104 results about "Parallel optimization" patented technology

The invention belongs to the technical field of additive manufacturing, and particularly relates to a multi-scale parallel optimization method for a lightweight structure based on lattice discrete optimization. According to the method, through multi-scale parallel optimization design, a macro structure and a micro lattice structure are described through geometric parameters of a movable deformation rod piece, an improved two-value coding parameterization method, called a BCP method for short, is adopted to solve discrete optimization of the micro lattice structure, and geometric parameters of an optimization result are easy to extract.

The present application belongs to the technical field of additive manufacturing, and particularly relates to a lightweight structure multi-scale parallel optimization method based on lattice discrete optimization. The present application uses the geometric parameters of movable and deformable rods to describe macroscopic structures and microscopic lattice structures through multi-scale parallel optimization design, and uses an improved double-value coding parameterization method, referred to as BCP method, to solve the discrete optimization of the microscopic lattice structure and easily extract the geometric parameters of the optimization results.

The invention provides an attitude control method, system and equipment of an underwater leveling machine and the underwater leveling machine, and relates to the technical field of ocean engineering. The method comprises the following steps: establishing a dynamic model; constructing a physical information neural network based on the dynamic model; constructing a digital twinborn body of the leveling machine, and performing simulation verification on the attitude error prediction result in the digital twinborn body; based on a simulation verification result, taking control parameters of the underwater leveling machine as optimization variables, performing parallel optimization on a plurality of control targets of the underwater leveling machine, and generating a control solution set; and determining a target control parameter combination from the control solution set, and generating a control instruction for controlling the underwater leveling machine to perform attitude adjustment based on the target control parameter combination. According to the method, the reliability of attitude control of the underwater leveling machine can be improved through prediction of the integrated physical information neural network, simulation verification of the digital twin and multi-objective optimization.

The invention provides a sewage plant water-energy-drug collaborative optimization method and system, and the method comprises the steps: obtaining the data of a technological process and a material transfer relationship of a sewage plant, constructing a graph network structure model with a technological unit as a node and material flow as an edge, and carrying out the preprocessing, thereby obtaining dynamic coupling graph structure data; a dynamic coupling graph neural network model containing a node feature coding layer, a time sequence coding module, a space message passing layer and an attention mechanism layer is constructed based on the data, and a prediction model capable of representing the dynamic coupling relation of the process unit is obtained through historical data training; then, a multi-objective optimization function which takes the lowest ton water treatment cost as an objective and covers water quality standard reaching, energy consumption and medicament dosage constraints is constructed; and in combination with the prediction model and the optimization function, the optimal operation parameters are solved through an optimization algorithm, and a whole-plant collaborative optimization decision scheme is generated. According to the invention, the dynamic coupling GNN prediction model and the virtual element multi-domain parallel optimization technology are integrated, and intelligent operation management of the sewage treatment plant is realized.

The invention discloses a Fortran program parallel optimization method based on intelligent dependency analysis. The method comprises the following steps: firstly, constructing a parallel optimization system consisting of a loop extraction module, a loop nested relation analysis module, a semantic analysis module, an intelligent dependency analysis engine, a variable classification module and an instruction generation and injection module; the loop extraction module analyzes a nested relation and a variable action range of loops; the loop nesting relation analysis module determines a nesting relation between loops; the semantic analysis module constructs a row-level data access view; the variable classification module identifies private variables and reduction variables; the intelligent dependency analysis engine executes loop type check, I / O operation check and data dependency check; and the instruction generation and injection module generates a parallelization instruction and inserts the parallelization instruction into the source code to obtain a parallelized program source code. The method can solve the problems that an existing parallelization method is low in cyclic dependency relation recognition accuracy and safety, and parallelization errors cannot be accurately recognized.

The application relates to the technical field of server data storage, and discloses a processing system for improving server data storage speed, which comprises a data blocking module, a parallel optimization module, a performance monitoring module, a strategy matching module and a storage execution module, and can comprise a load self-learning module. The data blocking module blocks data streams according to a dynamic strategy and generates a distribution queue; the parallel optimization module generates an optimal storage node combination and a scheduling strategy according to the queue; the performance monitoring module collects real-time performance data and screens for abnormalities; the strategy matching module determines a target adjustment rule through multidimensional matching; the storage execution module drives a node to correct a path; and the load self-learning module optimizes a scheduling factor according to feedback data. The system realizes dynamic blocking, adaptive scheduling and real-time optimization, improves storage efficiency and stability, and is suitable for a server data storage scene.

The present application relates to a kind of efficient parallel optimization method and device considering the actual engineering constraint of horizontal well, method includes the following steps: S1 constructs water drive reservoir production optimization mathematical control model;S2 based on horizontal well engineering application condition and reservoir numerical model, the constraint judgment method of horizontal well is constructed;S3 based on the population in Latin hypercube sampling initialization algorithm, using the constraint judgment method of horizontal well is handled, database is constructed;S4 using the excellent individual in database forms temporary population, based on temporary population, Gaussian model is constructed;S5 based on differential evolution algorithm with different strategies obtains three sub-populations, and using the constraint judgment method of horizontal well is handled;S6 based on the pseudo-update strategy of temporary population, the potential individual of sub-population is obtained;S7 based on Euclidean distance, select the individual closest to current optimal individual, and carry out parallel numerical simulation;S8 updates database.

The invention provides a method for reasoning by using a large language model, a control device and a storage medium, and belongs to the technical field of large language models. The method comprises the steps that in the reasoning process of a target large language model, a weight matrix and input data of the target large language model are segmented into a plurality of data blocks; sequentially loading each data block in the plurality of segmented data blocks into a memory, and carrying out parallel optimization calculation on each data block; and combining the calculation result corresponding to each data block to obtain a reasoning result. Through block parallel computing, the reasoning performance of the large language model in a CPU environment can be remarkably improved, and an efficient and flexible solution is provided for large language model reasoning in a resource-constrained environment.

The invention discloses a permanent magnet synchronous motor multi-target model prediction control method based on parallel optimization, which relates to the field of motor control, and sequentially comprises the following steps: establishing a permanent magnet synchronous motor mathematical model, performing Euler discretization-based model prediction torque control, and selecting an optimal vector based on a comprehensive optimization mechanism. Compared with the prior art, the optimal vector is selected on the basis of a comprehensive optimization mechanism, so that the control system adaptively switches the priority of multiple targets under high-speed and low-speed working conditions so as to realize multi-dimensional variable global optimization and rapid dynamic coordination, and the control is enabled to meet the performance requirements of preliminary constraint conditions, so that the control efficiency is improved. Ripples of torque and flux linkage are reduced, unstable operation of the motor and extra noise interference are overcome, and the control precision and steady-state performance of a control system are improved.

The application relates to an SGD-PSO-based lens parameter generation method and device and a storage medium, wherein the method comprises the following steps: S1, a basic configuration of a lens in an optical system is acquired, the number of lenses, design requirements and a parameter set are obtained based on the basic configuration of the lens; S2, an optimization target is constructed based on the design requirements, a fitness function is defined according to an optimized objective function, and a particle is defined based on the parameter set, wherein the position of each particle represents a group of values of the parameter set; S3, optimization is performed in an SGD-PSO mode to obtain an optimization result; and S4, the values of the parameters in the parameter set are obtained based on the optimization result. Compared with the prior art, the global information prompt multi-particle parallel optimization algorithm proposed in the application can fully explore the solution space and improve the efficiency of the computing power, and the optimization result, efficiency, stability and diversity are greatly improved through the multi-system independent parallel local optimization algorithm, the heuristic global search algorithm and the gradient-enhanced global algorithm.

The application discloses a two-purpose wrench production full-process intelligent management method and system under an industrial internet of things architecture, and relates to the technical field of industrial internet of things and intelligent manufacturing. The method comprises the following steps: taking the use characteristics of a target two-purpose wrench as a constraint, obtaining single-function production sample data and full-process historical production data; constructing and training a plurality of function performance evaluation bodies based on the data to form an evaluation body set; constructing a bidirectional master-slave optimization channel comprising a first master-slave optimization channel, a second master-slave optimization channel and an information exchange channel according to the evaluation body set; performing bidirectional parallel optimization iteration in combination with the full-process historical production data and the bidirectional master-slave optimization channel, and outputting an optimal process scheme; and finally applying the optimal scheme to a full production environment to realize intelligent management. The application effectively solves the coupling conflict problem of dual-function end production process parameters through a bidirectional optimization architecture, realizes global collaborative optimization, improves product quality and production efficiency, and reduces production cost and the waste product rate.

This invention discloses a parallel optimization system for massive video processing. The system adopts a four-layer integrated parallel and collaborative processing architecture, comprising, from top to bottom: a parallel task layer for video stream access, grouping, splitting, encapsulation, and queue management; a resource abstraction layer for unified hardware modeling, status acquisition, topology construction, and capability assessment; a parallel scheduling layer for task-hardware matching, load balancing, priority scheduling, and dynamic adjustment; and an execution optimization layer for data stream localization, parallel read / write, cache optimization, and zero-copy processing. The parallel task layer, resource abstraction layer, parallel scheduling layer, and execution optimization layer form a complete parallel processing link: task input, resource awareness, accurate scheduling, and optimized execution. This invention implements concurrent rate limiting and smooth access for the input video stream to prevent traffic surges; and sets synchronization points and timing control for parallel tasks to ensure orderly output.

The present application relates to a kind of crack morphology parameter's multiple initial value subinterval nonlinear inversion method, it is applied to petroleum technical field, the multiple initial value subinterval nonlinear inversion method for explaining crack morphology parameter based on crack front fiber strain includes: based on three-dimensional displacement discontinuity method, crack front fiber strain calculation model is established;According to crack parameter and the crack front fiber strain calculation model, predicted strain is calculated;According to the strain residual of measured fiber strain and the predicted strain, nonlinear least square inversion model with parameter boundary constraint is established;The crack parameter includes crack half crack length, crack height, crack net pressure in crack, crack vertical offset;Under coarse grid division, multiple initial values are set in parameter feasible region and are carried out parallel optimization, and first candidate solution is obtained;Again with the crack height of the first candidate solution as center, crack height feasible region is recursively divided into multiple subintervals.

The application discloses a kind of based on machine learning's genome analysis parallel optimization method and system, the method of the present application includes the input BAM file is cut into multiple same size file blocks;For each file block, extract file block characteristics, and input the file block characteristics into the pre-trained machine learning model to obtain the predicted running time of the file block;The predicted running time is in the front part file block is divided into smaller file block;File block is input into HaplotypeCaller in parallel to carry out variation detection;The variation detection result generated by HaplotypeCaller for each file block is merged and output.The present application is aimed at solving the problem of serious calculation tilt and low resource utilization caused by unclear calculation complexity before HaplotypeCaller runs, improving the efficiency of genome analysis, shortening the variation detection duration.

This invention relates to the field of chemical engineering technology, and in particular discloses a parallel optimization method for multivariate alarm values ​​that integrates intelligent identification of operating conditions in chemical plants. The method includes: S1 multi-source data access and preprocessing to achieve unified integration and standardization of data from systems such as DCS and MES; S2 analysis of the preprocessed data based on machine learning to intelligently identify and classify steady-state operating conditions; S3 parallel multivariate correlation analysis, calculation of frequent fluctuation ranges, and effective intervention time for each steady-state operating condition; S4 generating optimized alarm setpoints for different operating conditions based on the calculation results, and displaying a comparison of alarm frequency and optimization effects through a graphical interface; S5 approval, deployment, and application of alarm schemes. This invention combines the advantages of data-driven approaches with the security and interpretability of static alarms, solving the problems of isolated optimization and data distortion, and significantly improving the accuracy and reliability of the alarm system without high-risk dynamic intervention.

The application provides an image data processing method and system based on a lightweight sparse neural network, and relates to the technical field of machine learning, comprising: obtaining image data to be processed; sequentially performing random flipping, random cropping data enhancement and tensor standardization preprocessing on the image data to obtain standard tensor data meeting model input requirements; inputting the standard tensor data into a lightweight sparse neural network model; the model extracts local edge and texture features and global semantic features from the image through multiple sparse convolution layers; the physical features extracted are mapped to class confidence through a full connection layer, and finally a classification result is output through a Softmax function. Through the gradient-guided differential mutation strategy and GPU parallel optimization, the application realizes efficient image processing on resource-limited mobile terminals, embedded devices and edge computing devices, reduces the calculation overhead and memory occupation, and at the same time maintains high-precision classification performance.

The application discloses a trajectory data preparation method based on federated learning, which utilizes a unified privacy protection framework designed by federated learning to realize trajectory data preparation in a federated environment by using the function of a large language model. Meanwhile, the application designs a trajectory privacy automatic encoder to ensure data transmission security and protect privacy, and introduces a trajectory knowledge enhancer to improve model learning of knowledge related to trajectory data preparation, so that development of the large language model for trajectory data preparation is realized. In addition, the application also proposes federated parallel optimization to improve training efficiency by reducing data transmission and realizing parallel model training.

The invention relates to the technical field of computer aided design, and discloses a precise mold structure parameterization rapid design system based on multi-source data fusion. The system comprises a data acquisition module, a time sequence alignment module, a parameter calculation module, a structure generation module and a verification feedback module, time sequence alignment and fusion are carried out on multi-source heterogeneous data, a multi-target genetic algorithm is adopted to optimize mold key parameters in parallel, and parameter closed-loop correction is realized based on virtual combined simulation. According to the method, the consistency, precision and reliability of mold structure design can be effectively improved.

The invention discloses a chassis lining rigidity parallel optimization method and system capable of resisting energy deviation, and relates to the technical field of whole vehicle NVH optimization, and the method comprises the steps: carrying out the modeling of a road noise CAE simulation model, carrying out the analysis of a transmission path, calculating the contribution amount, arranging contribution paths in a descending order, and generating a parameterized lining file according to the linings of the contribution paths; performing dynamic polycondensation on the non-optimized region to form a CMS super-unit, generating a super-unit file, constructing a super-unit hybrid architecture based on the super-unit file and the parameterized lining file, and performing precision verification; and calculating a sound pressure value of a target point in the vehicle, converting the sound pressure value into an A weighting sound pressure level, and constructing an anti-energy offset optimization objective function based on a dynamic weighting superscale function of a superstandard frequency point and an adjacent frequency point correlation penalty function.

The invention discloses a serial interference cancellation sequence configuration method and device based on probability iterative optimization, and relates to the technical field of wireless communication, and the method comprises the following steps: initializing a device set and algorithm parameters; discretizing decoding positions in the same frequency spectrum through binary coding, and compressing search dimensions; generating a candidate SIC sequence sample set based on the adaptive probability distribution; performing mapping and conflict processing on the samples, and screening feasible samples; evaluating the performance of a feasible sample by using a complex system model, and mapping the performance into a single evaluation scalar; screening elite samples and iteratively updating probability distribution; and judging probability distribution convergence, and outputting an optimal SIC sequence. Through a probability-driven sample generation and conflict processing mechanism, the method does not depend on an explicit mathematical model, can adapt to a complex system model, remarkably reduces the calculation complexity, meets the parallel optimization requirements in a multi-device and multi-spectrum-cluster scene, and further reduces the calculation complexity.

The invention discloses a picking robot-oriented redundant mechanical arm multi-path-point parallel inverse kinematics optimization method, belongs to the technical field of robot motion control and path planning, and solves the problem that an existing method cannot optimize an inverse kinematics solution of a redundant mechanical arm at multiple path points with trajectory constraints. The method comprises the following steps: establishing a kinematic model of the robotic arm, and constructing a multi-objective parallel optimization cost function J; a hybrid GA-GWO parallel optimization algorithm is adopted to initialize a population and encode the population, iterative optimization is carried out based on the hybrid GA-GWO parallel optimization algorithm in combination with a hierarchical hybrid strategy, and a joint angle sequence corresponding to an individual with the optimal fitness is output; according to the method, through a multi-waypoint parallel optimization mechanism, the attitude before grabbing and the grabbing attitude are taken as a whole for joint solving, so that the planning success rate in a complex limited space is improved, the generated tail end trajectory strictly conforms to linear constraint, and the collision risk caused by trajectory deviation is effectively eliminated.

The invention discloses a shear flow simulation method based on molecular dynamics, and relates to the technical field of computational fluid mechanics and molecular simulation. The method comprises the following steps: determining a target fluid; a non-equilibrium molecular dynamics core framework is constructed, and molecular motion is simulated according to a multi-scale coupling strategy and a parallel optimization algorithm; the unbalanced molecular dynamics core framework is composed of an MARTINI coarse graining model, a boundary driven shearing algorithm and an SLLOD algorithm; collecting related parameters of the molecular motion, and inputting the related parameters into a slip length prediction model for processing to obtain a slip simulation value of the target fluid; wherein the slip length prediction model is constructed according to the quantitative relation between the slip length and the wall surface wettability and the shear rate. The simulation efficiency can be improved.

This invention discloses a system and method for optimizing the configuration of a space target surveillance satellite constellation. The system includes: a constraint modeling and configuration module, which uniformly models the detection constraints and observation constraints in the space target surveillance process and constructs an observable judgment model; a surveillance performance evaluation index module, which constructs various surveillance performance evaluation indices; a multi-objective parallel optimization module based on a determinant-based point process, which, based on a non-dominated sorting genetic algorithm, introduces a determinant-based point process to model the diversity of the non-dominated solution set, optimizes the selection of individual quality and solution set diversity by maximizing the determinant of the subset of the kernel matrix, and completes efficient screening using a greedy approximate inference method; and a constellation configuration optimization solution module, which constructs a multi-objective optimization problem with the surveillance performance evaluation index as the optimization objective, calls the multi-objective parallel optimization module based on a determinant-based point process for joint optimization, and obtains the optimal or near-optimal constellation configuration scheme that satisfies multiple constraints.

The application discloses a kind of converter electromagnetic transient simulation parallel optimization method, equipment and storage medium, belong to power system electromagnetic transient simulation technical field, by merging significantly reduces unnecessary intermediate variable, shortens simulation calculation process;And since C, D matrix and Un, k matrix multiplication operation with Itotal of A, B matrix and multiplication are carried out simultaneously, and the calculation process between row and row is independent, increase the parallel degree of electromagnetic transient simulation calculation;Since parameter matrix does not change with the change of switch state, it is calculated and stored in advance in simulation initialization stage, only needs to consume register reading operation time in the calculation of single simulation step;Coefficient matrix A, B, C, D are all square matrix of scale Nb*Nb, the formula for calculating Ub and Ib is the same in structure, to lay the algorithm foundation for saving logic resources while improving the calculation efficiency of simulation program using pipeline and other hardware development techniques.

This invention relates to the field of power system energy storage technology, specifically to a method and device for site selection and capacity determination of computing-powered energy storage based on grid-connected green electricity direct connection. The method includes: acquiring 8760 hours of time-series operational data and line data from a computing center and new energy power plants, setting the connected grid as an infinite power source; using K-means clustering to obtain typical daily scenarios; constructing a parallel computing framework for three energy storage deployment scenarios, with the minimum life-cycle cost as the objective function, combined with constraints such as power balance; and solving the problem using particle swarm optimization and determining the optimal solution through economic comparison. The corresponding system includes modules for annual historical data acquisition, typical daily scenario clustering, parallel optimization framework, optimization problem solving, and solution acquisition, with each module collaboratively executing the above method. This invention fills a technological gap, improves computational efficiency, achieves optimal economic energy storage configuration, ensures power supply reliability, and facilitates the full absorption of green electricity.

The invention discloses a method and system for accelerating construction of a constellation model and a storage medium. The method comprises the following steps: decomposing an algorithm for constructing a constellation model into an algorithm serial part and an algorithm parallel part according to a preset decomposition mode; coding modification suitable for the CPU process is carried out on the serial part of the algorithm, and parallel coding modification suitable for the AI accelerator process is carried out on the parallel part of the algorithm. And merging and reconstructing the transformed algorithm serial part and algorithm parallel part, and executing the merged and reconstructed algorithm for constructing the constellation model by using an on-board computer carrying an AI accelerator. According to the method, the AI accelerator is used for parallel computing, the constellation model construction algorithm is optimized in parallel, and the number of serial cycles of a CPU is reduced, so that the problem of low operation efficiency of the model construction algorithm is solved, and the acceleration effect is achieved.

This invention discloses an automatic parallel optimization method for large-scale model training in hybrid heterogeneous clusters. This invention requires only real-time collection of a small number of system performance parameters to achieve efficient prediction of training time costs in hybrid heterogeneous cluster environments, while comprehensively considering factors such as GPU performance differences, communication overhead, and resource allocation. This invention formalizes the parallel strategy and parameter optimization problem for large-scale model training in hybrid heterogeneous clusters into a constrained optimization problem and establishes a bandwidth-aware training cost estimation model based on a theoretical model. It decomposes and estimates the time cost during training item by item, achieving rapid and accurate prediction of training costs under different heterogeneous configurations. This invention also designs a parallel strategy search and parameter optimization method based on intelligent optimization algorithms, which significantly reduces the traditional search space size while ensuring near-optimal solutions, improving search efficiency and reducing additional overhead.

The application discloses a kind of water conservancy engineering intelligent scheduling management systems, it is related to water conservancy intelligent scheduling technical field, including data acquisition module, feature extraction module, conflict measurement module and parallel optimization module;Extract time series scheduling dataset, according to scheduling time period to time series scheduling dataset is aggregated and is handled calculation, obtains water storage margin characteristic value and drainage urgency characteristic value, then the state classification of each time period reservoir is obtained, and time series characteristic vector set is obtained;Parallel reservoir adjacency matrix is constructed and calculated to obtain total inflow, the path conflict coefficient of reservoir;According to time series characteristic vector set and path conflict coefficient, scheduling classification reservoir label is set, and target reservoir is adjusted according to scheduling classification reservoir label Gate opening, generates scheduling adjustment instruction, effectively matches the timely demand of different reservoirs and the safety constraint of whole reservoir network, realizes the fine generation of control instruction.

The invention discloses a distributed energy aggregation-oriented multi-agent electric heat storage system collaborative optimization control method, system and device, and a medium, and belongs to the technical field of distributed comprehensive energy system scheduling and control. Comprising the following steps: constructing a layered-distributed cooperative control architecture to perform decoupling control and distributed execution of the multi-agent electric heat storage system; a multi-objective optimization function is designed to generate a multi-agent electrical heat storage system optimization objective, benefits of all agents are coordinated based on a Nash negotiation model, a real-time regulation and control mechanism is utilized to dynamically adjust an energy storage system, a distributed solution algorithm is adopted to decouple a multi-agent coupling optimization problem, and all the agents are optimized in parallel and executed autonomously. The method is high in optimization solution efficiency, high in system operation economy, excellent in resource regulation and control robustness, high in new energy consumption capability, high in real-time dynamic regulation capability and high in engineering application value, and can provide technical support for low-carbon operation and intelligent scheduling.

The invention discloses a DPA-NSGA-II-based beam string structure vibration control optimization method and system, and belongs to the technical field of beam string structure optimization. Constructing a multi-target vibration control optimization model of the beam string structure by taking structural acceleration response and energy consumption as targets according to the structural characteristics of the beam string structure; the multi-target vibration control optimization model of the beam string structure is converted into a mathematical optimization model, and a rapid elite multi-target genetic algorithm DPA-NSGA-II is improved by using a dynamic adaptive crossover operator based on population diversity and a mutation operator perceived by an evolution state to generate a Pareto optimal solution; and evaluating the Pareto optimal solution based on a TOPSIS decision method, and determining the vibration control optimal solution of the beam string structure. According to the method, vibration acceleration suppression and actuator energy consumption minimization serve as two parallel optimization targets, and a series of Pareto optimal solutions can be obtained through single calculation.