21 results about "FastICA" patented technology

The invention relates to a water turbine flow channel discharge noise monitoring and fault identification system and method, interference type optical fiber hydrophone arrays are arranged at a top cover between a rotating wheel and a movable guide vane, a volute mandoor and a draft tube straight cone section mandoor, discharge noise in a flow channel is directly measured, and after photoelectric conversion, signal amplification, analog-to-digital conversion and band-pass filtering, the discharge noise in the flow channel is detected. Blind source separation is carried out by using FastICA, features are extracted in combination with empirical mode decomposition (EEMD) and an energy method, audible sound and ultrasonic frequency bands are monitored at the same time, and feature type identification is carried out in combination with an existing fault feature database based on dynamic time warping (DTW) and a support vector machine (SVM) algorithm; the output signal-to-noise ratio and the anti-interference capability of the system are improved, multi-directional analysis of discharge noise characteristics is achieved, the monitoring range is wide, the fault recognition speed is high, the accuracy is high, and the system is mainly used for monitoring the running state of the water turbine in real time, finding and early warning faults in time and improving the running maintenance efficiency of the water turbine.

The invention discloses an automatic detection method for dynamic balance parameters of an automobile hub, and the method comprises the steps: employing a piezoelectric ceramic array self-sensing main shaft technology, enabling detection reference establishment and signal collection to be integrated in a main shaft body, and carrying out the fusion of a hub installation process and a detection reference establishment process; synchronous acquisition of multi-dimensional vibration signals is carried out; an original mixed signal is decomposed into modal components in a self-adaptive mode, and environmental vibration interference is eliminated; a FastICA-RLS hybrid algorithm is adopted to separate an independent source signal from the preliminarily purified signal, and non-stationary interference is tracked and compensated in real time; constructing a flexible rotor model based on modal parameter identification, and calculating an unbalance amount; self-adaptive Kalman filtering error compensation and prediction are carried out, and the amount of unbalance after compensation is output; and an optimal counterweight scheme is automatically calculated based on the compensated unbalance amount, and visual display and guidance are performed through a human-computer interface, so that the detection precision and the detection efficiency are improved.

The invention relates to an electroencephalogram signal preprocessing method based on particle swarm optimization (PSO), ensemble empirical mode decomposition (EEMD) and independent component analysis (ICA), and aims to improve the quality of electroencephalogram signals and enhance analyzability of the signals. The method comprises the following steps: firstly, carrying out EEMD (Ensemble Empirical Mode Decomposition) on an original electroencephalogram signal, generating a plurality of noise auxiliary signals by adding white noise with different intensities, carrying out EMD on each noise auxiliary signal, and then averaging intrinsic mode functions (IMF) of all the noise auxiliary signals to obtain a final IMF, thereby reducing the problem of mode aliasing and improving the reliability of the electroencephalogram signal. Useful components and noise components are preliminarily separated out; secondly, optimizing EEMD parameters by using a PSO algorithm; a particle swarm is initialized, each particle represents a possible parameter combination (such as Gaussian white noise standard deviation and noise adding times), a fitness function is defined, a quality index of a signal is taken as a target, positions and speeds of the particles are iteratively updated, the parameter combination is optimized, and finally optimal parameters are applied to perform EEMD decomposition, so that an optimized IMF is obtained. Therefore, the complexity of manual parameter adjustment is avoided, and the decomposition accuracy and stability are improved. Then, the sample entropy of each IMF is calculated, a sample entropy threshold value is set, the IMFs with the sample entropy values higher than the threshold value are screened out, IMF components with low information content are effectively removed, effective components with high information content are reserved, and the analyzability of the signals is further improved. And finally, combining the screened effective IMF component with the original electroencephalogram signal to generate a virtual multi-channel signal, performing Fast ICA (Independent Component Analysis), separating out independent electroencephalogram signal components, further removing noise, and improving the purity and the signal-to-noise ratio of the signal. Through the steps, the quality of the electroencephalogram signals can be remarkably improved, and a solid foundation is provided for subsequent signal analysis and application.

The invention discloses a graph signal blind source separation method in a white noise environment, and belongs to the technical field of signal processing. Aiming at the problem that the traditional blind source separation performance is reduced due to additive white noise, the invention provides a joint optimization scheme combining blind compression nonlinear noise suppression and image smoothing regularization. Firstly, a blind compression function is applied to a noisy observation image signal for preprocessing; secondly, carrying out mean value removal and whitening processing on the compressed signal; then, constructing a graph Laplacian matrix as a smoothing operator, establishing a joint diagonalization objective function fusing graph autocorrelation, a FastICA item and a graph smoothing regular item, and adopting a Givens rotation algorithm to iteratively optimize and solve a separation matrix; and finally, reconstructing a source signal through multiplication of the separation matrix and the whitening data. According to the method, noise interference is effectively suppressed through blind compression, signal structure consistency is maintained by using graph smoothing prior, and separation robustness, precision and convergence stability in a strong noise environment are remarkably improved.

The application particularly relates to a radar main lobe interference suppression method based on direction constraint, which realizes suppression of interference and separation of target echo in a main lobe suppression type interference scene by embedding array manifold vectors into FastICA fixed point iteration and applying direction constraint to the separation vectors after each update so that the separation vectors are always focused on a target angle subspace.

The application discloses a non-invasive load decomposition method based on CEEMDAN and FastICA, and belongs to the technical field of load monitoring. The method comprises the following steps: S1, collecting active power of total load and various single loads and preprocessing; S2, constructing a completely adaptive noise ensemble empirical mode decomposition (CEEMDAN) model, and decomposing total load power; S3, estimating the number of sources based on Bayesian information criterion; S4, performing dimension reduction by using maximum information coefficient (MIC); S5, performing load decomposition by using FastICA blind source separation; and S6, evaluating the approximation degree of decomposed signals and source signals. The application realizes load decomposition from the perspective of signal blind source separation, reduces the cumbersome load information feature extraction, and obtains complete load information through decomposition. Compared with deep learning, the model training time is greatly reduced.

The invention relates to the technical field of machining process parameter optimization, in particular to a surface quality dynamic optimization method for camshaft non-circular contour high-speed grinding. The method comprises the following steps: collecting a multi-source mixed signal in a camshaft non-circular contour high-speed grinding process; a signal processing method based on WPD-EWT-FastICA is adopted to separate out a source signal component corresponding to the state of the grinding machining process; a high-correlation feature set with high state correlation with the grinding machining process is optimized through ReliefF; dividing state labels; a recognition model based on CNN-LSTM-Attention is trained; a real-time multi-source mixed signal is obtained during machining, and a real-time grinding machining process state is output; and triggering a dynamic response mechanism, updating the multi-objective optimization function, and obtaining an optimal process parameter combination. According to the method, the optimal process parameter combination can be adaptively adjusted in real time, and the surface roughness, the residual stress and the material removal rate are dynamically optimized.

The invention discloses a DTWICA-based thought typewriting character signal template construction method, and relates to construction of a thought typewriting character signal template, which comprises the following steps of: firstly, constructing preprocessing, DTWICA and template construction on equipment, and visualizing and classifying aligned data; the electroencephalogram signals are sent into a DTWICA model, the model constructs a distortion function for the EEG signals of each trial through FastDTW, and signal decomposition is carried out based on FastICA to obtain time factors and neuron factors; carrying out mapping deformation on the decomposed time factor through a distortion function, and combining the time factor with a neuron factor to form a new EEG signal; then, a sliding time window is used for carrying out self-adaptive processing on the new EEG signal, and a deformed EEG signal is obtained; finally, averaging the deformed EEG signals of different test times to obtain each type of character signal template; according to the method, the performance of the template is verified in the imaginary sentence data set, and technical support is provided for real-time imaginary handwriting typing.

The application discloses a kind of sensor fault detection method and device of structural health monitoring system, it is related to structural health monitoring technical field, comprising: obtaining the nonlinear structure monitoring data of sensor system collected by structural health monitoring system, nonlinear structure monitoring data is handled using improved geometric PNL hybrid model, obtain linear to-be-separated mixed signal, using FastICA model to process to-be-separated mixed signal, obtain multiple independent elements, determine the separation matrix of improved geometric post-nonlinear independent component analysis model according to multiple independent elements, using the processing of improved geometric post-nonlinear independent component analysis to real-time collection nonlinear structure monitoring data, determine whether sensor system exists fault and the sensor of fault occurrence.The method can still complete linearization processing to sensor mixed signal under the condition that prior knowledge is unknown, compared with simple linear ICA analysis algorithm, it is more suitable for complex nonlinear structure.

The invention relates to the technical field of power system equipment fault diagnosis, in particular to a transformer substation GIS equipment partial discharge fault type identification method considering environmental interference compensation. The method comprises the steps of multi-source multi-mode signal acquisition, hybrid interference decoupling and preprocessing, physical information fusion dynamic interference compensation, fusion feature extraction, interpretable feature optimization and fault identification, and cloud edge collaborative closed-loop update full-process optimization, multi-mode signals are synchronously acquired through FPGA + edge nodes, and through preprocessing such as improved FastICA decoupling and db4 wavelet denoising, the cloud edge collaborative closed-loop update is obtained. The method comprises the following steps: constructing a fusion feature vector, carrying out dynamic compensation by using a random forest model, extracting 16-dimensional traditional features and 8-dimensional deep learning features, carrying out fusion, carrying out SHAP value analysis and genetic algorithm optimization, inputting into a transfer learning-PSO-ELM model for identification, and carrying out cloud edge collaboration to realize model updating. And the compensation signal distortion degree, the fault identification accuracy, the small sample generalization ability and the continuous operation accuracy are greatly improved.

Disclose a multiscale retinex and FASTICA combined infrared thermal image processing method, including the following steps: laser infrared thermal imaging; FASTICA image processing; multiscale Retinex image noise reduction; signal-to-noise ratio calculation.The FASTICA infrared thermal image processing method based on Retinex can enhance the infrared thermal image to display the key information of the image, and the effect of laser infrared nondestructive testing has obvious visual improvement, and other types of image optimization processing can also be carried out by this method.

This invention relates to a surface electromyography (SEMG) decomposition method based on FastICA and contractile force, belonging to the field of EEMG signal processing technology. First, a high-density array-type SEMG signal is pre-processed by filtering. Second, FastICA is used to decompose low-level contractile force EEMG signals, extracting the motor unit action potential sequence matrix from this level of muscle force signal. Then, FastICA is used to decompose high-level contractile force signals, and the extracted motor unit action potential sequence is subtracted from the action potential sequence matrix of the previous level. Finally, the above decomposition algorithm is iterated and optimized, with the constraint of minimizing the distance error between adjacent level potential sequence matrices. The improved decomposition method has high accuracy, is simple to implement, and can decompose a large number of motor units at medium to high contractile force levels.

The present application relates to the technical field of machining process parameter optimization, and particularly relates to a surface quality dynamic optimization method for camshaft non-circular contour high-speed grinding; the method comprises the following steps: collecting multi-source mixed signals in the camshaft non-circular contour high-speed grinding process; separating out source signal components corresponding to the grinding process state by using a signal processing method based on WPD-EWT-FastICA; selecting a high-correlation feature set with high correlation with the grinding process state by using ReliefF; dividing state labels; training a recognition model based on CNN-LSTM-Attention; acquiring real-time multi-source mixed signals during processing, and outputting real-time grinding process states; triggering a dynamic response mechanism, updating a multi-objective optimization function, and acquiring an optimal process parameter combination; the method can adaptively adjust the optimal process parameter combination in real time, and dynamically optimize surface roughness, residual stress and material removal rate.

The present invention provides a damage clustering reliability assessment method for single-lap adhesive joints of composite materials based on acoustic emission, comprising the following steps: 1. performing a tensile test to obtain acoustic emission signals; 2. proposing a noise separation method combining variational mode decomposition (VMD) and fast independent component analysis (FastICA) to remove environmental noise and invalid components; 3. proposing a statistically based credibility assessment method based on the characteristic parameters of the acoustic emission signal in the time and frequency domains, i.e., using the Pearson correlation coefficient to quantify neighborhood changes before and after signal dimensionality reduction, and evaluating the credibility of the event based on a credibility threshold; 4. identifying three types of damage, namely matrix cracking, adhesive debonding, and fiber breakage, based on a clustering algorithm, and analyzing their development mechanisms. The present invention innovatively combines VMD and FastICA for noise separation and proposes a cluster credibility assessment index, effectively improving the accuracy and reliability of acoustic emission data analysis, and providing technical support for damage assessment and health monitoring of single-lap adhesive joints of composite materials and other complex structures.

The invention relates to an acoustic feature enhancement method and device based on blind source separation and two-stage filtering and a medium, and the method comprises the steps: carrying out the FastICA blind source separation of a collected acoustic signal, and obtaining a plurality of separated sub-signals; the amplitude and the phase of each obtained sub-signal are corrected; the corrected sub-signals are screened through the energy ratio, and effective source signals are obtained; and filtering the effective source signal by adopting a self-adaptive two-stage filtering mode to obtain a target source signal. According to the invention, the target source signal can be effectively separated, the background noise is reduced, and the acoustic characteristics are obviously enhanced, thereby improving the effectiveness of acoustic fault recognition.

The application provides an oil and gas pipeline pulse eddy current response signal denoising method based on FastICA, which comprises the following steps: collecting signals generated during pulse eddy current detection of an oil and gas pipeline to obtain original signal data, wherein the original signal data comprises an oil and gas pipeline pulse eddy current response signal and electromagnetic noise; using an FIR adaptive filtering algorithm to pre-process the collected original signal data to obtain pre-processed signal data; using a MEMD algorithm to decompose the pre-processed signal data, and screening the decomposed components to obtain a final decomposition result; using an MDL criterion to estimate the total number of signal sources of the pre-processed signal data; and separating the oil and gas pipeline pulse eddy current response signal and the electromagnetic noise from the final decomposition result according to the estimated total number of signal sources and using a FastICA independent component analysis algorithm. The application adopting the above scheme can significantly improve the detection accuracy and provide more reliable technical support for the safety monitoring of the oil and gas pipeline.

This invention belongs to the field of online detection technology for metal impurities in food, and particularly relates to an intelligent sensing-based online detection method and system for metal impurities in food. It simultaneously transmits and receives multiple discrete frequency band electromagnetic signals, generating a multi-dimensional original signal matrix by leveraging the differences in their responses to metals, food, packaging, and the environment. Based on the fusion application of an improved FastICA blind source separation algorithm and wavelet packet transform, a multi-frequency feature-blind source separation dual-layer model is constructed. The effectiveness of signal separation is verified using a multi-scenario standard signal feature library and cosine similarity matching. After feature extraction and combined dimensionality reduction processing, the output is fused through a hybrid model of traditional machine learning and lightweight deep learning. The D-S evidence theory is introduced and combined with real-time parameters from the production line to dynamically adjust the decision rules.

The present invention belongs to the field of image and video processing technology, and specifically relates to a method for detecting moving targets based on fast independent component analysis combined with spatiotemporal differences, comprising: combining FastICA with a frame difference method, and obtaining a contour F by performing FastICA processing on the difference in the time domain between the current frame and the adjacent frames. R1 ; Send the video sequence to the Gaussian mixture model to generate the background image and the current frame to perform FastICA processing on the spatial domain to obtain the contour F R2 ; Through the fusion strategy, the contour F of the target in the time domain is transformed R1 And the target profile F in the spatial domain R2 The fused contours are morphologically processed to obtain accurate moving targets. This invention combines the temporal and spatial differences in the detection scene through fast independent component analysis, effectively making up for the shortcomings of the two methods in various fields, effectively avoiding these problems, and effectively improving the accuracy and effectiveness of the method.

The application provides a single-phase tree grounding fault detection method, comprising the following steps: collecting initial zero sequence voltage data in a power distribution network; screening the waveform amplitude of the initial zero sequence voltage data to obtain target zero sequence voltage with sudden change in waveform amplitude; judging whether the target zero sequence voltage has beat frequency characteristics; if yes, it is determined that the target zero sequence voltage has a first fault; otherwise, the next step is entered; separating the target zero sequence voltage by using a fixed point algorithm, and judging whether the separated zero sequence voltage has beat frequency characteristics; if yes, it is determined that the target zero sequence voltage has a second fault. The application also provides a single-phase tree grounding fault detection device. The single-phase tree grounding fault detection is based on the fixed point algorithm FastICA, so that the recognition speed is improved while the recognition ability is ensured, and the detection efficiency is effectively improved.

The present invention discloses a power grid disturbance identification method and system based on improved FastICA. The method includes: setting power grid disturbance types of different types and at different bus positions based on the structure of the target power grid and the parameters of its various components, respectively simulating them to obtain the change data of the voltage amplitudes of each bus in the target power grid under different disturbance types, that is, disturbance data; using the improved fast independent component analysis algorithm to denoise the disturbance data; extracting the features of the denoised disturbance data; generating a disturbance identification sample data set from the extracted features; constructing a power grid disturbance identification model; training the power grid disturbance identification model through the disturbance identification sample data set to obtain a trained power grid disturbance identification model; and using the trained power grid disturbance identification model to identify power grid disturbances and outputting a disturbance identification result including the disturbance type and the disturbance occurrence position. The present invention has advantages such as good denoising effect and high disturbance identification accuracy.

This invention discloses a single-channel mixed signal demodulation method based on oversampling. First, the mixed signal of BPSK and 16QAM signals received in a PCMA communication system is processed to construct a multi-channel blind source separation model. The signal from the multi-channel blind source separation model is then centered and whitened to reduce the computational cost of FastICA. Then, FastICA is used to extract independent components. Finally, based on the obtained independent components, a phase-locked loop is used to extract baseband information to obtain the baseband signal. This invention effectively solves the problem of high bit error rate after separating and demodulating mixed signals in a single-channel environment. This invention utilizes the orthogonality of carriers, which further improves the demodulation performance of single-channel mixed signals when the sampling factor is increased. Compared with time-domain filtering methods, frequency-domain notch filtering methods, empirical mode decomposition methods, and wavelet decomposition methods, this method has better demodulation performance and a lower bit error rate.