Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

70 results about "Gaussian sampling" patented technology

The Gaussian sampling strategy follows some of the same motivation for sampling on the boundary. In this case, the goal is to obtain points near by using a Gaussian distribution that biases the samples to be closer to , but the bias is gentler, as prescribed by the variance parameter of the Gaussian.

Rapid inversion method for Ia type super-novelty star spectral parameters based on KAN (Karan Area Network)

The invention discloses an Ia type super-novelty star spectral parameter rapid inversion method based on a KAN network, and the method comprises the following steps: S1, carrying out data preprocessing, generating simulation spectral data, including spectral lines and parameters, and carrying out the normalization; s2, building a KAN (Kernel Area Network) network model by using a python framework and a pytorch framework; s3, designing a loss function; s4, training the model, and testing the error of each parameter in the stage; s5, performing learning rate adjustment by using an Adam optimizer and a StepLR scheduler in the training process, and optimizing the model; s6, performing Gaussian sampling on an actually measured spectrum, and inputting the actually measured spectrum into the trained KAN model for prediction; s7, analyzing the result, and calculating the standard deviation of the output parameter data; by using the deep learning technology of the KAN network, key physical and chemical parameters are efficiently predicted from the celestial body spectrum, the problems that in the prior art, the calculation amount is too large, and an accurate calculation result cannot be efficiently obtained are solved, and the method has remarkable advantages in the aspects of improving the calculation efficiency and the processing capacity and is suitable for spectrum data analysis in the field of astronomical observation.
Owner:CHINA THREE GORGES UNIV

Ultrasonic enhanced robot obstacle avoidance path planning method and device

The invention discloses an ultrasonic enhanced robot obstacle avoidance path planning method and device, and belongs to the field of robot simulation. The method comprises the following steps: constructing an industrial robot model and a complex numerical control machining environment model; taking the starting point and the ending point as nodes on a tree, and gradually expanding new nodes on the tree from the starting point according to a self-adaptive expansion step length by using the industrial robot model and the processing environment model so as to obtain a collision-free path; in the process of expanding the new node, the expansion step length is adaptively updated according to the crowdedness degree of the space environment; and utilizing a Gaussian sampling strategy to carry out multiple times of iterative optimization on the collision-free path so as to obtain an optimal collision-free path. According to the invention, the path planning efficiency can be improved and collision can be reduced.
Owner:EAST CHINA UNIV OF SCI & TECH

Machine learning driven frame lightweight design method

The invention discloses a machine learning driven vehicle frame lightweight design method, which comprises the following steps: (1) performing vehicle frame parametric modeling in CATIA, constructing a vehicle frame modal analysis model and a stiffness analysis model in ABAQUS, and deducing a mathematical model of vehicle frame lightweight design; (2) performing Latin hypercube sampling and discrete Gaussian sampling on continuous and discrete parameters to generate a frame population, and establishing a database; (3) designing a variable type driven hybrid variation mechanism to obtain a candidate frame population; (4) constructing a radial basis function machine learning model, and deducing a comprehensive minimum statistical lower limit function to screen an optimal candidate frame; and (5) performing parametric modeling and simulation analysis on the optimal candidate frame, updating the frame population and the database, returning to the step (3) until the optimal frame meets the design requirement, and outputting the optimal frame. According to the method, adaptive variation strategies and screening functions are designed according to the vehicle frame parameter characteristics, and the vehicle frame lightweight design effect is good.
Owner:JIANGLING MOTORS

Method for accelerating point cloud registration based on 3DSC and improved ICP algorithm

The invention discloses a method for accelerating point cloud registration based on 3DSC and an improved ICP algorithm, and the method comprises the following steps: S1, reducing the data volume of a point cloud through the combination of voxel filtering and Gaussian sampling, guaranteeing the integrity of a point cloud structure, and improving the processing speed of the point cloud; and S2, performing feature descriptor extraction and feature descriptor matching by using a 3DSC algorithm. And S3, optimizing parameters such as neighbor search radius and point cloud density according to point cloud characteristics to obtain better matrix estimation. And S4, transmitting the coarse registration matrix estimation to an NICP algorithm, initializing an initial homogeneous transformation matrix of the NICP algorithm, and improving the registration precision. And S5, in the NICP algorithm, using the KD-Tree to accelerate the nearest neighbor search and point cloud matching process, and reducing the calculation time. And S6, matrix optimization is carried out in combination with an ICP improved algorithm, namely an NICP fine registration algorithm.
Owner:滨沅国科(秦皇岛)智能科技股份有限公司

Sample construction method and system for uncertainty analysis and storage medium

The invention discloses a sample construction method and system for uncertainty analysis and a storage medium, and relates to the field of nuclear reactor analogue simulation, and the method comprises the steps: determining an average value matrix and a covariance matrix corresponding to measurement data based on the measurement data of each input parameter in a numerical reactor, the input parameters are associated with a proxy model of the numerical reactor; constructing a Latin hypercube sampling matrix, and mapping the Latin hypercube sampling matrix to multivariate Gaussian distribution based on the average value matrix and the covariance matrix to obtain multivariate Gaussian sampling data; determining Shannon diversity index data corresponding to the multivariate Gaussian sampling data, and determining a sample minimum distance corresponding to the multivariate Gaussian sampling data; and determining an input sample of the proxy model based on the Shannon diversity index data and the sample minimum distance. According to the sample construction method for uncertainty analysis, the input parameters can be uniformly distributed while the distribution of the input parameters is accurately determined.
Owner:CHINA NUCLEAR POWER TECH RES INST CO LTD +1

A dynamic adaptive unmanned aerial vehicle path planning method based on RRT*

The application relates to a dynamic self-adaptive unmanned aerial vehicle path planning method based on RRT*, which solves the technical problems that the existing unmanned aerial vehicle path planning method based on the RRT* algorithm has slow convergence speed, a high degree of path tortuosity, poor smoothness and to-be-improved efficiency; the method generates sampling points through adaptive Gaussian sampling according to adaptive target bias probability, uses an intelligent variable step length mechanism to use different step lengths in different exploration periods, and further obtains an optimized path by introducing Pareto evaluation of the degree of path tortuosity. The path planned by the application has fast convergence speed, low degree of path tortuosity and good smoothness.
Owner:HARBIN INST OF TECH AT WEIHAI

Method for predicting residual life of wind power rolling bearing by applying incremental mapping and recursive estimation

The invention discloses a method for predicting the residual life of a wind power rolling bearing by applying incremental mapping and recursive estimation. The invention discloses a method for predicting the residual life of a wind power rolling bearing, and relates to renewable energy equipment state monitoring. The method comprises the following steps: firstly, obtaining a multiband envelope demodulation two-dimensional spectrogram by converting a vibration acceleration signal, calculating and normalizing the variation of the spectrogram, and obtaining a health index and a normalized degradation increment; secondly, extracting features from two dimensions of the two-dimensional spectrogram by using a channel attention module, and designing nonlinear mapping of a two-dimensional convolutional neural network learning spectrogram and a degradation increment; and then, establishing a recursive deep learning model, and calculating a posterior probability of a future health index in combination with model weight Gaussian sampling to realize rolling prediction. And finally, predicting the residual life according to the failure threshold. The method has the advantages that the health index based on the change increment of the multiband envelope demodulation two-dimensional spectrogram is constructed, the two-dimensional convolutional neural network containing the channel attention module is designed, recursive deep learning and weight Gaussian sampling are fused, and the prediction precision of the residual life of the wind power bearing is improved.
Owner:GUANGDONG HUADIAN FUXIN YANGJIANG OFFSHORE WIND POWER CO LTD +1

Grid-based public key prefix broadcast searchable encryption method, apparatus and device, and medium

The invention relates to a lattice-based public key prefix broadcast searchable encryption method and device, equipment and a medium, and the method comprises the steps: a sender integrates public keys of all target receivers to obtain an aggregated public key, carries out the hash processing of each to-be-encrypted keyword field in sequence, so as to construct a corresponding keyword vector, completes the extension, and carries out the encryption of the keyword vector; generating a broadcast ciphertext and uploading the broadcast ciphertext to a cloud server; the receiver calls a lattice base sampling algorithm to generate a private key lattice base corresponding to the aggregated public key based on the private key of the receiver, the aggregated public key and Gaussian sampling parameters in system public parameters; the receiver generates a search trap door according to the to-be-searched keyword information with the prefix search, and sends the search trap door to the cloud server; and the cloud server calculates a verification value according to the broadcast ciphertext and the search trap door, and returns a corresponding search result to the target receiver according to the verification value. According to the invention, sharp increase of storage overhead in a large-scale user scene can be avoided.
Owner:SOUTH CHINA AGRICULTURAL UNIVERSITY

Self-calibration self-adaptive sampling system and method for Gaussian sampling and matrix inversion

The invention relates to the technical field of probability calculation and mixed signal coprocessors, in particular to a self-calibration self-adaptive sampling system and method for Gaussian sampling and matrix inversion, and the technical scheme is characterized by comprising the steps of 1, task definition and input modeling, 2, matrix compilation mapping and discrete quantization, and 3, matrix compilation mapping and discrete quantization. The method comprises the following steps of 1, carrying out controllable noise injection and effective temperature setting, 4, carrying out on-line self-calibration and error compensation closed loop, 5, carrying out adaptive equalization / sampling control and stop criterion setting, and 6, outputting a result and carrying out system deployment. According to the method, the advantages of high throughput and low energy consumption of thermodynamic calculation are kept, and meanwhile, through online self-calibration and self-adaptive sampling control, controllable output distribution precision, verifiable errors and deployable and extensible Gaussian sampling and matrix inversion operators are achieved.
Owner:YISI GYROMAGNETIC (JIAXING) ELECTRONICS CO LTD

A method, apparatus, device, and storage medium for training a noise reduction neural network.

This application belongs to the field of signal denoising technology and discloses a training method, apparatus, device, and storage medium for a denoising neural network. The method includes the following steps: acquiring a noisy signal and inputting it into an initial neural network to obtain a sampling sequence; performing Gaussian sampling on the sampling sequence to obtain a noise signal; obtaining the original signal based on the noisy signal and the noise signal; calculating the error of the initial neural network based on the original signal; determining whether the error is less than a preset error threshold; if not, returning to the training step; if yes, ending the training and using the initial neural network as the denoising neural network. This application can obtain a denoising neural network without acquiring a clean signal, realizing unsupervised learning and having extremely high applicability in engineering applications.
Owner:广芯微电子(广州)股份有限公司

Financial product recommendation method, system and equipment based on semi-Gaussian sampling and medium

The invention relates to the technical field of time sequence prediction, and provides a financial product recommendation method, system and device based on semi-Gaussian sampling and a medium. The financial product recommendation method based on semi-Gaussian sampling comprises the steps of obtaining historical attribute data and historical popularity data of various financial products, performing preprocessing, constructing a sample set, dividing the sample set into a training set and a test set, mapping a sample sequence number in the training set into an x-axis interval [mu-3sigma, mu]] of semi-Gaussian distribution, calculating the weight of each sample, and converting the weight into a probability value; performing polynomial distribution by utilizing the probability value to extract a plurality of samples with replacement; and training the dynamic time convolution network based on the plurality of extracted samples and the true value corresponding to each sample to obtain a trained dynamic time convolution network for predicting the ranking of all financial products, and adding a plurality of financial products with the top ranking to recommend to a user. According to the invention, a more accurate recommendation result can be generated.
Owner:UNIV OF JINAN

GRU quantification method based on Gaussian distribution, storage medium and intelligent device

The invention provides a GRU quantification method based on Gaussian distribution, a storage medium and intelligent equipment. The GRU quantification method comprises the following steps: acquiring training data input into a GRU model, and calculating to obtain a variance of the training data; performing Gaussian sampling according to the variance of the training data to obtain a Gaussian sample set; sorting the elements in the Gaussian sample set; performing definition domain reconstruction on the activation function of the GRU model to limit the definition domain of the activation function of the GRU model to the ranked Gaussian sample set; performing interpolation calculation on the activation function after the definition domain reconstruction to obtain a corresponding value domain list; mapping elements in the value domain list to a fixed-point value domain list to complete quantization of the GRU model; and calculating a reset gate, an update gate, a candidate hidden state and a hidden state based on the quantized GRU model. On the basis, the GRU model can be suitable for non-fixed input data, and the calculation precision is high.
Owner:BESTECHNIC SHANGHAI CO LTD

An internet of things identity authentication method based on lightweight Falcon signature

The application discloses an Internet of Things identity authentication method based on a lightweight Falcon signature, and the Internet of Things comprises a device, an edge gateway and a cloud platform, and comprises the following steps: initializing the device to generate a Falcon-512 key pair, performing sparse compression on the Falcon-512 key pair, and performing fragmented encryption storage; the Falcon-512 key pair comprises a private key sk and a public key pk; sending a registration request CSR to a cloud certificate authority CA through the device to obtain a device certificate Cert; collecting communication data Data through the device, calculating a hash value H(Data||Timestamp) of the communication data Data, wherein Timestamp is a time stamp; inputting the hash value H into an iterative FFT, mapping and storing the hash value H through a plurality of butterfly operations in the iterative FFT to obtain an intermediate result C in the FFT; wherein, based on the independence of the butterfly operation in the iterative FFT, a plurality of butterfly operations are processed in parallel through the SIMD instruction of the ARM Cortex-M processor; and a random integer z conforming to a discrete Gaussian distribution is generated through Gaussian sampling.
Owner:CHENGDU UNIVERSITY OF TECHNOLOGY

Delay prediction method and device, electronic device, and storage medium

The disclosed embodiments relate to a method and apparatus for delay prediction, an electronic device, and a storage medium, and relate to the field of computer technology. The delay prediction method includes: performing mixed Gaussian sampling on a target model corresponding to a target operation to obtain multiple subnet network structures of the target model; performing a convolution operation on the multiple subnet network structures to perform delay prediction, determining a predicted result of the target model's delay information, and performing the target operation on a processing object based on the predicted result. The technical solutions in the disclosed embodiments can improve the prediction results of the target model's delay information and achieve universality.
Owner:GUANGDONG OPPO MOBILE TELECOMMUNICATIONS CORP LTD

Antenna signal decoding method and device

The embodiment of the invention provides an antenna signal decoding method and device, and the method comprises the steps: carrying out the channel estimation of a received signal, obtaining a channel estimation matrix, and enabling the received signal to be an uplink signal, received by a plurality of receiving antennas, of a plurality of transmitting antennas; determining an adjustment parameter which enables lattice Gaussian sampling probability search to converge; according to the adjustment parameter and the channel estimation matrix, carrying out lattice Gaussian sampling probability-based search decoding on the received signal to obtain a candidate vector set containing a plurality of candidate vectors; and determining a decoding result of the receiving antenna according to a plurality of candidate vectors in the candidate vector set. The problem that antenna signal decoding cannot meet high performance and low calculation complexity at the same time in the prior art can be solved at least.
Owner:ZTE CORP

Internet of Things identity authentication method based on lightweight Falcon signature

The invention discloses an Internet of Things identity authentication method based on a lightweight Falcon signature, the Internet of Things comprises a device, an edge gateway and a cloud platform, the method comprises the following steps: an initialization device generates a Falcon-512 key pair, the Falcon-512 key pair is subjected to sparse compression and fragmented encryption storage, and the Falcon-512 key pair comprises a private key sk and a public key pk; sending a registration request CSR to a cloud certificate authority CA through the device to obtain a device certificate Cert; the method comprises the steps that communication data are collected through the equipment, the hash value H (DataTimestamp) of the communication data is calculated, and the Timestamp is a timestamp; inputting the Hash value H into an iterative FFT (Fast Fourier Transform), and mapping and storing the Hash value H through a plurality of butterfly operations in the iterative FFT to obtain an FFT intermediate result C; wherein on the basis of independence of butterfly operations in the iterative FFT, a plurality of butterfly operations are processed in parallel through an SIMD instruction of the ARM Cortex-M processor; and generating a random integer z conforming to discrete Gaussian distribution through Gaussian sampling.
Owner:CHENGDU UNIVERSITY OF TECHNOLOGY

A method for predicting the residual life of a wind power rolling bearing by using incremental mapping and recursive estimation

The application discloses a method for predicting the residual life of wind power rolling bearings by using incremental mapping and recursive estimation, and relates to renewable energy equipment state monitoring. First, a multi-band envelope demodulation two-dimensional spectrum diagram is obtained by transforming a vibration acceleration signal, a spectrum diagram variation is calculated and normalized, a health index and a normalized degradation increment are obtained. Second, a channel attention module is used to extract features from two dimensions of the two-dimensional spectrum diagram, and a two-dimensional convolutional neural network is designed to learn the nonlinear mapping of the spectrum diagram and the degradation increment. Third, a recursive deep learning model is established, model weight Gaussian sampling is combined, the posterior probability of future health indexes is calculated, and rolling prediction is realized. Finally, the residual life is predicted according to a failure threshold. The application has the advantages that a health index based on a multi-band envelope demodulation two-dimensional spectrum diagram variation increment is constructed, a two-dimensional convolutional neural network containing a channel attention module is designed, recursive deep learning and weight Gaussian sampling are fused, and the prediction accuracy of the residual life of wind power bearings is improved.
Owner:GUANGDONG HUADIAN FUXIN YANGJIANG OFFSHORE WIND POWER CO LTD +1

Dynamic adaptive unmanned aerial vehicle path planning method based on RRT*

The invention relates to a dynamic self-adaptive unmanned aerial vehicle path planning method based on RRT *. The technical problems that an existing unmanned aerial vehicle path planning method based on an RRT * algorithm is low in convergence speed, a generated path is high in tortuosity degree and poor in smoothness, and efficiency needs to be improved are solved. Sampling points are generated through self-adaptive Gaussian sampling according to self-adaptive target offset probability selection, different step lengths are used in different exploration periods by adopting an intelligent variable step size mechanism, and Pareto evaluation on path tortuosity degree is introduced to further obtain an optimal path. The path planned by the method is high in convergence speed, and the generated path is low in tortuosity and good in smoothness.
Owner:HARBIN INST OF TECH AT WEIHAI

Gadget sub-Gaussian sampling method, device, equipment and storage medium

The present application relates to a gadget sub-Gaussian sampling method, device, equipment and storage medium, wherein the method comprises: obtaining an input parameter q of the gadget sub-Gaussian sampling; judging whether the input parameter q satisfies a first preset condition or a second preset condition; when judging that the input parameter q satisfies the first preset condition, obtaining the sub-Gaussian sampling by using a sub-Gaussian randomized nearest plane algorithm; when judging that the input parameter q satisfies the second preset condition, determining the last dimension x of the output vector by a random operation k‑1 The positive and negative of the last dimension is determined, and the first k-1 dimensions of the output vector are determined based on the positive and negative of the last dimension, and the sub-Gaussian sampling is obtained using the first k-1 dimensions. Thus, the sampling when the modulus q is not a power of b is converted to the sampling when the modulus q is a power of b, and the output vector is guaranteed to satisfy the sub-Gaussian distribution, and a smaller sub-Gaussian parameter can be achieved, thereby achieving higher quality sampling.
Owner:TSINGHUA UNIVERSITY

A training method for image classification model and image classification method

The embodiment of the present invention provides an image classification model training method, which includes: S1, obtaining an original image data set including multiple image samples and its corresponding feature matrix; S2, performing dimensionality reduction processing on the feature matrix corresponding to the original image data set, and Gaussian modeling on the feature matrix after dimensionality reduction processing to obtain the Gaussian distribution of each eigenvalue in the feature matrix; S3, based on the Gaussian distribution of each eigenvalue in step S2, Gaussian sampling is performed on it in sequence, so that all eigenvalues ​​obtained by Gaussian sampling constitute a new feature matrix, and a new image data set is generated based on the new feature matrix; S4, using the new image data set and the original image data set to form an enhanced image training set to train the image classification model until convergence. This method improves the performance of the image classification model in the case of few samples.
Owner:INST OF COMPUTING TECH CHINESE ACAD OF SCI

Information prediction method, training method of information prediction model and related device

The embodiment of the invention discloses an information prediction method, a training method of an information prediction model and a related device. The method comprises the following steps: encoding input information to obtain an encoding memory; hidden variable probability distribution is extracted for the coded memory, Gaussian sampling processing is carried out on the coded memory according to the hidden variable probability distribution, continuous hidden variables are obtained, and the continuous hidden variables represent hidden semantics corresponding to the input information; a plurality of discrete hidden variables are obtained, the continuous hidden variables and the discrete hidden variables are mixed, a plurality of mixed hidden variables are obtained, and the discrete hidden variables represent hidden semantics corresponding to the prediction results; and performing decoding processing according to the plurality of mixed hidden variables and the coding memory to obtain a plurality of prediction results. The embodiment of the invention can be suitable for application scenes meeting low-resource and time-limited conditions at the same time, and can ensure the diversity and accuracy of prediction results.
Owner:TENCENT TECHNOLOGY (SHENZHEN) CO LTD

Text steganography method for digital media protection based on variational autoencoder

ActiveCN113987129BAchieving visual indistinguishabilitySemantic analysisNeural architecturesAlgorithmNetwork model
The invention belongs to the field of information security, and particularly relates to a digital media protection text steganography method based on a variational autoencoder, comprising constructing a neural network model composed of an encoding network, Gaussian sampling and a decoding network to vectorize text; using the encoding network to respectively obtain features of global keywords and long sequences, and fusing the features of the global keywords and the long sequences to obtain global feature representation; using Gaussian sampling to perform Gaussian sampling on the global feature representation in the encoding network; using a decoding network to decode the sampling result of the Gaussian sampling to obtain the conditional probability distribution of the text; selecting K words with the largest conditional probabilities, and using Huffman coding to select a word corresponding to a secret bit stream to complete the steganography of a file; the invention can not only generate long and diverse steganographic texts, so that the steganographic texts can carry more secret information, but also realize visual indistinguishability, statistical indistinguishability and semantic indistinguishability between natural language and steganographic text.
Owner:JIANGXI JIE XUN ENTERPRISE MANAGEMENT CO LTD

First view angle incremental behavior identification method based on multi-mode adapter

ActiveCN119538049AImproved accuracy of average incremental recognitionimprove accuracyCharacter and pattern recognitionBiological modelsData setGyroscope
The invention discloses a first view angle incremental behavior identification method based on a multi-mode adapter, and relates to the field of video image processing. According to the method, the time sequence perception capability in the old task identification model is migrated to the current task identification model through the time sequence distillation loss, and the incremental identification of the classification layer on multiple tasks is realized through the feature Gaussian sampling of the old task identification model and the feature Gaussian sampling of the current task identification model in combination with the modal balance adapter. The method comprises the following steps: firstly, inserting a multi-mode time sequence sensing adapter in the front and back of forward network layers of all Transform encoders of a Vit-B / 16 model pre-trained on a frozen Image Net data set in a residual form, so as to extract time sequence information from the features of a visual mode, an acceleration mode and a gyroscope mode in the Transform encoders; and the time sequence information of the current task identification model and the old task identification model is close to each other as far as possible through the time sequence distillation loss. And finally, constructing the feature Gaussian distribution of the current task by calculating the feature mean value and variance of the current task recognition model, realizing the feature Gaussian sampling of multiple tasks in combination with the feature Gaussian distribution of the old task recognition model, and inserting a modal balance adapter in front of the classification layer of the current task. And training the classification layer of the current task again by using the sampling features and the modal balance adapter so as to improve the recognition capability of the classification layer on each task.
Owner:UNIV OF ELECTRONICS SCI & TECH OF CHINA

Millimeter wave movable array beam training method based on quantum correlated simulated annealing

The application discloses a millimeter wave movable array beam training method based on quantum correlation simulated annealing, and relates to the technical field of wireless communication.The application proposes a quantum correlation simulated annealing (QCSA) algorithm aiming at the high-dimensional, non-convex and strongly coupled beam-array position joint optimization problem caused by the movable array, and adopts an alternating optimization framework to decompose the joint optimization problem into iterative solving of the transmitting and receiving subspaces; in each subspace optimization, the physical correlation between parameters is learned through Gaussian sampling based on an adaptive covariance matrix, the global search capability is enhanced by using an acceptance criterion containing a thermodynamic term and a quantum tunneling term to escape from a local optimum, and a periodic oscillation thermal scheduling strategy is adopted to dynamically balance exploration and utilization, so that the application has the characteristics of low overhead, fast convergence and high robustness, and is suitable for 6G millimeter wave and terahertz movable array MIMO systems.
Owner:NORTHEASTERN UNIV AT QINHUANGDAO

Multi-working-condition tool wear prediction method based on reciprocal Gaussian sampling element learning

The invention discloses a multi-working-condition tool wear prediction method based on reciprocal Gaussian sampling element learning, and relates to the technical field of numerical control machining process monitoring, and the method comprises the steps: obtaining monitoring signals in a tool cutting process under different historical working conditions; feature screening is carried out for any historical working condition; initializing parameters of a task learner, a meta learner and a reciprocal Gaussian sampling mechanism; dividing the screened features into a support set and a query set based on an initialized reciprocal Gaussian sampling mechanism; obtaining the total support loss of the historical working condition on the support set; parameters of a task learner are updated based on the total support loss of the historical working conditions; obtaining the total query loss of the historical working condition on the query set; updating parameters of the meta-learner and parameters of a reciprocal Gaussian sampling mechanism according to the sum of the total query losses of all the historical working conditions; and the tool wear value is predicted based on the monitoring signals collected in real time under the current working condition and the final parameters of the meta-learner. According to the invention, the tool wear can be accurately measured.
Owner:SOUTHWEST JIAOTONG UNIV

Compact public key searchable encryption method, apparatus, system, device and medium

The present invention discloses a compact public key searchable encryption method, device, system, equipment and medium, belonging to the field of information encryption technology, and the method comprises the following steps: step 1, initializing the operating environment, generating the required public parameters; step 2, adopting the approximate trapdoor sampling method and the extensible output function, and running the key generation algorithm to generate the public key and the private key; step 3, generating the searchable ciphertext for the keyword attached to each encrypted file according to the public key, and uploading the generated searchable ciphertext and the ciphertext file together; step 4, based on the pre-set search keyword, adopting the non-spherical Gaussian sampling technology and the approximate trapdoor sampling algorithm, generating the retrieval trapdoor; step 5, retrieving all the searchable ciphertext according to the retrieval trapdoor, and returning the ciphertext files of all the matching searchable ciphertexts. The present invention has higher efficiency and lower public key and trapdoor storage, effectively resists the attack of quantum computers, and is better applied to the Internet of Things.
Owner:QUAN CHENG LABORATORY

Light coordinate mapping method for converting elliptical Gaussian beam into elliptical flat-topped beam

The invention discloses a mapping method between an incident ray coordinate and an emergent ray coordinate when a collimated elliptical Gaussian beam is converted into a collimated elliptical flat-topped beam and the size of an elliptical light spot is changed, and belongs to the field of laser shaping. The method for calculating the mapping relation between the output light coordinates and the input light coordinates comprises the following steps of: 1, utilizing the law of conservation of energy to ensure that the energy surrounded in an input surface ellipse is equal to the energy surrounded in an output surface ellipse; solving a mapping relation between the ellipse semi-major axis xiout and the ellipse semi-minor axis eta out of the emergent light beam and the ellipse semi-major axis xiin and the ellipse semi-minor axis eta in of the incident light beam; 2, solving a mathematical expression of a mapping relation between output light coordinates and input light coordinates: determining an included angle theta between a connecting line between a point (xin, yiin) and the center of an ellipse and an X axis according to the point (xin, yiin) on the ellipse circumferences of a semi-major axis xiin and a semi-minor axis eta in on an input surface, and solving an emergent point coordinate (xout, yout) corresponding to an incident point (xin, yiin) according to the theta, xiout and eta out; and 3, light intensity statistical simulation verification: carrying out Gaussian sampling on the coordinates (xin, yiin) of one million light rays, and carrying out distribution statistical simulation on all (xin, yiin) and (xout, yiout) by using matlab programming, so as to display the conversion from an elliptical Gaussian beam to an elliptical flat-topped beam.
Owner:QINGDAO LASENCE GRP CO LTD

A method for designing acoustic metamaterials based on PGN model

The application provides an acoustic metamaterial design method based on a PGN model, which not only overcomes the one-to-many mapping problem from spectral response to structural parameters in the traditional acoustic metamaterial design process, but also can significantly reduce the overall computing time and improve the design efficiency by predicting the solution immediately after the training phase. The design method first constructs a data set including the structural parameters of the acoustic metamaterial and the corresponding spectral response; then in a cascaded manner of a reverse GRU model in front and a forward DNN model pre-trained by the data set in back, a PGN model is constructed, and the PGN model is trained by the data set; wherein the candidate meta-structure obtained by Gaussian sampling of the output of the reverse GRU model is used as the input of the pre-trained forward DNN model; finally, the customized spectrum is input into the trained PGN model, and the Gaussian distribution predicted by the PGN model is probabilistically sampled to generate the structural parameters meeting the conditions.
Owner:NANJING UNIV

Virtual navigator path planning method based on SAC-RRT*

The invention provides a virtual navigator path planning method based on SAC-RRT *, which innovatively introduces formation geometric envelope parameters to map multi-agent entity constraints into virtual navigator geometric constraints, replaces traditional RRT uniform sampling with non-deterministic Gaussian sampling of an SAC algorithm, and realizes adaptive step length extension and formation-level collision detection in combination with a Critic network. Meanwhile, the path cost optimization quantity of RRT rewiring is creatively converted into rewards to be fed back to the SAC network to complete backward fusion, network training is enhanced in cooperation with a real and virtual dual-experience playback strategy, formation trafficability factors are fused in a state space, a multi-dimensional reward function optimization decision is designed, and finally, the formation trafficability is improved through three times of B spline curve smoothing processing. The multi-agent formation path planning considering safety, optimality and convergence efficiency is realized, and the problems that a traditional method ignores formation geometric constraints, sampling blindness is large, and reinforcement learning and sampling algorithm fusion is not deep are effectively solved.
Owner:CHINA THREE GORGES UNIV

Lattice-based fine-grained attribute encryption method supporting user and attribute double revocation

The invention discloses a lattice-based fine-grained attribute encryption method supporting user and attribute double revocation, and belongs to the technical field of information security and cryptography. The method comprises the following steps: S1, initializing a system; s2, public key distribution; s3, generating a private key; s4, encrypting and uploading the data; s5, requesting and decrypting data; s6, releasing the user revocation list; and S7, distributing the attribute revocation list. According to the method, fine-grained access control on encrypted data in a cloud environment is realized, user revocation and attribute revocation are supported at the same time, the attribute revocation does not need to re-execute a discrete Gaussian sampling algorithm to generate a user private key, and ciphertext updating is realized by cloud service without participation of an encryptor; the user revocation only needs to regenerate the ciphertext component related to the user revocation by the encryption party, and any user private key does not need to be updated; according to the double revocation mechanism, the system calculation amount and the communication burden are effectively reduced, and low-overhead and dynamic fine-grained authority management facing the cloud environment is realized.
Owner:NANJING UNIV OF SCI & TECH