69 results about "Density curve" patented technology

The invention relates to an electromechanical fault prediction and diagnosis method and system based on big data, and the method comprises the steps: collecting the operation state data of electromechanical equipment in real time, and synchronously obtaining historical associated data; performing dynamic feature extraction on the operation state data and the historical associated data, constructing a sliding mean value feature matrix, and calculating dynamic weights of feature parameters; generating a fusion weight coefficient according to the dynamic weight and a preset fault threshold interval; extracting a distribution density curve of a historical fault occurrence probability, and calculating a dynamic threshold value; performing weighted reconstruction on the sliding mean feature matrix based on the fusion weight coefficient, and outputting a fault type and an occurrence probability through a pre-trained lightweight residual neural network model; and when the fault occurrence probability exceeds a dynamic threshold value adjusted based on a historical fault occurrence probability distribution density curve, generating an electromechanical fault diagnosis result so as to realize the purposes of real-time monitoring of the operation state of the electromechanical equipment and accurate fault prediction and diagnosis.

The invention discloses a method for predicting a favorable area of a thin-layer stacked sand body, and relates to the technical field of exploration and development of oil and gas reservoirs, and the method comprises the steps: generating a synthetic seismic record through well-seismic combination by using interval transit time and a density curve, aligning the synthetic seismic record with an actual seismic record, and building a three-dimensional frame model of a target layer of a research area; performing abnormal value elimination and standardization processing on the logging curve, making a statistical histogram, comparing peak separation degrees of sandstone and mudstone, and selecting the most sensitive curve to distinguish the sandstone and mudstone; and realizing well-side seismic trace waveform dynamic clustering analysis through singular value decomposition, and determining the number of effective samples. Based on the result of waveform indication simulation, the method is combined with the optimized seismic attribute to predict the thin-layer stacked sand body together, a systematic method is provided, the multiplicity of solutions of predicting the thin-layer stacked favorable sand body through the seismic attribute and the uncertainty of depicting the sand body by the sedimentary facies in the area with few wells can be reduced, and the prediction accuracy of the thin-layer stacked favorable sand body is improved. Therefore, the precision and reliability of thin-layer stacked sand body prediction are improved.

The invention relates to the technical field of power equipment safety detection, and discloses a power equipment thermal fault diagnosis method and device, a storage medium and computer equipment, and the method comprises the steps: collecting on-site infrared image data of a transformer substation, constructing an infrared image data set, and carrying out the image enhancement processing; performing target identification and segmentation by using an instance segmentation model combining a cross-stage local network and an attention mechanism, and determining a power equipment region; inputting the gray value of the power equipment area into the deep neural network model to obtain a temperature value of the target power equipment, drawing a probability density curve of the temperature value by using a kernel density estimation method, obtaining a reference temperature of the target power equipment, and calculating a hot spot temperature; and calculating the relative temperature difference and determining a fault state based on a preset fault judgment standard. According to the method, the original data quality is improved, the power equipment is accurately identified and segmented, high-precision temperature prediction is carried out, and finally, full-automatic and reliable fault judgment of the power equipment is completed.

The invention discloses an equipment remaining service life prediction method and system, and belongs to the technical field of equipment state monitoring, and the method comprises the steps: obtaining the whole life cycle degradation data of equipment as a data set; performing data dimension reduction by using an automatic encoder to obtain an equipment health index HI; inputting the HI into a fusion type RUL prediction network model, and predicting a probability density curve of the RUL; wherein the model comprises a probability prediction network, a Wiener process model and a fusion module; the probability prediction network is used for obtaining a probability density curve in a numerical form; the Wiener process model is used for obtaining a probability density curve in an analysis form; and the fusion module is used for dynamically weighting the two curves to realize RUL prediction. According to the method, the RUL probability density curve prediction is realized, the problem that an existing prediction model based on machine learning can only obtain a point prediction result is solved, and the interpretability of the model is improved in combination with a random process.

The invention relates to a fault diagnosis method and device for an IGBT inverter, equipment and a storage medium, and the method comprises the following steps: diagnosing whether the IGBT inverter has a fault or not based on original electrical working parameters, and carrying out the probability distribution calculation of the original electrical working parameters if the IGBT inverter has the fault, and obtaining a fault feature probability density curve; carrying out topological structure analysis on the IGBT inverter based on the fault feature probability density curve to obtain a fault feature topological graph; performing graph attention network analysis on the fault feature topological graph to obtain a fault propagation path set; and performing causal reasoning analysis on the IGBT inverter based on the fault propagation path set to obtain a fault type identification result, and formulating a corresponding fault repair strategy based on the fault type identification result, thereby solving the problems that many methods are lack of dynamic response capability to electrical parameter change, cannot accurately capture early fault signals in real time, and are poor in reliability. And therefore, the hysteresis quality of fault diagnosis is caused.

The invention discloses a density attribute earthquake prediction method. The method comprises the following steps: forward modeling a longitudinal wave velocity, a transverse wave velocity and a density curve of saturated fluid rock; continuing forward modeling to generate a large number of longitudinal wave velocity, transverse wave velocity and density curves of different pseudo wells by using the bulk modulus and density parameters of the rock skeleton in the research area, and generating and synthesizing a pre-stack seismic gather under the guidance of a Zoeppritz equation; convolutional neural network training is carried out by using the longitudinal wave velocities, the transverse wave velocities and the density curves of the different pseudo wells and the synthetic pre-stack seismic gathers, and a trained convolutional neural network model is obtained; and applying the convolutional neural network model to an actual well and trace gather, correcting the trained convolutional neural network model, and performing density attribute prediction by using the corrected convolutional neural network model and applying the actual trace gather. Density attribute prediction is carried out through a deep learning method based on rock physical model sample expansion, and the problem that in the prior art, conventional elastic parameters cannot identify effective reservoirs under the medium-deep layer condition is solved.

The invention discloses a method for calculating fatigue life reliability of an asphalt pavement, and relates to the technical field of road engineering. According to the method, actually-measured temperature field historical data and traffic axle load data of an existing asphalt pavement structure are obtained through an asphalt pavement long-term performance observation network, the actually-measured temperature field historical data are extracted to construct a temperature field historical data set, and a temperature cumulative probability distribution curve is generated; then, utilizing traffic axle load data to generate an axle type cumulative probability density curve and an axle load interval cumulative probability density curve, and based on Monte Carlo simulation, performing inverse function sampling in each cumulative probability density curve during each simulation to obtain the structural response of the asphalt pavement structure to determine the standard fatigue life; and determining the reliability of the fatigue life by counting the probability that the standard fatigue life obtained by Monte Carlo simulation is greater than the standard cumulative action times within the design period. According to the method, the fatigue life of the asphalt pavement structure is accurately estimated, and the design reliability of the asphalt pavement structure is effectively guaranteed.

The invention relates to a probabilistic load flow calculation method for new energy and load correlation, and the method specifically comprises the steps: building a probabilistic model of new energy and load based on the distribution types and parameters of the new energy and load, inputting network structure parameters, the probabilistic model information of the new energy and load, and correlation coefficient information, the method comprises the following steps of: setting a median Latin hypercube sampling scale and an iteration number marking variable, then sampling mutually independent random variables such as new energy and load and random variables such as new energy and load with a correlation so as to generate random samples, and finally, based on a Newton-Raphson method, calculating a maximum iteration number of the new energy and the load so as to obtain a maximum iteration number of the new energy and the load. Performing power flow calculation on the obtained random samples in sequence, finally performing statistics on power flow results such as node voltage, line transmission power and line loss rate obtained by each power flow calculation, drawing a probability density curve, and performing probability power flow calculation of new energy and load correlation by using the probability density curve; the method has the advantages of being accurate in calculation, fast and efficient.

The invention provides a data reading method of a memory, storage equipment and a storage medium, and the data reading method of the memory comprises the following steps: carrying out reading operation on a programming unit to obtain target data; according to the bit flipping information of the target data, determining a corresponding failure bit counting exception type; wherein the failure bit counting exception type comprises a first type of exception caused by read voltage offset and a second type of exception caused by trough elevation of a probability density curve of a break-over voltage of a storage unit; in response to the first type of abnormality, adjusting the read voltage; or in response to the second type of exception, performing data refreshing operation or garbage collection operation. By means of the mode, the abnormal type causing reading voltage deviation is judged so as to judge the data state, then the rereading rate of a host is reduced, and the reading effectiveness is improved.

This invention discloses a method, system, and program product for identifying the fault severity of underwater robot thrusters under conditions of relatively mild faults, belonging to the field of underwater robot fault diagnosis technology. The invention first obtains a thrust deviation curve based on the bow angle and lateral thruster control voltage. Then, based on different parts of the curve, it uses difference and multi-time-window sliding Fourier transform methods in the time and frequency domains respectively to obtain multiple thrust losses. Next, it introduces a kernel density estimation method to obtain the time-domain and frequency-domain fault severity probability density curves based on the thrust losses obtained in the time and frequency domains. Finally, it uses a method of averaging at the same location to fuse the curves, and identifies the fault severity based on the fused fault severity probability density curves. This invention achieves high identification accuracy and is particularly suitable for fault severity identification in ocean current environments and under conditions of relatively mild faults.

This invention relates to a method, apparatus, medium, and equipment for analyzing the overpressure characteristic values ​​of combustible gas explosions in marine oil and gas production facilities. The analysis method includes the following steps: geometrically modeling the marine oil and gas production facility; calculating the congestion degree of the process area of ​​the marine oil and gas production facility; calculating the volume of the combustible gas cloud under different filling ratios in the process area of ​​the marine oil and gas production facility; performing transient CFD simulations of combustible gas explosions with varying gas cloud positions and ignition positions to obtain the overpressure value of the protected target changing over time; determining the maximum overpressure value of the protected target under combustible gas explosion conditions; expanding the sample and performing statistical analysis and characteristic value calculations to obtain probability density curves and characteristic values; and determining the explosion-proof level of the protected target according to protection requirements.

The present invention discloses an optimization scheduling method and terminal based on probability box and conditional risk value, which determine the prediction errors corresponding to the output of distributed renewable energy and the load demand in a preset period; determine the corresponding cumulative probability density curve based on each prediction error, and determine the interval set that meets the preset confidence from the cumulative probability density curve based on each prediction error according to the probability box theory; calculate the final uncertainty factor fluctuation interval corresponding to the interval set that meets the preset confidence based on the conditional risk value theory; construct an optimization scheduling model based on the final uncertainty factor fluctuation interval, and use an intelligent optimization algorithm to solve the optimization scheduling model to obtain an optimization scheduling plan, thereby ensuring the integrity of uncertain information, calculating uncertainty situations outside the confidence interval, avoiding the conservatism of traditional interval optimization methods while ensuring the sufficiency of safety constraints, and achieving an effective balance between the economy and safety of the distribution network.

The present application relates to the technical field of rock fracture modeling, and discloses a method and a processor for modeling a multi-scale fracture network of a tight reservoir. The method comprises: obtaining a strike attribute of a fracture; generating a fracture strike conforming to a preset mode according to a cumulative probability density curve of the strike; taking a fracture position interpreted from a core and an imaging logging as hard data of a fracture position in a fracture network; dividing the hard data into seed hard data and correction hard data; correcting a cumulative probability curve of fracture development intensity according to the seed hard data, and generating a fracture element set according to the corrected probability density curve; generating a discrete fracture network through similarity fusion criteria, mechanical cause fusion criteria and intersection criteria; comparing the established discrete fracture network with the correction hard data, and correcting the number of fracture elements and the cumulative probability curve. Through iterative inversion of the seed hard data and the correction hard data, a more optimal multi-scale fracture network model that is more consistent with the hard data is generated.

The invention discloses a threaded borehole density curve correction method based on numerical simulation and frequency domain filtering, and relates to the technical field of petroleum and natural gas development, and the process is as follows: obtaining stratum apparent density results under different thread depths by using threaded borehole stratum simulation models with different thread depths; performing frequency domain filtering processing on the stratum apparent density result to obtain a density curve frequency domain filtering processing result under each thread depth condition; then establishing a mathematical relationship between the deviation value and the corresponding thread depth to obtain a density deviation value chart suitable for frequency domain filtering under different thread depths; performing frequency domain filtering processing on the actually measured density curve, and outputting a frequency domain filtering result of the density curve; and finally, superposing the deviation value to a frequency domain filtering result of the actual measurement density curve to obtain a final correction result. According to the method, the density deviation value charts under different threaded well hole conditions are established through a numerical simulation method, the density curve after frequency domain filtering is corrected, and the accuracy of the density logging curve is improved.

The present invention discloses a channelized detection method based on robust noise floor estimation, belonging to the field of electronic reconnaissance. The radar signal is filtered, detected, and synthesized to output the detection result. The steps are as follows: First, after the radar signal is successively subjected to AD sampling and polyphase filtering, filtered data is obtained; according to the signal accumulation envelope sequence, an envelope probability density curve is constructed, the position of the sequence extreme point is found, the noise ratio of the sequence is calculated according to the position of the sequence extreme point, and the noise accumulation envelope sequence of each channel is obtained; using the noise accumulation envelope sequence, according to the expression of the cumulative distribution function of the Gaussian distribution, the noise floor mean estimate value and the noise floor variance estimate value are calculated; a reliable false alarm probability is determined by simulation experiments, and the detection threshold is calculated; by comparing the signal accumulation envelope with the detection threshold, the detection VP is obtained.

The invention discloses a method, a system and a program product for identifying the fault degree of an underwater robot propeller under the condition of weak fault degree, and belongs to the technical field of underwater robot fault diagnosis. According to the method, firstly, a thrust deviation curve is obtained according to a heading angle and lateral propeller control voltage, and then according to different parts of the curve, a difference value and a multi-time-window sliding Fourier transform method are adopted in a time domain and a frequency domain to obtain multiple thrust losses; secondly, introducing a kernel density estimation method to obtain thrust loss based on a time domain and a frequency domain, and respectively obtaining a time domain fault degree probability density curve and a frequency domain fault degree probability density curve; and finally, fusing by adopting a same position averaging method, and identifying the fault degree based on the fused fault degree probability density curve. The method can achieve high identification precision, and is especially suitable for fault degree identification in an ocean current environment and under the condition that the fault degree is weak.

The invention discloses a variable stiffness composite material wing aeroelastic analysis method containing multi-source uncertainty, which comprises the following steps: carrying out probability space division according to the condition of random uncertainty factors, and determining the value of a random vector at each sampling point and the probability value of a sub-domain; regarding the random vector at each sampling point as a deterministic parameter, and circularly calculating an interval boundary value of a to-be-analyzed aeroelastic response variable at each sample point; based on the interval boundary value, calculating the reliability of the structure at each sample point only under the influence of the interval factors; discretizing the initial condition of the generalized probability density evolution equation to obtain the probability density of the structural reliability in each sub-domain; superposing the probability density of each sub-domain to obtain a probability density curve of the structural reliability considering the multi-source uncertainty; and calculating the reliability of the structure based on the probability density curve. By adopting the technical scheme of the invention, the aeroelasticity analysis containing the multi-source uncertainty of the variable stiffness composite material wing is realized.

The application provides a curve reconstruction method and device based on deep learning, equipment and medium, the method comprises the following steps: obtaining the logging curve data of the non-expanded section, and establishing a training set and a test set; a density curve reconstruction model is established and the training set is used to train the density curve reconstruction model, the density curve reconstruction model is a deep learning model; the test set is used to test the density curve reconstruction model, and a reconstructed density curve is generated, the density curve reconstruction model is adjusted until the correlation of the reconstructed density curve and the original density curve of the non-expanded section and the relative error of the density curve reconstruction model reach a preset condition; the density curve of the target well is reconstructed by using the adjusted density curve reconstruction model. The application simplifies the operation process, especially in complex formation conditions, can accurately reflect the actual situation of the formation, and improves the calculation accuracy of the reservoir parameters.

The present invention proposes a two-dimensional feature selection and verification method for optical remote sensing image similarity assessment, which belongs to the field of optical remote sensing detection image simulation model accuracy assessment. The method first selects the measured optical remote sensing data product of the area to be evaluated, extracts the detection parameters from the measured data, uses the optical remote sensing imaging simulation model to simulate the image of the extracted data, respectively performs two-dimensional Fourier transform, two-dimensional spatial frequency domain filtering and two-dimensional inverse Fourier transform processing on the measured image and the simulated image, extracts the DC component, low-frequency component and high-frequency component in the two images, compares and calculates the amplitude difference estimation, feature difference estimation and global difference estimation pixel by pixel, forms three types of estimated two-dimensional images, and uses the statistical histogram and cumulative probability density curve of the estimated image to evaluate the similarity of the two images. The present invention has a high sensitivity to the micro-change features of the image, and can have a good recognition and evaluation of the texture details of the image.

The invention provides a method and system for determining the density of a reservoir in a less-well area, a storage medium and a processor, and belongs to the technical field of geophysical exploration. The method comprises the steps that a rock physical model is constructed based on logging, the incidence relation between a density curve and an elastic parameter of logging is determined through the rock physical model, the incidence relation between the density curve and the elastic parameter is a relational expression of density and a first parameter under the constraint of the longitudinal and transverse wave velocity ratio, and the first parameter is the ratio of the shear modulus to the longitudinal wave velocity; acquiring a low-frequency data volume and an intermediate-frequency data volume of the reservoir in the less-well area, and combining the low-frequency data volume and the intermediate-frequency data volume to obtain an absolute data volume which comprises longitudinal wave impedance, transverse wave impedance and a longitudinal-transverse wave velocity ratio; and according to the longitudinal wave impedance, the transverse wave impedance and the longitudinal and transverse wave velocity ratio, calculating a first parameter of the reservoir in the less-well area, and substituting the absolute data volume and the first parameter into an association relationship between the density curve and the elastic parameter to obtain the density of the reservoir in the less-well area. According to the embodiment of the invention, stable and reliable reservoir density data of the less-well area can be obtained.

The invention relates to the technical field of power system analysis, and particularly discloses a power distribution network load flow calculation method and system, electronic equipment and a storage medium, and the method comprises the steps: carrying out the load flow calculation through employing a Monte Carlo method, and obtaining a first probability density curve and a first cumulative distribution curve of a random variable of a power distribution network; by utilizing a semi-invariant and Gram series expansion method, selecting preset semi-invariants with different orders to carry out load flow calculation to obtain a second probability density curve and a second cumulative distribution curve of the random variable of the power distribution network under the preset semi-invariants of each order; and determining the order of the semi-invariant with the minimum curve difference value as the target order of the semi-invariant, and determining the second probability density curve and the second cumulative distribution curve of the random variable of the power distribution network under the semi-invariant of the target order as the load flow calculation result of the power distribution network. The calculation amount of power flow calculation of the power distribution network can be reduced, and valuable reference is provided for operation planning and decision making of the power distribution network.

The present invention relates to the field of data processing, and specifically to a method for improving the data accuracy of a gyro array on a stable platform. The method includes the following steps: S1. Primary screening: A low-pass filter is used to smooth and denoise the signal, and the performance of the gyro unit is quantified in the form of variance. The larger the variance, the more severe the gyro jitter; S2. Secondary screening: In the normality test, a frequency distribution histogram is plotted, the fitting Q-Q plot is visualized, and the probability density curve of the normal distribution is fitted; S3. Temperature drift compensation: The gyro drift value is obtained by the encoder differential velocity closed-loop. In the full temperature range, when the velocity obtained by the encoder differential is 0, the gyro data on the stable two-axis platform is sampled, and then the temperature drift curve after acquisition is fitted. The present invention uses the secondary screening algorithm and the temperature drift compensation algorithm to improve the output data accuracy of the gyro array, and greatly improves the control effect of the stable platform.

The invention discloses a short-term power load probability prediction method based on an improved neural network, and belongs to the technical field of power system load prediction. The method comprises the following steps: acquiring and preprocessing load and related data; the method comprises the following steps: constructing a DCS-CNN-BiLSTM-Attention-QR probability prediction model fusing a convolutional neural network, a bidirectional long short-term memory network, an attention mechanism, quantile regression and kernel density estimation; automatically optimizing key hyper-parameters of the model by utilizing a difference creation search algorithm; carrying out model training by adopting a quantile loss function to obtain load prediction values under different quantiles; and finally, generating a probability density curve of the prediction interval through kernel density estimation. According to the method, deterministic point prediction is expanded into probabilistic interval prediction, the uncertainty of load prediction can be quantified, the model precision and generalization ability are improved through an intelligent optimization algorithm, and richer and more reliable decision information is provided for power system scheduling and risk management.

This invention relates to the field of data processing technology, specifically to an efficient data processing method for communication networks. The method includes: acquiring a network data sequence to be processed; obtaining a feature value sequence based on the differences between adjacent network data in the network data sequence; obtaining suspected drift anomaly data based on the similarity between the normal distribution curve and probability density curve corresponding to the feature value sequence; obtaining an optimal second data window corresponding to each suspected drift anomaly data based on the difference between the feature value corresponding to each network data in each second data window and the mean corresponding to the feature value sequence; obtaining each drift anomaly data based on the first data window and the optimal second data window; and obtaining the target network data sequence based on the optimal second data window. This invention can eliminate the impact of drift anomaly data on detection accuracy, resulting in higher detection accuracy.

The invention relates to the field of data processing, in particular to a method for improving the precision of gyroscope array data of a stabilized platform. The method comprises the following steps: S1, carrying out primary screening: carrying out smooth denoising processing on a signal by adopting a low-pass filter, quantifying the performance of a gyroscope unit in a variance form, and showing that the gyroscope jittering degree is more severe when the variance is larger; s2, double screening: drawing a frequency distribution histogram in a normality test, visually fitting a Q-Q diagram, and fitting a normal distribution probability density curve; and S3, temperature drift compensation: performing closed loop on differential speed of the encoder to obtain a gyroscopic drift value, sampling gyroscopic data on the stable two-axis platform when the differential speed of the encoder is 0 in a full temperature range, and then fitting a collected temperature drift curve. The precision of output data of the gyroscope array is improved by utilizing a double screening algorithm and a temperature drift compensation algorithm, and the control effect of the stable platform is greatly improved.

The invention provides a distributed energy distribution network weak point identification method and system based on voltage distribution characteristics. The method comprises the following steps: obtaining a DG initial capacity coefficient matrix through a quasi-Monte Carlo method of a Sobol sequence; determining a convergence and dispersion discrimination result matrix of a part of the DG initial capacity coefficient matrix through load flow calculation, training the SVM model, and predicting load flow convergence and dispersion of the remaining part of the DG initial capacity coefficient matrix; screening all capacity coefficient matrixes of load flow convergence in the DG initial capacity coefficient matrix, gradually increasing the DG capacity according to a screening result until the voltage exceeds the upper limit, and recording the response voltage of the power grid system; voltage probability density curve charts of different nodes are obtained in combination with a probability distribution mapping technology and an error optimization fitting algorithm; calculating out-of-limit voltage difference indexes and out-of-limit voltage variance indexes of the nodes and sorting; and identifying the weak point of the power distribution network by combining the voltage probability density curve graph and the index sorting result. According to the method, the weak points of the distributed energy distribution network can be quickly and effectively identified.

The present invention discloses a method for assessing the structural morphology of engineering rock masses. The method involves percussing a hole in the engineering rock mass, recording and storing the drilling parameters in real time during the drilling process. After the hole is formed, a high-definition camera is used to image the entire hole along the depth of the hole. The drilling parameters are filtered. The principal component analysis method is used to reduce the dimension of the drilling parameters to determine two principal components and calculation parameters. The rock mass state in the hole is classified into several types based on the in-hole imaging photos. The principal component density map of each type of rock mass is plotted based on the calculation parameters to determine the first principal component. The probability density curve of the first principal component of each type of rock mass is plotted to divide the first principal component intervals of various types of rock mass. The first principal component value is calculated based on the drilling parameters to determine the rock mass state type. The present invention helps technicians use drilling measurement parameters to quickly and accurately assess the structural morphology of engineering rock masses, providing a decision-making basis for tunnel support, excavation, and mine planning and design.

The invention relates to a data analysis method and device for a wind turbine generator, electronic equipment, a storage medium and a computer program product. The data analysis method comprises the steps of obtaining historical data of target parameters of the wind turbine generator; calculating a target probability density curve of the target parameter based on the historical data of the target parameter; and evaluating the operation performance of the wind turbine generator based on the target probability density curve. Therefore, the evaluation of the operation performance of the wind turbine generator does not depend on any input information, but calculates the probability density curve corresponding to the parameters of the wind turbine generator based on the real operation data of the wind turbine generator in the historical operation process, and then evaluates the operation performance of the wind turbine generator based on the calculated probability density curve. According to the method, the actual operation rule of the wind turbine generator is fully combined, namely the design principle and the control characteristics of the wind turbine generator are fully utilized, the control mode of the wind turbine generator can be truly restored, and the accuracy of evaluating the operation performance of the wind turbine generator is improved.

The invention discloses a weight-free health degree analysis method for industrial equipment, which belongs to the technical field of electric digital data processing, and comprises the following steps: in a health state, according to a data index type for analyzing the health degree, respectively acquiring industrial time sequence data in at least one historical time window of the industrial equipment as reference data, collecting industrial time sequence data in the current time window as comparison data; according to the reference data and the comparison data, probability density curves are obtained through kernel density estimation one by one, so that probability distribution is obtained; and according to the data index type, obtaining JS divergence values for measuring data distribution similarity under the data index type according to the probability distribution of the comparison data and the probability distribution of the at least one piece of corresponding reference data, so as to obtain the health degree of the industrial equipment according to the plurality of JS divergence values. According to the invention, based on kernel density estimation and JS divergence, the distribution difference is quantified, weight setting is not needed, and the sensitivity and consistency of health degree assessment are improved.

The invention discloses a wind turbine generator bearing fault detection method and system. The method comprises the steps of collecting power, wind speed, generator bearing temperature and generator rotating speed of a wind turbine generator in different set days; performing data box separation and quartile distance anomaly detection based on power and wind speed; dividing the operation data matrix with abnormal values removed by day, and fitting the wind speed and power in each sub-matrix into a sigmoid curve as a wind speed and power curve of each day; eliminating data of days in a non-full power generation state based on a wind speed power curve of each day; dividing the operation data matrix in the normal state into a plurality of samples, fitting the generator rotating speed and the generator bearing temperature in each sample into a set nonlinear function, constructing a probability density curve of a degradation value of each sample, obtaining the bearing degradation probability, and if the bearing degradation probability exceeds a set degradation probability threshold, determining that the bearing is not damaged; if yes, determining that a fault exists. According to the invention, the overall operation and maintenance cost is saved, and the power generation efficiency is improved.