238 results about "Real arithmetic" patented technology

The invention discloses a sparse array antenna optimization method based on an improved zebra algorithm, is used for reducing the peak side lobe level of a sparse array antenna, and belongs to the technical field of radio frequency phased array antennas. The implementation process of the method comprises the following steps: firstly, adding a probability estimation model on the basis of an original zebra optimization algorithm, and converting a real number variable into a binary code; secondly, an adaptive T distribution disturbance strategy and a dynamic selection strategy are introduced, the global search capability of the algorithm is enhanced, the algorithm is prevented from falling into local optimum, and the convergence speed is effectively improved; and finally, an optimal solution is obtained through iterative optimization for a certain number of times. In order to verify the performance of the method, three groups of simulation experiments are carried out, and the result shows that compared with other optimization algorithms, the method can obtain a lower array antenna peak sidelobe level within shorter optimization time, so that the performance of the antenna is improved.

The invention relates to the technical field of artificial intelligence, and provides a medical knowledge question answering method, system and device based on a large language model, and the method comprises the steps: obtaining a question text inputted by a user; calling a quick question and answer language model to retrieve answers from a vector database according to questions proposed by a user, vector data in the vector database being real number vectors formed according to text fragments related to the digestive specialized knowledge, if the answer is retrieved, filling the question text and the retrieved answer into a prompt template to generate a standardized answer text, and outputting the standardized answer text to the user; and if no answer is retrieved, calling a deep reasoning language model to call an execution tool related to a reasoning task corresponding to the current question according to the question to perform online answer search, filling a question text and a search result into a prompt template to generate a second standardized answer text, and outputting the second standardized answer text to the user. Through cooperative work of the two intelligent models, the accuracy and the reply speed of medical knowledge questions and answers are considered.

The invention relates to the technical field of intelligent building data processing, and discloses a cantilever grouting beam construction process which is used for solving the problems of insufficient precision and potential safety hazards caused by data acquisition dispersion, model staticization and equipment control lag in traditional construction. The technology comprises the following steps: optimizing the node layout of an Internet of Things sensor based on a genetic algorithm, and generating high-coverage real-time monitoring data; constructing a digital twinborn model of multi-objective genetic optimization and exporting lightweight parameters; fusing Kalman filtering and LSTM prediction weight to generate an edge calculation control instruction; equipment parameters are cooperatively adjusted through real number coding, and the legality of the instruction is verified; optimizing data storage and a cross-chain tracing interface by adopting block chain integer coding; and dynamically updating the security early warning rule based on reinforcement learning. According to the method, dynamic acquisition, closed-loop optimization and credible storage of the construction data are realized, the construction precision and safety are improved, and cross-engineering experience reuse is supported.

The invention belongs to the technical field of intelligent optimization of intelligent robots, and particularly relates to a constraint optimization method for solving heterogeneous task allocation of multiple unmanned aerial vehicles. According to the method, a composite real number coding mode is adopted, and more information can be recorded in short individual codes; a self-adaptive alpha constraint level comparison strategy is adopted in the algorithm, relative feasible solution information can be effectively utilized in the iteration process, the convergence speed of the alpha level is adaptively influenced through the proportion of a relative feasible solution, the safety level is gradually improved, the search range of an optimal allocation scheme is limited within a satisfactory scheme, and the optimal allocation efficiency is improved. Therefore, an effective distribution scheme can be planned.

The invention provides a lead screw precision milling closed-loop optimization and real-time control process based on multi-axis linkage, which comprises the following steps of: acquiring geometric parameters of a workpiece, material attributes and kinetic parameters of a machine tool, designing a cutter and clamp scheme, and setting initial processing parameters based on a cutting database; constructing a lead screw spiral groove high-precision curved surface model by using an NURBS curve, and optimizing a path through curvature adaptive discretization and quintic spline interpolation; an idle stroke path is iteratively optimized by adopting a real number coding genetic algorithm, and the total time of the path and the reversing frequency are shortened; multi-axis synchronous control is achieved through inverse kinematics solution, and the cutter inclination angle is dynamically adjusted in real time based on the contact angle; and virtual simulation and finite element analysis are utilized to predict cutting force, thermal deformation and interference risks. According to the method, the continuous and smooth multi-axis synchronization instruction is generated, the positioning error of each axis is effectively reduced, meanwhile, the cutter inclination angle is dynamically corrected through force feedback of the contact angle, cutting force fluctuation is effectively restrained, and it is ensured that the tooth surface roughness of the spiral groove meets the requirement.

The embodiment of the invention provides a data transmission method and a related device. Specifically, a transmitting end acquires a superframe comprising a plurality of subframes, the subframes comprising training symbols and pilot symbols. Wherein in one polarization direction, the number of the training symbols in the sub-frame is NTS, the number of the pilot symbols in the sub-frame is NPS, both the NTS and the NPS are integers greater than 1, the sum of the NTS and the NPS is an odd number, and the sum of the NTS and the NPS is greater than or equal to 5. One symbol in the subframe is both a training symbol and a pilot symbol, each training symbol and each pilot symbol are one of four complex numbers of-A-Aj,-A + Aj, A-Aj and A + Aj, and A is a real number. Every continuous NPG symbols in the subframe comprise one pilot symbol located at a fixed position, and NPG is 96 or 128. And further, the transmitting end transmits the superframe. It can be seen that the interval of inserting the pilot symbols in the sub-frame is larger, and redundancy is reduced. In addition, one symbol in the sub-frame is both a training symbol and a pilot symbol, so that the structure of the sub-frame is simplified.

The embodiment of the invention discloses a precision positioning acceleration method and device based on a low earth orbit satellite / LSTM neural network. The method comprises the steps that a terminal obtains a wide lane phase decimal deviation product, a narrow lane phase decimal deviation product and a first GNSS combination observation value which are generated by a server based on an LSTM method and are subjected to coding processing; determining a first GNSS wide lane real number ambiguity according to the first GNSS wide lane real number Based on a precise point positioning algorithm, determining initial GNSS ionosphere elimination combination real ambiguity and other parameters; performing inter-satellite single difference on the first GNSS wide-lane real number ambiguity, and obtaining the integer ambiguity by combining wide-lane phase decimal deviation product fixation; performing inter-satellite single difference on the initial GNSS ionosphere-free combination real ambiguity, determining a first GNSS narrow-lane real ambiguity by combining the wide-lane integer ambiguity, and fixing the first GNSS narrow-lane real ambiguity by using a narrow-lane phase decimal deviation product through a partial ambiguity fixing method to obtain the integer ambiguity of the narrow-lane phase decimal deviation product; and calculating a user coordinate fixed solution in combination with related parameters.

The present subject relates to a method for proving a statement describing a functionality of an electronic circuit, using a quantum computer and a proof assistant.The procedure comprises: encoding a current proof state into a vector of real numbers of a fixed length, where the current proof state defines a task to prove at least part of the statement; encoding the vector into a quantum state of a quantum system of the quantum computer; using the quantum state as an input quantum state by a quantum machine learning model to provide an output quantum state, the measurement of which constitutes a proof step for the defined task; measuring the output quantum state, thereby obtaining the proof step for the defined task; providing the proof step to the proof assistant; and receiving a next proof state from the proof assistant in response to the provision of the proof step.

The invention discloses a satellite-ground fusion end-side cloud collaborative computing unloading method, and belongs to the technical field of wireless communication. In order to solve the problem of high task completion time delay caused by large ground processing time delay and limited satellite processing capability, the invention provides an end-side cloud collaborative computing unloading method through marginal computing. In the task modeling stage, modeling is carried out on the split calculation sub-task dependency relationship in combination with a linked list and a directed acyclic graph, and a generated task model can effectively reduce task processing delay. In the calculation unloading stage, task completion time delay is defined according to a node unloading strategy and a link transmission rate, and a time delay list based on task unloadable nodes is calculated and generated by adopting a complete unloading strategy; under the constraint of satellite remaining coverage time, node computing power and link unloading bandwidth, a real number coding genetic algorithm based on task completion time delay is provided to find an optimal unloading point of a computing task, and a formed optimal computing unloading strategy can effectively reduce time delay and improve the task completion rate.

The invention discloses a parameter serial and parallel mixed input complex multiplication implementation method, which is implemented by adopting a complex multiplier circuit, and the complex multiplier circuit comprises a serial input cache unit, a parallel input cache unit, a real number multiplication unit and a real number addition / subtraction unit. The method comprises the specific implementation steps that S1, serial input complex number parameters and parallel input complex number parameters are sequentially input from a serial input cache unit and a parallel input cache unit respectively; s2, the serial input complex parameters and the parallel input complex parameters are subjected to combined calculation through a real number multiplication unit and a real number addition / subtraction unit; and S3, outputting a real part and an imaginary part of a complex multiplication result in sequence in a serial mode. Compared with a traditional and optimized triple multiplier version complex multiplication operation circuit, only a small number of registers are additionally arranged, latch resources are conditionally selected, and the complex multiplication operation circuit has obvious advantages in hardware resource consumption.

The invention provides a data processing method for SIMD (Single Instruction Multiple Data)-oriented parallel iterative solution, which comprises the following steps: acquiring to-be-processed data (including 2m pattern matrixes in a complex form), storing each pattern matrix in a real number newCSR format (including a ValuesR array, a ValuesI array, a rowPtr array and a colIndices array), and vectorizing to obtain a vectorized fused newCSR format (including a newValuesR array, a newValuesI array, a rowPtr array and a colIndices array), and reading vectorized new CSR format fused data by using an SIMD extension component, carrying out parallel iterative solution, and outputting an iterative solution result. In this way, parallel iterative solution calculation can be carried out through the SIMD extension component in the SIMD architecture, and accelerated solution is achieved.

The invention relates to the technical field of transformers, in particular to a transformer parameterization design method, device and equipment, and the method comprises the following steps: solidifying a transformer design standard and specification in an Excel worksheet, and defining a control parameter through a set compiling grammar; the control parameters comprise a real number type, an integer type, a character string type and a Boolean type; one writing grammar is set for each parameter type; exporting control parameters in the Excel worksheet as an XML (Extensible Markup Language) file; creating a notebook file in automatic design software, and importing the control parameters in the XML file; based on the imported control parameters, parametric modeling is carried out in automatic design software, and a three-dimensional model of the transformer is generated; and automatically generating a two-dimensional engineering drawing of the transformer according to the three-dimensional model. Standardization of links such as design, manufacturing, installation and operation of the transformer is effectively ensured, and therefore performance and reliability of the transformer are greatly improved.

The invention discloses a space target monitoring constellation configuration optimization method and system considering complex constraints, and belongs to the technical field of aerospace mission design and optimization. The method comprises the following steps: firstly, constructing a discrete orbit modal library meeting regression orbit dynamic resonance conditions; constructing generalized hybrid design variables consisting of integer indexes and real number fine tuning quantities on the basis of the library; establishing an optimization model taking the space volume coverage rate, the maximum revisit time and the system construction cost as targets and taking illumination, dynamics and maneuvering capability as constraints; performing rapid analysis and coarse screening on invalid links through track geometry; and finally, performing global optimization by using an improved non-dominated sorting genetic algorithm to obtain a Pareto optimal constellation configuration set. According to the method, accurate evaluation of the three-dimensional space target coverage efficiency is realized, the physical feasibility and long-term stability of a configuration scheme are ensured, and a high-dimensional multi-target optimization problem is efficiently solved.

The invention discloses a method and a device for calculating Gamma correction in an RISC-V processor, which can improve the execution efficiency and the calculation precision of a Gamma correction algorithm on the RISC-V processor. Gamma correction is divided into six levels of pipeline structures, wherein the first level is composed of a floating point adder and a floating point shifter, and input pixel values in an integer range from 0 to 255 are mapped to a real number range from 0 to 1; the second stage is used for carrying out reciprocal operation on the Gamma value; in the third stage, logarithm operation is carried out on input pixel values, and a lookup table and piecewise linear approximation method is adopted; the fourth stage uses a floating point multiplier to perform product operation on the result of reciprocal operation and the result of logarithm operation; the fifth stage is used for performing exponential operation on the result output by the fourth stage; and the sixth level is composed of a floating point shifter and a floating point subtractor, and image pixel values after Gamma correction are remapped to an integer range of 0-255.

The invention discloses an indoor nuclear environment multi-information double-layer hybrid map modeling method. The method comprises the following steps: 1, acquiring a grid map containing grid elevation values; 2, establishing an IOP model to obtain a geometric grid map; 3, acquiring a radiation field grid map based on MCNP or a difference algorithm; 4, constructing a multi-information double-layer hybrid map: carrying out binary coding on the trafficability of each grid in the geometric grid map; coding the average radiation dose rate of each grid in the radiation field grid map by adopting a real number; and after the grid information is coded, the grid information is stored in a group number form to form a mapping relation. The invention provides a modeling method of a multi-information double-layer hybrid map for path planning of a mobile robot in a radioactive indoor environment, and the map is a grid map fusing trafficability information and environment radiation field information, and can describe multi-physical field coupling environment characteristics in an indoor nuclear environment. And a reliable environment map is provided for subsequent path planning of the nuclear robot.

A method, apparatus and computer program product for privacy-preserving homomorphic inferencing. In response to receipt of encrypted data, a ciphertext of real numbers is generated. Each real number has an associated sign that is desired to be maintained. A mask is then identified, preferably via an iterative algorithm that works on a trial and error basis to locate an appropriate solution. The mask comprises set of values randomly distributed over a given positive range and that remain positive after encoding under a fixed-point arithmetic and with a low scale value. Under homomorphic encryption, the ciphertext is then multiplied by the mask to generate a result comprising values corresponding to the real numbers in the ciphertext and that maintain their associated signs. The result is provided as a response to the encrypted data.

There is provided a method of computing lineage, comprising: managing a dataset of records, each record associated with set(s) of vectors of real numbers that encode an approximation of lineage of the respective record, the set(s) of vectors computed by an encoding process, obtaining result record(s) in response to executing a query on the dataset, computing set(s) of vectors for the result record(s), searching the set(s) of vectors on the records of the dataset to identify a record associated with a subset of vectors that are statistically similar to the set(s) of vectors for the result record(s), and providing a subset of the records corresponding to the identified subset of records, the subset of the records having a likelihood of contributing to the existence of the result record(s) in response to execution of the query.

The invention provides a multi-zone-area ordered charging scheduling method and system based on a simulated annealing genetic algorithm, and the method comprises the steps: generating an initial population through employing a real number coding mode, and enabling a population individual to represent a charging scheduling scheme; constructing a fitness function in the genetic algorithm by taking load balancing and network loss minimization as optimization targets; selecting individuals by adopting a roulette selection and elitism strategy, and reserving a plurality of individuals with the highest fitness to enter the next generation by adopting a self-adaptive crossover mutation operation; taking the optimal individual of the current population as an initial solution, and executing simulated annealing operation; after optimization of a preset number of iterations, an optimal charging scheduling scheme is output, and the load rate, the voltage and the charging pile state of each transformer area are monitored in real time; and when abnormality is detected, charging power distribution is dynamically adjusted, and the load of the overload area is transferred to the low-load area. Through fusion of simulated annealing and a genetic algorithm, multi-zone-area charging scheduling is optimized, load balancing and network loss minimization are realized, and the stability and dynamic adaptability of a power grid are improved.

The invention relates to the field of vehicle route planning, in particular to a dynamic multi-target vehicle route planning method and system based on prediction and recombination, and the method comprises the steps: building a mathematical model of dynamic multi-target vehicle route planning, initializing a population through real number coding, adjusting the population to meet a constraint, carrying out the iteration through a genetic algorithm, and obtaining a candidate solution set; a Pareto front set is obtained and clustered, and a clustering center is recorded as a special individual; and when environment change is detected, predicting individuals in a new environment by using a gray model or a least square support vector machine, and constructing a new population until an optimal solution set is output. According to the method, the convergence speed of the algorithm is high, the response to the environment change is more sensitive and timely, the dynamic multi-target vehicle routing problem is effectively solved, and the application effect and the economic benefit of the vehicle routing in the actual life are effectively promoted.

The invention provides a numerical control multi-target energy-saving optimization method and system and a storage medium, and the method comprises the steps: constructing a mixed integer linear programming model for multi-target flexible job shop scheduling, taking machine tool distribution, a machining sequence, tool distribution and a machining speed as decision variables, and taking total weighted tardiness and total energy consumption as optimization targets; a single-layer real number coding method is adopted, machine tool allocation, the machining sequence and the machining speed are coded in a one-dimensional real number sequence in a unified mode, an initial population is generated, the mixed integer linear programming model is solved through a multi-target optimization algorithm, and an algorithm optimal solution of the mixed integer linear programming model is obtained; decoding the optimal solution of the algorithm, distributing tools for each process, and generating a production plan table; and calculating the total weighted tardiness and the total energy consumption to obtain an algorithm energy-saving optimization scheduling scheme. According to the method, the dual-objective mixed integer linear programming model is constructed, and the innovative coding and decoding strategy and the dynamic optimization mechanism are combined, so that collaborative optimization of the production efficiency and the energy consumption is realized.

The invention provides a multi-component spectral feature selection method based on double-layer optimization, and aims to solve the problem of model performance reduction caused by high dimension, redundancy and noise interference in hyperspectral data. The method comprises the following steps: performing binary and real value vector double-layer coding on individuals, and randomly generating an initial population based on a fitness value; and then, calculating an importance score delta of each spectral variable in evolutionary iteration in combination with a filtering type Relief feature selection method so as to dynamically guide population optimization. A Gaussian function dynamic noise reduction method based on importance scores is designed in the optimization process, and interference noise existing in spectral variables is optimized and fitted through real number vectors; and finally, obtaining an optimal spectral feature subset through an environment selection strategy. According to the method, the data dimension can be effectively reduced, noisy and redundant spectral variables are eliminated, and the accuracy of the model is remarkably improved.

A method comprises creating an electronic circuit design having a plurality of electronic components, creating an analog simulation model of the electronic circuit design, and executing the analog simulation model to generate one or more simulation logs representing simulated operation of the electronic circuit design. The method also comprises generating a neural network model based on the one or more simulation logs, the neural network model comprising a plurality of weights and generating a mathematical simulation model based on the neural network model.

The invention discloses a probabilistic language three-way decision-making-based telecommunication potential customer evaluation method. The method comprises the steps of firstly grouping target information according to decision requirements, converting a real number data set into a probability language decision matrix, extracting probability language term set features and constructing an expectation-reliability score function; respectively obtaining subjective and objective weights of attributes through an analytic hierarchy process and a variable coefficient method, and solving a comprehensive weight by means of a cosine similarity maximization model; then, a relative loss function is constructed based on the score matrix, a comprehensive foreground value of the scheme is calculated, and (non) harmony indexes and net credibility indexes of the scheme are measured and calculated under each attribute in combination with attribute non-compensability and an ELCTRE thought; and finally, making a series of decision rules through the expectation relative loss function. According to the method, the telecommunication company can be assisted to efficiently classify and evaluate the users, scientific guidance is provided for decision making, and the accuracy of potential customer mining and management is improved.

Artificial-intelligence computer-implemented processes and machines predict whether personal data may be present in structured software based on metadata field(s) contained therein. Natural language processing preprocesses input strings corresponding to the metadata field(s) into normalized input sequence(s). Individual characters in the sequence(s) are embedded into fixed-dimension vectors of real numbers. Bidirectional LSTM(s) or other machine-learning algorithm(s) are utilized to generate forward and backward contextualization(s). Neural network output(s) are provided based on element-wise averaging or feed forwarding based on the contextualization(s) in order to predict whether one or more value fields corresponding to the metadata field(s) may contain personal data.

The invention discloses a field programmable gate array (FPGA) implementation method for Hermite matrix eigenvalue decomposition, which comprises the following steps of: performing real number processing on a Hermite matrix according to a corresponding relationship between the Hermite matrix and eigenvalues and eigenvectors; grouping the dimension real matrix obtained by the real number processing; calculating rotation matrixes of diagonal units and non-diagonal units; performing Jacobi rotation operation on the diagonal units and the non-diagonal units, and updating the feature vector matrix according to a rotation matrix; after a group of Jacobi rotation is completed, data exchange is carried out according to grouping of a parallel Jacobi algorithm; and multi-stage cleaning can be carried out according to actual requirements. According to the method, a parallel Jacobi algorithm is selected, a series of matrix rotation is adopted, the matrix is converted into a diagonal matrix, and therefore the eigenvalue and the eigenvector of the matrix are calculated. According to the method, the parallel computing advantages of the FPGA and the Jacobi algorithm are fully utilized, and the real-time performance of the algorithm is effectively improved.

The invention relates to the technical field of electromagnetism, in particular to a multi-layer slab waveguide mode processing method, device, equipment, medium and program, and the method comprises the steps: constructing a transfer matrix model of a multi-layer slab waveguide; obtaining boundary conditions of the infinite space, and determining a corresponding real number domain characteristic equation and a complex number domain characteristic equation according to the transfer matrix model and the boundary conditions so as to determine a guided mode and a leakage mode of the multilayer slab waveguide in the infinite space; introducing a perfect matching layer to convert a radiation condition of an infinite space into an absorption boundary in a finite computational domain, constructing a finite space characteristic equation according to the thicknesses of the perfect matching layer and the absorption boundary, and positioning a plurality of correction leakage modes according to the finite space characteristic equation to generate a Bereger mode; and generating complete mode distribution according to the guided mode, the leakage mode and the Bereger mode. Therefore, the problems of relatively poor calculation precision, relatively slow calculation speed, limitation of complete mode solving in the multilayer slab waveguide and the like in related technologies are solved.

A secret share value [y]=[δx2+ax] is obtained through secure computation using a secret share value [x] of a real number x, and a secret share value [func(x)]=[y(ζy+b)+cx] of an elementary function approximation value z=func(x) of the real number x is obtained and output through secure computation using secret share values [x] and [y]. Here, x, y, and z are real numbers, a, b, c, δ, and ζ are real number coefficients, and a secret share value of · is [·].

Disclosed is an operation method for executing an addition operation on quantized matrices. This operation method comprises the steps in which: an NPU matrix multiplication unit executes a matrix multiplication operation on a first matrix and a second matrix, wherein the first matrix has a size of R0*2C0 and is obtained by concatenating a first input tensor and a second input tensor, each of which is composed of quantized values and has a size of R0*C0, and the second matrix is obtained by concatenating a first diagonal matrix having a size of C0*C0 and having all main diagonal element values equal to integer N1 and a second diagonal matrix having a size of C0*C0 and having all main diagonal element values equal to integer N2; and an NPU vector processing unit calculates elements of a non-quantized tensor obtainable by adding the second input tensor to the first input tensor, by adding real number B to a product of real number R and a third matrix calculated by the matrix multiplication operation.

A method for detecting the cliques of evaluators in a decision-making process includes collecting a data matrix Sij to a computer system using an automatic input interface. The elements of the matrix Sij are the real numbers in a predefined range. Each of the elements of the matrix Sij represent a numerical evaluation provided by an evaluator j for an evaluated entity i. For each pair of evaluators j, where j1 and j2 (j1≠j2) from the data matrix Sij, calculating by the computer system the Pairwise Adjusted Distances EDj<sub2>1< / sub2>j<sub2>2< / sub2>. Applying by the computer system a nonlinear transformation to the Pairwise Adjusted Distances EDj<sub2>1< / sub2>j<sub2>2< / sub2>. Identifying cliques of evaluators by comparing by the computer system the transformed distancesEDj1⁢j2*to a robust lower threshold.

The embodiment of the invention provides a modulation method, a storage medium, an electronic device and a computer program product, and the method comprises the steps: shifting a first data sequence, and carrying out the addition operation of the first data sequence and the shifted first data sequence, and obtaining a second data sequence; multiplying each data element of the second data sequence by a negative real number, and sequentially inserting between every two adjacent data elements of the first data sequence to obtain a third data sequence; and transmitting the third data sequence. Through the embodiment of the invention, the problem that the low PAPR requirement of the communication system in the related technology cannot be met is solved, and the effect of reducing the PAPR of the communication system is achieved.