428 results about "Simulating annealing" patented technology

The invention provides a simulated annealing-based WSN hierarchical routing method. The method is based on a hierarchical network topology. The hierarchical structure is maintained by cluster nodes, cluster head nodes and sink nodes in a network. The method comprises the steps of firstly, dividing levels of the WSN, constructing the hierarchical structure of nodes, and then conducting data transmission. In the constructing phase of the hierarchical structure of nodes, cluster head nodes are elected and backed up according to energy. In the data transmission phase, a combination of a negotiation mechanism, a multi-hop energy multi-path routing mechanism and a simulated annealing algorithm is used, cluster nodes use the negotiation mechanism to transmit data to cluster head nodes, after data fusion is conducted by cluster head nodes, the data are transmitted to sink nodes, the multi-hop energy multi-path routing mechanism is used between cluster head nodes and sink nodes, the simulated annealing algorithm is used for seeking out an optimal path, and the data are transmitted from cluster head nodes to sink nodes along the optimal path.

The invention relates to a method for tracking multi-target vehicles by adopting the MCMC (Markov Chain Monte Carlo) algorithm, comprising the following steps: modeling the multi-target vehicle tracking process by adopting the MCMC algorithm, establishing various preselection states, traversing the preselection states by using the Metropolis-Hastings sampling based simulated annealing algorithm, and selecting the preselection state with the maximum connection probability as the optimal solution. In the method, a limit is set to the transfer equivalence for the first time, and the parameters in the data association are estimated by the experimental data regression fitting, thus the optimal vehicle motion trajectory with maximum posteriori probability significance is obtained, and the vehicles are tracked. The invention solves the problems of frequent shielding splitting of the multi-target vehicles and has the advantages of high multi-target vehicle tracking precision and good real-time performance.

The invention discloses a service request distribution method oriented to an edge computing environment. The method comprises the steps of (1) monitoring an available resource and a service request execution situation of each edge apparatus in a system in real time; (2) collecting service requests collected by all edge apparatuses in the system; (3) selecting an optimal edge server or a cloud server for all service requests by use of a GASD (Globally Aligned Spatial Distribution) algorithm; and (4) performing service request scheduling on all edge servers. Compared with the prior art, from twoaspects of distribution and scheduling of the service requests, the service response time is optimized to the utmost extent; meanwhile, a heuristic method is adopted, a temperature control mechanismin a simulated annealing method is introduced into a genetic algorithm, the convergence rate of the algorithm can be slowed down at the initial phase of the algorithm, thereby effectively avoiding local optimization; and the convergence rate is accelerated at the termination phase of the algorithm, thereby improving the efficiency of the algorithm.

The invention discloses a control system performer fault diagnosis method based on multi-swarm cooperative chaos simulated annealing particle swarm optimization-support vector machine (MCCSAPSO-SVM). The method comprises the steps of: by means of combined noise reduction and an improved empirical mode decomposition (EMD) method, carrying out feature extraction on collected output signals of a performer; utilizing MCCSAPSO to optimize structural parameters of the support vector machine; adopting a combined Kernel function to ensure good generalization capability and learning capability of the SVM; and utilizing training data to construct a binary tree SVM, and converting a complex polytomous problem into a plurality of dichotomous problems. According to the invention, common state signals, which can be easily obtained, of the control system are processed, whether the control system performer breaks down can be effectively judged in real time according to the output of the binary tree SVM, and the fault type can be accurately determined when the control system performer breaks down. The method is used for real-time fault diagnosis of a high-precision system.

The present invention discloses a main user positioning method based on sensor and quantum intelligent computing, which is achieved based on a wireless sensor network assisting a cognitive radio network, and includes the following steps: step 1, a network deploy stage; step 2, a positioning information collecting stage; step 3, a distance measurement stage, wherein a data fusion center averages the sampled signal strength as a received signal strength RSS of an anchor node, and estimates the distance between a main user and the anchor node according to the RSS in a lognormal shadow path loss wireless broadcasting environmental model; and step 4, a positioning stage, wherein the positioning problem is converted into an optimization problem, and the optimization problem is solved by using a quantum genetic simulated annealing algorithm, thereby achieving positioning the location of the main user in a two-dimensional space. On the premise that a good positioning performance is ensured, the present invention can achieve the effect of reducing complexity of the algorithm and saving the energy consumption of the battery at the same time; and accurate location information of the main user can be obtained via the positioning method based on the quantum genetic simulated annealing algorithm.

The invention discloses a mathematical model for optimal configuration of power distribution network filtering devices. Under the condition that harmonic voltage, condenser capacity and the like meet constraint conditions, in terms of system average voltage total distortion and investment cost, a comprehensive objective function is given in a linear weighting manner, a penalty function is added to the objective function so that a constrained optimization problem is converted into a unconstrained optimization problem, and an improved simulated annealing-particle swarm optimization (namely introduction of an adaptive inertia coefficient and a memorizer) is used for solving. The model can be adapted to the random variation of a harmonic source and network parameters, can ensure that the harmonicration and average voltage total harmonic distortion of each node of the network are in the specified limits, can optimize the installation type, installation location, installation quantity, capacity parameter and the like of active filters and passive filters in a centralized manner in the whole network; and by the mathematical model, the investment cost of filtering devices in the whole network is reduced to the minimum.

The invention provides a network security situation prediction method based on an SA _ SOA _ BP neural network. The method comprises the steps of collecting network security data information as experimental data for preprocessing; determining a network structure of the BP neural network by using a trial-and-error method according to the input quantity and the output quantity in the experimental data; introducing a simulated annealing algorithm into the crowd search algorithm to obtain an improved crowd search algorithm; initializing a simulated annealing algorithm, finding an optimal individual by adopting an improved crowd search algorithm, calculating the fitness value of the individual through a fitness function, and optimizing the connection weight and threshold of the BP neural network; and substituting the test sample into the BP neural network to obtain a predicted value of the network security situation. According to the method, the simulated annealing algorithm is introduced into the crowd search algorithm, so that the problems that the crowd search algorithm is easy to fall into local optimum and slow in convergence are solved; and the BP neural network is optimized and improved by utilizing the advantages of the improved crowd search algorithm in speed and global search.

An on-load tap-changer fault identification method based on vibration signal characteristic extraction is disclosed. The method is characterized by using a simulated annealing algorithm to carry out parameter selection; and adaptively selecting a modal number K and a second penalty factor alpha according to a vibration signal to obtain an optimal parameter combination. The effect of vibration signal modal decomposition is improved, which is good for extracting more significant characteristics and is convenient for subsequent state monitoring and fault diagnosis.

The invention discloses a distribution network energy storage optimization configuration method considering whole life cycle cost. A two-stage optimization solution method is adopted to solve the problems of distribution network energy storage addressing and constant volume; from the aspects of energy storage to stabilize voltage fluctuations caused by distributed power and load, with the whole life cycle cost of energy storage as an optimization goal, a two-stage optimization model is established; in a first stage, candidate nodes are screened through a genetic algorithm and a simulated annealing algorithm, and energy storage required mounting nodes and the energy storage operation strategy and the minimum mounting capacity of each node are outputted finally; and in a second stage, the service life of the energy storage is considered based on the optimization result in the first stage, and the energy storage capacity configuration with the lowest whole life cycle cost is optimized through the genetic algorithm and the simulated annealing algorithm. The method considers the random output fluctuation of the distributed power and the whole life cycle cost of the energy storage system, and can obtain a more economical energy storage location and a capacity allocation scheme for the distribution network, and a reference is provided for the distribution network energy storage optimization configuration method.

The invention discloses a simulated annealing particle swarm algorithm based hybrid power automobile parameter optimization method. A threshold value in a hybrid power automobile control strategy is converted into a group of particles to be optimized, the automobile fuel consumption rate and emissions are utilized as an optimization objective function, the simulated annealing process is performed on the particles in a parallel mode, the new state of every particle is selectively accepted according to the Metropolis criterion in the annealing process, the local optimum is jumped out through the jumping characteristic of a simulated annealing particle swarm algorithm, and the global optimal solution is achieved through convergence finally. According to the simulated annealing particle swarm algorithm based hybrid power automobile parameter optimization method, the problems that the setting process of hybrid power automobile control parameters is based on the experience and an optimal threshold value cannot be obtained are solved, the optimal threshold value can be obtained rapidly, and the vital significance is brought to the automobile energy conservation and emissions reduction and the theory research of hybrid power automobiles.

The invention provides a lithium battery capacity online prediction method based on K-means clustering and an Elman neural network. The method comprises the following steps: firstly, determining the model of a lithium ion battery to be tested, carrying out cyclic charging and discharging experiment by utilizing a battery with the same model as the battery to be tested, recording a lithium batterydischarging time sequence, carrying out K-means clustering on the lithium battery discharging time sequence, and establishing a data model; and then, introducing a simulated annealing genetic algorithm to optimize initial weight and threshold of the Elman neural network, training the Elman neural network by using the constructed data model, and establishing a lithium ion battery actual capacity prediction system offline. When capacity prediction is carried out online, the collected actual discharging time sequence data of the lithium ion battery to be tested is input into the prediction system, and the actual capacity of the battery is predicted while the normal work of the lithium ion battery is not influenced. According to the invention, online accurate prediction of the actual capacityof the lithium ion battery can be realized.

The present invention discloses a sliding mode robust control method of a discrete time delay uncertain system. A multi-population chaotic simulated annealing particle swarm optimization algorithm is used to optimize a support vector regression machine (SVR) structure parameter, a combined kernel function support vector regression machine is used to give a system prediction model, a control law is solved through a novel sliding mode reaching law, and the reachability of a sliding mode can be guaranteed. The method is used for the robust control of uncertain discrete system with state time delay.

The invention provides a multi-agent adversarial decision-making method based on cooperative reinforcement learning and transfer learning, and the method is characterized in that the method comprisesthe following steps: defining the state space S = {s1, s2,..., sn} of an agent; setting an action space A to be equal to {a1, a2,..., an}; setting a value function matrix of the agent reinforcement learning model; calculating a value function sequence corresponding to the current state st by using an action evaluator, and selecting a corresponding action at through an action selector based on simulated annealing and softmax strategy; meanwhile, the state of the intelligent agent is changed, and the intelligent agent is transferred to the next state st + 1. After the action at is executed, theintelligent agent obtains a reward signal rt from the environment; the loss of experience storage can be reduced through a weight sharing mode, and the adversarial decision-making efficiency is improved. Through the migration learning method based on the attenuation function, the agent can reuse previous experience with a gradually decreasing probability, and the migration learning migrates the previously trained action evaluator weight to more adversarial decision scenes, thereby improving the generalization of the learning model.

An AGV scheduling optimization model is built with goods cabinets, sorting tables and AGVs as scheduling research objects and minimum time for AGVs to complete a specified task as an optimization objective. The AGV scheduling optimization model is solved using a two-stage multi-population parallel genetic algorithm and a simulated annealing algorithm, and the algorithms are solved by means of parallel computation to get an optimal scheduling scheme of AGVs for a specified task. Populations can be prevented from converging too early. The computational efficiency is improved. The algorithms can deal with a large-scale scheduling optimization problem.

The invention discloses a photovoltaic array parameter identification method based on measured data. The method includes the steps that a feasible solution range of parameters to be acquired are determined in advance through a conventional parameter identification method; parameter identification is conducted on the measured data of a photovoltaic array through a genetic algorithm and a simulated annealing genetic algorithm. The parameter identification method is high in practicality. By compressing a region of search, identification efficiency of the genetic algorithm is effectively improved, and complexity of the algorithm is reduced; meanwhile, the measured parameters are identified by combining the genetic algorithm and the simulated annealing genetic algorithm, and therefore accuracy of identification results is effectively improved.

A methodology for optimizing the placement of network “Points of Presence” (POPs) across the carrier's entire network (i.e., a “global” solution”) utilizes carefully constructed customer clustering and simulated annealing methodology to create a cost-efficient solution. The customer base is first partitioned into a plurality of clusters such that the customers within each cluster are closer to its centroid than the centroid of any other cluster (e.g., applying the k-means clustering algorithm or any other suitable method of partitioning the customer base). A linear algorithm process is used to minimize the costs associated with the number of placement of POPs within each cluster. A simulated annealing (SA) process is then used to iterate the entire set of potential POP locations until a compact, steady-state solution is achieved (or, alternatively, until a given number of iterations has been performed). In a preferred embodiment, a number of iterations are performed at each “temperature” in the simulating annealing process to further improve the result (this iterative process referred to in the art as “intensification”).

The invention discloses a firefighting movable robot path planning method and system based on an improved artificial potential field method. The method comprises the steps that a fire point is acquired; the coordinate point of a firefighting movable robot near the fire point is acquired; a target point is set, and a corresponding relation between the target point and the firefighting movable robotis set up; environmental information is read, and resultant force corresponding to the firefighting movable robot is calculated through the improved artificial potential field method; whether the firefighting movable robot path gets into a local minimum point or not is judged, and if yes, the firefighting movable robot path gets away from the local minimum point through a simulated annealing algorithm, and then the resultant force is calculated again; if not, the calculated resultant force is sent to the corresponding firefighting movable robot, and whether the moved firefighting movable robot reaches the corresponding target point or not is judged; if the firefighting movable robot reaches the target point, fire extinguishing work is carried out; and if not, the resultant force is calculated again. According to the firefighting movable robot path planning method and system, the problem that during artificial potential field method path planning, the target cannot be reached, the firefighting movable robot gets into the local minimum point is solved, and meanwhile the firefighting movable robot avoids an obstacle and reaches the target point faster.

Described herein are systems and methods for classifying a data set using an ensemble classification technique. Classifiers are iteratively generated by applying machine learning techniques to a training data set, and training class sets are generated by classifying the elements in the training data set according to the classifiers. Objective values are computed based on the training class sets, and objective values associated with different classifiers are compared until a desired number of iterations is reached, and a final training class set is output.

The invention relates to a location method applicable to improvement of micro-earthquake location reliability. The method is characterized in that in view of shortcomings of a successive mesh generation micro-earthquake algorithm, in combination with a very fast simulated annealing algorithm, a micro-earthquake location algorithm is developed, which is based on joint very fast simulated annealing and successive mesh generation. According to the method, the very fast simulated annealing algorithm is employed to search for a highly energy-focused circular region in a three-dimensional target area, and then the successive mesh generation method is employed to search for a maximum energy-focused point in the area; thus, the problem of wrong location of the successive mesh generation method due to oversize primary mesh generation during location of a high-frequency seismic wave is effectively solved, and the location reliability is improved while the calculation efficiency is ensured.

The invention relates to a TSP problem path planning method. The method comprises the following steps: initializing; reading the position and calculating the distance; initializing a population through a greedy algorithm; replacing the worst individuals with randomly generated individuals; calculating the fitness; selecting; crossing; performing variation; randomly performing simulated annealing on the plurality of individuals; calculating the fitness; giving the contemporary optimal solution and the variant solution thereof to the first individual and the second individual respectively; and iterating until the termination condition is met. The population generated by the greedy algorithm has randomness and high quality, and optimization can be accelerated. A plurality of worst individualsare replaced by randomly generated individuals, so that the influence of differential solutions is reduced, and precocity is avoided. Some better solutions can be found through simulated annealing, and precocity and local optimization are avoided. The storage of the optimal solution and the variant solution of the optimal solution retains excellent information and increases population diversity.According to the invention, a shortest access path can be effectively and quickly planned, so that the method is an effective method capable of providing path planning for the TSP problem.

The invention discloses a modeling method of a support vector machine (SVM)-based software measurement instrument in a biological fermentation process. The modeling method uses an optimized SVM modelto estimate fermentation products in the biological fermentation process and comprises two steps: the SVM modeling of the biological fermentation process and the confirmation of an optimal model basedon a genetic simulated annealing algorithm and an akaike information criterion. The invention provides an optimal SVM software measurement instrument model for variables difficult in online measurement in the biological fermentation process, eliminates a phenomenon of untimely control caused by time lag of a traditional offline measuring method, lays a reliable foundation for establishing the SVMmodel in the biological fermentation software measurement and has the advantages of online measurement, automatic parameter optimization, automatic input variable selection, small sample demand and high precision.

The invention relates to a cloud computing cluster load scheduling method based on a GA algorithm, which comprises the steps of: S1: reading a task queue, and searching an operation parameter of eachtask from a service task execution time model table; S2: carrying out coding on the operation parameters, converting an operation parameter set into a chromosome bit string set, and producing an initial group Group(g), wherein g is equal to 0; S3: calculating a fitness value of each individual in the initial group; S4: judging an end condition that g id greater than or equal to Gmax (Gmax represents a maximum breeding algebra), if the condition is not met, executing the step S5, or turning to the step S8; S5: carrying out a selection operation to form a next-generation group Group(g), and adopting a selection algorithm which simulates annealing, wherein g+ is equal to 1; S6: carrying out an interlace operation by a probability Pc; S7: carrying out a mutation operation by a probability Pm,and turning to the step S3; and S8: ending the algorithm, and outputting a current optimal scheduling scheme. According to the invention, corresponding request task time is greatly shortened, GIS service quality can also be obviously improved in a case that a user amount is increased, and use satisfaction of a user is improved.