501 results about "Crossover" patented technology

The invention discloses an intelligent layout method used for a rectangular part. The method comprises the steps that S1 relative parameters of the genetic algorithm are initialized; S2 relative information of the rectangular part is extracted from a rectangular part bank to be laid out; S3 relative information of raw material boards is extracted from a board tank; S4 the obtained information is coded, and primary species are generated randomly; S5 one-by-one decoding is conducted on the primary species by means of the lowest horizontal line search algorithm to obtain solution using efficiency; S6 selection, crossover and mutation operation is conducted according to the genetic algorithm until iteration is finished, and the optimal layout scheme is output. According to the intelligent layout method, the process requirement of the rectangular part can be met well, the intelligent algorithm and the heuristic algorithm are combined, one optimizing scheme can be found rapidly and efficiently, and therefore the material using rate of an enterprise is greatly improved, layout time can be obviously shortened, and layout efficiency is improved.

The invention relates to a three-dimensional encasement novel genetic algorithm model under a multi-constrain condition. At present, the logistics transportation industry rapidly develops, but encasement plan decision models are not perfect; particularly, a three-dimensional encasement model under the multi-constrain condition has the following typical problems: 1, the time complexity is higher; 2, the space utilization ratio is lower; 3, a encasement plan cannot be perfect. The design considers the constraint conditions such as the space utilization ratio, the centre-of-gravity position and the load bearing in the encasement problem, combines an improved genetic algorithm with a Monte Carlo method, a gene injection algorithm and a non-dominated sorting algorithm, aims at improving the space utilization ratio of the encasement plan and reducing the time complexity of the algorithms under the multi-constrain condition, and belongs to the field of intelligent arithmetic optimization. The three-dimensional encasement novel genetic algorithm model is mainly characterized in that population is initialized through the Monte Carlo method based on normal distribution; the gene injection algorithm is used in the encasement decision mode; the probability of crossover, mutation and gene injection operators is fitness functions; an online space combining method is used.

The invention provides an optimization algorithm based on multi-objective resource-constrained project scheduling model. The scheduling model requires scheduling the start time of each activity to achieve the optimal goal under the condition of satisfying the relevant constraints. Based on the RCPSP model, the invention introduces the optimal resource balance as the objective, and expands the model to a multi-objective model. The solution of RCPSP is mainly based on the heuristic algorithm. When the task list is used to encode chromosomes in the heuristic algorithm, the task list initialized randomly may not satisfy the constraint relation between the top and bottom. The invention provides an individual generation mode based on control relation, which presents a new crossover operator andmutation operator based on the NSGA-II algorithm. The invention can greatly reduce the time complexity of the algorithm, realize the balanced allocation of resources, improve the production efficiencyand save the production cost while ensuring the solution precision of the algorithm, thereby improving the economic benefit of the resource scheduling production process.

The invention discloses a parameter optimization control method of semiconductor advance process control (APC). In semiconductor technological process, a traditional method uses a linear prediction model for the optimization control method of batch process. The parameter optimization control method of the semiconductor advance process control uses an optimized back propagation (BP) neural network prediction model based on genetic algorithm, optimizes the initial weight values and threshold values of the neural network through the genetic algorithm, uses selecting operation, probability crossover and mutation operation and the like according to the fitness function F corresponding to each chromosome, and outputs the optimum solution finally to determine the optimum initial weight value and the threshold value of the BP neural network. The performance of the BP neural network is improved with an additional momentum method and variable learning rate learning algorithm being used, so that the BP neural network after being trained can predict the non-linear model well. The genetic algorithm in the method has good global searching ability, a global optimal solution or a second-best solution with good performance is easy to obtain, and the genetic algorithm well promotes the improvement of modeling ability of the neural network.

The invention discloses an electric power Internet of Things task allocation method based on edge cooperation. The method comprises the steps that a two-point cooperation task allocation model is established according to all service requests and EN idle resources of the electric power Internet of Things at a certain moment; a task allocation problem model is constructed by taking the minimized average task completion delay as an optimization target; and the task allocation problem model is solved by applying a TENCTA-IPSO algorithm. According to the method, the task completion time delay of the power grid service is effectively reduced. According to the TENCTA-IPSO algorithm, a particle updating strategy in a particle swarm algorithm is improved by utilizing crossover and mutation operations of a genetic algorithm; the particle swarm diversity is improved while the self-learning ability of the particles is kept, premature is prevented from falling into local optimum, and the calculatedtask allocation scheme shortens the average task completion delay to the maximum extent.

The invention provides a self-crossover genetic algorithm for solving a flexible job-shop scheduling problem. The algorithm relates to the field of job-shop scheduling, and particularly relates to the field of flexible job-shop scheduling. The existing genetic algorithms are mostly amphilepsis, the coding mode is complex, crossover and variation are caused to be complex, and a non-feasible solution is easy to acquire. The invention provides monolepsis-based self-crossover whose coding, crossover and variation are performed on a uniparental chromosome. The coding uniparental chromosome is divided into a working procedure portion and an equipment portion, wherein the working procedure portion is coded based on the workpiece number, and the equipment portion represents selected equipment by using the probability. Self-crossover is performed on the working procedure portion, and the equipment portion also performs the same crossover transform along with the working procedure portion. Two types of variation operators are adopted, exchange type variation is adopted for the working procedure portion, and insertion type variation is adopted for the equipment portion. The self-crossover genetic algorithm provided by the invention has the characteristics of high practicability and wide application range.

Systems and methods are disclosed that utilize a genetic algorithm to search for an index configuration for a collection of data such as, e.g., a database. Genetic algorithms can include stochastic search heuristics that mimic processes of natural evolution including inheritance, mutation, crossover, and selection. A population of chromosomes representing candidate index configurations can evolve to increase or optimize the fitness of the population and to identify the best (e.g., most fit) index configuration. Fitness of a chromosome may be measured based at least in part on the cost of computer resources used for executing Structured Query Language (SQL) statements in the indexed database. In various implementations, virtual indexing may be used to simulate building an index, chromosomes may be encoded using non-bitmapped representations of index configurations, chromosomes may include genes representing a column in a table in a database, dropping an index from a table in a database, or a composite index for a database, and/or a participation pool may be used to select fitter genes for an initial population of chromosomes.

The invention provides a wing profile optimal design method of a parallel difference evolutionary algorithm based on open computing language (Open CL). The wing profile optimal design method is used for wing profile design. A standard wing profile function and a profile function are selected, the profile function serves as a design variable, an optimization objective function is determined, steps of the difference evolutionary algorithm are divided into different stages according to processed data, and all stages are packaged in different cores for operating based on Open CL. In population updating, mutation operators are used for generating test vectors and crossover operators are used for generating descendants; individuals in a population are restored to the wing profile shape and tested whether to meet geometric constraint; computational fluid dynamics (CFD) analysis is used for obtaining fitness of the individuals and seeking the optimum individual, and whether the optimum individual meets performance constraint is tested; and finally iteration is finished and the optimum result is copied back into a host internal storage. The wing profile optimal design method achieves parallel processing of a wing profile design process, performs sufficient search in effective space, shortens design period, achieves cross-platform wing profile optimal design, and improves design efficiency.

The invention provides a critical process-combined genetic local search algorithm for solving flexible job-shop scheduling and aims at solving the problems that an infeasible solution can be generated by a genetic algorithm in an operation and the solving result is unstable to cause insufficient local search capability due to the randomness of local search in the prior art. The disadvantages are overcome by combining the methods of a cross repairing procedure, a critical process-based local search method and the like. According to the genetic local search algorithm, a novel vector-based encoding mode is adopted when populations are initialized, a gene repairing procedure is carried out during crossover operation of the genetic algorithm and the search range is controlled by adopting critical process-based search. The flexible job-shop scheduling problem can be solved, and the genetic local search algorithm has the characteristic of high practicability.

The invention discloses an RFID network topology method based on the multi-agent evolutionary algorithm and belongs to the field of automatic control and information technology. The method combines a multi-agent system with evolutionary computation to solve the RFID network topology problem. The method is characterized in that each agent in an agent grid is initialized according to two algorithms, and then a neighborhood competition operator, a neighborhood right-angled crossover operator, a mutation operator and a self-learning operation are designed to optimize the agents. Verification results show that the method has great advantages in terms of solving the maximum coverage rate of reader-writers, the minimum number of the reader-writers and the minimum interference rate of the reader-writers when used for evaluating and solving the RFID network topology problem and is an effective method for solving the RFID network topology problem.

The invention relates to a method for optimizing a multi-grade light-splitting passive optical network of a distribution communication network. The method is that the network at the current grade bears the gene optimization result of the network at the previous grade through the cascading genetic algorithm. The method specifically comprises the steps of designing a gene code; creating constraint conditions for the gene code, and selecting adaptability functions; performing crossover and variation operation for the gene code on the premise that the constraint conditions for the gene code are met; performing multi-generation iterative crossover and variation operation to obtain the optimal gene code of the PON network at the current level; returning to step (1) to bear the gene code of previous level when the number of the network layers of the multi-grade light-slitting passive optical network is less than n, and updating the adaptability function; finishing the optimization when the number of the network layers of the multi-grade light-slitting passive optical network is more than 1. According to the method, the optimal light splitter and relatively high network star topology in the network are selected from the optimized multi-grade light-splitting network, and therefore, the network communication construction cost is saved; the network planning is constructed into a mathematical model, thus the expandability is improved, and the calculation complexity is reduced.

The invention discloses an unbalanced data set oversampling method based on a genetic algorithm and k-means clustering, and the method comprises the following steps: inputting an original unbalanced data set, and dividing the unbalanced data set into a training data set and a testing data set; dividing the training data set into a positive class sample set and a negative class sample set; clustering the positive class sample set by using a k-means clustering algorithm to obtain a plurality of different clusters; allocating corresponding sampling weights to the number of samples in each cluster; calculating the Mahalanobis distance of the sample data in each cluster, and dividing the sample data into two groups of parent class sample data sets according to the Mahalanobis distance; according to a crossover operator in the genetic algorithm, forming a new positive class sample by by utilizing the parent class sample data set; combining the newly synthesized positive class sample and theoriginal training data set into a balanced data set; training a classifier model by utilizing the balance data set; and evaluating the performance of the classifier model by utilizing the test data set. According to the method, the classification accuracy of the classifier model on the positive samples in the unbalanced data set can be effectively improved.

The invention discloses a BP neural network wind speed prediction method based on genetic algorithm optimization. The method comprises the following steps: firstly, collecting wind speed data of a wind power plant, establishing a BP neural network prediction model, and estimating an initial value range; then, performing real number coding on the weight and the threshold of the neural network, randomly generating a group of initial individuals to form an initial population, and each initial individual represents an initial solution of a problem; calculating the fitness of each individual in thepopulation, performing selection, crossover and mutation operations to form a next generation of population, evaluating the fitness of the individuals in the new population, judging convergence conditions, selecting an optimal individual, and taking the optimal individual as an initial weight and a threshold of the neural network; and finally, training by utilizing matlab to obtain a wind speed prediction value. According to the method, the wind speed prediction efficiency and accuracy of the BP neural network are improved.

The invention discloses a multi-target flexible job shop scheduling method based on an improved niche genetic algorithm. Constructing a production scheduling sequence according to the process data ofall the workpieces in the multi-target flexible job shop, taking the production scheduling sequence as an individual, and generating a primary population; calculating a total objective function valueof the individual, and calculating a fitness value of the individual by using an improved niche method; selecting an individual set in a roulette mode according to the fitness value; implementing crossover operation and mutation operation of the genetic algorithm; forming a new population by the obtained individuals and the individuals with the highest fitness value in the generation population; repeating the steps until a termination condition is met, outputting an optimal individual in the last generation population, and arranging processing treatment by adopting a scheduling sequence of theoptimal individual, so as to realize multi-target flexible job shop scheduling. The improved ecological niche genetic algorithm is adopted to solve the scheduling problem in the production process, ahigh-quality scheduling result can be stably obtained, workshop resource allocation is optimized, and therefore the production efficiency of a workshop is improved.

The invention provides a DNN task unloading method in an edge-cloud hybrid computing environment and a terminal. According to the types and the number of computing nodes, the number of DNN tasks to be unloaded and the number of layers of each DNN task to be unloaded, calculating the number of the DNN tasks to be unloaded; establishing a target function based on total cost minimization; determining a corresponding constraint condition. The influences of conditions such as computing power and time delay constraints of different types of nodes are considered, the feasibility of the obtained optimal solution is guaranteed, when the optimal solution is solved, crossover operation and mutation operation in the genetic algorithm are introduced into the particle swarm algorithm, a specific algorithm is given, and the problem that the particle swarm algorithm is prone to falling into local optimum in the optimal solution solving process is effectively solved.

The invention discloses a hybrid genetic algorithm-based calculation unloading method in mobile edge calculation, which comprises the following steps: S1, establishing a system model to obtain calculation time delay of a sub-task set in each processor and transmission time delay among the processors, and determining each task layer value in the sub-task set according to a constraint relationship of the sub-task set; S2, initializing a population according to the determined task layer value and a random strategy to obtain initial population individuals of the sub-task set, performing symbol coding to obtain a task scheduling sequence, and optimizing the individuals in the initial population; S3, constructing a fitness evaluation function, and performing selection operation on individuals inthe optimized initial population; S4, constructing a crossover mechanism, and crossover individuals in the new population by using crossover operation based on a tabu table search algorithm; S5, performing mutation operation on individuals in the new population by using mutation operation based on a simulated annealing algorithm; and S6, judging whether the iteration step length is reached or not, and if not, repeating the step S3-S5; and if so, outputting a globally optimal solution.

The invention provides an agent modeling method for multi-objective compilation optimization sequence selection, aims to provides a solution for solving a calculation cost constraint problem of the multi-objective compilation optimization sequence selection and belongs to the field of compiler optimization. The method comprises the following steps of: firstly, performing binary coding on a compilation optimization sequence; respectively designing fitness functions for the scale and the running speed of two optimization target codes; generating a child population after selection and crossover operations; combining the child population with a parent population; performing quick non-dominated sorting to generate a new population; and finally obtaining a Pareto optimal solution set. In the search iteration process, the compilation optimization sequence and the two corresponding target fitness values are used for constructing an agent model, the agent model is used for calculating approximate fitness values for the child population generated by evolution operation, and actual fitness values are calculated for excellent solutions, so that the evolution efficiency is improved. According to the method, the compilation optimization sequence meeting multiple objectives (such as running speed and code scale) can be effectively selected for a program to be compiled, and the problem of calculation cost constraint caused in the iteration process is solved.

The invention provides a workpiece intelligent scheduling production scheduling method based on a genetic algorithm. The method comprises the steps of randomly generating an initial population according to order data, and randomly combining an order number, a to-be-processed process, a processing machine and an operator list in the initial population to obtain a first initial population; accordingto a preset time sequence rule, setting starting time of the to-be-processed process in the first initial population to obtain a second initial population; sequentially performing first quick sortingprocessing, crossover operation processing, mutation operation processing and second quick sorting processing on the second initial population to obtain a new initial population, and decoding order data in the new initial population to obtain a scheduling plan; calculating the fitness corresponding to the production scheduling plan according to the production scheduling plan and a preset fitnessfunction; and when the fitness meets a preset optimal condition, outputting the production scheduling plan as an optimal production scheduling plan. According to the workpiece intelligent scheduling production scheduling method based on the genetic algorithm, the obtained optimal production scheduling plan is more accurate.

The invention discloses a crossover probability factor adjustable differential evolution algorithm. The algorithm comprises the following steps of (1) performing population initialization; (2) performing mutation operation; (3) performing crossover operation; (4) selecting an operator; and (5) stopping iteration. According to the crossover probability factor adjustable differential evolution algorithm provided by the invention, in a crossover link, different individuals are endowed with different crossover probability factor values according to the previous-generation fitness values of the different individuals; and compared with a conventional DE algorithm, when the crossover probability factor adjustable differential evolution algorithm is used for performing numerical optimization, theneeded iteration frequency is remarkably reduced, an expected optimal value can be found more quickly, and the calculation time and the memory consumption are effectively shortened and reduced.