197 results about "Multi kernel" patented technology

A system, method and computer program product that is designed to support high-availability, rapid fault recovery, out of band condition signaling and / or other quality of service assurances and security in a networked environment. In one aspect, a method of the invention includes the step of providing a processing system with a dual-kernel or multi-kernel software operating system. The operating system includes a supervisory operating system and a secondary operating system that provides network functions to user applications. The method also includes the step of providing a Network Control Software (NCS) in the supervisory operating system. The NCS is configured to transparently monitor and control network operations in the secondary operating system.

A system, method and computer program product that is designed to support high-availability, rapid fault recovery, out of band condition signaling and / or other quality of service assurances and security in a networked environment. In one aspect, a method of the invention includes the step of providing a processing system with a dual-kernel or multi-kernel software operating system. The operating system includes a supervisory operating system and a secondary operating system that provides network functions to user applications. The method also includes the step of providing a Network Control Software (NCS) in the supervisory operating system. The NCS is configured to transparently monitor and control network operations in the secondary operating system.

A chip used for realizing communication between multi-kernel in chip is prepared as setting a sharing storage in chip, providing data transceiving buffer region and data staring channel, carrying out data transceiving and interacting in said sharing storage to realize communication mechanism of data-sharing and data-interacting as well as interruption-interacting by using ARM kernel and DSP kernel to write data in address unit preset by ARM kernel / DSP kernel through said sharing storage and informing the opposite party to fetch data.

The invention discloses a multi-kernel processor for implementing self- adaptive scheduling, comprising buffering message receiving unit, above one kernel processing unit and the connected message transmitting unit; and also comprising kernel scheduling unit for counting each use parameter of each kernel processing unit and according to the counted results, processing the kernel processing units in groups; message allocating unit, for feeding the message received by the message receiving unit into a group of kernel processing units according to the current grouping information of kernel scheduling units. And it also discloses a multi-kernel processing method for implementing self-adaptive scheduling. And it gives full play to advantages of the multi-kernel processor; increasing system throughput and balancing numbers of messages processed by kernel processing units and raising processing efficiency of the processor.

The invention discloses a multi-kernel-learning-based multi-mode emotion identification method. According to the method, extraction of emotion features like an expression feature, a voice feature and a physiological feature is carried out on sample data of each mode in a multi-mode emotion database; several different kernel matrixes are constructed for each mode respectively; the kernel matrix groups corresponding to different modes are fused to obtain a fused multi-mode emotion feature; and a multi-kernel support vector machine is used as a classifier to carry out training and identification. Therefore, basic emotions like angering, disgusting, fearing, delighting, upsetting, and surprising and the like can be identified effectively.

The invention discloses an oil soaked transformer fault diagnosis method, including the following steps: a sample is firstly obtained, and the concentration data of five gases in the sample is in normalized treatment, so that a training sample set and a test sample set are formed; the number of elementary kernel functions and the parameters of each basis kernel are determined, and the optimal punitive parameters are determined by using the method of cross validation; according to the optimal punitive parameters, corresponding disaggregated model is obtained by using training sample and multi-classification multi-kernel learning method; and the disaggregated model after trained is used for carrying out fault diagnosis of the sample to be tested in verification centralization. The invention can guarantee very high accuracy rate of diagnosis, and has very good practicability and generalization.

The invention discloses a consistency maintenance device for a multi-kernel processor and a consistency interaction method, mainly solving the technical problem of large directory access delay in a consistency interaction process for processing read-miss and write-miss by a Cache consistency protocol of the traditional multi-kernel processor. According to the invention, all kernels of the multi-kernel processor are divided into a plurality nodes in parallel relation, wherein each node comprises a plurality of kernels. When the read-miss and the write-miss occur, effective data transcription nodes closest to the kernels undergoing the read-miss and the write-miss are directly predicted and accessed according to node predication Cache, and a directory updating step is put off and is not performed until data access is finished, so that directory access delay is completely concealed and the access efficiency is increased; a double-layer directory structure is beneficial to conversion of directory storage expense from exponential increase into linear increase, so that better expandability is achieved; and because the node is taken as a unit for performing coarse-grained predication, the storage expense for information predication is saved compared with that for fine-grained prediction in which the kernel is taken as a unit.

The invention discloses a multi-kernel support vector machine multi-instance learning algorithm applied to pedestrian re-identification. The algorithm includes two main steps, namely multi-feature description and a multi-kernel SVM model multi-instance learning algorithm. According to the algorithm, HSV color features and SIFT local features of two pictures, taken under a camera A and a camera B, of the same pedestrian are extracted to construct a word bag, and difference vectors of the two kinds of the features represent the conversion relation under the two cameras to serve as two instance samples and are encapsulated as a bag; then a multi-kernel support vector machine model is optimized, the bag is trained by means of linear fusion of the Gaussian kernel and a polynomial kernel, optimal parameters are obtained through multi-instance learning, and a high identification rate is achieved.

The invention discloses a multi- kernel DSP (digital signal processing) reconfigurable special integrated circuit system, belonging to the technical field of digital signal processing, and comprising: an internal bus, and a control processor kernel, an enhanced direct memory access, an input and output cache, a DSP multi-kernel array, a configuration information cache, a reconfigurable logic unitand an internal cache which are all connected with the internal bus, wherein the DSP multi-kernel array is connected with the configuration information cache and the reconfigurable logic unit througha reconfigurable on-chip interconnection mode and transmits the configuration information and reconfigurable information. The multi-kernel DSP reconfigurable special integrated circuit system can be well combined with an IP multiplexing technology of SoC (System on a Chip), a multi-kernel DSP reconfigurable ASIC (Application Specific Integrated Circuit) takes the DSP multi-kernel array as a core,and simultaneously, integrates IP modules, such as a logic control, an embedded memory, a data interface and the like, thereby being capable of flexibly and efficiently implement large scale computing.

The invention discloses a remote sensing image scene classification method based on a deep convolutional neural network and multi-kernel learning. The remote sensing image scene classification methodincludes the steps: training remote sensing scene images through the deep convolutional neural network, and taking the output of two full connection layers which are obtained through learning as the characteristics of the remote sensing scene images; utilizing multi-kernel learning to the kernel function being suitable for the characteristics of the two full connection layers, so as to map the extracted characteristics of the two full connection layers to a higher space to realize adaptive fusion of the characteristics of the two full connection layers in the higher space; and finally, designing a multi-kernel learning-support vector machine to effectively classify the remote sensing scene images. The remote sensing image scene classification method based on a deep convolutional neural network and multi-kernel learning performs characteristic extraction on the remote sensing images through the deep convolutional neural network, and the deep characteristic coverage information obtainedthrough learning is complete and has relatively higher discriminating performance, at the same time, the characteristics are fused into the multi-kernel learning framework, so that preferable classification performance can be obtained.

A performance offset-based aeroengine fault diagnosis method is disclosed. The invention relates to the performance offset-based aeroengine fault diagnosis method. The performance offset-based aeroengine fault diagnosis method is used for solving a problem that individual difference between aeroengines is neglected via technologies of the prior art, aeroengine fault sample data is small in quantity and conventional methods are low in practicality due to usual adoption of simulation data. The performance offset-based aeroengine fault diagnosis method comprises the following steps: in a first step, aeroengine air passage performance data is obtained and divided into groups; in a second step, according to a grouping result of the first step, an aeroengine performance offset model is built and performance offset is obtained via solving operation; in a third step, according to a solved result of the performance offset, a support vector machine multi-kernel function can be determined and fault classification can be performed. In terms of fault classification accuracy and generalization, the performance offset-based aeroengine fault diagnosis method is better than a traditional fault diagnosis method based on time series fitting. The performance offset-based aeroengine fault diagnosis method can be applied to the technical field of aeroengine maintenance and optimization.

The invention relates to an electromyographic signal classification method based on a multi-kernel support vector machine. For a sample with complex distribution, based on the classification performance of a single-kernel support vector machine, the classification accuracy and the quantity of support vectors are easily influenced. The method combines a multi-kernel support vector machine method with a binary tree combination strategy and comprises the following specific steps of: collecting electromyographic signals of the lower limbs of a human body through an electromyographic signal acquisition instrument; denoising the electromyographic signals containing interference noise by using a wavelet coefficient inter-scale correlation denoising method; extracting the features of the denoised electromyographic signals to obtain the features of the electromyographic signals by using denoised wavelet coefficients; and classifying on the basis of the multi-kernel support vector machine. The method can well meet the multi-classification requirement of lower extremity prosthesis control, and takes into account both accuracy and instantaneity, and has broad application prospects in the multi-movement mode recognition of intelligent prosthesis control.

The invention discloses a parallel data processing method based on the hidden Dirichlet distribution model, which belongs to the data mining field. The method includes three solutions, including the multi-process parallel processing, the multi-thread parallel processing and the composite multi-process multi-thread processing; the data DM for being processed is divided into data segments in equal or unequal length in the three solutions; each data segment is provided with an index; each computer process / thread processes the corresponding data segment through applying the index, so as to obtain the subject information of each data item and generate the local sufficient statistic; when the whole DM is processed, the global sufficient statistic is obtained through the merge of the local sufficient statistics so that the current Mi model is obtained through the estimation until the model becomes convergence. The parallel data processing method can utilize the multi-kernel parallel frame of a single computer and the cluster large-scale parallel capability of multi-computer to realize the high-speed processing of the large-scale text sets and effectively reduce the memory usage during the parallel processing process.

The invention relates to the technical field of tunnel excavators, in particular to a TBM cutting tool life prediction method based on a data driven support vector regression machine. The method comprises the following steps that 1, data of the TBM cutting tool excavation site is collected; 2, driving factors that affect the TBM cutting tool life is determined, and a sample data set of the driving factors is established as a training set; 3,a prediction model of a multi-kernel support vector regression machine is constructed, the training set is input, training is conducted on the prediction model to determine corresponding optimal parameters and penalty function C and insensitive loss function parameter epsilon; 4, an optimal kernel function of the prediction model is determined; 5, the sample data set of the driving factors of the cutting tools to be predicted serves as a prediction sample set, the prediction model is input, and a prediction result is obtained. According to the TBM cutting tool life prediction method based on the data driven support vector regression machine, a large amount of data of the excavation site is selected as parameters, and on the basis of a model based on support vectors regression machine is constructed, and the accuracy of a prediction tool is improved.

The invention discloses an SVM (support vector machine) parameter selection method based on multi-kernel function adaptive fusion. Compared with the prior art, the method achieves the specific analysis of the characteristics of a local kernel function, a global kernel function, a mixed kernel function and a multi-kernel function. The method comprises the steps: combining all fusion coefficients, kernel function parameters and regression parameters of the multi-kernel function together to serve as a parameter state vector, thereby enabling a model selection problem to be converted into a state estimation problem of a nonlinear system; carrying out the parameter estimation through fifth-order volume Kalman filtering, and achieving the adaptive fusion of weighted coefficients of the multi-kernel function and the selection of kernel parameters and regression parameters.

The invention provides a BP neural network parallelization method oriented to a multi-core computing environment. Including: parallel calculation task division and mapping method; cache setting method for storing intermediate calculation results of each layer; parallel training method of BP neural network. For the hardware platform containing P computing units, the task division method combines the hidden layer and output layer computing tasks into a larger granularity task as a whole, so as to improve parallelism; the cache setting method, in a training process, The same variable can be used in the next subtask after being accessed by the previous subtask, and the next subtask will not cause Cache loss; in the BP neural network training method, the samples are divided into K batches to enter the network training, and the K value is designed on a comprehensive computing platform The size of the secondary cache, combined with the cache settings, maximizes hardware performance and is suitable for BP neural network applications under multi-core computing platforms.

The invention discloses a heterogeneous multi-kernel power capping method through coordination of DVFS and task mapping. The method comprises the steps that firstly, computational node power consumption, CPU power consumption and GPU power consumption scripts can be measured after program execution is completed for a heterogeneous system, then, selected parallel test benchmark programs are modified for obtaining the execution time of different kernel functions; different frequencies are set for a CPU and a GPU, application programs are operated only on the CPU and the GPU, detailed operation information is obtained and comprises the total execution time, the execution time of each kernel function, computational node power consumption, CPU power consumption and GPU power consumption; on thebasis of the operation information, a predicted model is designed and includes a predicted execution time model and a power consumption model; finally, on the basis of the predicted model, system power consumption and execution time under different CPU frequencies, GPU frequencies and task distribution schemes are obtained to be filled in a configuration table, and according to an improved greedyalgorithm, the best configuration scheme is found. By adopting the heterogeneous multi-kernel power capping method, the system power consumption budget is limited while the system performance can beimproved.

The invention discloses a configurable processor. Ordered-connection function units are configured to form a local assembly line, multiple function units are guaranteed to simultaneously participate in operation to realize a specified function, and the resource utilization rate can be improved. In addition, the multi-kernel configurable processor kernel structure can be used for realizing the function required by users through configuration; and meanwhile, the configuration mode and control information are determined or changed at any time, thus being very flexible. The interconnection structure ensures that the processor can efficiently collaborate to realize the complex function under the condition that the processor contains a plurality of simple or configurable kernels.

The invention discloses a multi-kernel support vector machine classification method for remote sensing images, and belongs to a support vector machine classification method for the remote sensing images. The method comprises the following steps of: performing principal component transform on original data; taking first four principal components to represent spectral information, performing wavelet texture feature extraction on the first principal component, and combining the spectral feature and spacial feature by adopting two independent radial basis functions; and finally performing classification by utilizing a multi-kernel support vector machine method. The wavelet texture feature and the spectral feature are combined thorough a plurality of basis functions, so the spectral feature extracted by principal component analysis is fully utilized, the wavelet texture feature is fused, the support vector machine is optimized, and the limitation that the traditional method separately adopts the spectral feature for classification is overcome; therefore, the classification accuracy is effectively improved. The method has the main advantage of improving the classification accuracy by combining the spectral information and the spacial information through the plurality of basis functions.

An embodiment of the invention discloses a method for invoking interfaces. The method includes receiving interface invoking information transmitted by an optional browser kernel in a multi-kernel browser; converting formats of parameters; utilizing the converted parameters as input parameters and invoking target interface functions corresponding to interface identifiers. The interface invoking information contains the interface identifiers and the parameters which are required for invoking the target interfaces, and data types of the parameters are supported by the optional browser kernel. Data types of the converted parameters are preset first data types. The embodiment of the invention further discloses a device and a terminal for invoking the interfaces. The method, the device and the terminal in the embodiment of the invention have the advantage that the interface invoking efficiency can be improved by the aid of the method, the device and the terminal.

The invention discloses a method and a device for switching different versions of IE (internet explorer) kernels on the basis of a multi-kernel browser. The browser comprises at least two kernels, wherein one kernel is the IE kernel. The method comprises the following steps of: receiving request information of a loaded webpage, wherein the request information comprises URL (uniform resource locator) information; learning the adaptation relationship between the requested webpage and a browser kernel; selecting the corresponding browser kernel; when the selected browser kernel is the IE kernel, judging the version of the current adaptive IE kernel; if the current adaptive IE kernel is the IE6 kernel, invoking the redirected module of the IE kernel to load a preset IE6 kernel file; and if the current adaptive IE kernel is a novel IE kernel, invoking the redirected module of the IE kernel to load a preset novel IE kernel file. According to the invention, the problems of compatibility and suitability of the IE6 kernel browser and the IE8 kernel browser when the browser comprises the at least two kernels and one of the kernels is the IE kernel can be solved.

The invention relates to a method for establishing a partition system based on a multi-kernel MOS (Module Operating System). The method is applicable to multi-core / many-core heterogeneous processor platforms. According to the principle of load balance and minimum communication latency, system partitions are mapped to proper processor cores respectively; an MOS kernel matching with ISA (Instruction Set Architecture) of the corresponding processor core is started to run on the processor core, a communication channel and an accessory management data structure are established and initialized for the partition mapped to the processor core, and the partitions are communicated by means of messaging; the MOS kernel starts operating systems and tasks inside the partitions mapped to the processor cores where the MOS kernel is located. The method allows the processor cores in varieties of ISAs to run in the same partition system, and timeliness and accuracy in program running are guaranteed to the maximum extent while system efficacy and extensibility are improved effectively.

The invention relates to a roller bearing service life predicting model of a self-adaptive multi-kernel combination relevance vector machine. The roller bearing service life predicting model comprises the following steps of utilizing particle filter to initialize a combination kernel function weighting matrix, so as to obtain a combination kernel function set; establishing a multi-kernel combination relevance vector machine set; performing iteration predicting, weight updating and re-sampling self-adaption to obtain an optimal multi-kernel combination relevance vector machine model; finally, predicting the running state and remaining life of the rolling bearing. The roller bearing service life predicting model has the advantages that the excellent characteristics of a plurality of single kernel functions are adaptively integrated, the reliance of the single kernel function relevance vector machine model on the parameters is reduced, the predicting accuracy is improved, the predicting stability is better, the robustness of the model is higher, and the engineering application value is higher.

The invention discloses a Structure sparse multi-kernel learning-based multi-modal data feature screening and classification method. The method comprises the following steps of: extracting data features and normalizing the features; constructing a kernel matrix for each data feature by utilizing a kernel function; grouping all the data features and the corresponding kernel functions according to the data features; training a Structure sparse multi-kernel classification model by utilizing training data with class labels, and optimizing parameters of the model; and classifying test data by using the trained Structure sparse multi-kernel classification model. According to the method, the data feature selection and data modal fusion are modeled in the uniform structure sparse multi-modal classification model, and the structure sparse feature selection and classifier learning expressed on the basis of an optimal kernel are carried out at the same time, so that a multi-modal data feature screening, fusion and classification method is provided.

The invention discloses an automatic election method for a master boot kernel in heterogeneous multi-kernel system, comprising the following steps: after switched on, a multi-kernel CPU calls an election mechanism which is used for determining the master boot kernel; each kernel in the multi-kernel CPU executes the election mechanism, and judges whether itself is a master boot kernel; at the same time, all slavery kernels in normal state are recorded; the determined master boot kernel loads an operating system, and initializes peripheral hardware equipments and the operating system; the slavery kernels enter into sleep state; the master boot kernel wakes up all slavery kernels in normal state, the slavery kernels load and initialize respective operating system code segments after waken up, and then the master boot kernel and the slavery kernels enter into normal running state. The method is capable of avoiding the defect of startup failure of the whole system, which is caused by the fault occurrence of the master boot kernel.

The invention provides a hyperspectrum classification method based on a composite kernel function. The hyperspectrum classification method comprises the steps of inputting a set of hyperspectrum images in N classes, taking a support vector machine as a base classifier, and meanwhile, randomly selecting S samples from every classes of the hyperspectrum images to form a training set and forming a test set with the left samples, determining the change range of each parameter, next, determining the optimal performance parameters, including a penalty factor and a kernel of the support vector machine by virtue of cross validation for K times, constructing a composite kernel function by use of a composite kernel construction policy and training the support vector machine, and circulating for N times by use of the parameters of a support vector machine decision function obtained in the training process to obtain decision function values of which the test set belongs to every classes and to form a matrix as shown in the specification, and then determining multiple classifier policies, namely finding the maximum values of every columns of the matrix as shown in the specification. The hyperspectrum classification method based on the composite kernel function has the characteristics of better description of distribution features of a data set, relatively high classification accuracy and the like. The time taken by parameter optimization of the hyperspectrum classification method is also relatively short in contrast with a traditional multi-kernel learning method.

The present invention discloses a preparation method of new-type condensed polynuclear aromatic resin carbon base solid acid catalyst. Said preparation method includes the following steps: using polycyclic aromatic compound or its mixture as raw material, using aldehyde compound as cross-linking agent, using inorganic strong acid as catalyst, making condensation reaction to obtain B-stage condensed polynuclear aromatic resin; adding inorganic strong acid and aldehyde compound in the B-stage condensed polynuclear aromatic resin, making continuous reaction at high temperature to obtain C-stage condensed polyunuclear aromatic resin; then making the C-stage condensed polynuclear romatic resin undergo the processes of sulfonation, washing by using water, filtering, removing sulfonating agent and drying so as to obtain the invented target product.

The invention discloses a flight control distributed type real-time simulation system, relates to the field of flight control simulation, and solves the problems that a model cannot be installed with a large quantity of I / O (Input / Output) physical systems and real-time simulation cannot be realized in the prior art. The flight control distributed type real-time simulation system comprises one or two development host units and multiple target machines, wherein the development main units are system dynamics, navigation and control model development and simulation state monitoring platforms; the target machines are real-time simulation processors; each target machine can be regarded as a node in a real-time network and operates a QNX real-time operating system; the main units are connected with the target machines through a TCP / IP network; the target machines are connected with one another through an IEEE-1394 bus. By the adoption of multi-kernel and multi-model distributed type simulation, parallel simulation of multiple models is realized, and the simulation efficiency is obviously improved; synchronous real-time simulation of the simulation time and the actual physical time is realized; all the models are independent from one another but also cooperated with one another, so that the flexibility of module configuration is improved.

The invention discloses a novel genotype analysis method based on multimodal brain images. With utilization of information, including a structural image and a functional image, of two kinds of modes, regression on the rs429358 locus genotype linked to the Alzheimer's disease closely is realized. The method comprises: extracting voxel features from a brain structure images and extracting connecting network features from a brain function image; carrying out feature extraction on the two kinds of features by using an elastic network and keeping brain area and network connection with a high association degree; and then carrying out fusion and genotype regression on selected nodes and side attributes by using a multi-kernel support vector. The method has the following advantages: for the brain image with the complicated structure and functions and genotype data, a novel regression analysis method for a single locus genotype based on structural and functional multimodal brain images is put forward; the analysis effect is good; and complementary information of the two modes is utilized fully.

The invention relates to the technical field of network safety and discloses a multi-detection engine load balancing method applied to a network intrusion defensive system. In the invention, a multi-kernel network service processor is used as a hardware basis, a plurality of processing kernels of the multi-kernel network service processor are used as detection engines, and load balancing among a plurality of kernels, the detection accuracy and the processing demands of gigabit processing line speed are ensured by utilizing a dynamic load balancing method based on a stream. The dynamic load balancing method of network intrusion detection comprises the following steps of updating and maintaining a detection engine real-time load table, sending a data package, updating and maintaining a data package sending table and balancing a load. Through concrete implementation of the steps, the integrity of the stream can be ensured so as to enable subsequent detection or other processing to be more accurate. The dynamic load balancing method is simultaneously adopted so as to enable the loading amounts of each processing engine to have equal distribution and fast response. The invention is particularly suitable for intrusion detection in the gigabit network environment.