38results about How to "Expand the solution space" patented technology

A heterogeneous multi-core energy-saving task schedule method based on improved genetic algorithm comprises the improved genetic algorithm used for determining task priority and energy-saving schedule algorithm based on zooming priority. The flow of the heterogeneous multi-core energy-saving task schedule method includes (1), initializing population information; (2), entering a loop body and determining the task priority by the aid of genetic algorithm; (3), determining schedule sequence of tasks on a processor according to a task DAG (directed acyclic graph) and division strategies; (4), realizing dynamic voltage zooming on the basis of feasible task schedule according to relation of saving energy of the tasks and prolonging time; (5), calculating fitness of a current population and sorting the current population; and (6), updating the population by the aid of the improved genetic algorithm, determining new task priority, quitting if termination conditions are met, and continuing iteration if the termination conditions are not met.

The invention relates to a data center energy saving method based on a central processing unit (CPU) dynamic frequency modulation technology. The method includes two stages of initialization arrangement and dynamic management of a virtual machine. The method is different from methods in the prior art, a virtual machine arrangement technology used by energy saving of a data center is divided into the two stages of initialization arrangement and dynamic management. The initialization arrangement is applicable to the occasions of operation beginning and regular integral adjustment of the data center. The adopted method achieves the overall effect and the long-term effect due to time permission. The dynamic management is used for local management or moving facing different tasks during data center operation. The adopted method relates to local resource and ensures real-time performance.

The invention provides a calculation method of an optimal power flow based on a class extension variable interior point method. The calculation method comprises the following steps of importing a class extension variable in a non-convex nonlinear programming, and determining a lagrangian function after the class extension variable is imported; constructing a Karush-Kuhn-Tucker (KKT) condition equation of the lagrangian function after the class extension variable is imported; solving the KKT condition equation; and calculating the optimal power flow according to a prediction-correction method. According to the calculation method, when an original optimization problem has no solution for the limits of constraint conditions, a range needing to open the constraints and the corresponding optimal solution are provided; and a solution is also provided for some optimization problems which cannot be converged by using the traditional prediction-correction method; the solution space of the optimization problem is widened; and the convergence is improved. A practical solution is provided to numerical problems in a large number of equations for a new method by the calculation method, and the effectiveness is verified by an example of an electrical power system power flow optimization problem.

The invention discloses a variable neighborhood search method and system on a cloud computing platform. The which are variable neighborhood search method and system on the cloud computing platform used for overcoming the defects, of excessively-quick convergence and being weakened in the capacity of fleeing a local extreme value, existing in conducted variable neighborhood search based on a single living example. The method includes the steps of presetting a data collection and a plurality of initial solutions and storing the initial solutions to the data collection;, using a plurality of living examples and being based on at least one of the initial solutions to conduct neighborhood search on a solving space; for any living example, when a locally optimal solution which is obtained in a searching mode is superior to a worst solution in the data collection, using the locally optimal solution to update the worst solution, wherein the worst solution is the solution of a minimum value among initial solutions and/or locally optimal solution in the data collection. The living examples are adopted to simultaneously conduct search and results can be searched mutually among the living examples, therefore, solving space can be effectively expanded, the possibility of fleeing locally optimal possibility is improved, and furthermore a better overall optimal solution can be obtained.

The invention relates to an on-line string matching method without gap constraint, which relates to the technical field of electric digital data processing. The on-line mode is used for processing thepattern matching problem without gap constraint and without overlap, that is, characters at the same position in the sequence string can be matched at different positions of the pattern string. The method comprises the following steps: reading a pattern string P to establish a plurality of queues; reading a given sequence string S according to a pre-sequence and a post-sequence; determining whether queue i is capable of creating a node; determining whether the occurrence of a non-overlapping condition can be constituted, and when the occurrence of a condition is constituted, outputting on thedisplay until all the characters in the sequence string S are processed. The invention overcomes the defect that it is difficult to effectively control the space cost and the time cost on the basis of ensuring the completeness of the prior art, and not only improves the solving speed, but also ensures the completeness of the solution.

The invention discloses a logistics distribution line planning method, device and equipment and a storage medium. According to the method, the mathematical relationship between the distribution line and the constraint condition of the mathematical model is calculated to continuously adjust the planning of the transfer center in the distribution line; specifically, a genetic crossover algorithm andan iterative neighborhood search algorithm are adopted to continuously screen and replace a transfer center; therefore, the solution space can be increased, the optimal transfer center can be selected to find the optimal distribution route of logistics, the optimal distribution route of the logistics distribution route can be quickly found, the total length of the obtained route length is the shortest, and the logistics distribution cost is effectively reduced.

The invention discloses an ultra-short-term wind power combination prediction method based on a support vector machine, and the method comprises the steps: firstly carrying out the linear interpolation replacement of to-be-processed wind power historical data according to the data of an adjacent time period, and carrying out the normalization of the preprocessed data; secondly, decomposing the processed wind power data into an eigenfunction sequence and a residual error sequence by using empirical mode decomposition; secondly, establishing a quantum particle swarm-support vector machine model for the eigenfunction sequence and the residual sequence obtained by decomposition, and performing training optimization to obtain a predicted value of each sequence; and finally, superposing the prediction values of the sequences to obtain a final wind power prediction value, and carrying out error evaluation analysis. Compared with a support vector machine direct prediction result or a result without data feature decomposition, the prediction result of the method is improved, and meanwhile the situation that local errors are too large does not occur. Compared with an existing wind power prediction scheme, the method is higher in robustness, higher in calculation speed, less in data requirement and better in prediction effectThe invention discloses an ultra-short-term wind power combination prediction method based on a support vector machine, and the method comprises the steps: carrying out the linear interpolation replacement of to-be-processed wind power historical data according to the data of an adjacent time period, and carrying out the normalization of the preprocessed data; secondly, decomposing the processed wind power data into a cost characteristic function sequence and a residual sequence by utilizing empirical mode decomposition; secondly, establishing a quantum particle swarm-residual sequence for the intrinsic function sequence and the residual sequence obtained by decomposition; carrying out training optimization on the support vector machine model to obtain a predicted value of each sequence; and finally, superposing the predicted values of the sequences to obtain a final wind power predicted value, and carrying out error evaluation analysis. Compared with a result of direct prediction of a support vector machine or no data feature decomposition, the prediction result of the method is improved, and meanwhile, the situation of overlarge local error does not occur. Compared with an existing wind power prediction scheme, the method is higher in robustness, higher in calculation speed, less in data demand and better in prediction effect.

The invention discloses a slope stability prediction method based on an improved PSORBF algorithm. The invention discloses a slope stability prediction model based on an improved PSORBF algorithm, andbelongs to the technical field of slope stability prediction. According to the method, the radial basis function neural network is initialized, the particle swarm optimization algorithm based on thenormal attenuation inertia weight factor is adopted to optimize the radial basis function neural network parameters, and the new radial basis function neural network prediction model is constructed according to the optimal parameters, calculated by the optimization algorithm, of the radial basis function neural network; and the slope stability is predicted by using the improved radial basis function neural network prediction model. According to the method, a radial basis function extension speed control factor is added on the basis of a Gaussian function used by a traditional hidden layer, thefactor can adjust the change trend of parameters of the neural network in the iteration process, sudden change in the iteration process is avoided, and the prediction precision of the trained model is higher.

The invention relates to a design method of an escape orbit starting from a Halo track and used for detecting a deep space target. The method is especially suitable for design of the escape orbit starting from the Halo track and used for conducting a deep space detecting task and belongs to the technical field of aerospace engineering. First, a minor celestial body periapsis state of interior extent unstable disturbance manifold of a task given Halo track is obtained through calculation; then, according to minor celestial body escaping hyperbola overspeed required by interplanetary transfer required by the task, maneuvering pulse required to be exerted at the position of disturbance manifold periapsis is solved, and flying time from the periapsis to a minor celestial body affecting ball is solved; and finally, contour maps of velocity increment and flying time are drawn respectively, and the escape orbit meeting a task requirement is selected from the contour maps. On the basis of low fuel consumption, flying time for escaping from the Halo track is shortened, and ephemeris constraint of other celestial bodies is not required to be introduced. The degree of freedom of the escape orbit starting from the Halo track is increased by introducing the disturbance manifold, and solution space of a traditional design method is expanded.

The invention provides a satellite task scheduling method and system based on time window segmentation, and relates to the field of satellite scheduling. The method comprises the following steps: obtaining a visible time window of a satellite task; discretizing the visible time window to obtain an observation time window; performing real number encoding on the satellite task based on the observation time window to obtain a satellite task sequence; acquiring a satellite task population based on the satellite task sequence; performing crossover operation on the satellite task population based ona preset two-generation competitive optimization algorithm; and performing mutation operation on the satellite task population after the crossover operation based on a double-gene mutation method toobtain a satellite task scheduling scheme. According to the invention, the efficiency of satellite observation tasks is improved.

The invention provides an unmanned aerial vehicle path planning method for minimizing the number of unmanned aerial vehicles, which mainly solves the problem of minimizing the number of unmanned aerial vehicles in the existing unmanned aerial vehicle path planning scheme, and comprises the following implementation steps: 1) initializing parameters; 2) constructing a to-be-accessed sensor set and initializing a local optimal path set; 3) calculating flight energy consumption, hovering energy consumption and communication energy consumption of the unmanned aerial vehicle; 4) planning flight paths for a given number of unmanned aerial vehicles by adopting an improved ant colony algorithm, and determining a local optimal path set; and 5) obtaining an unmanned aerial vehicle path planning result according to the characteristics of the local optimal path set. According to the invention, on the basis of accessing all the sensors, the minimum number of unmanned aerial vehicles are used to complete the task, the system cost is reduced, and the total energy consumption of the unmanned aerial vehicles is reduced.

The invention belongs to the field of chip manufacturing, and provides an optical proximity correction mask method based on adaptive hybrid conjugate gradient descent. The method provides a self-adjustable hybrid conjugate gradient descent strategy. According to the invention, the step length and the direction of gradient search are automatically adjusted by guiding a conjugate gradient method which is specifically adopted in a next cycle correction process by utilizing experience of a previous cycle, so that an optimal solution is better searched. The solution space is larger, and the newly designed optimization function can enable the quality of the generated mask to be better; a photoetching acceleration model is realized, and the algorithm efficiency is high; and the generated mask can more effectively resist process deviation. The chip manufacturing mask is optimized, so that the pattern after photoetching has higher fidelity.

The invention relates to a 5G communication system resource allocation method based on an improved grey wolf optimization algorithm. The method comprises the following steps: firstly, initializing thebasic parameters of a system, initializing a grey wolf population identifying a resource allocation scheme according to a good point set principle, and calculating and sequencing fitness values of the grey wolf population by taking the minimum total interference of the system as a target function; secondly, selecting the wolves with the first three fitness values according to a sequencing result,and calculating the positions of the other wolves relative to the wolves with the first three fitness values; updating the positions of the other wolves, updating the positions of all wolves and calculating the fitness value of each wolf; and finally, determining whether the number of iterations reaches the set maximum number of iterations or not, and if the iterations are completed, obtaining anoptimal resource allocation strategy. According to the method, a solution space is increased through the good point set principle, an algorithm nonlinear convergence factor calculation mode is improved, an individual position updating equation is improved in combination with a fitness value, the total interference of the system is effectively reduced, then the D2D user access number of a cell isincreased, the system throughput is improved, and the method is easy to implement.

The invention discloses a distribution network path optimization method based on a condensation hierarchical clustering method. Under the background of energy Internet, a large-scale renewable energysource can be connected into a power distribution network through various electric energy devices such as an electric energy router, and electric energy generated by the electric energy router is alsodistributed through the electric energy router. The path optimization of the distribution network containing a distributed energy source can effectively reduce the electric energy loss in the electric energy transmission process, meanwhile, the voltage distribution condition of the distribution network can be improved, and the node voltage deviation is effectively reduced. According to the distribution network path optimization method, through modeling and load flow calculation of the power distribution network containing the renewable energy source and the electric energy router, an improvedimmune clone algorithm based on aggregation hierarchical clustering is provided, so that an optimal power supply path is searched by taking the minimum total consumption value of active power and theminimum total deviation value of node voltage as targets.

The invention relates to a method for exchanging adjacent processes of quasi-critical path and equipment, which comprises the following steps of determining a scheduling sequence of each process in aproduct processing process tree by adopting a quasi-critical path strategy, and determining scheduling time of each process by adopting a first adaptation strategy to form an original scheduling scheme; on the basis of the original scheduling scheme, searching for interchangeable adjacent procedures sequentially on each machining equipment; adjusting an interchange process scheduling sequence according to an adjacent process interchange strategy, and meanwhile, adjusting the influenced process; obtaining a new product scheduling scheme; and searching next interchangeable adjacent processes insequence to generate a new product scheduling scheme. And after all the schemes are generated, selecting the scheme with the minimum total processing time as the final product scheduling scheme. According to the algorithm, on the basis of a quasi-critical path strategy, an adjacent process exchange strategy and an adjacent process exchange adjustment strategy are provided, so that the parallel processing efficiency between parallel processes in the scheduling problem is improved and the scheduling result is optimized while the compactness between the serial processes is ensured.

The invention relates to a layout method for three-dimensional FPGA chip design. The method comprises the following steps of: simulating a three-dimensional chip by using a planar layout method of thetwo-dimensional chip and a circuit area and the component capacity of the three-dimensional chip, performing hierarchical division of the three-dimensional chip after a two-dimensional planar layoutresult is generated, and vertically superposing all the divided hierarchical layouts into a three-dimensional chip layout. Compared with an existing method, the method has the advantages that the layout quality is better, and the layout result has more solutions; the main workload is concentrated to a two-dimensional simulation part, so that the complexity of software engineering is reduced; the current software process can be better reused, and the development progress of three-dimensional layout software is accelerated.

The invention discloses a parallel task energy-saving scheduling method based on linear programming, including the following steps: firstly, the initial scheduling sequence of tasks and the mapping relationship between tasks and processor nodes are obtained by the existing performance-first task scheduling methods; then the execution process of the task is divided into time slices according to thefrequency of the processor, and a linear energy-saving objective function and constraints are constructed; finally, a linear programming tool is used to solve the reasonable time slice value. The invention does not limit the number and range of frequencies to be selected from the overall consideration, and has lower energy consumption, and is suitable for a plurality of performance-first task scheduling methods, and does not increase the scheduling length.

The invention discloses an MIMO radar orthogonal waveform design method based on improved GA-SQP, and the method comprises the steps: optimizing a phase coding matrix through an improved GA-SQP algorithm, and designing an orthogonal waveform through the optimized phase coding matrix; according to the method, after crossover and variation of a traditional genetic algorithm, an SQP optimization stage is added, and a certain number of individuals are selected for SQP multiple iterations, so that the local optimization performance of the algorithm is improved; meanwhile, on the basis of an original genetic algorithm, crossover operation and mutation operation continue to be refined, an operation object is changed into a chromosome from an individual, the solution space is expanded, the convergence of the whole algorithm is further improved, and the optimization speed of a phase encoding matrix is increased.

The invention provides a data processing method, device and equipment, and the method comprises the steps: obtaining an original execution plan which comprises a plurality of nodes; selecting a targetnode which cannot be processed by a data source from the nodes of the original execution plan; performing equivalent transformation on a target node in the original execution plan to obtain an equivalent execution plan; determining a target execution plan according to the original execution plan and the equivalent execution plan; and sending the target execution plan to a data source, so that thedata source executes the target execution plan. Through the technical scheme of the invention, the computing power of the data source can be reasonably utilized, the data transmission quantity is reduced, and higher query performance is obtained.