814 results about "Mechanical failure" patented technology

A system and method for ΔI-powered engine diagnostics combine OBD-II data, high-frequency sound, and vibration analysis to detect engine wear and mechanical faults. A detachable puck sensor in the engine bay captures sound and vibration signals, while an OBD-II module collects engine performance metrics. A smartphone app collects and uploads data to a backend AI engine, where Dynamic Time Warping (DTW) aligns event-driven or asynchronous data and time series data from multiple sources, ensuring accurate feature fusion. Machine learning models then detect engine wear, belt degradation, knocking, and bearing faults, generating a diagnostic report with severity assessments and predictive maintenance recommendations. By integrating multi-modal sensor data, this system enhances early-stage fault detection beyond traditional OBD-II diagnostics, offering greater accuracy in assessing physical wear conditions and optimizing vehicle maintenance.

The invention relates to a mechanical equipment fault data identification method based on artificial intelligence, and belongs to the technical field of data processing and artificial intelligence. The problems of insensitive signal processing, insufficient feature extraction, significant noise interference, low model training efficiency and the like in fault diagnosis in the prior art are solved. According to the method, a self-adaptive normalization method based on energy density is provided, the influence of non-stationary signals is effectively inhibited, and key features are reserved; through mixed energy entropy feature extraction and inter-band energy jump penalty terms, the sensitivity to a complex fault mode is significantly enhanced; constructing a fault feature enhancement strategy based on a Gaussian potential well, and reinforcing the response of a fault feature accumulation area; a weighted loss function and a gradient directional correction mechanism are adopted, so that the robustness and accuracy of the model are improved; and in combination with an entropy weighted learning rate and a covariance scaling strategy, adaptive training is realized, and the convergence speed and adaptability are improved. According to the method, the precision and efficiency of mechanical fault diagnosis are improved, and support is provided for industrial intelligent development.

The invention relates to a simulation prediction method for the plugging life of a contact element of an electric connector, and the method comprises the steps: building a three-dimensional simulation model of a plug end and a socket end of the electric connector, and defining the material attributes and contact relationships of all parts in the three-dimensional simulation model; performing finite element grid discretization on the three-dimensional simulation model; applying a fixed constraint to the socket end, applying a displacement load for simulating the plugging process to the plug end, and performing statics solution to obtain plugging force in the plugging process of the plug end and the socket end and contact pressure data of the plug end contact piece and the socket end contact piece; and mapping a statics solution result into a fatigue analysis module, and simulating, calculating and outputting a life distribution result of the plug end contact element and the socket end contact element under repeated plugging and unplugging loads. The method has the effect of analyzing the stress-strain state of the internal contact element of the electric connector and the fatigue damage of the internal contact element under repeated plugging and unplugging so as to help monitoring personnel to discover the potential mechanical failure problem in time.

The invention discloses a wind turbine generator monitoring and early warning system based on the Internet of Things sensing technology, and relates to the technical field of the Internet of Things, and the system comprises a data collection module which collects the operation state information of key parts of a wind turbine generator in real time, and transmits the operation state data to an edge calculation unit; the prediction module is used for identifying an abnormal mode, analyzing time domain and frequency domain characteristics of the vibration signal in combination with fast Fourier transform, detecting a mechanical fault, and obtaining a residual life prediction value according to a residual life prediction model; the preliminary early warning module is used for determining an equipment health state score based on fuzzy reasoning, and performing preliminary early warning locally according to the equipment health state score; and the grading early warning module is used for carrying out grading alarm on the wind turbine generator equipment according to the mechanical fault condition in combination with the equipment health state score, and providing maintenance decision suggestions. By integrating the Internet of Things technology, edge computing and artificial intelligence, intelligent monitoring and early warning of the wind turbine generator are achieved, and the reliability and safety of equipment operation are improved.

The invention relates to the technical field of fault detection, in particular to a rotating machine fault identification method based on multi-source information fusion, which comprises the following steps: collecting noise and vibration data of a target rotating machine; identifying and separating interference noise data generated by other sound production parts outside the target rotating machine; correcting the noise data of the target rotating machine by using the dry noise data; performing fusion processing on the corrected noise data and the vibration data, generating a comprehensive fault feature, matching the comprehensive fault feature with a known fault mode, and identifying a specific fault type; and generating a fault diagnosis report based on the fault type. Therefore, the accuracy of fault detection can be improved, the maintenance cost can be effectively reduced, and the service life of equipment is prolonged.

The invention relates to a large transformer vibration and insulation on-line monitoring device and method, and the method comprises the following steps: processing an electric signal outputted by an optical fiber sensor, and obtaining a vibration signal and a partial discharge signal of a transformer; and carrying out detection and evaluation according to the vibration signal and the partial discharge signal of the transformer. The sensors are arranged in the iron core and the winding of the transformer respectively, the same sensor can monitor two parameters of vibration and partial discharge at the same time, partial discharge defects caused by various mechanical faults of the transformer can be accurately monitored and recognized by monitoring the two parameters, the partial discharge defects are provided for operation and maintenance personnel in time for precise maintenance, and the maintenance efficiency is improved. The maintenance time is greatly shortened, and the sudden power failure fault of the transformer is effectively prevented.

The invention belongs to the technical field of mechanical fault diagnosis, and relates to a bolt fault diagnosis model construction method, a bolt fault diagnosis method, a bolt fault diagnosis system and a bolt fault diagnosis device. According to the method, a vibration signal of a bolt operation state is obtained, and a data set is constructed according to the vibration signal of the operation state; preprocessing the vibration signal in the operation state to obtain a preprocessed vibration signal; extracting fault feature data of the preprocessed vibration signal; performing hierarchical processing on each data segment of the fault feature data to obtain a plurality of low-frequency and high-frequency components, constructing a graph model for each low-frequency and high-frequency component, and calculating a graph similarity score of the graph model and the data set, combining the graph similarity scores of all the low-frequency and high-frequency components into an integral similarity index through an adaptive input weighted fusion technology; a convolutional neural network algorithm is combined with a similarity index to carry out model training on an existing data set, a bolt fault diagnosis model with relatively high precision is obtained, and accurate detection and diagnosis of a bolt fault are realized.

The invention relates to the technical field of mechanical fault diagnosis, in particular to a rotating machine fault diagnosis method and system based on a time sequence large model, and the method comprises the following steps: extracting a time sequence value of a vibration signal, measuring the frequency and amplitude fluctuation of a time period, and recognizing an extreme point based on the operation data of a rotating machine; and calculating the fluctuation amplitude difference between the extreme point and the adjacent segment, measuring the time interval between the amplitude peak values, and generating the frequency change segment amplitude. According to the method, the time sequence value of the vibration signal is extracted, frequency and amplitude fluctuation extreme point recognition and adjacent segment amplitude difference calculation are combined, the fine-grained analysis capability of the dynamic characteristics is improved, key frequency point time distribution and amplification value change are detected, and amplitude fluctuation time interval and frequency amplification correlation analysis is combined; deep analysis of frequency and amplitude interaction characteristics, dynamic change characteristic positioning of an interaction ratio and rate quantification are realized, and diagnosis accuracy and evaluation depth are comprehensively improved.

The invention discloses a bridge laying high-voltage cable mechanical fatigue evaluation method and system, and relates to the field of cable fatigue evaluation. Existing methods for evaluating cable fatigue caused by factors such as bridge expansion and vibration are lacked. The method comprises the following steps: arranging a test cable on a mobile test platform, and collecting sensor data to obtain bending, stretching and compression stress of the cable; data are repeatedly acquired by moving the movable box body, and the stress condition of the cable is simulated; carrying out a slice test on the cable, and measuring the elongation at break of a cable sheath and insulation; in combination with sensor data and cable material fatigue characteristics, shape parameters and scale parameters of Weibull joint distribution are fitted, and the fatigue life, the fault probability and the mechanical failure risk of the cable are predicted based on a fitting result, so that whether the fatigue failure of the cable occurs or not is judged. According to the technical scheme, through combination of the test platform and the Weibull distribution model, the mechanical fatigue of the bridge laying high-voltage cable is accurately evaluated, and the accuracy of cable life prediction is improved.

The invention relates to a GIS disconnecting switch mechanical fault identification method and device, and the method comprises the steps: obtaining bus structure parameters of a source side and a load side of a GIS disconnecting switch, and calculating bus electrical parameters according to the bus structure parameters; constructing a travel curve discrete time sequence according to the plurality of motion states of the moving contact of the GIS disconnecting switch; constructing a whole-process VFTO discrete time sequence based on the travel curve discrete time sequence, and tagging time sequence data in the whole-process VFTO discrete time sequence; extracting a plurality of transient characteristics from the whole-process VFTO discrete time sequence, and constructing a labeled sample set with a supervision model according to each transient characteristic and a corresponding label; and through the trained supervised model and the real-time whole-process VFTO discrete time sequence, identifying the mechanical fault of the target GIS isolation switch. According to the method, the supervised model is trained through the discrete time sequence based on the mechanical state of the GIS disconnecting switch, so that the accuracy of fault detection is improved.

The invention discloses a TCN-SVM rolling bearing fault diagnosis method fusing an SE attention mechanism, and belongs to the technical field of mechanical fault intelligent diagnosis. Aiming at the problems of feature redundancy, noise sensitivity, insufficient Softmax classifier generalization and the like existing in a traditional time sequence convolutional network (TCN), the invention provides a solution for collaborative optimization of a deep network and a support vector machine. The method comprises the following steps: acquiring a vibration signal of the rolling bearing; constructing a multi-fault sample set, and processing an original signal; constructing an SE-TCN feature extraction network, capturing multi-scale time sequence features by adopting expansion causal convolution, and embedding an SE module into a residual module to realize channel adaptive weighting; and a support vector machine (SVM) classifier decision function is constructed, and fault classification is completed through the RBF kernel SVM. Experiments show that the method has high fault recognition accuracy and robustness, and the problem of confusion of composite fault features of the rolling bearing is effectively solved.

The invention provides a vibration-temperature correlation fault early warning method based on multi-sensor fusion, and relates to the technical field of rotating machine fault diagnosis, and the method achieves the efficient diagnosis of the rotating machine fault through the three-field coupling analysis of vibration, temperature and torque. Multi-dimensional health indexes are introduced, sensor data of different dimensions are normalized into a monitorable numerical value, and the health state of equipment is reflected in real time through a dynamically optimized weight coefficient. A core depth time sequence neural network model is combined with an attention mechanism and a Transform coding layer, the problem of long-term dependence is effectively solved, and the fault recognition precision is improved. According to the method, future innovation directions such as cross-device collaborative diagnosis, quantum computing acceleration and digital twinborn visualization are also exhibited, and the method has important industrial application value.

The invention relates to the technical field of fault diagnosis, in particular to a rotating machine fault diagnosis method, device and equipment under variable working conditions, and the method comprises the steps: collecting the operation parameters of a target rotating machine under a plurality of preset working conditions, and carrying out the feature normalization processing of the operation parameters, thereby obtaining a sample data set; the sample data set considers a plurality of preset working conditions and minimizes signal characteristic changes caused by working condition changes; a deep multi-scale conditional adversarial network model is constructed, a test data set is input into the deep multi-scale conditional adversarial network model trained by a training data set to output a fault diagnosis result, and domain self-adaption of an output space is achieved on different feature levels by using time sequence features. The accuracy of mechanical fault diagnosis in an actual production scene is improved; meanwhile, as the constructed multi-scale conditional adversarial network model belongs to an unsupervised transfer learning category, the dependency on training label samples is relatively weak, and the time and resources invested into label marking are greatly reduced.

The invention discloses a wind power gear box online fault diagnosis method based on multi-source data fusion, and particularly relates to the field of mechanical fault detection, and the method comprises the steps: S1, collecting high-frequency dynamic, medium-frequency working condition and low-frequency thermal state data, and outputting a standardized data set through time alignment, quality verification and physical constraint verification; s2, performing multi-scale decomposition to retain a fault sensitive frequency band, inverting physical parameters such as gear contact stress and the like, constructing a physical cause and effect graph, and determining fault sensitive characteristics and threshold values; s3, constructing a multi-modal feature tensor, and obtaining a low-dimensional health representation vector through CP decomposition fusion, graph neural network reasoning and variational auto-encoder dimension reduction; s4, calculating a weight by using an entropy weight method, calculating a dynamic health degree in combination with a health benchmark, predicting a trend by using LSTM, and establishing a five-level health system; s5, judging a fault mode through double-layer identification, analyzing a root cause and formulating a hierarchical operation and maintenance suggestion; the method is based on multi-source fusion and data mechanism dual drive, and precise diagnosis and operation and maintenance guidance are achieved.

The invention discloses a mechanical fault diagnosis method and system based on deep fusion of multi-modal monitoring data. The method comprises the following steps: randomly selecting two kinds of modal target monitoring data from multi-modal target monitoring data; a corresponding target mechanical fault diagnosis model is determined according to the two kinds of modal target monitoring data, the target mechanical fault diagnosis model is obtained by training a sample monitoring data set with the same modal as the two kinds of modal target monitoring data, and the sample monitoring data set is acquired multi-modal monitoring data in the operation process of multiple kinds of mechanical equipment; the target mechanical fault diagnosis model comprises a symmetric fusion module and a fault classifier, the symmetric fusion module is used for carrying out bidirectional symmetric interaction on two kinds of modal target monitoring data, and the fault classifier is used for classifying mechanical faults; and inputting the two kinds of modal target monitoring data into the target mechanical fault diagnosis model to obtain a mechanical fault diagnosis result. According to the invention, the accuracy of mechanical fault diagnosis can be improved.

The invention relates to the technical field of power equipment fault diagnosis, and particularly discloses a pole-mounted circuit breaker mechanical fault diagnosis method and system, and the method comprises the steps: firstly, synchronously collecting a dynamic force-displacement signal, a high-frequency acoustic emission signal and a broadband vibration signal in the switching-on and switching-off process of a circuit breaker operating mechanism; then, the acoustic emission and vibration signals are decoupled into an impact source component and a friction source component which are statistically independent through a blind source separation algorithm; performing envelope spectrum analysis on the impact source component to extract impact characteristics, and performing energy calculation on the friction source component to obtain friction noise energy characteristics; finally, on the basis of dynamic force-displacement curve fitting analysis, in combination with impact characteristics and friction noise energy characteristics, distinguishing and delimiting of mechanical wear and lubrication degradation faults are achieved. According to the invention, the problem that the existing diagnosis technology cannot effectively identify and distinguish the coupling fault of mechanical wear and lubrication deterioration is solved, and early warning and accurate diagnosis of the mechanical fault of the pole-mounted circuit breaker are realized.

The invention relates to the technical field of new energy vehicles, and discloses a new energy vehicle air conditioner compressor fault diagnosis system based on deep learning. The system comprises a data acquisition module, a model construction module, a model activation module and a result output module. The data acquisition module collects real-time operation data of the compressor to form an initial fault feature set containing a vibration spectrum feature sequence and a current fluctuation trend graph. The model construction module constructs a first diagnosis model and a second diagnosis model, the first diagnosis model reflects correlation mapping of vibration characteristics and mechanical faults, and the second diagnosis model comprises nonlinear classification rules of current characteristics and electrical faults. And the model activation module starts a corresponding target model and generates a diagnosis result according to the real-time vibration deviation and the current fluctuation abnormal data. And the result output module transmits the result to a vehicle control system to trigger a fault early warning signal or a maintenance suggestion instruction. The system can comprehensively cover mechanical and electrical fault diagnosis through dual-model collaborative operation, and is adaptive to a dynamic operation scene of the compressor.

Autonomous vehicles (AVs) may have processors to perform compute operations. These processors may be susceptible to overheating, due to one or more mechanical failures that may occur in the assembly that encloses the processors. To proactively address potential overheating of these processors, temperature of the processors can be monitored over time, e.g., at launch and during AV operation. A temperature profile line can be fitted to envelop the temperature data samples. The temperature profile line can indicate an extent of overheating, the rate of regression of the assembly's ability to cool the processors and predict when the processor may reach a critical temperature in the future. Based on one or more parameters that define the temperature profile line, it is possible to determine a regression profile of the processor. The regression profile can dictate whether to take an action to compensate for the regression.

The invention discloses an intelligent mechanical fault rapid diagnosis method. The method comprises the steps of multi-source data acquisition, data optimization processing, mechanical abnormity rapid detection, mechanical fault rapid identification, model performance optimization and mechanical fault intelligent diagnosis. The invention relates to the technical field of mechanical fault data processing, in particular to an intelligent mechanical fault rapid diagnosis method, which innovatively proposes a mechanical fault rapid diagnosis technology for abnormal trigger key sequence extraction, and improves the sensitivity and discrimination precision of mechanical fault features. A random convolution disturbance mechanism is introduced, and a multi-equipment adaptability enhanced loss function is designed, so that rapid, accurate and intelligent diagnosis of mechanical faults under complex working conditions and multi-equipment conditions is realized; an optimization algorithm is improved by adopting a disturbance guide extension strategy, a jump drive fine tuning mechanism and a multi-channel fine tuning candidate strategy, a model performance optimal parameter combination is obtained, and the stability and recognition precision of a mechanical fault recognition model are enhanced.

The invention belongs to the technical field of wind turbine generators, and particularly relates to a wind turbine generator fault early warning and abnormal parameter inspection method comprising the following steps: collecting historical operation data of a wind turbine generator, analyzing the data, and establishing a data model; the historical data comprises operation data of a generator, a gear box, a main bearing, a variable pitch system and an electrical system; vibration spectrums of a gearbox and a generator rotating part are monitored through an acceleration sensor to identify whether bearing pitting corrosion, gear tooth breakage mechanical faults and frequency are abnormal or not; an on-line monitoring device is installed in an oil way of the wind turbine generator to monitor the oil state in real time. The temperature of the gearbox and the bearing is detected through the thermal infrared imager, and whether the temperature is abnormal or not is monitored; abnormal sound signals of blade cracks or bearing early damage are captured through an ultrasonic sensor, and the blade is judged.

The invention discloses a motor fault feature extraction and fault diagnosis method. The method comprises the following steps: step 1, collecting an MCAVFCNN training data set, wherein motor fault types comprise two major classes, namely mechanical faults and electrical faults, namely 9 subclasses; 2, an MCAVFCNN model is built, a dynamic random erasure layer, a self-adaptive weighted sound and vibration fusion layer and a self-adaptive POOLING layer are provided, and the data set collected in the step 1 is used for model training till convergence; and step 3, the trained model is used online, and the trained MCAVFCNN model is applied online. According to the method, state monitoring and fault detection can be accurately carried out on the running motor in real time, and unnecessary loss is reduced while safe running of the motor is guaranteed.

The utility model discloses a three-way pipe fitting injection mold which comprises a lower mold, a lower mold cavity is formed in the top of the lower mold, a mold core structure is arranged in the lower mold cavity in a sliding mode, the mold core structure comprises a first mold core and a second mold core which are perpendicular to each other, and a reset part is arranged at the bottom of the mold core structure. The reset part comprises a second spring used for pulling the mold core structure out of the lower mold cavity. The upper mold is arranged above the lower mold, and an upper mold cavity is formed in the position, corresponding to the lower mold cavity, of the upper mold. The mold has the beneficial effects that the first positioning inclined plate and the second positioning inclined plate are utilized to ensure that the first mold core and the second mold core accurately move to required positions, so that deviation and defects in production can be reduced; the size and the shape of each product are ensured to meet strict quality standards; the automatic reset function of the mold core is achieved through the second spring, the production process is simplified, and the risk of mechanical faults is reduced.

The invention discloses a transport robot motor torque evaluation method, and relates to the technical field of torque evaluation, and the method comprises the steps: collecting a motor current signal of a transport robot, an actual torque signal of an output shaft, and a driving wheel vibration signal; constructing a time-frequency matrix of the actual torque signal and the driving wheel vibration signal, and extracting a fault sensitive factor through singular value decomposition; order analysis is carried out on the motor current signal to obtain net abnormal torque; and when the net abnormal torque is larger than a first preset threshold value and the fault sensitive factor is larger than a second preset threshold value, it is judged that a mechanical fault occurs in the motor of the transportation robot. According to the invention, the accuracy of motor torque evaluation of the transportation robot is improved.

The invention discloses a shaftless disc type turbine self-powered flow monitoring device. The device comprises a static assembly, a rotating assembly, a bearing and a data collection module which are coaxially arranged. Magnetic induction coils are distributed in the shell in the circumferential direction; the shell end cover is provided with a guide vane; the turbine comprises shaftless fluid power blades and permanent magnets which are uniformly distributed in the circumferential direction, the magnet yoke is rigidly connected with the impeller assembly, and the permanent magnets are arranged in a Halbach array; the dynamic sealing interface adopts a three-stage labyrinth sealing structure; the blade has an NACA airfoil section; the magnetic induction coil is of a three-layer close winding structure. Self-power-generation is achieved through the electromagnetic induction principle, an external power source is not needed, and cost is reduced; the shaftless design reduces mechanical faults, improves the reliability and prolongs the service life; the data processing module can realize high-precision measurement; a three-stage labyrinth sealing structure prevents fluid leakage; the optimized blade parameters adapt to fluid flow; the design of the permanent magnet assembly and the magnetic induction coil enhances the magnetic field and improves the power generation efficiency and the anti-interference capability.

The utility model provides sectional material storage equipment, which relates to the field of sectional material storage, and adopts the technical scheme that the sectional material storage equipment comprises a material storage frame and a support frame, and further comprises a sectional material correction unit, the sectional material correction unit comprises a lifting beam, a conveying belt assembly and a positioning assembly, the lifting beam is arranged along the sectional material conveying direction and is connected with a driving part, and the conveying belt assembly is connected with the positioning assembly. The driving part can drive the lifting beam to move in the vertical direction, the driving part is connected with the input assembly, the multiple conveying belt assemblies are arranged on the lifting beam in parallel, the conveying direction of the conveying belt assemblies is perpendicular to the conveying direction of the input assembly, and the tops of the conveying belt assemblies are higher than the input assembly after the conveying belt assemblies are lifted. The positioning column is located at the conveying final position of the conveying belt assembly, the lower portion of the positioning column is connected with a supporting plate, and the top of the supporting plate is not lower than the top of the conveying belt assembly after descending. The sectional material correcting device can automatically correct sectional materials, avoids deflection and mechanical faults, and achieves full-automatic warehouse-in and warehouse-out.

The invention discloses a rotating machine fault diagnosis method and device and a storage medium in the technical field of mechanical fault diagnosis. The method comprises the following steps: acquiring a discretized demodulation signal; solving the discretized demodulation signal through an augmented Lagrangian multiplier method and an alternating direction multiplier method, and obtaining an iteratively optimized demodulation signal; calculating an instantaneous frequency increment based on the demodulated signal after iterative optimization, applying smoothness constraint and low-rank constraint to the instantaneous frequency increment, and solving through an iterative solution algorithm to obtain an estimated instantaneous frequency; and based on a pre-acquired theoretical fault characteristic frequency and the estimated instantaneous frequency, determining the fault type of the rotating machine according to a matching result. According to the method, the technical problems that the instantaneous frequency presents a non-stationary feature and is easily influenced by background noise, so that the accurate instantaneous frequency is difficult to extract, and the potential fault of the rotating machine cannot be accurately diagnosed can be solved.

The invention relates to the technical field of wind turbine generator system fault diagnosis, and discloses a wind turbine generator system gear box output end bearing comprehensive diagnosis method based on multi-model collaboration, and the method comprises the steps: predicting the bearing temperature trend through a dynamic power-bearing temperature model, and carrying out the temperature early warning through a power-temperature difference model; confidence calculation is carried out in cooperation with multi-dimensional data such as mechanical vibration characteristics (time domain / frequency domain impact energy reflects bearing mechanical damage), bearing temperature trend and temperature difference analysis (eliminating environment and load interference to position real fault temperature rise), and mechanical faults such as roller peeling and inner ring cracks and operation faults such as heat dissipation failure can be distinguished; and a traditional'single-point alarm 'mode is broken through, a'root analysis-unit power reduction / trigger shutdown' whole-process closed loop is constructed, and the safety, reliability and operation and maintenance efficiency of unit operation are remarkably improved.

The invention provides a bearing fault diagnosis method and system based on a twin neural network under a small sample, and relates to the technical field of mechanical fault diagnosis and artificial intelligence. The method comprises the following steps: collecting X-axis, Y-axis and Z-axis vibration signals of a bearing, slicing, and generating a time-frequency grey-scale map through short-time Fourier transform to enhance feature expression; the method comprises the following steps: constructing same-class and different-class sample pair training sets, training by adopting a weight-shared twin neural network model which comprises two same sub-networks, and optimizing model parameters by calculating the Euclidean distance of sample pair features and combining a cross entropy loss function; in the test stage, unknown samples are matched based on One-shot and N-shot strategies, and the fault category is judged with the maximum similarity probability. According to the method, the accuracy of 95% or above is achieved under 70 training samples, the noise immunity and generalization ability under the small sample condition are remarkably improved, and the method is suitable for low-cost intelligent diagnosis of industrial equipment.

The invention discloses a rotating machine fault diagnosis method based on an expansion residual network, and relates to the technical field of mechanical faults, and the method comprises the steps: collecting a continuous time sequence vibration signal of a rotating machine through an acceleration sensor, obtaining a high-order harmonic energy sequence set after fast Fourier transform, and determining an abnormal trend segment; then, obtaining and updating a structural anomaly candidate set based on the anomaly trend segment and the operating environment state of the rotating machine, generating a first drift vector set according to the updated structural anomaly candidate set, and identifying a structural offset section according to the first drift vector set and historical operating data; and backtracking the structural offset section to an original time domain signal to establish a second drift vector set, and finally, inputting the first drift vector set and the second drift vector set into the expansion residual network to analyze whether the rotating machinery is evolved into a fault state.

The invention discloses a lightweight rotating machine fault diagnosis method based on SEFormer. The method comprises the following steps: 1) data acquisition: acquiring vibration signals of key transmission parts (such as a bearing and a gear) of a rotating machine in various health states; 2) data preprocessing: segmenting the vibration signals into sample sets through sliding window sampling, and dividing the sample sets into a training set, a verification set and a test set; 3) model construction: designing and developing a separable multi-scale deep convolution SMDC and an efficient self-attention ESA, and constructing an SEFormer model; 4) model training, verification and evaluation: training the model on the training set and the verification set, and evaluating the model performance on the test set; and 5) fault diagnosis: performing fault diagnosis on the rotating machine by using the trained model. According to the method, a lightweight SEFormer model is constructed for rotating machine fault diagnosis, multi-scale features are extracted and integrated from different channel dimensions of the vibration signals, key fine-grained features of the vibration signals are captured in a global range, and the method has the advantages of robustness, generalization ability and lightweight.