CAJ | 학술논문

在汽轮机振动信号采样点数较少时，使用经典著谱估算法进行分析的效果差。针对经典算法存在的这一问题，文中将现代谱估计中的自回归模型算法引入到汽轮机振动信号谱分析中，对算法的具体原理及自回归模型阶次确定的信息论准则进行了详细讨论。结合工业现场的实测信号，对算法的实际性能进行了分析与仿真，并与经典算法性能进行了对比。仿真结果表明：自回归模型功率谱估计算法适合于汽轮机振动信号的频谱分析，在采样点数较少的情况下，其性能优于经典算法。同时，信息论准则为算法应用中最优阶次的确定提供了有效的理论依据。
재기륜궤진동신호채양점수교소시，사용경전저보고산법진행분석적효과차。침대경전산법존재적저일문제，문중장현대보고계중적자회귀모형산법인입도기륜궤진동신호보분석중，대산법적구체원리급자회귀모형계차학정적신식론준칙진행료상세토론。결합공업현장적실측신호，대산법적실제성능진행료분석여방진，병여경전산법성능진행료대비。방진결과표명：자회귀모형공솔보고계산법괄합우기륜궤진동신호적빈보분석，재채양점수교소적정황하，기성능우우경전산법。동시，신식론준칙위산법응용중최우계차적학정제공료유효적이론의거。
The analysis effect of classical spectrum estimation algorithms is worse when the sample length of turbine’ s vibration signal is relatively short . Aiming at the problem existing in classical methods ,this paper introduces the AR model algorithm in the field of modern power spectrum estimation into the spectrum analysis of turbine’s vibration signal and discusses the principle of the method and the Akaike’s Information criteria (AIC ) used in the determination of the AR model’ s order in detail . Combining with the turbine’s vibration signals collected in industrial site ,the actual performance of the AR model method is simulated and analyzed .And the performance of the method is compared with the performance of classical methods .Theoretical analysis and simulation result shows that AR model power spectrum estimation algorithm is suitable for the spectrum analysis of turbine ’s vibration signal and the performance is better than classical methods when the data length is short .At the same time ,the AIC is proved to be an effective theoretical basis for the determination of the optimal order .