论文部分内容阅读
在研究安全投资规律及影响因素的复杂性、不确定性的基础上,提出基于分形理论与模拟退火算法的安全投资预测新方法。该方法首先运用重标极差法分析了安全投资时间序列预测的可行性,然后利用模拟退火算法并结合对原始数据的一系列变换求解满足预测误差精度的平移量,得到安全投资时间序列的分形维数,最后根据分形统计模型可以预测出下一时间单位的安全投资。通过实例证明该方法的可行性和有效性,提高了决策过程的科学性,对企业制定未来安全投资总量具有借鉴作用。
On the basis of studying the complexity and uncertainty of safety investment law and influencing factors, a new method of safety investment forecasting based on fractal theory and simulated annealing algorithm is proposed. Firstly, the method of re-calibration is used to analyze the feasibility of time-series forecast of safety investment. Then, a simulated annealing algorithm is combined with a series of transformations of the original data to find the translation which satisfies the precision of prediction error. The fractal of safety investment time series is obtained. Dimension, and finally based on the fractal statistical model can predict the next time the unit of safety investment. The feasibility and effectiveness of this method are proved by examples, which improves the scientific decision-making process and provides a reference for enterprises in formulating the total amount of safety investment in the future.