基于趋势滤波的瞬态曲线协同分段线性化方法研究

An Investigation of Trend Filter Based Collaborative Piecewise Linearization Method for Transient Curve

核电厂瞬态曲线由数量巨大的数据点构成,无法直接用于疲劳分析计算。因此,合理地对瞬态曲线进行分段线性化处理是进行疲劳分析的前提条件。本文提出一种基于趋势滤波算法的核电厂瞬态协同分段线性化方法,并分别采用折线参考瞬态和真实瞬态曲线对该方法进行了验证。结果表明,本文提出的方法能够准确、合理地对含噪声的瞬态数据进行处理,准确实现瞬态曲线的分段线性化。

In nuclear power plants, both the calculated design transient and monitored real transient are constructed by massive data. Due to the high computational cost, the raw transient data can not be used for fatigue analysis directly. Therefore, how to piecewise linearize the raw data is a key issue for fatigue analysis. A trend filter based collaborative piecewise linearization method for transient curve is proposed in the present work. An assumed reference transient and a real monitored transient are used to verify the present method. The results show that the present method can linearize the real transient curve accurately and efficiently.