Medical isolation power supply system fault analysis method and device, and storage medium

Abstract

A medical isolation power supply system fault analysis method and device and a storage medium belong to the technical field of medical power supplies, and the medical isolation power supply system fault analysis method comprises the following steps: monitoring the frequency spectrum change of a signal injected into a medical isolation power supply system in real time, analyzing the injected signal based on the frequency spectrum change of the signal, and dynamically filtering out a noise interference signal; monitoring the injection signal after the noise interference is filtered out, and performing signal reconstruction on the injection signal to obtain a target signal; and analyzing the target signal based on a pre-trained neural network model, and determining the fault type and position of the medical power supply system. Signals injected into the medical isolation power supply system are analyzed according to signal frequency spectrum changes, noise interference signals are dynamically filtered out, real-time suppression of noise is achieved, then the signals after noise interference are effectively recovered through multi-channel signal reconstruction, target signals are analyzed through a neural network model, and the target signals are obtained. And the fault diagnosis of the medical power supply system can be efficiently and accurately realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical power supplies, and in particular relates to a method, a device and a storage medium for analyzing faults of a medical isolation power supply system. Background Art

[0002] The fault location of the existing medical isolation power system mainly relies on the cooperation of the signal generator and the insulation fault locator. The signal generator generates a detection signal and injects it into the medical isolation power system. Then the insulation fault locator performs signal analysis to determine the fault location. However, when the signal is injected into the medical isolation power system, it usually encounters noise interference in the medical environment, which affects the quality of the signal, causing the signal to be distorted or distorted, affecting the accuracy of fault location. Therefore, how to ensure the integrity of the injected signal in a complex medical environment with noise and interference to improve the accuracy of fault location is a technical problem that needs to be solved urgently. Summary of the invention

[0003] In view of this, an embodiment of the present invention provides a medical isolation power system fault analysis method, device and storage medium, aiming to solve the problem of signal aberration and distortion caused by noise interference in traditional technical solutions, thereby improving the accuracy and reliability of medical isolation power system fault diagnosis and positioning.

[0004] A first aspect of an embodiment of the present invention provides a method for analyzing a fault in a medical isolation power supply system, comprising: Monitor the changes in the signal spectrum injected into the medical isolation power system in real time, analyze the injected signal based on the changes in the signal spectrum, and dynamically filter out noise interference signals; Monitor the injected signal after filtering out noise interference, reconstruct the injected signal, and obtain the target signal; The target signal is analyzed based on the pre-trained neural network model to determine the fault type and location of the medical power system.

[0005] In one embodiment, analyzing the injected signal based on the signal spectrum change and dynamically filtering out the noise interference signal includes: If the amplitude of the signal spectrum in the frequency band is greater than the preset amplitude threshold, the noise interference signal in the frequency band is filtered out; Alternatively, if the frequency spectrum correlation coefficient at adjacent moments is less than a preset correlation threshold, the noise interference signal at the adjacent moments is filtered out.

[0006] In one embodiment, if the amplitude of the signal spectrum in a frequency band is greater than a preset amplitude threshold, filtering out the noise interference signal in the frequency band includes: If the amplitude of the signal spectrum in the frequency band is greater than a preset amplitude threshold, a transfer function of the bandpass filter is determined based on the frequency range in the frequency band; Based on the transfer function, a bandpass filter is used to filter out the noise interference signal within the frequency band.

[0007] In one embodiment, if the spectrum correlation coefficient of adjacent moments is less than a preset correlation threshold, filtering out the noise interference signal of the adjacent moments includes: If the spectrum correlation coefficient at adjacent moments is less than a preset correlation threshold, the frequency response of the adaptive filter is determined based on the spectrum amplitude of the current signal; Based on the frequency response, an adaptive filter is used to filter out the noise interference signals at adjacent moments.

[0008] In one embodiment, monitoring the injected signal after filtering out noise interference, reconstructing the injected signal, and obtaining the target signal includes: Monitor the injected signal after filtering out noise interference, decompose the injected signal based on multi-scale wavelet transform, and extract the low-frequency components and high-frequency components of the injected signal at different transformation scales; According to the low-frequency components and high-frequency components of each channel signal, each channel signal is weighted and fused to obtain a fused signal; The fused signal is subjected to inverse wavelet transform to obtain the target signal.

[0009] In one embodiment, the training data set of the pre-trained neural network model includes signal data corresponding to different fault types and locations of the medical power system.

[0010] In one embodiment, a pre-trained neural network model includes an input layer, multiple hidden layers, an output layer and a loss function; the input layer transmits different features of the input signal to multiple hidden layers through neurons, and the neurons in each hidden layer perform nonlinear transformation on the features transmitted by the input layer through an activation function, and then transmit them to the output layer; the output layer judges the features transmitted by the hidden layer through a classification function to obtain the fault type and location; the number of neurons in the output layer is equal to the number of categories of fault type and location, and classification is performed using a classification function; the loss function is used to evaluate the accuracy of the pre-trained neural network model in judging the fault type and location.

[0011] A second aspect of an embodiment of the present application provides a medical isolation power system fault analysis device, comprising: An analysis module is used to monitor the spectrum changes of the signal injected into the medical isolation power system in real time, analyze the injected signal based on the spectrum changes of the signal, and dynamically filter out the noise interference signal; A reconstruction module is used to monitor the injected signal after the noise interference is filtered out, and to reconstruct the injected signal to obtain the target signal; The determination module is used to analyze the target signal based on the pre-trained neural network model to determine the fault type and location of the medical power supply system.

[0012] In one embodiment, the analysis module is used to: If the amplitude of the signal spectrum in the frequency band is greater than the preset amplitude threshold, the noise interference signal in the frequency band is filtered out; Alternatively, if the frequency spectrum correlation coefficient at adjacent moments is less than a preset correlation threshold, the noise interference signal at the adjacent moments is filtered out.

[0013] In one embodiment, the analysis module includes: A first determination unit, configured to determine a transfer function of a bandpass filter based on a frequency range within the frequency band if the amplitude of a signal spectrum within the frequency band is greater than a preset amplitude threshold; The first filtering unit is used to filter out noise interference signals within a frequency band using a bandpass filter based on a transfer function.

[0014] In one embodiment, the analysis module includes: A second determination unit, configured to determine a frequency response of the adaptive filter based on a spectrum amplitude of a current signal if a spectrum correlation coefficient at adjacent moments is less than a preset correlation threshold; The second filtering unit is used to filter out noise interference signals at adjacent moments by using an adaptive filter based on the frequency response.

[0015] In one embodiment, the reconstruction module includes: An extraction unit is used to monitor the injected signal after the noise interference is filtered out, decompose the injected signal based on multi-scale wavelet transform, and extract the low-frequency components and high-frequency components of the injected signal at different transformation scales; A fusion unit, used for performing weighted fusion on the signals of each channel according to the low-frequency components and high-frequency components of the signals of each channel to obtain a fused signal; The transform unit is used to perform inverse wavelet transform on the fused signal to obtain the target signal.

[0016] In one embodiment, the training data set of the pre-trained neural network model includes signal data corresponding to different fault types and locations of the medical power system.

[0017] In one embodiment, a pre-trained neural network model includes an input layer, multiple hidden layers, an output layer and a loss function; the input layer transmits different features of the input signal to multiple hidden layers through neurons, and the neurons in each hidden layer perform nonlinear transformation on the features transmitted by the input layer through an activation function, and then transmit them to the output layer; the output layer judges the features transmitted by the hidden layer through a classification function to obtain the fault type and location; the number of neurons in the output layer is equal to the number of categories of fault type and location, and classification is performed using a classification function; the loss function is used to evaluate the accuracy of the pre-trained neural network model in judging the fault type and location.

[0018] A third aspect of an embodiment of the present application provides a medical isolation power system fault analysis device, comprising: a processor, a memory, and a computer program stored in the memory and executable on the processor; when the processor executes the computer program, the steps of the medical isolation power system fault analysis method described in the first aspect above are implemented.

[0019] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the medical isolation power system fault analysis method described in the first aspect above are implemented.

[0020] The beneficial effects of the embodiments of the present application are as follows: by analyzing the signal injected into the medical isolation power system according to the change of the signal spectrum, the noise interference signal is dynamically filtered out to achieve real-time noise suppression, and then through multi-channel signal reconstruction, the signal after noise interference is effectively restored, and then the target signal is analyzed through a neural network model, so that the fault diagnosis of the medical power system can be achieved efficiently and accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0022] Figure 1 A schematic diagram of the implementation flow of a method for analyzing a fault in a medical isolation power system provided in one embodiment of the present application; Figure 2 for Figure 1 Schematic diagram of the specific implementation process of S102; Figure 3 is a schematic diagram of a medical isolation power system fault analysis device provided in the second embodiment of the present application; Figure 4It is a schematic diagram of a medical isolation power system fault analysis device provided in the third embodiment of the present application. DETAILED DESCRIPTION

[0023] The following embodiments of the technical solution of the present application are described in detail in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present application, and are therefore only used as examples, and cannot be used to limit the scope of protection of the present application.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" in the specification and claims of this application and the above-mentioned figure descriptions and any variations thereof are intended to cover non-exclusive inclusions.

[0025] In the description of the embodiments of the present application, the technical terms "first", "second", etc. are only used to distinguish different objects, and cannot be understood as indicating or implying relative importance or implicitly indicating the number, specific order or primary and secondary relationship of the indicated technical features. In the description of the embodiments of the present application, the meaning of "multiple" is more than two, unless otherwise clearly and specifically defined.

[0026] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0027] In the description of the embodiments of the present application, the term "and / or" is only a description of the association relationship of the associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0028] In the description of the embodiments of the present application, the term "multi-frame" refers to more than two (including two).

[0029] In the description of the embodiments of the present application, the technical terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc. indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, which are only for the convenience of describing the embodiments of the present application and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the embodiments of the present application.

[0030] An embodiment of the present invention provides a method for analyzing faults of a medical isolation power supply system. The method analyzes the signal injected into the medical isolation power supply system according to the change of the signal spectrum, dynamically filters out the noise interference signal, realizes real-time suppression of the noise, and then effectively restores the signal after noise interference through multi-channel signal reconstruction. The target signal is then analyzed through a neural network model, thereby realizing efficient and accurate fault diagnosis of the medical power supply system.

[0031] See also Figure 1 , Figure 1 A schematic diagram of the implementation flow of a method for analyzing a medical isolation power system failure provided in an embodiment of the present application. Figure 1 It can be seen that the medical isolation power system fault risk method provided in the embodiment of the present application includes steps S101 to S104. The details are as follows: S101: monitor the spectrum changes of the signal injected into the medical isolation power system in real time, analyze the injected signal based on the spectrum changes of the signal, and dynamically filter out the noise interference signal.

[0032] In a complex medical environment, noise interference usually affects the signal in the form of high or low frequency. This noise may come from electromagnetic interference, signal leakage from other electrical equipment, or other unstable factors. Noise interference can cause changes in the signal spectrum. Specifically, changes in noise intensity can be captured by peaks or abnormal fluctuations in the spectrum. Noise interference often covers or overlaps with the effective frequency range of the signal. This causes the spectrum of the signal to change, usually manifested as distortion or expansion of the spectrum. Noise may cause an increase in amplitude or a change in shape in the signal spectrum, which is significantly different from the spectrum of a pure signal. Therefore, by analyzing the changes in the signal spectrum injected into the medical isolation power system, the noise interference signal can be effectively filtered out.

[0033] In one embodiment, the injected signal is analyzed based on the change in the signal spectrum, and the noise interference signal is dynamically filtered out, including: if the amplitude of the signal spectrum in the frequency band is greater than a preset amplitude threshold, the noise interference signal in the frequency band is filtered out; or, if the spectrum correlation coefficient at adjacent moments is less than a preset correlation threshold, the noise interference signal at adjacent moments is filtered out.

[0034] Specifically, the signal spectrum can be obtained through Fourier transform, and the frequency band of the noise can be identified through the signal spectrum. The change of the signal amplitude in the signal frequency band spectrum can reflect whether the frequency band is interfered by noise. In addition, in the present application, spectrum monitoring is not a static process, but by tracking the changes in the signal spectrum in real time, it can dynamically respond to changes in environmental noise, flexibly adjust and suppress noise.

[0035] In one embodiment, if the amplitude of the signal spectrum in a frequency band is greater than a preset amplitude threshold, the noise interference signal in the frequency band is filtered out, including: if the amplitude of the signal spectrum in the frequency band is greater than the preset amplitude threshold, determining the transfer function of the bandpass filter based on the frequency range in the frequency band; based on the transfer function, using the bandpass filter to filter out the noise interference signal in the corresponding frequency band.

[0036] For example, assuming that in the frequency band to If the amplitude of the signal spectrum is greater than the preset amplitude threshold, it is determined that the frequency band is interfered by noise. and The transfer function of the bandpass filter is determined as: This filter can remove low frequencies (frequencies below ) and high frequencies (frequencies above ) noise, retaining the signal in the frequency band not affected by noise. The filtered signal is expressed as: ;in, is the signal spectrum, is the transfer function of the filter, is the signal after filtering.

[0037] The design and adjustment of the bandpass filter of the present application is customized based on the results of real-time spectrum analysis, rather than a preset fixed filter. This dynamic adaptability can more accurately suppress noise interference. Through accurate spectrum monitoring and dynamic filtering, noise interference can be effectively removed, the effective information of the original signal can be retained, and the signal quality can be improved.

[0038] In one embodiment, if the spectral correlation coefficient of adjacent moments is less than a preset correlation threshold, the noise interference signal of the adjacent moments is filtered out, including: if the spectral correlation coefficient of adjacent moments is less than the preset correlation threshold, based on the spectral amplitude of the current signal, determining the frequency response of the adaptive filter; based on the frequency response, using the adaptive filter to filter out the noise interference signal of the adjacent moments.

[0039] For example, at time and The spectrum below and , the corresponding spectral correlation coefficient in this application is expressed as: in, and It is adjacent time and The signal spectrum, yes The complex conjugate of , Re represents the real part. This formula measures the similarity of the two signal spectra in the frequency domain. The closer the corresponding correlation coefficient is to 1, the more similar the two signal spectra are; if the correlation coefficient is smaller, it means that there is a large difference between the two signals, indicating that the spectrum shape has changed significantly, which may be signal distortion caused by noise.

[0040] In this application, the frequency response of the adaptive filter is determined based on the spectrum amplitude of the current signal, which is expressed as: in, is the frequency response of the adaptive filter, is the spectrum amplitude of the current signal, is a sign function, which is used to ensure that the filter maintains the phase information of the signal and avoids phase distortion when adjusting the frequency response. | is the spectrum amplitude of the target signal.

[0041] By dynamically adjusting the frequency response according to the ratio of the spectrum amplitude of the current signal to the spectrum amplitude of the target signal, the noise can be effectively filtered out and the signal can be enhanced, so that the target signal can still be accurately restored in a noisy environment. ) to ensure that the output signal of the filter not only has the correct amplitude change in the frequency domain, but also retains the phase structure of the original signal.

[0042] Based on the frequency response, an adaptive filter is used to filter out the noise interference signal at adjacent moments, which can be expressed as: in, For the moment The signal spectrum is the signal obtained after filtering out the noise interference signal. is an adaptive filter, is the inverse Fourier transform.

[0043] In the present application, the correlation threshold can be dynamically adjusted according to the historical data of the signal and the statistical characteristics of the noise. For example, an adaptive correlation threshold can be calculated based on the spectrum correlation at the previous moment, so as to respond to noise interference more sensitively. The frequency response of the adaptive filter can be dynamically adjusted according to the spectrum amplitude of the signal to avoid the blind spots caused by fixed filter parameters. The adaptive filter can consider the information in the time and frequency domains at the same time, making the noise filtering process more refined. Combining the changes in time and the changes in frequency characteristics, different types of noise can be removed more efficiently.

[0044] S102: Monitor the injected signal after noise interference is filtered out, and reconstruct the injected signal to obtain a target signal.

[0045] By reconstructing the injected signal after filtering out the noise interference, the problem of signal loss caused by noise interference during the propagation process can be solved, the integrity of the signal can be restored, and finally the target signal that can be used for subsequent analysis and fault diagnosis can be obtained.

[0046] In practical applications, the injection signal is usually collected by multiple sensors or signal channels at the same time. Each channel may be interfered by noise to varying degrees, so the signals of these different channels need to be integrated. After filtering out the noise, the signal in each channel can be restored to a content close to the original signal, but due to the randomness of the noise and the diversity of interference, the signals of different channels may still have some differences. Therefore, by combining the signals of each channel, a more accurate target signal can be restored.

[0047] For example, Figure 2 As shown, Figure 2 for Figure 1 The specific implementation flow diagram of S102 is shown in FIG. Figure 2 It can be seen that S102 includes: S1021: monitor the injected signal after the noise interference is filtered out, decompose the injected signal based on the multi-scale wavelet transform, and extract the low-frequency components and high-frequency components of the injected signal at different transformation scales.

[0048] In the present application, the injected signal after filtering out noise interference is decomposed based on multi-scale wavelet transform, and the low-frequency component (referring to the frequency part lower than or equal to the preset frequency threshold, the low-frequency component can reflect the change trend) and high-frequency component (referring to the frequency part higher than the preset frequency threshold, the high-frequency component can reflect the signal mutation point) of the signal can be extracted at different scales. Specifically, the injected signal is decomposed based on multi-scale wavelet transform, which is expressed as: in, (a, b) are the wavelet coefficients of the signal, is the injected signal after filtering out noise interference, is the wavelet basis function, a is the scale factor, and b is the translation factor.

[0049] S1022: Perform weighted fusion on the signals of each channel according to the low-frequency components and high-frequency components of the signals of each channel to obtain a fused signal.

[0050] After wavelet transform, multi-channel signals are fused. The useful information (including low-frequency components and high-frequency components) provided by each signal channel can be regarded as a redundant source of the signal. The quality and noise level of the signal are determined based on the useful information included in each channel, and the weight of each channel signal is adaptively adjusted to maximize the reconstruction quality of the signal.

[0051] The signals of each channel are weightedly fused according to the low-frequency components and high-frequency components of the signals of each channel to obtain a fused signal, which can be expressed as: the fusion weight of each channel signal is determined according to the low-frequency components and high-frequency components of the signals of each channel; based on the fusion weight of each channel signal, the signals of each channel are weightedly fused to obtain a fused signal.

[0052] Specifically, the fusion weight of each channel signal is determined according to the low-frequency component and high-frequency component of each channel signal, which is expressed as: in, is the total energy of channel i, is the weight ratio of low-frequency components, is the weight ratio of high frequency components, is the low frequency energy, It is high frequency energy.

[0053] In this application, the noise and signal spectra may change by adjusting the weight ratio of low-frequency components and high-frequency components in real time (i.e., by adjusting and ), the fusion weights of individual channel signals can be optimized in real time.

[0054] By analyzing the spectrum of each channel signal, the energy of low-frequency components and high-frequency components can be calculated, and the fusion weight can be determined according to the ratio of these energies. In this way, the effective components of each signal channel can be effectively extracted, and noise can be suppressed in the signal fusion process to improve the quality of the target signal.

[0055] S1023: Perform inverse wavelet transform on the fused signal to obtain the target signal.

[0056] By performing multi-scale wavelet analysis on the signal, the details and trends of the signal can be accurately extracted at different scale levels. Adaptive weights are used to fuse multi-channel signals, and the weights of each channel are adjusted according to the real-time quality of the signal to achieve optimal signal reconstruction.

[0057] S103: Analyze the target signal based on the pre-trained neural network model to determine the fault type and location of the medical power system.

[0058] In one embodiment, the training data set of the pre-trained neural network model includes signal data corresponding to different fault types and locations of the medical power system. Specifically, in order to train the neural network, a large amount of signal data of the medical power system needs to be collected, and these data should include signal samples of different fault types and locations. For example, different types of faults (such as overload, short circuit, open circuit, insulation fault, etc.) can be simulated and the corresponding signal changes can be recorded. Each signal data needs to be provided together with the corresponding fault type (such as overload, short circuit) and fault location (such as left side, right side) label. The label data is the supervisory signal for neural network training.

[0059] In one embodiment, a pre-trained neural network model includes an input layer, multiple hidden layers, an output layer and a loss function; the input layer transmits different features of the input signal to multiple hidden layers through neurons, and the neurons in each hidden layer perform nonlinear transformation on the features transmitted by the input layer through an activation function, and then transmit them to the output layer; the output layer judges the features transmitted by the hidden layer through a classification function to obtain the fault type and location; the number of neurons in the output layer is equal to the number of categories of fault type and location, and classification is performed using a classification function; the loss function is used to evaluate the accuracy of the pre-trained neural network model in judging the fault type and location.

[0060] Exemplarily, the classification function of the output layer is expressed as: ;in, is the weighted sum of the i-th category faults at the k-th location, is the total number of categories of fault type and location, represents the fault at position k, represents a category i fault, is the predicted probability of the i-th category and the k-th position.

[0061] The neural network model provided in this application outputs the probability distribution of fault type and location at the last layer of the network through the classification function. For each input signal, the neural network will output a probability of fault type and location, and select the one with the highest probability as the prediction result.

[0062] From the above analysis, it can be seen that the medical isolation power system fault analysis method provided by the present application includes: real-time monitoring of the signal spectrum changes injected into the medical isolation power system, analyzing the injected signal based on the signal spectrum changes, and dynamically filtering out the noise interference signal; monitoring the injected signal after filtering out the noise interference, reconstructing the injected signal to obtain the target signal; analyzing the target signal based on the pre-trained neural network model to determine the fault type and location of the medical power system. By analyzing the signal injected into the medical isolation power system according to the signal spectrum changes, dynamically filtering out the noise interference signal, and realizing real-time noise suppression, and then effectively restoring the signal after the noise interference through multi-channel signal reconstruction, and then analyzing the target signal through the neural network model, the fault diagnosis of the medical power system can be realized efficiently and accurately.

[0063] See also Figure 3 As shown, Figure 3 Schematic diagram of a medical isolation power system fault analysis device provided in the second embodiment of the present application. The units included are used to perform Figure 1 to Figure 2 For details, please refer to the steps in the corresponding embodiment. Figure 1 to Figure 2 For the convenience of explanation, only the parts related to this embodiment are shown. Figure 3 , a medical isolation power system fault analysis device 30, comprising: The analysis module 310 is used to monitor the changes in the signal spectrum injected into the medical isolation power system in real time, analyze the injected signal based on the changes in the signal spectrum, and dynamically filter out noise interference signals.

[0064] The reconstruction module 320 is used to monitor the injected signal after the noise interference is filtered out, and to reconstruct the injected signal to obtain the target signal.

[0065] The determination module 330 is used to analyze the target signal based on the pre-trained neural network model to determine the fault type and location of the medical power system.

[0066] In one embodiment, the analysis module 310 is used to: If the amplitude of the signal spectrum in the frequency band is greater than the preset amplitude threshold, the noise interference signal in the frequency band is filtered out; Alternatively, if the frequency spectrum correlation coefficient at adjacent moments is less than a preset correlation threshold, the noise interference signal at the adjacent moments is filtered out.

[0067] In one embodiment, the analysis module 310 includes: A first determination unit, configured to determine a transfer function of a bandpass filter based on a frequency range within the frequency band if the amplitude of a signal spectrum within the frequency band is greater than a preset amplitude threshold; The first filtering unit is used to filter out noise interference signals within a frequency band using a bandpass filter based on a transfer function.

[0068] In one embodiment, the analysis module 310 includes: A second determination unit, configured to determine a frequency response of the adaptive filter based on a spectrum amplitude of a current signal if a spectrum correlation coefficient at adjacent moments is less than a preset correlation threshold; The second filtering unit is used to filter out noise interference signals at adjacent moments by using an adaptive filter based on the frequency response.

[0069] In one embodiment, the reconstruction module 320 includes: An extraction unit is used to monitor the injected signal after the noise interference is filtered out, decompose the injected signal based on multi-scale wavelet transform, and extract the low-frequency components and high-frequency components of the injected signal at different transformation scales; A fusion unit, used for performing weighted fusion on the signals of each channel according to the low-frequency components and high-frequency components of the signals of each channel to obtain a fused signal; The transform unit is used to perform inverse wavelet transform on the fused signal to obtain the target signal.

[0070] In one embodiment, the training data set of the pre-trained neural network model includes: signal data corresponding to different fault types and locations of the medical power system.

[0071] In one embodiment, a pre-trained neural network model includes an input layer, multiple hidden layers, an output layer and a loss function; the input layer transmits different features of the input signal to multiple hidden layers through neurons, and the neurons in each hidden layer perform nonlinear transformation on the features transmitted by the input layer through an activation function, and then transmit them to the output layer; the output layer judges the features transmitted by the hidden layer through a classification function to obtain the fault type and location; the number of neurons in the output layer is equal to the number of categories of fault type and location, and classification is performed using a classification function; the loss function is used to evaluate the accuracy of the pre-trained neural network model in judging the fault type and location.

[0072] See also Figure 4 , Figure 4 4 is a schematic diagram of a medical isolation power system fault analysis device provided in the third embodiment of the present application. The medical isolation power system fault analysis device 400 includes: a processor 410, a memory 420, and a computer program 430 stored in the memory 420 and executable on the processor 410; when the processor 410 executes the computer program 430, the steps in the above-mentioned medical isolation power system fault analysis method embodiment are implemented, such as Figure 1Alternatively, when the processor 410 executes the computer program 430, the functions of each module / unit in the above-mentioned device embodiments are implemented, for example Figure 3 Functions of modules 310 to 330 are shown.

[0073] Exemplarily, the computer program 430 may be divided into one or more modules / units, one or more modules / units are stored in the memory 420 and executed by the processor 410 to complete the present application. One or more modules / units may be a series of computer program instruction segments capable of completing specific functions, and the instruction segments are used to describe the execution process of the computer program 430 in the medical isolation power system fault analysis device 400. For example, the computer program 430 may be divided into an analysis module, a reconstruction module, and a determination module, and the specific functions of each module are as follows: The analysis module 310 is used to monitor the changes in the signal spectrum injected into the medical isolation power system in real time, analyze the injected signal based on the changes in the signal spectrum, and dynamically filter out noise interference signals.

[0074] The reconstruction module 320 is used to monitor the injected signal after the noise interference is filtered out, and to reconstruct the injected signal to obtain the target signal.

[0075] The determination module 330 is used to analyze the target signal based on the pre-trained neural network model to determine the fault type and location of the medical power system.

[0076] The medical isolation power system fault analysis device 400 provided in this embodiment may include, but is not limited to, a processor 410 and a memory 420. Those skilled in the art will appreciate that Figure 4 It is only an example of a medical isolation power system fault analysis device and does not constitute a limitation of the medical isolation power system fault analysis device. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the medical isolation power system fault analysis device may also include input and output devices, network access devices, buses, etc.

[0077] The processor 410 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0078] The memory 420 may be an internal storage unit of the medical isolation power system fault analysis device 400, such as a hard disk or memory of the medical isolation power system fault analysis device. The memory 420 may also be an external storage device of the medical isolation power system fault analysis device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. equipped on the medical isolation power system fault analysis device. Further, the medical isolation power system fault analysis device may also include both an internal storage unit and an external storage device. The memory 420 is used to store computer programs and other programs and data required by the medical isolation power system fault analysis device. The memory 420 may also be used to temporarily store data that has been output or is to be output.

[0079] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0080] An embodiment of the present application also provides a network device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor implements the steps in any of the above-mentioned method embodiments when executing the computer program.

[0081] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0082] An embodiment of the present application provides a computer program product. When the computer program product runs on a mobile terminal, the mobile terminal can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0083] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the camera / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, RandomAccess Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0084] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0085] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0086] In the embodiments provided in the present application, it should be understood that the disclosed devices / network equipment and methods can be implemented in other ways. For example, the device / network equipment embodiments described above are merely schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0087] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0088] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for analyzing faults in a medical isolation power supply system, characterized in that: include: Monitor the changes in the signal spectrum injected into the medical isolation power system in real time, analyze the injected signal based on the changes in the signal spectrum, and dynamically filter out noise interference signals; Monitoring the injected signal after noise interference is filtered out, and performing signal reconstruction on the injected signal to obtain a target signal; The target signal is analyzed based on a pre-trained neural network model to determine the fault type and location of the medical power system.

2. The medical isolation power system fault analysis method according to claim 1, characterized in that: The step of analyzing the injected signal based on the signal spectrum change and dynamically filtering out the noise interference signal includes: If the amplitude of the signal spectrum in a frequency band is greater than a preset amplitude threshold, filtering out the noise interference signal in the frequency band; Alternatively, if the frequency spectrum correlation coefficient at adjacent moments is less than a preset correlation threshold, the noise interference signal at the adjacent moments is filtered out.

3. The medical isolation power system fault analysis method according to claim 2, characterized in that: If the amplitude of the signal spectrum in a frequency band is greater than a preset amplitude threshold, filtering out the noise interference signal in the frequency band includes: If the amplitude of the signal spectrum in a frequency band is greater than a preset amplitude threshold, determining a transfer function of the bandpass filter based on the frequency range in the frequency band; Based on the transfer function, a bandpass filter is used to filter out noise interference signals within the frequency band.

4. The medical isolation power system fault analysis method according to claim 2, characterized in that: If the spectrum correlation coefficient at adjacent moments is less than a preset correlation threshold, filtering out the noise interference signal at the adjacent moments includes: If the spectrum correlation coefficient at adjacent moments is less than a preset correlation threshold, the frequency response of the adaptive filter is determined based on the spectrum amplitude of the current signal; Based on the frequency response, an adaptive filter is used to filter out the noise interference signal at the adjacent time.

5. The medical isolation power system fault analysis method according to claim 1, characterized in that: The monitoring of the injected signal after the noise interference is removed, and performing signal reconstruction on the injected signal to obtain the target signal includes: Monitoring the injected signal after filtering out noise interference, decomposing the injected signal based on multi-scale wavelet transform, and extracting low-frequency components and high-frequency components of the injected signal at different transformation scales; According to the low-frequency components and the high-frequency components of the channel signals, the signals of each channel are weightedly fused to obtain a fused signal; The fused signal is subjected to inverse wavelet transform to obtain the target signal.

6. The medical isolation power system fault analysis method according to claim 1, characterized in that: The training data set of the pre-trained neural network model includes signal data corresponding to different fault types and locations of the medical power supply system.

7. The medical isolation power system fault analysis method according to claim 6, characterized in that: The pre-trained neural network model includes an input layer, multiple hidden layers, an output layer and a loss function; the input layer transmits different features of the input signal to the multiple hidden layers through neurons, and the neurons of each hidden layer perform nonlinear transformation on the features transmitted by the input layer through an activation function, and then transmit the features to the output layer; the output layer judges the features transmitted by the hidden layer through a classification function to obtain the fault type and location; the number of neurons in the output layer is equal to the number of categories of the fault type and location, and classification is performed using a classification function; The loss function is used to evaluate the accuracy of the pre-trained neural network model in determining the fault type and location.

8. A medical isolation power system fault analysis device, characterized in that: include: An analysis module, used for real-time monitoring of changes in the spectrum of the signal injected into the medical isolation power supply system, analyzing the injected signal based on the changes in the spectrum of the signal, and dynamically filtering out noise interference signals; A reconstruction module is used to monitor the injected signal after the noise interference is filtered out, and to reconstruct the injected signal to obtain a target signal; A determination module is used to analyze the target signal based on a pre-trained neural network model to determine the fault type and location of the medical power system.

9. A medical isolation power system fault analysis device, characterized in that: include: A processor, a memory, and a computer program stored in the memory and executable on the processor; When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images