FFT与二乘复合算法在曲线趋势预测领域内的研究

The Forecast Research on the Curve’S Trend Based on the Composite Algorithm of the FFT Means and the Least Squares Fitting

快速傅里叶变换（FFT）属于数字信号处理中最基础的运算之一,广泛应用于通讯、医学电子学、雷达或无线电天文学等领域。由于单纯的FFT算法不适用于此类部件振动曲线的预测,为提高预测精度,此文主要探讨了使用FFT与二乘拟合的复合算法来对该类曲线趋势进行预测的研究情况,仿真结果表明短时预测效果明显,从而验证了此方法的可行性和有效性。

FFT is a kind of most basic operation in digital signal processing and it has been widely used in area of communica-tions, medical electronics, radar or radio astronomy and so on. Because the simple FFT algorithm is not suitable for the predic-tion of the vibration curve of this kind of component, so this article mainly says the research phenomenon of using the compos-ite algorithm of FFT means and the least squares fitting to increase the forecasting precision of the vibration curve’s trend of gun. The simulation results shows that the effect of the short-time forecast is obvious and the feasibility and effectiveness of this solution are verified.