64 results about "Crowding distance" patented technology

The invention discloses a many-objective optimized scheduling method for combined operation of cascade hydropower stations, and aims at solving main problems in engineering application of standard quantum-behaved particle swarm optimization and problems in solving single-objective optimized scheduling. A multi-population evolution strategy is realized by external file set, advantageous individual selection and a chaotic mutation operator strategies, diversity of individuals is ensured, calculation of the method is accelerated, and an approximate Pareto optimal leading edge with sound distribution is obtained. An external file set is introduced to store elite individuals, dynamic update and maintenance of the file set are realized via non-inferior layered ordering and crowd distance, and distribution of the individual is kept uniform; and a chaotic mutation operator is used to carry out local disturbance on a non-control solution, and the neighborhood exploration capability of the individuals is enhanced. According to the invention, the particle swarm optimization is improved, and effectively applied to making the many-objective optimized scheduling scheme of combined operation of the cascade hydropower stations, and a feasible and high efficiency calculating method is provided for many-objective optimized scheduling of the cascade hydropower stations.

The invention provides a multi robot path planning method based on a multi-target artificial bee colony algorithm and belongs to the technical field of path planning. The method includes path planning problem environment modeling, multi-target artificial bee colony algorithm parameter initialization, three-variety bee iteration optimization path and non-inferior solution determination, good path reservation by sequencing and optimum path set outputting. By means of the method, the standard artificial bee colony algorithm is improved based on the concept of non-domination sequence of Pareto domination and crowd distance, and the multi-target artificial bee colony algorithm applicable to solving the multi-target optimization problem is provided. In the path planning process, multiple performance indexes of path length, smoothness and safety are considered in the algorithm, and a group of Pareto optimum paths can be acquired through one-step path planning. The path planning method belongs to meta-heuristic intelligent optimization methods, is different from the traditional single-target path planning method, and can well adapt to path planning tasks in complex environment.

The invention discloses a method for selecting a multi-objective immune optimization multicast router path, which mainly solves the problem of the optimization of multicast routers. The method has the following steps of: (1) determining an optimized target to generate a network model, and setting running parameters to generate an initial population; (2) eliminating an individual path loop; (3) calculating the individual target value to generate a current non-dominant population; (4) judging finishing conditions, if the finishing conditions are met, the current non-dominant population is output, or else, carrying out step (5); (5) calculating the individual crowding distance of the current non-dominant distance to generate an active population; (6) carrying out cloning and local searching operation on the active population; (7) carrying out recombination, variation and local search operation on the cloned population; (8) combining the current non-dominant population with the populations obtained in the steps (6) and (7) to eliminate the individual path loop; (9) calculating the individual value to refresh the current non-dominant population, and carrying out the step (4). The invention has the advantage of providing a flexible optimization scheme and is suitable for selecting the multicast router path.

The invention relates to a meme evolution multiobjective optimization scheduling method based on objective importance decomposition, comprising following steps: randomly generating an initial population with volume of N; in every generation of the algorithm, selecting by the current population through binary championship, generating an offspring population by a genetic operator; carrying out fine search to the offspring population by a partial search strategy to obtain improved population; combining the current population, the offspring population and the improved population to generate a population, carrying out mutation operation to the individuals in the population with identical objective size; sorting the individuals in the population through using the rapid nondominated sorting and crowding distance method in the NSGA-II (Nondominated Sorting Genetic Algorithm II), thus selecting N best solutions as the next generation populations. According to the invention, a multiobjective flexible working shop system can be scheduled effectively; the scheduling effect is superior to the existing advanced algorithm; and the method of the invention can be widely applied in the computer application technical field and the production scheduling field.

The invention provides an intelligent algorithm for solving a multi-objective MDVRPTW (Multi-Depot Vehicle Routing Problem with Time Window), which comprises: a first step: by combining a non-dominant sorting genetic algorithm with an elitist strategy on the basis of an extreme value crowding distance with local searching, searching an extremal solution in a decision space; and a second step: by combining a multi-objective evolutionary algorithm based on decomposition with local searching, on the basis of a final population solved in the first step, further optimizing to obtain a group of solutions with both convergence and diversity. The two-step process enables convergence and diversity of the algorithm to be well balanced, and promotes quality of solving the multi-objective MDVRPTW by the algorithm.

The invention relates to a power distribution network multi-target reactive-power optimization method based on a non-dominated neighbor-domain immune algorithm. Active power loss and reactive compensation input are regarded as targets to be optimized, and a power distribution network multi-target reactive-power optimization model considering constraints such as active balance, reactive balance, power distribution network power limits, node voltage limits, reactive compensation capacity limits, transformer tap limits, compensation node limits and line transmission power limits is established. The power distribution network multi-target reactive-power optimization model is solved by utilizing the non-dominated neighbor-domain immune algorithm. According to the algorithm, the non-inferiority and distributivity of a finally obtained Pareto solution are ensured by adopting proportional cloning, combination, variation and other operations and selection based on crowding distances. The specific configurations of a reactive compensation device having minimum active power loss and lowest compensation input cost can be rapidly and reliably obtained, and the optimization method has a better engineering application prospect.

An embodiment of the invention provides a multi-target active dispatching solving method and system. The multi-target state variable constraint is processed by a processing mode based on constraint priority non-dominated sorting, so that the problem of penalty coefficient selection difficulty in the conventional processing mode is reduced; and according to the algorithm, in the multi-target optimization process, multiple non-domination solutions can be obtained according to the domination relation and congestion distance determined by the non-dominated sorting, namely a Pareto optimal solution set can be obtained. Information interaction and resource sharing between multiple groups based on the constraint priority non-dominated sorting are provided to obtain the more accurate and more uniform Pareto optimal solution set; and in addition, one best compromising solution is determined from the Pareto optimal solution set by adopting a fuzzy subordination method in the fuzzy mathematics to be used as the solution scheme of the multi-target optimization power flow problem.

The invention discloses a sequential test optimization method based on a multi-objective genetic programming algorithm. The method comprises the steps of firstly initializing to obtain a fault diagnostic tree population, using the multi-objective genetic programming method to select, cross and mutate a fault diagnostic tree, during the iteration of each generation, grouping the fault diagnosis tree individuals, wherein, the adaptability of each individual is calculated by using grouping adaptability and a crowding distance; after multiple iterations, selecting non-domination individuals from a final generation population as a non-domination fault diagnosis tree of the sequential test of the system. The sequential test optimization method based on the multi-objective genetic programming algorithm can be used for acquiring a Pareto optimum solution of the fault diagnosis tree of the sequential test for which multiple test indexes can be used as optimized targets, the optimum solutions can be selected by testers to provide guidance for system testers.

The invention relates to an assembly line multi-target modeling method, a particle swarm algorithm and an optimization scheduling method, wherein the optimization scheduling method comprises the steps of S1 constructing a assembly line multi-target optimization model; S2 performing multi-target optimization design to the assembly line multi-target optimization model by the particle swarm algorithm and screening the optimization result to reconstruct the assembly line. According to the reconstructible assembly line optimization scheduling method, a crowding distance calculation method and an elite strategy are referred to; diversity maintenance and global optimal value update are conducted on the basis of individual crowding distance ordering; the complex fitness calculating process is avoided; and a small probability random mutation mechanism is introduced, thereby enhancing the global searching optimization capability gratly.

The invention discloses a multi-objective multi-modal particle swarm optimization method based on Bayesian adaptive resonance. The method comprises the following steps: dividing all particles into a plurality of populations by using a Bayesian adaptive resonance theory; sorting the particles of each population according to a non-dominated sorting method and the special congestion distance; updating the particles in the population by using the individual optimization of the particles and the global optimization of the population; connecting the non-dominated solution sets of various groups endto end to form a closed loop topology, and performing local exploration by using a particle swarm optimization algorithm based on the loop topology; and repeating the two updating and exploring processes until a termination condition is met, and outputting all the non-dominated solution sets and the Pareto frontier. The method is suitable for optimization of solving the multi-target multi-modal problem, the distribution of the Pareto leading edge can be found in the target space, the corresponding Pareto optimal solution set can be found in the decision variable space, a redundant backup method is provided, and the reliability of engineering practice activities is improved.

The present invention discloses an improved crisscross optimization algorithm-based multi-objective reactive power optimization method and system. The method comprises the steps of calculating target values of each particle in an initial population, wherein the target values at least comprise target values of an active power network loss, a voltage offset and a voltage stability margin; performing horizontal cross and vertical cross on the initial population so as to generate sub-generation W and sub-generation R; screening the sub-generation R to obtain an excellent particle population; and combining the initial population, the sub-generation W and the excellent particle population so as to generate a population pool, selecting a new generation of population by using non-dominated sorting and crowding distance, and outputting a final result when an iteration number of times is greater than a preset threshold. In the method, the active power network loss, voltage offset and voltage stability margin are all considered in reactive power optimization of the system, and the system is optimized by using the improved crisscross optimization algorithm, so that multi-objective reaction power optimization is realized, and the algorithm is less likely to optimize locally.

The invention provides a steel making factory multi-target scheduling plan compiling method considering molten iron supply time and molten iron resource utilization. The method comprises the following steps that a multi-target function and constraint conditions considering the molten iron supply condition are established, multiple Pareto optimal solutions about a decision variable are acquired by utilizing a multi-target genetic algorithm based on Pareto for iterative operation, and one concrete iterative process is that a matching scheme between charge and molten iron bottles is adopted to indicate chromosomes, and feasible solutions of all the chromosomes in the current population are acquired by utilizing a decoding heuristic method; a non-dominated solution construction method is designed to calculate non-dominated solutions of the feasible solutions; non-dominated hierarchical ranking is performed on the chromosomes corresponding to all the solutions and crowding distance between the solutions is calculated, and a parent population is selected; and selection, intersection and mutation are performed on the chromosomes in the parent population so that progeny populations are obtained. The problem of scheduling plan compiling considering the molten iron supply condition can be solved, and multiple Pareto optimal solutions are acquired so that decision makers are enabled to select more appropriate solutions to be applied to actual production.

The invention provides an optical waveguide array-optical fiber array automatic butt-coupling parallel index optimization method. The optical waveguide array-optical fiber array automatic butt-coupling parallel index optimization method comprises that mapping between an optical waveguide array-optical fiber array auto-aligning physical parameter and a parallel index optimization model is established, and a first objective function and a second objective function are obtained; a maximum evolutional generation is set, input single-mode fiber arrays and output single-mode fiber arrays enter specified spatial positions in a specified direction and attitude according to the regulation of each individual, optical power values of two or more than two waveguide channels are read, recorded and stored, and the Pareto sequence and the crowding distance are calculated; evaluation of each individual in one generation cluster is completed, a next generation cluster is generated by means of intersection, variation and lowliest place elimination, the evolutional generation is accumulated, whether the evolutional generation reaches the maximum evolutional generation or not is determined, if the evolutional generation reaches the maximum evolutional generation, the process is finished; if the evolutional generation does not reach the maximum evolutional generation, the intersection step is returned. According to the invention, the low-loss rapid butt-coupling automation degree and the working efficiency of an optical fiber array-waveguide device-optical fiber array system are effectively improved.

The invention discloses an image cutting method based on a multi-target intelligent body evolution clustering algorithm. The problems that the image cutting technology is prone to local optimum and an algorithm is not high in robustness are mainly solved. The image cutting problem is converted into a global optimization clustering problem. The process includes the steps of extracting gray information of pixel points of an image to be cut, initiating parameters and establishing an image intelligent body network, calculating the energy of an image intelligent body, conducting non-domination sequencing, conducting neighborhood competition operation, conducting Gaussian mutation operation, calculating the energy of the image intelligent body, conducting non-domination sequencing, conducting self-learning operation, selecting the optimal clustering result according to the crowding distance, outputting a clustering label, and achieving image cutting. Multiple targeting is achieved for the image processing process, the convergence effect is good, the robustness of the method is enhanced, the image cutting quality can be improved, the cutting effect stability can be enhanced, and the extraction, recognition and other subsequent processing of the image targets are facilitated.

The invention discloses a multi-objective particle swarm optimization algorithm based on a collaborative mutation method. According to the algorithm, an initial population with the scale of N is randomly generated in a decision space of a to-be-solved problem; in the iteration process, a population is randomly divided into two equal-scale sub-populations P1 and P2. Multi-mode differential variation and non-uniform variation are made on the two sub-populations according to the probability to generate descendant sub-populations P1'and P2'. The descendant sub-populations P1'and P2' are combined to form a temporary intermediate population. Then elite individuals are scrrened from the temporary intermediate population by using a rapid non-dominated sorting method to update an external archive set, and the diversity of the external archive set is maintained by using a crowding distance strategy. A method effectively solves the problems that a multi-objective particle swarm optimization algorithm is prone to premature convergence and unbalanced in search capability.

The invention discloses a data-driven material reverse design method and system, and the method comprises the following steps: firstly, obtaining the material data of a formed material sample; then, according to the material data, respectively selecting a machine learning model of a corresponding relationship between each performance parameter and a design parameter; searching and selecting parameters of each machine learning model by adopting a cross validation method based on the material data to obtain a modified machine learning model of a corresponding relationship between each performance parameter and a design parameter, and taking the modified machine learning model as a fitness function of each performance parameter of the genetic algorithm; and finally, carrying out optimizationsolution by adopting a genetic algorithm. According to the invention, a non-dominated sorting mode and a crowded distance sorting mode are adopted in the genetic algorithm; multi-objective optimization solution is realized, a penalty factor does not need to be introduced, the technical defects that all possibilities cannot be exhausted in a search space in traditional grid search and a design result is inaccurate due to a penalty factor introduction process are overcome, and the comprehensiveness and the accuracy of material reverse design are improved.

The invention provides a BP neural network and MBFO algorithm-based aluminum electrolysis energy conservation and emission reduction control method. The method comprises the following steps: carrying out modeling on the aluminum electrolysis production process by utilizing a BP neural network; and optimizing the production process model by utilizing a crowding distance-based multi-target bacterial foraging optimization algorithm so as to obtain a group of optimum solutions of each decision variable as well as current efficiencies and greenhouse gas emissions corresponding to the optimum solutions, wherein the crowding distances of the non-inferior solutions need to be calculated during the optimization of the production process model and the external files are updated according to the crowding distances so that the floras rapidly move toward the target to ensure quick convergence under the premise of ensuring the population diversity. According to the method, the optimum values of the process parameters in the aluminum electrolysis production process are determined, the current efficiency is effectively improved and the greenhouse gas emission is decreased, so that the aim of conserving energy and reducing emission are really achieved.

The invention discloses a multi-target weapon-target distribution method based on an artificial fish swarm algorithm, aiming to overcome the defect that solving for a WTA problem in the prior art deviates from a real Pareto leading edge greatly. The multi-target weapon-target distribution method comprises the following steps: firstly, randomly generating an initial population, calculating a non-dominated solution set of the initial population, and sorting according to a congestion distance to obtain a globally optimal solution of the initial population; then, according to the clustering behavior of the artificial fish swarm algorithm, enabling other individual fishes in the fish swarm to approach the optimal solution to obtain a new population and a previous non-dominated solution set, andcalculating a new non-dominated solution set; and finally, carrying out crossover variation on the clustered population to increase population diversity, combining with the previous non-dominated solution set again, and carrying out multiple iterations to obtain a final Pareto leading edge. The multi-target weapon-target distribution method is mainly used in the field of fire fighting decision making, is closer to the real Pareto frontier compared with the prior art, has small dependence on parameters, and has great application value in multi-target weapon-target distribution.

The invention discloses a multi-broadcasting route optimization searching method based on improving the clone niche algorithm in the technical field of communication, which comprises obtaining network information and generating an alternative route base, randomly generating a first antibody aggregation according to the first alternative route base, determining an antibody in a memory pool, calculating the affinity degree of the first antibody aggregation, forming final n groups of antibody clusters according to the affinity degree, carrying about the clone proliferation for n groups of the antibody clusters and variation then, selecting a Pareto solution in each group of the antibody clusters and then putting the Pareto solution in a second antibody aggregation, processing the second antibody aggregation, selecting out the Pareto solution of the second antibody aggregation and then putting the Pareto solution in the memory pool, carrying out the similarity suppression and gradient decision, determining whether the partial crowding distance of the antibody in the memory pool is smaller than or equal to a preset upper-limit threshold value, operating a partial crowding system if the partial crowed distance is larger than the threshold value, determining whether reaching iteration times, and completing the optimization searching if reaching the iteration times. The invention can optimize a plurality of QoS parameters in the optimizing process of multi-broadcasting route.

The invention relates to a power distribution network reactive power optimization method based on an improved NAGAII algorithm, and the method comprises the steps: S1, determining an optimization target, and building a target function; S2, determining the number of control variables and constraint conditions; S3, initializing control parameters of the improved NSGAII algorithm, and generating an initial parent population; S4, performing genetic operation on the parent population individuals, performing updating, generating new individuals to generate a test population, and calculating target function values and overall constraint violation degrees of the test population individuals; S5, enabling individuals in the parent population and the test population to compete to generate a temporarypopulation; S6, performing non-dominated solution sorting and congestion distance sorting on individuals in the temporary population, and selecting a new parent population; S7, judging whether the number of iterations is greater than the maximum number of iterations or not, if so, stopping iteration, and outputting a reactive power optimization result; and if not, returning to S4. According to the method, the problems that the original NAGAII algorithm is easy to fall into local optimum and the convergence speed is low are effectively solved.