Abnormity detection positioning method and system for electromagnetic test

Abstract

The invention provides an abnormal detection positioning method and system for an electromagnetic test, and the method comprises the steps: carrying out the electromagnetic interference test of a target device, and obtaining abnormal position information and abnormal test parameters when a detection abnormality occurs; obtaining an equipment abnormal area and a plurality of implicit abnormal elements in the equipment abnormal area according to the abnormal position information and an index, and collecting an infrared image of the equipment abnormal area; according to the plurality of implicit abnormal elements, carrying out integration level analysis to obtain an integration level, and configuring a detection positioning coefficient; and according to the infrared image, identifying a plurality of temperature anomalies of the plurality of implicit abnormal elements, respectively correcting the detection positioning coefficients, and respectively carrying out array anomaly identification according to the anomaly test parameters to obtain a plurality of element anomaly probabilities as anomaly detection positioning results. The technical problem that in the prior art, when electronic equipment is subjected to electromagnetic interference testing, abnormal element detection and positioning are not accurate is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of electromagnetic test, in particular to an abnormality detection positioning method and system for electromagnetic test. BACKGROUND

[0002] In the process of electromagnetic interference test on electronic equipment, the core is to judge whether the electromagnetic interference emitted by the equipment meets the standard. Excessive electromagnetic interference is usually caused by abnormal single element, so it is necessary to accurately locate the abnormal element. However, the existing technical method (such as spectrum analysis method or near-field scanning) has the problem of inaccurate detection and positioning of abnormal elements. SUMMARY

[0003] The present application provides an abnormality detection positioning method and system for electromagnetic test, which solves the technical problem of inaccurate detection and positioning of abnormal elements in the prior art electromagnetic interference test on electronic equipment.

[0004] The technical solution of the present application to solve the above technical problem is as follows:

[0005] In a first aspect, the present application provides an abnormality detection positioning method for electromagnetic test, comprising:

[0006] Performing electromagnetic interference test on a target device, and obtaining abnormal position information and abnormal test parameters when detection abnormality occurs;

[0007] According to the abnormal position information, obtaining a device abnormal area and a plurality of implicit abnormal elements in the device abnormal area by indexing, and collecting an infrared image of the device abnormal area;

[0008] According to the plurality of implicit abnormal elements, performing integration degree analysis to obtain integration degree, and configuring detection positioning coefficient;

[0009] According to the infrared image, identifying a plurality of temperature abnormality degrees of the plurality of implicit abnormal elements, respectively correcting the detection positioning coefficient, and respectively performing array abnormality identification according to the abnormal test parameters to obtain a plurality of element abnormality probabilities as abnormality detection positioning results.

[0010] In a second aspect, the present application provides an abnormality detection positioning system for electromagnetic test, comprising:

[0011] A data acquisition module is configured to perform electromagnetic interference test on a target device, and obtain abnormal position information and abnormal test parameters when detection abnormality occurs;

[0012] An infrared image acquisition module is configured to obtain a device abnormal area and a plurality of implicit abnormal elements in the device abnormal area by indexing according to the abnormal position information, and collect an infrared image of the device abnormal area;

[0013] An integration degree analysis module is configured to perform integration degree analysis according to the plurality of hidden abnormal elements, and obtain an integration degree and a detection positioning coefficient;

[0014] An identification output module is configured to identify a plurality of temperature abnormal degrees of the plurality of hidden abnormal elements according to the infrared image, correct the detection positioning coefficient respectively, perform array abnormality identification according to the abnormal test parameters respectively, and obtain a plurality of element abnormal probabilities as abnormal detection positioning results.

[0015] The present application has the following advantages:

[0016] Firstly, the present application obtains electromagnetic interference test parameters of a plurality of positions of a target device by performing electromagnetic interference test on the target device, and judges whether the test parameters meet the electromagnetic interference test standard of the target device, and obtains abnormal position information and abnormal test parameters when detection abnormality occurs, thereby providing necessary data support for subsequent accurate positioning. Secondly, based on a device area classifier, the obtained abnormal position information is input, and a more specific device abnormal area is output, and then a list of a plurality of hidden abnormal elements in the device abnormal area and an infrared image of the device abnormal area are collected, and potential abnormal elements are identified, thereby providing necessary data support for subsequent accurate positioning of abnormal elements. Thirdly, by obtaining installation coordinates of the plurality of hidden abnormal elements on the target device, the distance between each two closest hidden elements is calculated, and the average of the distances of the plurality of hidden elements is calculated as the integration degree between the elements, and finally the ratio of the reference hidden element distance to the integration degree is calculated as the detection positioning coefficient, thereby improving the detection efficiency under the premise of ensuring the accuracy of detection positioning by dynamically regulating the detection positioning coefficient. Finally, according to the infrared image, the abnormal temperature of the plurality of hidden abnormal elements is obtained, and then the ratio of the average element temperature under normal operation to the abnormal temperature is calculated to obtain the temperature abnormal degree, the detection positioning coefficient is corrected by the plurality of temperature abnormal degrees respectively, and finally the corrected detection positioning coefficient is input into the electromagnetic abnormality identifier to output the abnormal probability of the plurality of elements as the abnormal detection positioning result.

[0017] Through the above technical solutions, the present application accurately positions the abnormal elements, and improves the accuracy of detection positioning of abnormal elements in the process of electromagnetic interference test of electronic devices. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 A flowchart of an abnormal detection positioning method of electromagnetic test provided by the present application is shown in the figure.

[0019] Figure 2 A structure diagram of an abnormal detection positioning system of electromagnetic test provided by the present application is shown in the figure.

[0020] In the drawings, the components represented by the respective reference numerals are as follows:

[0021] The data acquisition module 11, the infrared image acquisition module 12, the integration degree analysis module 13, and the identification output module 14. DETAILED DESCRIPTION

[0022] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person skilled in the art without creative work fall within the scope of protection of the present application.

[0023] In the description of the present application, the terms "first", "second" are used only for the purpose of description, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly specified.

[0024] In the description of the present application, the term "for example" is used to indicate "as an example, illustration or explanation". Any embodiment described as "for example" in the present application is not necessarily interpreted as more preferred or more advantageous than other embodiments. The following description is given in order to enable any person skilled in the art to implement and use the present application. In the following description, details are listed for the purpose of explanation. It should be understood that a person skilled in the art can realize the present application without using these specific details. In other examples, well-known structures and processes will not be described in detail in order to avoid unnecessary details making the description of the present application obscure. Therefore, the present application is not intended to be limited to the shown embodiments, but is consistent with the broadest scope of principles and features disclosed.

[0025] Embodiment one, as shown in the present application, provides an abnormality detection positioning method for electromagnetic test, which comprises: Figure 1

[0026] S10: performing electromagnetic interference test on the target device, and acquiring abnormal position information and abnormal test parameters when detection abnormality occurs;

[0027] During the electromagnetic interference test on the electronic device, electromagnetic interference exceeding the standard is usually caused by abnormality of a single element, so it is necessary to accurately position the abnormal element causing electromagnetic interference.

[0028] ​To solve the above problems, the application obtains electromagnetic interference test parameters of multiple positions of the target device by performing electromagnetic interference test on the target device, and determines whether the test parameters meet the electromagnetic interference test standard of the target device. When an abnormality is detected, abnormal position information and abnormal test parameters are obtained.

[0029] Specifically, step S10 in the method comprises:

[0030] Performing electromagnetic interference test on multiple positions of the target device to obtain test parameters;

[0031] Obtaining the electromagnetic interference test standard of the target device;

[0032] Determining whether the test parameters do not meet the electromagnetic interference test standard, and if so, obtaining abnormal position information and abnormal test parameters.

[0033] In the embodiment of the application, first, electromagnetic interference test is performed on multiple positions of the target device to obtain test parameters, wherein the test parameters include multi-dimensional electromagnetic characteristic parameters such as frequency, electric field intensity, and harmonic distortion rate. Specifically, a mechanical arm or an XYZ guide rail system is used to carry an electromagnetic probe array to perform stereoscopic scanning on the surface of the target device according to a preset path and a preset stepping precision (such as 10 mm interval) to obtain an electromagnetic parameter space point array of the target device and establish a one-to-one mapping relationship between the test parameters and the space point array. For example, a mechanical arm is used to carry an electromagnetic probe array to perform stereoscopic scanning on the target device according to a preset path and a stepping precision of 10 mm interval to obtain a three-dimensional coordinate (+102 mm, +33 mm, -203 mm) and its corresponding test parameters [frequency 85 MHz, electric field intensity 60 dBμV / m, and harmonic distortion rate 3%].

[0034] Secondly, the electromagnetic interference test standard of the target device is obtained. Specifically, the electromagnetic interference test standard (such as CISPR 32, EN 55011, etc.) corresponding to the device characteristic code (such as FCC ID) is automatically associated. For example, the target device is a medical device, and then the standard IEC 60601-1-2 is obtained, which includes important test standards such as electric field intensity limit value 30 dBμV / m.

[0035] Finally, it is determined whether the test parameters do not meet the electromagnetic interference test standard, and if so, abnormal position information and abnormal test parameters are obtained. Specifically, by comparing the test parameters with the electromagnetic interference test standard limit value in real time, when the test parameters exceed the limit value, the system automatically records the abnormal position coordinates (x, y, z) and the abnormal test parameters (such as frequency, electric field intensity, and harmonic distortion rate, etc.) to form a structured abnormal event report. In this way, the approximate direction of the abnormal element and the corresponding abnormal characteristic parameters are determined.

[0036] In summary, the target device is subjected to electromagnetic interference testing, electromagnetic interference test parameters of multiple positions of the target device are obtained, and it is determined whether the test parameters meet the electromagnetic interference test standards of the target device. When detection abnormalities occur, abnormal position information and abnormal test parameters are obtained. In this way, based on the technical route of spatial discrete sampling, standard parameter comparison, and abnormal information acquisition, approximate abnormal position information and abnormal test parameters are obtained, which provides necessary data support for subsequent accurate positioning.

[0037] S20: According to the abnormal position information, index acquisition device abnormal area and multiple implicit abnormal elements in the device abnormal area are obtained, and an infrared image of the device abnormal area is collected;

[0038] When performing accurate positioning of abnormal elements, it is necessary to determine a more accurate abnormal area according to the abnormal position information (i.e., the approximate direction of the abnormality). By analyzing whether the electronic elements in the area are abnormal, the abnormal elements can be accurately positioned. Further, temperature is also an important basis for measuring whether electronic elements are abnormal. When electronic elements have electromagnetic interference abnormalities, there is a high probability of temperature abnormal phenomena such as overheating.

[0039] To solve the above problems, the present application inputs the obtained abnormal position information (i.e., the approximate direction of the abnormality) based on the device area classifier, and outputs a more specific device abnormal area. Then, a list of multiple implicit abnormal elements in the device abnormal area and an infrared image of the device abnormal area are collected.

[0040] Specifically, step S20 in the method comprises:

[0041] The abnormal position coordinates of the abnormal position information are input into the device area classifier to obtain a device abnormal area, wherein the device area classifier includes a mapping relationship between sample abnormal position information and sample device abnormal areas;

[0042] Obtain multiple elements in the device abnormal area as multiple implicit abnormal elements;

[0043] Collect an infrared image in the device abnormal area.

[0044] In the embodiments of the present application, first, the abnormal position coordinates of the abnormal position information are input into a device area classifier to obtain a device abnormal area, wherein the device area classifier includes a mapping relationship between sample abnormal position information and sample device abnormal area. Further, each device abnormal area has a uniquely determined coordinate range (mm), for example, the coordinate range of the power management area is (x: 0-100, y: 0-50, z: 50-100), and the coordinate range of the radio frequency front-end area is (x: 100-200, y: 50-100, z: 100-120). The device abnormal area in which the abnormal position coordinates are located can be obtained by searching the device area classifier to determine in which coordinate range of the device abnormal area the position coordinates of the abnormal position information fall. Illustratively, the coordinates (20, 25, 60) of the abnormal position information are input into the device area classifier, and the output obtained device abnormal area is the power management area. In this way, the device abnormal area in which the abnormal element is located is determined more accurately.

[0045] Secondly, a plurality of element lists in the device abnormal area are obtained as a plurality of implied abnormal elements. Illustratively, the element list in the device abnormal area is extracted by querying a CAD design file or a PCB assembly drawing, for example, the implied abnormal element list is obtained by querying the CAD design file of the power management area, including capacitors and ICs.

[0046] Finally, an infrared image in the device abnormal area is collected. This is because when an electronic element has electromagnetic interference abnormality, the working frequency is probably high, and when the working frequency increases, the resistance, inductance and other elements in the circuit will generate more heat due to the increase of the current change rate, causing the temperature to rise. For example, when a microprocessor is running at an ultra-high frequency, the working frequency is greatly increased, the power consumption is significantly increased, and the chip temperature is rapidly increased. Therefore, by collecting the infrared image in the device abnormal area, the element with electromagnetic interference abnormality can be identified. Further, the infrared image in the device abnormal area can be accurately collected by professional infrared equipment (such as FLIRT series or Testo 8 series handheld thermal imager) cooperating with a standardized collection process, and the specific overheated element can be accurately located by analyzing the infrared image. In this way, reliable thermodynamic data is provided for accurate positioning of the abnormal element.

[0047] In summary, compared with the prior art, the present application is based on the device area classifier, the input obtained abnormal position information (i.e. the approximate position of the abnormality) can output a more specific device abnormal area, and then a plurality of implied abnormal elements and infrared images in the device abnormal area are collected. In this way, necessary data support is provided for subsequent accurate positioning of the abnormal element.

[0048] S30: According to the plurality of implied abnormal elements, integrated degree analysis is performed to obtain an integrated degree and configure a detection positioning coefficient;

[0049] The density between elements largely determines the difficulty of anomaly detection positioning, and a fixed detection positioning coefficient can result in resource waste or low accuracy. Specifically, the more dense the elements are, the more difficult the detection positioning is, and a larger detection positioning coefficient should be configured to improve accuracy; the more sparse the elements are, the lower the difficulty of detection positioning is, and a smaller detection positioning coefficient should be configured to improve efficiency. Therefore, when positioning and detecting abnormal elements, the detection positioning coefficient needs to be dynamically adjusted according to the density between elements to improve accuracy and efficiency.

[0050] To solve the above problems, the present application obtains the installation coordinates of a plurality of implicit abnormal elements on the target device, calculates the distance between each two closest implicit element coordinates, and calculates the average of the distances of the plurality of implicit elements as the integration degree of the elements. Finally, the ratio of the reference implicit element distance to the integration degree is calculated as the detection positioning coefficient.

[0051] Specifically, step S30 in the method comprises:

[0052] Obtaining the installation coordinates of the plurality of implicit abnormal elements on the target device to obtain a plurality of implicit element coordinates;

[0053] Calculating the distance between each two closest implicit element coordinates to obtain a plurality of implicit element distances, and calculating the average to obtain the average implicit element distance as the integration degree;

[0054] Configuring the detection positioning coefficient according to the integration degree.

[0055] In the embodiment of the present application, first, the installation coordinates of the plurality of implicit abnormal elements on the target device are obtained to obtain a plurality of implicit element coordinates. Exemplarily, the coordinate information (mm) of a plurality of implicit abnormal elements (DC-DC converter, PWM controller, high-frequency switching element) in the device abnormal area (power management area) is (20, 25, 60), (30, 33, 73), and (40, 45, 85).

[0056] Secondly, the distance between each two closest implicit element coordinates is calculated to obtain a plurality of implicit element distances, and the average is calculated to obtain the average implicit element distance as the integration degree. The integration degree represents the distance between elements and reflects the density of element distribution. The smaller the integration degree is, the more dense the elements are, and the larger the integration degree is, the more sparse the elements are. The calculation of the element distance can be performed by a spatial topology optimization algorithm, and the KD-Tree data structure is used to optimize the nearest neighbor calculation, and the time complexity is reduced from O(n 2) to O(n log n), realizing the acceleration of the near neighbor search. Exemplarily, the distance between the coordinates of the implied elements (20, 25, 60) and (30, 33, 73) is calculated as 18.25 mm, the distance between (30, 33, 73) and (40, 45, 85) is calculated as 19.70 mm, and the average value is calculated as 18.98 mm, obtaining the average implied element distance as the integration.

[0057] Finally, according to the integration, the ratio of the reference implied element distance to the integration is calculated to obtain the detection positioning coefficient, i.e., the detection positioning coefficient = reference implied element distance / integration, wherein the smaller the integration, the larger the detection positioning coefficient obtained by calculation. Specifically, the smaller the integration represents the greater the density between elements, and the greater the difficulty of detecting and positioning the element, and a larger detection positioning coefficient should be configured to improve the accuracy; the greater the integration represents the smaller the density between elements, and the lower the difficulty of detecting and positioning the element, and a smaller detection positioning coefficient should be configured to improve the efficiency. Exemplarily, the reference implied element distance of a certain industrial equipment is 1 mm, and the calculated integration is 9.1 mm, so the detection positioning coefficient configured therefor is: 1 mm / 9.1 mm = 0.11.

[0058] Specifically, the “configuring a detection positioning coefficient according to the integration” comprises:

[0059] obtaining a reference implied element distance by acquiring a set of all historical implied element distances of a plurality of devices subjected to electromagnetic interference tests in a historical time and screening a minimum value;

[0060] calculating the ratio of the reference implied element distance to the average implied element distance to obtain a detection positioning coefficient.

[0061] First, a reference implied element distance is obtained by acquiring a set of all historical implied element distances of a plurality of devices subjected to electromagnetic interference tests in a historical time and screening a minimum value. Specifically, a database containing historical test data of a plurality of devices is constructed, which records the three-dimensional coordinate information of elements in various electronic devices and the corresponding electromagnetic interference test results. By deeply mining and analyzing the historical data, the minimum distance between all adjacent elements in each type of device is calculated, and the global minimum value is screened as the reference implied element distance, which represents the maximum density level of the element layout in the historical time (for example, the reference implied element distance of a smart phone mainboard can be as low as 0.15 mm, and the reference implied element distance of an industrial control device is about 0.8 mm).

[0062] Secondly, a ratio of the reference implied element distance and the average implied element distance is calculated to obtain a detection positioning coefficient. The smaller the integration degree is, the greater the density between elements is, the greater the detection positioning coefficient should be configured to improve the accuracy; the greater the integration degree is, the smaller the density between elements is, the smaller the detection positioning coefficient should be configured to improve the efficiency. For example, the reference implied element distance of a certain power management area is 1mm, the calculated integration degree is 9.1mm, the ratio of the reference implied element distance (1mm) and the average implied element distance (9.1mm) is 1mm / 9.1mm=0.11, which is taken as the detection positioning coefficient.

[0063] In summary, compared with the prior art, the present application obtains the installation coordinates of multiple implied abnormal elements on the target device, calculates the distance between every two closest implied element coordinates, calculates the average of multiple implied element distances as the integration degree between elements, and finally calculates the ratio of the reference implied element distance and the integration degree as the detection positioning coefficient. The greater the density between elements is, the greater the detection positioning coefficient should be configured to improve the accuracy; the smaller the density between elements is, the smaller the detection positioning coefficient should be configured to improve the efficiency. In this way, the detection positioning coefficient is dynamically regulated based on the integration degree, and the detection efficiency is improved under the premise of ensuring the detection positioning accuracy.

[0064] S40: According to the infrared image, multiple temperature abnormal degrees of the multiple implied abnormal elements are identified, the detection positioning coefficients are respectively corrected, and array abnormality recognition is performed according to the abnormal test parameters to obtain multiple element abnormality probabilities as abnormal detection positioning results.

[0065] Temperature is also an important basis for measuring whether an electronic element is abnormal. When an electronic element has electromagnetic interference abnormality, it is likely to have temperature abnormality such as overheating. Therefore, in order to improve the accuracy of positioning detection, the detection positioning coefficient needs to be further corrected based on the temperature abnormality, and the abnormal element is identified and positioned based on the corrected detection positioning coefficient.

[0066] To solve the above problems, the present application obtains the abnormal temperature of multiple implied abnormal elements according to the infrared image, then calculates the ratio of the average element temperature under normal operation and the abnormal temperature to obtain the temperature abnormal degree, corrects the detection positioning coefficient by multiple temperature abnormal degrees, and finally inputs the corrected detection positioning coefficient into the electromagnetic abnormality recognizer to output the abnormal probability of multiple elements as the abnormal detection positioning result.

[0067] Specifically, step S40 in the method comprises:

[0068] cropping the infrared image according to the plurality of implied element coordinates of the plurality of implied abnormal elements to obtain a plurality of element infrared images;

[0069] extracting a plurality of element temperatures of the plurality of element infrared images;

[0070] obtaining a plurality of average element temperatures under normal operation of the plurality of implied abnormal elements;

[0071] calculating a ratio of the plurality of element temperatures and the plurality of average element temperatures to obtain a plurality of temperature abnormality degrees.

[0072] In the embodiments of the present application, first, the infrared image is cropped according to the plurality of implied element coordinates of the plurality of implied abnormal elements to obtain a plurality of element infrared images. Specifically, the corresponding region is accurately cropped from the infrared image according to the three-dimensional coordinates of the abnormal element, for example, the region with the element coordinate point as the center ±5mm is cropped to obtain the infrared image of the element.

[0073] Secondly, a plurality of element temperatures of the plurality of element infrared images are extracted. Specifically, the maximum pixel temperature of the element infrared image is extracted as the temperature of the element, wherein the maximum pixel temperature reflects the hottest point of the element. For example, the maximum pixel temperature of the infrared image of a certain element is 50°C, which is taken as the temperature of the element.

[0074] Thirdly, a plurality of average element temperatures under normal operation of the plurality of implied abnormal elements are obtained. Specifically, a plurality of average element temperatures under normal operation of the element are obtained based on a historical database. For example, the average temperature of a certain element under normal working conditions is extracted from the historical database as 40°C.

[0075] Finally, the ratio of the plurality of element temperatures and the plurality of average element temperatures is calculated to obtain a plurality of temperature abnormality degrees, that is, temperature abnormality degree=abnormal element actual temperature / average temperature of abnormal element under normal working conditions. Wherein, the greater the temperature abnormality degree, the greater the probability of abnormal temperature of the element, and the greater the probability of electromagnetic interference. For example, the actual temperature of a certain element is 50°C, and the average temperature under normal working conditions in the historical data is 40°C, so the temperature abnormality degree=50°C / 40°C=1.25. In this way, the probability of electromagnetic interference of the element is reflected by the quantitative data, which provides necessary data support for subsequent correction of detection positioning coefficients.

[0076] Further, the "respectively correcting the detection positioning coefficients, and respectively performing array abnormality identification according to the abnormal test parameters to obtain a plurality of element abnormality probabilities as abnormal detection positioning results" includes:

[0077] According to the plurality of temperature anomaly degrees, respectively, the detection positioning coefficient is corrected and calculated to obtain a plurality of corrected detection positioning coefficients;

[0078] The plurality of element electromagnetic anomaly recognizers corresponding to the plurality of hidden abnormal elements are called, wherein each element electromagnetic anomaly recognizer includes T element electromagnetic anomaly recognition branches, and T is a positive integer.

[0079] According to the plurality of corrected detection positioning coefficients and the branch number T, a plurality of branch calling numbers are calculated and obtained.

[0080] According to the plurality of branch calling numbers, a plurality of element electromagnetic anomaly recognition branch combinations are randomly called, the abnormal test parameters are respectively input, and a plurality of element probability sets are recognized and output to obtain a plurality of element anomaly probabilities as abnormal detection positioning results.

[0081] In the embodiment of the application, first, according to the plurality of temperature anomaly degrees, respectively, the detection positioning coefficient is corrected and calculated to obtain a plurality of corrected detection positioning coefficients, wherein the correction calculation is performed by temperature anomaly degree * detection positioning coefficient. Specifically, the greater the temperature anomaly degree, the greater the probability of electromagnetic interference of the element, and the greater the corrected detection positioning coefficient. The greater detection positioning coefficient can improve the accuracy of positioning detection. For example, the detection positioning coefficient of a certain element is 0.11, and the temperature anomaly degree is 1.25. Then the corrected detection positioning coefficient = 0.11*1.25 = 0.1375. In this way, the initial detection positioning coefficient is increased by the temperature anomaly degree, and the accuracy of positioning detection is improved.

[0082] Secondly, the plurality of element electromagnetic anomaly recognizers corresponding to the plurality of hidden abnormal elements are called, wherein each element electromagnetic anomaly recognizer includes T element electromagnetic anomaly recognition branches, and T is a positive integer. Specifically, the electromagnetic anomaly recognizer includes a plurality of electromagnetic anomaly recognition branches, each branch inputs an abnormal test parameter and outputs an abnormal probability. Further, the electromagnetic anomaly recognizer adopts a heterogeneous integrated learning architecture, including four functional branches of time domain analysis (1D-CNN+LSTM), frequency domain analysis (2D-CNN), modulation analysis (ResNet-18) and spatial correlation (GNN). First, the historical EMI test data is used to extract multi-dimensional electromagnetic feature information such as frequency, electric field intensity and harmonic distortion rate, and the corresponding fault elements are labeled to form a training set. Then, a multi-task learning framework is used to train a shared feature encoding layer and an independent branch network, and each branch focuses on the abnormal pattern recognition of a specific feature mode. In actual application, the system dynamically determines the number of branches to be called according to the correction coefficient (0-1 range) calculated by the temperature anomaly degree and the element integration degree, and finally realizes fault positioning by weighted fusion of the probabilities output by each branch.

[0083] Again, a plurality of branch call numbers are calculated according to a plurality of correction detection positioning coefficients and the branch number T respectively, wherein the branch call number is obtained by rounding up the correction detection positioning coefficient * the total branch number T. For example, the correction detection positioning coefficient is 0.1375, the total branch number T is 15, and the branch call number = [0.1375 * 15] = 3, wherein [] is the rounding up calculation.

[0084] Finally, a plurality of element probability sets are obtained by respectively inputting the abnormal test parameters into a plurality of element electromagnetic anomaly recognizer branch combinations called randomly according to the plurality of branch call numbers, and calculating the mean value of the element anomaly probability as the abnormal detection positioning result. For example, the branch call number is 3, the total branch number T is 15, 3 electromagnetic anomaly recognition branches are randomly called from 15 branches, and the abnormal test parameters (such as the frequency point exceeding the standard and the field strength value) are input. Each branch outputs an abnormal probability of [0.72, 0.85, 0.91], the mean value of the abnormal probability is 0.83, and the abnormal probability of the element is obtained as the abnormal detection positioning result.

[0085] Further, the training step of the plurality of element electromagnetic anomaly recognizers comprises:

[0086] In the electromagnetic interference test data of the same device in the historical time, a sample abnormal test parameter set is collected, and the proportion of the occurrence of the anomaly of a plurality of elements under different sample abnormal test parameters is collected, and a plurality of sample element anomaly probability sets are labeled and obtained;

[0087] The sample abnormal test parameter set and the plurality of sample element anomaly probability sets are divided to obtain a plurality of element electromagnetic anomaly recognition training data sets, wherein each element electromagnetic anomaly recognition training data set includes T element electromagnetic anomaly recognition training data.

[0088] A plurality of element electromagnetic anomaly recognizers are trained using the plurality of element electromagnetic anomaly recognition training data sets respectively, wherein each element electromagnetic anomaly recognizer includes T element electromagnetic anomaly recognition branches.

[0089] In the embodiments of the present application, first, in the electromagnetic interference test data of the same device in the historical time, a sample abnormal test parameter set is collected, and the proportion of the abnormality of multiple elements under different sample abnormal test parameters is collected, and a plurality of sample element abnormal probability sets are obtained by labeling. Specifically, the correspondence between the test parameters in the historical abnormal event and the abnormal probability of multiple elements is labeled, and a data set with probability labels is constructed. For example, electromagnetic interference test data of the same type of device (such as the same type of data base station) is selected from the historical EMC test database, a sample abnormal test parameter set (exceeding frequency 150 kHz, exceeding amplitude 45 dBμV / m) is collected, and the proportion of the abnormality of multiple appearing elements (power amplifier U1, filter capacitor C102) under abnormal test parameters (such as 38 / 50, 12 / 50) is collected, and the abnormal probability of multiple elements under this abnormal test parameter is calculated to be 0.76, 0.24, and the abnormal probability of multiple elements under abnormal test parameters is labeled manually.

[0090] Secondly, the sample abnormal test parameter set and the plurality of sample element abnormal probability sets are divided to obtain a plurality of groups of element electromagnetic abnormality recognition training data sets, wherein each group of element electromagnetic abnormality recognition training data set includes T element electromagnetic abnormality recognition training data. Further, when the data set is divided, the data set can be simply divided into T parts according to the number, or the data set can be grouped according to the feature type of the abnormal test parameter, that is, the data set is divided into T groups according to the type of the abnormal test parameter (such as time domain, frequency domain, modulation domain, etc.), and each group corresponds to the training data of one recognition branch. For example, the abnormal test parameter is divided into: time domain group: containing time domain parameters such as pulse rise time and peak-to-peak value, frequency domain group: containing spectral parameters such as fundamental and harmonic amplitude, frequency deviation, etc. Different branches realize processing of different types of data, and different electromagnetic abnormality recognition branches are trained and optimized through different types of abnormal test parameter sets, which can realize independent recognition and output of different branches for different data types. For example, a recognizer including 5 branches is trained, the time domain feature branch focuses on pulse noise features, and the frequency domain branch learns harmonic distortion features. In this way, the accuracy of abnormal probability output can be improved.

[0091] Finally, the plurality of element electromagnetic anomaly recognition training data sets are used to train a plurality of element electromagnetic anomaly recognizers, each of which includes T element electromagnetic anomaly recognition branches. Specifically, the plurality of element electromagnetic anomaly recognition training data sets are divided into a training set, a validation set, and a test set according to a ratio of 7.5:1.5:1.5. A course learning strategy is used to train and optimize the T electromagnetic anomaly recognition branches of the electromagnetic anomaly recognizer. The model is trained using the training set, and is fine-tuned by expanding difficult samples using a generative adversarial network (GAN). Bayesian hyperparameter search is introduced in the model optimization stage, and the model finally achieves an accuracy of more than 90% on the test set, which is considered to be converged.

[0092] In summary, compared with the prior art, the present application obtains the abnormal temperatures of a plurality of hidden abnormal elements according to the infrared image, then calculates the ratio of the average element temperature under normal operation to the abnormal temperature to obtain the temperature abnormality degree, corrects the detection positioning coefficient through a plurality of temperature abnormality degrees, and finally inputs the corrected detection positioning coefficient into the electromagnetic anomaly recognizer to output the abnormal probability of a plurality of elements as the abnormal detection positioning result. In this way, the accuracy of the abnormal detection positioning result is improved.

[0093] In summary, the embodiments of the present application have at least the following technical effects:

[0094] Compared with the prior art, the present application first performs electromagnetic interference testing on the target device to obtain electromagnetic interference testing parameters at a plurality of positions of the target device, and determines whether the testing parameters meet the electromagnetic interference testing standards of the target device. When a detection anomaly occurs, abnormal position information and abnormal testing parameters are obtained. In this way, based on the technical route of spatial discrete sampling, standard parameter comparison, and abnormal information acquisition, approximate abnormal position information and abnormal testing parameters are obtained, which provide necessary data support for subsequent accurate positioning.

[0095] Secondly, based on the device area classifier, the obtained abnormal position information is input to output a more specific device abnormal area, and then a plurality of hidden abnormal element lists in the device abnormal area and an infrared image of the device abnormal area are collected. In this way, potential abnormal elements are identified, and necessary data support is provided for subsequent accurate positioning of abnormal elements.

[0096] Again, by acquiring the installation coordinates of multiple hidden abnormal elements on the target device, the distance between each two closest hidden element coordinates is calculated, and the average of the distances of multiple hidden elements is calculated as the integration between elements, and finally the ratio of the reference hidden element distance to the integration is calculated as the detection positioning coefficient, wherein the greater the density between elements, the more difficult it is to detect and position the element, and a larger detection positioning coefficient should be configured to improve accuracy; the smaller the density between elements, the lower the difficulty of detecting and positioning the element, and a smaller detection positioning coefficient should be configured to improve efficiency. In this way, by dynamically regulating the detection positioning coefficient, the detection efficiency is improved under the premise of ensuring the accuracy of detection and positioning.

[0097] Finally, according to the infrared image, the abnormal temperature of multiple hidden abnormal elements is acquired, and then the ratio of the average element temperature under normal operation to the abnormal temperature of the abnormal element is calculated to obtain the temperature abnormality degree. The detection positioning coefficient is corrected by multiple temperature abnormality degrees, and finally the corrected detection positioning coefficient is input into the electromagnetic anomaly recognizer to output the abnormal probability of multiple elements as the abnormal detection and positioning result. In this way, the accurate coordinate information of the abnormal element is obtained, and the accuracy of the abnormal element detection and positioning result is improved.

[0098] Through the above technical solutions, the present application acquires the abnormal detection and positioning result of the target device, obtains the detection positioning coefficient through the original element integration, corrects the detection positioning coefficient through the temperature abnormality degree, and finally inputs the corrected detection positioning coefficient into the electromagnetic anomaly recognizer to output the abnormal probability of multiple elements as the abnormal detection and positioning result. In this way, the abnormal element is accurately positioned, and the accuracy of the abnormal element detection and positioning in the electromagnetic interference test process of the electronic device is improved.

[0099] Embodiment two, as shown in Figure 2 Based on the same inventive concept of the electromagnetic test abnormal detection and positioning method provided in embodiment one, the present application embodiment further provides an electromagnetic test abnormal detection and positioning system, comprising:

[0100] The data acquisition module 11 is used for electromagnetic interference test of the target device, and when an abnormality is detected, abnormal position information and abnormal test parameters are acquired;

[0101] The infrared image acquisition module 12 is used for indexing the device abnormal area and multiple hidden abnormal elements in the device abnormal area according to the abnormal position information, and acquiring the infrared image of the device abnormal area;

[0102] The integration analysis module 13 is used for integration analysis according to the multiple hidden abnormal elements to obtain the integration and configure the detection positioning coefficient;

[0103] The identification output module 14 is configured to identify temperature abnormality degrees of the plurality of hidden abnormal elements according to the infrared image, correct the detection positioning coefficients respectively, and perform array abnormality identification according to the abnormality test parameters respectively to obtain element abnormality probabilities as abnormality detection positioning results.

[0104] The data collection module 11 is configured to:

[0105] perform electromagnetic interference tests on a plurality of positions of the target device to obtain test parameters;

[0106] obtain electromagnetic interference test standards of the target device;

[0107] determine whether the test parameters do not meet the electromagnetic interference test standards, and if so, obtain abnormal position information and abnormal test parameters.

[0108] The infrared image collection module 12 is configured to:

[0109] input abnormal position coordinates of the abnormal position information into a device area classifier to obtain a device abnormal area, wherein the device area classifier includes a mapping relationship between sample abnormal position information and sample device abnormal areas;

[0110] obtain a plurality of elements in the device abnormal area as a plurality of hidden abnormal elements;

[0111] collect infrared images in the device abnormal area.

[0112] The integration degree analysis module 13 is configured to:

[0113] obtain installation coordinates of the plurality of hidden abnormal elements on the target device to obtain a plurality of hidden element coordinates;

[0114] calculate distances between every two closest hidden element coordinates to obtain a plurality of hidden element distances, and calculate a mean value to obtain an average hidden element distance as an integration degree;

[0115] configure a detection positioning coefficient according to the integration degree.

[0116] Further, the “configuring a detection positioning coefficient according to the integration degree” includes:

[0117] obtain a reference hidden element distance by collecting all historical hidden element distance sets of electromagnetic interference tests on a plurality of devices in a historical time and screening a minimum value;

[0118] calculate a ratio of the reference hidden element distance to the average hidden element distance to obtain a detection positioning coefficient.

[0119] The identification output module 14 is specifically configured to:

[0120] The infrared image is cropped according to the plurality of element coordinates of the plurality of implied abnormal elements to obtain a plurality of element infrared images;

[0121] The plurality of element temperatures of the plurality of element infrared images are extracted;

[0122] A plurality of average element temperatures under normal operation of the plurality of implied abnormal elements are obtained;

[0123] The ratio of the plurality of element temperatures to the plurality of average element temperatures is calculated to obtain a plurality of temperature abnormality degrees.

[0124] Further, the "respectively correcting the detection positioning coefficients, and respectively performing array anomaly identification according to the abnormal test parameters to obtain a plurality of element abnormality probabilities as the abnormal detection positioning results" includes:

[0125] The detection positioning coefficients are respectively corrected and calculated according to the plurality of temperature abnormality degrees to obtain a plurality of corrected detection positioning coefficients;

[0126] A plurality of element electromagnetic anomaly identifiers corresponding to the plurality of implied abnormal elements are called, and each element electromagnetic anomaly identifier includes T element electromagnetic anomaly identification branches, where T is a positive integer;

[0127] A plurality of branch calling numbers are calculated according to the plurality of corrected detection positioning coefficients and the branch number T;

[0128] A plurality of element probability sets are respectively identified and output according to the plurality of branch calling numbers, and a plurality of element abnormality probabilities are calculated as the abnormal detection positioning results.

[0129] Further, the training steps of the plurality of element electromagnetic anomaly identifiers include:

[0130] A sample abnormal test parameter set is collected from electromagnetic interference test data of the same device in a historical time, and a proportion of abnormalities of a plurality of elements under different sample abnormal test parameters is collected to label and obtain a plurality of sample element abnormality probability sets;

[0131] The sample abnormal test parameter set and the plurality of sample element abnormality probability sets are divided to obtain a plurality of groups of element electromagnetic anomaly identification training data sets, and each group of element electromagnetic anomaly identification training data set includes T element electromagnetic anomaly identification training data;

[0132] The plurality of element electromagnetic anomaly recognizers are obtained by training using the plurality of element electromagnetic anomaly recognition training data sets respectively, wherein each element electromagnetic anomaly recognizer comprises T element electromagnetic anomaly recognition branches.

[0133] In summary, the embodiments of the present application have at least the following technical effects:

[0134] The data acquisition module obtains electromagnetic interference test parameters of the target device at multiple positions by performing electromagnetic interference tests on the target device, and determines whether the test parameters meet the electromagnetic interference test standards of the target device. When a detection anomaly occurs, the abnormal position information and abnormal test parameters are obtained. In this way, necessary data support is provided for subsequent accurate positioning. The infrared image acquisition module inputs the obtained abnormal position information based on the device area classifier, and outputs a more specific device abnormal area. Then, a list of multiple hidden abnormal elements in the device abnormal area and an infrared image of the device abnormal area are acquired. In this way, potential abnormal elements are identified, and necessary data support is provided for subsequent accurate positioning of abnormal elements. The integration analysis module obtains the installation coordinates of the multiple hidden abnormal elements on the target device, calculates the distance between the coordinates of the two closest hidden elements, and calculates the average distance of the multiple hidden elements as the integration degree between the elements. Finally, the ratio of the reference hidden element distance to the integration degree is calculated as the detection positioning coefficient. In this way, by dynamically adjusting the detection positioning coefficient, the detection efficiency is improved while ensuring the accuracy of the detection positioning. The identification output module obtains the abnormal temperature of the multiple hidden abnormal elements based on the infrared image, and then calculates the ratio of the average element temperature under normal operation to the abnormal temperature to obtain the temperature abnormality degree. The detection positioning coefficient is corrected by the multiple temperature abnormality degrees, and finally the corrected detection positioning coefficient is input into the electromagnetic anomaly recognizer to output the abnormal probability of the multiple elements as the abnormal detection positioning result. In this way, the accurate coordinate information of the abnormal elements is obtained, and the accuracy of the abnormal element detection positioning result is improved.

[0135] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0136] Those skilled in the art will understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program code.

[0137] The present application is described in reference to the flowchart illustrations and / or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart Figure 1 one or more functions specified in the flowchart Figure 1 one or more functions specified in the flowchart

[0138] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart Figure 1 one or more functions specified in the flowchart Figure 1 one or more functions specified in the flowchart

[0139] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart Figure 1 one or more functions specified in the flowchart Figure 1 one or more functions specified in the flowchart

[0140] Although preferred embodiments of the application have been described, additional modifications and changes can occur to those skilled in the art once they gain an understanding of the basic inventive concepts.

[0141] It will be apparent to those skilled in the art that various modifications and variations can be made in the present application without departing from the spirit or scope of the application. Thus, it is intended that the present application cover the modifications and variations of this application and its equivalent.

Claims

1. A method of anomaly detection localization for electromagnetic testing, characterized in that, The method comprises: Performing electromagnetic interference test on a target device, obtaining abnormal position information and abnormal test parameters when detection abnormalities occur; According to the abnormal position information, index acquisition device abnormal area and a plurality of implicit abnormal elements in the device abnormal area are obtained, and an infrared image of the device abnormal area is collected; According to the plurality of implicit abnormal elements, integration degree analysis is performed to obtain integration degree, and detection positioning coefficient is configured; According to the infrared image, a plurality of temperature abnormal degrees of the plurality of implicit abnormal elements are identified, the detection positioning coefficient is corrected respectively, and array abnormality recognition is performed according to the abnormal test parameters respectively, a plurality of element abnormal probabilities are obtained as abnormal detection positioning results.

2. The method of claim 1, wherein, Performing electromagnetic interference test on a target device, obtaining electromagnetic interference test parameter sequence, obtaining abnormal position information and abnormal test parameters when detection abnormalities occur, comprising: Performing electromagnetic interference test on a plurality of positions of the target device to obtain test parameters; Obtaining electromagnetic interference test standard of the target device; Determine whether the test parameters do not meet the electromagnetic interference test standard, if so, obtain abnormal position information and abnormal test parameters.

3. The method of claim 1, wherein, According to the abnormal position information, index acquisition device abnormal area and a plurality of implicit abnormal elements in the device abnormal area are obtained, and an infrared image of the device abnormal area is collected, comprising: Input the abnormal position coordinates of the abnormal position information into the device area classifier to obtain the device abnormal area, wherein the device area classifier includes the mapping relationship between the sample abnormal position information and the sample device abnormal area; Obtain a plurality of elements in the device abnormal area as a plurality of implicit abnormal elements; Collect infrared image in the device abnormal area.

4. The method of claim 1, wherein, According to the plurality of implicit abnormal elements, integration degree analysis is performed to obtain integration degree, and detection positioning coefficient is configured, comprising: Obtain the installation coordinates of the plurality of implicit abnormal elements on the target device to obtain a plurality of implicit element coordinates; Calculate the distance between every two closest implicit element coordinates to obtain a plurality of implicit element distances, and calculate the mean value to obtain the average implicit element distance as the integration degree; According to the integration degree, configure the detection positioning coefficient.

5. The method of claim 4, wherein, According to the integration degree, configure the detection positioning coefficient, comprising: Obtain all historical implicit element distance sets of electromagnetic interference test on a plurality of devices in a historical time, and select the minimum value to obtain the reference implicit element distance; Calculate the ratio of the reference implicit element distance to the average implicit element distance to obtain the detection positioning coefficient.

6. The method of claim 1, wherein, According to the infrared image, a plurality of temperature abnormal degrees of the plurality of implicit abnormal elements are identified, comprising: According to the plurality of implicit element coordinates of the plurality of implicit abnormal elements, the infrared image is cropped to obtain a plurality of element infrared images; Extracting a plurality of element temperatures of the plurality of element infrared images; Obtain a plurality of average element temperatures under normal operation of the plurality of implicit abnormal elements; Calculate the ratio of the plurality of element temperatures to the plurality of average element temperatures to obtain a plurality of temperature abnormal degrees.

7. The method of claim 1, wherein, The detection positioning coefficients are respectively corrected according to the abnormal test parameters, and array abnormality identification is respectively performed according to the abnormal test parameters, so as to obtain a plurality of element abnormality probabilities as abnormal detection positioning results, including: According to the plurality of temperature abnormality degrees, the detection positioning coefficients are respectively corrected and calculated to obtain a plurality of corrected detection positioning coefficients; The plurality of element electromagnetic abnormality recognizers corresponding to the plurality of hidden abnormal elements are called, wherein each element electromagnetic abnormality recognizer includes T element electromagnetic abnormality recognition branches, and T is a positive integer; According to the plurality of corrected detection positioning coefficients and the branch number T, a plurality of branch calling numbers are calculated and obtained; According to the plurality of branch calling numbers, a plurality of element electromagnetic abnormality recognition branch combinations are randomly called, the abnormal test parameters are respectively input, a plurality of element probability sets are identified and output, and a plurality of element abnormality probabilities are calculated and obtained by taking the mean value, as abnormal detection positioning results.

8. The method of claim 7, wherein, The training steps of the plurality of element electromagnetic abnormality recognizers include: In the electromagnetic interference test data of the same device in the historical time, a sample abnormal test parameter set is collected, and the proportion of the occurrence of the abnormality of a plurality of elements under different sample abnormal test parameters is collected, and a plurality of sample element abnormality probability sets are labeled and obtained; The sample abnormal test parameter set and the plurality of sample element abnormality probability sets are divided to obtain a plurality of sets of element electromagnetic abnormality recognition training data sets, wherein each set of element electromagnetic abnormality recognition training data set includes T element electromagnetic abnormality recognition training data. The plurality of element electromagnetic abnormality recognizers are trained by using the plurality of sets of element electromagnetic abnormality recognition training data sets, wherein each element electromagnetic abnormality recognizer includes T element electromagnetic abnormality recognition branches.

9. An anomaly detection localization system for electromagnetic testing, characterized in that, The method for performing any one of claims 1-8 includes: A data acquisition module for electromagnetic interference test of a target device, and acquiring abnormal position information and abnormal test parameters when a detection abnormality occurs; An infrared image acquisition module for indexing a plurality of hidden abnormal elements in a device abnormal area according to the abnormal position information, and acquiring an infrared image of the device abnormal area; An integration analysis module for performing integration analysis according to the plurality of hidden abnormal elements to obtain an integration degree and configure a detection positioning coefficient; An identification output module for identifying a plurality of temperature abnormality degrees of the plurality of hidden abnormal elements according to the infrared image, correcting the detection positioning coefficients, and performing array abnormality identification according to the abnormal test parameters to obtain a plurality of element abnormality probabilities as abnormal detection positioning results.

Images