91 results about "Random method" patented technology

Disclosed herein are exemplary methods, apparatus, and systems for performing timing-aware automatic test pattern generation (ATPG) that can be used, for example, to improve the quality of a test set generated for detecting delay defects or holding time defects. In certain embodiments, timing information derived from various sources (e.g. from Standard Delay Format (SDF) files) is integrated into an ATPG tool. The timing information can be used to guide the test generator to detect the faults through certain paths (e.g., paths having a selected length, or range of lengths, such as the longest or shortest paths). To avoid propagating the faults through similar paths repeatedly, a weighted random method can be used to improve the path coverage during test generation. Experimental results show that significant test quality improvement can be achieved when applying embodiments of timing-aware ATPG to industrial designs.

The present invention provides a method of detecting segment boundaries for a series of successive frames in a video sequence. The method includes the steps of acquiring color information from each frame, determining a color histogram for each frame, and applying a boundary detection technique utilizing the color histograms. Finally, the method includes segmenting the frames of the video sequence into uniform color segments. In addition, a system is provided for detecting segment boundaries for a series of successive frames in a video sequence. The system includes means for acquiring color information from each frame, means for determining a color histogram for each frame, and means for applying a boundary detection technique utilizing the color histograms. Finally, the system includes means for segmenting the frames of the video sequence into uniform color segments. The boundary detection techniques include a family color histogram method, a weighted average color histogram method, a successive color histogram method, a stochastic method, a shot-based color histogram method, a mosaic color histogram method, and a computable macro-segment boundary method.

Analog computational arrays for matrix-vector multiplication offer very large integration density and throughput as, for instance, needed for real-time signal processing in video. Despite the success of adaptive algorithms and architectures in reducing the effect of analog component mismatch and noise on system performance, the precision and repeatability of analog VLSI computation under process and environmental variations is inadequate for some applications. Digital implementation offers absolute precision limited only by wordlength, but at the cost of significantly larger silicon area and power dissipation compared with dedicated, fine-grain parallel analog implementation. The present invention comprises a hybrid analog and digital technology for fast and accurate computing of a product of a long vector (thousands of dimensions) with a large matrix (thousands of rows and columns). At the core of the externally digital architecture is a high-density, low-power analog array performing binary-binary partial matrix-vector multiplication. Digital multiplication of variable resolution is obtained with bit-serial inputs and bit-parallel storage of matrix elements, by combining quantized outputs from one or more rows of cells over time. Full digital resolution is maintained even with low-resolution analog-to-digital conversion, owing to random statistics in the analog summation of binary products. A random modulation scheme produces near-Bernoulli statistics even for highly correlated inputs. The approach has been validated by electronic prototypes achieving computational efficiency (number of computations per unit time using unit power) and integration density (number of computations per unit time on a unit chip area) each a factor of 100 to 10,000 higher than that of existing signal processors making the invention highly suitable for inexpensive micropower implementations of high-data-rate real-time signal processors.

The invention develops a software test data evolution generation system facing a path, which has the advantages that the test data passing a target path can be automatically generated by using a genetic algorithm according to the test path selected by users, in addition, various optional genetic algorithm individual coding modes, adaptive value calculation methods, selection modes, crossing modes, variation modes and the like can be provided in the process of generating test data by using the genetic algorithm, and users can conveniently select different test data generation methods according to the tested program in different types. In addition, the system also provides the traditional random method test data generation method as the comparison, and the automation generation of the software test data is really realized. When the system disclosed by the invention is utilized, the software test efficiency is greatly improved, a large number of software development resources are also saved, the technical support is provided for the credible software research and development of the national relevant plan and the engineering, and the development of software industry in China is promoted.

The invention provides a fracture network 3D digital core modeling method comprising the steps as follows: step 1, using a CT scanning method or numerical method to establish a rock matrix 3D digital core; step 2, using a random method to generate a fractal discrete fracture network FDFN based on the assumption that fracture distribution satisfies the fractal characteristics; and step 3, discretizing the FDFN and superposing the FDFN and the rock matrix digital core voxel by voxel to obtain a fracture network 3D digital core. By adopting the fracture network 3D digital core modeling method of the invention, the problem that a representative core can be hardly obtained for a fractured reservoir is solved, and a research platform is provided for numerical simulation research of the physical properties of rocks in the fractured reservoir. Therefore, the method is worth promotion.

A signal scrambling method based on combined symbol rotation and inversion (CSRI) for the peak-to-average power ratio (PAPR) reduction of OFDM signals. By dividing OFDM sequences into subblocks and performing symbol rotation and inversion in each subblock, high degrees of freedom are available to offset the possibility of encounting poor sequences with large PAPR. In order to reduce the complexity of this scheme, two suboptimal CSRI schemes are disclosed, one based on a successive approach combined with symbol grouping, and the other based on a random approach combined with threshold control.

The invention puts forward a finite-element reconstruction method for space weighting of auto-fluorescence imaging. The method uses a diffuse optical tomography (DOT) technique to precisely reconstruct optical parameters of main biological tissue. The method limits a possible real light-source region, and divides a whole reconstruction region into a light-source possible region and a light-source impossible region so as to improve the numerical stability and effectiveness of reconstruction problems and reduce the ill-conditioned property of BLT reconstruction problems. In order to avoid famous 'Inverse Crime' problems, the method adopts a Monte Carlo (MC) random method to simulate the process of transmitting light in the biological tissue. Then the invention puts forward a finite-element method based on a space-weighting unit and adopts a tolerant algorithm for linear constrained optimization problems to reconstruct light-source information.

A method of providing delay information in a multi-antenna transmission system is provided. A time delay value is selected from a set of time delay values. A mobile station is informed of information related to the selected time delay value. The same data is transmitted from a first antenna and from a second antenna. The data is transmitted from the second antenna after the data is transmitted from the first antenna according to the selected time delay value. The time delay value is selected based on a deterministic method or a random method.

The invention proposes a traffic flow prediction method based on firefly algorithm and RBF neural network. The method comprises: performing normalization to the sample data so that the input data and output data are on the same order of magnitude; initializing the firefly algorithm parameters; utilizing the random method to initialize the firefly populations and encoding each individual in the populations; using the firefly algorithm to train the RBF neural network to obtain the best individual in the populations; decoding the best individual in the populations to obtain the trained RBF neural network; and utilizing the trained RBF neural network to predict the traffic flow data sample. Compared with the traditional traffic flow prediction method, the method of the invention makes full use of the advantages of the firefly algorithm in the RBF neural network training so that the RBF network possesses a more accurate prediction capability, achieves even faster training efficiency and better generalization capability. The invention belongs to the traffic transportation information engineering technology field and can be used for the predictions of road traffic flows in an intelligent traffic system.

A signal scrambling method based on combined symbol rotation and inversion (CSRI) for the peak-to-average power ratio (PAPR) reduction of OFDM signals. By dividing OFDM sequences into subblocks and performing symbol rotation and inversion in each subblock, high degrees of freedom are available to offset the possibility of encounting poor sequences with large PAPR. In order to reduce the complexity of this scheme, two suboptimal CSRI schemes are disclosed, one based on a successive approach combined with symbol grouping, and the other based on a random approach combined with threshold control.

The invention relates to a multi-seismic resource parallel motivation collecting and hybrid seismic record separating method. All measuring line shot points in an exploration area are divided into multiple corresponding sets, different scan signals are adopted by different seismic sources, motivation signals of all seismic sources are designed based on a pseudo-random method, wave detectors cover the receiving arrays corresponding to all the shot points, the wave detectors are distributed in a one-off mode, and roller paths are not needed. In the construction process, a seismograph continuously records the motivation signals to generate hybrid seismic records, after collection is completed, the hybrid seismic records of all the shot points are cut out according to the motivation times of all the shot points, and the seismic source scan signals adopted by all the shot points are used as reference signals for cross-correlation detection, namely the seismic records of each single seismic source of each shot point can be separated. Compared with an existing single-seismic-source motivation collecting method, the seismic exploration efficiency is substantially improved, n seismic sources are adopted to work in parallel, the actual construction time is 1 / n of the time of the existing single-seismic-source motivation collecting method, production efficiency is improved by n times, construction cost is lowered, and a low-cost data collection method is provided.

The invention discloses a test data generating method and a test data generating device. The test data generating method comprises the following steps that the preset test data definition information is obtained, and fields contained in test data to be generated are determined; further, relevant fields to be generated are determined according to the relevant definition information in the definition information of the test data, wherein the relevant fields are data fields with the relevant relationship with external data; the external data is read, and the relevant field data value of the test data is generated according to the content of the external data; and the nonrelevant field data value of the test data is generated by a random method. When the technical scheme disclosed by the invention is applied, the data fields relevant to the external data can be fully considered in the data generating process, so the relevant relationship among data is ensured, and the test requirement of a complicated system is met.

A method for distributing resource in mass storage system includes evaluating resource in SAN according to volume rule, carrying out resource distribution as per evaluation value based on probability random manner for realizing resource utilization and performance optimization and simultaneously supporting addition and deletion of magnetic disc in SAN.

Systems and methods are disclosed for stochastic block allocation for improved wear leveling for a system having non-volatile memory (“NVM”). The system can probabilistically allocate a block or super block for wear leveling based on statistics associated with the block or super block. In some embodiments, the system can select a set of blocks or super blocks based on a pre-determined threshold of a number of cycles (e.g., erase cycles and / or write cycles). The block or super block can then be selected from the set of super blocks. In other embodiments, the system can use a fully stochastic approach by selecting a block or super block based on a biased random variable. The biased random variable may be generated based in part on the number of cycles associated with each block or super block of the NVM.

The invention provides a random walk model-based zero-reference article recommendation method and system. The method comprises the following steps of 1, constructing an academic network model, and obtaining feature values corresponding to a first author, a conference or periodical, a mechanism and a published time of each paper through a random walk method; 2, establishing a sorting model, and selecting paper data processed in the step 1 to construct a training set; 3, sorting the training set through a weak classifier; 4, judging whether the sorting result of the weak classifier is matched with a real sorting result of the training set or not, so as to obtain an optimal sorting model; and 5, recommending zero-reference literatures required by the users through the sorting model. According to the method and system provided by the invention, a brand new paper sorting through is used, so that the newly published papers can be recommended more effectively and then the users can obtain the most relevant new papers.

A scalable computer architecture capable of performing fully scalable simulations includes a plurality of processing elements (PEs) and a plurality of interconnections between the PEs. In this regard, the interconnections can interconnect each processing element to each neighboring processing element located adjacent the respective processing element, and further interconnect at least one processing element to at least one other processing element located remote from the respective at least one processing element. For example, the interconnections can interconnect the plurality of processing elements according to a fractal-type method or a quenched random method. Further, the plurality of interconnections can include at least one interconnection at each length scale of the plurality of processing elements.

The application form relates to software test core problems, namely test case automatic generation, and provides an invariant technology-combined random test improvement method, namely an invariant-booted random test case automatic generation method, aiming at the defects of currently widely-applied automatic technology, namely a random test method. Two new methods are mainly provided and comprise: filtering invalid test cases by utilizing invariants of the extracted programs, and retaining valid cases; reducing invalid case space by utilizing the invariants of the extracted programs, and improving the probability of selecting valid cases. The methods can be used for reducing the redundancy rate, meeting the coverage requirement and extremely solving the defects of the traditional random methods, are full-automatic in the whole case generation process, without other any input and manual interference except for necessary tested programs.

The invention belongs to the field of satellite-to-earth observation and relates to a satellite observation task planning method based on case matching. The method comprises steps of: (S1), according to a historical task planning result, generating a case according to a case generation method and establishing a case library; (S2), computing a satellite regression period and setting a planning parameter and a comprehensive evaluated value function; (S3), reading the case library and performing similar target matching on the target set of a task set to be planned and the target set of the case set; (S4) performing case matching on the task set to be planned; (S5) performing case correction on a candidate case set; (S6), forming a case method initial group and a random method initial group according to a case method initialization strategy and a random method initialization strategy; (S7) performing first-stage evolution; (S8) combining the evolved groups into one group to be subjected to second-stage evolution to obtain a final group, and selecting an optimal solution as a final result; and (S9), matching the final result and storing the final result in the case library if the final result is not matched.

The invention discloses a renminbi selectable multiple-characteristic-point transmission authenticity-identifying method based on terahertz time-domain spectroscopy, comprising the following steps of: firstly transmitting total or selected characteristic points of hundred-yuan authorized renminbi by utilizing terahertz waves of discrete pulses to establish a dactylogram database of the terahertz absorption coefficients of the hundred-yuan authorized renminbi; then selecting only 3-5 transmission points to be detected on suspected renminbi to be identified by utilizing a random method according to the classified data of the dactylogram database, and enabling the terahertz waves of the discrete pulses to respectively transmit through each selected transmission point to obtain the terahertz absorption coefficient spectrums of the transmission points; and then comparing and analyzing the spectrums and the dactylogram database to judge the ascription on the authenticity of each transmission point and then judge the authenticity of the suspected renminbi to be identified. The accuracy of an identifying result is ensured through the fingerprint characteristics of the terahertz waves, thetransmission of paper sheets and the randomness of the selected transmission points to be detected; and in addition, by means of the renminbi selectable multiple-characteristic-point transmission authenticity-identifying method, overall scanning is omitted and a small quantity of characteristic points only need to be scanned, therefore the identifying time is greatly shortened.

The invention discloses a test data automatic generation method based on linear fitting function driving. According to the method, a path set is obtained through path analysis, then test data capable of covering each path are searched for each path, and a test data set of source codes to be tested is generated. The step of searching for the test data capable of covering each path for each path is achieved in the method that an input vector set is constructed through a random method firstly, then operational analysis is carried out on the input vector set, the input vector set is expanded through linear fitting means, and the test data capable of covering the paths can be obtained through repeated iteration. The method is highly automated, unit test efficiency can be improved as much as possible, test cost is reduced, and test coverage criteria oriented to paths are supported.

The invention discloses an asymmetric carrier double-random modulation vibration reduction method of a motor. The implementation system of the method comprises a voltage source inversion bridge, a permanent-magnet synchronous motor, a rotor position detection module, a coordinate conversion module, a PI regulation module, an asymmetric carrier double-random modulation space vector pulse width modulation (SVPWM) module, wherein the asymmetric carrier double-random modulation SVPWM module comprises a fan-shaped judgment module, a voltage vector calculation module, a zero vector calculation module, a random number generator, a zero vector random allocation module, a carrier generator, a modulation wave generator and a PWM generator, the zero vector random allocation module and the carrier generator are used for achieving double-random mechanism, the zero vector random allocation module is used for achieving random allocation of a zero vector, and the carrier generator is used for achieving randomness of an asymmetric carrier period. By the asymmetric carrier double-random method, the purpose of preventing full-frequency band vibration of the motor can be achieved, and an effective control method is provided for preventing vibration of the motor.

The present invention discloses a computer aided training method for mouth-shape recognition ability which solves the technique problem of increasing training efficiency, saving time of teacher and parents. The method includes steps as follows: the user can select self-selection method or random method, when self-selection method is selected, the user watches pronouncing mouth-shape video and simulates pronouncing mouth-shape action and pronunciation; when random method is selected, computer plays pronouncing mouth-shape video, and interface displays letter, the user judges and computer gives correct answer. By comparison with present technique, the method can make computer replacing teacher or parents for training mouth-shape recognition ability by using computer multimedia function, so the method can save time of teacher or parents and has active action for increasing viability of deaf and dumb people and other hearing impaired people. The method is especially suitable for training mouth-shape recognition ability of hearing impaired people, such as deaf and dumb people, half-deafness people.

The invention discloses a method for identifying key proteins based on characteristics of a structural domain. For all proteins of species, the method comprises the following steps of: obtaining structural domain information through using a database or carrying out analysis of protein sequences, counting the frequency of occurrence of each structural domain type in the proteins on the basis of the structural domain information, and calculating the weight of each protein according to the statistical result, wherein the bigger is the weight, the more possibility of the protein is to form the key protein. The method is simple and effective, is only based on the structural domain information of the protein and does not rely on the network information of protein interactions and the like, so that the huge amount of manpower and material resources consumed in biological experiments can be avoided. Compared with a random method, the method provided by the invention can be used for accurately identifying key proteins.

The invention discloses an unmanned aerial vehicle flight trajectory optimization method based on machine learning in an edge computing network. The method comprises the following steps: in a cycle, an unmanned aerial vehicle departs from a position where a server is deployed, sequentially passes through all ground terminals and completes a communication task, and then returns to the initial position; when the unmanned aerial vehicle takes off, a flight scheduling table is calculated according to the position distribution of the ground terminals, and then the unmanned aerial vehicle flies through all the ground terminals in sequence according to the flight scheduling table; and in the flying process of the unmanned aerial vehicle, a local optimal trajectory of the unmanned aerial vehicle is calculated by utilizing a random method and a machine learning method in an iteration mode, the local optimal trajectory is added into a global optimal trajectory q, and the unmanned aerial vehicleflies along the global optimal trajectory q. According to the method, the flight distance and the flight time of the unmanned aerial vehicle can be effectively reduced and shortened, the flight cyclecan be shortened, and the transmission efficiency can be improved.

The invention discloses a random write method and system based on an erasure code. The method comprises the steps of generating a random write file and uniquely marking the random write file and setting a file data security level of the random write file; inputting random write data in the random write file, striping the input random write data according to the erasure code and the file data security level, and calculating a random write data size and random write offset; obtaining a stripe list spanned by the random write of this time; calculating a range of data storage units of each strip needing to be updated and updating the data storage units in the range; and asynchronously recalculating and writing check data of all stripes updated by the random write file. The method and the system have the beneficial effects that when data write is carried out on the data storage units of the stripes, only partial data storage units are updated, and it is unnecessary to read the data of the whole stripes for updating, so calculation loads are reduced, and the random write amplification problem is solved by asynchronously recalculating the check data.

A method to install software by receiving a partially linked version of the software including a plurality of code modules, randomly ordering the code modules, assigning non-volatile memory addresses for the software and generating an executable image of the software. A system to install a software application having a partially linked object code version of the software application including a plurality of code modules and an installation program configured to randomly order the code modules, assign non-volatile memory addresses for the software, and generate an executable image of the software, wherein the installation program performs the ordering, assigning, and generating on a target computing device that is different from a development computing device that partially linked the software application.

The invention discloses a wireless sensor network rolling horizon signal-to-noise ratio estimation method with packet loss. The method comprises the following steps of: 1) obtaining a state space model of a wireless sensor network by use of a signal-to-noise ratio formula, a power control algorithm and a flow rate control algorithm; 2) modeling the packet loss into a random process of Bernoulli distribution according to the characteristics of packet loss to obtain a wireless sensor network model with packet loss; 3) transforming the rolling horizon signal-to-noise ratio estimation problem into an equivalent minimization problem according to the set rolling horizon window length N and the weight matrixes Pi, Q and R; and 4) solving the minimization problem based on a first-order optimization principle and a random method to obtain a wireless sensor network rolling horizon signal-to-noise ratio estimator, and obtaining the optimal estimated value and the predicted value of the next moment of the signal-to-noise ratio by the estimator. The method disclosed by the invention has a reasonable model and has functions of compensating packet loss and performing online calculation.

The present invention relates to a physical layout algorithm for circular pieces, comprising the following steps: firstly initializing algorithm parameters; reading radius sizes of the circular pieces and corresponding quantity information; numbering the circular pieces one by one; reading width information of a plate; setting a point in the lower left corner of the plate as a coordinate origin; setting a horizontal direction as x-axis; establishing a Cartesian coordinate system; disordering the numbered circular pieces with a random method; placing the correspondingly numbered circular pieces in the plate by using a circuitous arrangement algorithm according to a disordered sequence; establishing a rubber band to surround all the circular pieces to be laid out, so that the circular pieces are gathered toward the bottom of the plate under the shrinkage action of the rubber band to gradually realize a compact layout; calculating the plate utilization rate; judging whether a current layout result meets requirements or not or whether the number of iterations exceeds the maximum number of iterations or not; and if yes, outputting an optimal layout result. According to the present invention, the physical layout algorithm has the average plate utilization rate of more than 80%, is easy in realization of programming, and has better adaptability to various engineering application examples.

A random data generator for reproducing a random data stream producible by an identical generator at another location, comprising: an input means for receiving a bit stream, a random selector for selecting random individual bits from said bit stream to form a random data stream, and wherein said random selector is randomized by a previous segment of said random data stream. The generator allows process running at two separate locations to make use of the same random data.

The invention relates to a method for generating Gomoku playing data. The method comprises the following steps of setting a Gomoku forbidden rule and the size of a Gomoku chessboard; selecting AI which ranks top R in an existing Gomoku AI game, and setting the performance of the AI; designating different opening games; under the designated opening games, using the selected Gomoku AI to play Gomokuto generate the Gomoku playing data. According to the method for generating the Gomoku playing data, different AI with the high Gomoku playing ability in a Gomoku AI game is adopted, different high-level Gomoku AI plays Gomoku to generate the Gomoku playing data, and the quality and diversity of the Gomoku playing data in the later period can be improved. According to the method for generating the Gomoku playing data, different opening games are designated, a random method is adopted to improve the diversity of the opening games, and the diversity of the Gomoku playing data generated by the AI can be improved. Gomoku playing data of the chessboards with different sizes can be generated, and the generated Gomoku playing data can be increased according to requirements. The generated Gomokuplaying data can be used for training to obtain an effective neural network.