資源描述:
《基于rbf神經(jīng)網(wǎng)絡(luò)整定的pid控制器設(shè)計及仿真設(shè)計》由會員上傳分享,免費在線閱讀,更多相關(guān)內(nèi)容在學(xué)術(shù)論文-天天文庫。
1、華北電力大學(xué)畢業(yè)設(shè)計(論文)系別專業(yè)班級學(xué)生姓名秦術(shù)員指導(dǎo)教師基于RBF神經(jīng)網(wǎng)絡(luò)整定的PID控制器設(shè)計及仿真題目年月II華北電力大學(xué)科技學(xué)院本科畢業(yè)設(shè)計(摘要)I華北電力大學(xué)科技學(xué)院本科畢業(yè)設(shè)計(摘要)I基于RBF神經(jīng)網(wǎng)絡(luò)整定的PID控制器設(shè)計及仿真摘要目前,因為PID控制具有簡單的控制結(jié)構(gòu),可通過調(diào)節(jié)比例積分和微分取得基本滿意的控制性能,在實際應(yīng)用中又較易于整定,所以廣泛應(yīng)用于過程控制和運動控制中,尤其在可建立精確模型的確定性控制系統(tǒng)中應(yīng)用比較多。然而隨著現(xiàn)代工業(yè)過程的日益復(fù)雜,對控制要求的逐步增高(如穩(wěn)定性、準(zhǔn)確性、快速性等),經(jīng)典控制理論面臨著嚴(yán)重的挑戰(zhàn)。對工業(yè)控制領(lǐng)域中非線性系統(tǒng),采
2、用傳統(tǒng)PID控制不能獲得滿意的控制效果。采用基于梯度下降算法優(yōu)化RBF神經(jīng)網(wǎng)絡(luò),它將神經(jīng)網(wǎng)絡(luò)和PID控制技術(shù)融為一體,既具有常規(guī)PID控制器結(jié)構(gòu)簡單、物理意義明確的優(yōu)點,同時又具有神經(jīng)網(wǎng)絡(luò)自學(xué)習(xí)、自適應(yīng)的功能。因此,本文通過對RBF神經(jīng)網(wǎng)絡(luò)的結(jié)構(gòu)和計算方法的學(xué)習(xí),設(shè)計一個基于RBF神經(jīng)網(wǎng)絡(luò)整定的PID控制器,構(gòu)建其模型,進(jìn)而編寫M語言程序。運用MATLAB軟件對所設(shè)計的RBF神經(jīng)網(wǎng)絡(luò)整定的PID控制算法進(jìn)行仿真研究。然后再進(jìn)一步通過仿真實驗數(shù)據(jù),研究本控制系統(tǒng)的穩(wěn)定性,魯棒性,抗干擾能力等。關(guān)鍵詞:PID;RBF神經(jīng)網(wǎng)絡(luò);參數(shù)整定ⅡSETTINGOFTHEPIDCONTROLLERBASED
3、ONRBFNEURALNETWORKDESIGNANDSIMULATIONAbstractAtpresent,becausethePIDcontrolhasasimplecontrolstructure,throughadjustingtheproportionalintegralanddifferentialgainbasicsatisfactorycontrolperformance,andisrelativelyeasytosettinginpracticalapplication,sowidelyusedinprocesscontrolandmotioncontrol,especia
4、llyintheaccuratemodelcanbebuiltmoredeterministiccontrolsystemapplication.Withtheincreasinglycomplexofthemodernindustrialprocess,however,increasedstepbysteptocontrolrequirements(e.g.,stability,accuracyandquickness,etc.),classicalcontroltheoryisfacedwithseverechallenges.Non-linearsystemsinindustrialc
5、ontrolfield,usingthetraditionalPIDcontrolcannotobtainsatisfactorycontroleffect.OptimizedRBFneuralnetworkbasedongradientdescentalgorithm,itwillbeintegratedneuralnetworkandPIDcontroltechnology,withaconventionalPIDcontrollerhassimplestructure,physicalmeaningisclearadvantages,atthesametimewithneuralnet
6、workself-learning,adaptivefunction.Therefore,thisarticlethroughtotheRBFneuralnetworkstructureandthecalculationmethodoflearning,todesignasettingofthePIDcontrollerbasedonRBFneuralnetwork,constructsitsmodel,andthenwriteMlanguageprogram.UsingtheMATLABsoftwaretodesigntheRBFneuralnetworksettingofPIDcontr
7、olalgorithmsimulationresearch.Dataandthenfurtherthroughsimulationexperiment,thecontrolsystemstability,robustness,anti-interferenceability,etc.Keywords:PID;RBFneuralnetwork;ParametersettingⅡ目錄摘要ⅠAbstractⅡ1緒論