Anti-interference signal processing method and related equipment for FPC sensing module

Abstract

The present invention relates to an anti-interference signal processing method and related equipment for an FPC sensing module, including the following steps: performing time-frequency domain conversion on the original signal to obtain a signal time-frequency spectrum diagram; identifying whether there is an abnormal interference area in the signal time-frequency spectrum diagram, and if so, marking the abnormal interference area; performing feature analysis and classification on the marked abnormal interference area based on the signal time-frequency spectrum diagram to obtain the interference signal type; predicting the interference source distribution for the target FPC sensing module based on the interference signal type and the interference signal corresponding to the interference signal type to obtain an interference source feature mapping matrix; performing anti-interference processing on the interference signal in the original signal based on the interference source feature mapping matrix to obtain a target working signal, solving the technical problem that the interference signal not only affects the accuracy of the FPC sensing module, but also causes system misjudgment or function failure, seriously affecting the reliability and stability of the equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of FPC sensing modules, and particularly to an anti-interference signal processing method and related equipment for an FPC sensing module. Background Art

[0002] FPC (Flexible Printed Circuit) sensing modules are widely used in modern electronic devices. Their high sensitivity and flexibility make them indispensable in various application scenarios. However, with the complexity of electronic devices and the variability of the environment, FPC sensing modules face increasingly serious interference problems in practical applications. These interference signals not only affect the accuracy of the sensors but may also cause system misjudgment or functional failure, seriously affecting the reliability and stability of the devices.

[0003] Currently, the methods for dealing with interference signals mainly rely on hardware filtering or shielding techniques, but these methods have certain limitations. For example, hardware filters are ineffective in dealing with high-frequency and wide-band interference, and they increase the cost and complexity of the devices. In addition, traditional shielding methods cannot completely isolate certain environmental interferences, such as electromagnetic interference (EMI), radio frequency interference (RFI), etc. The complexity and diversity of these interference signals have prompted researchers to seek more efficient and intelligent anti-interference signal processing methods.

[0004] Therefore, it is particularly important to study an algorithm that can intelligently identify and process interference signals. By combining signal processing techniques and machine learning algorithms, the anti-interference ability of FPC sensing modules can be improved without increasing the hardware cost. However, existing algorithms still have deficiencies in terms of real-time performance, accuracy, and the identification and classification of different types of interference signals. How to effectively extract the characteristics of interference signals, classify them, and implement targeted anti-interference processing is a key issue in current research. Summary of the Invention

[0005] The main objective of the present invention is to provide an anti-interference signal processing method and related equipment for an FPC sensing module, which solves the technical problem that interference signals not only affect the accuracy of the FPC sensing module but also cause system misjudgment or functional failure, seriously affecting the reliability and stability of the devices.

[0006] To achieve the above objective, the present invention provides an anti-interference signal processing method for an FPC sensing module, including the following steps:

[0007] Collect the original signal of the target FPC sensing module through a preset signal sensor, and perform time-frequency domain conversion on the original signal through a preset Morlet wavelet transform technology to obtain a signal time-frequency spectrum diagram;

[0008] Identify whether there is an abnormal interference area in the time-frequency spectrum diagram of the signal through a preset abnormal recognition model. If so, mark the abnormal interference area; where the abnormal interference area represents an interference signal.

[0009] Through a preset spectral clustering algorithm, perform feature analysis and classification on the marked abnormal interference area based on the time-frequency spectrum diagram of the signal to obtain the type of interference signal.

[0010] Based on the type of interference signal and the interference signal corresponding to the type of interference signal, predict the distribution of interference sources for the target FPC sensing module to obtain an interference source feature mapping matrix.

[0011] Through a preset signal processing technology, perform anti-interference processing on the interference signal in the original signal based on the interference source feature mapping matrix to obtain a target working signal.

[0012] Further, the time-frequency domain conversion of the original signal through a preset Morlet wavelet transform technology to obtain a time-frequency spectrum diagram of the signal includes:

[0013] Perform scale discretization processing on the original signal to obtain different scale discretized signals; where the original signal includes an original electrical signal and an original communication signal.

[0014] Through a preset Morlet wavelet transform technology, perform time-frequency domain conversion and wavelet transform on different scale discretized signals to obtain a series of wavelet signals; where the series of wavelet signals represents the energy distribution of the original signal in different frequency bands and time windows.

[0015] Perform dimensionality reduction processing on the wavelet signal to obtain a dimensionality-reduced signal.

[0016] Perform spatial interpolation on the dimensionality-reduced signal through a preset bilinear interpolation algorithm to obtain an interpolated signal.

[0017] Adopt a preset gray mapping algorithm to convert the interpolated signal into a time-frequency spectrum diagram of the signal.

[0018] Further, the identification of whether there is an abnormal interference area in the time-frequency spectrum diagram of the signal through a preset abnormal recognition model. If so, mark the abnormal interference area, including:

[0019] Extract local binary pattern features from the time-frequency spectrum diagram of the signal to obtain an LBP feature vector sequence.

[0020] Based on a preset sliding window, perform time segmentation on the LBP feature vector sequence to obtain multiple LBP feature vector subsequences; where each LBP feature vector subsequence represents a local area of the time-frequency spectrum diagram of the signal.

[0021] Perform singular value decomposition on each of the LBP feature vector subsequences to obtain corresponding singular value vectors; wherein, the singular value vectors are used to reflect the energy distribution characteristics of the LBP feature vector subsequences.

[0022] Based on the singular value vectors, use a preset anomaly recognition model to identify whether there are abnormal interference regions in the signal time-frequency spectrum diagram.

[0023] If there are any, calculate the degree of abnormality of the existing abnormal interference regions based on a preset degree-of-abnormality evaluation algorithm to obtain corresponding abnormality scores; wherein, the higher the abnormality score, the greater the possibility of interference signals existing in the local region of the signal time-frequency spectrum diagram.

[0024] Combine and organize the abnormal interference regions in the signal time-frequency spectrum diagram with abnormality scores greater than a preset threshold, and label the combined and organized abnormal interference regions. Among them, the combination and organization include removing overlapping regions and merging adjacent regions, and the abnormal interference regions include the start time, duration, and frequency range of the interference signals.

[0025] Further, based on the signal time-frequency spectrum diagram, perform feature analysis and classification on the labeled abnormal interference regions through a preset spectral clustering algorithm to obtain interference signal types, including:

[0026] Input the labeled abnormal interference regions into a preset region-signal mapping algorithm to obtain the interference signals corresponding to the labeled abnormal interference regions.

[0027] Through high-order spectral estimation technology, extract the energy distribution characteristics of the interference signals in different frequency bands on the signal time-frequency spectrum diagram.

[0028] Construct a spectral feature map based on the energy distribution characteristics.

[0029] In the spectral feature map, use a preset spectral clustering algorithm to analyze the feature similarity of each interference signal to obtain a similar feature distribution result; wherein, the similar feature distribution result includes signal frequency aggregation degree, energy density distribution, and band correlation coefficient.

[0030] Perform dimensionality reduction processing on the similar feature distribution result to obtain a dimensionality-reduced similar feature distribution result.

[0031] Input the dimensionality-reduced similar feature distribution result into a preset support vector machine for classification to obtain interference signal types.

[0032] Further, predicting the interference source distribution for the target FPC sensing module based on the interference signal type and the corresponding interference model number of the interference signal type to obtain an interference source feature mapping matrix includes:

[0033] Extract the interference signal corresponding to the interference signal type;

[0034] Perform high-order Volterra series analysis on the interference signal corresponding to the interference signal type to obtain a non-linear time-varying interference feature sequence; wherein, the non-linear time-varying interference feature sequence includes amplitude features, phase features, and frequency features;

[0035] Obtain the physical layout distribution map corresponding to the target FPC sensing module, and perform probability distribution analysis on the physical layout distribution map based on the non-linear time-varying interference feature sequence through a non-parametric kernel density estimation method to obtain the interference source distribution spatial density; wherein, the interference source distribution spatial density includes the position probability and intensity probability of the interference source;

[0036] Input the interference source distribution spatial density into a preset deep belief network for interference source prediction to obtain a predicted interference source;

[0037] Adopt an adaptive Kalman particle filtering algorithm to perform dynamic trajectory prediction on the predicted interference source to obtain the interference source propagation trajectory characteristics; wherein, it includes the movement direction characteristics and speed characteristics of the interference source;

[0038] Perform spatio-temporal correlation analysis on the interference source propagation trajectory characteristics to obtain the interference source influence range;

[0039] Based on the interference source influence range and the interference signal type, obtain an interference source feature mapping matrix; wherein, the interference source feature mapping matrix includes rows, columns, and elements, the rows represent the interference source type, the columns represent the physical positions of the FPC sensing module, and the elements represent the feature intensity of the interference signal.

[0040] Further, performing probability distribution analysis on the physical layout distribution map based on the non-linear time-varying interference feature sequence to obtain the interference source distribution spatial density includes:

[0041] Perform feature extraction on the non-linear time-varying interference feature sequence to obtain non-linear dynamic features; wherein, the non-linear dynamic features include the non-linear coefficient, time constant, and frequency response characteristics of the system;

[0042] Perform multi-scale entropy analysis on the non-linear dynamic features to obtain the complexity characteristics of the interference signal; wherein, the complexity characteristics represent the complexity and regularity of the interference signal at different time scales;

[0043] Based on the non-linear dynamic features and the complexity features, perform spectral envelope analysis on the physical layout distribution diagram through a preset spectral envelope estimation algorithm to obtain a spectral envelope distribution;

[0044] Use the spectral envelope distribution to perform probability distribution analysis on the physical layout distribution diagram to obtain the spatial density of the interference source distribution.

[0045] Further, the anti-interference processing of the interference signal based on the interference source feature mapping matrix through a preset signal processing technology to obtain a target working signal includes:

[0046] Perform spatial correlation analysis on the interference signal based on the interference source feature mapping matrix to obtain the position corresponding to the interference signal;

[0047] At the position, perform notch processing on the interference signal through a preset adaptive notch filter bank and a preset signal processing technology to obtain a notch processed signal;

[0048] Extract a preset signal from the notch processed signal through a preset signal subspace projection algorithm, and separate the preset signal from the notch processed signal to obtain a preliminary working signal and a remaining interference signal; wherein, the preset signal is the signal required for the FPC sensing module during operation;

[0049] Perform signal enhancement on the remaining interference signal to obtain an enhanced signal;

[0050] Perform frequency selective processing on the enhanced signal to obtain a selective filtering scheme;

[0051] Use the selective filtering scheme to perform anti-interference signal processing on the remaining interference signal to reduce the contamination of the remaining interference signal to the preliminary working signal and obtain a target working signal.

[0052] The present invention also provides an anti-interference signal processing device for an FPC sensing module, including:

[0053] A conversion module for collecting the original signal of the target FPC sensing module through a preset signal sensor, and performing time-frequency domain conversion on the original signal through a preset Morlet wavelet transform technology to obtain a signal time-frequency spectrogram;

[0054] An identification module for identifying whether there is an abnormal interference area in the signal time-frequency spectrogram through a preset abnormal identification model, and if so, marking the abnormal interference area; wherein, the abnormal interference area represents an interference signal;

[0055] A classification module, configured to perform feature analysis and classification on the labeled abnormal interference region based on the signal time-frequency spectrogram through a preset spectral clustering algorithm, so as to obtain the types of interference signals;

[0056] A prediction module, configured to perform interference source distribution prediction on the target FPC sensing module based on the types of interference signals and the interference signals corresponding to the types of interference signals, so as to obtain an interference source feature mapping matrix;

[0057] An anti-interference module, configured to perform anti-interference processing on the interference signals in the original signal based on the interference source feature mapping matrix through a preset signal processing technology, so as to obtain a target working signal.

[0058] The present invention further provides a computer device, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps of the method described in any one of the above are implemented.

[0059] The present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method described in any one of the above are implemented.

[0060] The anti-interference signal processing method for an FPC sensing module provided by the present invention includes the following steps: collecting an original signal of a target FPC sensing module through a preset signal sensor, and performing time-frequency domain conversion on the original signal through a preset Morlet wavelet transform technology to obtain a signal time-frequency spectrogram; identifying whether there is an abnormal interference region in the signal time-frequency spectrogram through a preset abnormal recognition model, and if so, labeling the abnormal interference region; wherein the abnormal interference region represents an interference signal; performing feature analysis and classification on the labeled abnormal interference region based on the signal time-frequency spectrogram through a preset spectral clustering algorithm to obtain the types of interference signals; performing interference source distribution prediction on the target FPC sensing module based on the types of interference signals and the interference signals corresponding to the types of interference signals to obtain an interference source feature mapping matrix; performing anti-interference processing on the interference signals in the original signal based on the interference source feature mapping matrix through a preset signal processing technology to obtain a target working signal. Through the above technical means, the technical problem that interference signals not only affect the accuracy of the FPC sensing module, but also cause system misjudgment or functional failure, seriously affecting the reliability and stability of the device is solved. It is realized that by using signal processing technology and an interference source feature mapping matrix, targeted anti-interference processing can be performed on the interference signals in the original signal to obtain a target working signal closer to the real one. This processing method not only improves the quality of the signal, but also enhances the reliability and stability of the FPC sensing module. Description of the Drawings

[0061] Figure 1 It is a schematic diagram of the steps of the anti-interference signal processing method of the FPC sensing module in an embodiment of the present invention;

[0062] Figure 2 It is a block diagram of the structure of the anti-interference signal processing device of the FPC sensing module in an embodiment of the present invention;

[0063] Figure 3 It is a schematic block diagram of the structure of a computer device in an embodiment of the present invention.

[0064] The realization, functional characteristics and advantages of the object of the present invention will be further described with reference to the embodiments and the accompanying drawings. Detailed implementation manners

[0065] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0066] As Figure 1 shown, Figure 1 It is a schematic diagram of the steps of an anti-interference signal processing method of an FPC sensing module in an embodiment of the present invention;

[0067] An embodiment of the present invention provides an anti-interference signal processing method for an FPC sensing module, including the following steps:

[0068] Step S1, collect the original signal of the target FPC sensing module through a preset signal sensor, and perform time-frequency domain conversion on the original signal through a preset Morlet wavelet transform technology to obtain a signal time-frequency spectrum diagram.

[0069] Specifically, when implementing the step of "collecting the original signals of the target FPC sensing module through a preset signal sensor and performing time-frequency domain conversion on the original signals through a preset Morlet wavelet transform technology to obtain a signal time-frequency spectrogram", it is first necessary to use the preset signal sensor to collect signals from the FPC sensing module. These sensors may be sensing elements integrated on the FPC module, specifically designed to capture and record various electrical signals generated by the module during operation. Suppose in an industrial automation scenario, the FPC sensing module is used to monitor the vibration of machine equipment. Then the sensor will capture the electrical signals generated by the machine vibration. Once the original signals are collected, the preset Morlet wavelet transform technology will be used to process these signals next. The Morlet wavelet transform is a time-frequency analysis method that can convert signals in the time domain to the time-frequency domain, thereby obtaining the spectral information of the signal changing with time. Specifically in the operation, the Morlet wavelet transform performs a convolution operation on the signal by using the Morlet wavelet basis function with high time resolution and frequency resolution. In this way, a corresponding time-frequency spectrogram can be obtained for each time point of the signal. Through this process, the transient characteristics, frequency components in the original signal, and the changes of these components with time can be clearly presented. For example, in the above industrial automation scenario, the machine equipment may generate abnormal vibrations due to mechanical failures or external interferences during operation. Through the Morlet wavelet transform, we can identify the frequencies and time points corresponding to these abnormal vibrations in the signal time-frequency spectrogram. Suppose a high-frequency vibration occurs in the machine equipment within a certain period of time. This will be shown as a prominent high-frequency energy peak in the signal time-frequency spectrogram, and the time position corresponds to the time point when the vibration occurs. Through this step, we can not only accurately locate the moment when the vibration occurs, but also analyze its frequency characteristics, which helps in further fault diagnosis and maintenance plan formulation. Thus, it can be seen that the Morlet wavelet transform provides a powerful tool for the time-frequency analysis of signals, ensuring the effectiveness of the anti-interference signal processing method of the FPC sensing module in practical applications.

[0070] Step S2, identify whether there is an abnormal interference area in the signal time-frequency spectrogram through a preset abnormal recognition model. If so, mark the abnormal interference area; where the abnormal interference area represents interference signals.

[0071] Specifically, in the step of "identifying whether there is an abnormal interference region in the time-frequency spectrum diagram of the signal through a preset abnormal recognition model, and if so, marking the abnormal interference region; wherein, the abnormal interference region represents an interference signal", it is first necessary to establish an abnormal recognition model. This model is usually based on machine learning or deep learning algorithms and learns the feature differences between normal working signals and interference signals in the time-frequency spectrum diagram of the signal through training data. In an industrial automation scenario, assuming that the FPC sensing module is used to monitor the vibration of machine equipment, the abnormal recognition model can identify the feature of the vibration signal under normal operating conditions by learning historical data. Once the time-frequency spectrum diagram of the signal is generated, the abnormal recognition model will analyze it. The model will scan the entire time-frequency spectrum diagram to find regions that do not match the previously learned normal signal features, and these regions may represent interference signals. Assuming that when the machine equipment is working normally, its vibration signal is mainly concentrated in a specific frequency range, but due to external interference or internal failure, high-frequency or irregular vibrations occur, which will be manifested as obvious abnormal regions on the time-frequency spectrum diagram. The recognition model determines whether it is an abnormal interference region by comparing the differences between these regions and the normal signal features. For example, in practical applications, if a sudden high-frequency vibration occurs in the machine equipment at a certain time point, the abnormal recognition model will identify this high-frequency vibration region as an abnormal interference region by comparing the frequency features of this region with the normal signal features in the historical data. The model will automatically mark these regions on the time-frequency spectrum diagram of the signal, usually highlighting these abnormal regions through colors or other visual marks for subsequent analysis and processing. Through this step, the FPC sensing module can not only identify abnormal interference signals, but also provide accurate positioning and marking for subsequent feature analysis and classification, ensuring the effectiveness and pertinence of interference signal processing.

[0072] Step S3, based on the time-frequency spectrum diagram of the signal, perform feature analysis and classification on the marked abnormal interference regions through a preset spectral clustering algorithm to obtain the types of interference signals.

[0073] Specifically, in the step of "performing feature analysis and classification on the labeled abnormal interference region based on the signal time-frequency spectrogram through a preset spectral clustering algorithm to obtain the interference signal type", it is first necessary to preset a spectral clustering algorithm. Spectral clustering is a clustering method based on graph theory, which is particularly suitable for processing data with non-linear structures. In an industrial automation scenario, assume that the FPC sensing module is used to monitor the vibration of machine equipment, and the abnormal recognition model has labeled the abnormal interference regions in the signal time-frequency spectrogram. Next, the spectral clustering algorithm will perform further feature extraction and classification based on these marked regions. The spectral clustering algorithm represents each abnormal region in the signal time-frequency spectrogram by constructing a graph, where each region is regarded as a node, and the connection weight between nodes represents their similarity or distance. By performing spectral decomposition on this graph, the algorithm can find an optimal cutting method to divide the nodes into different clusters. Assume that the machine equipment may be affected by different types of interference during operation, such as electromagnetic interference (EMI) or mechanical vibration interference, which will be manifested as different frequency and time-domain characteristics on the time-frequency spectrogram. The spectral clustering algorithm can divide the abnormal interference regions into different categories by analyzing the similarities and differences of these characteristics, and each category corresponds to a type of interference signal. For example, if an irregular low-frequency vibration and a high-frequency electromagnetic interference occur during the operation of the machine equipment, the spectral clustering algorithm will classify them into different clusters according to their characteristic differences on the time-frequency spectrogram. The low-frequency vibration may be identified as interference caused by mechanical faults, while the high-frequency electromagnetic interference may be generated by electrical equipment or nearby electronic devices. Through this classification, the FPC sensing module can not only identify the presence of interference signals but also accurately distinguish the types of interference, providing an accurate basis for subsequent interference processing and prevention. This method not only improves the intelligent level of signal processing but also provides strong data support for equipment maintenance and optimization.

[0074] Step S4: Predict the interference source distribution of the target FPC sensing module based on the interference signal type and the interference signal corresponding to the interference signal type, and obtain the interference source feature mapping matrix.

[0075] Specifically, in the step of "predicting the interference source distribution for the target FPC sensing module based on the interference signal type and the corresponding interference signal of the interference signal type to obtain the interference source feature mapping matrix", it is first necessary to use the interference signal types obtained by the spectral clustering algorithm before for prediction. Assume that in an industrial automation scenario, the FPC sensing module is used to monitor the vibration of machine equipment, and the spectral clustering algorithm has classified different types of interference signals. Next, based on these classification results, we will predict the distribution of the interference source on the FPC sensing module. The prediction process usually involves the spatial positioning and intensity analysis of the interference signal. By analyzing the characteristics of different types of interference signals, such as frequency, amplitude, duration, etc., the location and influence range of the interference source can be inferred. For example, if the spectral clustering algorithm identifies a low-frequency mechanical vibration interference, then this interference may come from a certain component of the machine equipment, such as a bearing or a gear, and the positions and operating states of these components can be considered during prediction. Similarly, if a high-frequency electromagnetic interference is identified, this interference may come from nearby electrical equipment or electromagnetic field sources, and the position and distance of the FPC sensing module relative to these interference sources need to be considered during prediction. Based on these analyses, an interference source feature mapping matrix is generated. This matrix not only records the distribution of different types of interference signals on the FPC sensing module, but also includes the characteristics of each interference source, such as the frequency, amplitude, propagation path, etc. of the interference signal. For example, during the operation of the machine equipment, if a high-frequency electromagnetic interference is identified, the interference source may be located in the upper left corner of the FPC sensing module, the frequency of the interference signal is 100 MHz, and the amplitude is 10 dBm. These information will be recorded in the interference source feature mapping matrix, and the possible influence area of the interference source will be marked. In this way, the FPC sensing module can not only identify the type of the interference signal, but also predict the distribution of the interference source, providing precise guidance for subsequent anti-interference processing. This not only improves the accuracy of signal processing, but also provides strong data support for the optimization and maintenance of the equipment.

[0076] Step S5, through a preset signal processing technology, perform anti-interference processing on the interference signal in the original signal based on the interference source feature mapping matrix to obtain the target working signal.

[0077] Specifically, in the step of "performing anti-interference processing on the interference signals in the original signal based on the interference source feature mapping matrix through a preset signal processing technique to obtain the target working signal", it is first necessary to perform signal processing using the previously generated interference source feature mapping matrix. In an industrial automation scenario, assume that the FPC sensing module is used to monitor the vibration of machine equipment. Through spectral clustering algorithms and interference source distribution prediction, the types of interference signals and their distribution information on the FPC sensing module have been obtained. The signal processing technique plays a key role here. Its purpose is to remove or weaken the interference signals from the original signal and retain or enhance the target working signal. Based on the interference source feature mapping matrix, the signal processing algorithm can accurately locate the position and characteristics of the interference signals, and thus adopt targeted processing methods. For example, if the interference signal is high-frequency electromagnetic interference, the signal processing technique may use filters or adaptive noise cancellation algorithms to suppress these high-frequency components. Assume that during the operation of machine equipment, a low-frequency mechanical vibration interference is identified. The signal processing algorithm may use frequency-domain filtering techniques to suppress the signals in this frequency range. Specifically, the signal processing technique may include, but is not limited to: time-domain filtering, frequency-domain filtering, adaptive filtering, waveform matching, blind source separation, etc. Assume that in an industrial automation scenario, the original signal monitored by the FPC sensing module contains the vibration signals generated by the normal operation of machine equipment and the electromagnetic interference caused by nearby electrical equipment. The signal processing algorithm will identify the characteristics of the electromagnetic interference on the time-frequency spectrogram based on the interference source feature mapping matrix, and use appropriate filtering techniques to filter out or weaken these interference signals, thereby obtaining a target working signal that is closer to the actual machine vibration situation. For example, during the processing, if a sudden electromagnetic interference is identified, the signal processing algorithm may remove the signal components in the interference frequency band through frequency-domain cutting methods to ensure the quality of the target working signal. In this way, the FPC sensing module can effectively suppress interference signals, improve the reliability and stability of the signals, and provide accurate data support for the normal operation and maintenance of machine equipment.

[0078] In a specific embodiment, the performing time-frequency domain conversion on the original signal through a preset Morlet wavelet transform technique to obtain a signal time-frequency spectrogram includes:

[0079] Performing scale discretization processing on the original signal to obtain different scale discretized signals; wherein, the original signal includes an original electrical signal and an original communication signal;

[0080] Performing time-frequency domain conversion and wavelet transform on different scale discretized signals through a preset Morlet wavelet transform technique to obtain a series of wavelet signals; wherein, the series of wavelet signals represents the energy distribution of the original signal under different frequency bands and time windows;

[0081] Perform dimensionality reduction processing on the wavelet signal to obtain the signal after dimensionality reduction;

[0082] Perform spatial interpolation on the signal after dimensionality reduction through a preset bilinear interpolation algorithm to obtain the interpolated signal;

[0083] Adopt a preset gray-scale mapping algorithm to convert the interpolated signal into a signal time-frequency spectrum diagram.

[0084] Specifically, in the step of "performing time-frequency domain conversion on the original signal through a preset Morlet wavelet transform technology to obtain a signal time-frequency spectrogram", it is first necessary to perform scale discretization processing on the original signal. In an industrial automation scenario, assuming that the FPC sensing module is used to monitor the vibration of machine equipment, the original signal may include electrical signals and communication signals generated during machine operation. The purpose of scale discretization processing is to convert a continuous time signal into a series of discrete scale signals. Specifically, the original signal is sampled according to a preset scale. For example, different time window lengths or sampling frequencies are selected to obtain discrete signals of different scales. Assuming that when the machine equipment is operating normally, the frequency components of the vibration signal are mainly concentrated in the low frequency band, while when a fault occurs, high frequency components may appear. Through scale discretization, it is possible to better capture the signal characteristics in these different frequency ranges. Next, wavelet transform and time-frequency domain conversion are performed on these scale-discretized signals through a preset Morlet wavelet transform technology. The Morlet wavelet transform is a time-frequency analysis method that uses a special wavelet basis function (i.e., the Morlet wavelet) to perform convolution operations on the signal to obtain a series of wavelet signals. These wavelet signals represent the energy distribution of the original signal in different frequency bands and time windows. For example, during the operation of machine equipment, if a sudden high-frequency vibration occurs, the Morlet wavelet transform can clearly display the duration and frequency position of this high-frequency vibration on the time-frequency spectrogram, providing a basis for subsequent anomaly recognition. After obtaining the wavelet signals, it is necessary to perform dimensionality reduction processing on these signals. The purpose of dimensionality reduction is to simplify the data structure, reduce computational complexity, and at the same time retain key information. In an industrial automation scenario, the FPC sensing module may generate a large amount of vibration data. Dimensionality reduction processing can compress this data from a high-dimensional space to a low-dimensional space. For example, the main components in the wavelet signals can be extracted through principal component analysis (PCA) or other dimensionality reduction algorithms to reduce redundant information and ensure the efficiency and accuracy of subsequent processing. After dimensionality reduction processing, a preset bilinear interpolation algorithm is used to perform spatial interpolation on the dimensionality-reduced signal. Bilinear interpolation is an interpolation technique that interpolates the signal in a two-dimensional space to generate a more detailed interpolated signal. In the application of the FPC sensing module, the interpolation technique can help restore the details of the signal on the time-frequency spectrogram, making the boundaries of the abnormal interference area clearer. For example, if a short-term high-frequency vibration occurs in the machine equipment at a certain time point, through interpolation processing, it is possible to more accurately locate the time and frequency range where this vibration occurs. Finally, through a preset gray-scale mapping algorithm, the interpolated signal is converted into a signal time-frequency spectrogram. The gray-scale mapping algorithm maps the energy distribution of the signal to gray levels to generate an intuitive image representation.In an industrial automation scenario, assuming that an irregular vibration occurs during the operation of a machine device, the gray-scale mapping algorithm can display this vibration as different gray-scale regions on the time-frequency spectrum diagram. The brightness represents the energy intensity of the vibration, and the color change represents the frequency change. In this way, the operator can visually observe the vibration conditions of the machine device at different time points, identify the characteristics of the abnormal interference region, and provide visual assistance for subsequent fault diagnosis and maintenance. For example, in practical applications, if a high-frequency vibration occurs in a machine device during a certain period, after Morlet wavelet transform, dimensionality reduction processing, and interpolation, a high-brightness region will be displayed on the generated signal time-frequency spectrum diagram, representing the time and frequency positions of this high-frequency vibration. Through this processing, the FPC sensing module can not only identify abnormal interference signals but also provide accurate positioning and marking for subsequent feature analysis and classification, ensuring the effectiveness and pertinence of interference signal processing.

[0085] In a specific embodiment, identifying whether there is an abnormal interference region in the signal time-frequency spectrum diagram through a preset abnormal recognition model, and if so, marking the abnormal interference region, includes:

[0086] Performing local binary pattern feature extraction on the signal time-frequency spectrum diagram to obtain an LBP feature vector sequence;

[0087] Based on a preset sliding window, segmenting the LBP feature vector sequence in time to obtain multiple LBP feature vector subsequences; wherein, each LBP feature vector subsequence represents a local region of the signal time-frequency spectrum diagram;

[0088] Performing singular value decomposition on each LBP feature vector subsequence to obtain a corresponding singular value vector; wherein, the singular value vector is used to reflect the energy distribution characteristics of the LBP feature vector subsequence;

[0089] Through a preset abnormal recognition model, identifying whether there is an abnormal interference region in the signal time-frequency spectrum diagram based on the singular value vector;

[0090] If so, calculating the degree of abnormality of the existing abnormal interference region based on a preset degree-of-abnormality evaluation algorithm to obtain a corresponding abnormality score; wherein, the higher the abnormality score, the greater the possibility of interference signals existing in the local region of the signal time-frequency spectrum diagram;

[0091] Combining and organizing the abnormal interference regions in the signal time-frequency spectrum diagram with an abnormality score greater than a preset threshold, and marking the abnormal interference regions after region combination and organization, wherein the region combination and organization include removing overlapping regions and merging adjacent regions, and the abnormal interference region includes the start time, duration, and frequency range of the interference signal.

[0092] Specifically, in the step of "identifying whether there is an abnormal interference region in the time-frequency spectrum diagram of the signal through a preset abnormal recognition model, and if so, marking the abnormal interference region", it is first necessary to extract local binary pattern (LBP) features from the time-frequency spectrum diagram of the signal. In an industrial automation scenario, assume that the FPC sensing module is used to monitor the vibration of machine equipment, and the time-frequency spectrum diagram of the signal shows the vibration frequency and energy distribution of the machine equipment at different time periods. LBP feature extraction is an image processing technology that generates a series of feature vector sequences by comparing the value of each pixel point with the values of its surrounding neighbor pixel points. These feature vectors can capture the texture information of local regions in the time-frequency spectrum diagram of the signal. For example, if an irregular vibration occurs during the operation of the machine equipment, LBP feature extraction can identify the characteristic differences of this vibration on the time-frequency spectrum diagram. Next, time segmentation is performed on these LBP feature vector sequences based on a preset sliding window. The sliding window is a method of moving average or segmentation. By sliding a fixed-size time window over the time-frequency spectrum diagram of the signal, multiple LBP feature vector subsequences are obtained. Each subsequence represents a local region of the time-frequency spectrum diagram of the signal, allowing for a detailed analysis of different time periods of the signal. For example, during the operation of the machine equipment, the sliding window can capture the vibration characteristics at different time points, thereby identifying possible abnormal vibration regions. Singular value decomposition (SVD) is performed on each LBP feature vector subsequence to obtain the corresponding singular value vector. Singular value decomposition is a matrix decomposition method used to reflect the energy distribution characteristics of the LBP feature vector subsequence. Through SVD, the local features of the signal can be compressed into several main singular values, which reflect the degree of concentration of the signal energy. For example, if the vibration signal during a certain time period shows a high-energy concentration area on the time-frequency spectrum diagram, then through singular value decomposition, a high singular value can be obtained, indicating that there may be abnormal interference in this area. Through a preset abnormal recognition model, based on these singular value vectors, it is identified whether there is an abnormal interference region in the time-frequency spectrum diagram of the signal. The abnormal recognition model may be a machine learning-based classifier that learns the characteristic differences between normal working signals and interference signals through training data. For example, when the machine equipment is operating normally, its vibration signal is mainly concentrated in the low-frequency band, while when a fault occurs or interference is received, high-frequency components will appear in the signal. The abnormal recognition model can identify these abnormal characteristics by comparing the singular value vectors. If an abnormal interference region is identified, it is necessary to calculate the degree of abnormality of these regions based on a preset abnormal degree evaluation algorithm to obtain the corresponding abnormal score. The higher the abnormal score, the greater the possibility of interference signals existing in this local region. For example, during the operation of the machine equipment, if a sudden high-frequency vibration is identified, the abnormal degree evaluation algorithm can calculate a high abnormal score based on its duration, frequency range and other characteristics, indicating that this area needs special attention.Finally, combine and organize the abnormal interference regions with abnormal scores greater than the preset threshold, and label the abnormal interference regions after combination and organization. The combination and organization of regions include removing overlapping regions and merging adjacent regions to ensure that the identified abnormal regions are accurate and coherent. When labeling the abnormal interference regions, it is necessary to record the start time, duration, and frequency range of the interference signal. For example, in an industrial automation scenario, if it is identified that a machine device continuously exhibits abnormal vibrations within a certain period of time, this information will be marked on the signal time-frequency spectrum diagram and provided to the operator or maintenance personnel as a basis for fault diagnosis and maintenance. Through these steps, the FPC sensing module can not only identify abnormal interference signals, but also accurately locate and mark the characteristics of these signals, providing a clear target for subsequent interference processing. This method not only improves the intelligent level of signal processing, but also provides strong data support for the maintenance and optimization of equipment.

[0093] In a specific embodiment, through the preset spectral clustering algorithm, based on the signal time-frequency spectrum diagram, perform feature analysis and classification on the labeled abnormal interference regions to obtain the types of interference signals, including:

[0094] Input the labeled abnormal interference regions into a preset region-signal mapping algorithm to obtain the interference signals corresponding to the labeled abnormal interference regions;

[0095] Through high-order spectral estimation technology, extract the energy distribution characteristics of the interference signal in different frequency bands on the signal time-frequency spectrum diagram;

[0096] Based on the energy distribution characteristics, construct a spectral feature map;

[0097] In the spectral feature map, use the preset spectral clustering algorithm to analyze the feature similarity of each interference signal to obtain a similar feature distribution result; wherein, the similar feature distribution result includes signal frequency aggregation degree, energy density distribution, and band correlation coefficient;

[0098] Perform dimensionality reduction processing on the similar feature distribution result to obtain a dimensionality-reduced similar feature distribution result;

[0099] Input the dimensionality-reduced similar feature distribution result into a preset support vector machine for classification to obtain the types of interference signals.

[0100] Specifically, in the step of "performing feature analysis and classification on the labeled abnormal interference regions based on the signal time-frequency spectrogram through a preset spectral clustering algorithm to obtain the types of interference signals", it is first necessary to input the labeled abnormal interference regions into a preset region-signal mapping algorithm to obtain the interference signals corresponding to these regions. In an industrial automation scenario, assume that the FPC sensing module is used to monitor the vibration of machine equipment, and the abnormal recognition model has labeled the abnormal interference regions on the signal time-frequency spectrogram. The region-signal mapping algorithm analyzes the time-domain and frequency-domain characteristics of these regions and converts them into corresponding interference signals. For example, if a high-frequency vibration occurs in a machine equipment at a certain time point, through the region-signal mapping algorithm, the specific manifestation of this high-frequency vibration on the time-frequency spectrogram can be extracted. Next, the high-order spectrum estimation technology is used to extract the energy distribution characteristics of these interference signals in different frequency bands. High-order spectrum estimation is a signal processing technology used to analyze the non-linear characteristics of signals. It can identify the energy concentration degree of interference signals in different frequency bands. For example, during the operation of machine equipment, if a low-frequency mechanical vibration interference occurs, the high-order spectrum estimation technology can analyze the energy distribution of this vibration in the low-frequency band, providing a basis for subsequent feature analysis. Based on these energy distribution characteristics, a spectral feature map is constructed. The spectral feature map visually shows the energy distribution of different frequency bands. For example, if the machine equipment is subject to electromagnetic interference during operation, this interference may show high-energy peaks in the high-frequency band, and the spectral feature map will clearly display these features, providing an intuitive reference for subsequent analysis and classification. In the spectral feature map, a preset spectral clustering algorithm is used to analyze the feature similarity of each interference signal. Spectral clustering is a clustering method based on graph theory, suitable for processing data with non-linear structures. It constructs a graph, regards each interference signal as a node, and the connection weight between nodes represents their similarity or distance. For example, in an industrial automation scenario, if the interference signals received by the machine equipment are mainly of two types, one is low-frequency mechanical vibration interference and the other is high-frequency electromagnetic interference, the spectral clustering algorithm will classify them into different clusters according to the feature differences of these signals on the spectral feature map. Through analysis, a similar feature distribution result is obtained. The similar feature distribution result includes signal frequency aggregation degree, energy density distribution, and frequency band correlation coefficient. For example, low-frequency vibration interference may show a higher frequency aggregation degree and an energy density distribution in the low-frequency band, while high-frequency electromagnetic interference has obvious energy aggregation and frequency band correlation in the high-frequency band. Next, dimensionality reduction processing is performed on these similar feature distribution results. The purpose of dimensionality reduction is to simplify the data structure, reduce the computational complexity, and at the same time retain key information. For example, through principal component analysis (PCA) or other dimensionality reduction algorithms, the main components in the similar feature distribution results can be extracted to ensure the efficiency and accuracy of subsequent processing. The dimensionality-reduced similar feature distribution results are input into a preset support vector machine (SVM) for classification.Support vector machine is a powerful classification algorithm that, by learning the features of data, finds the optimal decision boundary to separate different types of interference signals. For example, during the operation of a machine device, if a sudden electromagnetic interference is identified, the support vector machine can classify it as an electromagnetic interference type based on its features in the spectral feature map. Through this process, the FPC sensing module can not only identify abnormal interference signals but also accurately distinguish the types of interference. For example, if a low-frequency mechanical vibration interference and a high-frequency electromagnetic interference occur during the operation of a machine device, through spectral clustering algorithm and support vector machine classification, the FPC sensing module can classify these interference signals as mechanical fault interference and electromagnetic interference respectively, providing an accurate basis for subsequent interference processing and prevention. This method not only improves the intelligent level of signal processing but also provides strong data support for equipment maintenance and optimization.

[0101] In a specific embodiment, the method of predicting the interference source distribution of the target FPC sensing module based on the interference signal type and the interference model corresponding to the interference signal type to obtain an interference source feature mapping matrix includes:

[0102] Extract the interference signal corresponding to the interference signal type;

[0103] Perform high-order Volterra series analysis on the interference signal corresponding to the interference signal type to obtain a non-linear time-varying interference feature sequence; wherein, the non-linear time-varying interference feature sequence includes amplitude feature, phase feature, and frequency feature;

[0104] Obtain the physical layout distribution map corresponding to the target FPC sensing module, and through non-parametric kernel density estimation method, perform probability distribution analysis on the physical layout distribution map based on the non-linear time-varying interference feature sequence to obtain the interference source distribution space density; wherein, the interference source distribution space density includes the position probability and intensity probability of the interference source;

[0105] Input the interference source distribution space density into a preset deep belief network for interference source prediction to obtain a predicted interference source;

[0106] Adopt an adaptive Kalman particle filter algorithm to perform dynamic trajectory prediction on the predicted interference source to obtain the interference source propagation trajectory feature; wherein, it includes the motion direction feature and speed feature of the interference source;

[0107] Perform spatio-temporal correlation analysis on the interference source propagation trajectory feature to obtain the interference source influence range;

[0108] Obtain an interference source feature mapping matrix based on the influence range of the interference source and the type of the interference signal; wherein, the interference source feature mapping matrix includes rows, columns and elements, the rows represent the types of interference sources, the columns represent the physical positions of the FPC sensing modules, and the elements represent the feature intensities of the interference signals.

[0109] Specifically, in the step of "predicting the interference source distribution for the target FPC sensing module based on the interference signal type and the corresponding interference model number of the interference signal type to obtain the interference source feature mapping matrix", it is first necessary to extract the interference signals corresponding to the interference signal type. In an industrial automation scenario, assuming that the FPC sensing module is used to monitor the vibration of machine equipment, different types of interference signals have been identified through spectral clustering algorithms and support vector machine classification. After extracting these signals, further in-depth analysis is required. Perform high-order Volterra series analysis on the interference signals corresponding to the interference signal type to obtain the non-linear time-varying interference feature sequence. The high-order Volterra series is a non-linear system analysis method that can describe the dynamic behavior of signals in a non-linear system. It extracts the amplitude characteristics, phase characteristics, and frequency characteristics of the interference signal through a series of convolution integrals. For example, if a machine equipment is subjected to a low-frequency mechanical vibration interference during operation, the Volterra series analysis can reveal the non-linear changes of this vibration signal at different time points. Then, obtain the physical layout distribution map corresponding to the target FPC sensing module, and perform probability distribution analysis on the physical layout distribution map based on the non-linear time-varying interference feature sequence through the non-parametric kernel density estimation method. The non-parametric kernel density estimation is a statistical method used to estimate the probability density function of data. It calculates the probability density at each position by placing kernel functions around the data points, thereby obtaining the distribution space density of the interference source on the FPC sensing module. For example, during the operation of machine equipment, if an electromagnetic interference is identified, the non-parametric kernel density estimation can help predict the possible location and intensity distribution of this interference source. Input the interference source distribution space density into a preset deep belief network for interference source prediction. The deep belief network is a deep learning model that is good at processing complex non-linear data. By learning the characteristics of the interference source distribution space density, the deep belief network can predict the location and intensity of the interference source. For example, if a machine equipment is interfered with during a certain period of time, the deep belief network can identify that the interference source may come from a specific location of the equipment. Use the adaptive Kalman particle filtering algorithm to perform dynamic trajectory prediction on the predicted interference source. The Kalman particle filtering is a fusion filtering technology that combines the advantages of Kalman filtering and particle filtering and is suitable for dynamic trajectory prediction of non-linear and non-Gaussian systems. It provides a dynamic prediction model by tracking the motion direction characteristics and speed characteristics of the interference source. For example, during the operation of machine equipment, if an interference signal is identified as moving, the adaptive Kalman particle filtering algorithm can predict its future motion trajectory. Perform spatio-temporal correlation analysis on the interference source propagation trajectory characteristics to obtain the influence range of the interference source. The spatio-temporal correlation analysis analyzes the possible physical regions affected by considering the motion trajectory and time distribution of the interference source.For example, if the interference source appears on the left side of the FPC sensing module and moves to the right side, spatio-temporal correlation analysis can determine the area affected during the entire movement process. Finally, based on the influence range of the interference source and the type of interference signal, an interference source feature mapping matrix is obtained. The interference source feature mapping matrix is a two-dimensional matrix, where the rows represent the types of interference sources, the columns represent the physical positions of the FPC sensing module, and the elements represent the characteristic intensities of the interference signals. For example, if a machine device is subjected to low-frequency mechanical vibration interference and high-frequency electromagnetic interference during operation, the interference source feature mapping matrix will record the distribution and intensity of these interference signals on the FPC sensing module, providing an intuitive view to help understand the impact of the interference source on the device. Through this procedure, the FPC sensing module can not only identify abnormal interference signals but also predict the distribution, dynamic trajectory, and influence range of the interference source, providing an accurate basis for subsequent interference processing and prevention. This method not only improves the intelligent level of signal processing but also provides strong data support for the maintenance and optimization of the device.

[0110] In a specific embodiment, the probability distribution analysis of the physical layout distribution map based on the non-linear time-varying interference feature sequence to obtain the interference source distribution space density includes:

[0111] Feature extraction is performed on the non-linear time-varying interference feature sequence to obtain non-linear dynamic features; wherein, the non-linear dynamic features include the non-linear coefficient, time constant, and frequency response characteristics of the system;

[0112] Multi-scale entropy analysis is performed on the non-linear dynamic features to obtain the complexity features of the interference signal; wherein, the complexity features represent the complexity and regularity of the interference signal on different time scales;

[0113] Through a preset spectrum envelope estimation algorithm, spectrum envelope analysis is performed on the physical layout distribution map based on the non-linear dynamic features and the complexity features to obtain the spectrum envelope distribution;

[0114] The probability distribution analysis of the physical layout distribution map is performed using the spectrum envelope distribution to obtain the interference source distribution space density.

[0115] Specifically, in the step of "performing a probability distribution analysis on the physical layout distribution diagram based on the nonlinear time-varying interference feature sequence to obtain the spatial density of the interference source distribution", it is first necessary to extract features from the nonlinear time-varying interference feature sequence. The nonlinear time-varying interference feature sequence has obtained the amplitude characteristics, phase characteristics and frequency characteristics of the interference signal through high-order Volterra series analysis, and these characteristics reflect the changes of the interference signal at different time points. For example, in an industrial automation scenario, if the FPC sensor module detects that the machine equipment is subjected to a low-frequency mechanical vibration interference during operation, the nonlinear coefficient, time constant and frequency response characteristics of the vibration signal at different time points can be obtained through feature extraction. Then, these nonlinear dynamic features are subjected to multi-scale entropy analysis. Multi-scale entropy analysis is a nonlinear signal processing method used to evaluate the complexity and regularity of signals. Through this analysis, the complexity characteristics of the interference signal at different time scales can be obtained. For example, if the machine equipment is subjected to a high-frequency electromagnetic interference during operation, the multi-scale entropy analysis can reveal the complexity changes of the interference signal at different time scales and help understand the essential characteristics of the interference. Through the preset spectrum envelope estimation algorithm, the spectrum envelope analysis of the physical layout distribution map is performed based on nonlinear dynamic characteristics and complexity characteristics. Spectrum envelope analysis is a signal processing technology that can reveal the distribution of the spectrum characteristics of the signal by performing envelope processing on the spectrum of the signal. For example, during the operation of the machine equipment, if a sudden vibration interference is identified, the spectrum envelope analysis can display the envelope shape of the interference signal on the spectrum, providing an intuitive view to help locate the interference source. The probability distribution analysis of the physical layout distribution map is performed using the spectrum envelope distribution to obtain the spatial density of the interference source distribution. The probability distribution analysis maps the spectrum envelope distribution to the physical layout distribution map through the non-parametric kernel density estimation method, and calculates the location probability and intensity probability of the interference source on the FPC sensor module. For example, if the machine equipment is subjected to a low-frequency mechanical vibration interference during operation, the spectrum envelope analysis can identify that the interference is more obvious in a specific area of ​​the equipment, and the non-parametric kernel density estimation can further calculate the distribution of the interference source on the FPC sensor module. Through this process, the FPC sensor module can not only identify abnormal interference signals, but also predict the distribution of interference sources on the equipment. For example, if a machine is subject to a sudden high-frequency electromagnetic interference during operation, the FPC sensor module can identify the nonlinear characteristics and complexity characteristics of the interference signal through feature extraction, multi-scale entropy analysis and spectrum envelope analysis, and then determine the possible location and intensity distribution of the interference source through spectrum envelope distribution and probability distribution analysis, providing an accurate basis for subsequent interference processing and prevention. This method not only improves the intelligence level of signal processing, but also provides strong data support for equipment maintenance and optimization.For example, in an industrial automation scenario, the FPC sensing module can identify abnormal vibrations during the operation of machinery and equipment, and through a series of non-linear signal processing techniques, provide reliable vibration monitoring data, ensuring the operational stability and safety of the equipment, reducing false alarms or missed alarms caused by interference signals, and enhancing the intelligent monitoring ability of the equipment.

[0116] In a specific embodiment, the anti-interference processing of the interference signal based on the interference source feature mapping matrix through a preset signal processing technique to obtain a target working signal includes:

[0117] Performing spatial correlation analysis on the interference signal based on the interference source feature mapping matrix to obtain the position corresponding to the interference signal;

[0118] At the position, performing notch filtering on the interference signal through a preset adaptive notch filter bank and a preset signal processing technique to obtain a notch-filtered signal;

[0119] Extracting a preset signal from the notch-filtered signal through a preset signal subspace projection algorithm, and separating the preset signal from the notch-filtered signal to obtain a preliminary working signal and a remaining interference signal; wherein, the preset signal is the signal required by the FPC sensing module during operation;

[0120] Performing signal enhancement on the remaining interference signal to obtain an enhanced signal;

[0121] Performing frequency-selective processing on the enhanced signal to obtain a selective filtering scheme;

[0122] Using the selective filtering scheme to perform anti-interference signal processing on the remaining interference signal to reduce the contamination of the remaining interference signal to the preliminary working signal and obtain a target working signal.

[0123] Specifically, in the step of "performing anti-interference processing on the interference signal based on the interference source feature mapping matrix through a preset signal processing technology to obtain the target working signal", it is first necessary to perform spatial correlation analysis on the interference signal based on the interference source feature mapping matrix. The interference source feature mapping matrix has recorded the distribution of different types of interference signals on the FPC sensing module. Through spatial correlation analysis, the sources of these interference signals can be accurately located. For example, in an industrial automation scenario, if the FPC sensing module detects a low-frequency mechanical vibration interference during the operation of a machine device, spatial correlation analysis can identify that this interference signal is mainly concentrated in a specific area of the device. After identifying the location of the interference signal, notch processing is performed on these interference signals through a preset adaptive notch filter bank and a preset signal processing technology. An adaptive notch filter is a dynamic signal processor that can automatically adjust the notch frequency according to the change of the input signal, thereby effectively suppressing interference signals within a specific frequency range. For example, if a high-frequency electromagnetic interference is identified, the adaptive notch filter can perform signal notch processing on this frequency band to generate a notch-processed signal. Next, a preset signal in the notch-processed signal is extracted using a preset signal subspace projection algorithm, and the preset signal is separated from the notch-processed signal to obtain a preliminary working signal and a remaining interference signal. In an industrial automation scenario, the preset signal refers to the vibration signal required for the normal operation of a machine device. Through the signal subspace projection algorithm, the interference signal and the normal working signal can be separated. For example, if the normal working signal of a machine device is mainly concentrated in the low-frequency band, while the interference signal is in the high-frequency band, then the signal subspace projection algorithm can extract the low-frequency band signal as the preliminary working signal and identify the remaining high-frequency signal as the remaining interference signal. Signal enhancement processing is performed on the remaining interference signal. Signal enhancement is a signal processing technology used to enhance the intensity of useful signals and reduce the influence of noise and interference. For example, through adaptive noise cancellation technology or frequency domain filtering, the useful components in the remaining interference signal can be enhanced to obtain an enhanced signal. Then, frequency selective processing is performed on the enhanced signal to obtain a selective filtering scheme. Frequency selective processing analyzes the spectral characteristics of the enhanced signal and selectively suppresses or enhances signal components within a specific frequency range. For example, if there are still some low-frequency mechanical vibration interferences in the enhanced signal, frequency selective processing can design a filtering scheme for these low-frequency interferences. Finally, anti-interference signal processing is performed on the remaining interference signal using the selective filtering scheme to reduce the contamination of the remaining interference signal to the preliminary working signal and obtain the target working signal. Through a carefully designed filtering scheme, the harmful components in the remaining interference signal will be weakened or removed, making the preliminary working signal purer.For example, if the remaining interference signal contains the noise of the surrounding environment of the machine equipment, the selective filtering scheme effectively suppresses these noise components, so that the finally obtained target working signal can reflect the true working state of the machine equipment as much as possible. Through this step, the FPC sensing module can not only identify abnormal interference signals, but also effectively process these interferences through a series of signal processing technologies, so as to obtain a target working signal closer to the actual vibration of the machine equipment. This method not only improves the intelligent level of signal processing, but also provides accurate data support for the normal operation and maintenance of the equipment. For example, in the industrial automation scenario, the FPC sensing module can identify abnormal vibrations during the operation of the machine equipment, and through anti-interference processing, provide reliable vibration monitoring data to ensure the operation stability and safety of the equipment.

[0124] The anti-interference signal processing method of the FPC sensing module in the embodiment of the present invention is described above. Next, the anti-interference signal processing device of the FPC sensing module in the embodiment of the present invention will be described. Please refer to Figure 2 , an embodiment of the anti-interference signal processing device of the FPC sensing module in the embodiment of the present invention includes:

[0125] A conversion module 21, configured to collect an original signal of a target FPC sensing module through a preset signal sensor, and perform time-frequency domain conversion on the original signal through a preset Morlet wavelet transform technology to obtain a signal time-frequency spectrogram;

[0126] An identification module 22, configured to identify whether there is an abnormal interference area in the signal time-frequency spectrogram through a preset abnormal identification model. If so, mark the abnormal interference area; wherein, the abnormal interference area represents an interference signal;

[0127] A classification module 23, configured to perform feature analysis and classification on the marked abnormal interference area based on the signal time-frequency spectrogram through a preset spectral clustering algorithm to obtain an interference signal type;

[0128] A prediction module 24, configured to perform interference source distribution prediction on the target FPC sensing module based on the interference signal type and the interference signal corresponding to the interference signal type to obtain an interference source feature mapping matrix;

[0129] An anti-interference module 25, configured to perform anti-interference processing on the interference signal in the original signal based on the interference source feature mapping matrix through a preset signal processing technology to obtain a target working signal.

[0130] In this embodiment, for the specific implementation of each unit in the above device embodiment, please refer to that described in the above method embodiment, and details are not described herein again.

[0131] Refer toFigure 3 , in an embodiment of the present invention, a computer device is further provided. The internal structure of the computer device may be as Figure 3 shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface, and a database connected through a system bus. Among them, the processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. The computer program, when executed by the processor, implements the above method.

[0132] Those skilled in the art can understand that Figure 3 the structure shown in

[0133] is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0134] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium provided by the present invention and used in the embodiments can include non-volatile and / or volatile memories. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or an external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0135] It should be noted that in this article, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, device, article or method. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, device, article or method including that element.

[0136] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structural or equivalent process transformation made by using the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall equally be included in the patent protection scope of the present invention.

Claims

1. An anti-interference signal processing method for an FPC sensor module, characterized in that: The following steps are involved: The original signal of the target FPC sensor module is collected by a preset signal sensor, and the original signal is converted into a time-frequency domain by a preset Morlet wavelet transform technology to obtain a signal time-frequency spectrum diagram; By presetting an abnormality recognition model, identifying whether there is an abnormal interference area in the spectrum diagram of the signal, and if so, marking the abnormal interference area; wherein the abnormal interference area represents an interference signal; By using a preset spectral clustering algorithm, characteristic analysis and classification are performed on the marked abnormal interference area based on the signal time-frequency spectrum diagram to obtain the interference signal type; Based on the interference signal type and the interference signal corresponding to the interference signal type, the interference source distribution of the target FPC sensor module is predicted to obtain an interference source feature mapping matrix; By using a preset signal processing technology, anti-interference processing is performed on the interference signal in the original signal based on the interference source feature mapping matrix to obtain a target working signal; The method of converting the original signal into a time-frequency domain by using a preset Morlet wavelet transform technology to obtain a signal time-frequency spectrum diagram includes: Performing scale discretization processing on the original signal to obtain discretized signals of different scales; wherein the original signal includes an original electrical signal and an original communication signal; Through the preset Morlet wavelet transform technology, the time-frequency domain conversion and wavelet transform are performed on the discretized signals of different scales to obtain a series of wavelet signals; wherein the series of wavelet signals represent the energy distribution of the original signal in different frequency bands and time windows; Performing dimensionality reduction processing on the wavelet signal to obtain a dimensionality reduced signal; Performing spatial interpolation on the dimension-reduced signal by a preset bilinear interpolation algorithm to obtain an interpolated signal; The interpolated signal is converted into a signal time-frequency spectrum diagram by using a preset grayscale mapping algorithm.

2. The anti-interference signal processing method of the FPC sensor module according to claim 1, characterized in that: The method of identifying whether there is an abnormal interference area in the signal spectrum diagram by using a preset abnormality identification model, and if so, marking the abnormal interference area, includes: Extracting local binary pattern features from the signal time-frequency spectrum to obtain an LBP feature vector sequence; Based on a preset sliding window, the LBP feature vector sequence is time segmented to obtain a plurality of LBP feature vector subsequences; wherein each of the LBP feature vector subsequences represents a local area of ​​the signal time-frequency spectrum diagram; Performing singular value decomposition on each of the LBP feature vector subsequences to obtain a corresponding singular value vector; wherein the singular value vector is used to reflect the energy distribution characteristics of the LBP feature vector subsequence; By presetting an abnormality recognition model, identifying whether there is an abnormal interference area in the spectrum graph of the signal based on the singular value vector; If it exists, the abnormality degree of the abnormal interference area is calculated based on a preset abnormality degree assessment algorithm to obtain a corresponding abnormality score; wherein the higher the abnormality score, the greater the possibility that an interference signal exists in the local area of ​​the signal spectrum diagram; The abnormal interference areas in the signal time-frequency spectrum diagram whose abnormality scores are greater than a preset threshold are combined and sorted, and the abnormal interference areas after the area combination and sorting are marked, wherein the area combination and sorting include removing overlapping areas and merging adjacent areas, and the abnormal interference areas include the start time, duration and frequency range of the interference signal.

3. The anti-interference signal processing method of the FPC sensor module according to claim 1, characterized in that: The preset spectral clustering algorithm is used to perform feature analysis and classification on the annotated abnormal interference area based on the signal time-frequency spectrum diagram to obtain the interference signal type, including: Inputting the annotated abnormal interference region into a preset region-signal mapping algorithm to obtain an interference signal corresponding to the annotated abnormal interference region; Extracting energy distribution characteristics of the interference signal in different frequency bands on the signal spectrum diagram through high-order spectrum estimation technology; Constructing a frequency spectrum feature graph based on the energy distribution feature; In the frequency spectrum feature graph, a preset spectrum clustering algorithm is used to analyze the feature similarity of each interference signal to obtain a similar feature distribution result; wherein the similar feature distribution result includes signal frequency concentration, energy density distribution and frequency band correlation coefficient; Performing dimensionality reduction processing on the similar feature distribution result to obtain a similar feature distribution result after dimensionality reduction; The similar feature distribution result after dimensionality reduction is input into a preset support vector machine for classification to obtain the interference signal type.

4. The anti-interference signal processing method of the FPC sensor module according to claim 1, characterized in that: The interference source distribution prediction of the target FPC sensor module based on the interference signal type and the interference signal corresponding to the interference signal type is performed to obtain an interference source feature mapping matrix, including: Extracting an interference signal corresponding to the interference signal type; Performing high-order Volterra series analysis on the interference signal corresponding to the interference signal type to obtain a nonlinear time-varying interference feature sequence; wherein the nonlinear time-varying interference feature sequence includes amplitude features, phase features and frequency features; Obtain a physical layout distribution map corresponding to the target FPC sensor module, and perform a probability distribution analysis on the physical layout distribution map based on the nonlinear time-varying interference feature sequence using a nonparametric kernel density estimation method to obtain a spatial density of interference source distribution; wherein the spatial density of interference source distribution includes a location probability and an intensity probability of the interference source; Inputting the interference source distribution spatial density into a preset deep belief network to predict the interference source, thereby obtaining a predicted interference source; Adopting an adaptive Kalman particle filter algorithm to perform dynamic trajectory prediction on the predicted interference source, and obtaining the propagation trajectory characteristics of the interference source, including the movement direction characteristics and speed characteristics of the interference source; Performing spatiotemporal correlation analysis on the propagation trajectory characteristics of the interference source to obtain the influence range of the interference source; An interference source characteristic mapping matrix is ​​obtained based on the interference source influence range and the interference signal type; wherein the interference source characteristic mapping matrix includes rows, columns and elements, the rows represent the interference source type, the columns represent the physical location of the FPC sensor module and the elements represent the characteristic intensity of the interference signal.

5. The anti-interference signal processing method of the FPC sensor module according to claim 4, characterized in that: The performing probability distribution analysis on the physical layout distribution diagram based on the nonlinear time-varying interference feature sequence to obtain the interference source distribution space density includes: Extracting the nonlinear time-varying interference feature sequence to obtain nonlinear dynamic features; wherein the nonlinear dynamic features include the nonlinear coefficient, time constant and frequency response features of the system; Performing multi-scale entropy analysis on the nonlinear dynamic characteristics to obtain complexity characteristics of the interference signal; wherein the complexity characteristics represent the complexity and regularity of the interference signal at different time scales; By using a preset spectrum envelope estimation algorithm, based on the nonlinear dynamic characteristics and the complexity characteristics, spectrum envelope analysis is performed on the physical layout distribution diagram to obtain spectrum envelope distribution; The frequency spectrum envelope distribution is used to perform probability distribution analysis on the physical layout distribution diagram to obtain the interference source distribution space density.

6. The anti-interference signal processing method of the FPC sensor module according to claim 1, characterized in that: The method of performing anti-interference processing on the interference signal based on the interference source feature mapping matrix by using a preset signal processing technology to obtain a target working signal includes: Performing spatial correlation analysis on the interference signal based on the interference source feature mapping matrix to obtain a position corresponding to the interference signal; At the position, performing notch processing on the interference signal by using a preset adaptive notch filter group and a preset signal processing technology to obtain a notch processed signal; By using a preset signal subspace projection algorithm, a preset signal in the notch processing signal is extracted, and the preset signal is separated from the notch processing signal to obtain a preliminary working signal and a residual interference signal; wherein the preset signal is a signal required by the FPC sensor module when working; Performing signal enhancement on the remaining interference signal to obtain an enhanced signal; Performing frequency selective processing on the enhanced signal to obtain a selective filtering scheme; The selective filtering scheme is used to perform anti-interference signal processing on the remaining interference signal to reduce the contamination of the preliminary working signal by the remaining interference signal, thereby obtaining a target working signal.

7. An anti-interference signal processing device for an FPC sensor module, characterized in that: include: A conversion module is used to collect the original signal of the target FPC sensor module through a preset signal sensor, and perform time-frequency domain conversion on the original signal through a preset Morlet wavelet transform technology to obtain a signal time-frequency spectrum diagram; An identification module, used to identify whether there is an abnormal interference area in the spectrum diagram of the signal through a preset abnormal identification model, and if so, mark the abnormal interference area; wherein the abnormal interference area represents an interference signal; A classification module, used to perform feature analysis and classification on the annotated abnormal interference area based on the signal time-frequency spectrum diagram by using a preset spectral clustering algorithm to obtain the interference signal type; A prediction module, used to predict the interference source distribution of the target FPC sensor module based on the interference signal type and the interference signal corresponding to the interference signal type, and obtain an interference source feature mapping matrix; An anti-interference module is used to perform anti-interference processing on the interference signal in the original signal based on the interference source feature mapping matrix through a preset signal processing technology to obtain a target working signal; The method of converting the original signal into a time-frequency domain by using a preset Morlet wavelet transform technology to obtain a signal time-frequency spectrum diagram includes: Performing scale discretization processing on the original signal to obtain discretized signals of different scales; wherein the original signal includes an original electrical signal and an original communication signal; Through the preset Morlet wavelet transform technology, the time-frequency domain conversion and wavelet transform are performed on the discretized signals of different scales to obtain a series of wavelet signals; wherein the series of wavelet signals represent the energy distribution of the original signal in different frequency bands and time windows; Performing dimensionality reduction processing on the wavelet signal to obtain a dimensionality reduced signal; Performing spatial interpolation on the dimension-reduced signal by a preset bilinear interpolation algorithm to obtain an interpolated signal; The interpolated signal is converted into a signal time-frequency spectrum diagram by using a preset grayscale mapping algorithm.

8. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images