A multi-channel synchronous signal acquisition system

Abstract

This invention discloses a multi-channel synchronous signal acquisition system, relating to the field of high-voltage equipment monitoring technology, which solves the problems of asynchronous acquisition of acoustic and temperature monitoring signals and fragmented spatial data in GIS equipment. The system includes a signal acquisition and storage module, a data polling module, a data processing module, and a data fusion module. Each signal acquisition channel alternately samples acoustic and temperature signals, adding timestamps and channel identifiers. The central aggregation node of the data polling module polls and reads data from each channel, combining timestamp-aligned samples into multi-dimensional array snapshots. Beamforming is performed on the acoustic array snapshot to obtain an acoustic field energy distribution map, and spatial interpolation is performed on the temperature array snapshot to obtain a temperature field distribution map. The two maps are spatially aligned and fused to generate a GIS equipment status feature map that simultaneously contains acoustic and temperature information. This achieves strict synchronization of acoustic and temperature multi-physical quantity acquisition and precise spatial fusion characterization.

Description

Technical Field

[0001] This invention belongs to the field of high-voltage equipment monitoring technology, specifically a multi-channel synchronous signal acquisition system. Background Technology

[0002] GIS (Gas Insulated Switchgear) is a critical piece of equipment in power systems, and its internal partial discharge and abnormal temperature rise are important indicators of insulation degradation and contact faults. Currently, online monitoring of GIS equipment generally adopts a technical approach that separates acoustic and temperature monitoring. Acoustic monitoring typically uses an array of ultrasonic sensors arranged on the casing surface to locate partial discharges, while temperature monitoring often employs periodic inspections with infrared thermal imagers or the installation of discrete temperature sensors. These two monitoring methods are independent in data acquisition, each with its own sampling system and time reference.

[0003] Existing separate monitoring schemes have shortcomings. The acoustic array and temperature sensing system cannot achieve strict synchronization at the sampling time, resulting in a time-separation discrepancy between the acquired acoustic and temperature signals, making it difficult to determine the true correspondence between the sound source location and the temperature field at the same moment. Furthermore, discrete temperature point data cannot form a complete temperature field matching the spatial resolution of the acoustic array, and the two physical quantities are spatially fragmented. This prevents operators from accurately correlating and comprehensively analyzing the sound source hotspots generated by partial discharge with physical temperature hotspots caused by poor contact or other reasons, hindering the development of accurate fault diagnosis and early warning capabilities based on multi-parameter fusion. Therefore, it is necessary to solve the technical challenge of strictly synchronizing sound and temperature data during acquisition and achieving spatial alignment and fusion. Summary of the Invention

[0004] This invention aims to solve at least one of the technical problems existing in the prior art;

[0005] Therefore, this invention proposes a multi-channel synchronous signal acquisition system, comprising:

[0006] The signal acquisition and storage module has multiple signal acquisition channels arranged in a ring on the surface of the GIS equipment housing. The acoustic wave digital sample sequence and temperature digital sample sequence obtained by alternating sampling in each signal acquisition channel are temporarily stored in the corresponding channel-specific buffer memory. The acoustic wave digital sample sequence and temperature digital sample sequence are attached with channel identification code and timestamp generated by global synchronization clock counting.

[0007] The data polling module is used to set up a central aggregation node. The central aggregation node polls the channel-specific buffer memory of all signal acquisition channels in sequence according to a preset period, and reads the sound wave and temperature digital sample data packets with timestamps and channel identification codes.

[0008] The data processing module is used to combine timestamp-aligned digital acoustic sample sequences from different signal acquisition channels into a multidimensional acoustic array snapshot within the central aggregation node, and to combine timestamp-aligned digital temperature sample sequences from different signal acquisition channels into a multidimensional temperature array snapshot. Spatial beamforming calculations are performed on each multidimensional acoustic array snapshot to obtain a sound field energy distribution map characterizing the spatial distribution of the sound source on the surface of the GIS equipment housing. Spatial interpolation calculations are performed on each multidimensional temperature array snapshot to obtain a continuous temperature field distribution map covering the surface of the GIS equipment housing.

[0009] The data fusion module is used to spatially align the sound field energy distribution map and the temperature field distribution map at the same timestamp, and fuse them to generate a GIS equipment status feature map with sound and temperature superposition information.

[0010] Furthermore, the arrangement of multiple signal acquisition channels in a ring on the surface of the GIS equipment housing includes:

[0011] Multiple acoustic wave sensor probes are arranged in a ring on the surface of the GIS equipment housing, and a temperature sensor head is configured next to each acoustic wave sensor probe to form multiple physically adjacent acoustic and temperature sensing units.

[0012] Each set of acoustic temperature sensing units is assigned an independent signal acquisition channel. Each signal acquisition channel contains an adjustable gain amplification stage for amplifying acoustic signals and an isolation conversion stage for conditioning temperature signals.

[0013] A global synchronization clock signal with a fixed frequency is generated, and the global synchronization clock signal is distributed to the analog-to-digital converter controllers in all signal acquisition channels in a star topology.

[0014] The analog-to-digital converter controller is a control chip integrated on the signal conditioning circuit board of each signal acquisition channel, used to convert analog signals into digital signals.

[0015] Within each signal acquisition channel, the received global synchronization clock signal drives the analog-to-digital converter controller to perform strict, equally spaced alternating sampling of the acoustic signal after the adjustable gain amplification stage and the temperature signal after the isolation conversion stage.

[0016] The method involves arranging multiple acoustic wave sensing probes in a ring on the surface of the GIS equipment housing, and configuring a temperature sensing head next to each acoustic wave sensing probe, forming multiple physically adjacent acoustic-temperature sensing units. Specifically:

[0017] Obtain the three-dimensional geometric model of the GIS equipment casing to be monitored, and extract its generatrix length and cross-sectional perimeter parameters;

[0018] The maximum number of sensor units that can be theoretically arranged is calculated by dividing the perimeter of the cross section by the preset minimum spacing between sensor units; according to the acoustic spatial sampling theorem, the minimum number of sensor units required to avoid spatial overlap on the annular cross section is calculated; in actual arrangement, the determined number of sensor units should not be less than the minimum number of sensor units and not greater than the maximum number of sensor units.

[0019] The coordinates of the installation points of each actual sensor unit are calculated at equal intervals on the same annular cross section of the GIS equipment housing to ensure that all installation points are located on the same geometric plane.

[0020] At each installation point, a broadband acoustic wave sensor probe is installed, with the central axis of the broadband acoustic wave sensor probe perpendicular to the tangent direction of the outer surface of the GIS equipment housing.

[0021] Install a contact temperature sensor head on the side of each broadband acoustic wave sensor probe housing, no more than five centimeters from the center of the acoustic wave sensor probe, and ensure that the temperature sensing surface of the temperature sensor head is in close contact with the surface of the GIS equipment housing.

[0022] The broadband acoustic wave sensing probe and the contact temperature sensing head at each location are physically fixed as an integral component, which is defined as a set of acoustic temperature sensing units.

[0023] Each set of acoustic temperature sensing units is assigned a unique physical channel number, which is also associated with the position index of the annular cross section where it is located.

[0024] Furthermore, each group of sound temperature sensing units is allocated an independent signal acquisition channel. Each signal acquisition channel includes an adjustable gain amplification stage for amplifying the sound wave signal and an isolation conversion stage for conditioning the temperature signal, specifically:

[0025] Connect an independent signal conditioning circuit board to the output interface of each group of acoustic temperature sensing units;

[0026] On the signal conditioning circuit board, the raw voltage signal from the broadband acoustic wave sensing probe is connected to an adjustable gain operational amplifier circuit to form an adjustable gain amplification stage. The gain value of the adjustable gain operational amplifier circuit is set by a digital potentiometer, which receives remote configuration from the central aggregation node.

[0027] Meanwhile, on the signal conditioning circuit board, the original resistance signal from the contact temperature sensor head is connected to a four-wire constant current source drive and voltage measurement circuit to form an isolation conversion stage. The isolation conversion stage can linearly convert the resistance signal into a differential voltage signal and eliminate common-mode interference through an isolation amplifier.

[0028] The analog signals output from the adjustable gain amplification stage and the isolation conversion stage are sent in parallel to different input terminals of a multiplexer on the signal conditioning circuit board.

[0029] Furthermore, within each signal acquisition channel, the received global synchronization clock signal drives the analog-to-digital converter controller to perform strictly equal-interval alternating sampling of the acoustic signal after adjustable gain amplification and the temperature signal after isolation conversion. Specifically:

[0030] Each signal conditioning circuit board has an analog-to-digital converter controller with an internal sampling sequence generator.

[0031] The sampling sequence generator receives a global synchronization clock signal from the central aggregation node and uses it as its own reference clock.

[0032] The sampling sequence generator is pre-configured with an alternating sampling mode, which instructs the analog-to-digital converter controller to sequentially sample and convert the acoustic analog signal and the temperature analog signal input from the multiplexer;

[0033] At the rising edge of each global synchronization clock signal, the analog-to-digital converter controller controls the multiplexer to switch the input channel once and start an analog-to-digital conversion once, according to the alternating sampling mode;

[0034] After the analog-to-digital conversion is completed, the digital quantity is latched into the output data register inside the analog-to-digital converter controller. At the same time, the sampling sequence generator records a local clock count value for the digital quantity. The local clock count value is accumulated based on the global synchronization clock signal and serves as the precise sampling time of the digital quantity.

[0035] During a complete sampling period, the acoustic analog signal and the temperature analog signal are each sampled once, thereby generating a pair of acoustic digital samples and temperature digital samples that are strictly correlated in time.

[0036] Furthermore, the process of temporarily storing the acoustic wave digital sample sequence and temperature digital sample sequence obtained by alternating sampling within each signal acquisition channel in the corresponding channel-specific buffer memory, and attaching a channel identifier code and a timestamp generated by a global synchronization clock, specifically involves:

[0037] Each signal conditioning board is equipped with a dual-port buffer memory as a channel-specific buffer memory;

[0038] Each time the analog-to-digital converter completes the acquisition of a pair of acoustic digital sample sequences and temperature digital sample sequences, it writes the acoustic digital sample and temperature digital sample sequences, along with their local clock count values, as a data packet into a designated queue of the channel-specific buffer memory.

[0039] An independent formatting microprocessor reads data packets from the channel-specific buffer queue, adds the physical channel number information of this channel to each data packet, and forms a channel identification code;

[0040] The formatting microprocessor uses the local clock count value carried in the data packet to query a time mapping table corrected by a global synchronization clock signal, and converts the local clock count value into a unified timestamp based on absolute time.

[0041] The data packets, with added channel identifiers and unified timestamps, are rewritten into another send queue area of ​​the channel-specific buffer, awaiting reading by the central aggregation node.

[0042] Furthermore, the setting of a central aggregation node involves the central aggregation node sequentially polling the channel-specific buffer memory of all signal acquisition channels according to a preset period, and reading the sound wave and temperature digital sample data packets containing timestamps and channel identification codes. Specifically:

[0043] The central aggregation node contains a main control processor and a multi-channel digital interface controller;

[0044] The main control processor is configured with a polling schedule, which specifies the order and time interval for reading data from each signal acquisition channel;

[0045] According to the polling schedule, the multi-channel digital interface controller sequentially accesses the dual-port buffer memory on each signal conditioning board via a high-speed serial bus.

[0046] When accessing a specified channel, the multi-channel digital interface controller reads one or more complete data packets from the transmit queue area of ​​the channel-specific buffer memory of the specified channel until the marker at the end of the queue is read;

[0047] Each read data packet is transmitted to the main memory of the central aggregation node via a high-speed serial bus and stored according to its channel identifier code;

[0048] After completing a round of polling all signal acquisition channels, the main control processor checks the data packets from different channels in the main memory that have the same or very close timestamps, and marks them as data sets belonging to the same synchronous acquisition time.

[0049] Furthermore, within the central aggregation node, the timestamp-aligned digital acoustic sample sequences from different signal acquisition channels are combined into a multidimensional acoustic array snapshot, and the timestamp-aligned digital temperature samples from different signal acquisition channels are combined into a multidimensional temperature array snapshot. Specifically:

[0050] In the main memory of the central aggregation node, a multidimensional array storage area is allocated for acoustic data and temperature data respectively. The dimensions of the multidimensional array include time dimension, channel dimension and sample value dimension.

[0051] The main control processor traverses a set of data acquired at a synchronous time and determines which channel index position of the multidimensional array should be stored for the acoustic digital sample sequence and temperature digital sample sequence based on the channel identifier code in each data packet.

[0052] For acoustic data, the digital sample sequences of all channels at the same time are arranged into a one-dimensional acoustic sample vector according to the circular arrangement order of the channels. The one-dimensional acoustic sample vector is a snapshot of a multi-dimensional acoustic array.

[0053] For temperature data, the temperature digital sample sequence of all channels at the same time is arranged into a one-dimensional temperature sample vector according to the circular arrangement order of the channels. The one-dimensional temperature sample vector is a multi-dimensional temperature array snapshot.

[0054] The obtained one-dimensional acoustic wave sample vector and one-dimensional temperature sample vector are saved to the corresponding time index in the acoustic wave data multidimensional array and the temperature data multidimensional array, respectively, thereby constructing a time-varying acoustic wave array snapshot sequence and temperature array snapshot sequence.

[0055] Furthermore, the spatial beamforming calculation is performed on each multidimensional acoustic array snapshot to obtain a sound field energy distribution map characterizing the spatial distribution of the sound source on the surface of the GIS equipment housing, specifically:

[0056] The main control processor of the central aggregation node loads a pre-calculated beamforming weight matrix, which is calculated based on the annular arrangement coordinates of all acoustic wave sensing probes on the surface of the GIS equipment housing and a preset scanning angle grid.

[0057] For each multidimensional acoustic array snapshot in the multidimensional array of acoustic data, that is, a one-dimensional acoustic sample vector, the main control processor multiplies it with the beamforming weight matrix.

[0058] The result of the multiplication operation is a two-dimensional complex matrix, where the number of rows corresponds to the number of scanning angle grids and the number of columns corresponds to the number of frequency components.

[0059] The square of the modulus of each element of the two-dimensional complex matrix is ​​used to obtain the signal power value at each scanning angle and each frequency point;

[0060] The signal power values ​​at all scanning angles are arranged according to their azimuth angles corresponding to the annular cross section of the GIS equipment housing, and a polar coordinate energy distribution map is drawn with angle as the horizontal axis and energy intensity as the color. The polar coordinate energy distribution map is the sound field energy distribution map.

[0061] Add a timestamp to the corresponding multidimensional acoustic array snapshot for each sound field energy distribution map.

[0062] Furthermore, the spatial interpolation calculation performed on each multidimensional temperature array snapshot yields a continuous temperature field distribution map covering the surface of the GIS equipment housing, specifically as follows:

[0063] The main control processor of the central aggregation node loads the three-dimensional geometric model of the GIS equipment shell and extracts the annular cross-sectional contour line where the acoustic temperature sensing unit is located from the three-dimensional geometric model.

[0064] For each multidimensional temperature array snapshot in the multidimensional array of temperature data, i.e. a one-dimensional temperature sample vector, the main control processor binds each sample value to the position coordinates of the acoustic temperature sensing unit corresponding to its channel identifier on the annular contour line.

[0065] A scattered interpolation algorithm based on radial basis functions is used to calculate the temperature estimate of any point on the entire annular cross-section region enclosed by the annular cross-section profile, based on the temperature sample values ​​of known coordinate points on the annular cross-section profile.

[0066] The entire annular cross-sectional region is discretized into a high-resolution two-dimensional grid, and the temperature estimate of each grid node is calculated.

[0067] Based on the temperature estimates of all grid nodes, a pseudo-color image is generated using a color map, where colors represent temperature levels. This pseudo-color image is a continuous temperature field distribution map.

[0068] Add a timestamp to the corresponding multidimensional temperature array snapshot for each continuous temperature field distribution map.

[0069] Furthermore, the step of aligning the sound field energy distribution map and temperature field distribution map at the same timestamp in spatial coordinates and fusing them to generate a GIS equipment status feature map with sound and temperature superposition information specifically involves:

[0070] The main control processor retrieves a sound field energy distribution map and a continuous temperature field distribution map with the same timestamp from the channel-specific buffer memory;

[0071] The main control processor first resamples the sound field energy distribution map to the same two-dimensional grid coordinate system as the continuous temperature field distribution map through coordinate transformation;

[0072] In a unified two-dimensional grid coordinate system, each pixel simultaneously has an energy intensity value from the resampled sound field energy distribution map and a temperature value from the continuous temperature field distribution map;

[0073] Define a fusion function that takes energy intensity value and temperature value as input and calculates a fusion feature value. The specific form of the fusion function is to perform a weighted summation of the normalized energy intensity value and the normalized temperature value.

[0074] The calculated fusion feature values ​​are mapped onto a new color bar to generate a single-channel fusion image;

[0075] The single-channel fused image is overlaid with the original continuous temperature field distribution map in pseudo-color, where the brightness of the fused image represents the fusion feature value and the hue of the temperature field distribution map represents the temperature value, ultimately generating a GIS equipment status feature map that simultaneously visualizes sound field energy and temperature information.

[0076] The generated GIS device status feature map, along with the timestamp, is stored in the non-volatile memory of the central aggregation node.

[0077] Compared with the prior art, the beneficial effects of the present invention are:

[0078] Each signal acquisition channel alternately acquires acoustic wave samples and temperature samples, and uses the same global synchronization clock to attach a uniform timestamp to all samples across all channels. This design integrates the two originally independent types of sensor signals into a unified time-series coordinate system at the acquisition source, providing a precisely comparable time reference for acoustic wave samples, temperature samples, and samples across physical quantities from different channels. It eliminates the inherent random time errors in traditional separate monitoring systems caused by different clock sources and asynchronous triggering, laying a precise time foundation for achieving rigorous synchronous analysis and correlation of acoustic and temperature signals.

[0079] By combining acoustic wave samples from all channels at the same time into a multi-dimensional array snapshot and performing spatial beamforming calculations, a sound field energy distribution map of the GIS equipment casing surface is directly generated. Similarly, temperature samples at the same time are combined and spatially interpolated to generate a continuous temperature field distribution map. The two distribution maps are then spatially aligned and fused. This processing workflow integrates acoustic array processing technology and temperature field reconstruction technology on a unified spatial grid. The output fused feature map directly presents the spatial overlap and correlation between acoustic emission locations and temperature distribution patterns. This processing transforms the representation of equipment status from single-parameter, independent point and line information to multi-parameter correlated two-dimensional spatial image information, enhancing the intuitiveness and interpretability of status characteristics and revealing implicit complex fault modes such as "whether partial discharge is accompanied by local overheating." Attached Figure Description

[0080] Figure 1 This is a timing diagram of the multi-channel synchronous signal acquisition system described in this invention;

[0081] Figure 2 A flowchart illustrating the signal acquisition channel layout and synchronous sampling;

[0082] Figure 3 A flowchart attached to data storage and timestamps;

[0083] Figure 4 Polar coordinate distribution cloud map of sound field energy of sound temperature sensing unit during abnormal stage of GIS equipment;

[0084] Figure 5 This is an interpolated distribution map of the continuous temperature field across the annular cross-section of the GIS equipment casing. Detailed Implementation

[0085] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0086] See Figure 1 The system features multiple signal acquisition channels arranged in a ring on the surface of the GIS equipment housing, each operating independently. The signal acquisition and storage module alternately samples acoustic and temperature signals within each channel, generating digital sample sequences for both acoustic and temperature signals. These sample data, along with their channel identifiers and timestamps generated based on a global synchronization clock, are stored in the channel's dedicated buffer memory. A data polling module uses a central aggregation node that accesses the buffer memories of all channels at fixed intervals, reading tagged acoustic and temperature digital sample data packets. The data processing module operates within the central aggregation node. It combines acoustic sample sequences from different channels at the same time into a multidimensional acoustic array snapshot, and combines temperature sample sequences at the same time into a multidimensional temperature array snapshot. Subsequently, it performs spatial beamforming calculations on each acoustic array snapshot, outputting a sound field energy distribution map, and performs spatial interpolation calculations on each temperature array snapshot, outputting a continuous temperature field distribution map. The data fusion module aligns the acoustic energy distribution map and the temperature distribution map at the same moment in space and overlays the information, and finally merges them to generate a GIS equipment status feature map that contains both acoustic and temperature information.

[0087] See Figure 2In one embodiment of the present invention, multiple signal acquisition channels are arranged in a ring on the surface of the GIS equipment housing. A three-dimensional geometric model of the GIS equipment housing to be monitored is acquired, and its generatrix length and cross-sectional perimeter parameters are extracted. The maximum theoretically achievable number of sensor units is calculated by dividing the cross-sectional perimeter by a preset minimum sensor unit spacing. Based on the acoustic wave spatial sampling theorem, the minimum number of sensor units required to avoid spatial aliasing on the ring-shaped cross-section is calculated. The determined number of sensor units should be no less than the minimum number and no greater than the maximum number. The coordinates of the installation points of the actual number of sensor units are calculated at equal intervals on the same ring-shaped cross-section of the GIS equipment housing, ensuring that all installation points are located on the same geometric plane. At each installation point, a broadband acoustic wave sensor probe is installed, with its central axis perpendicular to the tangent direction of the outer surface of the GIS equipment housing. Adjacent to the side of each broadband acoustic wave sensor probe housing, at a distance of no more than five centimeters from the probe center, a contact temperature sensor head is installed, ensuring that the temperature sensing surface of the temperature sensor head is in close contact with the surface of the GIS equipment housing. The broadband acoustic wave sensing probe and the contact temperature sensing head at each location are physically fixed as a single assembly, defined as a group of acoustic-temperature sensing units. Each group of acoustic-temperature sensing units is assigned a unique physical channel number, which is also associated with its corresponding annular cross-sectional location index. An independent signal conditioning circuit board is connected to the output interface of each group of acoustic-temperature sensing units. On the signal conditioning circuit board, the raw voltage signal from the broadband acoustic wave sensing probe is input to an adjustable gain operational amplifier circuit, forming an adjustable gain amplification stage. The gain value of this adjustable gain operational amplifier circuit is set via a digital potentiometer, which receives remote configuration from the central convergence node. Simultaneously, on the signal conditioning circuit board, the raw resistance signal from the contact temperature sensing head is input to a four-wire constant current source drive and voltage measurement circuit, forming an isolation conversion stage. This isolation conversion stage linearly converts the resistance signal into a differential voltage signal and eliminates common-mode interference through an isolation amplifier. The analog signals output from the adjustable gain amplification stage and the isolation conversion stage are sent in parallel to different inputs of a multiplexer on the signal conditioning circuit board. A global synchronization clock signal with a fixed frequency is generated and distributed in a star topology to the analog-to-digital converters (ADCs) in all signal acquisition channels. The ADCs are control chips integrated on the signal conditioning circuit boards of each signal acquisition channel, used to convert analog signals to digital signals. Each ADC on the signal conditioning circuit board contains a sampling sequence generator. The sampling sequence generator receives the global synchronization clock signal from the central aggregation node and uses it as its own reference clock. The sampling sequence generator is pre-configured with an alternating sampling mode, which instructs the ADCs to sequentially sample and convert the acoustic analog signal and temperature analog signal input from the multiplexer.At the rising edge of each global synchronization clock signal, the analog-to-digital converter (ADC) controller, according to the alternating sampling mode, controls the multiplexer to switch the input channel once and initiates an analog-to-digital conversion. The digital quantity after the ADC conversion is completed is latched into the ADC controller's internal output data register. Simultaneously, the sampling sequence generator records a local clock count for this digital quantity. This local clock count is accumulated based on the global synchronization clock signal and serves as the precise sampling time for the digital quantity. Within a complete sampling cycle, the acoustic analog signal and the temperature analog signal are each sampled once, thereby generating a pair of time-dependent acoustic digital samples and temperature digital samples.

[0088] In practical implementation, after obtaining the three-dimensional geometric model of the GIS equipment shell to be monitored, its generatrix length and cross-sectional perimeter parameters are extracted. The maximum theoretically achievable number of sensor units is calculated by dividing the cross-sectional perimeter by the preset minimum sensor unit spacing. Based on the acoustic spatial sampling theorem, the minimum number of sensor units required to avoid spatial aliasing on the annular cross-section is calculated. The formula used is:

[0089]

[0090] in: This represents the radius of the annular cross-section of the GIS equipment housing. This represents the minimum wavelength in the acoustic signal to be monitored. This indicates rounding up; the number of sensing units determined in the actual layout should be no less than the minimum number of sensing units and no more than the maximum number of sensing units. In some embodiments, the coordinates of the installation points of the actual number of sensing units are calculated at equal intervals on the same annular cross-section of the GIS equipment housing, ensuring that all installation points are located on the same geometric plane. A broadband acoustic wave sensing probe is installed at each installation point, with the central axis of the broadband acoustic wave sensing probe perpendicular to the tangent direction of the outer surface of the GIS equipment housing. It can be understood that a contact temperature sensing head is installed on the side of the housing of each broadband acoustic wave sensing probe, at a distance of no more than five centimeters from the center of the probe, ensuring that the temperature sensing surface of the contact temperature sensing head is in close contact with the surface of the GIS equipment housing.

[0091] In practical implementation, the broadband acoustic wave sensing probe and the contact temperature sensing head at each location are physically fixed as a single component, defined as a group of acoustic-temperature sensing units. Each group of acoustic-temperature sensing units is assigned a unique physical channel number, which is also associated with its corresponding annular cross-sectional position index. An independent signal conditioning circuit board is connected to the output interface of each group of acoustic-temperature sensing units. On the signal conditioning circuit board, the raw voltage signal from the broadband acoustic wave sensing probe is input to an adjustable gain operational amplifier circuit to form an adjustable gain amplification stage. The gain value of the adjustable gain operational amplifier circuit is set via a digital potentiometer, which receives remote configuration from the central convergence node. Optionally, the raw resistance signal from the contact temperature sensing head is input to a four-wire constant current source drive and voltage measurement circuit on the signal conditioning circuit board to form an isolation conversion stage. This stage linearly converts the resistance signal into a differential voltage signal and eliminates common-mode interference through an isolation amplifier. The analog signals output from the adjustable gain amplification stage and the isolation conversion stage are sent in parallel to different inputs of a multiplexer on the signal conditioning circuit board. It can be understood that a global synchronization clock signal with a fixed frequency is generated and distributed to the analog-to-digital converters (ADCs) in all signal acquisition channels in a star topology. Optionally, each ADC on the signal conditioning circuit board has a sampling sequence generator inside. The sampling sequence generator receives the global synchronization clock signal from the central aggregation node and uses it as its own reference clock. The sampling sequence generator is pre-configured with an alternating sampling mode, which instructs the ADC to sequentially sample and convert the acoustic analog signal and temperature analog signal input to the multiplexer. On the rising edge of each global synchronization clock signal, the ADC controls the multiplexer to switch the input channel once and initiate an analog-to-digital conversion according to the alternating sampling mode. The digital quantity after the analog-to-digital conversion is latched into the output data register inside the ADC, while the sampling sequence generator records a local clock count value for the digital quantity. The local clock count value is accumulated based on the global synchronization clock signal to obtain the precise sampling time of the digital quantity. Within a complete sampling cycle, the acoustic analog signal and the temperature analog signal are each sampled once, thereby generating a pair of acoustic digital samples and temperature digital samples that are strictly correlated in time.

[0092] See Figure 3In one embodiment of the invention, each signal conditioning circuit board is equipped with a dual-port buffer memory as a channel-specific buffer memory. Each time the analog-to-digital converter completes the acquisition of a pair of acoustic and temperature digital samples, it writes the acoustic and temperature digital samples, along with their local clock count values, as a data packet into a designated queue in the channel-specific buffer memory. A separate formatting microprocessor reads the data packets in the channel-specific buffer memory queue, adds the physical channel number information of the channel to each data packet, forming a channel identifier. The formatting microprocessor uses the local clock count value carried in the data packet to query a time mapping table corrected by a globally synchronized clock signal, converting the local clock count value into a unified timestamp based on absolute time. The data packet with the added channel identifier and unified timestamp is then rewritten into another transmission queue area of ​​the channel-specific buffer memory, awaiting reading by the central aggregation node.

[0093] In implementation, a dedicated formatting microprocessor reads data packets from a designated queue in the channel-specific buffer at fixed intervals. The formatting microprocessor adds the physical channel number information of the channel to each read data packet. This physical channel number information, together with the original local clock count value of the data packet, constitutes the channel identification code. Optionally, after adding the channel identification code, the formatting microprocessor accesses a time mapping table pre-stored in local read-only memory. The time mapping table records the correction relationship between the global synchronization clock signal and the absolute time reference. The formatting microprocessor uses the local clock count value carried by the data packet as an index to query the time mapping table and obtain the corresponding absolute time offset. The formatting microprocessor adds the absolute time offset to a preset epoch start time to calculate the unified timestamp corresponding to the data packet based on absolute time. The formula for calculating the unified timestamp is:

[0094]

[0095] in: This represents the calculated unified timestamp. This represents the pre-defined absolute time of the start of the era. This represents the local clock count value included in the data packet. This represents the number of seconds that each count represents, determined by the frequency of the global synchronization clock signal.

[0096] In practice, the formatting microprocessor rewrites data packets with added channel identifiers and unified timestamps into another transmission queue area of ​​the channel-specific buffer memory. This transmission queue area is physically isolated from the designated queue in the dual-port buffer memory, and the data packet order in the transmission queue area is consistent with the order of the unified timestamps. The transmission queue area of ​​the channel-specific buffer memory is equipped with a status flag register, which sets a marker for the tail of the queue. When the formatting microprocessor writes a new data packet, it updates the marker position at the tail of the queue. When the central aggregation node reads a data packet, it determines the end point of the read based on the marker at the tail of the queue. The dual-port design of the channel-specific buffer memory allows the formatting microprocessor and the multi-channel digital interface controller of the central aggregation node to access different areas simultaneously. The formatting microprocessor continuously writes data packets with unified timestamps to the transmission queue area, while the multi-channel digital interface controller periodically reads data packets from the transmission queue area. The data packet storage format in the channel-specific buffer memory includes a fixed-length header and a payload. The header continuously stores the channel identifier and unified timestamp, while the payload stores the acoustic wave digital sample and temperature digital sample. The local clock count value is no longer retained after the data packet is written to the transmission queue area.

[0097] In one embodiment of the present invention, the central aggregation node includes a main control processor and a multi-channel digital interface controller. The main control processor is configured with a polling schedule that specifies the order and time interval for reading data from each signal acquisition channel; the polling schedule defines a fixed polling period. In each polling cycle, the multi-channel digital interface controller must complete a full access to the dedicated buffer memory of all channels. The polling schedule also divides each polling cycle into several equal time slots. Each time slot is assigned to a specific signal acquisition channel for data reading operations. The multi-channel digital interface controller accesses the dual-port buffer memory on each signal conditioning board sequentially via a high-speed serial bus according to a polling schedule. The access process follows a strict time sequence, and when the system has… When acquiring one signal channel, the polling cycle is... With time slot The relationship satisfies:

[0098]

[0099] in: This represents the necessary bus switching and protocol overhead time within a polling cycle. This indicates a fixed polling period. This represents the equal time slots divided in each polling cycle. It can be understood that when accessing a specified channel, the multi-channel digital interface controller reads one or more complete data packets from the transmit queue area of ​​the channel-specific buffer memory of that channel. The reading operation continues until the multi-channel digital interface controller detects a pre-set queue tail marker in the transmit queue area. Each read data packet is transmitted to the main memory of the central aggregation node via a high-speed serial bus. The main memory has multiple independent buffer blocks, each uniquely associated with a physical channel number. The multi-channel digital interface controller stores the data packet into the corresponding buffer block in the main memory based on the channel identifier carried by the data packet, and arranges the data packets in the buffer block according to the ascending order of a unified timestamp. Optionally, after completing one round of polling of all signal acquisition channels, the main control processor starts a timestamp alignment process. The timestamp alignment process checks the most recently stored data packets in all buffer blocks in the main memory; the timestamp alignment process traverses the timestamp field of the header data packets of all channels' corresponding buffer blocks, and finds the largest timestamp as the base timestamp. The timestamp alignment process ensures that the timestamps in the header packets of all channel buffer blocks meet the required standards. Data packets are marked as belonging to the same data set acquired at the same synchronous acquisition time. In the formula... A timestamp representing a specific data packet. This represents the preset time tolerance threshold. For data packets whose timestamps exceed the tolerance range, the timestamp alignment process marks them as asynchronous data and stores them in a separate buffer area for subsequent analysis, while the data sets successfully marked as synchronous acquisition times are passed to the subsequent data processing module for unified processing.

[0100] In one embodiment of the present invention, a multidimensional array storage area is allocated in the main memory of the central aggregation node for acoustic data and temperature data, respectively. The dimensions of the multidimensional array include time dimension, channel dimension, and sample value dimension. The main control processor traverses the data set of a synchronous acquisition moment and determines which channel index position of the multidimensional array should store the acoustic digital sample and temperature digital sample in each data packet according to the channel identifier code. For acoustic data, the acoustic digital samples of all channels at the same moment are arranged in a circular arrangement order of their channels to form a one-dimensional acoustic sample vector, which is a multidimensional acoustic array snapshot. For temperature data, the temperature digital samples of all channels at the same moment are also arranged in a circular arrangement order of their channels to form a one-dimensional temperature sample vector, which is a multidimensional temperature array snapshot. The obtained one-dimensional acoustic sample vector and one-dimensional temperature sample vector are saved to the corresponding time index in the acoustic data multidimensional array and the temperature data multidimensional array, respectively, thereby constructing a time-varying acoustic array snapshot sequence and a temperature array snapshot sequence. The main control processor at the central aggregation node loads a pre-calculated beamforming weight matrix, which is calculated based on the annular arrangement coordinates of all acoustic wave sensor probes on the surface of the GIS equipment housing and a preset scanning angle grid. For each multidimensional acoustic array snapshot in the multidimensional acoustic data array, i.e., a one-dimensional acoustic sample vector, the main control processor multiplies it with the beamforming weight matrix. The result of the multiplication is a two-dimensional complex matrix, where the number of rows corresponds to the number of scanning angle grids, and the number of columns corresponds to the number of frequency components. The square of the modulus of each element of this two-dimensional complex matrix is ​​taken to obtain the signal power value at each scanning angle and each frequency point. The signal power values ​​at all scanning angles are arranged according to their azimuth angles corresponding to the annular cross-section of the GIS equipment housing, and plotted as a polar coordinate energy distribution map with angle as the horizontal axis and energy intensity as the color. This polar coordinate energy distribution map is the sound field energy distribution map. A timestamp of the corresponding multidimensional acoustic array snapshot is added to each sound field energy distribution map. The main control processor at the central aggregation node loads the 3D geometric model of the GIS equipment housing and extracts the annular cross-sectional contour line where the acoustic temperature sensing unit is located from the model. For each multidimensional temperature array snapshot in the multidimensional temperature data array, i.e., a one-dimensional temperature sample vector, the main control processor binds each sample value to the position coordinates of the sensing unit corresponding to its channel identifier on the annular contour line. Using a radial basis function-based scattered interpolation algorithm, based on the temperature sample values ​​of known coordinate points on the annular contour line, the estimated temperature value at any point in the entire annular cross-sectional area enclosed by the contour line is calculated. The entire annular cross-sectional area is discretized into a high-resolution two-dimensional grid, and the estimated temperature value of each grid node is calculated.Based on the temperature estimates of all grid nodes, a pseudo-color image is generated using a color map, where colors represent temperature levels. This pseudo-color image represents the continuous temperature field distribution. A timestamp is appended to each continuous temperature field distribution image to represent a snapshot of the corresponding multidimensional temperature array.

[0101] In specific implementation, the main memory of the central aggregation node allocates a multidimensional array storage area for acoustic data and temperature data respectively. The dimensions of the multidimensional array include time dimension, channel dimension, and sample value dimension. The main control processor traverses the data set of a synchronous acquisition moment and determines which channel index position of the multidimensional array should be stored for the acoustic digital sample and temperature digital sample based on the channel identifier code in each data packet. In some embodiments, for acoustic data, the acoustic digital samples of all channels at the same moment are arranged in a circular arrangement order of their channels to form a one-dimensional acoustic sample vector, which is a multidimensional acoustic array snapshot; for temperature data, the temperature digital samples of all channels at the same moment are similarly arranged to form a one-dimensional temperature sample vector, which is a multidimensional temperature array snapshot. The obtained one-dimensional acoustic sample vector and one-dimensional temperature sample vector are saved to the corresponding time index in the acoustic data multidimensional array and temperature data multidimensional array respectively, thereby constructing a time-varying acoustic array snapshot sequence and temperature array snapshot sequence.

[0102] The main control processor of the central aggregation node loads a pre-calculated beamforming weight matrix. The beamforming weight matrix is ​​calculated based on the annular arrangement coordinates of all acoustic wave sensor probes on the surface of the GIS equipment housing and the preset scanning angle grid. The storage structure of the beamforming weight matrix is ​​shown in Table 1.

[0103] Table 1: Storage Structure of Beamforming Weight Matrix

[0104]

[0105] For each multidimensional acoustic array snapshot in the multidimensional array of acoustic data, i.e., a one-dimensional acoustic sample vector, the main control processor first performs a short-time Fourier transform on the one-dimensional acoustic sample vector to obtain the frequency components, and then multiplies them with the beamforming weight matrix. The result of the multiplication operation is a two-dimensional complex matrix, where the number of rows corresponds to the number of scanning angle grids, and the number of columns corresponds to the number of frequency components. The formula for the multiplication operation is expressed as:

[0106]

[0107] in: Represents the first complex number in a two-dimensional complex matrix. The scanning angle and the first Signal power values ​​at each frequency point The first element in the beamforming weight matrix represents the... The scanning angle and the first Complex weights for each channel, The first one represents the first sound wave sample vector after a short-time Fourier transform. The first channel in the Complex values ​​at each frequency point Represents the total number of channels. This indicates a modulo operation. In some embodiments, the signal power values ​​at all scanning angles are arranged according to their azimuth angles corresponding to the annular cross-section of the GIS device housing, and a polar coordinate energy distribution map is drawn with angle as the horizontal axis and energy intensity as the color. The polar coordinate energy distribution map is the sound field energy distribution map; a timestamp of the corresponding multidimensional acoustic array snapshot is added to each sound field energy distribution map.

[0108] The main control processor at the central aggregation node loads the 3D geometric model of the GIS equipment housing and extracts the annular cross-sectional contour line where the acoustic temperature sensing unit is located from the model. For each multidimensional temperature array snapshot in the multidimensional temperature data array, i.e., a one-dimensional temperature sample vector, the main control processor binds each sample value to the position coordinates of the sensing unit corresponding to its channel identifier on the annular contour line. This can be understood as using a radial basis function-based scattered interpolation algorithm to calculate the temperature estimate at any point on the entire annular cross-sectional area enclosed by the contour line, based on the temperature sample values ​​of known coordinate points on the annular contour line. The entire annular cross-sectional area is discretized into a high-resolution two-dimensional grid, and the temperature estimate for each grid node is calculated. Based on the temperature estimates of all grid nodes, a pseudo-color image is generated using a color mapping table, where color represents temperature level; the pseudo-color image is a continuous temperature field distribution map. A timestamp of the corresponding multidimensional temperature array snapshot is appended to each continuous temperature field distribution map.

[0109] See Figure 4During the abnormal operation phase of the GIS equipment (data acquisition time 40), a polar coordinate distribution cloud map of the sound field energy was calculated using the spatial beamforming of the multi-channel synchronous signal acquisition system. This map visually presents the spatial distribution of sound source energy on the annular cross-section of the GIS equipment casing in polar coordinates. The color mapping represents the normalized sound field energy intensity (from 0.2 for darker colors to 1.0 for lighter colors), and the circular white dots represent the physical installation locations of the sound temperature sensing units. As can be seen from the map, there are obvious high-energy sound source regions in the direction of channel 12 (approximately 225° azimuth) and channel 8 (approximately 180°-225° azimuth), characterized by a significant light-colored, bright distribution, with a normalized energy intensity close to 1.0. This highly matches the sound wave radiation characteristics under abnormal operating conditions such as partial discharge inside the GIS equipment. The sound field energy intensity in other azimuth regions is at a lower level (predominantly darker colors), and the overall distribution exhibits significant directional characteristics, reflecting the spatial radiation pattern of abnormal sound sources on the surface of the GIS equipment casing. The figure provides crucial acoustic field characteristics for subsequent multi-physics field fusion analysis combining temperature field information and for locating internal faults in GIS equipment.

[0110] In one embodiment of the present invention, the main control processor retrieves a sound field energy distribution map and a continuous temperature field distribution map with the same timestamp from storage. Since the sound field energy distribution map is based on a polar coordinate system, while the continuous temperature field distribution map is based on a Cartesian coordinate system, the main control processor first resamples the sound field energy distribution map to the same two-dimensional grid coordinate system as the continuous temperature field distribution map through coordinate transformation. In the unified two-dimensional grid coordinate system, each pixel simultaneously has an energy intensity value from the resampled sound field energy distribution map and a temperature value from the continuous temperature field distribution map. A fusion function is defined, which takes the energy intensity value and the temperature value as inputs and calculates a fusion feature value. The specific form of the fusion function is a weighted sum of the normalized energy intensity value and the normalized temperature value. The calculated fusion feature value is mapped onto a new color bar to generate a single-channel fused image. The fused single-channel image is overlaid with the original continuous temperature field distribution map using pseudo-color overlay. The brightness of the fused image represents the fusion feature value, and the hue of the temperature field distribution map represents the temperature value. This results in a GIS device status feature map that simultaneously visualizes sound field energy and temperature information. The generated GIS device status feature map, along with a timestamp, is stored in the non-volatile memory of the central aggregation node.

[0111] In specific implementation, the main control processor retrieves a sound field energy distribution map and a continuous temperature field distribution map with the same timestamp from storage. The retrieval is based on a precise match of the timestamp fields attached to the sound field energy distribution map and the continuous temperature field distribution map. In some embodiments, since the sound field energy distribution map is based on a polar coordinate system while the continuous temperature field distribution map is based on a Cartesian coordinate system, the main control processor first resamples the sound field energy distribution map to the same two-dimensional grid coordinate system as the continuous temperature field distribution map through coordinate transformation. The coordinate transformation process calculates the polar radius and polar angle corresponding to each pixel in the sound field energy distribution map in the continuous temperature field distribution. Figure 2 The image uses horizontal and vertical coordinates in a 2D grid coordinate system, and employs bilinear interpolation to calculate the energy intensity value at the new coordinate location. Optionally, the image resolution of the resampled sound field energy distribution map is kept consistent with the image resolution of the continuous temperature field distribution map, ensuring that each pixel has the same spatial size and physical coordinate correspondence. In this unified 2D grid coordinate system, each pixel simultaneously possesses an energy intensity value from the resampled sound field energy distribution map and a temperature value from the continuous temperature field distribution map. The main control processor maintains a data pair containing the energy intensity value and temperature value for each pixel. This can be understood as defining a fusion function that uses the energy intensity value and temperature value as input to calculate a fusion feature value; for an image with coordinates of... The normalized energy intensity value of the pixel. and normalized temperature value The eigenvalues ​​are obtained through their respective linear normalization functions and then fused. The calculation formula is:

[0112]

[0113] in: Preset weighting coefficients representing energy intensity values, The preset weighting coefficients representing temperature values, and satisfying the following conditions: In some embodiments, the calculated fusion feature values ​​are mapped onto a new color bar, which defines a continuous color gradient from low to high fusion feature values, thereby generating a single-channel fused image; the grayscale or luminance value of each pixel in the fused image is determined by the fusion feature value of that pixel. The mapping result on the color bar is determined.

[0114] The fused image from a single channel is overlaid with the original continuous temperature field distribution map using a pseudo-color overlay display. This pseudo-color overlay employs HSV color space transformation. Temperature values ​​from the continuous temperature field distribution map are mapped to the hue component of the HSV color space, while luminance values ​​from the fused image from the single channel are mapped to the lightness component of the HSV color space. The saturation component is set to a constant value. Optionally, by converting from the HSV color space back to the RGB color space, a GIS device status feature map simultaneously visualizing sound field energy and temperature information is generated. In the generated GIS device status feature map, temperature information is expressed through color hue, and the fused sound field energy information is expressed through regional brightness or lightness / darkness variations. In specific implementations, the generated GIS device status feature map and its timestamp are stored together in the non-volatile memory of the central aggregation node. The storage format includes a file header and a data body. The file header records the timestamp, image size, and color space parameters, while the data body stores RGB pixel values ​​in row-major order.

[0115] See Figure 5 In temperature field monitoring of GIS equipment, the figure shows a continuous temperature field distribution map generated by temperature field interpolation. Specifically, the original temperature data is collected by a ring-shaped arrangement of acoustic temperature sensing units, with each unit's temperature sample bound to the corresponding annular cross-sectional coordinates. A radial basis function-based scattered interpolation algorithm is used, based on the temperature values ​​of discrete sensing points, to interpolate the entire annular cross-sectional area of ​​the GIS equipment casing, obtaining continuous temperature estimates. After discretizing the annular cross-sectional area into a high-resolution two-dimensional grid, a pseudo-color image is generated through color mapping, where the color gradient corresponds to a temperature range of 30.0℃ to 47.5℃. The black dashed line in the figure represents the annular cross-sectional outline of the GIS equipment, and the black dots represent the installation positions of the acoustic temperature sensing units. It can be clearly seen that the temperature field exhibits a significant high-temperature concentration in the area below the cross-section (light-colored area), while the temperature is relatively low in other areas (dark-colored area), intuitively reflecting the temperature distribution characteristics of the GIS equipment casing surface.

[0116] 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 multi-channel synchronous signal acquisition system, characterized in that, include: The signal acquisition and storage module has multiple signal acquisition channels arranged in a ring on the surface of the GIS equipment housing. The acoustic wave digital sample sequence and temperature digital sample sequence obtained by alternating sampling in each signal acquisition channel are temporarily stored in the corresponding channel-specific buffer memory. The acoustic wave digital sample sequence and temperature digital sample sequence are attached with channel identification code and timestamp generated by global synchronization clock counting. The data polling module is used to set up a central aggregation node. The central aggregation node polls the channel-specific buffer memory of all signal acquisition channels in sequence according to a preset period, and reads the sound wave and temperature digital sample data packets with timestamps and channel identification codes. The data processing module is used to combine timestamp-aligned digital acoustic wave sample sequences from different signal acquisition channels into a multidimensional acoustic wave array snapshot within the central aggregation node, and to combine timestamp-aligned digital temperature sample sequences from different signal acquisition channels into a multidimensional temperature array snapshot. Spatial beamforming calculations are performed on each multidimensional acoustic wave array snapshot to obtain a sound field energy distribution map characterizing the spatial distribution of the sound source on the surface of the GIS equipment housing. Spatial interpolation calculations are performed on each multidimensional temperature array snapshot to obtain a continuous temperature field distribution map covering the surface of the GIS equipment housing. The data fusion module is used to spatially align the sound field energy distribution map and the temperature field distribution map at the same timestamp, and fuse them to generate a GIS equipment status feature map with sound and temperature superposition information.

2. The multi-channel synchronous signal acquisition system according to claim 1, characterized in that, The arrangement of multiple signal acquisition channels in a ring on the surface of the GIS equipment housing includes: Multiple acoustic wave sensor probes are arranged in a ring on the surface of the GIS equipment housing, and a temperature sensor head is configured next to each acoustic wave sensor probe to form multiple physically adjacent acoustic and temperature sensing units. Each set of acoustic temperature sensing units is assigned an independent signal acquisition channel. Each signal acquisition channel contains an adjustable gain amplification stage for amplifying acoustic signals and an isolation conversion stage for conditioning temperature signals. A global synchronization clock signal with a fixed frequency is generated, and the global synchronization clock signal is distributed to the analog-to-digital converter controllers in all signal acquisition channels in a star topology. The analog-to-digital converter controller is a control chip integrated on the signal conditioning circuit board of each signal acquisition channel, used to convert analog signals into digital signals. Within each signal acquisition channel, the received global synchronization clock signal drives the analog-to-digital converter controller to perform strict, equally spaced alternating sampling of the acoustic signal after the adjustable gain amplification stage and the temperature signal after the isolation conversion stage. The method involves arranging multiple acoustic wave sensing probes in a ring on the surface of the GIS equipment housing, and configuring a temperature sensing head next to each acoustic wave sensing probe, forming multiple physically adjacent acoustic-temperature sensing units. Specifically: Obtain the three-dimensional geometric model of the GIS equipment casing to be monitored, and extract its generatrix length and cross-sectional perimeter parameters; The maximum number of sensor units that can be theoretically arranged is calculated by dividing the perimeter of the cross section by the preset minimum spacing between sensor units; according to the acoustic spatial sampling theorem, the minimum number of sensor units required to avoid spatial overlap on the annular cross section is calculated; in actual arrangement, the determined number of sensor units should not be less than the minimum number of sensor units and not greater than the maximum number of sensor units. The coordinates of the installation points of each actual sensor unit are calculated at equal intervals on the same annular cross section of the GIS equipment housing to ensure that all installation points are located on the same geometric plane. At each installation point, a broadband acoustic wave sensor probe is installed, with the central axis of the broadband acoustic wave sensor probe perpendicular to the tangent direction of the outer surface of the GIS equipment housing. Install a contact temperature sensor head on the side of each broadband acoustic wave sensor probe housing, no more than five centimeters from the center of the acoustic wave sensor probe, and ensure that the temperature sensing surface of the temperature sensor head is in close contact with the surface of the GIS equipment housing. The broadband acoustic wave sensing probe and the contact temperature sensing head at each location are physically fixed as an integral component, which is defined as a set of acoustic temperature sensing units. Each set of acoustic temperature sensing units is assigned a unique physical channel number, which is also associated with the position index of the annular cross section where it is located.

3. The multi-channel synchronous signal acquisition system according to claim 2, characterized in that, Each group of sound temperature sensing units is allocated an independent signal acquisition channel. Each signal acquisition channel includes an adjustable gain amplification stage for amplifying the sound wave signal and an isolation conversion stage for conditioning the temperature signal. Specifically: Connect an independent signal conditioning circuit board to the output interface of each group of acoustic temperature sensing units; On the signal conditioning circuit board, the raw voltage signal from the broadband acoustic wave sensing probe is connected to an adjustable gain operational amplifier circuit to form an adjustable gain amplification stage. The gain value of the adjustable gain operational amplifier circuit is set by a digital potentiometer, which receives remote configuration from the central aggregation node. Meanwhile, on the signal conditioning circuit board, the original resistance signal from the contact temperature sensor head is connected to a four-wire constant current source drive and voltage measurement circuit to form an isolation conversion stage. The isolation conversion stage can linearly convert the resistance signal into a differential voltage signal and eliminate common-mode interference through an isolation amplifier. The analog signals output from the adjustable gain amplification stage and the isolation conversion stage are sent in parallel to different input terminals of a multiplexer on the signal conditioning circuit board.

4. The multi-channel synchronous signal acquisition system according to claim 3, characterized in that, Within each signal acquisition channel, the received global synchronization clock signal drives the analog-to-digital converter controller to perform strictly equal-interval alternating sampling of the acoustic signal after adjustable gain amplification and the temperature signal after isolation conversion. Specifically: Each signal conditioning circuit board has an analog-to-digital converter controller with an internal sampling sequence generator. The sampling sequence generator receives a global synchronization clock signal from the central aggregation node and uses it as its own reference clock. The sampling sequence generator is pre-configured with an alternating sampling mode, which instructs the analog-to-digital converter controller to sequentially sample and convert the acoustic analog signal and the temperature analog signal input from the multiplexer; At the rising edge of each global synchronization clock signal, the analog-to-digital converter controller controls the multiplexer to switch the input channel once and start an analog-to-digital conversion once, according to the alternating sampling mode; After the analog-to-digital conversion is completed, the digital quantity is latched into the output data register inside the analog-to-digital converter controller. At the same time, the sampling sequence generator records a local clock count value for the digital quantity. The local clock count value is accumulated based on the global synchronization clock signal and serves as the precise sampling time of the digital quantity. During a complete sampling period, the acoustic analog signal and the temperature analog signal are each sampled once, thereby generating a pair of acoustic digital samples and temperature digital samples that are strictly correlated in time.

5. A multi-channel synchronous signal acquisition system according to claim 1, characterized in that, The process involves temporarily storing the alternating acoustic wave digital sample sequence and temperature digital sample sequence obtained in each signal acquisition channel in the corresponding channel-specific buffer memory, and attaching a channel identifier code and a timestamp generated by the global synchronization clock. Specifically: Each signal conditioning board is equipped with a dual-port buffer memory as a channel-specific buffer memory; Each time the analog-to-digital converter completes the acquisition of a pair of acoustic digital sample sequences and temperature digital sample sequences, it writes the acoustic digital sample sequences and temperature digital sample sequences, along with their local clock count values, into a designated queue in the channel-specific buffer memory as a data packet. An independent formatting microprocessor reads data packets from the channel-specific buffer queue, adds the physical channel number information of this channel to each data packet, and forms a channel identification code; The formatting microprocessor uses the local clock count value carried in the data packet to query a time mapping table corrected by a global synchronization clock signal, and converts the local clock count value into a unified timestamp based on absolute time. The data packets, with added channel identifiers and unified timestamps, are rewritten into another send queue area of ​​the channel-specific buffer, awaiting reading by the central aggregation node.

6. The multi-channel synchronous signal acquisition system according to claim 1, characterized in that, The central aggregation node is configured to sequentially poll the channel-specific buffer memory of all signal acquisition channels according to a preset period, and read the sound wave and temperature digital sample data packets containing timestamps and channel identifiers. Specifically: The central aggregation node contains a main control processor and a multi-channel digital interface controller; The main control processor is configured with a polling schedule, which specifies the order and time interval for reading data from each signal acquisition channel; According to the polling schedule, the multi-channel digital interface controller sequentially accesses the dual-port buffer memory on each signal conditioning board via a high-speed serial bus. When accessing a specified channel, the multi-channel digital interface controller reads one or more complete data packets from the transmit queue area of ​​the channel-specific buffer memory of the specified channel until the marker at the end of the queue is read; Each read data packet is transmitted to the main memory of the central aggregation node via a high-speed serial bus and stored according to its channel identifier code; After completing a round of polling all signal acquisition channels, the main control processor checks the data packets from different channels in the main memory that have the same or very close timestamps, and marks them as data sets belonging to the same synchronous acquisition time.

7. A multi-channel synchronous signal acquisition system according to claim 6, characterized in that, Within the central aggregation node, timestamp-aligned digital acoustic sample sequences from different signal acquisition channels are combined into a multidimensional acoustic array snapshot, and timestamp-aligned digital temperature sample sequences from different signal acquisition channels are combined into a multidimensional temperature array snapshot. Specifically: In the main memory of the central aggregation node, a multidimensional array storage area is allocated for acoustic data and temperature data respectively. The dimensions of the multidimensional array include time dimension, channel dimension and sample value dimension. The main control processor traverses a set of data acquired at a synchronous acquisition time and determines which channel index position of the multidimensional array should be stored for the acoustic digital sample sequence and temperature digital sample sequence based on the channel identifier code in each data packet. For acoustic data, the digital sample sequences of all channels at the same time are arranged into a one-dimensional acoustic sample vector according to the circular arrangement order of the channels. The one-dimensional acoustic sample vector is a snapshot of a multi-dimensional acoustic array. For temperature data, the temperature digital sample sequence of all channels at the same time is arranged into a one-dimensional temperature sample vector according to the circular arrangement order of the channels. The one-dimensional temperature sample vector is a multi-dimensional temperature array snapshot. The obtained one-dimensional acoustic wave sample vector and one-dimensional temperature sample vector are saved to the corresponding time index in the acoustic wave data multidimensional array and the temperature data multidimensional array, respectively, thereby constructing a time-varying acoustic wave array snapshot sequence and temperature array snapshot sequence.

8. A multi-channel synchronous signal acquisition system according to claim 7, characterized in that, The spatial beamforming calculation is performed on each multidimensional acoustic array snapshot to obtain a sound field energy distribution map characterizing the spatial distribution of the sound source on the surface of the GIS equipment housing, specifically: The main control processor of the central aggregation node loads a pre-calculated beamforming weight matrix, which is calculated based on the annular arrangement coordinates of all acoustic wave sensing probes on the surface of the GIS equipment housing and a preset scanning angle grid. For each multidimensional acoustic array snapshot in the multidimensional array of acoustic data, that is, a one-dimensional acoustic sample vector, the main control processor multiplies it with the beamforming weight matrix. The result of the multiplication operation is a two-dimensional complex matrix, where the number of rows corresponds to the number of scanning angle grids and the number of columns corresponds to the number of frequency components. The square of the modulus of each element of the two-dimensional complex matrix is ​​used to obtain the signal power value at each scanning angle and each frequency point; The signal power values ​​at all scanning angles are arranged according to their azimuth angles corresponding to the annular cross section of the GIS equipment housing, and a polar coordinate energy distribution map is drawn with angle as the horizontal axis and energy intensity as the color. The polar coordinate energy distribution map is the sound field energy distribution map. Add a timestamp to the corresponding multidimensional acoustic array snapshot for each sound field energy distribution map.

9. A multi-channel synchronous signal acquisition system according to claim 8, characterized in that, The spatial interpolation calculation performed on each multidimensional temperature array snapshot yields a continuous temperature field distribution map covering the surface of the GIS equipment housing, specifically: The main control processor of the central aggregation node loads the three-dimensional geometric model of the GIS equipment shell and extracts the annular cross-sectional contour line where the acoustic temperature sensing unit is located from the three-dimensional geometric model. For each multidimensional temperature array snapshot in the multidimensional array of temperature data, i.e. a one-dimensional temperature sample vector, the main control processor binds each sample value to the position coordinates of the acoustic temperature sensing unit corresponding to its channel identifier on the annular contour line. A scattered interpolation algorithm based on radial basis functions is used to calculate the temperature estimate of any point on the entire annular cross-section region enclosed by the annular cross-section profile, based on the temperature sample values ​​of known coordinate points on the annular cross-section profile. The entire annular cross-sectional region is discretized into a high-resolution two-dimensional grid, and the temperature estimate of each grid node is calculated. Based on the temperature estimates of all grid nodes, a pseudo-color image is generated using a color map, where colors represent temperature levels. This pseudo-color image is a continuous temperature field distribution map. Add a timestamp to the corresponding multidimensional temperature array snapshot for each continuous temperature field distribution map.

10. A multi-channel synchronous signal acquisition system according to claim 9, characterized in that, The process of aligning the sound field energy distribution map and temperature field distribution map at the same timestamp with spatial coordinates and fusing them to generate a GIS equipment status feature map with sound and temperature superposition information is as follows: The main control processor retrieves a sound field energy distribution map and a continuous temperature field distribution map with the same timestamp from the channel-specific buffer memory; The main control processor first resamples the sound field energy distribution map to the same two-dimensional grid coordinate system as the continuous temperature field distribution map through coordinate transformation; In a unified two-dimensional grid coordinate system, each pixel simultaneously has an energy intensity value from the resampled sound field energy distribution map and a temperature value from the continuous temperature field distribution map; Define a fusion function that takes energy intensity value and temperature value as input and calculates a fusion feature value. The specific form of the fusion function is to perform a weighted summation of the normalized energy intensity value and the normalized temperature value. The calculated fusion feature values ​​are mapped onto a new color bar to generate a single-channel fusion image; The single-channel fused image is overlaid with the original continuous temperature field distribution map in pseudo-color, where the brightness of the fused image represents the fusion feature value and the hue of the temperature field distribution map represents the temperature value, ultimately generating a GIS equipment status feature map that simultaneously visualizes sound field energy and temperature information. The generated GIS device status feature map, along with the timestamp, is stored in the non-volatile memory of the central aggregation node.

Images