论文部分内容阅读
将计算机发生的“扰动-响应”曲线的基值与正态分布伪随机数迭加,可以获得人工模拟的“扰动-响应”实验曲线。本文分别运用矩量法、加权矩量法、传递函数法、傅立叶分析法及时间域最小二乘拟合法对模拟实验曲线加以处理与参数估值,并采用 Monte-Carlo 法对这五种数据处理方法作出比较与评价。模拟实验计算结果表明,傅立叶分析法和时间域最小二乘拟合法精度最高,是值得推荐的数据处理方法。
By adding the base value of the “disturbance-response” curve of the computer to the pseudo-random number with normal distribution, the experimentally simulated “disturbance-response” experimental curve can be obtained. In this paper, we use the method of moment, weighted moment method, transfer function method, Fourier analysis and time domain least square fitting method to deal with the simulation experimental curve and parameter estimation, and Monte-Carlo method for these five data processing Method to make a comparison and evaluation. The simulation results show that Fourier analysis and time-domain least-squares fitting method have the highest accuracy and are the recommended data processing methods.