A step voltage online monitoring system and method

Abstract

The present invention relates to the field of measurement and detection technology, and in particular to a step voltage online monitoring system and method. It includes: real-time acquisition of step voltage data, and converting the step voltage data into analog voltage signals, and then converting the analog voltage signals into digital voltage signals, introducing multi-level collaborative filtering and trend capture algorithms to intelligently process the digital voltage signals; calculating the real-time step voltage value based on the step voltage digital signal after intelligent processing to obtain a step voltage value sequence set; and analyzing the change trend of the step voltage in the step voltage value sequence set through a time-frequency domain hybrid similarity trend matrix algorithm to obtain step voltage data and fluctuation analysis results, and then monitoring the step voltage data in real time according to a preset dynamic threshold. It solves the technical problem of inaccurate processing and analysis of step voltage signals when monitoring step voltage online in a complex power environment.

Description

Technical Field

[0001] The present invention relates to the field of measurement and detection technology, and in particular to an online step voltage monitoring system and method. Background Art

[0002] In the power system, step voltage refers to the potential difference formed on the ground surface by the current caused by the ground fault when the power equipment or cable fails. The existence of step voltage poses a serious threat to the surrounding environment and personal safety, especially in a humid surface or highly conductive soil environment, where step voltage may cause serious electric shock accidents. Therefore, real-time and accurate monitoring of step voltage changes and timely warning of abnormal conditions are of great significance for ensuring the safe operation of the power system and the safety of personnel. Traditional step voltage monitoring methods mainly rely on fixed voltage monitoring equipment, which usually focuses on analog signals, and it is difficult to achieve full-time and comprehensive coverage of step voltage. With the increase in the scale and complexity of power systems, a single fixed monitoring method can no longer meet the needs of modern power systems. At the same time, step voltage monitoring faces the challenges of noise and interference. Complex electromagnetic interference in the power environment often affects the step voltage signal, making the step voltage signal unstable or even causing false alarms. In order to improve the accuracy and anti-interference ability of step voltage monitoring, modern step voltage monitoring is gradually developing towards digitalization and intelligence.

[0003] In summary, the existing step voltage detection method has the problem of inaccurate processing and analysis of step voltage signals when monitoring the step voltage online in a complex power environment. Summary of the invention

[0004] The present invention provides a step voltage online monitoring system and method to solve the technical problem of inaccurate step voltage signal processing and analysis when the step voltage is monitored online in a complex power environment.

[0005] A step voltage online monitoring system and method of the present invention specifically includes the following technical solutions:

[0006] A step voltage online monitoring method comprises the following steps:

[0007] S1. Collect step voltage data in real time, convert the step voltage data into analog voltage signals, and then convert the analog voltage signals into digital voltage signals. Multi-level collaborative filtering and trend capture algorithms are introduced to intelligently process the digital voltage signals to obtain the step voltage digital signals after intelligent processing.

[0008] S2. Calculate the real-time step voltage value based on the step voltage digital signal after intelligent processing to obtain the step voltage value sequence, merge the step voltage value sequence to obtain the step voltage value sequence set; and analyze the change trend of the step voltage in the step voltage value sequence set through the time domain-frequency domain mixed similarity trend matrix algorithm to obtain the step voltage data and fluctuation analysis results, and then monitor the step voltage data in real time according to the preset dynamic threshold.

[0009] Preferably, the S1 specifically includes:

[0010] The multi-level collaborative filtering and trend capture algorithm processes the digital voltage signal in two dimensions of time and frequency based on filters of different levels and recursive structures, including: basic filtering, dynamic weighted noise reduction, trend tracking, and abnormal fluctuation correction processing.

[0011] Preferably, the S1 specifically includes:

[0012] In the process of implementing the multi-level collaborative filtering and trend capture algorithm, the digital voltage signal is subjected to multi-level filtering to obtain the filtered voltage signal output. The specific implementation is as follows:

[0013] ,

[0014] in, It is The voltage signal output after level filtering; is the initial step voltage signal; is the direct signal input weight of the current layer; It is the cumulative weight of each level of filtering on the historical signal; It is The historical signal of the level filter output; is the sampling period, which indicates the time step between different filter layers; It is Adaptive adjustment coefficient of the historical signal output by the first-order filtering; yes The step voltage signal at the moment; It is Estimated standard deviation of the voltage signal after level filtering.

[0015] Preferably, the S1 specifically includes:

[0016] In the process of realizing the multi-level collaborative filtering and trend capture algorithm, the voltage signal output after multi-level filtering is subjected to dynamic weighted noise reduction processing in the time domain to obtain the voltage signal output after dynamic weighted noise reduction.

[0017] Preferably, the S1 specifically includes:

[0018] In the process of implementing the multi-level collaborative filtering and trend capture algorithm, the voltage signal output after dynamic weighted noise reduction is subjected to an overall trend analysis, the abnormal pattern of voltage change is identified, and abnormal fluctuation correction processing is performed to obtain the step voltage digital signal after intelligent processing.

[0019] Preferably, the S2 specifically includes:

[0020] In the process of realizing the time domain-frequency domain hybrid similarity trend matrix algorithm, each step voltage value sequence in the step voltage value sequence set is subjected to a hybrid processing of the time domain and the frequency domain to generate a dynamic time domain-frequency domain hybrid matrix.

[0021] Preferably, the S2 specifically includes:

[0022] In the process of implementing the time domain-frequency domain hybrid similarity trend matrix algorithm, a similarity matrix is ​​constructed based on the dynamic time domain-frequency domain hybrid matrix, and the similarity matrix is ​​mapped to the trend quantification index to obtain the trend quantification index of the step voltage; the trend quantification index of the step voltage is used to represent the changing trend of the step voltage.

[0023] Preferably, the S2 specifically includes:

[0024] In the process of implementing the time domain-frequency domain hybrid similarity trend matrix algorithm, the trend quantification index is judged by the dynamic threshold method; the dynamic threshold is set based on the mean, standard deviation and trend change rate of the trend quantification index. When the trend quantification index exceeds the dynamic threshold, it is considered that there is an abnormal step voltage, and the abnormal step voltage data and trend quantification index are output.

[0025] A step voltage online monitoring system includes the following parts:

[0026] Step voltage monitoring sensor module, signal processing module, data calculation and analysis module, data storage and transmission module, early warning module;

[0027] The step voltage monitoring sensor module collects step voltage data in real time, converts the step voltage data into analog voltage signals, and sends the analog voltage signals to the signal processing module;

[0028] The signal processing module converts the analog voltage signal into a digital voltage signal, performs intelligent processing on the digital voltage signal to obtain an intelligently processed step voltage digital signal, and sends the intelligently processed step voltage digital signal to the data calculation and analysis module;

[0029] The data calculation and analysis module calculates the real-time step voltage value based on the step voltage digital signal after intelligent processing, and analyzes the change trend of the step voltage through the time domain-frequency domain hybrid similarity trend matrix algorithm to obtain the step voltage data and fluctuation analysis results, and sends the step voltage data and fluctuation analysis results to the data storage and transmission module;

[0030] The data storage and transmission module stores and backs up the step voltage data and fluctuation analysis results, and sends the step voltage data and fluctuation analysis results to the early warning module;

[0031] The early warning module monitors the step voltage data in real time according to the preset dynamic threshold, step voltage data and fluctuation analysis results, obtains the monitoring results, implements early warning based on the monitoring results, and performs online monitoring of the step voltage.

[0032] The beneficial effects of the technical solution of the present invention are:

[0033] 1. Multi-level collaborative filtering and trend capture algorithms are adopted to filter layer by layer in different frequency bands, and the step voltage signal fluctuation is adaptively adjusted through dynamic weighted noise reduction processing to remove interference signals and reduce abnormal noise, and finally a stable step voltage signal output is obtained, which effectively improves the signal-to-noise ratio. The step voltage online monitoring system can maintain the ability to accurately process voltage signals in complex noise environments, and is especially suitable for real-time monitoring in harsh electromagnetic environments.

[0034] 2. Use the time domain-frequency domain mixed similarity trend matrix algorithm to analyze the step voltage value, combine the time and frequency characteristics to capture the changing trend of the step voltage signal, and obtain the step voltage data and fluctuation analysis results. This trend analysis can not only identify the short-term drastic fluctuations of the step voltage signal, but also detect the long-term trend of the step voltage, obtain the potential step voltage abnormal risk, and provide strong data support for subsequent safety warnings. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a structural diagram of a step voltage online monitoring system according to the present invention;

[0036] Figure 2 The present invention is a flowchart of a step voltage online monitoring method. DETAILED DESCRIPTION

[0037] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0039] The specific scheme of a step voltage online monitoring system and method provided by the present invention is described in detail below with reference to the accompanying drawings.

[0040] See attached Figure 1 , which shows a structural diagram of a step voltage online monitoring system provided by an embodiment of the present invention, the system includes the following parts:

[0041] Step voltage monitoring sensor module, signal processing module, data calculation and analysis module, data storage and transmission module, early warning module;

[0042] The step voltage monitoring sensor module uses a multi-point sensor array to cover the target area, collects the step voltage data in the target area environment in real time, converts the step voltage data into an analog voltage signal, and sends the analog voltage signal to the signal processing module;

[0043] The signal processing module converts the analog voltage signal into a digital voltage signal through an analog-to-digital converter, performs intelligent processing on the digital voltage signal to obtain an intelligently processed step voltage digital signal, and sends the intelligently processed step voltage digital signal to the data calculation and analysis module;

[0044] The data calculation and analysis module calculates the real-time step voltage value based on the step voltage digital signal after intelligent processing, and analyzes the change trend of the step voltage through the time domain-frequency domain hybrid similarity trend matrix algorithm to obtain the step voltage data and fluctuation analysis results, and sends the step voltage data and fluctuation analysis results to the data storage and transmission module;

[0045] The data storage and transmission module stores and backs up the step voltage data and fluctuation analysis results to ensure the integrity of the data, and sends the step voltage data and fluctuation analysis results to the early warning module;

[0046] The early warning module monitors the step voltage data in real time according to the dynamic threshold, step voltage data and fluctuation analysis results, obtains the monitoring results, implements early warning based on the monitoring results, and realizes online monitoring of the step voltage.

[0047] See attached Figure 2 , which shows a flow chart of a step voltage online monitoring method provided by an embodiment of the present invention, the method comprising the following steps:

[0048] S1. Collect step voltage data in real time, convert the step voltage data into analog voltage signals, and then convert the analog voltage signals into digital voltage signals. Multi-level collaborative filtering and trend capture algorithms are introduced to intelligently process the digital voltage signals to obtain the step voltage digital signals after intelligent processing.

[0049] Professional technical staff will arrange the step voltage monitoring area A sensor is arranged on the ground where step voltage may occur. The sensor collects step voltage data in real time and directly generates an analog voltage signal, which is then converted into a digital voltage signal using an analog-to-digital converter (such as a 12-bit or 16-bit ADC).

[0050] Furthermore, a multi-level collaborative filtering and trend capture algorithm is introduced to perform intelligent processing on the digital voltage signal to obtain the step voltage digital signal after intelligent processing. The multi-level collaborative filtering and trend capture algorithm is based on multiple levels of filters and recursive structures to perform complex processing on the digital voltage signal in two dimensions: time and frequency. The entire processing process is divided into four main steps: basic filtering, dynamic weighted noise reduction, trend tracking, and abnormal fluctuation correction; accurate extraction of effective signals in a noisy environment to achieve high-precision digital voltage signal processing. The specific implementation process is as follows:

[0051] First, the digital voltage signal is subjected to multi-level filtering. The noise signals of different frequency bands are processed by multi-layer filtering to obtain a clearer voltage signal output. Voltage signal output after level filtering It is expressed as:

[0052] ,

[0053] in, It is The voltage signal output after level filtering reflects the current moment go through The voltage signal after the secondary filtering; is the initial step voltage signal, the original signal containing noise and interference; is the direct signal input weight of the current layer, with a value range of , used to adjust the weight of the current layer input; It is the cumulative weight of each level of filtering on the historical signal, which is used to adjust the influence of different levels of filtering on the historical signal to meet ; It is The historical signal of the level filter output is used to build a multi-level filtering structure; is the sampling period, which indicates the time step between different filter layers; It is The adaptive adjustment coefficient of the historical signal output by the first-level filter reflects the influence of the long-time span signal and satisfies ; yes The step voltage signal at the moment; It is The standard deviation estimate of the voltage signal after level filtering is used to reduce the interference of high-frequency noise.

[0054] Multi-layer filtering combines the filtered voltage signal output with the time delay signal and applies it to multiple filtering layers to remove noise in multiple frequency bands in the frequency domain. Through multi-layer collaborative filtering, a more stable voltage signal output is initially obtained. , this result will be used as the input signal for the subsequent dynamic weighted denoising.

[0055] Furthermore, dynamic weighted noise reduction processing is performed in the time domain. In the dynamic weighted noise reduction stage in the time domain, the adaptive weight of the fluctuation characteristics is introduced to process the local abnormal fluctuations in the filtered voltage signal output, thereby further optimizing the filtering effect. The voltage signal output after multi-layer filtering is As input, a dynamically weighted and denoised voltage signal is generated as output , specifically expressed as:

[0056] ,

[0057] in, It is the voltage signal output after dynamic weighted noise reduction; is the time window size, which is used to calculate the amount of historical data for weighted denoising; is the dynamic weighting coefficient at the current moment, satisfying ; It is the standard deviation of the signal fluctuation within the time window, estimates the noise characteristics, and affects the weight distribution at each moment; is a small constant used to avoid division by zero errors; yes The passage of time The voltage signal after the secondary filtering; is the coefficient of the phase correction term, which is used to adjust the local offset; is a periodic interference term, represents the frequency of the periodic component; is the time interval used to control the length of the weighted window and the time range of the denoising window.

[0058] The purpose of dynamic weighted noise reduction is to improve the smoothness of the signal and reduce the interference of abnormal values ​​according to the signal fluctuation, so as to obtain a more stable voltage signal output after dynamic weighted noise reduction. This result will be used for subsequent frequency domain trend capture.

[0059] Furthermore, the voltage signal output after dynamic weighted noise reduction is Perform overall trend analysis to identify potential abnormal patterns of voltage changes in order to achieve abnormal fluctuation correction processing. In this stage, the signal trend is modeled by polynomial fitting method, and phase difference tracking of the signal is added to accurately capture the instantaneous trend of voltage fluctuation. Introduce the fitting function of trend capture , whose expression is:

[0060] ,

[0061] in, It is a fitting function for trend capture, used to capture the real-time change trend of voltage; , , , is the polynomial fitting coefficient, which is fitted by the least square method and the calculation process can be obtained by using a recursive iterative algorithm; It is the signal amplitude coefficient, which is used to balance the fitting function of trend capture and the voltage signal amplitude after dynamic weighted noise reduction, so as to avoid the oversensitivity of the fitting process to abnormal fluctuations. Through the processing of the fitting function of the above trend capture, the voltage change trend can be captured with higher accuracy.

[0062] Furthermore, the phase difference of the signal is calculated to identify the mutation trend:

[0063] ,

[0064] in, It indicates the phase difference between the noise reduction signal and the trend signal, reflecting the degree of deviation of the signal during the instantaneous change process; It is the derivative of the voltage signal output after dynamic weighted noise reduction, that is, the instantaneous rate of change; It is the derivative of the trend capturing fitting function and represents the rate of change of the trend.

[0065] By monitoring the dramatic changes in the phase difference between the noise reduction signal and the trend signal, abnormal fluctuations in the voltage signal can be detected and marked as potential voltage abnormal areas. The phase difference between the noise reduction signal and the trend signal It will be passed to the next stage of abnormal fluctuation correction process.

[0066] The abnormal fluctuation correction process uses an adaptive regression model, which combines the trend capture fitting function with historical data to automatically correct abnormal fluctuations. The expression is:

[0067] ,

[0068] in, is the corrected output signal, indicating the final voltage signal obtained after abnormal fluctuation correction, that is, the step voltage digital signal after intelligent processing; is the window size of the regression model, which is set to the time span of processing historical data, i.e., the voltage signal data after dynamic weighting and noise reduction within a dynamic window; is the adaptive regression coefficient; is an exponential decay coefficient, which is used to control the influence of historical data and decay the weight of older data; yes The step voltage signal after multi-level filtering and weighted noise reduction at each moment; Indicates forward tracking A time point with a time step length, that is, a historical time within a dynamic window; is the time interval used to define the time step of the adaptive regression model and helps calculate the weighted average.

[0069] S2. Calculate the real-time step voltage value based on the step voltage digital signal after intelligent processing to obtain the step voltage value sequence, merge the step voltage value sequence to obtain the step voltage value sequence set; and analyze the change trend of the step voltage in the step voltage value sequence set through the time domain-frequency domain mixed similarity trend matrix algorithm to obtain the step voltage data and fluctuation analysis results, and then monitor the step voltage data in real time according to the preset dynamic threshold.

[0070] Based on the step voltage digital signal after intelligent processing, the real-time step voltage value is calculated according to the voltage difference between every two adjacent monitoring points as a step voltage value, and stored in an array according to time series to obtain a step voltage value sequence, and the step voltage value sequences are further merged to obtain a step voltage value sequence set.

[0071] Furthermore, the time domain-frequency domain hybrid similarity trend matrix algorithm is used to analyze the step voltage change trend of the step voltage value sequence set to obtain the step voltage data and fluctuation analysis results, and then the step voltage data and fluctuation analysis results are monitored in real time according to the preset dynamic threshold. The time domain-frequency domain hybrid similarity trend matrix algorithm combines the time series information with the frequency domain characteristics by establishing a dynamic time domain and frequency domain similarity matrix, gradually calculates the trend quantitative index of the step voltage, and finally realizes the detection of abnormal fluctuations through dynamic threshold analysis. The specific implementation process is as follows:

[0072] First, each step voltage value sequence in the step voltage value sequence set is mixed in the time domain and frequency domain to generate a dynamic time domain-frequency domain mixed matrix , which is used to maintain both time series information and frequency information. The formula for this matrix is:

[0073] ,

[0074] in, is a time-frequency domain mixing matrix, representing the and frequency Step voltage mixing characteristics under ; For time The sequence of step voltage values ​​at ; It's time and frequency The cosine components acting together are used to capture the time domain features; For time The sequence of step voltage values ​​at ; Represents the frequency domain component, at time The role of is used to capture frequency domain features; It is a time-integrated variable, which indicates the time-integrated range of the step voltage signal and is used to calculate the change of the step voltage signal over time. Through the time-frequency domain hybrid matrix, the instantaneous signal of the step voltage and its characteristics at different frequencies can be captured together to form a joint feature space of the time-frequency domain, thereby providing complete input data for subsequent similarity calculations.

[0075] Furthermore, in order to analyze the changing trend of the step voltage signal at different time points, the time domain-frequency domain hybrid matrix Construct a similarity matrix , to quantify the feature similarity between time points, thereby reflecting the correlation characteristics of the step voltage signal at different time points. The calculation formula for any element in the similarity matrix is:

[0076] ,

[0077] in, Indicates time point and The similarity between them reflects the matching degree of the two time points in the time-frequency domain mixed characteristics; and is the time domain-frequency domain mixed matrix at time point and The value of Represents the difference in step voltage characteristics between different time points; is the similarity decay parameter, which controls the similarity decay rate; is the total number of frequency components; It is a distance-based similarity measure that uses an exponential decay function, which can effectively reduce the impact of large differences in time points on similarity; Represents the dot product of the time-frequency domain mixing matrix at all frequencies, which is used to calculate the feature matching degree in the frequency domain;

[0078] After generating the similarity matrix, in order to better capture the dynamic trend of the step voltage, the similarity matrix is ​​mapped to the trend quantification index , used to represent the gradual change of step voltage over time. Its calculation formula is as follows:

[0079] ,

[0080] in, For a specific time The trend quantitative index of the step voltage at the position is used to obtain the trend information of the step voltage signal by integrating the similarity characteristics in the time window and frequency dimensions; It is the time window size of trend calculation, which controls the range of cumulative similarity and ensures that the trend quantitative indicators not only represent a single time point, but also include the overall changes of surrounding time points; Indicates time point With time point similarities between; is the attenuation coefficient, which is used to reduce the impact of data far away from the current time point and further enhance the locality of the trend; It is the integration of the similarity matrix in frequency, which represents the aggregation of similarity features in the entire frequency domain, thus ensuring the sensitivity of trend quantitative indicators to signal fluctuations; is the maximum frequency, indicating the upper limit of the integral range. The calculation of the trend quantification index combines the similarity accumulation in the time domain and the frequency domain, so that the trend quantification index can fully describe the change trend of the step voltage signal.

[0081] Getting trend quantitative indicators Finally, in order to identify the abnormal fluctuation of step voltage, the trend quantitative index is determined by the dynamic threshold method. The setting depends on the mean, standard deviation and trend change rate of the trend quantitative indicator. The specific calculation formula is as follows:

[0082] ,

[0083] in, It's in time The dynamic threshold value represents the abnormal judgment standard of the step voltage at this time point; Is the trend quantitative indicator at the current time The average value before, indicating the step voltage trend quantitative index at the current time The previous average trend level; is the sensitivity adjustment parameter, which controls the looseness of the dynamic threshold; It is the weight coefficient of the trend change rate, ensuring that the dynamic threshold is adjusted dynamically as the trend changes; The central difference is used to approximate the instantaneous rate of change of the step voltage signal; For a specific time Quantitative indicator of the trend of step voltage at For a specific time Quantitative indicator of the trend of step voltage at For in time The trend quantitative index of the step voltage at the step point. By setting the dynamic threshold, the significant fluctuation of the step voltage can be captured in real time, and sensitive detection of abnormal signals can be achieved.

[0084] When the trend quantitative indicator Dynamic Threshold Exceeded When the voltage is abnormal, it is considered that there is an abnormal step voltage at that time point. The records and trend quantification indicators of the abnormal points The result is taken as output for further response by the early warning module. The final output data contains trend quantitative indicators , used to describe the change of step voltage trend; specifically: when When the value of rises or falls rapidly in a short period of time, it means that the step voltage signal has a violent fluctuation, indicating that the step voltage has a short-term instability problem, and an early warning begins; If the average value continues to rise, it means that the step voltage signal is in an upward trend; on the contrary, if If the average value of continues to decrease, it indicates that the step voltage signal tends to decrease or stabilize. When the instantaneous change rate is much greater than 0, it indicates that the step voltage trend change is accelerating; when the instantaneous change rate is much less than 0, it indicates that the step voltage trend change is rapidly decreasing; when the instantaneous change rate is close to 0, it indicates that the step voltage trend change remains stable, and the step voltage trend change is small or unchanged.

[0085] In summary, a step voltage online monitoring system and method are completed.

[0086] The order of the embodiments of the invention is for description only and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0087] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

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

Claims

1. A step voltage online monitoring method, characterized in that: The following steps are involved: S1. Collect step voltage data in real time, convert the step voltage data into analog voltage signals, and then convert the analog voltage signals into digital voltage signals. Introduce multi-level collaborative filtering and trend capture algorithms to intelligently process the digital voltage signals, and obtain the step voltage digital signals after intelligent processing. The multi-level collaborative filtering and trend capture algorithms process the digital voltage signals in two dimensions, time and frequency, based on filters of different levels and recursive structures, including: basic filtering, dynamic weighted noise reduction, trend tracking, and abnormal fluctuation correction processing. In the process of implementing the multi-level collaborative filtering and trend capture algorithm, the digital voltage signal is subjected to multi-level filtering to obtain the filtered voltage signal output. The specific implementation is as follows: ， in, It is The voltage signal output after level filtering; is the initial step voltage signal; is the direct signal input weight of the current layer; It is the cumulative weight of each level of filtering on the historical signal; It is The historical signal of the level filter output; is the sampling period, which represents the time step between different filter layers; It is Adaptive adjustment coefficient of the historical signal output by the first-order filtering; yes The step voltage signal at the moment; It is The estimated value of the standard deviation of the voltage signal after level filtering; S2. Calculate the real-time step voltage value based on the step voltage digital signal after intelligent processing to obtain a step voltage value sequence, merge the step voltage value sequence to obtain a step voltage value sequence set; and analyze the change trend of the step voltage in the step voltage value sequence set through a time domain-frequency domain hybrid similarity trend matrix algorithm to obtain step voltage data and fluctuation analysis results, and then monitor the step voltage data and fluctuation analysis results in real time according to a preset dynamic threshold; In the process of realizing the time domain-frequency domain hybrid similarity trend matrix algorithm, each step voltage value sequence in the step voltage value sequence set is subjected to a hybrid processing of the time domain and the frequency domain to generate a dynamic time domain-frequency domain hybrid matrix; A similarity matrix is ​​constructed based on a dynamic time domain-frequency domain hybrid matrix, and the similarity matrix is ​​mapped to a trend quantification index to obtain a trend quantification index of the step voltage; the trend quantification index of the step voltage is used to represent the changing trend of the step voltage.

2. A step voltage online monitoring method according to claim 1, characterized in that: The S1 specifically includes: In the process of realizing the multi-level collaborative filtering and trend capture algorithm, the voltage signal output after multi-level filtering is subjected to dynamic weighted noise reduction processing in the time domain to obtain the voltage signal output after dynamic weighted noise reduction.

3. A step voltage online monitoring method according to claim 2, characterized in that: The S1 specifically includes: In the process of implementing the multi-level collaborative filtering and trend capture algorithm, the voltage signal output after dynamic weighted noise reduction is subjected to an overall trend analysis, the abnormal pattern of voltage change is identified, and abnormal fluctuation correction processing is performed to obtain the step voltage digital signal after intelligent processing.

4. The method for online monitoring of step voltage according to claim 1, characterized in that: The S2 specifically includes: In the process of implementing the time domain-frequency domain hybrid similarity trend matrix algorithm, the trend quantification index is judged by the dynamic threshold method; the dynamic threshold is set based on the mean, standard deviation and trend change rate of the trend quantification index. When the trend quantification index exceeds the dynamic threshold, it is considered that there is an abnormal step voltage, and the abnormal step voltage data and trend quantification index are output.

5. A step voltage online monitoring system, applied to the step voltage online monitoring method according to claim 1, characterized in that: Includes the following parts: Step voltage monitoring sensor module, signal processing module, data calculation and analysis module, data storage and transmission module, early warning module; The step voltage monitoring sensor module collects step voltage data in real time, converts the step voltage data into analog voltage signals, and sends the analog voltage signals to the signal processing module; The signal processing module converts the analog voltage signal into a digital voltage signal, performs intelligent processing on the digital voltage signal to obtain an intelligently processed step voltage digital signal, and sends the intelligently processed step voltage digital signal to the data calculation and analysis module; The data calculation and analysis module calculates the real-time step voltage value based on the step voltage digital signal after intelligent processing, and analyzes the change trend of the step voltage through the time domain-frequency domain hybrid similarity trend matrix algorithm to obtain the step voltage data and fluctuation analysis results, and sends the step voltage data and fluctuation analysis results to the data storage and transmission module; The data storage and transmission module stores and backs up the step voltage data and fluctuation analysis results, and sends the step voltage data and fluctuation analysis results to the early warning module; The early warning module monitors the step voltage data in real time according to the preset dynamic threshold, step voltage data and fluctuation analysis results, obtains the monitoring results, implements early warning based on the monitoring results, and performs online monitoring of the step voltage.

Images