51 results about "Time scale decomposition" patented technology

A depolarization waveform classifier based on the Modified lifting line wavelet Transform is described. Overcomes problems in existing rate-based event classifiers. A task for pacemaker / defibrillators is the accurate identification of rhythm categories so correct electrotherapy can be administered. Because some rhythms cause rapid dangerous drop in cardiac output, it's desirable to categorize depolarization waveforms on a beat-to-beat basis to accomplish rhythm classification as rapidly as possible. Although rate based methods of event categorization have served well in implanted devices, these methods suffer in sensitivity and specificity when atrial / ventricular rates are similar. Human experts differentiate rhythms by morphological features of strip chart electrocardiograms. The wavelet transform approximates human expert analysis function because it correlates distinct morphological features at multiple scales. The accuracy of implanted rhythm determination can then be improved by using human-appreciable time domain features enhanced by time scale decomposition of depolarization waveforms.

The invention discloses a rolling bearing failure diagnosis method base on vibration temporal frequency analysis. The method comprises the following steps: utilizing a vibration acceleration sensor to collect vibration signals of the rolling bearing under a normal condition and a failure condition; utilizing a modified inherent time scale resolving method to resolve the collected vibration signals, and generating a plurality of inherent time scale components and residual signals; calculating relativity of the time scale components and the vibration signals, selecting the inherent time scale components of which the relativity is ranked top 5 as related components, and rejecting noise signals and false components; calculating Wigner distribution of the related components respectively, and conducting linear stack to obtain the Wigner temporal frequency figure of the original signal; extracting difference fractal box dimensionality of the Wigner temporal frequency figure and the image entropy as failure characteristics; utilizing mahalanobis distance to build mapping relation of the failure characteristics and failure types to realize failure diagnosis. According to the invention, interference of Wigner distribution cross terms is avoided; two kinds of representative failure characteristics of the difference fractal box dimensionality and the image entropy are confirmed.

The invention discloses a time scale function decomposition based hypersonic aircraft actuator saturation control method. The method is used for solving the technical problem of difficulty in engineering realization under the existing hypersonic aircraft actuator saturation condition. The method includes: obtaining a high-speed slow variable subsystem, a speed slow variable subsystem and an attitude fast variable subsystem by time scale decomposition, and building a discrete form of an original system through an Eulerian method; regarding the height and the speed in a fast subsystem design process as constants so as to achieve model simplification; considering actuator saturation limitations, and importing auxiliary control variables to design throttling valve openness and the controlpiston deflexion angle; and designing an updating law of a neural network by importing an auxiliary error variable. The time scale function decomposition based hypersonic aircraft actuator saturation control method has the advantages that computer control characteristics are combined, a discrete model is built, the subsystems are designed according to time scale function decomposition, the actuator saturation condition is fully considered during controller design, and the method is suitable for engineering application.

The invention discloses a rotor system fault diagnosis method and device based on vibration analysis. A sensor acquires normal conditions of a rotor system and vibration signals under fault conditions; the acquired vibration signals are decomposed by an improved inherent time scale decomposition method to generate a plurality of rotational components and residual signals; related rotational components capable of reflecting fault information are selected from the rotational components; energy of each related rotational component is calculated; related vector machine multi-classification models are built by an improved directed acyclic method; fault characteristics are inputted to the related vector machine multi-classification models for training and fault diagnosis. A motor, a first bearing block, a second bearing block and a third bearing block are arranged on a test bed base, the first bearing block, the second bearing block and the third bearing block respectively support a first rotating shaft and a second rotating shaft which are sequentially connected with an output shaft of the motor, both the first rotating shaft and the second rotating shaft are provided with a disk, and a sensor group is arranged at the end of the second rotating shaft. Rotor system fault types can be rapidly and accurately recognized, and the method and the device are applicable to online diagnosis of the rotor system.

The invention provides a method for recognizing and separating magnetotelluric signal and noise. Firstly, a magnetotelluric signal collected in a strong electromagnetic interference environment is equidistantly segmented, by optimizing inherent time-scale decomposition, each magnetotelluric signal segment is subjected to adaptive decomposition to obtain a rotational component, and a sample entropyand a fuzzy entropy of the rotational component are extracted. Then, the sample entropy and the fuzzy entropy are taken as joint feature parameters to carry out fuzzy C-means clustering, according tothe feature parameters and a clustering method, a measured magnetotelluric signal sequence is divided into two kinds including a useful signal and a strong interference signal. Finally, only a magnetotelluric signal segment which is recognized as the strong interference signal is subjected to noise suppression processing by using a wavelet threshold, and a reconstructed magnetotelluric signal isobtained after superposition. The method has high reliability.

A diesel engine fuel oil system fault diagnosis method based on a least square support vector machine comprises the steps of: collecting vibration acceleration signals of a diesel engine under conditions of normal work and various kinds of faults; utilizing an inherent time scale decomposition algorithm to decompose the vibration acceleration signals, and generating a plurality of rotation components and residual error signals; calculating typical frequency domain characteristics of first N-order rotation components, and using the typical frequency domain characteristics as fault characteristics; dividing training samples and test samples; utilizing a hybrid algorithm of a difference evolution algorithm and a particle swarm algorithm to optimize a punishment factor and a kernel function parameter of the least square support vector machine, and obtaining an optimal punishment factor and an optimal kernel function parameter; and utilizing the obtained optimal punishment factor and optimal kernel function parameter to train the least square support vector machine for carrying out fault diagnosis. By adopting the method provided by the invention, the operation state of the fault diagnosis can be rapidly and accurately judged, and the method is applicable to online diagnosis of the diesel engine.

The invention discloses a pseudo-random code estimation method of a direct sequence spread spectrum system, which mainly overcomes the shortcomings of high requirement on sampling precision, great error of carrier frequency estimation and poor universality of the prior method. The invention comprises the following steps: 1. obtaining a pseudo code period and a symbol period from a secondary spectrum and a secondary moment of related functions; 2. determining a starting point of the pseudo code of a period according to an autocorrelation matrix of a received signal; 3. taking a pseudo code starting point as the starting point for the received signal, taking a plurality of pseudo code periods as window size for subsection, carrying out inherent time-scale decomposition on each section through carrier frequency and a first zero-crossing point frequency, and obtaining respective instantaneous amplitude value; 4. subtracting one instantaneous amplitude value from the other instantaneous amplitude value to obtain a difference signal, and making a column diagram to set two thresholds; 5. recording the moment when a minimum point is lower than a smaller threshold in a pseudo code period ofthe difference signal; and 6. setting the polarity of starting code strip, determining the polarity of each code strip from the second code strip, and determining the pseudo code according to pseudocode property. The invention has the advantages of high estimation precision and strong universality, and is used for monitoring frequency spectrum.

The invention discloses an on-line detection method for industrial control loop multi-period oscillation. The method includes the following steps that in a control loop to be detected, a set of progress data are collected in real time in an on-line mode; improved intrinsic time-scale decomposition is performed on the progress data in real time in an on-line mode, and monitoring statistics corresponding to all obtained decomposition sub-signals are calculated in real time; whether all the monitoring statistics exceed a set threshold value omega or not is judged, and an on-line detection result is obtained according to all judgment results. By the utilization of the on-line detection method, multi-period oscillation behaviors of the industrial control loop can be detected in a quantified mode to obtain regularity degrees and periods of all oscillation components in the multi-period oscillation. Rich data supports are provided for evaluating the oscillation behaviors and diagnosing a fault source.

The invention discloses a bearing fault diagnosis method based on ITD and improved morphological filtering, which is a bearing fault diagnosis method based on intrinsic time scale decomposition (ITD)and improved morphological filtering, and belongs to the field of mechanical fault diagnosis. The method includes the following steps: firstly, a bearing vibration signal is subjected to ITD to obtaina series of rotation components (Proper rotation, PR), the effective PR with rich fault information is screened out based on the criterion of kurtosis, and each effective component is subjected to improved morphological filtering-ACDIF to extract the impact component and then reconstruct the signal, and finally, spectrum analysis is used to extract fault features in the reconstructed signal. Themethod utilizes the advantages of the adaptive decomposition capability for the signal of the ITD method and the effective filtering and retaining of impact components of the ACDIF, thereby improvingthe accuracy of the filtering output. An ACDIF filter not only has a simple structure design but also can be applied to the fault diagnosis of vibration signals, and the signal impact characteristicsafter filtering with the method are completely preserved.

The invention relates to a water turbine tail water pipe dynamic characteristic extraction method. The method comprises the following steps that pressure fluctuation signals of a water turbine tail water pipe under three states of no vortex strip, vortex strips and serious vortex strips are collected by a field test which is carried out by a hydroelectric generating set; the collected pressure fluctuation signals are subjected to resample under the three states of the water turbine tail water pipe, and high-frequency interference in the pressure fluctuation signals is removed; the pressure fluctuation signals subjected to resample under the three states are resolved by adopting an intrinsic time-scale decomposition method, and a monotonous baseline vector and a plurality of intrinsic rotational components are obtained corresponding to the pressure fluctuation signals under the three states; the dynamic characteristics of the water turbine tail water pipe are extracted by the approximate entropy of the intrinsic rotational components, which is respectively obtained by calculation, of the water turbine tail water pipe under the three states. The water turbine tail water pipe dynamic characteristic extraction method has the advantages of efficiency, strong instantaneity and the like, and can be widely applied to the fields of running guarantee of the hydroelectric generating set.

The invention discloses an ITD, spectral kurtosis, smooth iteration envelope analysis method of a rolling bearing. The ITD, spectral kurtosis, smooth iteration envelope analysis method comprises steps that an original signal is decomposed by adopting an intrinsic time-scale decomposition method, and a noise component and a trend term of a decomposition result are eliminated by using data rearrangement and replacement operation, and then a signal after primary filtering is analyzed by adopting a spectral kurtosis method, and then a central frequency and a central bandwidth of an optimal filter are acquired; the above mentioned optimal filter is used for the secondary filtering of the signal after the primary filtering, and then the envelope analysis of the signal after the secondary filtering is carried out by adopting the smooth iteration envelope analysis, and at last, the fault type of the rolling bearing is determined according to an envelope spectrum. The ITD, spectral kurtosis, smooth iteration envelope analysis method is suitable for processing complicated rolling bearing fault signals, and is capable of determining the fault type of the rolling bearing accurately, and has a good anti-noise performance and good robustness, and is convenient for engineering application.

A diesel engine fault diagnosis method based on vibration analysis includes acquiring normal conditions and vibration signals in fault of the diesel engine by a vibration acceleration sensor; decomposing the acquired vibration signals by means of improved inherent time scale and generating a plurality of inherent time scale components and residual signals; computing energy of the inherent time scale components, sequencing the inherent time scale components according to energy, and selecting former N components with accumulative energy larger than original signal energy by 95% as main components; computing kurtosis and waveform factor of the main components as fault characteristics; constructing a multi-classification correlation vector machine model by an error correcting and encoding method based on probability output for fault diagnosis. By the diesel engine fault diagnosis method, kurtosis and waveform factor as sensitive fault characteristics are determined, and an error correcting and encoding correlation vector machine multi-classification model based on probability output is provided; the diesel engine fault diagnosis method can accurately judge the operating states of a diesel engine and is applicable to fault diagnosis of diesel engines.

The invention discloses a transmission line insulator leakage current online monitoring method and a system in order to solve the technical problem that the existing insulator leakage current monitoring system has the disadvantages of high data communication cost and low accuracy of leakage current monitoring results. The method includes the following steps: a leakage current signal of each insulator on a transmission line is obtained in real time, and transmitted to a remote monitoring center through a wireless communication network; the remote monitoring center receives and stores each leakage current signal, separates a non-periodic component and a periodic component of the leakage current signal through Fourier transformation, removes inherent frequency interference in the periodic component, removes the noise in the non-periodic component through time scale decomposition, and uses the de-noised non-periodic component and the periodic component to reconstruct the leakage current signal; and if there is a leakage current signal greater than a preset leakage current threshold, the remote monitoring center issues a warning.

The present invention discloses a Beidou navigation satellite system carrier phase cycle slip detection and repairing method. The method comprises the steps of: employing a Beidou carrier phase observation amount and a pseudorange observation amount to construct cycle slip signals; employing an improved inherent time scale decomposition method, namely, an improved ITD method, to perform decomposition of the cycle slip signals to obtain a plurality of mutually separated inherent rotation component signals, namely PR components; screening out the PR components including the cycle slip; performing Hilbert spectral analysis for the PR components including the cycle slip to detect the epoch when the cycle slip occurs; and performing repairing of the PR component signals including the cycle sliptaken as training samples, employing the particle swarm optimization algorithm to perform parameter optimization of a least square support vector machine, namely an LS-SVM, employing the optimized LS-SVM to perform regression prediction, and calculating the difference of an actually measured value and a prediction value to determine the size of the cycle slip and repair the cycle slip. The methodcan effectively detect and repair the mini cycle slip appearing about one week in the Beidou single-frequency observation data.

The invention discloses an improved variation mode decomposition diagnosis method for engine main shaft bearing fault diagnosis. Firstly, a fault original signal is input into inherent time scale decomposition, the signal is decomposed into a plurality of inherent rotation components and a residual term, the residual term is filtered out, and key components of the signal are completely retained while the original signal is denoised; secondly, each inherent rotation component is subjected to variation mode decomposition, the optimal component in each group of IMFs is selected according to the kurtosis principle, and the signal is reconstructed; finally, the reconstructed signal is subjected to hilbert envelope transformation to diagnose the fault types of bearings. According to the method,on the one hand, the signal is denoised by means of inherent time scale decomposition, and the signal-to-noise ratio is increased; on the other hand, each inherent rotation component is self-adaptively decomposed to be close to the respective center frequency by means of variable mode decomposition, and the optimal components are selected to reconstruct the signal. The method has good denoising capability, completely retains fault information, and has strong fault diagnosis advantages.

The invention relates to a rolling bearing fault diagnosis method based on the PCA (principal component analysis) and ELM (extreme learning machine) and belongs to the technical field of fault diagnosis and signal processing and analysis. According to the method, intrinsic time-scale decomposition (ITD) is utilized to decompose vibration signals, and inherent rotation (Pr) components are screened,and the entropy values and time-domain features of all selected PR components are calculated; principal component analysis (PCA) is adopted to perform dimensionality reduction processing on the obtained features, and the dimensionality-reduced features are adopted to build an ELM (extreme learning machine) fault diagnosis model; and therefore, the recognition of the state of a rolling bearing isrealized. With the method of the invention adopted, the problem that the single feature of a rolling bearing is unlikely to accurately reflect the state of the bearing and the problem of the decreaseof the performance of a pattern recognition algorithm due to dimensionality increase with many features containing much bearing state information can be solved. A bearing experiment shows that the method can effectively recognize bearing states, and is simple in principle and high in practicability.

The invention belongs to the field of automobile maintenance, and relates to a fault diagnosis method for an automobile generator bearing. The fault diagnosis method comprises the steps that firstly,four channels of generator bearing fault signals are collected, and improved intrinsic time-scale decomposition (IITD) is used for decomposing each channel of original signals into proper rotation components (PRC) and a monotonous trend item; the PRCs are reconstructed into one set according to the decomposing scale, autocorrelation coefficients of all the reconstructed PRCs are calculated, the maximum correlation component is selected to construct a Hankel matrix and conduct enhanced multi-resolution singular value decomposition (MRSVD), and a corresponding approximate signal and a corresponding detail signal are obtained; and finally, the optimum precise component is selected to conduct Hilbert envelope transform, and the fault type is determined. On the one hand, by utilizing improved IITD and fusing the multi-channel signals, the signal-to-noise ratio is effectively increased, and noise is inhibited; and on the other hand, the signals are further refined and purified through enhanced MRSVD, accurate fault information is obtained, and the bearing fault type is determined through Hilbert envelope demodulation.

A rolling bearing fault diagnosis method based on integral inherent time scale decomposition algorithm comprises the following steps of collecting a vibration signal of a rolling bearing via a displacement sensor, decomposing the collected vibration signal via the integral inherent time scale decomposition algorithm to generate a plurality of rotational components and residual error signals, selecting sensitive rotational components capable of reflecting fault information from all the rotational components, conducting an envelope spectrum analysis on the sensitive rotational components and determining fault types according to envelope spectrum amplitude values corresponding to the fault feature frequency. A modal mixing problem of inherent time scale decomposition algorithm can be solved and great foundation is provided for feature extraction; according to peakedness calculation, rotational components sensitive to the fault are selected; and the fault type is determined via the analysis of the sensitive component envelope spectrum amplitude values corresponding to the fault feature frequency. Rolling bearing faults can be accurately identified and the method is suitable for rolling bearing fault diagnosis.

The invention discloses a pipeline defect identification method of suppression-end intrinsic time scale decomposition, which aims at solving the problem that the intrinsic time scale decomposition iseasy for generating an endpoint effect, adopts symmetric extension to process an extremum sequence, and adopts the intrinsic time scale decomposition to process the processed extremum to obtain an intrinsic rotation component. The method comprises the following steps: subtracting the endpoint extension from an original signal, obtaining an intrinsic rotation component, obtaining a novel extremum sequence, repeating the previous steps, obtaining a series of intrinsic rotation components and a single trend item, introducing the endpoint effect evaluation index theta to quantitatively analyze theendpoint effect, and selecting the intrinsic rotation component and the single trend item of a recombinant pipeline scalar magnetic signal by combining kurtosis determination. The recombinant scalarmagnetic signal and the intrinsic rotation component are enveloped, the gradient processing is performed for the enveloped signal, a pipeline deformation index is solved, the frequency spectrum analysis is performed for the recombinant scalar magnetic signal and the intrinsic rotation component, and a pipeline defect is determined by analyzing a gradient abnormal signal and a frequency spectrum analysis result.

The invention discloses an industrial multi-loop oscillation detection method based on multidimensional essential time scale decomposition. The method comprises steps that in to-be-detected control loops, process output signals of all the to-be-detected control loops are acquired; the output signals are projected along a series of projection directions to acquire a set of signals; multidimensional signal extreme points are solved, and multidimensional baseline node extraction of the extreme points is carried out; a multidimensional linear transformation formula is utilized to acquire linearity extraction results; the linearity extraction results are integrated to acquire a mean value estimate and decomposition subsignals of the output signals; stop conditions are determined, if the stop conditions are satisfied, decomposition stops; zero crossing point regularity indexes of the decomposition subsignals are calculated, oscillation degrees of the subsignals and whether the industrial process has plant-level oscillation are determined according to the indexes. The method is advantaged in that quantitative detection on multi-loop oscillation behaviors in the industrial process can be carried out, the rule degree and the period of each oscillation component can be further acquired, and abundant data supports are provided for oscillation behavior evaluation and fault source diagnosis.

The invention discloses an envelope analysis method based on intrinsic time scale decomposition filtering. The method includes that, firstly, the original signal is decomposed by the intrinsic time scale decomposition method, and then the noise component and the trend item in the decomposition result are eliminated by the data rearrangement and substitution operation, the signal after the primary filtering is analyzed by the spectral kurtosis method to obtain the center frequency and bandwidth of an optimal filter, the filter filters the signal after the primary filtering secondarily, the signal after the secondary filtering is subjected to envelope analysis by means of the rational spline iteration smoothing envelope analysis method, and at the end, the fault type of a rotary machine can be determined based on the spectrum envelope. The method is suitable for dealing with complex rotary machine fault signals, can accurately determine the fault type of the rotary machine, is good in noise resistance and robustness, and is convenient for engineering application.

The invention provides a method and system for optimizing operation of a new energy combined cooling heating and power system according to time scales, and the method comprises the steps: carrying out the time scale decomposition of a system based on an operation optimization model of the new energy combined cooling heating and power supply system, and dividing the system into a quick-change subsystem and a slow-change subsystem according to the difference of a state variable in a time scale; for the decomposed fast-varying subsystem and slow-varying subsystem, continuously updating the source and load input information of the system, carrying out rolling operation optimization on electricity and cold / heat energy in different time scales, and adjusting the real-time output of cold / heat supply equipment. Therefore, collaborative optimization of the electric energy and the cold / heat energy on different time scales is realized, and the requirements of the actual operation condition of the system can be better met.

The invention provides a fault travelling wave detection method comprising the steps of: S1, installing a travelling wave signal detection device in each transformer substation; S2, when a circuit hasa fault, detected signals are transmitted back to a system master station for signal processing and analysis; S3, firstly building a travelling wave signal monostable stochastic resonance model, andadjusting the parameters a and b and the sampling step size h in the travelling wave signal model so that the parameters, travelling wave signals and noise signals achieve a synergic state and energytransmitted to the travelling wave signals by noise is maximized, and effectively separating useful signals submerged in the noise signals; S4, decomposing the travelling wave signals into a pluralityof intrinsic rotary components and trend components by using intrinsic time scale decomposition, calculating the instantaneous frequency of each component, and extracting travelling wave signals, wherein the sudden change of the instantaneous frequencies embodies the sudden change of traveling wave signals and the sudden change moments of the instantaneous frequencies are the arrival moments of travelling wave signals. The method has the advantages of simple principle, high detection precision and easy implementation.

The invention discloses a heterogeneous dense oil deposit seepage flow time scale analysis method. An implementation method system, a modal analysis method, a production data output mode are included. The implementation method system includes a seepage flow model integral transformation method, a Laplace space solving method and an integral transformation retrieval method. The modal analysis method includes a method for converting a generalized eigenvalue problem into a sparse diagonal matrix eigenvalue problem, a method for obtaining a sparse diagonal matrix eigenvalue and an eigenvector, and a method for refining a generalized eigenvalue and an eigenvector; the production data output mode includes a time scale distribution diagram, and a production data time domain expression. Production data is decomposed according to the time scale, meanwhile, the production data contributions and mutual coverage relations of time scales are analyzed, the method has the advantages of being precise, reflecting dense oil deposit anisotropism, reflecting covered and developed regions and the like, and the method can be popularized and promoted in oil enterprises, scientific research institutions, colleges and universities in the field of oil-gas field development.

The invention discloses a circumference alarm system intrusion signal identification signal based on capacitance disturbance, and relates to the technology of capacitance disturbance signal processingand identification. The method comprises the following steps: firstly performing de-noising processing on an acquired capacitance disturbance signal; and then extracting a training sample feature value by using intrinsic time scale decomposition algorithm ITD and HT transform; and finally constructing a PNN probability neural network, and providing effective characteristic quantity for the PNN neural network to train, training so as to adjust parameters of the PNN probability neural network, and then providing the adjusted parameters for a probability neural network classification module to use. The method disclosed by the invention has the advantages of being high in identification precision, good in efficiency, high in sensitivity, and high in security.

The invention provides a knee osteoarthritis diagnosis system based on inherent time scale decomposition, phase space reconstruction and neural network. The system comprises a gait data acquisition module, a gait characteristic extraction module, a modeling module based on an RBF neural network, a training bait mode database, a dynamic estimator module and a diagnosis module. Based on the extracted bait dynamics data of knee joint angle and displacement, through inherent time scale decomposition, an effective rotation component signal is extracted. Phase space reconstruction is performed on the effective rotation component signal. An Euclidean distance is calculated. An effective gait characteristic variable is acquired. Neural network modeling and identification are dynamically performedon the gait system of a healthy normal person and a training person who suffers from the knee osteoarthritis. A normal value neural network is used for constructing a dynamic estimator, thereby realizing auxiliary diagnosis to the knee osteoarthritis. Compared with diagnosis facilities such as magnetic resonance imaging and arthroscope operation, the system has advantages of noninvasive operation,time saving, cost saving, etc.

The invention provides a cycle slip detecting and repairing method based on HHT and a support vector machine. First, an EMD algorithm is used for carrying out multi-time scale decomposition on a code phase second difference time sequence, then IMF components obtained by decomposition are subjected to Hilbert converting; second, according to whether the Hilbert amplitude values of the IMF components change suddenly at the same time, cycle slip is detected; then for the IMF components in which the Hilbert amplitude values change suddenly, support vector machine forecast models are established respectively, the value when cycle slip happens is forecast; finally the forecast value and a practical value are compared, and the magnitude of the cycle slip is determined. The method can be used for cycle slip detecting and repairing of single-frequency or multi-frequency carrier phase observing values, and the sensitivity of cycle slip detecting and the accuracy of cycle slip repairing are improved.

A diesel engine fuel oil system fault diagnosis method based on a least square support vector machine comprises the steps of: collecting vibration acceleration signals of a diesel engine under conditions of normal work and various kinds of faults; utilizing an inherent time scale decomposition algorithm to decompose the vibration acceleration signals, and generating a plurality of rotation components and residual error signals; calculating typical frequency domain characteristics of first N-order rotation components, and using the typical frequency domain characteristics as fault characteristics; dividing training samples and test samples; utilizing a hybrid algorithm of a difference evolution algorithm and a particle swarm algorithm to optimize a punishment factor and a kernel function parameter of the least square support vector machine, and obtaining an optimal punishment factor and an optimal kernel function parameter; and utilizing the obtained optimal punishment factor and optimal kernel function parameter to train the least square support vector machine for carrying out fault diagnosis. By adopting the method provided by the invention, the operation state of the fault diagnosis can be rapidly and accurately judged, and the method is applicable to online diagnosis of the diesel engine.

The invention relates to an elastomer aircraft global finite time control method based on time scale decomposition, belongs to the field of aircraft control, and is used for solving the problem that the elastic mode and pneumatic uncertainty shown by an elastomer aircraft. According to the method, firstly, based on the singular perturbation theory, rigid-elastic coupling flight dynamics is converted into a rigid body slow-varying subsystem and an elastic quick-varying subsystem; then, a global finite time control method based on a monitoring mechanism is designed for the rigid body subsystem, a finite time composite learning strategy is constructed when the flight state is in an intelligent working domain, a finite time robust control strategy is constructed when the flight state deviates from the intelligent working domain, and a finite time sliding mode control method is designed for the elastic subsystem. According to the method, self-adaptive dynamic adjustment and robust tracking of the system are realized, rapid and stable convergence of the elastic mode is ensured, and the method is of great significance to safe and reliable flight of the elastomer aircraft.