{"title":"Simultaneous perturbation stochastic algorithm parameters effect towards multi agent robot broadcast controller","authors":"N. Sariff, Z. Ismail","doi":"10.1109/USYS.2017.8309456","DOIUrl":null,"url":null,"abstract":"This paper investigates the effect of Simultaneous Perturbation Stochastic Algorithm towards multi agent motion coordination performances. Parameters of a, A, a, c and y were applied based on the guidelines given and based on the optimal value found from investigation. Each parameter will be varied and the results will be observed to determine the optimal parameters for this case study. The results show that with optimal parameters of ak and ck gain of SPSA which are a=0.21, A=30, a=0.602, c=1, y=0.101 and with a quadratic loss function, the optimal result was proven achieved. The results of convergence time, percentage of agent movement and the number of iteration were recorded to verify that the agent capable to reach its desired target successfully. The reading shows that in average, the agent will reach its target in 14.53 sec with 138 number of iterations in a 5 times run.","PeriodicalId":137478,"journal":{"name":"2017 IEEE 7th International Conference on Underwater System Technology: Theory and Applications (USYS)","volume":"118 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2017-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2017 IEEE 7th International Conference on Underwater System Technology: Theory and Applications (USYS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/USYS.2017.8309456","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
Abstract
This paper investigates the effect of Simultaneous Perturbation Stochastic Algorithm towards multi agent motion coordination performances. Parameters of a, A, a, c and y were applied based on the guidelines given and based on the optimal value found from investigation. Each parameter will be varied and the results will be observed to determine the optimal parameters for this case study. The results show that with optimal parameters of ak and ck gain of SPSA which are a=0.21, A=30, a=0.602, c=1, y=0.101 and with a quadratic loss function, the optimal result was proven achieved. The results of convergence time, percentage of agent movement and the number of iteration were recorded to verify that the agent capable to reach its desired target successfully. The reading shows that in average, the agent will reach its target in 14.53 sec with 138 number of iterations in a 5 times run.