38 results about "Phase domain" patented technology

The invention relates to the technical field of precise geometric quantity measurement, and discloses an electronic component size and contour automatic metering method, which comprises the following steps: calculating a nominal truncation phase distribution diagram of a to-be-measured electronic component by using a virtual projection imaging model, controlling a projection unit to project a single-frequency sine stripe grating to the electronic component and collecting a deformed stripe image, calculating the gradient of the modulation degree distribution map and extracting edge contour features to solve the actual pose of the assembly; performing modulo 2pi differential operation on the nominal and actually measured truncated phase distribution maps, and eliminating high-gradient artifacts by using an edge mask to generate a phase residual map; and converting the phase residual image into a height deviation distribution diagram according to the phase height mapping relation, and by constructing a phase domain differential metering channel, the phase unwrapping ambiguity in discontinuous surface measurement is avoided.

The invention relates to the technical field of vibration monitoring, and discloses a steam turbine cold end fault diagnosis method and system based on a fusion time sequence convolutional network, and the method comprises the steps: eliminating the frequency spectrum drift caused by the rotation speed fluctuation through band-limited filtering and order domain resampling, and automatically extracting the specific low-frequency modulation component of a cold end through the envelope square and a periodogram, and finally, the asymmetric dynamic response characteristics of the rotor system are quantified by third-order spectrum phase bias, and the phase change rule is adaptively learned by the pre-trained convolutional network. According to the method, the problem that traditional amplitude spectrum analysis is high in misjudgment rate under the strong noise, variable speed and unsteady environment is effectively solved, the fault features show higher stability and distinguishability in the phase domain, and therefore high-robustness and high-precision automatic diagnosis of the turbine cold end rotor faults is achieved.

The invention discloses a titanium-aluminum alloy defect detection device and method.According to the titanium-aluminum alloy defect detection device, a micro-texture film is attached to the surface of a detected workpiece in the defect detection process, so that a known interference source is introduced, and the defect detection accuracy is improved by combining a forward and reverse constant-speed scanning strategy; signal processing is carried out after the forward phase domain response signal and the reverse phase domain response signal are obtained, and a real defect signal can be accurately extracted, so that a surface roughness interference signal can be effectively eliminated on the premise that the surface of a workpiece is not damaged, the detection rate of titanium-aluminum alloy surface and near-surface defects is improved, the false detection rate is reduced, and the detection efficiency is improved. And high resolution and high signal-to-noise ratio are maintained.

The invention provides a DDR wafer test anomaly detection method and device based on machine learning, and the method comprises the steps: obtaining a time sequence eye pattern synchronously collected in a DDR wafer test process and time sequence waveform data related to total jitter, deterministic jitter and random jitter, and carrying out the clock edge phase segmentation alignment of differential time sequence sampling points in the time sequence waveform data, phase domain dense sampling data is constructed, adaptive sparse coding is completed in combination with DDR jitter physical priori, noise and normal process fluctuation related components are removed, defect sensitive sparse components are obtained, variational mode decomposition is performed, components related to process drift, normal process fluctuation and abrupt change abnormity are obtained, and the defect sensitive sparse components are subjected to sparse coding; and the abrupt change anomaly correlation component is extracted as target detection data, a preset machine learning model is input, an anomaly mode correlation result is generated, a DDR wafer test anomaly detection result is output based on the result, and a corresponding test setting item adjustment instruction is generated. According to the invention, the overall efficiency and the result reliability of the DDR wafer test can be improved.

The invention discloses a laser powder bed sheet pattern distortion real-time correction system and method based on a phase type light valve, and relates to the technical field of metal additive manufacturing. The system comprises a polarization laser, a beam expanding and collimating unit, a reflecting mirror, a phase-type spatial light modulator, a reflecting mirror, a relay optical system, a galvanometer / field lens system, a movable spectroscope, a reference plane mirror and a forming cylinder. A double-loop correction system with a position sensitive detector as an absolute reference and a wavefront sensor as direct feedback is constructed, information of the position sensitive detector and the wavefront sensor is unified in a phase domain through a frequency domain fusion algorithm, and a comprehensive compensation phase diagram capable of completing pattern generation and distortion correction at the same time is generated. And loading to a phase type light valve to carry out high-power laser patterning modulation, and melting metal powder of a powder bed.

The application provides a multi-satellite signal acquisition method based on a graphics processor. Through the asynchronous pipeline architecture of CPU-GPU cooperation, after completing parameter configuration and data asynchronous uploading at the central processor end, the graphics processor end divides multiple CUDA streams based on the parallel subtasks of the CUDA stream based on the Doppler dimension, combines the pre-bound Fourier transform plan, accelerates the down-conversion through the frequency lookup table, performs frequency domain correlation and non-coherent accumulation in the code phase domain on multiple CUDA streams in parallel, and finally extracts the local peak value through one-level reduction and returns it asynchronously. The central processor end summarizes the results and judges the effectiveness based on the carrier-to-noise ratio. The design realizes the deep integration of the Doppler task-level pipeline and the code phase thread-level parallelism, avoids non-merged access of the GPU device memory, significantly improves the efficiency of multi-constellation signal acquisition, and supports multi-frequency point signal processing of GPS, Beidou and the like through a unified pre-configuration interface, thereby enhancing the architecture universality.

This invention discloses a method and system for analyzing circuit breaker breaking parameters based on phase domain trajectory analysis. The method includes: first, constructing a time-domain waveform expression for the short-circuit current; after acquiring the fault waveform, mapping the time-domain sampling points to the polar coordinate plane by defining reference points and sampling windows to form a phase domain trajectory and drawing it; then, performing waveform reproduction and similarity analysis, adjusting the initial phase of the sinusoidal AC component and the attenuated DC component to reproduce a high-precision time-domain waveform and improve the short-circuit current time-domain waveform expression; next, using vectorized matching technology based on digital channel names and kurtosis algorithms to screen out associated circuit breakers with displacement in the fault recording; then, calculating key parameters such as the DC time constant, fault current duration, and RMS value of the total current, and comparing and verifying them with the rated parameters of the circuit breaker. This improves the accuracy and real-time performance of circuit breaker breaking capacity assessment, providing an efficient and automated method for circuit breaker performance evaluation and anomaly diagnosis after power grid faults.

The application provides a multi-satellite signal acquisition method based on a graphics processor. Through the asynchronous pipeline architecture of CPU-GPU cooperation, after completing parameter configuration and data asynchronous uploading at the central processor end, the graphics processor end divides multiple CUDA streams based on the parallel subtasks of the CUDA stream based on the Doppler dimension, combines the pre-bound Fourier transform plan, accelerates the down-conversion through the frequency lookup table, performs frequency domain correlation and non-coherent accumulation in the code phase domain on multiple CUDA streams in parallel, and finally extracts the local peak value through one-level reduction and returns it asynchronously. The central processor end summarizes the results and judges the effectiveness based on the carrier-to-noise ratio. The design realizes the deep integration of the Doppler task-level pipeline and the code phase thread-level parallelism, avoids non-merged access of the GPU device memory, significantly improves the efficiency of multi-constellation signal acquisition, and supports multi-frequency point signal processing of GPS, Beidou and the like through a unified pre-configuration interface, thereby enhancing the architecture universality.

The invention discloses a power distribution network fault diagnosis method and system based on improved empirical mode decomposition and a hybrid neural network, and the method comprises the following steps: obtaining high-frequency fault disturbance waveform data, and dividing a fault disturbance waveform into a plurality of data segments through a moving time window; obtaining an intrinsic mode component set for each data segment by using an improved empirical mode decomposition method; a random component in the intrinsic mode component is screened and removed by using an improved run detection method containing an adaptive factor; mapping the processed three-phase voltage transient waveform to a complex field plane by using phase domain transformation; and constructing a power distribution network fault diagnosis model by using the multi-branch hybrid neural network, the input signal being the fault data converted to the complex field plane, and the output data being the fault type.

The invention provides a visual positioning measurement method and system based on a micro-coding target. The method comprises the following steps: acquiring a target image through a module camera matched with a miniature microscope lens, and firstly, realizing automatic estimation of a distortion center through minimizing a phase nonlinear residual by using global linear structure characteristics of an orthogonal dot matrix in a frequency domain and a phase domain; on the basis, low-order radial and tangential distortion parameters are solved by adopting closed weighted least squares, so that single-image lens distortion correction is realized; then, a plurality of target images at different local positions are shot from different positions, each image is decoded to obtain global position information of the photoetching plate corresponding to the feature points, self-calibration of the camera is achieved, and internal parameters of the camera are obtained; and finally, performing two-dimensional Fourier transform on the acquired target image, performing unwrapping by using phase information of a complex field, and fusing and outputting the absolute position of the camera in combination with decoding information of the image.

A single-ended wired transmitter based on adaptive amplitude-phase cross-domain equalization comprises a data unit and a clock unit, and the clock unit provides multi-frequency and multi-phase clock signals for a serialization and driving circuit of the data unit through frequency division, duty cycle calibration and quadrature phase calibration according to an external differential input clock; the data unit performs amplitude pre-emphasis processing according to the hopping information to generate an equalization control signal, and performs phase control on the four-phase clock according to the length of the continuous same code element to obtain a clock signal for phase domain equalization; through the adaptive amplitude-phase cross-domain equalization technology, the mixed-mode driver technology, the current multiplexing technology of the phase modulator and the power gating technology of the serializer, the power consumption and the power noise can be remarkably reduced and the equalization capability can be improved in the memory interface and core particle interconnection application scene.

The invention relates to the technical field of rotor misalignment fault identification, in particular to a wind turbine generator fault online monitoring method, device and equipment and a computer readable storage medium, and the method comprises the steps: installing a strain sensor measuring point on a diaphragm of a high-speed shaft coupling of a wind turbine generator, and constructing a strain measurement chain; acquiring a strain signal by using the strain measurement chain, and converting the strain signal into a digital signal; performing frequency domain and phase domain feature extraction on the digital signal to obtain a feature extraction result; and identifying a rotating shaft misalignment fault type based on a feature extraction result to obtain a fault monitoring result. The non-centering state is represented by directly measuring the strain of the coupler diaphragm, direct evidence reflecting the centering state of the rotating shaft is obtained, and the signal-to-noise ratio and the monitoring accuracy are remarkably improved; the frequency domain amplitude features and the phase domain phase features are combined, the accuracy and reliability of fault type recognition are greatly improved, and the method has significant significance in guaranteeing safe and stable operation of the wind turbine generator and reducing the operation and maintenance cost.

This application discloses a digital predistortion calculation method, apparatus, storage medium, and electronic device. The digital predistortion calculation method includes: acquiring the input signal and corresponding output signal of a power amplifier; determining the nonlinear distortion characteristics of the power amplifier in the amplitude and phase domains based on the input and output signals to identify symmetry breaking modes characterizing the nonlinear distortion characteristics; mapping the symmetry breaking modes to a Lie algebra space, and analyzing the distortion mechanism constituting the symmetry breaking modes based on the Lie algebra space to obtain Lie algebra elements; performing exponential mapping operations on the Lie algebra elements to generate the corresponding symmetry recovery transform function; constructing a performance manifold based on the physical constraints and linearization performance indicators of the power amplifier, and geometrically optimizing the parameters of the symmetry recovery transform function on the performance manifold to generate the target predistortion signal.

The application provides a detection system and an automatic focusing method and a storage medium thereof. The system comprises a light source assembly, an optical assembly, an imaging assembly and a processor. A surface of a to-be-detected object is provided with a feature structure for focus surface positioning. The processor is configured to coarsely adjust a relative position of the imaging assembly and the to-be-detected object, collect interference images at multiple positions, determine an initial focus surface position according to the collected interference images, then change the relative position of the imaging assembly and the to-be-detected object multiple times near the initial focus surface position and collect interference images at corresponding positions respectively, construct a sharpness index by using phase information corresponding to the feature structure, and compare index values at the positions to accurately position a best focus surface position. The application improves the focus adjustment basis from traditional image intensity analysis to phase domain analysis, utilizes high sensitivity of the phase to defocusing, realizes automatic focus adjustment with nanometer-level precision without increasing hardware, and provides reliable focus surface guarantee for high-precision semiconductor detection.

This invention relates to the field of precision geometric measurement technology and discloses an automated measurement method for the dimensions and contours of electronic components. The method includes: calculating the nominal truncated phase distribution map of the electronic component under test using a virtual projection imaging model; controlling a projection unit to project a single-frequency sinusoidal fringe grating onto the electronic component and acquiring deformed fringe images; calculating the modulation distribution map gradient and extracting edge contour features to solve the actual pose of the component; performing modulo-2π differential operations on the nominal and measured truncated phase distribution maps and using an edge mask to remove high-gradient artifacts to generate a phase residual map; and converting the phase residual map into a height deviation distribution map based on the phase height mapping relationship. This invention avoids the ambiguity of phase unwrapping in the measurement of discontinuous surfaces by constructing a phase domain differential measurement channel.

The application discloses a short offset transient electromagnetic method super-deep detection system and a signal analysis method, relates to the technical field of geological detection, and comprises a signal receiving and collecting adaptive module, an original signal preprocessing module, a geometric effect correction and deep gain reconstruction module, a layered response separation module, an inversion imaging and prior constraint module and an analysis result output and visualization module. The near-field coupling, amplitude and phase distortion and deep energy attenuation caused by the short offset are systematically compensated through the quantitative geometric effect correction and the depth self-adaptive gain reconstruction. The mechanism not only synchronously recovers the late and weak amplitude signal in the time domain and the phase domain, but also maintains the amplitude and phase consistency, ensures the use of the homologous and comparable data input for the subsequent layered separation and inversion, and fundamentally expands the applicability of the short offset transient electromagnetic method to the super-deep target. Compared with the method of increasing the transmission or expanding the array, the scheme can stably extract the deep response when the equipment and the field layout are limited, and fundamentally expands the applicability of the short offset transient electromagnetic method to the super-deep target.

The invention provides a same-tower four-circuit non-transposition parallel line zero-sequence parameter live line measurement method, and relates to the technical field of intelligent power grids, and the method comprises the steps: obtaining the power parameters of a multi-circuit line in a target connection mode; establishing a port equivalence equation according to the electric power parameters, and solving a pilot frequency phase domain impedance parameter matrix by adopting a least square method; and according to the pilot frequency phase domain impedance parameter matrix, converting to a power frequency and carrying out sequence domain transformation, and calculating to obtain a power frequency zero sequence parameter of each power transmission line. According to the invention, the complex electromagnetic characteristics of an untransposed line can be accurately matched, and the accuracy of zero-sequence parameter measurement is improved.

The invention discloses a system and a method for testing broadband damping characteristics of a network construction type energy storage converter, and relates to the technical field of power electronic equipment testing, the system comprises the following components: a physical layer, a data acquisition layer, a data processing layer and an application layer; according to the method, the sweep frequency disturbance signal is optimized through a dynamic amplitude adjustment strategy, the problems that traditional fixed amplitude disturbance is weak in signal in a high-impedance frequency band and prone to overload in a low-impedance frequency band are solved, tests show that after dynamic amplitude adjustment is adopted, the signal-to-noise ratio of data collected in each frequency band is averagely increased by 28%, the signal-to-noise ratio of the low-frequency band of 1-45 Hz is increased most remarkably, and meanwhile, the signal-to-noise ratio of the low-frequency band of 1-45 Hz is increased more remarkably. And the data acquisition layer adopts a method of combining timestamp calibration and phase compensation, the synchronization precision is improved to a phase domain less than or equal to 0.5 degree, and the phase deviation near a 50Hz rated frequency is controlled within 0.2 degree, so that an impedance calculation error caused by the phase deviation is effectively avoided, and the accuracy of the Bode diagram phase-frequency characteristic is ensured.

This invention provides a machine learning-based method and device for anomaly detection in DDR wafer testing. It acquires timing eye diagrams and timing waveform data related to total jitter, deterministic jitter, and random jitter synchronously during DDR wafer testing. The method performs clock edge phase segmentation alignment on the differential timing sampling points in the timing waveform data to construct dense sampling data in the phase domain. Combined with DDR jitter physical priors, adaptive sparse coding is performed to remove noise and components related to normal process fluctuations, obtaining defect-sensitive sparse components. Variational mode decomposition is then performed to obtain components related to process drift, normal process fluctuations, and abrupt anomalies. The components related to abrupt anomalies are extracted as target detection data, input into a preset machine learning model, and anomaly pattern association results are generated. Based on these, DDR wafer testing anomaly detection results are output, and corresponding test setting adjustment instructions are generated. This invention can improve the overall efficiency and reliability of DDR wafer testing results.

The invention relates to the field of power transmission of a power system, and provides a phase domain frequency change modeling method and device for a direct current transmission line, and the method comprises the steps: building a phase domain frequency change mathematical model based on the physical characteristics of the direct current transmission line; when the propagation time delay of the electromagnetic wave on the direct-current power transmission line and the simulation step length are in a non-integer multiple relationship, adaptively adjusting interpolation table parameters through the trained deep learning model, and performing interpolation reconstruction on the reflection current based on the interpolation table parameters; the deep learning model is used for representing a nonlinear mapping relation between interpolation table parameters and system dynamic response; and based on the reflection current after interpolation reconstruction and the phase domain frequency change mathematical model, adopting a recursive convolution algorithm to carry out time domain simulation under a non-linear sudden change working condition. According to the DC power transmission line phase domain frequency change modeling method and device provided by the invention, the modeling precision of the DC power transmission line in a high-frequency transient process is remarkably improved, and waveform distortion and simulation errors are reduced.

This invention discloses a titanium-aluminum alloy defect detection device and method. The titanium-aluminum alloy defect detection device of this invention introduces known interference sources by attaching a micro-textured film to the surface of the workpiece during the defect detection process. Combined with a forward and reverse isokinetic scanning strategy, after obtaining the forward phase domain response signal and the reverse phase domain response signal, signal processing is performed to accurately extract the real defect signal. Thus, without damaging the workpiece surface, the surface roughness interference signal can be effectively eliminated, thereby improving the detection rate of surface and near-surface defects of titanium-aluminum alloys, reducing the false detection rate, and maintaining high resolution and high signal-to-noise ratio.

The invention relates to the technical field of material testing, in particular to a state analysis system for parallel running multi-path structural members, and provides the following scheme: aiming at at least two path structural members arranged on the same base, the acceleration, torque and position signals of active path traction assemblies on each path are acquired; and determining an induction signal of the path structural member in the excitation working condition and a response signal of the response path. The energy coupling characteristics between different paths are quantified by projecting a signal from a time sequence domain to a position phase domain and extracting local frequency characteristics under energy flow redistribution. And extracting characteristic parameters reflecting the structural rigidity, the connection fastening state and the fatigue of the supporting piece based on the mutual interference matrix, and outputting the mechanical state of the multi-path structural piece. Spatial resolution diagnosis of a multi-path parallel system is achieved, the cross-path mutual interference rule can be accurately recognized under the non-stable working condition, and the dynamic stability of equipment operation and the structural health monitoring precision are improved.

The invention provides a music driving action generation method based on double-flow decoupling and continuous phase field modeling, and belongs to the technical field of artificial intelligence and multimedia generation. According to the method, an audio feature extraction strategy of amplitude domain and phase domain combined modeling is constructed, music semantics and energy information are represented by logarithmic Mel spectrum, and a periodic structure is described by phase rhythm features; introducing a continuous beat phase function to construct a dense periodic coordinate system, and mapping a discrete beat signal into a continuous phase field with a circumferential topological structure; a double-flow cross attention injection mechanism is adopted in the diffusion generation framework, and the beat control intensity is scheduled through a gating potential energy window; and the training stability is improved by adopting an original prediction parameterization form. According to the method, human body actions with accurate rhythm alignment, consistent semantics and continuous and natural time sequence are generated.

The invention provides a time phase domain random modulation ultra-long distance Brillouin optical time domain reflectometer, and belongs to the technical field of time phase domain random modulation Brillouin optical time domain reflectometers. In order to solve the technical problem that the spatial resolution and the sensing distance of an existing Brillouin optical time domain reflectometer cannot be increased at the same time, the adopted technical scheme is as follows: the output end of a narrow linewidth laser is connected to the input end a of a first optical fiber coupler; the output end b of the first optical fiber coupler is connected to the input end d of the pulse modulator; the output end f of the pulse modulator is connected to the input end g of the phase modulator; the output end A of the arbitrary waveform generator is connected to the input end e of the pulse modulator, and the output end B of the arbitrary waveform generator is connected to the input end h of the phase modulator; the output end i of the phase modulator is connected to the input end of the pulsed light amplifier; the output end of the pulse optical amplifier is connected to the input end of the first optical filter; the method is applied to the Brillouin optical time domain reflectometer.

The invention provides a laser epitaxial wafer. The laser epitaxial wafer comprises a Si substrate, an Al < x > Ga < 1-x > As seed layer, a GaAs buffer layer, a first overlapping layer, an n-type GaAs contact layer, an n-type Al < z > Ga < 1-z > As waveguide layer, a second overlapping layer, a p-type Al < g > Ga < 1-g > As waveguide layer and a p-type GaAs contact layer which are sequentially overlapped. According to the invention, through the growth of the Al < x > Ga < 1-x > As seed layer, the generation of an anti-phase domain (APB) is inhibited from the source, and then through the GaAs buffer layer, the existing anti-phase domain (APB) is annihilated in the GaAs buffer layer. Through the introduction of the InyGa1-yAs / GaAs superlattice, the dislocation density is greatly reduced, and the level of preparing a high-performance laser is achieved.

The application provides a voiceprint recognition method, which comprises the following steps: preprocessing a to-be-recognized voice signal to obtain an effective voice signal; determining a sentence-level voiceprint feature according to the effective voice signal based on a finite difference network; the finite difference network comprises a time-frequency transformation subnetwork and a phase transformation subnetwork, wherein the time-frequency transformation subnetwork is used for performing multi-scale filtering on a frequency domain spectrum of the effective voice signal to obtain a frequency spectrum domain voice signal representation; the phase transformation subnetwork is used for extracting a phase domain voice signal representation of the effective voice signal; and voiceprint recognition is performed according to the sentence-level voiceprint feature. The scheme uses an original audio as an input, obtains speaker identity description information from a frequency spectrum domain and a phase domain, and fuses the information, so that the problem of speaker identity information loss can be avoided, and the precision of voiceprint recognition is improved.

The application belongs to the technical field of metasurfaces, and relates to a core granulation metamaterial for electromagnetic wave regulation, which comprises: a plurality of arrayed metasurface units; a core granule is arranged in each gap of a radiation layer of the metasurface unit to form a circuit structure in which the core granule is connected in parallel with a capacitor; the core granule comprises: one or more core granule branches and a diode branch in parallel; the core granule branch comprises a resistor, an inductor and / or a capacitor; the diode branch comprises one diode; different configurations of the resistor, the inductor, the capacitor and the diode in the core granule correspond to different functions of the metamaterial in the frequency domain, the amplitude domain, the polarization domain and the phase domain, so that electromagnetic wave regulation of the core granulation metamaterial is realized. The application can decouple the physical structure of the metasurface and the core circuit of the electromagnetic response, and realize different electromagnetic functions in the frequency domain, the amplitude domain, the polarization domain and the phase domain by configuring different core granules.

The application discloses a kind of light automatic feed drilling device and control method thereof, device includes execution module, power module and control module being arranged at the one side of execution module, the middle part of the execution module is provided with vibration unit, the axial direction of the vibration unit is sequentially provided with stator, roller, cage and rotor, the contact surface side of the roller is corrugated surface, the other side is plane, the relative angle between the corrugated surface and the cage is phase angle, the power module includes main motor;Method includes by identifying the current following fluctuation characteristics of main motor caused by load fluctuation, predict the change of main motor current with phase angle or with time, i.e. in phase domain or time domain, so as to give the corresponding feed current compensation of main motor. By the method of current feed compensation, even under the condition of small inertia of power module, the smooth operation of motor can also be realized, the light weight of automatic feed drilling equipment is realized, and the labor load when worker holds operation is reduced.

The present application relates to the field of semiconductor laser, and particularly relates to a vertical cavity surface emitting laser and a preparation method thereof, the laser comprising a substrate; a buffer structure covering a surface of a germanium / silicon or germanium substrate, the buffer structure comprising: a nucleation layer in contact with the surface of the substrate, for providing nucleation points on the surface of the substrate; a first buffer layer covering the nucleation layer, for merging a three-dimensional island structure of the nucleation layer and forming a surface without anti-phase domains; a second buffer layer covering the first buffer layer, for providing a flat surface; a resonant cavity arranged on the buffer structure, comprising: a first Bragg mirror arranged on the second buffer layer, for reflecting light in the laser to form laser oscillation; a second Bragg mirror separated from the first Bragg mirror by an active layer, for reflecting light in the laser to form laser oscillation; the active layer for providing gain to generate laser; and an electrode arranged in the resonant cavity or in contact with the substrate, for current injection.

The application belongs to the technical field of power equipment insulation monitoring, and discloses a partial discharge comprehensive diagnosis system and method suitable for a substation environment, wherein the system comprises an acoustic sensor array module, a TEV sensor module, a synchronous data acquisition unit, a main control processing unit and a man-machine interaction and communication module. The system provides a unified global sampling clock for multiple signals through the synchronous data acquisition unit with FPGA as the core, and realizes the hardware-level microsecond-to-nanosecond synchronous acquisition. The method is based on the synchronous acquisition of acoustic waves and TEV signals, extracts time domain, frequency domain and phase domain features in parallel, carries out intelligent diagnosis and event type discrimination through a two-stage cascaded hybrid classification model, and realizes high-precision three-dimensional spatial positioning of the discharge source by using a hyperbolic equation set for solving after fusing acoustic TDOA positioning and TEV absolute time reference. The application has the advantages of edge intelligence, strong anti-interference capability and accurate positioning.