System and method for automatically monitoring switch pressing

Abstract

The invention provides a system and method for automatically monitoring switch pressing, and relates to the technical field of switch equipment monitoring, and the method comprises the steps: collecting a body acoustic signal and a mechanism acoustic signal in the pressing process of switch equipment through an acoustic sensor, carrying out the adaptive segmentation processing of the signals, extracting voiceprint feature parameters in each time period, and carrying out the detection of the voiceprint feature parameters. And constructing a double-domain voiceprint characteristic spectrum of the switch equipment, calculating the phase synchronization degree, constructing a collaborative voiceprint model of the switch equipment, further performing dynamic comparison with a standard voiceprint library, positioning a fault type, and outputting early warning information. And the dynamic acoustic characteristics in the operation process of the switch equipment can be reflected more comprehensively and accurately. And meanwhile, the collaborative voiceprint model is established by calculating the phase synchronization degree, and the dynamic matching relationship between the ontology and the mechanism is analyzed, so that the accuracy of fault positioning and the timeliness of early warning are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of switchgear monitoring technology, and more specifically, to a system and method for automatically monitoring switch voltage. Background Technology

[0002] With the rapid development of modern power systems, switchgear, as one of the core components of power systems, has received widespread attention for its operational reliability and condition monitoring technology. During the opening and closing operations of switchgear, key components such as energy storage mechanisms, transmission mechanisms, and locking mechanisms work together, and their condition directly affects the safety and stability of the power system. Traditional switchgear condition monitoring methods mainly rely on the acquisition and analysis of electrical characteristic parameters, such as current, voltage, and contact resistance. However, these methods have limitations and cannot comprehensively reflect the mechanical state and acoustic behavior characteristics of switchgear during dynamic operation, especially for potential faults caused by mechanical wear, spring fatigue, and transmission jamming, where traditional methods lack sensitivity. Furthermore, in recent years, vibration signal-based monitoring methods have been gradually applied in switchgear fault diagnosis, using accelerometers to collect vibration signals and analyze their spectral characteristics to determine the equipment condition. However, the frequency distribution of vibration signals is usually limited to the low and medium frequency bands and is greatly affected by external environmental noise, making it difficult to guarantee diagnostic accuracy under complex operating conditions.

[0003] Existing research indicates that acoustic signals, as high-frequency and sensitive state parameters, can reflect minute changes in mechanical motion, making them particularly suitable for monitoring the condition of dynamically operating components such as switchgear. Some technical solutions have begun to utilize acoustic sensors to collect and analyze acoustic signals during the operation of switchgear, for example, monitoring abnormal noises through a single acoustic characteristic parameter (such as sound pressure amplitude). However, these methods have significant shortcomings in data processing and fault mode analysis: First, existing methods often employ fixed-window segmentation for acoustic signal processing, which cannot adapt to the dynamic changes in signals, resulting in insufficient accuracy in feature extraction; second, the multidimensional features of acoustic signals (such as harmonic characteristics, energy distribution, and time-frequency variations) have not been effectively fused and analyzed, leading to an inability to comprehensively reflect the operating status of the switchgear; third, the lack of multi-domain collaborative analysis methods based on acoustic signals makes it difficult to reveal the dynamic coordination characteristics between the switchgear body and the operating mechanism, thus hindering the timely and accurate location of complex faults. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention is proposed. This invention provides a system and method for automatically monitoring switch voltage, which can, to some extent, solve the problem that existing technologies lack flexibility in processing acoustic signals during the dynamic operation of switchgear, leading to the inability to accurately extract key characteristic parameters and promptly locate fault types and output early warnings.

[0005] According to one aspect of the present invention, an automatic monitoring method for switch voltage is provided, comprising: Acoustic sensors are installed to collect the acoustic signals of the switchgear itself and the mechanism during the pressurization process. Adaptive segmentation processing is performed on the acoustic signals of the body and the acoustic signals of the mechanism to extract the acoustic feature parameters in each time period and construct a dual-domain acoustic feature map of the switching equipment. Based on the dual-domain acoustic signature feature map, the phase synchronization degree is calculated, and a cooperative acoustic signature model of the switching equipment is constructed based on the phase synchronization degree. The feature parameters of the collaborative voiceprint model are dynamically compared with the standard voiceprint library to locate the fault type and output early warning information.

[0006] Furthermore, the acoustic sensor includes a primary acoustic sensor and an auxiliary acoustic sensor; The main acoustic sensor is installed on the switchgear body to collect the acoustic signals of the switchgear body during the pressurization process; The auxiliary acoustic sensor is installed at the operating mechanism to collect the acoustic signals of the mechanism during the pressure application process of the switching equipment.

[0007] Furthermore, the adaptive segmentation process employs an improved dynamic segmentation algorithm, which determines the segmentation point by calculating the local energy change characteristics and signal abrupt change characteristics of the signal.

[0008] Furthermore, the improved dynamic segmentation algorithm is achieved by setting the time series as... Improved local energy change rate The calculation formula is: in, At the current time point, This is the local window length, used to determine the calculation range; the initial value is 20ms. The variable to be summed represents the time points within the window. The attenuation coefficient is... The signal energy value. These are the mutation weighting coefficients; when At that time, The point is marked as a segmentation point, where The adaptive threshold is expressed as: in, As the baseline threshold, The standard deviation of the current time window. For reference standard deviation, This is a correction factor.

[0009] Furthermore, the voiceprint characteristic parameters include harmonic ratio, energy entropy value, and sound pressure amplitude.

[0010] Furthermore, the harmonic ratio is calculated using an improved harmonic analysis method, expressed as: in, for Weighting coefficients for first harmonics, The complex spectrum is the frequency component. This represents the total harmonic number.

[0011] Furthermore, the energy entropy value is calculated using an improved wavelet packet decomposition method, assuming the first... Layer The energy of each frequency band is Normalized energy is The formula is expressed as: The improved formula for calculating energy entropy is: in, To improve the energy entropy value, The number of wavelet packet decomposition levels. Frequency band number, from 1 to , indicating different frequency bands For the first The total number of frequency bands after layer wavelet packet decomposition. For the frequency band weighting function, This is the weighting adjustment coefficient.

[0012] Furthermore, the sound pressure amplitude is calculated using a dynamic reference correction method, expressed by the following formula: Where RMS(t) is the root mean square value. For reference sound pressure level, For dynamic correction coefficients, The standard deviation of the current time window. This is the reference standard deviation under standard operating conditions. This is the correction factor.

[0013] Furthermore, the dual-domain voiceprint feature map includes a time-domain feature curve and a frequency-domain feature curve; The construction of the time-domain characteristic curve involves plotting a waveform curve with time as the horizontal axis and sound pressure amplitude as the vertical axis, while simultaneously overlaying and displaying the changing trend of energy entropy value to construct a two-parameter time-domain characteristic curve. The frequency domain characteristic curve is constructed with time as the horizontal axis and frequency as the vertical axis. The energy intensity is represented by the color intensity, and the harmonic ratio change trajectory is marked.

[0014] According to another aspect of the present invention, an automatic monitoring system for monitoring switch voltage is provided, comprising: The acquisition module is configured to use acoustic sensors to acquire the acoustic signals of the device itself and the mechanism during the pressurization process of the switching equipment; The segmentation module performs adaptive segmentation processing on the acoustic signals of the body and the acoustic signals of the mechanism, extracts the acoustic feature parameters in each time period, and constructs a dual-domain acoustic feature map of the switching equipment. The calculation module calculates the phase synchronization degree based on the dual-domain acoustic signature feature map, and constructs a cooperative acoustic signature model of the switching equipment based on the phase synchronization degree. The early warning module dynamically compares the feature parameters of the collaborative voiceprint model with the standard voiceprint library to locate the fault type and output early warning information.

[0015] Compared with existing technologies, this invention collects acoustic signals from the switchgear body and mechanism using primary and secondary acoustic sensors respectively, and extracts acoustic signature feature parameters using an adaptive segmentation processing method to construct a dual-domain acoustic signature feature map. This provides a more comprehensive and accurate reflection of the dynamic acoustic characteristics of the switchgear during operation. Simultaneously, by calculating phase synchronization to establish a cooperative acoustic signature model, the dynamic coordination relationship between the body and mechanism is analyzed, significantly improving the accuracy of fault location and the timeliness of early warning. Furthermore, this invention, combined with a dynamic comparison method using a standard acoustic signature library, not only improves diagnostic robustness under complex operating conditions but also enhances the intelligence level of switchgear operation status monitoring. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 This is a flowchart of an automatic monitoring method for monitoring switch voltage according to an embodiment of the present invention. Detailed Implementation

[0017] Hereinafter, exemplary embodiments according to the present invention will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments of the present invention. It should be understood that the present invention is not limited to the exemplary embodiments described herein.

[0018] As mentioned in the background section above, the existing technology has two main problems: First, during the dynamic operation of the switchgear, the acoustic signal processing method is inflexible and it is difficult to accurately extract key feature parameters, resulting in incomplete monitoring of the operating status; second, there is a lack of synergistic analysis of the dynamic coordination relationship between the switch body and the operating mechanism, making it impossible to locate the fault type and output early warning information in a timely and accurate manner.

[0019] Figure 1 This is a system block diagram of an automatic monitoring method for monitoring switch voltage according to an embodiment of the present invention. Figure 1 As shown, the monitoring method for automatic monitoring switch pressure includes: S1: Set acoustic sensors to collect the acoustic signals of the main body and mechanism of the switching equipment during the pressurization process.

[0020] Three main acoustic sensors are installed. Two of them are fixedly installed on the upper and lower ends of the switch body, respectively, at a distance of 50mm from the upper and lower ends of the switch body. The third main acoustic sensor is installed in the middle of the switch body, at the center line of the upper and lower ends.

[0021] The three main acoustic sensors are arranged in an equilateral triangle with sides of 300mm to achieve omnidirectional acquisition of acoustic signals from the switch body. The main acoustic sensors are ABC-100 wideband piezoelectric sensors with a sensitivity of 100mV / Pa, a sampling frequency of 100kHz, a frequency response range of 0.1Hz-50kHz, a signal-to-noise ratio greater than 70dB, and a dynamic range of 30-130dB.

[0022] Two auxiliary acoustic sensors are provided. One is installed at the transmission bearing of the switching equipment operating mechanism, 20mm from the center of the bearing, to collect the acoustic signals generated when the bearing moves. The other is installed at the energy storage spring, 30mm from the end of the spring, to collect the acoustic signals generated during the compression and release of the spring.

[0023] The auxiliary acoustic sensor is an XYZ-200 piezoelectric sensor with a sensitivity of 200mV / Pa, a sampling frequency of 50kHz, a frequency response range of 1Hz-25kHz, a signal-to-noise ratio greater than 75dB, and a dynamic range of 40-140dB. The auxiliary acoustic sensor is fixed to the operating mechanism via a magnetic base with an attraction force of no less than 50N, and the base surface is treated with rust prevention.

[0024] Both the main and auxiliary acoustic sensors are encapsulated in waterproof and dustproof housings with an IP65 protection rating to ensure stable operation under various environmental conditions. A silicone rubber gasket seals the sensor to the housing, and a dedicated waterproof connector is used for signal cable routing.

[0025] The sensor's output signal is transmitted to the data acquisition unit via a shielded cable. The shielded cable adopts a double-layer shielding structure, with the inner layer using copper foil shielding and the outer layer using metal braided mesh shielding. The grounding resistance of the shielding layer is less than 0.1Ω, effectively reducing the impact of external electromagnetic interference on signal acquisition.

[0026] The data acquisition unit employs a high-performance DSP processor to synchronously acquire both the body acoustic signals and the mechanism acoustic signals, with a sampling accuracy of 24 bits and a sampling clock error of less than ±1ppm. Simultaneously, the data acquisition unit integrates a real-time digital filter, capable of filtering out 50Hz power frequency interference and high-frequency noise above 25kHz. The filter order is configurable, supporting up to 128th-order FIR filtering.

[0027] Both the primary and secondary acoustic sensors require sensitivity calibration during installation. The calibration process consists of two steps: First, a 94dB, 1kHz standard acoustic signal is generated at a distance of 1 meter from the sensor's acoustic surface using a B&K 4231 sound level calibrator, and the sensor's output voltage value is recorded. Then, a wideband sound source is used to perform frequency response calibration within the 40Hz-20kHz range, obtaining the sensor's sensitivity correction coefficients at different frequencies. Sensor sensitivity calibration is performed quarterly, and the results are recorded in a dedicated calibration log. The sensor must be replaced when the sensitivity change exceeds ±1dB.

[0028] During installation, the mounting surfaces of the main acoustic sensor and the auxiliary acoustic sensor need to be surface treated to ensure that the flatness of the mounting surface is better than 0.1mm, and special thermal grease is used to improve the acoustic coupling between the sensor and the mounting surface.

[0029] S2: Adaptive segmentation processing is performed on the acoustic signal of the body and the acoustic signal of the mechanism to extract the acoustic feature parameters in each time period and construct a dual-domain acoustic feature map of the switching equipment.

[0030] Adaptive segmentation processing is performed on the acoustic signals of the body and the mechanism.

[0031] Specifically, an improved dynamic segmentation algorithm is used to determine the segmentation points by calculating the local energy change characteristics and signal abrupt change characteristics of the signal. Let the time series be... Improved local energy change rate The calculation formula is: in, At the current time point, This is the local window length, used to determine the calculation range; the initial value is 20ms. The variable to be summed represents the time points within the window. The attenuation coefficient is... The signal energy value. This represents the mutation weighting coefficient.

[0032] when At that time, The point is marked as a segmentation point, where The adaptive threshold is expressed as: in, As the baseline threshold, The standard deviation of the current time window. For reference standard deviation, This is a correction factor.

[0033] For each segmented signal, the window length adopts an improved adaptive window length, expressed as: in, The baseline window length is initially set to 20ms. For adjustment coefficients, For window adjustment functions, This is for adjusting the coefficient.

[0034] Based on the established segmented windows, voiceprint feature parameters, including harmonic ratio, energy entropy value and sound pressure amplitude, are extracted within each segmented window.

[0035] Harmonic ratio The calculation uses an improved harmonic analysis method, expressed as: in, for Weighting coefficients for first harmonics, The complex spectrum is the frequency component. This represents the total harmonic number.

[0036] Furthermore, weighting coefficients Considering the harmonic attenuation characteristics, the formula is expressed as: in, The attenuation factor is 0.1. The modulation depth.

[0037] The energy entropy value is calculated using an improved wavelet packet decomposition method. Let the first wavelet packet decomposition value be... Layer The energy of each frequency band is Normalized energy is The formula is expressed as: The improved formula for calculating energy entropy is: in, To improve the energy entropy value, The number of wavelet packet decomposition levels. Frequency band number, from 1 to , indicating different frequency bands For the first The total number of frequency bands after layer wavelet packet decomposition. For the frequency band weighting function, This is the weighting adjustment coefficient.

[0038] The sound pressure amplitude is calculated using a dynamic reference correction method, expressed by the following formula: Where RMS(t) is the root mean square value. For reference sound pressure level, For dynamic correction coefficients, The standard deviation of the current time window. This is the reference standard deviation under standard operating conditions. This is the correction factor.

[0039] A dual-domain voiceprint feature map is constructed based on the obtained voiceprint feature parameters.

[0040] The construction of the time-domain characteristic curve involves plotting the waveform curve with time as the horizontal axis and sound pressure amplitude as the vertical axis, while simultaneously overlaying and displaying the changing trend of energy entropy value, thus forming a two-parameter time-domain characteristic curve.

[0041] The frequency domain characteristic curve is constructed with time as the horizontal axis and frequency as the vertical axis. The energy intensity is represented by the color intensity, and the harmonic ratio change trajectory is marked.

[0042] Specifically, time-domain characteristic curves Feature fusion method is used: in, is the feature fusion coefficient.

[0043] Frequency domain characteristic curve The following was obtained through an improved time-frequency analysis method: in, It is a short-time Fourier transform. For adaptive enhancement function, This is the enhancement coefficient.

[0044] Based on the acquired feature curves, a dual-domain voiceprint feature map is constructed using a two-layer structure. The upper layer displays the time-domain feature curves, and the lower layer displays the frequency-domain feature curves, both sharing the same time reference. Simultaneously, feature markers are set within the map; when a feature parameter becomes abnormal, a red warning marker is displayed at the corresponding location.

[0045] S3: Based on the dual-domain acoustic signature feature map, calculate the phase synchronization degree, and construct a cooperative acoustic signature model of the switching equipment according to the phase synchronization degree.

[0046] Hilbert transforms were performed on the time-domain and frequency-domain characteristic curves in different frequency bands to obtain analytic signals. For the time-domain characteristic curve, its analytical signal expression is: in, This is the time-domain voiceprint feature curve. The instantaneous amplitude of the time-domain signal. The instantaneous phase of the time-domain signal.

[0047] The analytic signal of the energy trajectory of the frequency domain characteristic curve in frequency band f is: in, This is the frequency domain acoustic signature curve. For frequency domain signals in the frequency band Instantaneous amplitude, For frequency domain signals in the frequency band The instantaneous phase on.

[0048] Furthermore, an improved recursive quantization analysis method is used to calculate the phase synchronization degree. First, the phase difference matrix is ​​constructed: in, The phase difference between the time domain and frequency domain characteristics. , The parameter for the phase lock-in ratio. This indicates the phase difference modulo operation.

[0049] The phase difference matrix is ​​recursively quantized, and an adaptive threshold function is introduced to determine the significance of the phase difference: in, The phase difference threshold is dynamically adjusted. As the baseline threshold, This is the adjustment coefficient.

[0050] The phase synchronization index is defined as: in, The time window length, The attenuation coefficient is... For smoothing functions, This is the smoothing coefficient.

[0051] Based on the calculated phase synchronization index, a cooperative voiceprint model is constructed, including: Define the body-mechanism coupling strength: in, This is a dynamic adjustment coefficient. This refers to the frequency band weight.

[0052] And introduce a time correlation function: in, The time window length, This is the time decay coefficient.

[0053] Then, collaborative voiceprint model Represented as: in, For correction factors, For smoothing coefficients, This is an adaptive correction function.

[0054] S4: Dynamically compare the feature parameters of the collaborative voiceprint model with the standard voiceprint library to locate the fault type and output early warning information.

[0055] Based on the constructed collaborative acoustic signature model, the standard acoustic signature library is systematically layered according to the operating conditions of the switch pressurization process, including typical combinations of operating conditions such as energy storage status (rated energy storage / over-storage / under-storage), operation type (opening / closing / repeated operation), and ambient temperature (high temperature / normal temperature / low temperature). When the switch performs a pressurization action, the corresponding standard template library is called, and the dynamic changes in phase synchronization index (PSI) are calculated and compared in real time. When a change in operating condition is detected, such as from the energy storage stage to the opening / closing operation, the system switches to the corresponding standard template for comparison and analysis.

[0056] During dynamic comparison, a fixed-length sliding time window (the standard window length is the time of one complete opening and closing action) is used to continuously calculate the weighted average of the phase synchronization index (PSI). When the average synchronization index of three consecutive time windows is lower than a preset threshold, the coupling strength between the body and the mechanism is adjusted. Time-frequency analysis reveals that when the negative value of the first derivative consistently exceeds 30% of the standard operating time and lasts for more than three operating cycles, it indicates an abnormal pressing operation speed. When the variation of the second derivative within a single operating cycle exceeds twice the historical average, it indicates a sudden fault, such as mechanism jamming, spring breakage, or transmission obstruction. If the difference between the fluctuation frequency and the natural frequency of the switch body or operating mechanism is less than 10% and the duration exceeds one complete pressure cycle, then it is determined that there is a risk of pressure resonance.

[0057] Further analysis of time correlation functions The change pattern, if If the pressure drop exceeds 50% of the standard operating time within a single pressure test cycle, it indicates a sudden system failure, such as sudden failure of the energy storage spring, momentary breakage of the connecting rod, or momentary failure of the locking mechanism; if If the cumulative decrease exceeds 30% within ten consecutive pressure tests and shows an accelerating downward trend, it indicates that the pressure testing performance is deteriorating, such as due to poor lubrication, reduced spring force, or wear of the operating mechanism; if When the standard deviation exceeds 1.5 times the historical average within five consecutive pressure-pressurizing operations, it indicates that the pressure-pressurizing stability has deteriorated, and problems such as improper operation, inaccurate positioning, or unstable opening and closing times may occur.

[0058] Furthermore, combined with the adaptive correction function The response characteristics, if When the average of the second derivative term in the voltage regulation exceeds twice the historical level over three consecutive voltage suppression operations, it indicates abnormal voltage suppression dynamics, such as uneven energy storage speed, fluctuations in opening and closing speeds, or abnormal terminal buffering. If the correction amplitude increases continuously within five consecutive pressure tests and the increase exceeds 40%, stronger adaptive compensation is required, which may lead to problems such as excessive clearance in the operating mechanism or a significant increase in resistance torque; if The changing trend and If the correlation coefficient is less than -0.8 and continues for more than three consecutive operations, it indicates that the adaptive compensation effect is poor and there may be structural defects or serious performance degradation.

[0059] Furthermore, the local standard deviation of the phase difference is calculated. When the average of the five consecutive pressure-reduction operations increases by more than 40% compared to the historical level, it indicates a significant deterioration in system coordination. If the local standard deviation of the phase difference in a certain pressure-reduction operation exceeds twice the historical average, it is judged as a serious coordination failure. When the local standard deviation of the phase difference shows a peak value exceeding three times the historical average in the energy storage characteristic frequency band (30-50Hz) or the opening and closing characteristic frequency band (80-120Hz), it is used to locate the frequency range where the fault occurred, and then determine the specific faulty component.

[0060] Fault type localization is based on time-domain and frequency-domain characteristic curves. First, abnormal patterns in the time-domain characteristic curves are analyzed. When the curve exhibits regular periodic fluctuations, it indicates a vibration fault in the energy storage mechanism. The vibration frequency (within the energy storage motor's operating frequency range of 30-60Hz), amplitude (deviation from the normal energy storage current exceeding 50%), and duration (exceeding twice the standard energy storage time) are recorded. When the curve shows steep rises and falls with sharp peaks, it is identified as a switching impact fault, specifically manifested as an impact current exceeding twice the rated value, an action time less than 1 / 3 of the standard time, and a buffer time less than 50% of the design value. If the curve shows a unidirectional, continuous deviation from the reference value, and the stroke deviation exceeds 30% of the design value and shows a continuous increasing trend, it is determined to be a mechanism stroke deviation fault.

[0061] For the frequency domain characteristic curve, the frequency band is divided into three intervals: energy storage characteristic segment (0-100Hz), circuit breaker characteristic segment (100-500Hz), and buffer characteristic segment (above 500Hz). In the energy storage characteristic segment, the characteristic frequency components related to the energy storage mechanism are monitored. For example, if the amplitude of the motor operating frequency (standard value 45Hz) is abnormal (exceeding the reference value by 50%) or the sideband is enhanced (sideband amplitude exceeds the main frequency by 30%), it indicates that there is an anomaly in the energy storage system. In the circuit breaker characteristic segment, the mechanism's action characteristics are monitored. When there is a wideband energy increase (40% higher than the normal value) or a shift in the characteristic frequency (exceeding 10%), it indicates an abnormality in the circuit breaker action. In the buffer characteristic segment, the terminal buffer characteristics are analyzed. When the energy of the buffer frequency exceeds twice the warning value or there is a significant secondary impact, it indicates that there is a fault in the buffer mechanism.

[0062] Based on the above warning information, a tiered warning system is adopted: when the phase synchronization degree experiences a short-term, slight decrease (the decrease is less than 20% and occurs only during a single pressure test), and the body-mechanism coupling strength... When the fluctuation range is within the allowable range (energy storage time deviation not exceeding ±15%, and opening and closing time deviation not exceeding ±10%), a first-level warning is issued, requiring the pressure test interval to be shortened to 1 / 2 of the standard cycle. When the synchronization rate remains at a low level (a decrease of more than 20% and three consecutive abnormal suppressions), at the same time Significant anomalies occur (energy storage time deviation exceeds ±25% or opening / closing time deviation exceeds ±20%), or time correlation function... When an unfavorable trend is shown (operational stability decreases by more than 30%), a Level II warning is issued, and a special inspection is recommended to be carried out within 8 hours, along with a list of key areas to be inspected. When a clear fault characteristic pattern is detected, such as a continuous abnormality in PSI in a specific frequency band (the abnormality in the energy storage or circuit breaker characteristic frequency band continues to exceed 50%) or multiple characteristic parameters deteriorate simultaneously (more than 3 parameters exceed the warning value), a level 3 warning is issued, and a detailed fault diagnosis report is generated, including the fault type, specific fault location, development trend prediction, and the priority of suggested handling measures.

[0063] In summary, the system and method for automatically monitoring switch pressure based on embodiments of the present invention have been clarified. By setting primary and secondary acoustic sensors to collect acoustic signals from the switchgear body and mechanism respectively, and employing an adaptive segmentation processing method to extract acoustic signature feature parameters, a dual-domain acoustic signature feature map is constructed, which can more comprehensively and accurately reflect the dynamic acoustic characteristics of the switchgear during operation. Simultaneously, by calculating phase synchronization to establish a cooperative acoustic signature model, the dynamic coordination relationship between the body and mechanism is analyzed, significantly improving the accuracy of fault location and the timeliness of early warning. Furthermore, the present invention, combined with a dynamic comparison method using a standard acoustic signature library, not only improves diagnostic robustness under complex operating conditions but also enhances the intelligence level of switchgear operating status monitoring.

Claims

1. A method for automatically monitoring switch voltage, characterized in that, include: Acoustic sensors are installed to collect the acoustic signals of the switchgear itself and the mechanism during the pressurization process. Adaptive segmentation processing is performed on the acoustic signals of the body and the acoustic signals of the mechanism to extract the acoustic feature parameters in each time period and construct a dual-domain acoustic feature map of the switching equipment. Based on the dual-domain acoustic signature feature map, the phase synchronization degree is calculated, and a cooperative acoustic signature model of the switching equipment is constructed based on the phase synchronization degree. The feature parameters of the collaborative voiceprint model are dynamically compared with the standard voiceprint library to locate the fault type and output early warning information.

2. The method for monitoring automatic switch voltage as described in claim 1, characterized in that, The acoustic sensor includes a primary acoustic sensor and an auxiliary acoustic sensor; The main acoustic sensor is installed on the switchgear body to collect the acoustic signals of the switchgear body during the pressurization process; The auxiliary acoustic sensor is installed at the operating mechanism to collect the acoustic signals of the mechanism during the pressure application process of the switching equipment.

3. The method for monitoring automatic switch voltage as described in claim 1, characterized in that, The adaptive segmentation process employs an improved dynamic segmentation algorithm, which determines the segmentation points by calculating the local energy change characteristics and signal abrupt change characteristics of the signal.

4. The method for monitoring automatic switch voltage as described in claim 3, characterized in that, The improved dynamic segmentation algorithm is achieved by setting the time series as... Improved local energy change rate The calculation formula is: in, At the current time point, This is the local window length, used to determine the calculation range; the initial value is 20ms. The variable to be summed represents the time points within the window. The attenuation coefficient is... The signal energy value. These are the mutation weighting coefficients; when At that time, The point is marked as a segmentation point, where The adaptive threshold is expressed as: in, As the baseline threshold, The standard deviation of the current time window. For reference standard deviation, This is a correction factor.

5. The method for monitoring automatic switch voltage as described in claim 1, characterized in that, The acoustic signature parameters include harmonic ratio, energy entropy value, and sound pressure amplitude.

6. The method for monitoring automatic switch voltage as described in claim 5, characterized in that, The harmonic ratio is calculated using an improved harmonic analysis method, expressed as: in, for Weighting coefficients for first harmonics, The complex spectrum is the frequency component. This represents the total harmonic number.

7. The method for monitoring automatic switch voltage as described in claim 5, characterized in that, The energy entropy value is calculated using an improved wavelet packet decomposition method, let the first... Layer The energy of each frequency band is Normalized energy is The formula is expressed as: The improved formula for calculating energy entropy is: in, To improve the energy entropy value, The number of wavelet packet decomposition levels. Frequency band number, from 1 to , indicating different frequency bands For the first The total number of frequency bands after layer wavelet packet decomposition. For the frequency band weighting function, This is the weighting adjustment coefficient.

8. The method for monitoring automatic switch voltage as described in claim 5, characterized in that, The sound pressure amplitude is calculated using a dynamic reference correction method, expressed by the following formula: Where RMS(t) is the root mean square value. For reference sound pressure level, For dynamic correction coefficients, The standard deviation of the current time window. This is the reference standard deviation under standard operating conditions. This is the correction factor.

9. The method for monitoring automatic switch voltage as described in claim 8, characterized in that, The dual-domain voiceprint feature map includes a time-domain feature curve and a frequency-domain feature curve; The construction of the time-domain characteristic curve involves plotting a waveform curve with time as the horizontal axis and sound pressure amplitude as the vertical axis, while simultaneously overlaying and displaying the changing trend of energy entropy value to construct a two-parameter time-domain characteristic curve. The frequency domain characteristic curve is constructed with time as the horizontal axis and frequency as the vertical axis. The energy intensity is represented by the color intensity, and the harmonic ratio change trajectory is marked.

10. A monitoring system for automatically monitoring switch pressure, characterized in that, include: The acquisition module is configured to use acoustic sensors to acquire the acoustic signals of the device itself and the mechanism during the pressurization process of the switching equipment; The segmentation module performs adaptive segmentation processing on the acoustic signals of the body and the acoustic signals of the mechanism, extracts the acoustic feature parameters in each time period, and constructs a dual-domain acoustic feature map of the switching equipment. The calculation module calculates the phase synchronization degree based on the dual-domain acoustic signature feature map, and constructs a cooperative acoustic signature model of the switching equipment based on the phase synchronization degree. The early warning module dynamically compares the feature parameters of the collaborative voiceprint model with the standard voiceprint library to locate the fault type and output early warning information.

Images