39results about How to "Maintain population diversity" patented technology

The invention relates to a method for inverting anisotropy parameters using variable offset vertical seismic profile data. The method includes: calculating vertical longitudinal wave speed and vertical transverse wave speed of each layer of a vertical seismic profile; collecting symbol layer positions and determining intersections of each layer and well trajectory; building up a surface layer model and an observing system; randomly initiating the anisotropy parameters of layers to be determined and cubic coefficient of a top layer interface to be determined; forming anisotropy speed models to be optimized; screening receiver inter-layer points ; collecting shot first arrival to calculate polarizing angles; building up a target function using P wave first arrival travel time and the polarizing angles; using niche genetic algorithm to optimize the target function; finishing optimization of the anisotropy parameters of the layers to be determined and configuration of the top layer interface to be determined; calculating layer by layer 'from shallow deprocessing to deep deprocessing' to finish optimization of the anisotropy parameters of the whole model. By the method, a reflecting interface is close to actual stratum, linear convergence ray tracing is high in calculation efficiency, complicated internal relations of anisotropy media are simplified, iterative test ray tracing modeling is guided by guide search of the niche genetic algorithm, and the anisotropy parameters and interface configurations are inverted and optimized nonlinearly.

The invention provides a goods allocation optimization method applied to a Flying-V untraditional layout warehouse. The method is characterized by comprising the steps that S1, goods allocation relevant parameters of the Flying-V warehouse are set; S2, the goods allocation parameters are initialized; S3, a population is initialized according to an information list of to-be-stored goods; S4, an adaptive genetic algorithm is adopted to perform individual optimal selection on the population; S5, whether the number of algorithm termination iterations is reached is judged, if yes, the step S6 is performed, and otherwise the step S4 continues to be cycled; and S6, an optimal goods allocation scheme is output. The goods allocation optimization method is applied to the application occasion of "Flying-V untraditional warehouse layout", goods storage and delivery efficiency and a lowest gravity center of a goods shelf after the goods are placed on the goods shelf are optimized, and a multi-target optimization problem processing method with different dimensions is provided; by the adoption of the adaptive genetic algorithm, the crossover rate and the variation rate change along with adaptation values; and population diversity is maintained, and global convergence of the genetic algorithm is guaranteed.

The invention is applicable to the technical field of workshop production planning management, and provides a hybrid genetic algorithm-based MES (Manufacturing Execution System) production planning and scheduling method. According to an order priority generation mode and / or random generation mode, a workshop task sequence which meets constraint relations between tasks and a genetic algorithm coding rule is generated for a preset number of initial scheduling schemes; according to the best task starting and ending time and according to one or more distribution combination modes in resource load balancing principle distribution and random distribution, execution resources are set for each workshop task in the preset number of initial scheduling schemes; the preset number of initial scheduling schemes are converted into a series of chromosomes through a coding process to serve as an initial population for the hybrid genetic algorithm; and the initial population is introduced to the hybrid genetic algorithm, and a scheduling result after optimization is calculated according to a preset optimization target. High efficiency of the MES production planning and scheduling results in the prior art is improved.

The invention relates to an optimization scheduling method in a raw material leaching process for chemical industry production, belonging to the technical field of intelligent optimization scheduling in a chemical industry production process. According to the invention, a scheduling model and an optimization object in the raw material leaching process are determined; then, the optimization object is optimized by using the optimization scheduling method based on a self-adaptive distribution estimation algorithm, wherein the scheduling model is established according to the maximum completion time of raw materials processed on each leaching equipment; and the optimization object is to minimize the maximum completion time. By means of the invention, the raw material leaching process for chemical industry production is expressed clearly and accurately; the scheduling method is reasonable and effective; the evolution direction of populations can be guided easily by utilizing more excellent individual information; therefore, the global search depth of the algorithm is increased; and the scheduling problem in the raw material leaching process for chemical industry production can be effectively solved.

The invention discloses a heavy haul train operation curve multi-objective optimization method based on a hook buffer device model. Based on running line constraint conditions of the heavy haul train,a dynamic longitudinal dynamics model and a hook buffer device model in the train operation process are established; A multi-objective genetic algorithm is used to establish a train optimization control model. meanwhile, the premature phenomenon of the genetic algorithm is considered, the genetic algorithm parameters are dynamically adjusted through the self-adaptive algorithm, the self-adaptivegenetic algorithm combining the self-adaptive algorithm and the genetic algorithm can keep the population diversity, meanwhile, the convergence of the genetic algorithm is guaranteed, and a train operation optimization curve is obtained. For a complex nonlinear heavy haul train operation process, a dynamic longitudinal dynamics model and a hook buffer device model of the train operation process are established, a train optimization operation model is established by applying a multi-objective genetic algorithm, a train operation curve is optimized, and safe, punctual and energy-saving operationof a train is realized.

The invention discloses a reactive power optimization method of an electric power system based on an improved CSO algorithm. The algorithm is a swarm intelligent search algorithm based on an improved CSO (ICSO) algorithm. The reactive power optimization method mainly comprises a horizontal cross operator, a longitudinal cross operator and a differential mutation operator. In horizontal cross, every two of all particles in a population are non-repeatedly paired in the horizontal cross, and the paired particles and the edges thereof are searched and updated in real time; in longitudinal cross, all dimensions are paired and then subjected to arithmetic cross; in differential mutation, all particles are subjected to mutation disturbance and cross and finally subjected to preferential selection; the three operators update the population through the selection operation, so that the convergence rate is accelerated and the population diversity is kept. The reactive power optimization method has the beneficial effects that the convergence speed of the ICSO algorithm is high, information exchange among individuals in the population is complete, the global convergence capability is strong, the particle diversity is good, and the reaction power optimization method has good applicability aiming at the high-dimensionality, multi-constraint and nonlinear complicated practical problem of reactive optimization of the electric power system.

The invention discloses a method and a system for optimally dispatching cascade reservoirs on the basis of quantum-behaved particle swarm algorithms. The method includes acquiring initialized population according to established objective functions for solving problems for optimally dispatching the cascade reservoirs and utilizing the initialized population as parent-generation particles; constructing small-world networks to obtain adjacent matrixes; updating the parent-generation particles according to the adjacent matrixes and generating child-generation particles; computing the fitness of the child-generation particles according to fitness functions; comparing the fitness of the parent-generation particles to the fitness of the child-generation particles by the aid of competition operators, selecting the child-generation particles with the good fitness and utilizing the selected child-generation particles as parent-generation particles for next iteration; judging whether current iteration numbers are larger than the maximum thresholds or not; carrying out computation if the current iteration numbers are larger than the maximum thresholds and outputting cascade reservoir optimal dispatching computation results. The method and the system have the advantages that the quantum-behaved particle swarm algorithms are improved by small-world network models, so that the population diversity can be kept by improved algorithms, the shortcoming of easiness in trapping in local optimization of basic quantum-behaved particle swarm algorithms can be overcome, and effects of optimally dispatching the cascade reservoirs can be improved.

The invention discloses an improved genetic algorithm-based m-sequence radar signal waveform optimization method which is implemented through the following steps of: dividing randomly-generated populations to be evolved into a ruling class, a middle class and a bottom class according to fitness in a descending order; carrying out evolution on the three classes by using different evolutionary strategies, wherein the ruling class is evolved by using overall crossover after the accessibility of the ruling class is limited by using a hamming distance, the middle class is evolved by using a standard genetic algorithm, and the bottom class is evolved by carrying out replacement by using randomly-generated new individuals; after the three classes are respectively evolved, completing a populationupdating process; and through multiple population updating processes, obtaining an optimal m-sequence radar signal. By using the method disclosed by the invention, a situation that an optimization process is in a local suboptimal state can be prevented, an optimal m-sequence can be converged quickly, and a situation that a biphase coded radar signal encoded by the sequence has a big pulse-pressure main-to-side lobe ratio under the condition of a certain bandwidth product in the process of signal emission can be guaranteed, thereby facilitating the detection on dim targets in a strong target echo pulse pressure side lobe area.

The invention provides a flexible job shop scheduling method and system. The method comprises steps that S1, the initial population S is generated based on basic parameters of a flexible job shop scheduling problem FJSP, and the initial population S is taken as a parent population P; S2, the parent population P is selected, crossed and mutated to obtain a temporary progeny population T; S3, basedon the temporary progeny population T, the parent population P is subjected to niche pre-selection operation to obtain a progeny population C; S4, a fitness value of each individual in the progeny population C is calculated, and the individual with the highest fitness value in the progeny population C is taken as the optimal solution of the FJSP; and S5, based on the optimal solution, a job shop corresponding to the FJSP is scheduled. The method is advantaged in that the solution search space can be made to maintain population diversity, the obtained solution can be guaranteed to converge to the global optimum, and thereby the job shop scheduling effect can be improved.

The invention discloses a DG grid connection optimization configuration method based on the improved genetic algorithm. The method comprises the following steps: Step 1, establishing an objective function based on system losses, voltage deviations and economical performances; Step 2, establishing equality constraints according to the objective function; Step 3, establishing inequality constraints;and Step 4, using the reciprocal of the objective function as a fitness function, and calculating the optimal solutions through the genetic algorithm to obtain an installation position and an installed capacity of a DG. The method disclosed by the invention has the advantages that the problem in the prior art that voltages of nodes in a power distribution system cannot be effectively improved after adoption of an installation position and an installed capacity of the DG because the installation position and the installed capacity of the DG are determined mostly based on empirical equations and human judgments is solved, the system losses are reduced, and the system operation stability and the power supply safety are improved.

The invention discloses a method and system for predicting wind speed. The method includes: obtaining the original sequence of wind speed data; using a particle swarm algorithm to determine the optimal preset scale parameter and optimal bandwidth parameter of the variational mode decomposition method, and converting the original sequence Decompose into several modal function subsequences; use the differential evolution algorithm to determine the kernel parameters of the least squares support vector machine model of each modal function subsequence, the variation factor of the mutation operation decreases with the increase of the evolutionary algebra, and the generated mutant individuals It is related to the optimal individual of the previous generation, and the crossover probability factor of the crossover operation increases with the increase of the evolutionary algebra; according to the autocorrelation of each modal function subsequence and each kernel parameter, determine the least squares of each modal function subsequence Support vector machine wind speed prediction sub-model, and predict the decomposition wind speed of each sub-sequence through each wind speed prediction sub-model; determine the final wind speed prediction value according to each decomposition wind speed. The method and system provided by the invention can accurately predict wind speed.

An optimization method of CSP solder joint structural parameters for reducing stress in power cycle-harmonic response coupling is disclosed. The method includes: 1) establishing a simulation analysismodel; 2) obtaining the thermal stress value of the solder joint; 3) determining the influencing factors of thermal stress value; 4) determining the parameter level value of the influencing factors; 5) obtaining an experimental sample; 6) obtaining a functional relationship between the influence factors and the thermal stress value; 7) carrying out regression analysis on that functional relation;8) establishing the correctness of the function relation; 9) generating an initial population in a random manner; 10) obtaining the current iterative number and the optimal fitness value; 11) performing crossover operation, mutation operation and evolutionary reversal on that individual; 12) selecting the optimal individual of fitness value; 13) re-judging aft population updating. The method combines a response surface and a genetic algorithm to reduce the stress in the CSP solder joint under power cyclic harmonic response coupling loading, exhibits excellent robustness, is simple in calculation ,and greatly brings convenience for optimization design of the structural parameters of the CSP solder joint in the later period.

Disclosed is an adjacent intersection road coordinate control method based on a minimum spanning tree clustering improved genetic algorithm. The adjacent intersection road coordinate control method based on the minimum spanning tree clustering improved genetic algorithm comprises establishing an adjacent intersection road coordinate control model; performing individual coding and data initialization, and setting parameters; performing initialization of a population; computing the fitness values of individuals in the population; performing minimum spanning tree clustering on the population; selecting individuals in the population to participate in genetic operation; performing cross and mutation operation on the selected individuals; performing repeated iteration to obtain an optimal individual. The adjacent intersection road coordinate control method based on the minimum spanning tree clustering improved genetic algorithm takes total vehicle delay as the performance index, establishes the more perfect adjacent intersection road coordinate control model by analyzing the traffic light status before vehicle fleets pass by intersections and determining whether vehicles before the intersections are completed released during a relief period, takes the optimal genetic algorithm as an optimization algorithm, and meanwhile, takes the factors of cycle length, signal timing and phase difference as parameters to achieve optimization solution.

The invention discloses an adaptive cuckoo and fireworks hybrid algorithm-based pressure container optimization design method, and belongs to the field of improvement and application of intelligent optimization algorithms. On one hand, the search step length of cuckoo can be adjusted according to a gap between current and optimal environmental adaptation values of a bird nest, and a discovery probability of bird eggs is calculated according to a standard deviation of adaptation values of individuals, so that the search efficiency of a population is improved; and on the other hand, a ''segmented value-taking'' method is adopted for the explosion radius in a fireworks algorithm, the number of sparks is determined according to a search range of an explosion point, and the individuals can perceive a gap between the current explosion point and the optimal value, so that the fireworks algorithm can perform a jumping cross-regional search. Two sub-populations that evolve independently are fused with each other through a fixed number of generations, so that information communication among different individuals can be enhanced. The method combines the advantages of two intelligent algorithms and has a good optimization effect in pressure container optimization design.

The invention belongs to the code design field of DNA calculation, relates to the group intelligent optimization algorithm and DNA codes and particularly relates to a DNA sequence optimization methodof the improved gravitational search algorithm based on chaotic and hybrid Gaussian variation. The method is characterized in that firstly, all the DNA sequences are generated in the D-dimensional search space as the initial population, and the improved gravitational search algorithm based on chaotic and hybrid Gaussian variation is utilized for optimization, through continuous loop iteration, anoptimal solution to an optimization problem can be finally obtained; for the gravitational search algorithm, calculating the total combined force of each individual in different directions is needed,the individual's acceleration is then calculated according to the total combined force, the speed and the position of each individual are updated according to the acceleration, and so on, when the algorithm reaches the maximum number of iterations, algorithm search stops, and the better quality DNA sequence code can be finally constructed. The method is advantaged in that an optimal DNA code sequence satisfying multiple constraints can be searched.

The invention relates to a lower bound estimation strategy adaptive protein structure prediction method. Under the evolutionary algorithm framework, firstly, a population is initialized, according tothe method of roulette, different strategies are chosen to mutate and cross the conformation; secondly, according to a lower bound estimation function, a Rosetta energy function score3 and the Monte Carlo probability acceptance criterion, the conformation is selected to guide the conformation update process, and the strategy selection probability is dynamically updated according to the historicalinformation of conformation update. The method is advantaged in that not only can a problem of inaccuracy of the energy function be alleviated, but also the conformation with lower energy and the morereasonable structure can be obtained by using historical information-guided algorithm sampling, sampling efficiency is improved, and the method further has properties of relatively high sampling efficiency and prediction precision.

The embodiment of the present invention discloses a multi-regional economic scheduling method based on the improved quantum particle swarm algorithm. The improvement of the quantum particle swarm algorithm by using the NW small world network can improve the shortcoming that the basic quantum particle swarm is easy to fall into a local optimum in the optimization process. The method in the embodiment of the present invention includes: S1: Establishing the objective function of the multi-regional economic scheduling problem; S2: Optimizing the objective function using the NW small-world network improved quantum particle swarm optimization algorithm, specifically including: S2-1: Population initialization; S2 ‑2: Build a small-world network and get the adjacency matrix; S2‑3: Update individuals and update the population; S2‑4: Calculate the fitness based on the updated population; S2‑5: Use the competition operator to adapt the parent particles Compared with the fitness of offspring particles, the one with better fitness is reserved as the parent of the next iteration; S2‑6: If the number of iterations calculated reaches the preset maximum number of iterations, calculate and output the multi-regional economic dispatch calculation If not, go to step S2‑2.

The invention discloses a NSGA-II-based improved transtainer scheduling method. According to the method, a crowding degree calculation method and control on an elitist range are improved on the basisof the NSGA-II (Elitist Non-dominated Sorting Genetic Algorithm); so that a crowding degree calculation method containing more information of individuals and an elitist range-improved NSGA-II algorithm can be introduced; the improved NSGA-II algorithm is applied to container yard / transtainer scheduling, and therefore, the operation completion time of the container filling of a yard and the transition time of a transtainer can be decreased; the improved NSGA-II algorithm, the NSGA-II and an NSGA-II+IMP algorithm are tested on four test functions, and test results show that the Pareto solution set of the improved NSGA-II of the invention has better distributivity and convergence and can better maintain population diversity.

The invention provides a high-dimensional multi-objective evolution method based on decomposition. The method includes: generating reference vectors, decomposing a high-dimensional multi-objective optimization problem into a plurality of single-objective optimization sub-problems; constructing a sub-population of a single-target optimization sub-problem based on the reference vector; distributingindividuals to the sub-populations by using a distribution mechanism; the method comprises the following steps: constructing a neighborhood sub-population, selecting individuals for genetic evolutionby using the constructed neighborhood sub-population, selecting individuals with excellent performance in the population by using a designed local and global selection strategy to enter the next generation of population, and repeatedly executing an evolution process until the evolution process is ended and a Pareto solution set of a high-dimensional multi-objective optimization problem is obtained. According to the method, the problem solving complexity is effectively reduced, the problem that good balance between population convergence and diversity is difficult to guarantee through a multi-objective optimization algorithm is solved, a Pareto solution set with good diversity and convergence is obtained, the algorithm efficiency is effectively improved, and the global convergence and population diversity of the algorithm can be effectively guaranteed.

The invention discloses a service workflow scheduling method oriented to energy consumption and safety perception, which is characterized in that safety service overhead is calculated by using the workload of safety service, the safety service is quantified, and meanwhile, a QoR model is established, so that a user can adjust a QoR level q according to the quality of a required result, and the energy consumption and the time are jointly considered. The result quality is adjusted by using the adjustable QoR level, and when the QoR level is lower, the workload required by processing the same data by using an original processing method is smaller, and the processing time required by the task is shorter, so that the energy consumption of mobile equipment is reduced. And meanwhile, the NSGA-II algorithm is utilized to carry out comprehensive optimization, so that the work flow delay and the equipment energy consumption are reduced.

The invention relates to a neutron spectrum real number solution genetic algorithm with linear constraint, a constraint penalty function is introduced in the algorithm process, priori information of afew groups can be fully utilized, and the accuracy and the calculation speed of the spectrum solution method can be effectively improved by using valuable priori information. Solving the problems that simple genetic algorithm is not accurate, and cannot effectively use the initial information problem and the solution space discrete. At the same time, solving the SAND-II iterative algorithm requires accurate cluster corresponding initial spectrum problem and the SAND-II iterative algorithm is heavily dependent on the initial spectrum.

The invention provides a dynamic pricing cloud computing service resource scheduling method and system, which belong to the technical field of cloud computing. The method comprises the steps of obtaining dynamic pricing instance data of various specifications on a plurality of cloud service platforms in real time to obtain the price of a dynamic pricing instance at the next moment; obtaining a leasing scheme of the cloud computing service according to a self-adaptive differential evolution algorithm by taking minimum cost as a target and combining the workload demand, the price of the dynamicpricing instance at the next moment and the available instance type and quantity; and dynamically scheduling cloud computing service resources according to the obtained leasing scheme. For the dynamically priced cloud service instance, the instance combination renting problem and the bidding problem of a user are optimally designed by combining the type and the task load of the instance, so that the use experience of the user is improved, and the renting cost of the user is reduced.

The invention provides a genetic algorithm intelligent test paper composition method and device based on segmented real number coding and a readable storage medium, and the method comprises the steps:1, performing test question coding: mapping each set of test paper into an independent chromosome in a genetic algorithm, mapping test questions in the test paper into genes, wherein gene values needto be completed under the support of a coding mechanism; 2, generating an initial group, and adopting a random extraction mode to generate N sets of test papers, namely an initial population; and 3,calculating individual fitness, calculating fitness values of all individuals in the initial population obtained through initialization, and carrying out quality evaluation; 4, selecting operators, eliminating inferior individuals according to the fitness values, and reserving excellent individuals in the population; 5, generating a new operator, wherein the new operator is generated in a gene crossing and mutation mode; and 6, setting a termination condition, searching test questions matched with characteristic parameters in a test question bank according to a given constraint condition so asto extract the optimal test question combination, and selecting an individual with the highest fitness value as a final test paper.

Disclosed is an adjacent intersection road coordinate control method based on a minimum spanning tree clustering improved genetic algorithm. The adjacent intersection road coordinate control method based on the minimum spanning tree clustering improved genetic algorithm comprises establishing an adjacent intersection road coordinate control model; performing individual coding and data initialization, and setting parameters; performing initialization of a population; computing the fitness values of individuals in the population; performing minimum spanning tree clustering on the population; selecting individuals in the population to participate in genetic operation; performing cross and mutation operation on the selected individuals; performing repeated iteration to obtain an optimal individual. The adjacent intersection road coordinate control method based on the minimum spanning tree clustering improved genetic algorithm takes total vehicle delay as the performance index, establishes the more perfect adjacent intersection road coordinate control model by analyzing the traffic light status before vehicle fleets pass by intersections and determining whether vehicles before the intersections are completed released during a relief period, takes the optimal genetic algorithm as an optimization algorithm, and meanwhile, takes the factors of cycle length, signal timing and phase difference as parameters to achieve optimization solution.

A population protein structure prediction method based on global and local strategy cooperation is disclosed. The method is characterized by under the framework of a differential evolution algorithm,for each conformation, firstly executing a global mutation strategy to carry out the global detection of a conformation space so as to obtain a potential area; and then, executing a local mutation strategy to carry out local search on the detected area so as to obtain a better conformation. In a global detection process, segment exchange is performed through the randomly-selected conformation andsimultaneously the multiple conformations are generated and tested, and the test conformation with lowest energy is selected. In a local search process, the conformation which is better than the target conformation is selected to guide mutation so as to generate multiple test conformations, and the conformation with lower energy is selected. Through the writing of the global strategy and the localstrategy, a relationship between a balance diversity and a convergence speed is achieved. By using the population protein structure prediction method based on the global and local strategy cooperation, prediction precision and search efficiency are high.

The invention discloses a crowdsourcing task data recommendation method based on a decision model and a genetic matrix decomposition method. Comprising the following steps: 1) preprocessing historical data in a crowdsourcing platform to obtain preprocessed historical data, and establishing a task feature matrix and a user feature matrix based on the preprocessed historical data; 2) performing information matching on the task feature matrix and the user feature matrix by using a user knowledge fusion decision method to obtain an initial capability matching matrix; 3) establishing a matching decision model according to the task feature matrix and the user feature matrix; and 4) according to the initial capability matching matrix, solving the matching decision model by using a genetic matrix decomposition algorithm to obtain the matching degree between the user and the plurality of tasks, and performing crowdsourcing task recommendation for the user by the crowdsourcing platform based on the matching degree between the user and the plurality of tasks. According to the invention, the precision and efficiency of task recommendation on a crowdsourcing platform are improved, and the realistic problems of task overload and low task matching efficiency in a crowdsourcing scene are effectively solved.