电力自动化设备
電力自動化設備
전력자동화설비
ELECTRIC POWER AUTOMATION EQUIPMENT
2010年
4期
10-15
,共6页
师彪%李郁侠%何常胜%于新花%闫旺%孟欣%李鹏
師彪%李鬱俠%何常勝%于新花%閆旺%孟訢%李鵬
사표%리욱협%하상성%우신화%염왕%맹흔%리붕
水轮机调速系统%自适应人工鱼群%BP神经网络%仿真%模型参数辨识
水輪機調速繫統%自適應人工魚群%BP神經網絡%倣真%模型參數辨識
수륜궤조속계통%자괄응인공어군%BP신경망락%방진%모형삼수변식
hydraulic turbine governing system%adaptive artificial fish school algorithm%BP neural network%simulation%model parameter identification
为建立与电网稳定计算有关的水轮机调速系统数学模型及模型参数测量辨识,提出一种基于自适应人工鱼群一神经网络技术并适用于水轮机调速系统控制的新技术,建立智能调速系统数学模型,使之符合实际调节及微机优化控制.分析了该模型组成部分的传递函数,提出采用自适应人工鱼群算法来弥补人工鱼群和神经网络算法的不足,阐述了自适应人工鱼群算法一神经网络优化器的算法.给出了自适应人工鱼群优化算法参数辨识算法设计和实现步骤.利用Matlab和自适应人工鱼群算法进行模型参数辨识,对一次调频和二次调节试验过程进行仿真并与实测对比.结果表明,仿真值与实测值相当接近,所研制的自适应人工鱼群一神经网络优化器,达到了优化PID调节器控制输出量的目标;所建立的调速系统数学模型真实地反映调速系统在机组并网工况下的调节特性,说明该方法原理正确,可用于优化控制.
為建立與電網穩定計算有關的水輪機調速繫統數學模型及模型參數測量辨識,提齣一種基于自適應人工魚群一神經網絡技術併適用于水輪機調速繫統控製的新技術,建立智能調速繫統數學模型,使之符閤實際調節及微機優化控製.分析瞭該模型組成部分的傳遞函數,提齣採用自適應人工魚群算法來瀰補人工魚群和神經網絡算法的不足,闡述瞭自適應人工魚群算法一神經網絡優化器的算法.給齣瞭自適應人工魚群優化算法參數辨識算法設計和實現步驟.利用Matlab和自適應人工魚群算法進行模型參數辨識,對一次調頻和二次調節試驗過程進行倣真併與實測對比.結果錶明,倣真值與實測值相噹接近,所研製的自適應人工魚群一神經網絡優化器,達到瞭優化PID調節器控製輸齣量的目標;所建立的調速繫統數學模型真實地反映調速繫統在機組併網工況下的調節特性,說明該方法原理正確,可用于優化控製.
위건립여전망은정계산유관적수륜궤조속계통수학모형급모형삼수측량변식,제출일충기우자괄응인공어군일신경망락기술병괄용우수륜궤조속계통공제적신기술,건립지능조속계통수학모형,사지부합실제조절급미궤우화공제.분석료해모형조성부분적전체함수,제출채용자괄응인공어군산법래미보인공어군화신경망락산법적불족,천술료자괄응인공어군산법일신경망락우화기적산법.급출료자괄응인공어군우화산법삼수변식산법설계화실현보취.이용Matlab화자괄응인공어군산법진행모형삼수변식,대일차조빈화이차조절시험과정진행방진병여실측대비.결과표명,방진치여실측치상당접근,소연제적자괄응인공어군일신경망락우화기,체도료우화PID조절기공제수출량적목표;소건립적조속계통수학모형진실지반영조속계통재궤조병망공황하적조절특성,설명해방법원리정학,가용우우화공제.
In order to establish the mathematical model of hydraulic turbine governing system for grid stability calculation and model parameter idemification,a novel and suitable technique based on adaptive artificial fish school algorithm-neural network is proposed.The mathematical model is established to adapt to the practical regulation as well as the microcomputer-based control.Its transfer funotions are analyzed and the adaptive artificial fish school algorithm is proposed to avoid the shortage of the artificial fish school algorithm and neural network algorithm.The adaptive artificial fish school algorithm-BP neural network is expounded and the design and implementation of parameter identification algorithm are described.The model parameters are identified with Matlab and the proposed algorithm.The primary frequency regulation and secondary regulation are simulated and tested,and the comparison of results shows that,the simulative values are very close to the measurements:the developed controller based on the adaptive artificial fish school algorithm.BP neural network reaches its control objective of optimizing the control output of the PID regulator;and the established model truly reflects the performance of the governing system under the operating mode of unit connected to grid,which indicates its principle is correct and suitable for optimal control.