviation spectral stability is calculated as follows: based on spectral samples over a period of time, the amplitude standard deviation is calculated for data at the same frequency point at different times, generating a standard deviation spectrum; the proportion of frequency points in the standard deviation spectrum where the standard deviation value is within a set standard deviation interval is calculated as the standard deviation spectral stability. By comprehensively comparing the mode spectral stability and standard deviation spectral stability of each monitoring point, the monitoring point with the lowest spectral stability is determined.

4. The electromagnetic interference prediction method based on statistical analysis of electromagnetic environment monitoring data as described in claim 1, characterized in that, In step 3, determining whether the fluctuation characteristic spectrum of other monitoring points is positively correlated with that of monitoring point X is as follows: The frequency bands whose amplitude exceeds the fluctuation threshold in the fluctuation characteristic spectrum are obtained as the fluctuation frequency bands; Determine whether the fluctuation characteristic spectrum of other monitoring points also exceeds the fluctuation threshold in the fluctuation frequency band; if so, determine that the fluctuation trend of the corresponding monitoring point is positively correlated with that of monitoring point X.

5. The electromagnetic interference prediction method based on statistical analysis of electromagnetic environment monitoring data as described in claim 4, characterized in that, The fluctuation characteristic spectrum includes fluctuation characteristic spectra of various statistical quantities; The determination of whether the fluctuation characteristic spectrum of other monitoring points also exceeds the fluctuation threshold in the fluctuation frequency band is as follows: if the fluctuation characteristic spectrum of any statistic exceeds the fluctuation threshold in any fluctuation frequency band, then the corresponding monitoring point is determined to exceed the fluctuation threshold.

6. The electromagnetic interference prediction method based on statistical analysis of electromagnetic environment monitoring data as described in claim 4, characterized in that, In step 4, the method for determining whether the frequency range of the mutation spectrum matches the frequency range of the over-threshold fluctuation in the fluctuation characteristic spectrum of monitoring point X is as follows: Determine whether the frequency range of the sudden change spectrum is within the fluctuation frequency band, or whether the overlap with the fluctuation frequency band exceeds a set proportion; if so, it is determined that the frequency range is matched.

7. The electromagnetic interference prediction method based on statistical analysis of electromagnetic environment monitoring data as described in claim 1, characterized in that, The device used to further locate potential interference sources based on the operating condition information is: Based on the operating condition information traced over a period of time, determine whether there has been a change in operating conditions; if so, identify the equipment that caused the change in operating conditions as a potential source of interference; if the operating conditions are stable, conduct a correlation spectrum analysis of typical interference sources to search for potential source of interference equipment.

8. The electromagnetic interference prediction method based on statistical analysis of electromagnetic environment monitoring data as described in claim 1, characterized in that, The correlation spectrum analysis of the typical interference sources is as follows: The analysis samples were placed into a typical interference source spectrum feature library for retrieval and matching, and frequency point correlation analysis was carried out: peaks exceeding the set amplitude threshold were screened from the analysis samples and typical interference sources respectively; the frequency points corresponding to the peaks in the analysis samples were selected as the frequency points to be matched, and the frequency points corresponding to the peaks in the typical interference sources were selected as the interference frequency points; it was determined whether the frequency points to be matched and the interference frequency points matched and their correlation. If a match is found, the device corresponding to the analyzed sample is identified as a potential source of interference, and the most relevant typical source of interference is identified as the type of interference.

Images