13 results about "Bayesian compressive sensing" patented technology

A complex value deconvolution DOA estimation method based on block sparse BCS depth expansion comprises the following steps: constructing an array receiving signal model under single snapshot through a compressed sensing technology according to a sparse signal and an over-complete steering vector; constructing a complex value beam model according to the array receiving signal model under the single snapshot by using a traditional beam forming algorithm; constructing a coupling layered Gaussian prior Bayesian compressed sensing model according to complex output of the complex value beam model by using a Bayesian compressed sensing technology; building a block sparse Bayesian compressed sensing complex value deconvolution model to realize accurate recovery of block sparse signals; constructing a training data set, wherein the training data set comprises a real information source distribution function serving as a label and complex value beam forming output which corresponds to the label and serves as a sample; training a loss function by a deep expansion network based on block sparse Bayesian compressed sensing; inputting the training data set into the deep expansion network for model training; and inputting test data into the trained deep expansion network to generate direction-of-arrival estimation.

The application relates to the technical field of aero-engine aerodynamic acoustic detection, and discloses a pipe-in-sound mode integrated detection method considering abnormal gain of a microphone, which comprises the following steps: a microphone array unit, a rotating speed detection unit and a signal acquisition module are built to realize data acquisition; a core algorithm unit is built, a joint sparse model is constructed based on a blade passing frequency calculated from rotating speed data, and super parameter adaptive updating is completed through an EM algorithm until the super parameter meets a convergence condition; a mode coefficient matrix and an abnormal gain matrix are extracted based on the converged super parameter, elements of the abnormal gain matrix are arranged in ascending order, and the position of an abnormal microphone is determined according to a preset threshold; subsequently, channel data corresponding to the abnormal microphone is removed, and the mode coefficient is re-inverted by using a Bayesian compressive sensing method; and a standardized detection report is output by the system. The method realizes high-precision mode recognition and abnormal detection, does not need manual intervention, and improves detection efficiency and robustness.

A kind of unmanned aerial vehicle auxiliary sparse crowd-sensing-oriented collaborative sensing method and device, method includes: using unmanned aerial vehicle to carry out sensing data collection to the sub-region in the sensing area, the participant is evaluated for sensing quality, obtain first sensing quality;Using the participant who has been evaluated to carry out sensing data collection, the participant who has not been evaluated is evaluated for sensing quality, obtain second sensing quality and participant dependency;Using the sensing data collected by the participant who has been evaluated, the participant who has been evaluated is evaluated for sensing quality, according to participant dependency, update sensing quality, obtain participant sensing quality;Data weighting fusion is carried out to sensing data, obtain input data;Using bayesian compressive sensing algorithm to calculate input data, determine the sensing data corresponding to the sub-region not sensed.The present application improves the accuracy of prediction result, reduces the influence of real physical world participant low-quality data on reasoning algorithm.

The invention discloses a Bluetooth earphone signal interference suppression system based on a graph neural network, and the system comprises a data collection and processing module which is used for collecting operation data and carrying out the normalization processing; the interference topology construction module is used for constructing an interference topological graph according to the operation data; the graph neural network inference module is used for generating an interference characterization vector and an interference correlation characterization vector; the priori generation module is used for constructing a priori vector and generating a corresponding hyper-parameter set; the frequency hopping consistency construction module is used for constructing a frequency hopping consistency constraint; the signal reconstruction module is used for executing sparse Bayesian compressed sensing to solve and reconstruct a Bluetooth signal; the communication parameter generation module is used for generating a frequency hopping sequence adjustment parameter, a channel selection parameter, a transmitting power parameter and a coding parameter; and the link adjusting module is used for adjusting the wireless communication link of the Bluetooth earphone in real time and maintaining stable transmission. According to the invention, interference is suppressed by using the graph neural network and sparse compressed sensing, and the Bluetooth signal reconstruction and transmission stability is improved.

The application discloses a non-invasive chaos polynomial casing assembly structure uncertainty analysis method, and belongs to the field of uncertainty analysis of casing assembly structures. The method comprises the following steps: step SS1: casing component dynamics modeling; step SS2: casing assembly structure dynamics modeling; step SS3: acquiring the probability distribution model of the fitting size and the tightening torque according to the experimental sample; step SS4: selecting the chaos polynomial order P, and generating the chaos polynomial expansion term; step SS5: calculating the coefficients of each chaos polynomial expansion term by adopting the Bayesian compressive sensing method; step SS6: determining the variation range of each order modal frequency of the casing assembly structure based on the established non-invasive chaos polynomial model; and step SS7: if the determination result does not converge, then P=P+1, and the step SS4 is executed; if the determination result converges, then the result is outputted and the method is ended. The method is used for the uncertainty analysis of each order modal frequency of the casing assembly structure, and the calculation efficiency is improved while the analysis precision is ensured.

The application discloses a liquid crystal phased array fast-response bistable driving and low-power scanning method, equipment and medium, and relates to the technical field of liquid crystal phased array.The application compares the deviation matrix generated by real-time acquisition of liquid crystal molecule response data with the digital twin theory value, combines the deep reinforcement learning to output the electric field parameter, generates the non-uniform composite electric field through the reverse design of topological photonics, realizes the high-precision bistable steering of molecules, reconstructs the adaptive sparse scanning path based on the Bayesian compressed sensing, optimizes the energy distribution by matching the dynamic energy spectrum shaping strategy, predicts the molecule steering state through the physical information neural network embedded with physical constraints, generates the pre-control driving sequence combined with the event triggering mechanism, and outputs the optimal control instruction through the Lyapunov optimization, so that the steering response speed and stability of the liquid crystal phased array are effectively improved, and the energy utilization efficiency is greatly improved.

The invention discloses an energy storage battery cabin thermal runaway early warning method and system based on adaptive compressed sensing, and belongs to the technical field of thermal management and safety early warning of energy storage battery cabins. The method comprises the following steps: acquiring electrical state data and temperature state data of each battery unit in the energy storage battery cabin in real time; constructing a dynamic risk map of the battery compartment based on the electrical state data and the temperature state data, and obtaining a risk value of each region; based on the dynamic risk map, adopting a Bayesian compressed sensing framework to dynamically adjust a sensing matrix of infrared imaging, and implementing differentiated compressed sampling on the battery compartment to obtain compressed measurement data; and inputting the compression measurement data and the risk value of each region into a lightweight recurrent neural network, outputting a thermal runaway risk score, and executing a corresponding early warning operation based on the thermal runaway risk score. The overall thermal safety protection capability of the energy storage battery cabin is comprehensively improved.

The application provides a message passing based multi-task clustering sparse reconstruction method. The method uses the joint clustering sparse structure characteristics of sparse signals among different tasks, so that better sparse reconstruction performance is obtained under the condition of fewer observation samples. Specifically, the sparse structure characteristics of the clustering sparse signal are described by using a Markov Spike and Slab prior, a generalized approximate message passing algorithm is introduced to iteratively approximate the posterior mean of each unknown variable, and an expectation-maximization method is used to iteratively update the unknown parameters. Compared with the traditional single-task Bayesian compressive sensing algorithm, the method has a significant performance improvement under the condition of fewer observation samples.

The application relates to the technical field of geotechnical engineering and foundation pit supporting design, and particularly discloses a system and method for predicting spatiotemporal evolution of foundation pit deformation and reversely optimizing supporting parameters, which comprises the following steps: collecting sparse borehole geotechnical parameters and in-situ test data to construct a prior random field; adopting Bayesian compressed sensing to invert parameter space full-grid mean values and covariance matrices under posterior probability distribution; generating multiple groups of random fields to realize and input finite element simulation, forming a training sample set of parameter distribution and deformation field; constructing a deep Gaussian process nonlinear mapping and quantifying prediction variance; and reversely searching supporting parameter combinations by taking deformation control threshold and allowable overrun probability as constraints. The application can quantitatively influence the variability of geological parameter space, and obtain a robust supporting scheme meeting overrun probability requirements.

The complex-valued deconvolution DOA estimation method based on block sparse BCS deep unfolding includes: constructing an array receiving signal model under a single snapshot according to a sparse signal and a super-complete steering vector through a compressed sensing technology; constructing a complex-valued beam model according to the array receiving signal model under the single snapshot by using a traditional beam forming algorithm; constructing a coupled hierarchical Gaussian prior Bayesian compressed sensing model according to a complex number output of the complex-valued beam model by using a Bayesian compressed sensing technology; constructing a block sparse Bayesian compressed sensing complex-valued deconvolution model to realize accurate recovery of a block sparse signal; constructing a training data set, including a real source distribution function as a label and a complex-valued beam forming output as a sample corresponding to the label; a training loss function of a deep unfolding network based on a block sparse Bayesian compressed sensing; inputting the training data set into the deep unfolding network for model training; and inputting test data into the trained deep unfolding network to generate a direction of arrival estimation.

This invention discloses a high-precision multi-station target detection method that integrates time delay and frequency shift, relating to the field of signal processing technology. The implementation process is as follows: (1) Divide the space into subspaces uniformly according to the characteristics of the measurement space; (2) Calculate the theoretical time delay and theoretical Doppler frequency shift of targets at different speeds at the center position of each subspace relative to the radar, and save them as a dictionary matrix; (3) Utilize the time delay and Doppler frequency shift of moving targets in space relative to each radar, and obtain the possible weights of the targets in each subspace through a multi-task Bayesian compressed sensing algorithm; (4) Determine the location of the target based on the maximum weight, and detect and locate the target. Simulation data and experimental results demonstrate the effectiveness and superiority of the target detection method of this invention. The method of this invention uses time delay and Doppler frequency shift features to jointly estimate the target position. Simulation results show that it is better than the commonly used time delay-only estimation method, and can be used for space exploration of UAVs, aircraft, etc.

The application discloses a kind of unmanned plane target coherent accumulation detection methods based on bayesian compressive sensing, comprising: S1, the radar echo of unmanned plane is compressed in distance, and micro-doppler signal is separated using zero space tracking algorithm, obtain the radar echo of unmanned plane;S2, based on the obtained radar echo of unmanned plane, the distance of the envelope of unmanned plane is corrected using Keystone conversion to walk, and the slow-time data of the position of distance envelope peak value is extracted as the observation signal of sparse reconstruction by iteration;S3, design sensing matrix, and the observation signal is reconstructed using bayesian compressive sensing algorithm to estimate the motion parameter of each order of unmanned plane;S4, according to the motion parameter of each order of unmanned plane, construct motion phase compensation filter, the echo of unmanned plane is phase compensated in pulse-range frequency domain, and the frequency domain echo signal after compensation is coherent accumulation and is detected by CFAR.The application has the advantages of high parameter estimation accuracy, low computational complexity, strong robustness and the like.