Multi-dimensional electrical signal processing method and system

Abstract

The invention relates to the technical field of multi-signal processing, in particular to a multi-dimensional electrical signal processing method and system, and the method comprises the following steps: carrying out the synchronous collection of time, frequency and spatial features of electrical signals of differentiated target sources, and generating a preliminary time-frequency-space feature set through the measurement and comparison of features and the screening of key time-frequency-space data. According to the method, the information extraction efficiency is greatly improved in a complex signal environment through hierarchical analysis in combination with feature correlation revelation and optimization of the contrast and definition of data, voltage fluctuation and frequency anomaly can be rapidly identified based on matching and classification of known abnormal modes, efficient classification and processing of abnormal signals are achieved, and the method is suitable for popularization and application. The real-time performance and precision of signal monitoring and the dynamic adjustment and optimization processing of the signals are enhanced, the reliability and adaptability of the whole signal processing process are improved, challenges in a complex signal environment are effectively handled, the analysis capability and processing efficiency of the signals are remarkably improved, and the signal quality in practical application is enhanced.

Description

Technical Field

[0001] The present invention relates to the field of multi-signal processing technology, and in particular to a multi-dimensional electrical signal processing method and system. Background Art

[0002] The field of multi-signal processing technology involves processing and analyzing information extracted from multiple sources or types of electrical signals. It is usually used to improve signal quality, remove noise, classify signals and extract features, so that valuable data can be identified from complex signal environments. In practical applications, multi-signal processing can use analysis methods in multiple dimensions such as time, frequency, and space, and synchronously process and optimize signals through algorithms and mathematical models. It is widely used in communication systems, biomedical engineering, sound processing, image processing, and other fields that require signal integration and interpretation.

[0003] Among them, the multi-dimensional electrical signal processing method is a specific application in the field of multi-signal processing technology, which mainly focuses on how to effectively integrate and process electrical signals from different dimensions (such as time, frequency, and space) to improve the efficiency and accuracy of information extraction. The main uses include but are not limited to enhancing the clarity and reliability of communication signals, improving the accuracy of medical imaging technology, and improving the quality of audio and video content. Through multi-dimensional processing, signal behavior in complex environments can be analyzed more accurately, thus finding practical applications in various high-tech fields.

[0004] When processing electrical signals from multiple sources or types, existing technologies are often only able to analyze in a single dimension, lacking comprehensive processing of time, frequency and space information, resulting in insufficient integrity in signal feature extraction. In terms of data optimization, the lack of in-depth analysis of feature correlation makes it difficult for the contrast and clarity of information to achieve the expected results, especially in complex environments, where the accuracy of data extraction is limited. In addition, the anomaly identification and classification methods of existing technologies are relatively simple, lacking comprehensive capture and real-time processing of abnormal signals, affecting the timeliness and effectiveness of signal monitoring. In terms of dynamic signal adjustment and optimization, existing technologies are unable to fully cope with quality fluctuations and stability issues during transmission, which ultimately results in unsatisfactory signal processing results and is difficult to meet high-demand practical application scenarios. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a multi-dimensional electrical signal processing method and system.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a multi-dimensional electrical signal processing method, comprising the following steps:

[0007] S1: Synchronously collect the time, frequency and space characteristics of the electrical signals of the differentiated target source, screen the key time-frequency-space data by measuring and comparing the characteristics, and generate a preliminary time-frequency-space feature set;

[0008] S2: Based on the preliminary time-frequency-space feature set, perform data hierarchical analysis, calculate and analyze the relative importance of each feature, reveal the correlation between features based on the analysis results, select and adjust the contrast and clarity of the data, and obtain adjusted feature data;

[0009] S3: Based on the adjusted feature data, the newly collected electrical signal features are matched by comparing with a known abnormal pattern database, and voltage fluctuations or frequency anomalies in the electrical signal are identified according to the matching results, abnormal patterns of the electrical signal are captured and classified, and abnormal pattern analysis results are generated;

[0010] S4: According to the abnormal pattern analysis result, the electrical signal is dynamically adjusted, the signal is filtered and the gain is adjusted, the identified abnormality is corrected, the quality and stability of the signal transmission process are optimized, and the signal optimization execution result is obtained.

[0011] As a further solution of the present invention, the steps of obtaining the preliminary time-frequency-space feature set are:

[0012] S111: Use sensors to synchronously collect the time, frequency and spatial characteristics of the electrical signals of the differentiated target source, using the formula,

[0013] D i =a·T i +b·F i +c·S i

[0014] Perform feature weighted synthesis on each signal point and calculate the comprehensive feature value D of the i-th signal point i , generating comprehensive feature data, where a, b and c represent the weights of the time, frequency and space features of the electrical signal, respectively, T i 、F i and S i Respectively represent the time, frequency, and spatial characteristic values ​​of the i-th electrical signal point;

[0015] S112: Statistical analysis is performed on the comprehensive feature data using the formula:

[0016]

[0017] and

[0018]

[0019] Calculate the mean value μ and standard deviation σ, filter the data greater than μ+σ, and generate the filtered key feature data, where n represents the total number of signal points;

[0020] S113: Arrange and classify the filtered key feature data, perform structured processing and format unification, and generate a preliminary time-frequency-space feature set.

[0021] As a further solution of the present invention, the steps of calculating and analyzing the relative importance of each feature are:

[0022] S211: Based on the preliminary time-frequency-space feature set, extract the occurrence frequency of each feature to obtain feature frequency data;

[0023] S212: Using the characteristic frequency data, perform normalization processing, using the formula:

[0024]

[0025] Calculate the normalized score G of feature i i , generate normalized frequency data, where G i represents the normalized frequency score of feature i, f i represents the frequency of occurrence of feature i in the feature frequency data, ∑f i represents the sum of all feature frequencies;

[0026] S213: Weighting the normalized frequency data using the formula:

[0027]

[0028] Calculate the importance score W of feature i i , and obtain the feature importance score data.

[0029] As a further solution of the present invention, the step of acquiring the adjusted feature data is:

[0030] S221: Based on the feature importance score data, using the formula:

[0031]

[0032] Calculate the correlation score c between features i and j ij , construct and output the feature correlation matrix, where W i and W j are the importance scores of feature i and feature j, ∑W k Represents the sum of all feature importance scores;

[0033] S222: Using the feature correlation matrix, adjust the data contrast, using the formula:

[0034]

[0035] Calculate the display parameter d of the adjusted feature i i , get the adjustment parameter data, where c ii represents the correlation score between feature i and itself, and max(C) represents the maximum value of the correlation score in the feature correlation matrix;

[0036] S223: Apply the adjustment parameter data, using the formula:

[0037] s′ i =s i ·d i

[0038] Calculate the intensity value s′ of the adjusted feature i i , output the adjusted feature data, s i is the initial feature intensity value.

[0039] As a further solution of the present invention, the step of matching the newly collected electrical signal features is:

[0040] S311: Extract feature vectors based on the adjusted feature data and perform standardization processing using the formula:

[0041]

[0042] Calculate the normalized vector value v of feature i i , generate a standardized feature vector set, where μ s is the average value of the adjusted feature data, σ s is the standard deviation of the adjusted feature data;

[0043] S312: Based on the standardized feature vector set, similarity detection is performed with a known abnormal pattern database, using the formula:

[0044]

[0045] Calculate the distance d between feature i and pattern j in the known abnormal pattern database ij , generate similarity analysis results, where m ij is the ith pattern eigenvalue of the jth pattern in the known anomaly database, and α is the weighting coefficient used to adjust the influence of linear and nonlinear distances;

[0046] S313: According to the similarity analysis result, select the corresponding pattern with the lowest distance as the matching result, identify the pattern that best matches the current signal, and generate a preliminary matching result.

[0047] As a further solution of the present invention, the step of obtaining the abnormal pattern analysis result is:

[0048] S321: Based on the preliminary matching result, evaluate the voltage fluctuation or frequency anomaly, set the anomaly determination threshold θ, if min(d ij )≥θ, the output is 1, otherwise it is 0, where 1 indicates abnormality and 0 indicates normality, the abnormal state of the signal is determined, and the abnormal judgment result is generated;

[0049] S322: Classify and process the abnormality determination results using the formula:

[0050] F i =∑(H i ·w i )

[0051] Calculate the classification value F of anomaly type i i , generate the classified abnormal results, where H i Indicates the abnormal state output value, w i represents the weight of anomaly type i;

[0052] S323: Summarize the classified abnormal results, conduct comprehensive analysis, and output abnormal pattern analysis results.

[0053] As a further solution of the present invention, the steps of obtaining the signal optimization execution result are:

[0054] S411: Based on the abnormal mode analysis result, determine the signal adjustment requirement, using the formula:

[0055] β=1-e -λ·t

[0056] Calculate the filter parameter β, where λ represents the attenuation coefficient and t is the time constant;

[0057] S412: Apply the filtering parameters to filter the electrical signal using the formula:

[0058] S′=β·S+(1-β)·S prev

[0059] Calculate the signal strength S′ after filtering to obtain the signal data after filtering, where S is the signal strength before filtering and S prev is the previous signal strength;

[0060] S413: Perform gain adjustment according to the filtered signal data, using the formula:

[0061]

[0062] Calculate the optimized signal strength S opt , to ensure the quality and stability of the signal during transmission, and obtain the signal optimization execution result, where K is the gain coefficient, S max and S min is the maximum and minimum value of S′, V ref is the reference voltage.

[0063] Multi-dimensional electrical signal processing system, including:

[0064] The signal acquisition module uses sensors to synchronously collect the time, frequency and spatial characteristics of the electrical signals of the differentiated target sources, performs feature weighted synthesis on each signal point, performs statistical analysis, calculates the average value and standard deviation, screens key feature data, organizes and classifies them, and generates a preliminary time-frequency-space feature set;

[0065] The feature analysis module extracts the occurrence frequency of each feature based on the preliminary time-frequency-space feature set, performs normalization processing, calculates the normalized score and performs weighting to obtain feature importance score data;

[0066] The data optimization module calculates the correlation scores between the features based on the feature importance score data, adjusts the data contrast, calculates the display parameters of the adjusted features, adjusts the signal feature strength, and outputs the adjusted feature data;

[0067] The pattern recognition module extracts the feature vector according to the adjusted feature data, performs standardization processing, calculates the standardized vector value of the feature, performs similarity detection with the known abnormal pattern database, identifies the pattern that best matches the current signal, and generates a preliminary matching result;

[0068] The abnormality determination module evaluates the voltage fluctuation or frequency abnormality according to the preliminary matching result, sets the abnormality determination threshold to determine the abnormal state of the signal, performs classification processing and comprehensive analysis, and outputs the abnormal pattern analysis result;

[0069] The signal adjustment module determines the signal adjustment requirements based on the abnormal mode analysis results, calculates and applies filter parameters, performs filtering processing on the electrical signal, performs gain adjustment, and obtains signal optimization execution results.

[0070] Compared with the prior art, the advantages and positive effects of the present invention are:

[0071] In the present invention, hierarchical analysis is combined with feature correlation disclosure to optimize the contrast and clarity of data, thereby greatly improving the efficiency of information extraction in a complex signal environment. Based on the matching and classification of known abnormal patterns, voltage fluctuations and frequency anomalies can be quickly identified, and efficient classification and processing of abnormal signals can be achieved. The real-time and accuracy of signal monitoring are enhanced, and the dynamic adjustment and optimization processing of signals ensures the stability and quality of signal transmission, improves the reliability and adaptability of the entire signal processing process, effectively responds to challenges in a complex signal environment, significantly improves the signal analysis capability and processing efficiency, and enhances the signal quality in practical applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 It is a schematic diagram of the steps of the present invention;

[0073] Figure 2 This is a flow chart of obtaining a preliminary time-frequency-space feature set of the present invention;

[0074] Figure 3 A flow chart for calculating and analyzing the relative importance of each feature of the present invention;

[0075] Figure 4 This is a flow chart for obtaining the adjusted characteristic data of the present invention;

[0076] Figure 5 A flow chart of the steps of matching newly acquired electrical signal features in the present invention;

[0077] Figure 6 A flowchart for obtaining abnormal mode analysis results of the present invention;

[0078] Figure 7 The present invention is a flowchart for obtaining the signal optimization execution results. DETAILED DESCRIPTION

[0079] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with 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 intended to limit the present invention.

[0080] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0081] See also Figure 1 , a multi-dimensional electrical signal processing method, comprising the following steps:

[0082] S1: Synchronously collect the time, frequency and space characteristics of the electrical signals of the differentiated target source, screen the key time-frequency-space data by measuring and comparing the characteristics, and generate a preliminary time-frequency-space feature set;

[0083] S2: Based on the preliminary time-frequency-space feature set, perform data hierarchical analysis, calculate and analyze the relative importance of each feature, reveal the correlation between features based on the analysis results, select and adjust the contrast and clarity of the data, and obtain the adjusted feature data;

[0084] S3: Based on the adjusted feature data, the newly collected electrical signal features are matched by comparing with the known abnormal pattern database, and the voltage fluctuation or frequency abnormality in the electrical signal is identified according to the matching result, and the abnormal pattern of the electrical signal is captured and classified to generate the abnormal pattern analysis result;

[0085] S4: According to the abnormal pattern analysis results, the electrical signal is dynamically adjusted, the signal is filtered and the gain is adjusted, the identified abnormalities are corrected, the quality and stability of the signal transmission process are optimized, and the signal optimization execution results are obtained.

[0086] The preliminary time-frequency-space feature set includes time series data, frequency distribution characteristics and spatial coordinate information. The adjusted feature data includes feature importance measurement results, feature correlation diagrams and data clarity levels. The abnormal pattern analysis results include voltage anomaly type, frequency deviation degree and abnormal pattern classification results. The signal optimization execution results include signal filtering effect, gain adjustment level and optimized stability assessment records.

[0087] See also Figure 2 , the steps to obtain the preliminary time-frequency-space feature set are:

[0088] S111: Use sensors to synchronously collect the time, frequency and spatial characteristics of the electrical signals of the differentiated target source, using the formula,

[0089] D i =a·T i +b·F i +c·S i

[0090] Perform feature weighted synthesis on each signal point and calculate the comprehensive feature value D of the i-th signal point i , generating comprehensive feature data, where a, b and c represent the weights of the time, frequency and space features of the electrical signal, respectively, T i 、F i and S i Respectively represent the time, frequency, and spatial characteristic values ​​of the i-th electrical signal point;

[0091] S112: Perform statistical analysis on the comprehensive feature data using the formula,

[0092]

[0093] and

[0094]

[0095] Calculate the mean value μ and standard deviation σ, filter the data greater than μ+σ, and generate the filtered key feature data, where n represents the total number of signal points;

[0096] S113: Arrange and classify the filtered key feature data, perform structured processing and format unification, and generate a preliminary time-frequency-space feature set.

[0097] Collect data of an electrical signal point, whose time characteristic value T i =20 units, frequency characteristic value F i =15Hz, spatial eigenvalue S i =5m, and the weights are set to a=0.5, b=0.3, c=0.2.

[0098] Substitute into the formula to calculate the comprehensive eigenvalue of the signal point:

[0099] D i =0.5×20+0.3×15+0.2×5

[0100] =10+4.5+1=15.5

[0101] Comprehensive eigenvalue D i =15.5 represents the weighted composite value of the signal point after considering the time, frequency and space characteristics, which can be used to compare with the characteristic values ​​of other signal points to determine which signal points have more prominent characteristics.

[0102] Collect data from three signal points and calculate their comprehensive eigenvalues, which are {10, 15.5, 20} respectively.

[0103] Calculate the average:

[0104]

[0105] Calculate the standard deviation:

[0106]

[0107] The average value μ≈15.1667 and the standard deviation σ≈4.09 describe the overall distribution of the comprehensive characteristic values ​​of the signal points. The average value indicates the comprehensive characteristic level of the signal points in general, while the standard deviation indicates the degree of fluctuation of the comprehensive characteristic values ​​of the signal points.

[0108] According to the above calculation results, we can get:

[0109] μ+σ=15.1667+4.09=19.2567

[0110] After screening the data with values ​​greater than the threshold, only 20 data points met the criteria, indicating that among all the signal points examined, the third signal point was significantly higher than the average level plus one standard deviation in terms of comprehensive time, frequency and space characteristics, and was a key signal point that should be prioritized or further analyzed.

[0111] See also Figure 3 , the calculation and analysis steps for the relative importance of each feature are:

[0112] S211: Based on the preliminary time-frequency-space feature set, extract the occurrence frequency of each feature to obtain feature frequency data;

[0113] S212: Use the characteristic frequency data to perform normalization processing, using the formula,

[0114]

[0115] Calculate the normalized score G of feature i i , generate normalized frequency data, where G i represents the normalized frequency score of feature i, f i represents the frequency of occurrence of feature i in the feature frequency data, ∑f i represents the sum of all feature frequencies;

[0116] S213: Weight the normalized frequency data using the formula:

[0117]

[0118] Calculate the importance score W of feature ii , and obtain the feature importance score data.

[0119] Based on the preliminary time-frequency-space feature set, the feature set contains {A, A, B, C, A, B, D}.

[0120] The frequency of extracting feature A is 3, which means that among all the recorded features, A appears most frequently, indicating that feature A may be a key feature.

[0121] Similarly, the frequency of feature B extracted is 2, the frequency of feature C is 1, and the frequency of feature D is 1.

[0122] Combining the above feature frequency extraction results, we get:

[0123] ∑f i =3+2+1+1=7

[0124] Substitute into the formula and calculate:

[0125]

[0126] The normalized frequency score of feature A is 0.4286, which means that feature A accounts for about 42.86% of the frequency of all features and is an important feature in the analysis, followed by feature B, feature C, and feature D.

[0127] Taking feature A as an example, using the normalized score G A =0.4286, substitute into the formula and calculate:

[0128]

[0129] The importance score of feature A is about 1.2, which helps quantify the importance of feature A relative to other features and provides a basis for subsequent data processing and analysis.

[0130] See also Figure 4 , the steps to obtain the adjusted feature data are:

[0131] S221: Based on the feature importance score data, the formula is used.

[0132]

[0133] Calculate the correlation score c between features i and j ij , construct and output the feature correlation matrix, where W i and W j are the importance scores of feature i and feature j, ∑W k Represents the sum of all feature importance scores;

[0134] S222: Use the feature correlation matrix to adjust the data contrast, using the formula,

[0135]

[0136] Calculate the display parameter d of the adjusted feature i i , get the adjustment parameter data, where c ii represents the correlation score between feature i and itself, and max(C) represents the maximum value of the correlation score in the feature correlation matrix;

[0137] S223: Apply adjustment parameter data, using formula,

[0138] s′ i =s i ·d i

[0139] Calculate the intensity value s′ of the adjusted feature i i , output the adjusted feature data, s i is the initial feature intensity value.

[0140] Combined with the above calculation results W A ≈1.2, and W is calculated in the same way B ≈1.14,W C ≈1.07,W D ≈1.07.

[0141] Calculate the sum of all feature importance scores:

[0142] ∑W k =1.2+1.14+1.07+1.07=4.48

[0143] Calculate the correlation between features A and B:

[0144]

[0145] The correlation score of feature A and B is about 0.305, indicating that these two features have a relatively high association in the entire feature set and are used for further analysis and data adjustment.

[0146] Similarly, calculation:

[0147]

[0148] Combining the above calculation results, the correlation matrix is ​​{0.305, 0.272, 0.256, 0.287}.

[0149] Calculate the autocorrelation c of feature A AA = 0.321 and the maximum value max(c) = 0.305, then

[0150]

[0151] Display parameter d of feature A A ≈1, indicating that feature A should basically maintain the original contrast and clarity when displayed or further processed. Features B, C, and D are processed in the same way. It is calculated that d B ≈0.95, d C ≈0.84, d D ≈0.84.

[0152] Collect the original feature intensity values ​​s i ={100, 85, 70, 50}, combined with the above calculation results, substitute into the formula, and we get:

[0153] s′ A =100×1.0=100

[0154] s′ B =85×0.95=80.75

[0155] s′ C =70×0.84=58.8

[0156] s′ D =50×0.84=42

[0157] Adjusted feature intensity value s′ i ={100, 80.75, 58.8, 42} indicates that feature A maintains its original strength, while other features have their strengths adjusted according to the correlation scores, reflecting the relative importance of the features.

[0158] See also Figure 5 , the steps for matching the newly acquired electrical signal features are:

[0159] S311: Extract feature vectors based on the adjusted feature data and perform standardization, using the formula:

[0160]

[0161] Calculate the normalized vector value v of feature i i , generate a standardized feature vector set, where μ s is the average value of the adjusted feature data, σ s is the standard deviation of the adjusted feature data;

[0162] S312: Based on the standardized feature vector set, similarity detection is performed with the known abnormal pattern database using the formula,

[0163]

[0164] Calculate the distance d between feature i and pattern j in the known abnormal pattern database ij , generate similarity analysis results, where m ij is the ith pattern eigenvalue of the jth pattern in the known anomaly database, and α is the weighting coefficient used to adjust the influence of linear and nonlinear distances;

[0165] S313: According to the similarity analysis result, the corresponding pattern with the lowest distance is selected as the matching result, the pattern that best matches the current signal is identified, and a preliminary matching result is generated.

[0166] Combined with the above adjusted feature data s′ i ={100, 80.75, 58.8, 42}, average value μ s =70.3875, standard deviation σ s ≈22.

[0167] For s′1=100:

[0168]

[0169] For s′2=80.75:

[0170]

[0171] For s′3=58.8:

[0172]

[0173] For s′4=42:

[0174]

[0175] The set of standardized feature vectors [1.35, 0.47, -0.53, -1.29] represents the deviation of each feature from the mean, measured in units of standard deviation, so that the data is at the same standard when compared with the patterns in the database.

[0176] Collect known pattern eigenvalues ​​m ij =[1.2, 1.4, -0.8, -0.95], set the weighting coefficient α=0.5.

[0177] With d 1j For example, substitute the formula and get:

[0178]

[0179] The obtained distance d 1j≈0.312 indicates the closeness of the normalized feature vector to a pattern in the database. Lower values ​​indicate a high degree of match and can be used for further analysis and decision making.

[0180] Similarly, calculate d 2j ≈1.395, d 3j ≈0.405, d 4j ≈0.51, then we get d 1j is the minimum value, indicating that d 1j It is the closest match in the newly acquired signal to a known anomaly pattern in the database.

[0181] See also Figure 6 , the steps to obtain the abnormal pattern analysis results are:

[0182] S321: Based on the preliminary matching results, evaluate the voltage fluctuation or frequency anomaly and set the anomaly determination threshold θ. If min(d ij )≥θ, the output is 1, otherwise it is 0, where 1 indicates abnormality and 0 indicates normality, the abnormal state of the signal is determined, and the abnormal judgment result is generated;

[0183] S322: Classify and process according to the abnormality determination result, using the formula:

[0184] F i =∑(H i ·w i )

[0185] Calculate the classification value F of anomaly type i i , generate the classified abnormal results, where H i Indicates the abnormal state output value, w i represents the weight of anomaly type i;

[0186] S323: Summarize the classified abnormal results, conduct comprehensive analysis, and output the abnormal pattern analysis results.

[0187] Combined with the above calculation results, we can get min(d ij )=0.374, and set the abnormality determination threshold θ=0.5.

[0188] 0.374<0.5

[0189] Therefore, the output value is 0, indicating that the signal is normal, confirming that the electrical signal shows no signs of voltage fluctuation or frequency anomalies in this particular test. This conclusion is based on the relatively long distance from the abnormal pattern.

[0190] The above analysis converts the value to 0, and the corresponding weight w i =0.2 (low weight because the state is normal).

[0191] Substituting into the formula, we get:

[0192] F i =∑(0×0.2)=0

[0193] Classification value F i =0 indicates that no abnormal pattern was identified in this particular detection, that is, the signal status is good and no further intervention is required.

[0194] See also Figure 7 , the steps to obtain the signal optimization execution results are:

[0195] S411: Based on the abnormal pattern analysis results, determine the signal adjustment requirements, using the formula,

[0196] β=1-e -λ·t

[0197] Calculate the filter parameter β, where λ represents the attenuation coefficient and t is the time constant;

[0198] S412: Apply filter parameters to filter the electrical signal using the formula:

[0199] S′=β·S+(1-β)·S prev

[0200] Calculate the signal strength S′ after filtering to obtain the signal data after filtering, where S is the signal strength before filtering and S prev is the previous signal strength;

[0201] S413: According to the filtered signal data, gain adjustment is performed using the formula:

[0202]

[0203] Calculate the optimized signal strength S opt , to ensure the quality and stability of the signal during transmission, and obtain the signal optimization execution result, where K is the gain coefficient, S max and S min is the maximum and minimum value of S′, V ref is the reference voltage.

[0204] Set λ=0.1 and t=5 seconds.

[0205] Substituting into the formula, we get:

[0206] β=1-e -0.1×5 =1-e -0.5 ≈1-0.606=0.394

[0207] The calculated filter parameter β≈0.394 indicates the influence of the filter in signal processing. When the filter is applied, 39.4% of the new signal value is combined with 60.6% of the previous signal state to generate the updated value of the current signal, which is used to balance the response speed and smoothness of the signal. A low β value means that the signal response is more sensitive to new data, while a higher proportion of previous states helps suppress noise, making the output signal more stable and smooth.

[0208] Collect the current signal strength S = 100 units, the previous signal strength S prev =80 units, and β =0.394.

[0209] Substituting into the formula, we get:

[0210] S′=0.394×100+(1-0.394)×80

[0211] =39.4+48.48=87.88

[0212] The signal strength S′≈87.88 after filtering shows that after adopting the exponential filtering parameters, the signal is smoothed, the influence of noise is reduced, and the stability of the signal is improved. Reaching 87.88 means that the filter successfully smoothes the input signal according to the set attenuation parameters, reduces the influence of mutations or anomalies, and thus provides a more stable and reliable basic signal for further gain adjustment.

[0213] Set K = 1.0, S max =110, S min =70, V ref =100.

[0214] Substituting into the formula, we get:

[0215]

[0216] The final optimized signal strength S opt ≈123.032 indicates that through gain adjustment, the strength of the signal is boosted to ensure its quality and stability during transmission, adapting to a wider range of voltage fluctuations or frequency anomalies that may exist.

[0217] Multi-dimensional electrical signal processing system, including:

[0218] The signal acquisition module uses sensors to synchronously collect the time, frequency and spatial characteristics of the electrical signals of the differentiated target sources, performs feature weighted synthesis on each signal point, performs statistical analysis, calculates the average value and standard deviation, screens key feature data, organizes and classifies them, and generates a preliminary time-frequency-space feature set;

[0219] The feature analysis module extracts the occurrence frequency of each feature based on the preliminary time-frequency-space feature set, performs normalization, calculates the normalized score and weights it to obtain the feature importance score data;

[0220] The data optimization module calculates the correlation scores between features based on the feature importance score data, adjusts the data contrast, calculates the display parameters of the adjusted features, adjusts the signal feature strength, and outputs the adjusted feature data;

[0221] The pattern recognition module extracts the feature vector based on the adjusted feature data, performs standardization processing, calculates the standardized vector value of the feature, performs similarity detection with the known abnormal pattern database, identifies the pattern that best matches the current signal, and generates a preliminary matching result;

[0222] The abnormality determination module evaluates the voltage fluctuation or frequency abnormality based on the preliminary matching results, sets the abnormality determination threshold to determine the abnormal state of the signal, performs classification processing and comprehensive analysis, and outputs the abnormal pattern analysis results;

[0223] The signal adjustment module determines the signal adjustment requirements based on the abnormal pattern analysis results, calculates and applies the filter parameters, filters the electrical signal, adjusts the gain, and obtains the signal optimization execution results.

[0224] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A multi-dimensional electrical signal processing method, characterized in that: The following steps are involved: The time, frequency and space characteristics of the electrical signals of the differentiated target sources are collected synchronously, and the key time-frequency-space data are screened by measuring and comparing the characteristics to generate a preliminary time-frequency-space feature set; According to the preliminary time-frequency-space feature set, perform data hierarchical analysis, calculate and analyze the relative importance of each feature, reveal the correlation between features according to the analysis results, select and adjust the contrast and clarity of the data, and obtain adjusted feature data; Based on the adjusted characteristic data, the newly collected electrical signal characteristics are matched by comparing with a known abnormal pattern database, and voltage fluctuations or frequency anomalies in the electrical signal are identified according to the matching results, abnormal patterns of the electrical signal are captured and classified, and abnormal pattern analysis results are generated; According to the abnormal pattern analysis results, the electrical signal is dynamically adjusted, the signal is filtered and the gain is adjusted, the identified abnormalities are corrected, the quality and stability of the signal transmission process are optimized, and the signal optimization execution result is obtained.

2. The multi-dimensional electrical signal processing method according to claim 1, characterized in that: The steps for obtaining the preliminary time-frequency-space feature set are: The sensor is used to synchronously collect the time, frequency and spatial characteristics of the electrical signal of the differentiated target source, and the formula is used. D i =a·T i +b·F i +c·S i Perform feature weighted synthesis on each signal point and calculate the comprehensive feature value D of the i-th signal point i , generating comprehensive feature data, where a, b and c represent the weights of the time, frequency and space features of the electrical signal, respectively, T i 、F i and S i Respectively represent the time, frequency, and spatial characteristic values ​​of the i-th electrical signal point; The comprehensive characteristic data is statistically analyzed using the formula: and Calculate the mean value μ and standard deviation σ, filter the data greater than μ+σ, and generate the filtered key feature data, where n represents the total number of signal points; The screened key feature data are sorted and classified, structured and formatted to generate a preliminary time-frequency-space feature set.

3. The multi-dimensional electrical signal processing method according to claim 2, characterized in that: The calculation and analysis steps for the relative importance of each feature are: Based on the preliminary time-frequency-space feature set, extract the occurrence frequency of each feature to obtain feature frequency data; Using the characteristic frequency data, normalization is performed using the formula, Calculate the normalized score G of feature i i , generate normalized frequency data, where G i represents the normalized frequency score of feature i, f i represents the frequency of occurrence of feature i in the feature frequency data, ∑f i represents the sum of all feature frequencies; The normalized frequency data is weighted using the formula: Calculate the importance score W of feature i i , and obtain the feature importance score data.

4. The multi-dimensional electrical signal processing method according to claim 3, characterized in that: The steps for obtaining the adjusted feature data are: Based on the feature importance score data, the formula is used: Calculate the correlation score c between features i and j ij , construct and output the feature correlation matrix, where W i and W j are the importance scores of feature i and feature j, ∑W k Represents the sum of all feature importance scores; Using the feature correlation matrix, the data contrast is adjusted using the formula, Calculate the display parameter d of the adjusted feature i i , get the adjustment parameter data, where c ii represents the correlation score between feature i and itself, and max(C) represents the maximum value of the correlation score in the feature correlation matrix; Applying the adjustment parameter data, using the formula, s′ i =s i ·d i Calculate the intensity value s′ of the adjusted feature i i , output the adjusted feature data, s i is the initial feature intensity value.

5. The multi-dimensional electrical signal processing method according to claim 4, characterized in that: The step of matching the newly collected electrical signal features is as follows: According to the adjusted feature data, feature vectors are extracted and standardized, using the formula: Calculate the normalized vector value v of feature i i , generate a standardized feature vector set, where μ s is the average value of the adjusted feature data, σ s is the standard deviation of the adjusted feature data; Based on the standardized feature vector set, similarity detection is performed with the known abnormal pattern database using the formula: Calculate the distance d between feature i and pattern j in the known abnormal pattern database ij , generate similarity analysis results, where m ij is the ith pattern eigenvalue of the jth pattern in the known anomaly database, and α is the weighting coefficient used to adjust the influence of linear and nonlinear distances; According to the similarity analysis result, the corresponding pattern with the lowest distance is selected as the matching result, the pattern that best matches the current signal is identified, and a preliminary matching result is generated.

6. The multi-dimensional electrical signal processing method according to claim 5, characterized in that: The steps for obtaining the abnormal pattern analysis result are: According to the preliminary matching results, voltage fluctuation or frequency anomaly is evaluated and an abnormality judgment threshold θ is set. If min(d ij )≥θ, the output is 1, otherwise it is 0, where 1 indicates abnormality and 0 indicates normality, the abnormal state of the signal is determined, and the abnormal judgment result is generated; According to the abnormal determination results, classification processing is performed, using the formula: F i =∑(H i ·w i ) Calculate the classification value F of anomaly type i i , generate the classified abnormal results, where H i Indicates the abnormal state output value, w i represents the weight of anomaly type i; The classified abnormal results are summarized, a comprehensive analysis is performed, and the abnormal pattern analysis results are output.

7. The multi-dimensional electrical signal processing method according to claim 6, characterized in that: The steps for obtaining the signal optimization execution result are: Based on the abnormal pattern analysis results, the signal adjustment requirements are determined using the formula, β＝1-e -λ·t Calculate the filter parameter β, where λ represents the attenuation coefficient and t is the time constant; Apply the filtering parameters to filter the electrical signal using the formula, S′=β·S+(1-β)·S prev Calculate the signal strength S′ after filtering to obtain the signal data after filtering, where S is the signal strength before filtering and S prev is the previous signal strength; According to the filtered signal data, gain adjustment is performed using the formula: Calculate the optimized signal strength S opt , to ensure the quality and stability of the signal during transmission, and obtain the signal optimization execution result, where K is the gain coefficient, S max and S min is the maximum and minimum value of S′, V ref is the reference voltage.

8. A multi-dimensional electrical signal processing system, characterized in that: According to any one of claims 1 to 7, the multi-dimensional electrical signal processing method comprises: The signal acquisition module uses sensors to synchronously collect the time, frequency and spatial characteristics of the electrical signals of the differentiated target sources, performs feature weighted synthesis on each signal point, performs statistical analysis, calculates the average value and standard deviation, screens key feature data, organizes and classifies them, and generates a preliminary time-frequency-space feature set; The feature analysis module extracts the occurrence frequency of each feature based on the preliminary time-frequency-space feature set, performs normalization processing, calculates the normalized score and performs weighting to obtain feature importance score data; The data optimization module calculates the correlation scores between the features based on the feature importance score data, adjusts the data contrast, calculates the display parameters of the adjusted features, adjusts the signal feature strength, and outputs the adjusted feature data; The pattern recognition module extracts the feature vector according to the adjusted feature data, performs standardization processing, calculates the standardized vector value of the feature, performs similarity detection with the known abnormal pattern database, identifies the pattern that best matches the current signal, and generates a preliminary matching result; The abnormality determination module evaluates the voltage fluctuation or frequency abnormality according to the preliminary matching result, sets the abnormality determination threshold to determine the abnormal state of the signal, performs classification processing and comprehensive analysis, and outputs the abnormal pattern analysis result; The signal adjustment module determines the signal adjustment requirements based on the abnormal mode analysis results, calculates and applies filter parameters, performs filtering processing on the electrical signal, performs gain adjustment, and obtains signal optimization execution results.

Images