1988results about "Power supply testing" patented technology

The invention discloses an energy storage power supply residual life prediction method and system based on dynamic weight distribution, and belongs to the technical field of energy storage power supply health management. The method comprises the following steps: acquiring voltage, current and temperature time sequence signals of energy storage power supply operation through a multi-source sensor, and constructing a degradation characteristic sequence by adopting a sliding window method; a degradation inflection point division stage is detected and identified by using a curvature extreme value, a historical degradation mode is matched based on a dynamic time warping algorithm, and an optimal model group is selected; respectively carrying out dynamic weight distribution on the time step length and the feature dimension by adopting a dual-channel attention network, and fusing a learnable coefficient with space-time attention output to generate a life prediction result; residual errors and feature drift are monitored and predicted in combination with an incremental learning mechanism, and model parameters are updated and a degradation knowledge base is expanded by adopting an elastic weight consolidation algorithm. According to the method, the key information capturing capability is enhanced through a space-time attention mechanism, the model adaptability is optimized in combination with incremental learning, and the energy storage power supply life prediction precision and the working condition generalization performance are remarkably improved.

The invention provides a multi-dimensional grid-connected test system for a photovoltaic string inverter, and the system comprises a power grid simulation module which is used for constructing a power grid simulation environment based on a multi-dimensional coupling mechanism of voltage fluctuation, frequency deviation and harmonic distortion; the photovoltaic array simulation module is used for constructing IV characteristic dynamic deduction models of different irradiance and temperature gradient effects based on a photovoltaic array equivalent circuit physical model to realize a natural simulation environment; the loading module is used for loading the power grid simulation environment and the natural simulation environment to the photovoltaic string inverter to be tested; the data acquisition module is used for monitoring running state data of the photovoltaic string type inverter in real time and constructing a running state data set containing multi-physical-quantity coupling characteristics; and the diagnosis module is used for inputting the running state data set into the trained LSTM neural network model to obtain a test result of the photovoltaic string inverter. According to the invention, the efficiency and convenience of the grid-connected test of the inverter can be improved, and the accuracy and reliability of the grid-connected performance of the inverter are ensured.

The invention discloses an intelligent operation and maintenance fault early warning method for a rail transit variable-frequency power supply, and relates to the technical field of power electronics and intelligent operation and maintenance, and the method comprises the following steps: setting a multi-point synchronous collection device to carry out the high-precision real-time collection of voltage and current signals of a variable-frequency power supply system, and carrying out the digital coding processing of data; and performing frequency domain analysis on the acquired voltage and current signals by using a preset spectrum analysis algorithm, and extracting harmonic amplitudes and phase characteristic parameters of different orders. According to the invention, through high-precision synchronous acquisition and spectrum analysis, rapid and accurate identification of harmonic abnormity is realized; a machine learning model and dynamic threshold adjustment are fused, and the adaptability and intelligent recognition capability of the system to fault modes under multiple working conditions are enhanced; and a trend prediction and graphical early warning platform is introduced, so that the fault risk is pre-judged in advance and visually responded, and the operation and maintenance efficiency and the operation safety of the rail transit power supply system are remarkably improved.

The invention discloses an output voltage detection method and system based on an aviation power supply, and relates to the technical field of aviation power supply voltage detection. A differential voltage sensor is connected to the output end of an aviation power supply, a high-frequency suppression capacitor is connected in parallel, the sampling rate of 20-50 times of the working frequency is set according to the type of the power supply, and the signals are collected and converted into analog voltage signals; the analog signal is subjected to two-stage filtering, is amplified to 0-5V through an instrument amplifier, and is converted into a digital voltage sequence through a 16-bit ADC (Analog to Digital Converter); calculating AC power supply effective value, peak value and other parameters, DC power supply average value, ripple coefficient and other parameters; comparing the characteristic parameters with a preset threshold value, determining that the characteristic parameters exceed the threshold value and recording related information; and a detection result is transmitted to a monitoring terminal through an RS485, data is locally stored, and a switching value is output to an alarm device when the data is abnormal. The method adapts to aviation power supply load fluctuation and environment temperature change, detection precision is high, data transmission and storage are safe, abnormity is accurately identified and early warning is carried out in advance, and aviation power supply operation and maintenance reliability and efficiency are improved.

The invention discloses a remote real-time monitoring and fault early warning method for a power supply operation state, and relates to the technical field of power electronics and energy management, and the method comprises the steps: collecting the temperature data of a key part of a power supply through an optical fiber Bragg grating sensor array, and analyzing the wavelength offset; electrical parameters are collected through a differential MEMS current / voltage sensor, and a permalloy / ferrite composite shielding layer is arranged in the sensor; collecting a mechanical vibration signal through a piezoelectric ceramic vibration sensor; for device data from different manufacturers, Modbus and CAN heterogeneous protocols are matched and analyzed through a regular expression, and the Modbus and CAN heterogeneous protocols are converted into a unified RDF triple format; inputting temperature, current and vibration data into a space-time alignment module, and constructing a semantic mapping relation by adopting an ontology; calculating a sensor confidence coefficient weight, and performing weighted fusion to generate an equipment health index; operating a lightweight LSTM model at an edge node, and detecting current harmonic and vibration resonance frequency band abnormity in real time; cOMSOL multi-physical field simulation is combined at the cloud end, and the evaporation rate of the capacitor electrolyte is predicted.

The invention relates to the technical field of aircraft power supply systems, and discloses an intelligent power distribution and fault diagnosis method for an aircraft power supply system, which comprises the following steps: acquiring a real-time current signal and load working condition information of an aircraft power supply bus, a double-domain composite entropy feature is constructed by extracting a time domain entropy transition index representing the energy domain stability of a current signal and a phase synchronization index representing the time structural domain stability; according to the method, dual inspection of the energy domain and the time structure domain is carried out on the current signals, and cooperative judgment is carried out in combination with the load working condition, so that sensitive capture of real fault forebodes is realized; and meanwhile, tolerance is dynamically shown on normal loads in the airplane operation process, the restriction between safety and usability in the prior art is effectively avoided, and predictive health management is made possible.

A load connection state detection method for an energy storage power supply. The energy storage power supply comprises an inverter power supply. The method comprises the following steps: sampling an output voltage and an output current of the inverter power supply, and calculating output apparent power (S1); and determining whether the apparent power is less than a preset power threshold; if the apparent power is greater than the preset power threshold, a detection result being that a load is connected to the energy storage power supply; and if the apparent power is less than the preset power threshold, executing an active load online detection algorithm for secondary detection, and according to the secondary detection result, determining whether a load is connected (S2).

The invention relates to the technical field of photovoltaic inverter regulation and control, in particular to a device capable of adjusting, controlling and monitoring a photovoltaic inverter. Comprising an intelligent adjusting unit which adopts a multi-parameter dynamic matching algorithm based on power grid situation awareness; an adaptive protocol unit; a state monitoring unit; and a safety control unit. According to the method, a multi-parameter dynamic matching algorithm based on power grid situation awareness is adopted, power, current and voltage parameters output by the photovoltaic inverter and voltage, frequency real-time parameters and fluctuation trend parameters of the power grid are collected, and the power grid situation dynamic matching logic is combined, so that accurate dynamic adaptation between the output of the inverter and the state of the power grid can be realized; the problem of insufficient adaptability caused by the fact that the power grid situation is not considered in traditional single-parameter adjustment is solved; and PWM adjusting signals are generated through the parameter acquisition module, the power grid situation analysis module and the adjustment decision module to drive an IGBT switch tube, so that the accuracy and stability of output adjustment are guaranteed.

The invention discloses a state monitoring method and system for an outdoor emergency power supply. Firstly, original data such as battery internal resistance, charging efficiency, temperature fluctuation and load abrupt change are collected to form an initial data set; and then noise suppression and fusion processing are carried out on the data set to obtain a smoothed battery parameter sequence. And constructing a dynamic trend vector by analyzing the differential change of the sequence, judging a potential anomaly level according to the slope of the dynamic trend vector, and outputting an anomaly probability score. And identifying an inter-parameter interaction cluster based on clustering analysis, and performing causal reasoning on a dominant cluster to determine a fault dominant factor. And finally, in combination with the real-time state, the environmental parameters and historical defect information, constructing a modeling matrix, and executing multi-dimensional clustering to identify specific fault types such as bearing wear, insulation aging and the like. If the fault is caused by internal resistance abnormity, an interference source is further positioned, and an accurate fault position is output. According to the invention, the battery fault monitoring accuracy and positioning capability in a complex environment can be effectively improved.

The invention discloses a gas meter power supply abnormity monitoring system and method based on a signal acquisition technology, and relates to the technical field of gas meter power supplies, the system comprises a multi-dimensional sensing module, a dynamic adaptive module, a multi-dimensional fusion analysis module, an intelligent judgment module and a verification execution module; according to the invention, the multi-dimensional fusion analysis module accurately calculates key parameters such as a power supply ripple coefficient and load transient response time in a corresponding mode, the electromagnetic interference intensity is calculated through professional algorithms such as discrete Fourier transform in combination with an environment electric field and a magnetic field sequence, a threshold value is dynamically adjusted based on a working mode, and a comprehensive score is calculated through multi-parameter deviation weighted fusion. The power supply state is deeply and comprehensively evaluated; the intelligent judgment module scientifically divides anomaly levels according to comprehensive scores, quantifies and distinguishes real anomaly and environmental interference by means of an environmental interference contribution degree formula, solves the problem that traditional monitoring cannot accurately recognize power anomaly under complex working conditions, and improves anomaly judgment accuracy.

The invention relates to an energy storage converter performance evaluation method and device, equipment, a medium and a program product. The method comprises the following steps: acquiring performance index parameters of the energy storage converter under a plurality of evaluation dimensions; the evaluation dimensions at least comprise electrical performance, thermal performance, mechanical performance, environmental adaptability and control performance; performing correlation analysis processing on correlation degrees among the evaluation dimensions based on the performance index parameters to generate a correlation matrix; vectors in the incidence matrix are used for representing the correlation degree between every two evaluation dimensions; and performing statistical analysis and quantitative evaluation processing on the incidence matrix to obtain a performance evaluation result for the energy storage converter. By adopting the method, the accuracy of energy storage converter performance evaluation can be improved.

The invention discloses a POL power supply automatic test method based on modular design, and belongs to the technical field of power supply test. The method comprises the following steps: firstly, under the control of an upper computer, automatically reading POL power supply basic parameters including key characteristic parameters such as input / output voltage, rated current and frequency from a preset Excel table, and establishing a judgment reference for testing; then, instrument initialization is completed, communication connection with test equipment such as an oscilloscope, an electronic load and a direct-current power supply is established, and configuration of initialization function parameters is completed; then, executing a measurement task of each core parameter of the POL power supply to be tested, and in the test process, executing a parallel measurement management mechanism, namely configuring each instrument parameter and channel state in real time according to a specific test item, and triggering a retry or interrupt logic under the conditions of communication abnormality, data timeout and the like until the test recovers a running state; and finally, after the test is completed, performing Excel storage and interaction on the standardized test data and the waveform screenshot, and finally safely exiting. According to the POL power supply automatic test method based on the modular design, the whole test process is divided into a plurality of serial or parallel function processing units such as parameter reading, instrument initialization, test execution and test management, data storage and interaction and the like, and high efficiency and usability of a test system are achieved.

A test method, system, device and medium for a remote control power supply, the method comprising: performing initial detection on a to-be-detected power supply based on a first detection instruction sequence to obtain a first response behavior sequence; obtaining functional parameters of a to-be-detected power supply and environmental data of the position of the to-be-detected power supply; mapping the environment data to the first response behavior sequence to obtain a second response behavior sequence; identifying a chained behavior sequence in the second response behavior sequence; analyzing an abnormal coupling response relationship in the chain behavior sequence and calculating a coupling risk value; constructing a third response behavior sequence, and determining a detection instruction set corresponding to the third response behavior sequence; calculating the triggering probability of triggering the third response behavior sequence by each first instruction group in the detection instruction set; generating a second detection instruction sequence based on the triggering probability; and testing the to-be-detected power supply based on the second detection instruction sequence to obtain a detection result. The detection accuracy of the remote control power supply can be improved.

The invention discloses a vehicle-mounted power supply short-circuit heat prediction method based on big data analysis, and the method comprises the following steps: collecting real-time operation data, and storing the data in a data cache, thereby obtaining a cache time sequence signal; filtering the cached time sequence signal, and extracting a basic feature sequence; constructing an RC network model based on the basic feature sequence, and converting a fractional order operator into an equivalent integer order model; performing parameter estimation on the equivalent integer order model to generate a thermal prediction model; taking the thermal prediction model and the basic feature sequence as joint input and output power loss features; inputting the power loss characteristics into a quantile regression forest model to form a heat prediction result; comparing the heat prediction result with a heat threshold value and a temperature rise threshold value, and outputting a control trigger signal; and calling parameters in a control strategy library according to the control trigger signal to generate a control strategy. According to the invention, vehicle-mounted power supply short-circuit heat prediction is realized.

A power supply aging test system, which belongs to the technical field of power supply tests. The system comprises a control module, a load module, a data collection module, an environment simulation module, an early-warning module, a communication interface module, an identification module and a data management module. The system realizes comprehensive control and automatic management of a power supply aging test process by means of integrating a plurality of modules such as the control module, the load module and the data collection module; the control module can send to each module instructions, such as test parameters and environment setting parameters, so as to ensure accurate control of test conditions; moreover, the data collection module monitors the operating parameters of a power supply in real time, so as to ensure the real-time performance and accuracy of test data; and such an integrated and automated design greatly improves the test efficiency and reduces human operation errors, thereby making test results more reliable.

The invention relates to the technical field of electronic measurement, in particular to a multifunctional comprehensive test method and system for a direct-current power supply, and the method comprises the steps: initializing a test; a static performance test step; a dynamic performance test step: driving a direct-current electronic load module to execute a time sequence instruction sequence in the dynamic load test file, and acquiring an output voltage and an output current waveform through a high-speed data acquisition module in the execution process; a ripple noise test step: switching the high-speed data acquisition module to an alternating current coupling mode and acquiring a high-frequency noise waveform in a constant load current section in the time sequence instruction sequence; a correlation analysis step; and a test termination step: controlling the load to return to zero and closing the tested DC power supply. Through automatic configuration and test execution of the system, the complexity of manual operation is reduced, and the test efficiency is improved. And especially in a high-frequency test scene, the test time is remarkably saved through automatic execution.

The invention discloses a modularized multi-level converter fault positioning method based on TimesNet and a harmonic theory. The method comprises the following steps: carrying out bridge arm voltage sampling on a modularized multi-level converter to obtain a bridge arm voltage signal; carrying out harmonic parameter estimation on the bridge arm voltage signal based on a pre-trained Times Net network to obtain a harmonic amplitude, a harmonic phase and a background noise level; carrying out self-adaptive correction on a preset basic amplitude threshold according to the background noise level to obtain a dynamic amplitude threshold; and when the harmonic amplitude is greater than the dynamic amplitude threshold, determining that the modular multilevel converter has a fault, and positioning the fault sub-module according to the harmonic phase. The method improves the fault positioning accuracy and reliability of the modular multilevel converter, and can be applied to the technical field of DC power transmission and transformation.

The invention discloses a fault diagnosis method and system for a voltage regulation module, and belongs to the technical field of fault detection, and the method specifically comprises the steps: periodically collecting key operation parameters of the voltage regulation module, carrying out the preliminary anomaly detection of the collected key operation parameters through a multi-dimensional dynamic threshold model, and carrying out the fault diagnosis of the voltage regulation module; the initial anomaly detection is judged based on the deviation degree of the current operation state relative to a historical normal track, when the initial anomaly is judged, anomaly trend analysis in a time window is carried out, if the anomaly continuously exceeds a preset time threshold value, the current anomaly state is compared with a preset fault feature matrix, and if the anomaly does not continuously exceed the preset time threshold value, the current anomaly state is judged to be abnormal; identifying whether the voltage regulating module has a fault, and generating a corresponding fault label according to the identified fault type; according to the invention, latent, nonlinear or multi-dimensional coupling type faults can be accurately captured, the adaptive capability is high, early warning can be realized, and the operation reliability and maintainability of the voltage regulation module under complex working conditions can be significantly improved.

The invention discloses a machine room fault positioning dynamic environment monitoring system, relates to the technical field of machine room monitoring, and solves the problems of inaccurate power supply fault source positioning and response lag of a traditional system. According to the technical scheme, the method comprises the steps that a current sensor set is arranged at the input end of a cabinet-level power distribution unit in a machine room, current values are monitored at the interval of 15-30 seconds, a power supply path database stores electrical connection topology and coordinates of a mains supply input cabinet, a UPS output cabinet, an array cabinet and the cabinet-level power distribution unit, and a central processing unit compares the current values with a preset threshold value; and when the current of a certain unit continuously exceeds a threshold value for 20-60 seconds, a primary alarm signal is generated, an upstream equipment sequence column head cabinet shunt, a UPS output cabinet circuit breaker and a mains supply input cabinet switch corresponding to the fault unit are extracted, and an alarm signal and sequence information are output. The system is used for machine room power supply fault accurate positioning and rapid processing, and the maintenance efficiency and the system reliability are improved.

The invention discloses an analysis method and system for energy conservation and loss reduction of transformers, and the method comprises the steps: collecting the standby state data of a transformer group, recognizing a high-loss unit, and predicting the parameter drift trend of the high-loss unit; acquiring an electromagnetic impact strength signal in a working state; the parameter drift trend and the electromagnetic impact strength signal are fused, and the collaborative imbalance risk is evaluated; and according to a risk assessment result and a risk cause, dynamic voltage regulation and load balancing optimization are adaptively executed. The cross-working-condition correlation analysis and closed-loop control are carried out on the transformer group serving as the power supply core, so that the standby no-load loss can be effectively reduced, the equipment aging can be delayed, the operation cost can be reduced, the long-term stability and reliability of the power supply system can be ensured, and the method can be widely applied to the industrial fields of pulsed electric fields, precision manufacturing and the like.

Disclosed is a test device to test the operation of an MMC by fabricating only one or several SMs constituting the MMC testing under real-time operating conditions. The test device for a modular multi-level converter (MMC) includes a simulation model of an MMC having at least one arm to which sub modules (SMs) are serially connected, at least one test target SM among the serially connected SMs being replaced with a dependent voltage source; an arm current simulation circuit including an equivalent SM that implements the test target SM as actual physical hardware and an inverter that supplies current to the equivalent SM; and a control unit configured to control the arm current simulation circuit to correspond to an operation of the simulation model and set a voltage corresponding to a charge / discharge voltage of the equivalent SM of the arm current simulation circuit to the dependent voltage source.

The invention relates to the technical field of electrical sensor measurement, and provides a fault test method, system and equipment suitable for a fan converter, and the method comprises the steps: collecting current data in an IGBT module of the fan converter, and taking the current data as an original current signal; screening a plurality of dominant frequencies of the original current signal; decomposing the original current signal to obtain a plurality of component signals, and corresponding to each dominant frequency to obtain a plurality of dominant frequency signals; screening to obtain a reference signal; according to the variation trend difference between each component signal and the reference signal, the variation abnormity of each component signal is obtained; analyzing the energy difference between each component signal and the reference signal in the time-frequency energy matrix to obtain the energy difference index of each component signal; obtaining an anomaly index of each component signal; and performing fault detection on the fan converter based on the anomaly index of each component signal. The invention aims to solve the problem that high-frequency electromagnetic noise affects signal performance of a current sampling loop and covers real fault features.

The invention discloses an online early warning method and detection system for a power supply, and belongs to the technical field of power supply monitoring, and the method comprises the steps: collecting an electric stress parameter of a to-be-monitored power supply in real time based on an early warning sampling frequency, and carrying out the preprocessing; once the real-time electric stress value is greater than or equal to the early warning safety threshold value, determining that the event is a potential abnormal event, constructing an abnormal time window, carrying out abnormal feature analysis, and carrying out recessive parameter inversion in combination with a digital twinborn body to generate an abnormal feature set; and calculating a real-time health index, once a potential abnormal event is identified, correcting the health index, generating an event health index, calculating a health decline rate, predicting a health index in a future time period and carrying out health early warning by combining the abnormal feature set, so that cost increase of reversible abnormal excessive early warning is avoided, and damage caused by missed judgment and accumulation is prevented.

The invention relates to an on-line monitoring method for the on-off state and abnormal behavior of a synchronous rectification MOSFET of a power system. According to the method, a drain-source voltage VDS signal and a drain current Isd signal of a synchronous rectification MOSFET are collected in the working process of a system, and on-line monitoring is conducted on the on-off state and the abnormal state of the MOSFET. The method specifically comprises the following steps: conducting state monitoring: when VDS is smaller than or equal to a conducting judgment threshold value, judging to be conducted, and otherwise, judging to be turned off; monitoring an oscillation state, and judging switch oscillation when the VDS generates multiple complete oscillations in a set time window; early turn-off monitoring is carried out, early turn-off is judged according to whether I sd is higher than a threshold value or not during turn-off in the DCM mode, and early turn-off is judged according to the current descending trend and zero intersection point deviation in the CCM mode; withstand voltage monitoring: when the VDS transient peak value exceeds the rated withstand voltage or margin of the device, determining that the device is unsafe; and mis-opening monitoring: in the DCM mode, when the VDS is close to the opening threshold value during the logic turn-off period, determining that the mis-opening risk exists. The method can improve the reliability of the power supply system.

The invention relates to an alternating-current and direct-current integrated power supply system and a control method, a power conversion unit comprises a parallel high-frequency rectification module, an inversion module and a communication power supply module, and a direct-current side is connected in parallel to a common direct-current bus; the energy storage unit realizes energy interaction between a storage battery pack and a bus through a bidirectional DC / DC converter, and a matched health monitoring circuit acquires internal resistance, temperature and voltage data in real time; the dynamic voltage support unit is formed by connecting a filter network and a transient suppression device in series and is directly connected with a bus to suppress voltage fluctuation caused by load abrupt change; the electromagnetic protection unit integrates an input-stage surge protector, a cabinet body grounding copper bar and an anti-interference device, so that the electromagnetic compatibility of the system is guaranteed; the intelligent control unit is combined with the communication bus through the main controller, and the functions of multi-mode power supply switching, energy storage unit temperature compensation charging and cooperative fault protection are achieved. The system meets the AC / DC load hybrid power supply requirement through the modular design and the cooperative control strategy, and improves the reliability and adaptability of the power supply system.

The invention relates to a power network structure weakness detection method based on multi-diagonal-block matrix decomposition, and belongs to the technical field of super-large-scale integrated circuits. According to the method, an original large-scale sparse matrix is converted into a band edge diagonal block structure and divided into a plurality of sub-matrixes capable of being solved independently by constructing a layering and blocking strategy for eliminating tree drive, and redundancy of full-matrix calculation is avoided. Meanwhile, a local approximate inverse algorithm of column norm truncation is designed, and target elements are calculated on the premise that a preset error threshold value is met. According to the method, the parallel computing architecture of the multi-core processor is fully utilized, the computing complexity is reduced by 2-3 orders of magnitude, the computing efficiency is effectively improved, the memory occupation is remarkably reduced, and a high-precision and high-efficiency solution is provided for detecting the weakness of the power network structure of the super-large-scale integrated circuit.

A thermosensitive wire includes a pair of conductive wires each covered with an insulator that softens at a preset temperature, and the pair of conductive wires are configured to come into contact with each other to be short-circuited in response to the insulator softening. A power supply system includes: a second power source configured to supply power to the pair of conductive wires; a short circuit detector configured to detect that the pair of conductive wires are short-circuited; a disconnection detector configured to detect that the pair of conductive wires are disconnected; and a controller configured to perform, for power supply to a power supply wire, abnormality control that differs between in the case where the short circuit detector detects that the pair of conductive wires are short-circuited and in the case where the disconnection detector detects that the pair of conductive wires are disconnected.

The invention relates to the technical field of power supply testing, in particular to a power supply testing system and method. Comprising the following steps: performing power supply demand simulation based on a load transition mechanism: performing deep analysis on a correlation relationship among components of a load by using demand analysis, and performing power supply demand simulation according to a deep analysis result; and constructing a power supply response state set based on the power supply demand simulation result and in combination with the load response characteristics, and carrying out the adaptability test of the power supply behavior through a power supply adaptability clustering algorithm according to the distribution characteristics of the load states under different configurations. According to the method, the instantaneous peak value, the stable power consumption and the load transition characteristics are extracted, the power supply response model is constructed in combination with the load state, and the fitting degree of the real power supply demand of each component is improved; and through a multi-round clustering center iteration verification and adjustment mechanism, it is ensured that the test covers the power supply threshold of all components, and the comprehensiveness and reliability of the power supply test are improved.

The invention discloses a standby power supply fault prediction system and method based on deep learning, particularly relates to the field of communication machine room operation and maintenance, and is used for solving the problems of insufficient early warning lead and unstable prompt caused by late, repeated and rough alarm labeling. The method comprises the following steps: acquiring a unified time reference generation event window set of multi-source time sequence data of a standby power supply link, rolling forwards to extract an early symptom fragment set and retaining a mapping relation, combining repeated labels by adopting a time migration estimation algorithm to form a weak label set and an advance target, supervising and training a deep learning model solidification configuration by using the advance target, and finally obtaining an early symptom fragment set. In the reasoning stage, a fragment screening process is reused to output an advance and a risk result, a rule updating suggestion is formed by matching a real event through consistency verification, a sequence is calibrated again in an off-line training period, and a time offset estimation algorithm and a fragment screening rule are judged to be realized.

The invention discloses a voltage and current synchronous measurement method for a high-voltage power supply of a fusion device, which is suitable for accurately sampling and monitoring voltage and current signals of each key node of the high-voltage power supply of the fusion device. According to the method, firstly, a resistance-capacitance voltage division technology is adopted, a front-end high-voltage attenuation network is constructed by utilizing high-precision resistors and capacitors, and stable voltage division and linear attenuation of high-voltage signals are achieved; then, electrical isolation is realized through a small-signal voltage and current transformer, and multi-channel synchronous acquisition of voltage and current signals is completed respectively. Then, the ADC module adopts two multi-channel AC sampling special electric energy metering chips to realize ten-channel synchronous acquisition of high voltage and current signals. And finally, the FPGA receives, processes and analyzes the acquired data in real time, and transmits a result to an upper computer monitoring platform. According to the method, the synchronism, the precision and the anti-interference performance of a fusion power supply high-voltage measurement system are remarkably improved, and reliable data support is provided for safe and stable operation of a fusion device.