基于bp和rbf神經(jīng)網(wǎng)絡(luò)的溫室溫度pid控制.doc
約49頁DOC格式手機(jī)打開展開
基于bp和rbf神經(jīng)網(wǎng)絡(luò)的溫室溫度pid控制,摘 要 隨著現(xiàn)代工業(yè)向大型化、集成化方向發(fā)展,生產(chǎn)過程日趨復(fù)雜,過程嚴(yán)重非線性、時(shí)變性、不確定性及變量間的強(qiáng)耦合使許多系統(tǒng)缺乏精確的數(shù)學(xué)描述,難以用傳統(tǒng)的理論方法分析和控制,因此有必要研究新的智能控制策略。我國目前已有研究所綜合應(yīng)用自動(dòng)控制、計(jì)算機(jī)應(yīng)用、人工智能以及化學(xué)工程學(xué)科的理論和技術(shù),對基于神經(jīng)網(wǎng)絡(luò)的...
內(nèi)容介紹
此文檔由會(huì)員 違規(guī)15 發(fā)布
摘 要
隨著現(xiàn)代工業(yè)向大型化、集成化方向發(fā)展,生產(chǎn)過程日趨復(fù)雜,過程嚴(yán)重非線性、時(shí)變性、不確定性及變量間的強(qiáng)耦合使許多系統(tǒng)缺乏精確的數(shù)學(xué)描述,難以用傳統(tǒng)的理論方法分析和控制,因此有必要研究新的智能控制策略。我國目前已有研究所綜合應(yīng)用自動(dòng)控制、計(jì)算機(jī)應(yīng)用、人工智能以及化學(xué)工程學(xué)科的理論和技術(shù),對基于神經(jīng)網(wǎng)絡(luò)的軟測量和智能控制技術(shù)及控制軟件進(jìn)行了深入研究。神經(jīng)網(wǎng)絡(luò)控制的研究,可以更好的解決復(fù)雜的非線性、不確定、不確知系統(tǒng)在不確定、不確知環(huán)境中的控制問題。神經(jīng)網(wǎng)絡(luò)控制不需建立精確的數(shù)學(xué)模型,能夠自動(dòng)辨識(shí)被控過程參數(shù)、自動(dòng)整定控制參數(shù)、適應(yīng)被控過程參數(shù)的變化,是解決傳統(tǒng)PID控制器參數(shù)整定難、不能實(shí)時(shí)調(diào)整參數(shù)和魯棒性不強(qiáng)的有效措施。
本文在分析了神經(jīng)網(wǎng)絡(luò)控制系統(tǒng)之后,主要進(jìn)行了兩個(gè)方面的研究。其一:對基于BP神經(jīng)網(wǎng)絡(luò)整定的PID控制的研究;其二:是對基于RBF神經(jīng)網(wǎng)絡(luò)辨識(shí)整定的PID控制研究。編寫了基于BP和RBF網(wǎng)絡(luò)的仿真程序,仿真結(jié)果表明,應(yīng)用神經(jīng)網(wǎng)絡(luò)對常規(guī)PID控制器進(jìn)行改進(jìn)后提高了系統(tǒng)的魯棒性和動(dòng)態(tài)特性,有效的改善了系統(tǒng)的控制結(jié)果,達(dá)到了預(yù)期的目的。
Abstract
With the development into large-scale and integrationg of modern industry,the product process tends to become complex. Many systems are short of accurate mathematics description, because the process is nonlinear, time varying, uncertain and the srong coupling of variable. Then it is difficult o analyze and conrol with traditional method,so we need research new intelligent control strategy. At present, the Institute of Integrated application of automatic control, computer applications, artificial intelligence and the theory of chemical engineering and technology, it is have a depth research in Neural network,which based soft sensor and intelligent control technology and control software . In the study of neural network control,you can solve the control problem better,in the complex non-linear, uncertain,unknown system with uncertainty and unknown environment. Neural network control can automatically identify controlled process parameters, automatic tuning control parameters, to adapt to changes in controlled process parameters, but do not need to establish an accurate mathematical model. It can effectively solve the traditional PID controller parameter tuning difficultly, not adjust the parameters in real time and not stronger in robustness issues
After analyzing the control system of neural network, this paper mainly dose research in two aspecs.Firstly, Research based on the BP neural network setting PID control;Secondly, based on RBF neural network to identify setting PID control study. This paper gives the overall research plan for this system. Also given the simulation program and analysis of BP network and RBF network.
目 錄
摘 要 1
Abstract 2
引 言 1
1 緒論 2
1.1 智能控制的發(fā)展與展望 2
1.1.1 智能控制的興起 2
1.1.2 傳統(tǒng)控制和智能控制 3
1.1.3 智能控制的展望 4
1.2 神經(jīng)網(wǎng)絡(luò)的發(fā)展與展望 5
1.2.1 神經(jīng)網(wǎng)絡(luò)應(yīng)用的研究與發(fā)展 5
1.2.2 神經(jīng)網(wǎng)絡(luò)硬件的研究與發(fā)展 6
1.2.3 新型神經(jīng)網(wǎng)絡(luò)模型的研究 6
1.3 論文研究內(nèi)容 6
2 基于BP神經(jīng)網(wǎng)絡(luò)整定的PID控制 7
2.1 BP神經(jīng)網(wǎng)絡(luò) 7
2.2 基于BP神經(jīng)網(wǎng)絡(luò)整定的PID控制 10
2.2.1 基于BP神經(jīng)網(wǎng)絡(luò)整定控制系統(tǒng)結(jié)構(gòu) 10
2.2.2 算法 10
2.2.3 仿真程序和分析 11
3 基于RBF神經(jīng)網(wǎng)絡(luò)辨識(shí)的PID控制 14
3.1RBF神經(jīng)網(wǎng)絡(luò) 14
3.1 基于RBF神經(jīng)網(wǎng)絡(luò)辨識(shí)的PID控制 15
3.1.1 RBF神經(jīng)網(wǎng)絡(luò)結(jié)構(gòu) 15
3.1.2 RBF網(wǎng)絡(luò)PID辨識(shí)原理 15
3.1.3 仿真程序及分析 17
4 基于RBF神經(jīng)網(wǎng)絡(luò)辨識(shí)的單神經(jīng)元PID模型參考自適應(yīng)控制 19
4.1 神經(jīng)網(wǎng)絡(luò)模型參考自適應(yīng)控制原理 19
4.2 仿真程序及分析 20
5 基于RBF神經(jīng)網(wǎng)絡(luò)辨識(shí)的BP-PID控制 23
5.1 基于RBF神經(jīng)網(wǎng)絡(luò)辨識(shí)的BP-PID控制 23
5.2 仿真程序及分析 23
結(jié) 論 26
參 考 文 獻(xiàn) 27
附錄A 仿真程序 29
致 謝 45
隨著現(xiàn)代工業(yè)向大型化、集成化方向發(fā)展,生產(chǎn)過程日趨復(fù)雜,過程嚴(yán)重非線性、時(shí)變性、不確定性及變量間的強(qiáng)耦合使許多系統(tǒng)缺乏精確的數(shù)學(xué)描述,難以用傳統(tǒng)的理論方法分析和控制,因此有必要研究新的智能控制策略。我國目前已有研究所綜合應(yīng)用自動(dòng)控制、計(jì)算機(jī)應(yīng)用、人工智能以及化學(xué)工程學(xué)科的理論和技術(shù),對基于神經(jīng)網(wǎng)絡(luò)的軟測量和智能控制技術(shù)及控制軟件進(jìn)行了深入研究。神經(jīng)網(wǎng)絡(luò)控制的研究,可以更好的解決復(fù)雜的非線性、不確定、不確知系統(tǒng)在不確定、不確知環(huán)境中的控制問題。神經(jīng)網(wǎng)絡(luò)控制不需建立精確的數(shù)學(xué)模型,能夠自動(dòng)辨識(shí)被控過程參數(shù)、自動(dòng)整定控制參數(shù)、適應(yīng)被控過程參數(shù)的變化,是解決傳統(tǒng)PID控制器參數(shù)整定難、不能實(shí)時(shí)調(diào)整參數(shù)和魯棒性不強(qiáng)的有效措施。
本文在分析了神經(jīng)網(wǎng)絡(luò)控制系統(tǒng)之后,主要進(jìn)行了兩個(gè)方面的研究。其一:對基于BP神經(jīng)網(wǎng)絡(luò)整定的PID控制的研究;其二:是對基于RBF神經(jīng)網(wǎng)絡(luò)辨識(shí)整定的PID控制研究。編寫了基于BP和RBF網(wǎng)絡(luò)的仿真程序,仿真結(jié)果表明,應(yīng)用神經(jīng)網(wǎng)絡(luò)對常規(guī)PID控制器進(jìn)行改進(jìn)后提高了系統(tǒng)的魯棒性和動(dòng)態(tài)特性,有效的改善了系統(tǒng)的控制結(jié)果,達(dá)到了預(yù)期的目的。
Abstract
With the development into large-scale and integrationg of modern industry,the product process tends to become complex. Many systems are short of accurate mathematics description, because the process is nonlinear, time varying, uncertain and the srong coupling of variable. Then it is difficult o analyze and conrol with traditional method,so we need research new intelligent control strategy. At present, the Institute of Integrated application of automatic control, computer applications, artificial intelligence and the theory of chemical engineering and technology, it is have a depth research in Neural network,which based soft sensor and intelligent control technology and control software . In the study of neural network control,you can solve the control problem better,in the complex non-linear, uncertain,unknown system with uncertainty and unknown environment. Neural network control can automatically identify controlled process parameters, automatic tuning control parameters, to adapt to changes in controlled process parameters, but do not need to establish an accurate mathematical model. It can effectively solve the traditional PID controller parameter tuning difficultly, not adjust the parameters in real time and not stronger in robustness issues
After analyzing the control system of neural network, this paper mainly dose research in two aspecs.Firstly, Research based on the BP neural network setting PID control;Secondly, based on RBF neural network to identify setting PID control study. This paper gives the overall research plan for this system. Also given the simulation program and analysis of BP network and RBF network.
目 錄
摘 要 1
Abstract 2
引 言 1
1 緒論 2
1.1 智能控制的發(fā)展與展望 2
1.1.1 智能控制的興起 2
1.1.2 傳統(tǒng)控制和智能控制 3
1.1.3 智能控制的展望 4
1.2 神經(jīng)網(wǎng)絡(luò)的發(fā)展與展望 5
1.2.1 神經(jīng)網(wǎng)絡(luò)應(yīng)用的研究與發(fā)展 5
1.2.2 神經(jīng)網(wǎng)絡(luò)硬件的研究與發(fā)展 6
1.2.3 新型神經(jīng)網(wǎng)絡(luò)模型的研究 6
1.3 論文研究內(nèi)容 6
2 基于BP神經(jīng)網(wǎng)絡(luò)整定的PID控制 7
2.1 BP神經(jīng)網(wǎng)絡(luò) 7
2.2 基于BP神經(jīng)網(wǎng)絡(luò)整定的PID控制 10
2.2.1 基于BP神經(jīng)網(wǎng)絡(luò)整定控制系統(tǒng)結(jié)構(gòu) 10
2.2.2 算法 10
2.2.3 仿真程序和分析 11
3 基于RBF神經(jīng)網(wǎng)絡(luò)辨識(shí)的PID控制 14
3.1RBF神經(jīng)網(wǎng)絡(luò) 14
3.1 基于RBF神經(jīng)網(wǎng)絡(luò)辨識(shí)的PID控制 15
3.1.1 RBF神經(jīng)網(wǎng)絡(luò)結(jié)構(gòu) 15
3.1.2 RBF網(wǎng)絡(luò)PID辨識(shí)原理 15
3.1.3 仿真程序及分析 17
4 基于RBF神經(jīng)網(wǎng)絡(luò)辨識(shí)的單神經(jīng)元PID模型參考自適應(yīng)控制 19
4.1 神經(jīng)網(wǎng)絡(luò)模型參考自適應(yīng)控制原理 19
4.2 仿真程序及分析 20
5 基于RBF神經(jīng)網(wǎng)絡(luò)辨識(shí)的BP-PID控制 23
5.1 基于RBF神經(jīng)網(wǎng)絡(luò)辨識(shí)的BP-PID控制 23
5.2 仿真程序及分析 23
結(jié) 論 26
參 考 文 獻(xiàn) 27
附錄A 仿真程序 29
致 謝 45