{"title":"Identification of an ARMAX model based on a momentum-accelerated multi-error stochastic information gradient algorithm","authors":"Shaoxue Jing","doi":"10.1109/DDCLS52934.2021.9455685","DOIUrl":null,"url":null,"abstract":"The ARMAX model is widely used in industrial modeling. However, the traditional stochastic information gradient algorithm for ARMAX identification needs less computation, but its convergence speed is too slow. To accelerate the algorithm, we propose a two-step algorithm based on a gradient acceleration strategy. The first step is to replace the error scalar with the error vector, and the second step is to introduce a momentum related to the gradient. The simulation results show that the proposed algorithm can obtain more accurate estimation and the convergence speed is greatly improved.","PeriodicalId":325897,"journal":{"name":"2021 IEEE 10th Data Driven Control and Learning Systems Conference (DDCLS)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2021-05-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2021 IEEE 10th Data Driven Control and Learning Systems Conference (DDCLS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/DDCLS52934.2021.9455685","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
Abstract
The ARMAX model is widely used in industrial modeling. However, the traditional stochastic information gradient algorithm for ARMAX identification needs less computation, but its convergence speed is too slow. To accelerate the algorithm, we propose a two-step algorithm based on a gradient acceleration strategy. The first step is to replace the error scalar with the error vector, and the second step is to introduce a momentum related to the gradient. The simulation results show that the proposed algorithm can obtain more accurate estimation and the convergence speed is greatly improved.