论文部分内容阅读
针对粒子群算法易出现早熟,搜索精度低的问题,从惯性权重的确定和算法搜索精度两个方面进行了改进。其中惯性权重由随迭代次数非线性递减函数和一随机扰动项确定,利用这个扰动项的突变性来跳出极小值区域,同时为增加粒子的多样性,提高算法搜索精度,引入了变尺度混沌搜索,并将该方法和标准粒子群算法分别与小波去噪结合,预测地基累计沉降量并做了对比,实验表明本文方法具有良好的全局和局部搜索能力,预测精度高。
For PSO prone to premature, the search accuracy is low, the improvement from two aspects of inertia weight determination and algorithm search accuracy is improved. The inertia weight is determined by the nonlinear decreasing function with iteration number and a random perturbation term. The perturbation term mutation is used to jump out the minimum region. At the same time, in order to increase the diversity of particles and improve the searching accuracy of the algorithm, The method and standard particle swarm optimization are respectively combined with wavelet denoising to predict the cumulative settlement of foundation. The experimental results show that this method has good global and local search ability and high prediction accuracy.