ABSTRACT

Spectra-compatible artificial ground motions are used extensively in the time history analysis of nuclear power plants. Owing to reasons such as the dense controlling frequency points and stringent requirements for the number of small response points, it is difficult for the conventional matching methods to generate artificial ground motions that are highly compatible with multi-damping design spectra. In this paper, to resolve the problems of high precision and robustness, an improved hybrid method for constructing a multi-parameter adjustment curve in the time domain is proposed. Different from artificial intelligence methods, the improved hybrid method is a deterministic iterative method that combines the advantages of the simulated annealing algorithm (SAA) and the time domain adjustment method. The SAA is used to determine the optimal weights of the corrective time histories of all the damping ratios at a specific frequency, which controls the influences of the corrective time histories on the response spectra. Subsequently, based on the optimal weights, the artificial ground motion is adjusted in the time domain to reduce the fitting error of all the damping ratios at a specific frequency. Moreover, the multi-damping design spectra matching problem of frequencies and damping ratios is simplified to a one-dimensional problem of frequencies using the SAA. Numerical examples are presented to demonstrate the versatility of the proposed improved hybrid method.

摘要

光谱兼容的人工地面运动广泛用于核电站的时间历史分析。由于控制频点密集、小响应点数量要求严格等原因，传统匹配方法难以生成与多阻尼设计谱高度兼容的人工地震动。本文针对高精度和鲁棒性问题，提出了一种改进的混合时域多参数调整曲线构建方法。与人工智能方法不同，改进的混合方法是一种结合了模拟退火算法（SAA）和时域调整方法优点的确定性迭代方法。SAA用于确定特定频率下所有阻尼比的校正时程的最佳权重，从而控制校正时程对响应谱的影响。随后，基于最优权重，在时域上对人工地震动进行调整，以减小特定频率下所有阻尼比的拟合误差。此外，使用SAA将频率和阻尼比的多阻尼设计谱匹配问题简化为频率的一维问题。给出了数值例子来证明所提出的改进的混合方法的通用性。基于最优权值，在时域对人工地震动进行调整，减小特定频率下所有阻尼比的拟合误差。此外，使用SAA将频率和阻尼比的多阻尼设计谱匹配问题简化为频率的一维问题。给出了数值例子来证明所提出的改进的混合方法的通用性。基于最优权值，在时域对人工地震动进行调整，减小特定频率下所有阻尼比的拟合误差。此外，使用SAA将频率和阻尼比的多阻尼设计谱匹配问题简化为频率的一维问题。给出了数值例子来证明所提出的改进的混合方法的通用性。