灰色自记忆模型的高层建筑物沉降分析

针对GM(1,1)模型对非线性数据的沉降趋势及其波动特征无法进行准确地预测,而灰色残差模型和灰色马尔科夫模型又无法解决这个问题,提出了灰色自记忆预测模型。该模型利用了自记忆原理考虑过去和现在对未来的影响的记忆性特点,克服了GM(1,1)模型对初值比较敏感、预测精度低等局限性,提高了对波动性数据的预测能力。通过实例验证表明了灰色自记忆模型的可靠性和可行性。

Analysis of high-rise constructions settlement based on grey self-memory model

For theGM(1,1)modelcannot predict the subsidence trend and fluctuation characteristics of nonlinear data accurately,grey residual error model and grey-markov model cannot solve the problem,a forecast model named as grey self-memory combined mode was presented in this paper.The model utilizes the characteristic of self-memory model that considering the effect of past and now on the future,overcoming the limitation of the GM(1,1)such as be sensitive to initial value,low accuracy of prediction and improving the ability to predict volatility data.The example validations indicated that the forecast precision of the combined model was reliable and feasible.