40 results about "Function optimization problems" patented technology

The invention relates to a disturbance-based elite reverse learning particle swarm optimization implementation method. The technical scheme of the method comprises a first step of initializing a particle parameter, a second step of calculating particle fitness values and obtaining individual extrema and a global extremum, a third step of progressively decreasing inertia weight in a nonlinear manner, wherein a nonlinear progressive decrease manner not a linear progressive decrease manner is used to change the inertia weight so that a convergence speed and convergence precision of an algorithm are improved, a fourth step of determining a particle position updating mode, a fifth step of updating the individual extrema and the global extremum, and a sixth step of determining a particle continuing execution condition. The method is suitable for solving a function optimization problem, and the method has the high convergence speed and the high convergence precision and can effectively prevent from falling into a local optimum.

The invention relates to a method for movie recommendation on the basis of an orthogonal and cluster pruning based improved multi-objective genetic algorithm. An improved algorithm OTNSGA-II is provided aiming at defects in distributivity and convergence of NSGA-II (non-dominated sorting genetic algorithm-II) and can be used for solving various multi-objective function optimization problems. By design of fault multi-objective orthogonal experiment initialization population, distributive deficiency caused by individual nonuniformity is avoided; by application of self-adaptive cluster pruning strategies, a population evolution process is maintained, and inferior individuals in an appropriate quantity are removed to keep convergence and distributivity of the population. By combination with information mining of user behaviors and movie properties, the algorithm is applied to solving of a practical problem of personalized movie recommendation, universality and effectiveness of the algorithm are explained by test comparison with existing algorithms, better recommendation results are obtained, recommendation accuracy rate, recall rate and coverage rate are increased, rich recommendation scheme combinations are provided, and interest points of users can be mined beneficially to provide more reliable recommendation services.

The invention discloses an RFID indoor positioning algorithm based on a dual-label array phase difference, and the method comprises the following steps: acquiring a phase difference at the same frequency by using double labels to form an array; processing phase data to eliminate phase ambiguity; calculating the distance difference between the label array and the multiple antennas through the relationship between the phase information and the distance, and obtaining multiple hyperbolic curves for hyperbolic positioning; considering the deflection angle of the dual-label array, transforming a positioning problem into a multi-dimensional function optimization problem by constructing an objective function; and solving a final positioning result by a genetic algorithm. The obtained result has asmall error, the positioning accuracy is high and the positioning effect is good.

The present invention discloses a hybrid cuckoo search algorithm. Two different one-dimensional update strategies are introduced into a cuckoo search algorithm. Further, by using the long jump characteristic of Levy distribution, the correct selection between Levy flight random walk and the one-dimensional update strategies is realized by setting a limit value. Thus, the hybrid cuckoo search algorithm not only overcomes inter-dimensional interference and is improved in local search ability, but also increases a probability of jumping out of local optimum, thereby achieving a balance between exploration and development. The hybrid cuckoo search algorithm can be effectively improved in convergence speed and convergence precision, reduces time complexity and have good convergence performancefor solving a function optimization problem.

Provided is a traffic signal timing optimization method based on principal component analysis and a local search improvement orthogonality genetic algorithm. The algorithm is provided by analyzing internal relation among the genetic algorithm, image processing and mode recognition and can be used for solving various function optimization problems. By means of the algorithm, an improvement orthogonality cross operator based on principal component analysis is provided. The operator first conducts PCA projection on the population before cross, individual length is reduced during cross, orthogonal cross operation is implemented on the projection area, the projection is projected to the original space after cross, redundant individual number and calculation expenses caused by redundancy are reduced, algorithm convergence speed is further improved, and the local search strategy is further introduced. The algorithm is applied to single-crossing signal timing optimization. By means of testing comparison with the existing algorithm, the method improves algorithm generality and efficiency, effective timing time is acquired, and the number of the queuing vehicles in front of a crossing is reduced.

An optimization task scheduling method in mobile edge computing comprises the steps: estimating the number of task migration failures according to historical data, modeling a task scheduling problem into an optimization problem about minimization of computing resource allocation variables and task scheduling variables, and converting the optimization problem into a set function optimization problem only about the task scheduling variables; obtaining a primary scheduling strategy suitable for all conditions according to the constructed linear approximation function of the target function, and further obtaining a secondary scheduling strategy according to the constructed sub-module approximation function of the target function when the computing power of the user is weaker than the computingpower of the server; and finally, obtaining an optimized task scheduling strategy through the primary scheduling strategy and the secondary scheduling strategy. According to the method, the delay ofthe computing task can be kept at a relatively low level when possible hardware and software faults occur.

The invention discloses a multi-target community discovering method integrating structure clustering and attributive classification. The method comprises the steps as follows: establishing a network adjacent matrix and an attribute matrix; establishing objective function modularity for measuring structure quality of community division; establishing objective function homogeneity for measuring attribute quality of the community division; initializing a network community division population; using cross and mutation operation to update the community division population; combining a mutated community division population and an external dominance population; finding all dominance community division in a final community division population. The method of the invention designs a function for balancing node attribute classification quality based on Shannon information entropy theory and models an attribute classification problem as an objective function optimization problem. A multi-objective optimization strategy is used to optimize a modularity function for balancing structure clustering quality and a homogeneity function for balancing attribute classification quality to obtain a group of community structures, which are suitable for different applications corresponding to different balances between structure clustering and attribute classification.

The invention relates to the technical field of non-intrusive load decomposition, in particular to a non-intrusive load decomposition method based on an improved differential evolution algorithm. A window function is designed for switching event detection and load steady-state operation time zone positioning. In order to solve the problem that the load identification rate of power load characteristics in a traditional NILD is not high, steady-state current harmonics of load equipment are selected as load identification characteristics. A load identification problem is converted into an objective function optimization problem, and optimization is performed based on a differential evolution algorithm (DE), and an improved DE algorithm of an adaptive cross factor is provided for solving the problems that a traditional DE is prone to falling into local optimization, and the convergence rate in the later period of evolution is low and the like. Experiments show that under different noise backgrounds, the load identification rate and the convergence rate of the improved DE algorithm based on steady-state current harmonics are obviously improved.

The invention discloses a KM (Kermack-Mckendrich) infectious disease model-based function optimization method, i.e., an SIR (Susceptible-Infective-Removed) algorithm. Based on an KM infectious disease model, suppose that N biological individuals exist in a certain ecological system, and each biological individual is represented by n characteristics; the ecological system has an infectious disease and the infectious disease is infected among the N biological individuals; the biological individuals exchange information with one another through an infection operator, a pathological operator, a cure operator, an immune operator and an activity operator; the individual with high PPI (Population Physique Index) transfers information on strong characteristic to the individual with low PPI through the pathological operator and the immune operator, so that the individual with the low PPI index can grow towards a good direction; one individual acquires average characteristic information of some other individuals through the infection operator and the cure operator, so that the probability that the individual gets in locally optimal solution is reduced; the search scope is expanded by improving the activity of the individual through the activity operator; and the algorithm has the characteristics of strong search capacity and global convergence and provides a method for solving a complex function optimization problem.

The invention puts forward a structural damage identification method based on a glowworm group intelligent algorithm. The structural damage identification method comprises the following specific steps: utilizing finite element modeling to obtain the overall quality matrix and the global stiffness matrix of a structure, and furthermore, obtaining the frequency and the modal parameter of the structure; through a frequency residual error and modal guaranteeing criterion, constructing a target function, i.e., a target function to be optimized, of a damage structure; and utilizing the glowworm group intelligent algorithm to find an optimized optimal solution. The method utilizes the glowworm group intelligent algorithm, adopts frequency domain data to research structural damage identification, converts a damage problem into a target function optimization problem and only needs to obtain the frequency and the modal parameters of the first multiple orders of the structure to be detected to obtain a damage identification result. The method has the advantages of high identification precision and the like and does not depend on an initial value.

The invention discloses a mobile user spectrum allocation method based on a cuckoo search algorithm. For the problem that in spectrum power allocation, network throughput is low, firstly, awareness based on an interference distance is built, and a spectrum power allocation problem is converted to a function optimization problem; maximized network throughput is converted to solving a total data volume completed in maximization available time, a target function is built, a spectrum allocation available is mapped to a position of a cuckoo nest, the cuckoo search algorithm is adopted for solving,finally, spectrum allocation with the network throughput higher than that of a genetic algorithm is obtained, and high secondary user effective channel capacity can be obtained.

The invention discloses a method for obtaining interference alignment precoding based on a genetic algorithm, which comprises the following steps: coding each group of precoding matrices and interference suppression matrices of a receiving end in a sequence number coding mode; initializing the precoding matrices and the interference suppression matrices; adopting a fitness function to evaluate thefunction value of the fitness of each chromosome; adopting a roulette mode to randomly select some chromosomes except the chromosome with the highest fitness function value to form a new population;adding the chromosome with the highest fitness function value to the new population; performing cross combination on the two chromosomes which are randomly selected in the new population by adopting adual-daughter-chromosome amphiphilic cross method; carrying out a mutation operation on each bit of each chromosome in the new population by using a variation probability; judging whether the function value of the fitness meets a termination condition or not; and decoding and outputting the precoding matrices and the interference suppression matrices when the termination condition is met. According to the method, a nonlinear multi-objective function optimization problem is solved through the genetic algorithm.

Disclosed is a population dynamics optimization method with vertical-structure nutrition chains. The population dynamics optimization method with the vertical-structure nutrition chains is a PDO-NCVS algorithm. A solution space of an optimization problem is regarded as an ecosystem which has three vertical-structure nutrition chain types of an opened nutrition chain, a closed nutrition chain and a branch nutrition chain, the ecosystem is divided into a plurality of different sub-systems, and each sub-system has a specific vertical-structure nutrition chain type. For each sub-system, a plurality of populations live in each sub-system. The populations can not be transmitted among the sub-systems, and information transfer exists in the kindred populations among the sub-systems which have the same vertical-structure nutrition chain types. The populations living in a sub-system are connected in a predator-prey circulating mode or in a resource-consumption circulating mode. Behaviors during a population acts in a sub-system are constructed into evolution operators which are used for constructing evolution strategies of the populations. The algorithm has the advantages of being strong in search capability and having global convergence, and a solution scheme is provided for the solution of a complex function optimization problem.

The invention relates to the technical field of micro inertial measurement device parameter calibration, discloses an MEMS accelerometer turntable-free calibration method based on an improved fruit fly optimization algorithm, and mainly aims to solve the problem of accelerometer parameter turntable-free calibration by applying an improved swarm intelligence optimization algorithm. An accelerometeroutput model is established according to an MEMS accelerometer error form and a defined coordinate system, and an accelerometer input and output equation is established through multi-position staticobservation. A nonlinear equation set solving problem containing accelerometer calibration parameters is converted into a nonlinear function optimization problem by utilizing a modular observation principle. Directed at the defects that only positive parameters can be searched and the search step length is fixed in a classic fruit fly optimization algorithm, a taste concentration judgment value and the search step length are improved, so that the improved fruit fly optimization algorithm has two properties of global parameter search and variable step length. And the improved fruit fly optimization algorithm is applied to a nonlinear function containing to-be-calibrated parameters of the accelerometer, and optimization solution is carried out on the to-be-calibrated parameters.

A population competition dynamics optimization method with horizontal nutrition structures is a PCDO-HNS algorithm. A solution space of an optimization problem is regarded as an ecosystem which has a plurality of different sub-systems, each sub-system has a specific horizontal nutrition structure type which is one of a normal distribution type, a discrete distribution type, a nearest type, an even type and a monotone decreasing type, populations living in each sub-system mutually compete, and mutual learning behaviors, mutual influence behaviors, mutation occurring behaviors and close keeping behaviors also exist in an accompanying mode. A population competition dynamics model with the horizontal nutrition structures are used for constructing five population evolution operators of the normal distribution type, the discrete distribution type, the nearest type, the even type and the monotone decreasing type, and the operators are used for constructing evolution strategies of the populations. The algorithm has the advantages of being strong in search capability and having global convergence, and a solution scheme is provided for the solution of a complex function optimization problem.