Font Size: a A A

Application Research Of Symbiotic Organisms Search Algorithm

Posted on:2018-09-24Degree:MasterType:Thesis
Country:ChinaCandidate:H Z WuFull Text:PDF
GTID:2348330512487084Subject:Computer system architecture
Abstract/Summary:PDF Full Text Request
Symbiotic organisms search(SOS)algorithm is a novel meta-heuristic optimization algorithm which simulates the symbiotic relationships adopted by organisms to survive and propagate in the ecosystem.The structure of symbiotic organisms search algorithm is simple,and it does not use tuning parameters,equips with strong robustness,powerful search ability and easy execution.According to the advantages,scholars at home and abroad got extensive interested in the algorithm and widely applied it to complex optimization and engineering calculation,and so on.In the process of the research,the scholars discover the drawbacks that the algorithm performs in slow convergence speed and low accuracy for some particular problems,which limit its application range.In this paper,symbiotic organisms search algorithm was applied in practical problems.In addition,the algorithm was improved in structure and the evolutionary strategys on account of the drawbacks.The main purpose to improve the algorithm is to consolidate its theoretical basis and expand its range of application.The main work of this paper includes the following three aspects:(1)For training the feedforward neural network,gradient-based training algorithm like back propagation algorithm is easy to fall into local optimum and sensitivity to the initial value.Symbiotic organisms search algorithm performs well in exploration and exploitation.By means of analysis and comparison,the simulation results show that the algorithm performed in high accuracy and was efficiency for optimizing the weights and biases.(2)The traditional methods,such as k-means,are sensitivity to the initial value and likely trapped into local optimal in solving clustering analysis problems.While,symbiotic organisms search algorithm enhances the global search ability,and avoids falling into local optimal,meanwhile improves the execution efficiency.Simulation results indicate that the algorithm improves the clustering ability in data sets.(3)To enhance the local search ability of symbiotic organisms search algorithm,the strategy of harmony search was adopted to make up a hybrid algorithm which consists of four phases: the mutualism phase,commensalism phase,parasitism phase and harmony phase.In solving the 0-1 knapsack problem,the hybrid algorithm was discretized and the greedy search strategy was adopted to modify the feasible solution.The aim of the mechanism aforementioned is to improve the ability of the algorithm both in local search and global search.
Keywords/Search Tags:Symbiotic organisms search, Harmony search, Greedy strategy, Clustering analysis, 0-1 knapsack problem, Feedforward neural network, Meta-heuristic algorithm
PDF Full Text Request
Related items