计算机工程与设计
計算機工程與設計
계산궤공정여설계
COMPUTER ENGINEERING AND DESIGN
2014年
10期
3613-3616,3684
,共5页
神经网络%径向基函数%发酵温度%模糊控制%规则校正
神經網絡%徑嚮基函數%髮酵溫度%模糊控製%規則校正
신경망락%경향기함수%발효온도%모호공제%규칙교정
neural networks%radial basis function%fermentation temperature%fuzzy control%rule emendation
针对啤酒发酵过程中罐内温度控制问题，研究神经网络对模糊控制规则的优化方法，利用径向基函数神经网络对模糊控制规则进行优化，提高其自适应能力。以啤酒生产过程中主发酵阶段的数据作为输入样本，通过径向基函数神经网络进行学习训练，校正模糊控制规则，优化模糊控制器。优化前与优化后响应特性曲线的比较结果表明， RB F神经网络学习能力强，收敛速度快；模糊控制规则的完备性和一致性明显改善，控制器的响应速度快、超调量小、稳定性强、控制效果好。
針對啤酒髮酵過程中罐內溫度控製問題，研究神經網絡對模糊控製規則的優化方法，利用徑嚮基函數神經網絡對模糊控製規則進行優化，提高其自適應能力。以啤酒生產過程中主髮酵階段的數據作為輸入樣本，通過徑嚮基函數神經網絡進行學習訓練，校正模糊控製規則，優化模糊控製器。優化前與優化後響應特性麯線的比較結果錶明， RB F神經網絡學習能力彊，收斂速度快；模糊控製規則的完備性和一緻性明顯改善，控製器的響應速度快、超調量小、穩定性彊、控製效果好。
침대비주발효과정중관내온도공제문제，연구신경망락대모호공제규칙적우화방법，이용경향기함수신경망락대모호공제규칙진행우화，제고기자괄응능력。이비주생산과정중주발효계단적수거작위수입양본，통과경향기함수신경망락진행학습훈련，교정모호공제규칙，우화모호공제기。우화전여우화후향응특성곡선적비교결과표명， RB F신경망락학습능력강，수렴속도쾌；모호공제규칙적완비성화일치성명현개선，공제기적향응속도쾌、초조량소、은정성강、공제효과호。
Aiming at the problem in the process of beer fermentation tank temperature control ,the optimized method of fuzzy control rules based on the neural network was studied ,fuzzy control rules were optimized by using the radial basis function neu-ral network to improve its adaptive ability .In the beer production process ,data of the main fermentation phase were taken as the input sample and trained by the radial basis function neural network ,and fuzzy control rules were revised ,the fuzzy controller was revised .Comparing the response characteristic curves before the optimization and that after ,the results of the comparison show the RBF neural network has strong learning ability and fast convergence speed .The completeness and uniformity of fuzzy control rules are obviously improved ,the controller has fast response speed ,small overshoot ,strong stability and good control effects .
