外文翻譯：現(xiàn)代控制理論的發(fā)展-資料下載頁
2025-05-12 04:59本頁面
【導讀】機器,實現(xiàn)其目標的自動控制。智能控制的注意力并不放在對數(shù)學公式的表達、計算和處理。上,而放在對任務和模型的描述,符號和環(huán)境的識別以及知識庫和推理機的設(shè)計開發(fā)上。同時具有以知識表示的非數(shù)學廣義模型和以數(shù)學模型表示的混合控制過程;提高系統(tǒng)的學習能力和自主能力;與線性系統(tǒng)平行。情況是不能適用的。李雅普諾夫方法:它是迄今為止最完善、最一般的非線性方法,但是由于它的一般性,性、準線性控制以及空間測度被破壞等。境的變化,達到所要求的控制性能指標。李雅普諾夫方法,對不確定性以狀態(tài)空間模式出現(xiàn)時是一種有利工具;模糊控制既不是指被控過程是。統(tǒng),通過計算機實現(xiàn)模糊控制往往能取得很好的結(jié)果?；灸：刂破?一旦模糊控制表確定之后,控制規(guī)則就固定不變了;模糊控制有待進一步研究的問題:模糊控制系統(tǒng)的功能、穩(wěn)定性、最優(yōu)化問題的評價;神經(jīng)網(wǎng)絡(luò)在控制系統(tǒng)中可充當對象的模型,還可充當控制器。
【正文】 was accused of process control effect is improved, and achieve the desired effect。 (3) Intelligent fuzzy controller, which, artificial intelligence and neural work linked to three, to realize prehensive information processing, make the system not only has the flexible reasoning mechanism, heuristic knowledge and production rules said, and has many layers, various types of control laws of choice. The characteristics of the fuzzy control is not to need to accurate mathematical model, robust, the control effect is good, easy to overe the influence of nonlinear factors, the control method is easy to grasp. Recently, someone put forward fuzzy neuralInter3 fusion control model, that is, the fusion structure, fusion algorithm and control is an organic whole to carry on the design. Again someone proposed by homotopy BP work memory fuzzy rules to lenovo way use these experience. Fuzzy control problems to be further studied: the fuzzy control system function, stability, optimization evaluation。 Complicated nonlinear system of fuzzy modeling and the fuzzy rules and the establishment of fuzzy inference。 Find out to follow the general design principles. 6. The Neural Network Control (Neural Network Control) Neural work is the simple unit by socalled neurons in parallel structure after adjustable connection power consists of the work. There are many kinds of neural work, control the monly used have multilayer feedforward BP work, the RBF work, the Hopfield puter and adaptive resonance theory model (ART), etc. The neural work control by using neural work is the tool from mechanism on the simple structure of the new simulation control and identification methods. Neural work in the control system can serve as a object model, also can serve as a controller. Common neural work control structure are: (1) Parameter estimation adaptive control system。 (2) Internal model control system。 (3) Predictive control system。 (4) Model reference adaptive system。 (5) Variable structure control systems. The neural work control of key features are: can describe any nonlinear system。 Used for nonlinear system identification and estimate。 For a plicated uncertainty of the adaptive。 Quick optimization calculation capacity。 With a distributed storage capacity, which can realize the online and offline learning. Recently, someone to put forward the Hopfield puter implementation of a multiresolution stered coordination algorithm, based on the level of the fusion of automatic way finish by coarse to fine until the full resolution matching and established. Again someone put forward a kind of selforganization work controller, the change of the steepest slope gradient descent learning algorithm, the application of the nonlinear tracking control in. In the future the problem of further discussion is to improve the work learning speed, put forward the new work structure, create more suitable to the control of special neural work.
畢業(yè)設(shè)計相關(guān)推薦
鄂ICP備17016276號-1