【摘要】：大型模鍛成形過程是一個復雜的非線性時變過程,包括鍛件流變成形過程與液壓系統(tǒng)驅動過程,以及還存在油液泄漏等眾多不確定性因素,導致精準鍛造過程控制異常困難。為此,在結合基于機理模型控制與數(shù)據(jù)控制優(yōu)點的基礎上,提出了基于物理模型結合在線順序極限學習機的智能控制方法。該方法首先使用已知的系統(tǒng)信息推導出名義控制律;其次,針對模型不確定性部分,使用在線順序極限學習機設計出該在線模型的補償控制律;最后,建立了基于機理模型與數(shù)據(jù)模型的集成控制器,獲得了最佳控制律。仿真結果表明,新方法能有效地控制復雜的鍛造過程,且比現(xiàn)有的方法有更好的控制精度。
[Abstract]:Large-scale die forging process is a complex nonlinear time-varying process, including forging rheological forming process and hydraulic system driving process, as well as the existence of oil leakage and many other uncertain factors, resulting in the precision forging process control is extremely difficult. Based on the advantages of mechanism-based model control and data control, an intelligent control method based on physical model and on-line sequential learning machine is proposed. Firstly, the nominal control law is derived by using the known system information, secondly, the compensation control law of the online model is designed by using the on-line sequential limit learning machine for the uncertain part of the model. Finally, an integrated controller based on mechanism model and data model is established, and the optimal control law is obtained. The simulation results show that the new method can effectively control the complex forging process and has better control precision than the existing methods.
【作者單位】： 中南大學機電工程學院高性能復雜制造國家重點實驗室;
【基金】：國家重點基礎研究發(fā)展計劃(“973”計劃)項目(2011CB706802) 國家自然科學基金資助項目(51205420) 新世紀人才計劃基金(NCET-13-0593) 湖南省自然科學基金資助項目(14JJ3011)
【分類號】：TG316.3
，

本文編號：2375454