60 results about "Fault detection and identification" patented technology

The invention discloses a fault detection and identification method for civil aircraft system based on LSTM-AE depth learning framework and relates to the technical field of condition monitoring and fault diagnosis of complex civil aircraft system, and can be used to realize the monitoring and identification of flight faults. The invention comprises the following steps: selecting time series dataof multi-state parameters of an aircraft in flight under a certain stable condition, and according to the characteristics of the monitored object, the time series data of state parameters under suitable conditions are selected for the training of the system reconstruction model, then the fault-free state of civil aircraft system is modeled and reconstructed by making full use of the long-time series-dependent memory ability of LSTM model. The fault monitoring and identification are realized by further analyzing the reconstruction error of its state parameters. The invention solves the problemof insufficient fault monitoring means of civil aircraft system, utilizes the advantage of deep learning in big data analysis to mine massive operation and maintenance data of civil aircraft, and provides important support for fault monitoring of civil aircraft system and route fault isolation.

A system to control a dynamic system, comprising actuators, sensors, a control module and a plurality of observers to monitor the state of the actuators is disclosed. Certain embodiments of the invention comprise a global failure detection and identification module, a baseline controller, and a retrofit controller. Methods to control the dynamic system are disclosed. Some of the methods utilize a second-order mathematical model of the dynamics of the actuators. In some embodiments, this mathematical model is parameterized by a loss-of-effectiveness (LOE) parameter and a lock-in-place (LIP) parameter. The method may further comprise the step of detecting disturbance conditions and structural damage conditions.

The invention discloses a high-voltage line patrol fault detection system based on a visual attention mechanism. The system comprises an image sensor, an ADC analog-to-digital conversion module, an FPGA module, a DDR storage module, a DSP image processing module, an automatic control module and a communication module. The automatic control module is mainly used for receiving an instruction signal of the communication module and controlling operation of the other modules, and when receiving a fault signal from the DSP, notifying the communication module to send the fault information to a control station; the image sensor is used for carrying out video image acquisition, and after ADC analog-to-digital conversion, video images are subjected to video sampling and image preprocessing by the FPGA module; and the DSP image processing module is used for carrying out fault detection and identification on video data in an RAM based on the visual attention mechanism. The invention also discloses a high-voltage line patrol fault detection method based on the visual attention mechanism. Power transmission line characteristics are extracted through an image identification method, and then, fault detection is carried out, and thus the problems of high intensity, high cost and low efficiency and the like of a conventional patrol mode can be solved.

The present invention relates to the field of fault detection and identification of a flexible DC grid, and particularly relates to a fault identification method for a flexible DC grid based on a convolutional neural network. The method comprises: constructing a convolutional neural network model with a branch structure; simulating to obtain fault case data, training the model and adjusting the model parameters; saving the model structure and parameters with high identification accuracy and a small loss function in the training verification; setting a fault detection and identification starting criterion, starting a fault identification program, sampling a data window in 2ms at a sampling signal detection point, and collecting the positive and negative voltages and currents of the line under the actual working condition; and performing normalized processing on the data and identifying an actual fault type by using the model. According to the method provided by the present invention, window information is sampled by fully using the 2ms, the comprehensive utilization of various dimension fault features can be realized by using the branch structure of the model, the accuracy of faultidentification in the flexible DC grid can be improved, the resistance to transition can be improved, and quickness, selectivity, and sensitivity requirements of fault identification can be satisfied.

The invention discloses a flexible spacecraft active fault tolerant control method based on dynamic output feedback control, which belongs to the field of spacecraft attitude control. The method comprises the steps that a dynamic equation of a flexible spacecraft attitude control system is transformed into a general state space equation; a fault model with additive sensor measurement offset is established; a fault detection and identification module composed of an unknown input observer and a filter is established; real-time detection and on-line estimation are carried out on unknown sensor fault; and a fault tolerant controller based on dynamic output feedback is designed based on acquired fault estimation information. According to the invention, when a flexible spacecraft can normally reach an expected posture in the case of an additive sensor measurement offset fault; the influence of disturbance generated by modeling uncertainty and flexible accessory is considered in a design process; and a fault diagnosis and identification module and the fault tolerant controller can be independently designed, and can be easily realized by a project.

The invention belongs to the field of industrial monitoring, particularly relates to the application of an empirical mode decomposition method and a probabilistic neural network in the aspect of fault diagnosis of flow passages of heat-engine plant steam turbines, and provides a fault diagnosis method for a flow passage of a steam turbine on the basis of empirical mode decomposition and the probabilistic neural network. The method includes the steps that firstly, fault symptom parameter data of the flow passage are collected when the heat-engine plant steam turbine operates normally and has different faults, fault features are extracted from the corresponding symptom parameter data by the utilization of the advantages of empirical mode decomposition in processing non-stationary and nonlinear data, fault detection and recognition are carried out by the utilization of the powerful nonlinear mode classification performance of the probabilistic neural network, and then faults of the flow passage of the steam turbine are effectively diagnosed. By means of the fault diagnosis method, the faults of the flow passage of the heat-engine plant steam turbine can be diagnosed rapidly and accurately.

The present invention discloses an omnidirectional fault automatic identification method of an isolation switch. The visible light image information and infrared detection image information are collected through each direction of the isolation switch, then the corresponding image information is subjected to preprocessing, the algorithm fusion is carried out, the feature point extraction and description are carried out and a description vector is generated, then the description vector is subjected to feature matching, a wrong matching feature point is eliminated, an identification area and the coordinate of an edge point are determined, the obtained edge point is subjected to affine transformation are is marked in the infrared fusion image of the isolation switch, thus the automatic identification of the isolation switch infrared fusion image is completed, the template image of the isolation switch is subjected to mode identification, and thus whether the isolation switch has a fault or not is judged. According to the method, the defect in current isolation switch fault detection technology is solved, and the method has the advantages of non-contact precise identification, isolation switch component comprehensive identification and fully automatic fault detection and identification.

A method of open-switch fault detection in a two-level voltage source power converter comprises the steps of:i. acquiring the waveforms of three-phase alternating current of the two-level voltage source power converter;ii. calculating at least a first time derivative of each acquired waveform; andiii. using the at least first time derivatives to determine whether an open-switch fault is present in the two-level voltage source power converter.

A device for processing faults, that detects and identifies faults, including, for at least part of the identified faults: a mechanism storing a log of occurrences of faults having a same identification, in a course of a plurality of flights; a mechanism determining an impact of each identified fault, on availability of aircraft in a predetermined duration, a mechanism associating a hierarchical attribute with each fault detected in a course of a flight, the hierarchical attribute being dependent on the log and on the impact; and a mechanism displaying identified faults associated with part of the hierarchical attributes. The device enables, in particular, maintenance operations or operations preparatory to maintenance to be better scheduled by virtue of hierarchization of faults displayed.

The invention discloses a method and system for monitoring the local integrity of train satellite positioning. The method comprises the steps of collecting track line geospatial information and trainrunning plan information of a target train; dividing the running process of the target train to obtain spatial-temporal local area grids; collecting satellite positioning observation performance information in real time in the running process of the target train, and calculating a satellite positioning local area observation model of each spatial-temporal local area grid; integrating parameters ofthe models to construct a satellite positioning observation model parameter set; extracting model parameters of the located spatial-temporal local area grid by using the satellite positioning observation data received by the train in the running process, and calculating on-line and off-line positioning error level protection grades; judging the availability state of integrity monitoring, carryingout fault detection and identification on the satellite positioning observation set under an available condition, performing reconstruction on the train satellite positioning observation set in realtime, and implementing positioning calculation. The method can realize effective adaptation to the uncertainty of satellite signal observation conditions in the actual running of the train.

The present disclosure provides methods, systems, and computer-readable media for the fault detection and identification in an aircraft that may occur in real time during a flight, or any time the aircraft is operating. For example, a controller may receive and calculate various parameter values at various times during an aircraft flight, and compare those values to baseline values in order to determine if a fault has occurred. Additionally, the controller may identify a fault that has occurred by comparing a calculated fault signature value with a fault signature database comprising fault signatures and their associated faults.

The invention provides a gradual fault backtracking fault-tolerant method for an inertial and satellite integrated navigation system, and the method comprises the following steps of performing initialalignment of the inertial navigation system; collecting and sliding the stored sensor data; performing gradual fault detection and identification on the satellite navigation position information by using a state chi-square and residual chi-square hybrid detection method, cutting into an inertia / satellite integrated navigation mode when the satellite navigation has no fault, and fusing the satellite navigation output information; performing the inertial navigation backtracking algorithm and the Kalman backtracking algorithm when the satellite navigation has a gradual fault to obtain the attitude matrix, speed and position of the integrated satellite gradual fault information at the k-1 time; backtracking to the satellite navigation fault point, performing pure inertial navigation trackingand recalculation, recursively stepping to the kth sampling time, and outputting the attitude, velocity and position information after isolating the satellite historical fault information. The methodprovided by the invention can effectively judge satellite navigation gradual faults, and perform the fault tolerance processing on the fault by the inertial navigation and the Kalman backtracking algorithm to avoid historical fault information pollution.

The invention provides an air conditioner fault detection and identification method and device based on a prediction and classification mode. The method comprises the steps that S101, an operation prediction model and a fault classification model are built respectively through normal state operation data and fault operation data according to the normal state operation data and the fault operationdata of an air conditioner; S102, whether the air conditioner breaks down or not is judged through the operation prediction model, if the air conditioner breaks down, the step S103 is executed, and ifthe air conditioner does not break down, the step S101 is executed; and S103, data segment where the air conditioner breaks down is located through the operation prediction model, and a fault type corresponding to the data segment is obtained through the fault classification model. According to the method, the operation prediction model and the fault classification model are built respectively through the normal state operation data and the fault operation data of the air conditioner, the state prediction model and the classification model are creatively fused, high detection and identification accuracy are achieved, meanwhile, the continuously accumulated operation data can be fully utilized, and through continuous machine learning, fault diagnosis accuracy is improved.

The invention, which belongs to the technical field of the industrial process monitoring and diagnosis, discloses an improved-principle-component-tracking-based industrial process monitoring method and application. Decomposition of a low-rank-matrix-expression-based principle component tracking method is carried out on industrial collection data to obtain a low-rank coefficient matrix including all variable relations of the process; and on the basis of the low-rank coefficient matrix as well as correlated coefficient weights of variables in a training matrix, an L2 statistic is constructed to carry out fault detection and identification. According to the invention, according to the principles of the low-rank matrix expression and the principle component tracking method, the low-rank matrix expression algorithm is fused into the principle component tracking to construct a low-rank-matrix-expression-based principle component tracking algorithm model; and the model is used for carrying out on-line monitoring. Therefore, the correlated relations between variables in the training matrix and effective information included by the variables in the training matrix are utilized fully. With the method, the accuracy of industrial fault detection and identification with abnormal values is high.

The invention discloses a multi-dimensional fault detection and recognition method, wherein the method comprises the following steps: S1, constructing a fault detection statistic, and carrying out thedetection of a fault based on chi-square detection; S2, when a fault is detected, constructing a fault set by fusing identification results of a data detection method and a correlation coefficient method; and S3, identifying and eliminating faults according to the fault set. Two typical RAIM algorithms of a data detection method and a correlation analysis method are fused to identify and processa fault. Under the condition of multi-dimensional faults, compared with independently adopted two RAIM algorithms of the data detection method and the correlation analysis method, the method has higher fault recognition accuracy and recognition efficiency, and the integrity and reliability of a satellite navigation system can be further improved.

The invention discloses a rapid self-adaptive fault detection and identification method for a micro-grid line section. The method comprises the following steps: collecting node three-phase current atnodes at two ends of the line section; calculating an instantaneous phase difference value and a phase threshold value of a three-phase current high-frequency component of the line section, and a nodezero-sequence current effective value and a zero-sequence current threshold value at nodes at two ends of the line section; judging a fault based on comparison of an instantaneous phase difference value and a phase threshold value of a three-phase current high-frequency component of a line section, and on this basis, further obtaining a specific fault type based on comparison of node zero-sequence current effective values and zero-sequence current threshold values at nodes at the two ends of the line section. According to the method, various faults occurring in the microgrid line section canbe detected, and the method can be universally applied to microgrids in grid-connected operation and island operation states; according to the method, the fault judgment threshold value can be updatedin real time, and good adaptability is achieved; the method is high in detection speed, and fault detection can be completed within tens of milliseconds and the fault type can be correctly identified.

The invention relates to an electric transmission line helicopter routing inspection image classification method. Traditional manual inspection is multiple in potential safety hazards and low in efficiency. According to the method, firstly, LoG transformation is conducted on an input image I to extract bar-type shapes in the image and obtain a BarMask; secondly, an edge is extracted to the image under the BarMask by using an edge detecting method, straight line sections of the edge are detected by using a Hough transformation class method, the fact that whether a pole exists in the image I is judged according to direction, number and length of the straight line sections; then the fact that whether the image under the BarMask is a guide line image is judged; and finally the fact that whether element components or other objects are on a guide line is judged. The electric transmission line helicopter routing inspection image classification method can classify and filter a large number of routing inspection images, and simplifies subsequent fault detect and identification works achieved by manual work or software.

The invention discloses an electronic checkpoint fault detection and identification method, and the method comprises the steps: detecting a real-time voltage value and a real-time current value of anelectronic checkpoint, and obtaining a first detection result; detecting the working state of an electronic police in the electronic checkpoint, and obtaining a second detection result; detecting theworking state of the traffic signal machine in the electronic checkpoint, and obtaining a third detection result; according to the first detection result, the second detection result and the third detection result, comprehensively judging whether the electronic checkpoint has a fault or not and a specific fault type. By detecting the voltage value and the current value of the electronic checkpoint, the working state of the electronic police and the working state of the traffic signal lamp, the working states of the components can be monitored in real time, and if one component breaks down, thefault type can be accurately determined at the first time; and the fault types can be uploaded to the background, so workers can timely and accurately know the fault types and implement solutions, the workers do not need to manually find out specific faults on site, the working efficiency is improved, and the manpower and the cost are saved.

The invention discloses an improved GIS equipment abnormal sound vibration identification method based on a particle swarm algorithm, which belongs to the technical field of GIS equipment fault detection and identification, and comprises the following steps: optimizing parameters of an LSSVM by adopting an ILPSO algorithm to obtain an ILPSO-LSSVM algorithm, and constructing a GIS equipment isolation switch contact state identification model by combining the ILPSO-LSSVM algorithm and a training sample set combined feature matrix; and carrying out validity verification on the GIS equipment isolation switch contact state identification model by adopting the constructed test sample set combination characteristic matrix, and the like. The improved least square support vector machine model basedon the particle swarm algorithm is constructed through optimization design of a traditional SVM model, and the model is high in recognition accuracy and small in average standard deviation.

The invention discloses a photovoltaic panel fault detection and identification method. The method comprises the following steps of preprocessing a photovoltaic panel infrared video, and extracting frame by frame to obtain the photovoltaic panel image data after perspective transformation; determining a photovoltaic array area; extracting to obtain a photovoltaic panel array image; for the extracted photovoltaic panel array image, removing invalid lines line by line by adopting a threshold method, and extracting to obtain the small photovoltaic panel area images; calculating a pixel mean valueand a variance of each small photovoltaic panel area image, and carrying out fault detection and fault category judgment by adopting a threshold method; numbering the faults by calculating a fault photovoltaic panel characteristic function value, obtaining the fault position information through calculation in combination with the unmanned aerial vehicle route data and a photovoltaic panel numbersequence, wherein the fault position information comprises the number and the position of the photovoltaic panel where a fault is located. According to the invention, the photovoltaic panel fault detection can be carried out automatically, the fault images, fault types and fault positions are outputted, and the workers of a photovoltaic power station can determine the fault information timely andconveniently for photovoltaic panel maintenance.

The invention provides an autonomous underwater robot propeller weak fault feature extraction method, which belongs to the technical field of underwater robot fault diagnosis and comprises two parts: fault feature enhancement and feature fusion. Firstly, Gaussian properties of all modes of a multi-source state signal and a control signal are judged through negentropy, so that parameters are optimized, noise reduction is completed, and fault features are extracted and enhanced based on a modified Bayesian algorithm. Afterwards, the characteristic signals are divided into a plurality of time intervals, faults occurring in all the intervals serve as focal elements, first-time characteristic fusion is conducted on all the signals except for the longitudinal speed, second-time fusion is conducted on the result of the first-time fusion and the characteristic signals of the longitudinal speed, and the fault characteristics are further enhanced; and meanwhile, monotonicity is presented between the fault feature and the fault degree. The method can provide a basis for subsequent fault detection and identification, and is especially suitable for state monitoring of an autonomous underwater robot propeller.

The invention relates to a ground-based augmentation system integrity fault identification method and system, belongs to the technical field of satellite navigation, and solves problems that a fault identification method in the prior art is low in precision and cannot better identify potential integrity faults. The method comprises the following steps: acquiring historical measured data provided by a ground reference receiver and an airborne receiver; calculating a differential positioning error and a protection level of the airborne receiver based on the historical measured data; extracting afault data set from the historical measured data based on the differential positioning error and the protection level; and performing fault feature extraction on the fault data set and determining afault type. The accuracy of fault detection and identification is improved, and the requirement for higher-level integrity fault identification is met.

The invention discloses an underground cable early fault detection and identification method based on a DAE-CNN. The method comprises the steps of carrying out the analog simulation of the early fault of an underground cable, so as to obtain the simulation data of a current, performing feature extraction on the simulation data by adopting a noise reduction automatic encoder to obtain dimension-reduced current data, training a convolutional neural network by using the current data after dimension reduction, and generating a discriminator which can be used for detecting and identifying an early fault of the underground cable, and performing feature extraction on the current data of the underground cable to be detected by adopting a noise reduction automatic encoder to obtain dimension-reduced current data, taking the dimension-reduced current data as the input of a discriminator, and outputting an early fault identification result of the underground cable by utilizing the discriminator. Accurate diagnosis of early faults of the power distribution network can be realized.

The invention provides a sampling frequency recommendation method and device, equipment and a storage medium, and relates to the technical field of communication. The sampling frequency recommendationmethod comprises the steps of obtaining a network key performance index of a to-be-analyzed data stream; sampling the network key performance index according to a plurality of different sampling frequencies to obtain an experience quality sequence corresponding to each sampling frequency, the plurality of different sampling frequencies comprising a standard sampling frequency and at least two to-be-tested sampling frequencies, and the standard sampling frequency being greater than each to-be-tested sampling frequency; And determining a matching degree between the experience quality sequence corresponding to each to-be-tested sampling frequency and the standard experience quality sequence, determining a recommended sampling frequency according to the matching degree between the experiencequality sequence corresponding to each to-be-tested sampling frequency and the standard experience quality sequence, and enabling the matching degree corresponding to the recommended sampling frequency to meet an expected condition. By using the technical scheme of the invention, the network experience quality, the network fault detection and identification capability and the network performance can be balanced.

The invention discloses a real-time fault detection method for a large-scale grid-connected solar photovoltaic power station. The method comprises the steps of simulating a theoretical photovoltaic power station in MATLAB / Simulink, achieving the theoretical photovoltaic power station, enabling the theoretical photovoltaic power station to generate a direct current similar to an actual photovoltaic power station in real time, obtaining the irradiance and photovoltaic module temperature data in real time from the theoretical photovoltaic power station in the Simulink, simulating the generated photovoltaic output, carrying out abnormal value detection by using a developed statistical tool Z-score, comparing the direct current power of a single photovoltaic string of an actual photovoltaic power station with the average direct current power photovoltaic string of the theoretical photovoltaic power station, and carrying out fault detection and identification through a Z-score value and a comparison result based on a Z-score threshold value.

The invention discloses a satellite fault identification method, a satellite fault identification device and a software receiver, and relates to the technical field of satellite navigation and positioning. The method comprises the following steps: acquiring receiver position data and clock correction, constructing an observation matrix according to the receiver position data and the clock correction, and decomposing the observation matrix by adopting an odd-even space algorithm to generate an odd-even space matrix; performing matrix operation on the odd-even space matrix to obtain the slope ofeach satellite participating in positioning, and extracting the maximum slope value; calculating a similar radial error protection value of the positioning satellite based on the maximum slope valueand the false alarm detection value, and when the similar radial error protection value is not less than a preset value, calculating a check statistic according to satellite deviation data corresponding to the odd-even space matrix; detecting whether a satellite participating in positioning has a fault or not according to the verification statistics, the false alarm detection value and a preset threshold factor. The method has the characteristics of low time complexity and low space complexity, and can adapt to satellite fault detection and identification of normal flight and high dynamic flight.

The invention discloses a photovoltaic power generating assembly fault detection and identification method based on digital twinning. The method comprises the following steps: detecting and outputting a characteristic quantity y (t) in a physical entity of a to-be-detected photovoltaic power generating assembly, wherein the photovoltaic power generating assembly comprises a solar cell assembly and a DC-DC converter; constructing a digital twin body having the same physical entity structure as the to-be-detected photovoltaic power generating assembly, and calculating and outputting the measurement characteristic quantity z (t) of the photovoltaic power generating assembly in the digital twin body; calculating and outputting a residual vector gamma (t) according to the characteristic quantity y (t) and the measured characteristic quantity z (t); outputting a detection result according to the residual vector gamma (t); when a fault exists in the detection result, calculating and outputting an L2 inner product according to the residual vector gamma (t) and a fault characteristic value fi, wherein the fault characteristic value fi is calculated by the residual vector gamma (t) and a 2-norm gamma (t) 2 of the residual vector gamma (t); and outputting a fault type according to the L2 inner product. The method can be used for detecting whether the photovoltaic power generating assembly has a fault or not and identifying the type of the generated fault. The reliability and the practicability are high.

The present invention relates to the field of fault detection and identification of a flexible DC grid, and particularly relates to a fault identification method for a flexible DC grid based on a convolutional neural network. The method comprises: constructing a convolutional neural network model with a branch structure; simulating to obtain fault case data, training the model and adjusting the model parameters; saving the model structure and parameters with high identification accuracy and a small loss function in the training verification; setting a fault detection and identification starting criterion, starting a fault identification program, sampling a data window in 2ms at a sampling signal detection point, and collecting the positive and negative voltages and currents of the line under the actual working condition; and performing normalized processing on the data and identifying an actual fault type by using the model. According to the method provided by the present invention, window information is sampled by fully using the 2ms, the comprehensive utilization of various dimension fault features can be realized by using the branch structure of the model, the accuracy of faultidentification in the flexible DC grid can be improved, the resistance to transition can be improved, and quickness, selectivity, and sensitivity requirements of fault identification can be satisfied.