Operating state monitoring system of interposition cabinet based on electrical interlocking box

Abstract

The application discloses a supervisory system for the operation state of a cubical switchgear based on an electrical interlocking box, and relates to the technical field of electrical equipment monitoring. The application solves the technical problem that there is a lack of effective collection and analysis means, and only an alarm can be sent after a fault occurs, so that predictive maintenance cannot be achieved. The application establishes an operation time baseline by using a statistical method, and establishes a vibration feature baseline by using an unsupervised learning algorithm. The baselines are automatically generated based on the health history data of the equipment itself, can adapt to individual differences between different equipment, make the monitoring results more accurate and reliable, fuse the operation time deviation degree and the vibration feature deviation degree, establish a two-dimensional state space, accurately distinguish vibration-sensitive abnormalities, time-sensitive abnormalities and serious abnormalities through cross-validation of the two baselines, and, in cooperation with frequency band energy analysis and segmented time analysis, can further infer specific fault types, so that early warning information is upgraded from simple abnormal alarm to an intelligent diagnostic report with guiding significance.

Description

Technical Field

[0001] This invention relates to the field of electrical equipment monitoring technology, specifically to a monitoring system for the operation status of a central switchgear based on an electrical interlock box. Background Technology

[0002] As an indispensable control and protection device in the power system, the accurate monitoring of the operating status of the intermediate frequency switchgear is crucial for the safety and stability of the power grid. Currently, monitoring systems for intermediate frequency switchgear mainly rely on the electrical contacts of auxiliary switches to determine the position of the handcart, the status of the grounding switch, and the opening and closing status of the circuit breaker.

[0003] Traditional solutions rely solely on a single auxiliary switch contact signal to determine the handcart's position. Over time, these auxiliary switches are prone to problems such as broken links, jamming, contact oxidation, or poor contact, leading to a misalignment between the electrical and mechanical positions – a major cause of operational errors.

[0004] Existing monitoring systems mostly stop at binary logic judgments of "in place" or "not in place," which cannot quantitatively reflect the dynamic characteristics of the mechanical operation process. They lack effective means of collecting and analyzing key information reflecting the health status of equipment, such as the gradual extension of operation time and abnormal impacts and vibrations during operation. At the same time, because they cannot obtain dynamic process data of mechanical operation, existing technologies can only issue alarms after a fault occurs, and cannot achieve predictive maintenance. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a mid-voltage switchgear operation status monitoring system based on electrical interlock boxes, which solves the problem of lacking effective data collection and analysis methods, only issuing alarms after a fault occurs, and being unable to achieve predictive maintenance.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a mid-voltage switchgear operation status monitoring system based on an electrical interlock box, comprising: The intelligent acquisition unit is used to collect the mechanical position signal and electrical contact signal of the central switch cabinet. It combines the mechanical position signal and electrical contact signal to determine whether the position of the central switch cabinet is correct and generates position correct information or position abnormal information. The normal monitoring and analysis unit is used to monitor and analyze the generated normal position information. It takes the operation command as the starting point of the timing and the stable change of the auxiliary contact in the final position state as the ending point of the timing. It records the operation time and simultaneously triggers the vibration sensor to record the original vibration waveform when the operation command is issued. It performs time domain and frequency domain analysis on the original vibration waveform, extracts key feature values, and transmits the obtained key feature values ​​to the operation status analysis unit. The operation status analysis unit is used to construct the health baseline of the central cabinet based on the acquired key characteristic values. The health baseline includes the operation time baseline and the vibration characteristic baseline, and transmits the two to the anomaly early warning analysis unit. The abnormal early warning analysis unit is used to perform abnormal monitoring and processing based on the acquired operation time baseline and vibration characteristic baseline. It compares the characteristic value of the current operation with the two baselines to obtain a quantified deviation index, and performs hierarchical early warning processing based on the deviation index. The hierarchical early warning processing information is then transmitted to the monitoring and management output unit. The monitoring and management output unit is used to display the acquired location anomaly information or hierarchical early warning processing information to the corresponding management personnel.

[0007] As a further aspect of the present invention, the method for determining the position of the central control cabinet by combining the integrated mechanical position signal and the electrical contact signal includes: When the handcart starts moving, monitoring of the two signals is initiated; for electrical contact signals, the on / off changes of auxiliary contacts are detected, and software filtering is used to continuously sample multiple times, and only when the states are consistent is it determined to be a valid change; For the mechanical position signal, a jitter-reducing filter is applied, and the two signals are aligned on the time axis, marking the change moments within the same operating cycle; Extract the contact state Sc corresponding to the electrical contact signal and the sensor state Ss corresponding to the mechanical position signal respectively, calculate the contact confidence Pc based on the contact action history, and dynamically calculate the sensor confidence Ps based on the sensor signal quality; Based on the values ​​of Pc, Ps, Sc, and Ss and their interrelationships, the final position state is output through a preset fusion decision rule.

[0008] As a further aspect of the present invention, the fusion decision rule includes: If Pc > 0.9 and Ps > 0.9 and Sc = Ss, then directly output Sc or Ss as the final position; If Pc>0.9 and Ps<0.5, then the contact signal is the main output Sc, and the sensor signal is marked as abnormal. If Pc>0.9 and Ps>0.9 but Sc≠Ss, the conflict handling mechanism is triggered. The original waveform data of the two signals are read to check if they are in the critical zone. If it is confirmed that they are not in the critical zone, the output position is unknown and a dual signal conflict alarm is triggered. If Pc < 0.3 and Ps < 0.3 but Sc = Ss, then output the common state and mark both signals as unreliable.

[0009] As a further aspect of the present invention, the method by which the normal monitoring and analysis unit extracts key feature values ​​includes: Raw acceleration data ax(t), ay(t), and az(t) along the X, Y, and Z axes are collected. The magnitude of the composite vector A(t) along the three axes is calculated, and A(t) is smoothed using a sliding window to obtain At. filtered (t); Then during operation period A filtered Find all local maxima in the (t) sequence and filter out those greater than the threshold Th. peak The maximum value among the values ​​is taken as the peak acceleration; For A filtered Generate the upper envelope E(t), starting the search from the time t0 when the operation command is issued. When E(t) first continuously exceeds Th... start The time is recorded as the vibration start time T. start From the peak time T peak The search then continues when E(t) first and lasts below T. stop The time is recorded as the vibration end time T. stop Calculate the vibration duration Tc=T stop -T start ; Extract from T start To T stop The vibration waveform segment is subjected to a Fast Fourier Transform to obtain the spectrum X(f), and the power spectrum P(f) is calculated. Preset key frequency bands and calculate the corresponding frequency band energy for each frequency band.

[0010] As a further aspect of the present invention, the operation status analysis unit constructs the operation time baseline in the following ways: A high-precision timer is integrated into the intelligent electrical interlock box, connecting the command signal, the initial movement signal A of the moving contact, and the final arrival signal B. The time from the issuance of the command signal to the change of the initial moving signal A of the moving contact is recorded as the response time T1, and the time from the initial moving signal A of the driven contact to the final position signal B is recorded as the travel time T2. Record T1 and T2 for each operation, use the extreme value removal average filtering method to obtain effective samples, calculate the mean and standard deviation of all effective samples, and set the normal range and warning range to establish an initial baseline.

[0011] As a further aspect of the present invention, the method by which the operating state analysis unit constructs the vibration characteristic baseline includes: The original vibration waveform is processed in real time to extract multidimensional feature values, including time-domain features, frequency-domain features, and time-frequency-domain features. The extracted feature vectors are normalized, and a Gaussian mixture model is used to perform cluster analysis on the normalized feature vectors so that the data from normal operation naturally cluster into one or more dense regions. The center point of each dense region is obtained and denoted as the cluster center point C. The Mahalanobis distance from each sample to its respective cluster center is calculated, and a preset quantile is taken as the boundary threshold D. th The vibration characteristic baseline is obtained based on the cluster center points, covariance matrix and boundary threshold.

[0012] As a further aspect of the present invention, the time-domain features include peak value, effective value, kurtosis and impulse factor; the frequency-domain features include the frequency band energy corresponding to the key frequency band and the centroid frequency; the time-frequency domain features are obtained by performing wavelet decomposition on the original vibration waveform and calculating the proportion of each frequency band energy to the total energy.

[0013] As a further aspect of the present invention, the anomaly early warning analysis unit performs anomaly monitoring processing in the following ways: According to the formula Calculate the deviation of operation time T current Indicates the current operation time. This represents the average operation time. Indicates the standard deviation of operation time; formula The vibration characteristic deviation was calculated. V current Let C represent the current vibration feature vector, and D be the cluster center. th This is the boundary threshold; Using the obtained operating time deviation and vibration characteristic deviation as two coordinate axes, a two-dimensional state space is established. Based on the different partitions where the operating time deviation and vibration characteristic deviation are located, corresponding to different equipment health states and fault types, hierarchical early warning processing is carried out based on different state descriptions, and hierarchical early warning processing information is generated.

[0014] As a further aspect of the present invention, the graded early warning processing includes: If the operation time deviation zone is normal and the vibration characteristic deviation zone is normal, then the corresponding health status will not trigger an alarm. If the operation time deviation zone is normal and the vibration characteristic deviation zone is of concern or abnormal, then the corresponding vibration sensitivity is abnormal, triggering a level two warning, and fault diagnosis is performed based on the frequency band energy anomaly. If the operation time deviation zone is marked as concerning or abnormal, and the vibration characteristic deviation zone is marked as normal, then the corresponding time-sensitive abnormality will trigger a level-two warning, and fault diagnosis will be performed based on the specific segment of the operation time. If the operation time deviation zone is marked as "concerned" or "abnormal", and the vibration characteristic deviation zone is marked as "concerned" or "abnormal", then a severe abnormality is detected, triggering a Level 1 warning. The most likely severe fault type is then output by combining historical data and the fault database.

[0015] This invention provides a mid-voltage switchgear operation status monitoring system based on an electrical interlock box. Compared with the prior art, it has the following advantages: This invention, by constructing a dual signal acquisition and intelligent decision-making mechanism, not only solves the problem of false position caused by the failure of a single signal source, but also achieves health diagnosis of the signal circuit itself by introducing contact confidence and sensor confidence. When there is a signal conflict, the system can intelligently distinguish whether it is a critical state, sensor failure or contact failure, avoiding the blindness of alarms caused by signal contradictions in traditional solutions, and truly realizing reliable confirmation of position status, eliminating the risk of misoperation from the source.

[0016] This invention employs statistical methods to establish an operating time baseline and an unsupervised learning algorithm to establish a vibration characteristic baseline. The baseline is automatically generated based on the equipment's own health history data, which can adapt to individual differences between different devices. At the same time, a temperature compensation mechanism is introduced to eliminate the influence of ambient temperature on operating time, avoiding false alarms caused by seasonal changes, making the monitoring results more accurate and reliable. The operating time deviation and vibration characteristic deviation are fused to establish a two-dimensional state space. Through cross-validation of the two baselines, vibration-sensitive anomalies, time-sensitive anomalies, and severe anomalies can be accurately distinguished. Combined with frequency band energy analysis and segmented time analysis, the specific fault type can be further inferred, upgrading the early warning information from a simple abnormal alarm to a guiding intelligent diagnostic report. Attached Figure Description

[0017] Figure 1 This is a block diagram of the central switch cabinet operation status monitoring system of the present invention; Figure 2 This is a flowchart illustrating the implementation of the central switch cabinet operation status monitoring system of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] First Embodiment Please see Figure 1 and Figure 2 This application provides a monitoring system for the operation status of a central switchgear based on an electrical interlock box, including: The intelligent data acquisition unit collects mechanical position signals and electrical contact signals from the central switchgear. The mechanical position signal is obtained by integrating a non-contact position sensor inside the electrical interlock box, in addition to retaining the original electrical circuit contacts. The electrical contact signal is obtained from the open / close position signal, located in the auxiliary switch box on the side or bottom of the circuit breaker body. The correct position of the central switchgear is determined by combining the mechanical and electrical position signals, and the specific determination method is as follows: When the trolley starts moving, monitoring of two signals is initiated. For electrical contact signals, the on / off changes of auxiliary contacts are detected and software filtering is applied. Only when the states are consistent after three consecutive samplings are a valid change is determined. For mechanical position signals, de-jitter filtering is also applied, and the two signals are aligned on the time axis to mark the change time within the same operating cycle. Next, the contact state Sc and sensor state Ss corresponding to the electrical contact signal and mechanical position signal are extracted respectively. The contact confidence level Pc is calculated based on the contact action history, and the sensor confidence level Ps is dynamically calculated based on the sensor signal quality. If Pc > 0.9 and Ps > 0.9 and Sc = Ss, then Sc or Ss is directly output as the final position. If Pc > 0.9 and Ps < 0.5, then Sc is output based on the contact signal, but the sensor signal is marked as abnormal. If Pc > 0.9 and Ps > 0.9 but Sc ≠ Ss, then the conflict handling mechanism is triggered. Immediately read the original waveform data of the two signals and check if they are in the critical zone. If they are not in the critical zone, the output position is unknown and a dual signal conflict alarm is triggered. If Pc < 0.3 and Ps < 0.3 but Sc = Ss, then output the common state and mark both signals as unreliable. Based on the above determination process, either correct location information or abnormal location information is generated. For abnormal location information, it is directly transmitted to the monitoring and management output unit, while for normal location information, it is transmitted to the normal monitoring and analysis unit.

[0020] The monitoring and management output unit is used to display the acquired location anomaly information to the corresponding management personnel.

[0021] Second Embodiment As a second embodiment of the present invention, it is implemented based on the first embodiment, and the difference from the first embodiment is as follows: The normal monitoring and analysis unit performs monitoring and analysis based on the generated normal position information. Specifically, it uses the operation command as the timing start point, such as the electrical signal for electric operation or the inductive switch signal installed near the operation hole during manual operation, and the stable change of the auxiliary contact in the final position state as the timing end point, such as the working / test position of the handcart or the open / closed position of the grounding switch. It simultaneously captures the command signal and the position signal, records the time difference between them using the timing module, i.e., the operation time T, and performs filtering to remove interference caused by contact jitter. When the operation command is issued, it synchronously triggers the vibration sensor to start recording a fixed-duration waveform as the original vibration waveform. It then performs time-domain and frequency-domain analysis on the original vibration waveform to extract key feature values, including peak acceleration, vibration duration, and energy spectral density. The obtained key feature values ​​are then transmitted to the operating status analysis unit, and the specific extraction method is as follows: Collect raw acceleration data ax(t), ay(t), and az(t) along the X, Y, and Z axes, and calculate the magnitude of the composite vector along the three axes: Simultaneously, a sliding window of length 5 points is used to smooth A(t) to eliminate high-frequency noise interference. The formula is as follows: ; For peak acceleration, A during the operating period filtered In the sequence (t), find all local maxima, where a local maximum is a value greater than the values ​​of the 10 points before and after it. From all local maxima, select those greater than a threshold Th. peak The value of , and Th peak =5×N floor , where N floor The noise floor is represented by the maximum value among the noise floors, which is taken as the peak acceleration PeakA for this operation and then normalized. Regarding the duration of vibration, for A filtered (t) Perform a Hilbert transform or use a simplified moving absolute value integral to generate the upper envelope E(t) of the signal. Start searching backward from the time t0 when the operation command is issued. When the envelope E(t) first continuously exceeds Th start =3×N floor Record this moment as the vibration start time T. start From the peak time T peak The search then begins when the envelope E(t) first and sustainably falls below Th. stop =2×N floor Record this moment as the vibration end time T. stop According to the formula Tc=T stop -T start The duration of vibration was calculated. For the out-of-spectral density, extract from T startTo T stop A complete vibration waveform segment, with a length of N sampling points (where the specific value of N is set by the operator), is subjected to an N-point Fast Fourier Transform to obtain the spectrum X(f), and the power spectrum is calculated. Based on the typical resonant frequency of the central power cabinet's mechanical structure, three key frequency bands are pre-defined: a low-frequency band (BL) of 20Hz-500Hz, a mid-frequency band (BM) of 500Hz-2kHz, and a high-frequency band (BH) of 2kHz-5kHz. The corresponding frequency band energy is calculated based on these key frequency bands. , , Simultaneously, normalization processing is performed, and a three-dimensional vector PSD=[EL, EM, EH] is output based on the obtained frequency band energy.

[0022] The operational status analysis unit is used to construct a health baseline for the central control cabinet based on the acquired key characteristic values. The health baseline includes an operating time baseline and a vibration characteristic baseline, and transmits both to the anomaly early warning analysis unit. The operating time baseline is constructed as follows: A high-precision timer is integrated into the intelligent electrical interlock box and connected to three types of signals, specifically including a command signal, an initial movement signal A of the moving contact, and a final position signal B. Then, the time from the issuance of the command signal to the change of the initial movement signal A of the moving contact is recorded as the response time T1, and the time from the initial movement signal A of the moving contact to the final position signal B is recorded as the travel time T2. Next, the response time T1 and travel time T2 for each operation were recorded. An extreme value removal averaging filter was used to remove abnormal peak values ​​caused by contact jitter, obtaining all valid samples and calculating the average of all valid samples. and standard deviation At the same time, set the normal range [ -2 , +2 and warning range +3 An initial baseline is established, and then a temperature compensation coefficient is introduced. The ambient temperature is recorded simultaneously with each operation. If T2 increases linearly as the temperature decreases, a compensation formula is established. T 标准 Indicates standard temperature; The vibration characteristic baseline is constructed by acquiring the original vibration waveform and processing it in real time to extract multidimensional feature values, including time-domain features, frequency-domain features, and time-frequency-domain features. The time-domain features include peak value, effective value, kurtosis, and impulse factor. The peak value represents the maximum absolute value of the waveform, the effective value represents the root mean square value of the waveform, and the kurtosis represents the impact of the waveform. Normal waveforms are approximately sinusoidal, with a kurtosis of ≈3. When early spalling or cracks occur, the kurtosis will increase significantly. The impulse factor represents the ratio of the peak value to the absolute average value. Then, the frequency band energy and centroid frequency corresponding to the key frequency band are acquired, and the centroid frequency represents the centroid position of the spectrum. At the same time, wavelet decomposition is performed on the original vibration waveform, and the proportion of each frequency band energy to the total energy is calculated. After extraction, a feature vector containing 15-20 feature values ​​is generated. Next, the obtained feature vectors are normalized, and a Gaussian mixture model is used to perform cluster analysis on the normalized feature vectors. The data in normal operation will naturally cluster into one or more dense regions. Then, the center point of each dense region is obtained and recorded as the cluster center point. At the same time, the Mahalanobis distance from each sample to its respective cluster center is calculated, and the 95th quantile is taken as the boundary threshold. Based on the obtained cluster center points, covariance matrix and boundary threshold, the vibration feature baseline is obtained.

[0023] Third Embodiment As a third embodiment of the present invention, it is implemented based on the second embodiment, and the difference from the second embodiment is as follows: Anomaly warning and analysis unit: This unit is used to perform anomaly monitoring and processing based on the acquired operation time baseline and vibration characteristic baseline. It compares the characteristic value of the current operation with the two baselines to obtain a quantified deviation index, which includes operation time deviation and vibration characteristic deviation. Operation time deviation is based on the formula Calculations are performed to obtain the operation time deviation. T current Indicates the current operation time. This represents the average operation time. This represents the standard deviation of operation time. Both are read from the healthy baseline of operation time and have corresponding thresholds set. If ≤2, it is classified as the normal region; if 2< If ≤3, it is classified as a region of interest. If the value is greater than 3, it is classified as an abnormal area; Vibration characteristic deviation is based on the formula The vibration characteristic deviation was calculated. And the formula for calculating the Mahalanobis distance is: ,in Denotes the inverse matrix of the covariance matrix. V represents the transpose symbol. current This represents the current vibration feature vector, extracted from the original vibration waveform. C is the cluster center, and D... th Both are boundary thresholds, extracted from the vibration feature baseline, and corresponding thresholds are set. If ≤1, it is classified as the normal region; if 1 < If ≤1.5, it is classified as a region of interest. If the value is greater than 1.5, it is classified as an abnormal area; Using the obtained operating time deviation and vibration characteristic deviation as two coordinate axes, a two-dimensional state space is established. Different coordinate regions correspond to different equipment health states and fault types. If both the operating time deviation and vibration characteristic deviation regions are normal, it corresponds to a healthy state. If both the operating time deviation and vibration characteristic deviation regions are of concern or abnormal, it corresponds to a vibration-sensitive abnormality. If both the operating time deviation and vibration characteristic deviation regions are of concern or abnormal, it corresponds to a time-sensitive abnormality. If both the operating time deviation and vibration characteristic deviation regions are of concern or abnormal, it corresponds to a severe abnormality. Based on the different state descriptions described above, a graded early warning process is implemented. For healthy states, no early warning is triggered. For vibration-sensitive anomalies, a level-two early warning is triggered, and fault diagnosis is performed based on the frequency band energy anomalies. For example, if the high-frequency band energy increases significantly, it is inferred to be poor lubrication or early wear; if the mid-frequency band energy increases significantly, it is inferred to be poor gear meshing or bearing failure; if the low-frequency band energy increases significantly, it is inferred to be loose components. For time-sensitive anomalies, a level-two early warning is triggered, and fault diagnosis is performed based on the specific segments of the analysis operation time. For example, if the response time T1 is prolonged, it is inferred to be a slow response of the control loop; if the travel time T2 is prolonged, it is inferred to be an increase in the resistance of the transmission mechanism. For severe anomalies, a level-one early warning is triggered. By combining historical data and fault database matching, the most likely severe fault type is output, and the generated graded early warning processing information is transmitted to the monitoring and management output unit.

[0024] The monitoring and management output unit is used to display the acquired hierarchical early warning information to the corresponding management personnel.

[0025] Fourth embodiment As a fourth embodiment of the present invention, the focus is on combining the implementation processes of the first, second and third embodiments.

[0026] Some of the data in the above formulas are numerical calculations with dimensions removed, and the contents not described in detail in this specification are all prior art known to those skilled in the art.

[0027] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A mid-voltage switchgear operation status monitoring system based on electrical interlock boxes, characterized in that, include: The intelligent acquisition unit is used to collect the mechanical position signal and electrical contact signal of the central switch cabinet. It combines the mechanical position signal and electrical contact signal to determine whether the position of the central switch cabinet is correct and generates position correct information or position abnormal information. The normal monitoring and analysis unit is used to monitor and analyze the generated normal position information. It takes the operation command as the starting point of the timing and the stable change of the auxiliary contact in the final position state as the ending point of the timing. It records the operation time and simultaneously triggers the vibration sensor to record the original vibration waveform when the operation command is issued. It performs time domain and frequency domain analysis on the original vibration waveform, extracts key feature values, and transmits the obtained key feature values ​​to the operation status analysis unit. The operation status analysis unit is used to construct the health baseline of the central cabinet based on the acquired key characteristic values. The health baseline includes the operation time baseline and the vibration characteristic baseline, and transmits the two to the anomaly early warning analysis unit. The abnormal early warning analysis unit is used to perform abnormal monitoring and processing based on the acquired operation time baseline and vibration characteristic baseline. It compares the characteristic value of the current operation with the two baselines to obtain a quantified deviation index, and performs hierarchical early warning processing based on the deviation index. The hierarchical early warning processing information is then transmitted to the monitoring and management output unit. The monitoring and management output unit is used to display the acquired location anomaly information or hierarchical early warning processing information to the corresponding management personnel.

2. The monitoring system for the operation status of a central switchgear based on an electrical interlock box according to claim 1, characterized in that, The methods for determining the position of the central control cabinet by combining the integrated mechanical position signal and electrical contact signal include: When the handcart starts moving, monitoring of the two signals is initiated; for electrical contact signals, the on / off changes of auxiliary contacts are detected, and software filtering is used to continuously sample multiple times, and only when the states are consistent is it determined to be a valid change; For the mechanical position signal, a jitter-reducing filter is applied, and the two signals are aligned on the time axis, marking the change moments within the same operating cycle; Extract the contact state Sc corresponding to the electrical contact signal and the sensor state Ss corresponding to the mechanical position signal respectively, calculate the contact confidence Pc based on the contact action history, and dynamically calculate the sensor confidence Ps based on the sensor signal quality; Based on the values ​​of Pc, Ps, Sc, and Ss and their interrelationships, the final position state is output through a preset fusion decision rule.

3. The mid-voltage switchgear operation status monitoring system based on an electrical interlock box according to claim 2, characterized in that, The fusion decision rules include: If Pc > 0.9 and Ps > 0.9 and Sc = Ss, then directly output Sc or Ss as the final position; If Pc>0.9 and Ps<0.5, then the contact signal is the main output Sc, and the sensor signal is marked as abnormal. If Pc>0.9 and Ps>0.9 but Sc≠Ss, the conflict handling mechanism is triggered. The original waveform data of the two signals are read to check if they are in the critical zone. If it is confirmed that they are not in the critical zone, the output position is unknown and a dual signal conflict alarm is triggered. If Pc < 0.3 and Ps < 0.3 but Sc = Ss, then output the common state and mark both signals as unreliable.

4. The mid-voltage switchgear operation status monitoring system based on electrical interlocking boxes according to claim 1, characterized in that, The normal monitoring and analysis unit extracts key feature values ​​in the following ways: Raw acceleration data ax(t), ay(t), and az(t) along the X, Y, and Z axes are collected. The magnitude of the composite vector A(t) along the three axes is calculated, and A(t) is smoothed using a sliding window to obtain At. filtered (t); Then during operation period A filtered Find all local maxima in the (t) sequence and filter out those greater than the threshold Th. peak The maximum value among the values ​​is taken as the peak acceleration; For A filtered Generate the upper envelope E(t), starting the search from the time t0 when the operation command is issued. When E(t) first continuously exceeds Th... start The time is recorded as the vibration start time T. start From the peak time T peak The search then continues when E(t) first and lasts below T. stop The time is recorded as the vibration end time T. stop Calculate the vibration duration Tc=T stop -T start ; Extract from T start To T stop The vibration waveform segment is subjected to a Fast Fourier Transform to obtain the spectrum X(f), and the power spectrum P(f) is calculated. Preset key frequency bands and calculate the corresponding frequency band energy for each frequency band.

5. The mid-voltage switchgear operation status monitoring system based on an electrical interlock box according to claim 1, characterized in that, The operation status analysis unit constructs the operation time baseline in the following ways: A high-precision timer is integrated into the intelligent electrical interlock box, connecting the command signal, the initial movement signal A of the moving contact, and the final arrival signal B. The time from the issuance of the command signal to the change of the initial moving signal A of the moving contact is recorded as the response time T1, and the time from the initial moving signal A of the driven contact to the final position signal B is recorded as the travel time T2. Record T1 and T2 for each operation, use the extreme value removal average filtering method to obtain effective samples, calculate the mean and standard deviation of all effective samples, and set the normal range and warning range to establish an initial baseline.

6. The mid-voltage switchgear operation status monitoring system based on an electrical interlock box according to claim 1, characterized in that, The operational status analysis unit constructs the vibration characteristic baseline in the following ways: The original vibration waveform is processed in real time to extract multidimensional feature values, including time-domain features, frequency-domain features, and time-frequency-domain features. The extracted feature vectors are normalized, and a Gaussian mixture model is used to perform cluster analysis on the normalized feature vectors so that the data from normal operation naturally cluster into one or more dense regions. The center point of each dense region is obtained and denoted as the cluster center point C. The Mahalanobis distance from each sample to its respective cluster center is calculated, and a preset quantile is taken as the boundary threshold D. th The vibration characteristic baseline is obtained based on the cluster center points, covariance matrix and boundary threshold.

7. The mid-voltage switchgear operation status monitoring system based on an electrical interlock box according to claim 6, characterized in that, The time-domain features include peak value, RMS value, kurtosis, and impulse factor; Frequency domain characteristics include the band energy corresponding to the key frequency band and the centroid frequency; The time-frequency domain characteristics are obtained by performing wavelet decomposition on the original vibration waveform and calculating the proportion of energy in each frequency band to the total energy.

8. The mid-voltage switchgear operation status monitoring system based on an electrical interlock box according to claim 1, characterized in that, The anomaly early warning analysis unit performs anomaly monitoring and processing in the following ways: According to the formula Calculate the deviation of operation time T current Indicates the current operation time. This represents the average operation time. Indicates the standard deviation of operation time; formula The vibration characteristic deviation was calculated. V current Let C represent the current vibration feature vector, and D be the cluster center. th This is the boundary threshold; Using the obtained operating time deviation and vibration characteristic deviation as two coordinate axes, a two-dimensional state space is established. Based on the different partitions where the operating time deviation and vibration characteristic deviation are located, corresponding to different equipment health states and fault types, hierarchical early warning processing is carried out based on different state descriptions, and hierarchical early warning processing information is generated.

9. The mid-voltage switchgear operation status monitoring system based on an electrical interlock box according to claim 8, characterized in that, The tiered early warning process includes: If the operation time deviation zone is normal and the vibration characteristic deviation zone is normal, then the corresponding health status will not trigger an alarm. If the operation time deviation zone is normal and the vibration characteristic deviation zone is of concern or abnormal, then the corresponding vibration sensitivity is abnormal, triggering a level two warning, and fault diagnosis is performed based on the frequency band energy anomaly. If the operation time deviation zone is marked as concerning or abnormal, and the vibration characteristic deviation zone is marked as normal, then the corresponding time-sensitive abnormality will trigger a level-two warning, and fault diagnosis will be performed based on the specific segment of the operation time. If the operation time deviation zone is marked as "concerned" or "abnormal", and the vibration characteristic deviation zone is marked as "concerned" or "abnormal", then a severe abnormality is detected, triggering a Level 1 warning. The most likely severe fault type is then output by combining historical data and the fault database.

Images