167 results about "Premature convergence" patented technology

The invention provides a power distribution network fault positioning method based on improvement of a binary particle swarm algorithm, the conventional binary particle swarm algorithm is improved, and the method is applied to positioning of power distribution network faults. The method comprises following steps: firstly, determining parameters including the particle swarm scale and the maximum iteration frequency etc.; then forming an expectation function of a switch according to fault information of the switch, and constructing a fitness function of power distribution network fault positioning; initializing a particle swarm, setting particle positions, and setting the speed of the particles as 0; calculating the fitness values of the particles according to the fitness function, and setting an initial global extremum; updating an individual extremum and the initial global extremum; updating the speed and position of the particle swarm; and stopping calculation when reaching the maximum iteration frequency, and outputting the global optimal position of the particle swarm, namely the practical fault state of each feed line section of a target power distribution network. According to the method, the problem of premature convergence of the conventional method can be overcome, and the convergence and the stability of the algorithm can be further improved.

The invention discloses a cascade hydropower station group optimized dispatching method based on an improved quantum-behaved particle swarm algorithm. The problems that local optimum happens to the quantum-behaved particle swarm algorithm at the later iteration period due to premature convergence for the reason that population diversity is decreased, and an obtained hydropower station group dispatching scheme is not the optimal scheme are mainly solved. The hydropower station group optimized dispatching method based on the improved quantum-behaved particle swarm algorithm is characterized by comprising the steps that first, power stations participating in calculation are selected, and the corresponding constraint condition of each power station is set; then, a two-dimensional real number matrix is used for encoding individuals; afterwards, a chaotic initialization population is used for improving the quality of an initial population, the fitness of each particle is calculated through a penalty function method, the individual extreme value and the global extreme value are updated, an update strategy is weighed, the optimum center location of the population is calculated, neighborhood mutation search is conducted on the global optimum individual, the positions of all the individuals in the population are updated according to a formula, and whether a stopping criterion is met or not is judged. The hydropower station group optimized dispatching method based on the improved quantum-behaved particle swarm algorithm is easy to operate, small in number of control parameters, high in convergence rate, high in computation speed, high in robustness, reasonable and effective in result, and applicable to optimized dispatching of cascade hydropower station groups and optimal allocation of water resources.

The invention belongs to the computer field, and discloses a method for scheduling a flow shop based on a multi-swarm hybrid particle swarm algorithm, which solves the problems that the flow shop scheduling method based on the hybrid particle swarm algorithm is easy to result in premature convergence and local optimum. The method comprises the following steps of: setting parameters and generating S sub-swarms; judging whether the terminal condition is satisfied, if so, outputting a current optimum scheduling scheme, otherwise, updating positions of particles in each sub-swarm with the particle swarm algorithm, carrying out a local search on odd and even sub-swarms respectively by using searching operators 1 and 2 to obtain an optimum scheduling sequence of each sub-swarm; sharing information of the obtained optimum scheduling sequence by using a statistics-based probability model; and optimizing an optimum working sequence with a simulated annealing algorithm. In the invention, multiple swarms are added, the local search is carried out by using different searching operators, a good flow shop scheduling scheme is obtained, the production time is shortened, and the method can be used for the selection of the job shop scheduling scheme.

The invention relates to a real-time path planning method of AUV (Autonomous Underwater Vehicle), in particular to a method for carrying out online, real-time local path planning according to an online map in an AUV real-time collision preventation process. The method comprises the steps of: setting the quantity of small populations according to the quantity of path points of the AUV, initializing; carrying out immune selection on each small population to obtain subgroups; carrying out genetic manipulation on one subgroup, carrying out cell cloning on the other subgroup; then clustering through a vaccination and an antibody to form the next generation of small population, judging whether the next generation of small population meets the conditions or not; if yes, selecting optimal individuals of the small populations; and selecting the optimal individuals from the set consisting of all optimal individuals to be used as a planning path. According to the invention, the diversity of the population is maintained by using an antibody clustering principle, the premature convergence of an algorithm is avoided, and the global optimization is facilitated. The established immune genetic algorithm is used for clustering and analyzing generated filial generations by adopting a self-regulating mechanism, and the diversity of the population is ensured.

The invention provides a polyclone artificial immunity network algorithm for multirobot dynamic path planning, and relates to an improved polyclone artificial immunity network algorithm. According to the polyclone artificial immunity network algorithm, a polyclone artificial immunity network is applied to multiple mobile robot dynamic path planning, mutual influence among robots and influence on the robots of mobile barriers are considered, a computational formula of antibody concentration is defined, diversity of antibodies is increased through clonal operators, crossover operators, mutation operators and selection operators, and the problem of premature convergence of a traditional immune network is solved. Specific antibodies corresponding to antigens in a specific environment are stored, initial concentration of the specific antibodies is increased, response time is shortened, and multiple mobile robot dynamic path planning in an unknown environment is effectively achieved.

The invention discloses an improved fuzzy neural network bus intelligent scheduling method based on a chaos theory, and belongs to the field of intelligent transportation. According to the improved particle swarm bus intelligent scheduling method based on the chaos theory, advantages and complementarity of various algorithms are fully utilized, a series of improvement measures are also introduced, such as conjugate gradient optimization, and inertia factor and constraint factor of the particle swarm algorithm etc., the mechanism and the search performance are researched from the theoretical and practical perspectives, problems of poor global search capability and premature convergence of the conventional optimization algorithm are fundamentally solved, the diversity of population can be obviously increased, the global search capability is obviously improved, the problem of fuzzy information can be effectively dealt with, the convergence speed is fast, and a new high-efficiency method is provided for bus intelligent scheduling.

The present invention provides a logistics distribution path optimization method. The method is configured to effectively optimize the logistics distribution path with multiple targets. The method comprises: through analysis of influence factors on customer satisfaction in the logistics activity, putting forwarding a logistics system customer satisfaction target function, combining a distance target function, and putting forwarding a total target function. The global optimization is realized by local optimization through combination of a kruskal algorithm, and it is proposed that the kruskal algorithm is organically combined to a traditional heredity algorithm crossover operator to solve the problem that a traditional heredity algorithm is liable to premature convergence. An order terminal distribution model is constructed by using a VRP model through combination of the customer satisfaction function, the order terminal distribution model is solved by using the method provided by the invention to obtain an optimization path. According to the invention, the logistics distribution path is effectively reduced, and three logistics distribution schemes consisting of the shortest path, the highest customer satisfaction and the optimum total objective function are provided so as to create conditions for reducing the logistics cost and improve the logistics enterprise vitality.

The invention discloses an improved particle swarm optimization (PSO) algorithm of solving the zero-waiting flow shop scheduling problem. Firstly, parameter initialization and population initialization are carried out, wherein initial workpiece sequences are generated, then a factorial encoding method is used to map all permutations to integers to form an initial population, and finally, a feasible initial velocity set is randomly generated; particles are moved; the population is updated through an original PSO population updating strategy, a new population is mapped to corresponding workpiecesequences, and work completion time of each new workpiece sequence is evaluated; an improved variable neighborhood search (VNS) algorithm is used for a local search, and results obtained by the search are used for replacement; a population adaption (PA) operator is used to increase diversity of the population; and checking of a termination condition is carried out, if the termination condition ismet, a process is stopped, and values of variables and corresponding sequences are returned to be used as a final solution, and otherwise, particle velocity is continuously updated. The method has the advantages of improving a particle swarm optimization algorithm, improving global search capability, and avoiding too early convergence.

The invention discloses a multi-objective optimization method based on an improved gravitation search algorithm. According to the algorithm, a memory strategy is introduced into a universal gravitation search algorithm, so that particle swarm information and information of previous generations and next generations of particles are shared, the global search capability and the local search capability of the particles are balanced, and the premature convergence problem is solved. On this basis, a diversity enhancement mechanism is introduced into the algorithm, namely, particle speed and position are controlled each iteration, so that the diversity loss is relieved, the particle diversity is improved, and diversity and distributivity of non-dominated solution sets are enhanced. Therefore, by means of the multi-objective optimization method based on the improved gravitation search algorithm, the phenomenon that multi-objective optimization is caught in a local extremum can be effectively avoided, and convergence, diversity and distributivity of non-dominated solutions are remarkably improved when the gravitation search algorithm is applied to the field of multi-objective optimization.

The invention relates to a sentiment classification method based on multi-source field instance migration. For the phenomenon that the migration efficiency of a TrAdaBoost algorithm in migration learning may be lowered, multi-source learning is introduced, by trying sample migration from different source fields or combining the features of multiple source fields, migration learning is stable and effective, much initial weight is distributed to a target sample, so that weight mismatching is relieved, and the phenomenon of quote imbalance is improved by resampling data at each step; and for the phenomenon of source field weight premature convergence in the TrAdaBoost algorithm, dynamic factors are added, and the problem that weight entropy is transferred from a source sample to the target sample is solved. According to the method, premature convergence of the weight of the source field sample which is small in the correlation with the target field is avoided, a learning target task is helped together, and knowledge of all source fields is fully used.

The invention discloses an ordered binary decision diagram modeling method for distribution network fault recovery. The method includes a first step of carrying out partition on a power distribution network, and using an on-off state as a decision variable, a second step of forming an adjacent matrix of the power distribution network according to a forming method of the adjacent matrix of an undirected graph, a third step of forming a reachable matrix of the power distribution network, a fourth step of forming boolean functions of each restriction of fault recovery, a fifth step of combining to generate a final binary decision diagram, and obtaining a contractive solution space, a sixth step of carrying out load flow calculation on feasible solutions one by one in the solution space after contraction, and carrying out safety verification, and a seventh step of obtaining an optimal recovery scheme according to priority levels of optimized objective functions. The modeling method has the advantages of effectively reducing the solution complexity of fault recovery problems, overcoming the defect that a traditional artificial intelligence algorithm is prone to premature convergence in local optimal solution, and improves understanding reliability and the like.

The invention relates to the field of electric systems and discloses a weighting Voronoi diagram substation planning method based on a chaotic and genetic strategy. The method aims at solving the problems that a prior algorithm is low in rate of convergence, poor in capacity of local optimization and sensitive in initial value, premature convergence exists, the unreasonable phenomenon caused by division of power-supply districts according to the principle of proximity exists, the load rate of a planned substation can not be controlled, and the like, and optimizing site selection of the substation and division of the power-supply districts by means of certain algorithms. The method comprises the steps of setting parameters; chaotic initialization and generating initial population including N individuals; carrying out the site selection of the substation and load distribution on the N individuals; judging whether end criterion is satisfied; calculating the fitness variance sigma 2 of the population; chaotic search; and executing and saving an optimized genetic algorithm and then returning to the fourth step. The weighting Voronoi diagram substation planning method based on the chaotic and genetic strategy is mainly applicable to the electric systems.

The invention discloses an optimal configuration method for an electric automobile charging pile. The method comprises the following steps: predicting the charging power demand of a planning area by a Monte Carlo simulation method on the basis of analysis of various electric automobile behavior characteristics; building a bi-level planning model of charging station investment profit and user perception effect under the consideration of constraint conditions such as a power grid, a charging station and an investor budget; and introducing a KKT (Karush-Kuhn-Tucker) condition to realize equivalent conversion of a double-layer model and a single-layer model, and solving by adopting a variable neighborhood search-particle swarm mixed algorithm with a convergence polymerization degree. Through adoption of the method, the problem of premature convergence of particles is avoided effectively; population diversity is increased; the optimization capacity of the particles and the convergence speed of the algorithm are improved and increased remarkably; the calculation speed and the calculation accuracy of optimal configuration of the charging station are increased; and important references are provided for investors to plan and build the charging station under an enterprise-dominant pattern.

The invention discloses a particle swarm optimization method based on survival of the fittest and step-by-step selection. The method mainly comprises the steps that particles are divided to two sets in the operating process, the first set of particles are superior particles, and the second set of particles are inferior particles. A solution space is searched for in the global scope through a species, and the global searching ability is enhanced; after each time of evolution, m the best particles in the species are maintained, the position spaces of better particles of the m particles are selected as new solution spaces, new particles are selected to replace the position of poor particles in the species in the new solution spaces, in this way, the optimal particles can be approached step by step, and the optimal solution is found out. The searching ability of the particle swarm optimization method is enhanced, the defects that a fundamental particle swarm optimization algorithm easily falls into local extremum or premature convergence or stagnation are overcome, and the optical values of parameters to be optimized can be more accurately and quickly found out.

The invention discloses a method for solving the traveling salesman problem based on an improved genetic algorithm. The steps include: aiming at the TSP problem, encoding the path using a decimal number string; calculating the total length, and then judging the total length; after encoding On the search space U of the decimal number string path, define the fitness function f(x), and define the population size n, the crossover probability Pc, the mutation probability Pm and the number of iterations T; in the search space U, randomly generate n individuals s1, s2, s3, ..., sn, constitute the initial population S0 = {s1, s2, s3, ..., sn}, set the current iteration number t = 0; according to the fitness function f(x), evaluate the individual fitness in the population , if t<T, then end the step, otherwise perform the genetic operation step; the individual with the highest fitness obtained through the genetic operation step is the optimal solution of the traveling salesman problem solving method. Based on the traditional genetic algorithm, the present invention optimizes the traveling salesman problem to achieve the purpose of improving the shortcoming that the algorithm is prone to premature convergence and optimizing the search efficiency.

The invention discloses a semi-supervised intrusion detection method combining the improved grey wolf algorithm, and belongs to the technical field of network information security. The method can effectively avoid the defect that the basic GWO algorithm is prone to premature convergence, and balance the global exploration and local development ability of the GWO algorithm; and the cloud GWO algorithm is used to optimize the K-means algorithm to mark data having similar features, and realization of generation of the large-scale accurate mark training data set is realized on the basis of manually marking the data in a small amount. The proportion of unmarked data and marked data is provided, the phenomenon of "under-fitting" and "over-fitting" of the model is avoided, and the detection accuracy of the model is ensured. The optimized semi-supervised learning method is combined with the cloud GWO algorithm to optimize parameters of a single-class support vector machine model. Compared withthe unilateral optimization, the invention achieves higher detection precision.

The invention relates to a multi-population simulated annealing hybrid genetic algorithm based on similarity expelling. The multi-population simulated annealing hybrid genetic algorithm includes the following steps: coding is carried out; initialization parameters are set; initial populations are created; fitness values are calculated; selecting operation is carried out; interlace operation is carried out; mutation operation is carried out; gene overturning operation is carried out; simulated annealing Metropolis rules are judged; migration operation based on similarity expelling is carried out; optimal storage is carried out; judgment is ended. The migration operation based on similarity expelling particularly includes the following steps: calculating the fitness values of individuals in a source population and a target population; selecting the individual with the largest fitness value from the source population to serve as the individual to be immigrated; conducting similarity calculation; conducting expelling replacement. The multi-population genetic algorithm with simulated annealing operation can improve the local search capability of the multi-population genetic algorithm, and the algorithm can search for approximate solutions even though optimal solutions to a larger extent. The individual similarity judgment is additionally carried out, attention is paid to differences between the individuals, the diversity of populations is maintained, premature convergence of the genetic algorithm is avoided, the solving quality of the algorithm is improved, and the algorithm is closer to the optimal solutions.

The invention relates to a pre-stack non-linear fluid identification method for a fuzzy neural network of a chaotic quantum-behaved particle swarm. Fluid identification is always a key point and difficult point problem in the oil-gas exploration field. By aiming at deficiency in the common fluid identification method at present, a multi-attribute angle gather combination fluid identification factor is built by researching an AVO (amplitude versus offset) response characteristic comprising different fluids; a chaos search mechanism, a quantum-behaved particle swarm and a fuzzy system theory are organically combined to fully perform respective advantages and complementarities of the chaos search mechanism, the quantum-behaved particle swarm and the fuzzy system theory; a novel group intelligent optimization algorithm of a ''chaotic quantum-behaved particle swarm fuzzy system'' is developed and researched, and a mechanism and an optimizing performance of the pre-stack non-linear fluid identification method are researched from two aspects of the theory and practicality; problems of poor global search capability, premature convergence and the like in a traditional optimization algorithm are fundamentally improved; the optimization algorithm is introduced into fluid identification to form the pre-stack non-linear fluid identification method for the fuzzy neural network of the chaotic quantum-behaved particle swarm; the problem existing when a traditional fluid detection means is used for carrying out fluid identification is effectively solved; fluid identification precision is improved; and a new scientific and effect technical method is provided for the fluid identification.

The invention provides a brain part MRI image segmentation method. The brain part MRI image segmentation method is characterized in that a gray level image of a brain part MRI image to be segmented can be acquired; the gray values of different pixel points of the brain part MRI image can be used as the clustering centers, which are used to form the clustering center sets as the particles, and the optimization of the clustering center sets can be carried out by adopting the particle swarm optimization algorithm; every pixel point of the brain part MRI image belongs to the category having the maximum membership, and then the gray values of the pixel points of the same category are equal to the same gray value, and the brain part MRI image segmentation can be completed. The brain part MRI image segmentation method is advantageous in that according to the chaotic characteristic and the logic self-mapping function, the uniformly-distributed particle swarms can be initialized by adopting the logic self-mapping function, and then the quality of the initial solution, the stability of the PSO algorithm, the speed and the precision of the image segmentation can be improved; the chaotic searching can be carried out, when the particles are in the premature convergence state, and the premature convergence phenomenon caused by the stagnated state of the particles during the iteration process can be prevented, and the optimal solution in the range of the whole situation can be realized, and then the speed and the precision of the image segmentation can be improved.

The invention discloses a chemical engineering process modeling method of an RNA genetic algorithm with remodeling operation, which has the following steps: 1) actual input and output sampling data is obtained by on-site operation or experiment, and then the sum of the absolute value of the error between the estimated output and the actual output of a chemical engineering process model is used as the objective function of the RNA genetic algorithm for optimization research; 2) algorithm control parameters is set; 3) the RNA genetic algorithm with remodeling operation is operated to do estimation for the unknown parameters of the chemical engineering process model, then the estimated value of the unknown parameters of the chemical engineering process model is obtained by the minimized objective function, and the estimated value of the unknown parameters is substituted into the chemical engineering process model to form a mathematical model of the chemical engineering process. The excellent genes of original species group is preserved when the diversity of the species group is increased effectively, thereby avoiding the defects of the premature convergence of traditional genetic algorithm and the convergence of locally optimal solution better.

The invention discloses an automobile spare part loading optimization method based on an improved secondary particle swarm algorithm. The decision-making elements of cost, resources, service quality, and the like are comprehensively considered during the logistic transportation loading of automobile spare parts, and an automobile spare part loading optimization model is established. The secondary particle swarm algorithm is introduced for solution, and the improved secondary particle swarm algorithm is adopted aiming at the defect of low early particle diversity searched by the secondary particle swarm algorithm, a variation idea and an interchangeable updating mechanism of a genetic algorithm are adopted to enhance the diversity of a population so as to avoid premature convergence, and the optimization solution effect is improved. Simulation examples show that the particle diversity and the algorithm solution efficiency during the algorithm solution process are obviously better than those before improvement, and the probability for searching optimal overall situation is higher. The invention provides a method support for optimizing and improving a loading scheme of enterprise automobile spare parts and related goods in the logistics industry.

The invention provides a hydrological model parameter calibration method based on a PSO (particle swarm optimization)-GA (genetic algorithm) mixed algorithm, and belongs to the field of hydrological model parameter calibration. The method includes inputting rainfall, all parameters of a hydrological model, and maximum and minimum values of each parameter in an initialization phase of the hydrological model,; executing a hydrological model algorithm program to acquire an output forecasting runoff value; comparing the calculated forecasting runoff value with an actual value to acquire a certainty coefficient, and performing judgment; when the certainty coefficient is larger than or equal to 0.2, indicating that an error between the forecasting runoff value and the actual value is larger than or equal to 20%, and performing parameter calibration according to the PSO-GA mixed algorithm. The method has the advantages that optimizing characteristics of GA and PSO algorithms are combined, premature convergence is avoided, and searching accuracy is improved.

The invention discloses an intelligent microgrid building load power dispatching method improving gravitational search. First, all power using loads in an intelligent microgrid building are classified, an objective function and constraint conditions of building load dispatching are clarified, and a load power dispatching model is built. Then a gravitational search algorithm is subjected to binary discretization and a parasite population and a host population are established. At the same time, the memory and population information sharing capabilities of particles in a particle swarm algorithm are introduced, and parasitic behaviors of organisms are simulated. Through co-evolution of the two populations, the convergence speed and search precision of the algorithm are improved, the diversity of the populations and the global search ability are improved, and the shortcomings of the original gravitational search algorithm such as premature convergence and low optimization accuracy are overcome. The method considers the objectives including resident economy, comfort and stability of a power grid and adopts the improved gravitational search algorithm for multi-objective optimization, so that the least electricity cost, the most comfort and the least impact on the power grid are achieved for residents.

The invention provides an improved solving algorithm for a flexible flow shop scheduling problem. Particle velocity and position correlation operators are redefined. A coding matrix and a decoding matrix are introduced to represent the relationship among workpiece, machine and scheduling. In order to improve the initial group quality of the improved discrete particle swarm algorithm for the flexible flow shop scheduling problem, a shortest time decomposition policy algorithm based on an NEH algorithm is proposed for the first time by analyzing the relationship between initial machine selection and total completion time of scheduling. The improved discrete particle swarm algorithm is used for global optimization. The algorithm has the advantages that the quality of the initial solution of a particle population is improved; and the time for particles to find the optimal solution is shortened; and an inertial weighting method with exponential decrease is used to effectively avoid the premature convergence of the particles and enhance the early global search ability of the particles.

The invention discloses a vehicle route optimization method. The vehicle route optimization method includes that defining a vehicle route problem into a graph model, solving an inter-cluster cost route from a global view to acquire a feasible solution space, and optimizing the feasible solution space through a Monte Carlo method, genetic manipulation, a quantum rotating gate adaptive strategy and the like. The vehicle route optimization method is capable of solving the vehicle route problem in the global connection, optimizing the time and space complexity in the problem solving process, and avoiding premature convergence. The vehicle route problem refers to that a certain number of customers have different numbers of goods demands, a distribution center provides goods for the customers, one motorcade is responsible for distributing goods, and proper driving routes are organized to meet the demands of the customers and achieve the aims of shortest journey, lowest cost, shortest time consumption and the like under a certain constraints.

The invention mainly belongs to the technical field of multi-target optimization, and particularly relates to a novel multi-target particle swarm optimization method. The method is based on shared learning and Cauthy mutation. Shared learning factors are adopted in the method to change the speed and position updating formula of particles, and the optimal position of the particle individuals and external files are updated by adopting Cauthy mutation operators. According to the method, the global search capability and the local optimization search precision of the particles can be enhanced, and the algorithm is enabled to fast approach Pareto leading edge and premature convergence of the algorithm can be avoided. According to the method, solution convergence, diversity and distributivity of the multi-target particle swarm algorithm for processing the problem of multi-target optimization can be enhanced.

Provided is a pose recognition method of a leaner. The method comprises steps of separation of a portrait and a background; using mathematical morphological operation to extract a profile image of a learner for a binary image; using a Zernike matrix to carry out feature extraction; adopting a support vector machine to train the feature vectors; and recognizing poses of the learner. According to the invention, by introducing the Levy flight mechanism in a bee colony algorithm, according to different flight step lengths, different searching modes are adopted, so the variety of the population can be enriched; premature convergence and falling into the local optimum can be avoided; a dynamic weighting mechanism is adopted in the bee colony algorithm, and an individual bee searching mode is adjusted according to the evolutionary rate of the population, so global searching and local searching ability are dynamically balanced; and the method is advantaged by good segmentation effects and high recognition rate, and can be used for recognition of poses of the learners and recognition and classification of other images.

The invention discloses a gradient adaptive particle swarm optimization method based on swarm aggregation effect. The method comprises the following steps of firstly, setting initialization parameters, initializing the speed and position of a particle swarm, then initializing a population extreme value and an individual extreme value, then clustering the particle swarm by adopting a K-Means clustering algorithm according to the relative positions of particles in a search space to obtain a clustering result, and then calculating a clustering extreme value and a corresponding position accordingto the clustering result; adaptively adjusting the calculation parameter of each particle according to the descent gradient of the fitness value of the target function of the particle, and calculatingthe fitness value of the particle at the current position according to the current position of the particle and the target function; and finally, updating the individual extremum, the clustering extremum and the global extremum according to the fitness value of the particle at the current position, and updating the speed and the position of the particle. The method provided by the invention can effectively solve the problems of premature convergence, local optimum and the like of the existing particle swarm method, and greatly improves the optimization capability of the algorithm.

The invention provides a genetic algorithm using an improved coding method to solve a distributed flexible job shop scheduling problem. The algorithm is suitable for the field of flexible job shop scheduling. The distributed flexible job shop scheduling problem refers to production activities carried out in several factories and manufacturing units, contains the information of flexible job shop scheduling problems, and contains the selection of suitable factories and flexible manufacturing units. The allocation of specific workpieces to different factories produces different production scheduling, which affects a supply chain. The existing technology has little research on the problem. The invention proposes the genetic algorithm to solve the problem. A traditional genetic algorithm is easy to fall into local optimum and leads to premature convergence, and the coding method produces infeasible solution and other problems. According to the invention, the improved coding method based on probability is provided; the crossover of a tabu table is introduced; the problems are avoided based on the variation of neighborhood search; and the genetic algorithm takes into account the actual production situation, and has the characteristics of high practicability and the like.

The present invention discloses a particle swarm optimization algorithm based on a clustering degree of a swarm. The algorithm comprises the following steps of carrying out initialization; updating the swarm; judging whether a number of iterations is greater than a preset number of iterations and executing a corresponding step; judging whether a number of update iterations is greater than a preset number of times of stagnation and executing a corresponding step; calculating a particle clustering degree of each particle and a particle clustering degree of a swarm optimal position so as to acquire a distance between each particle and the swarm optimal position; according to a fitness of each particle, selecting a plurality of particles of which the number accords with a swarm scale to form a current swarm; and carrying out iterative optimization and updating until the maximum number of iterations is reached. According to the particle swarm optimization algorithm disclosed by the embodiment of the present invention, different evolutionary strategies can be adopted for different particles according to the progress of the optimizing process and the particle clustering degree so as to reduce the possibility of falling into the local minimum, improve the global searching ability of the algorithm and effectively avoid premature convergence.