基于混沌时间序列的资源管理大数据挖掘方法

Mining Method for Resource Management Big Data Based on Chaotic Time Series

目的为了实现资源管理大数据的精准挖掘,保证挖掘结果的真实性与有效性,提出一种基于混沌时间序列的资源管理大数据挖掘方法。方法首先,通过相空间重构一个等价的多维状态空间,利用相空间的混沌吸引子寻找预测点的相关函数关联,实现对被测目标区域位置的准确判断;其次,运用经验模态分解将资源管理大数据回波信号分解,按照回波信号间的关联性融合软阈值和粗糙惩罚手段,对资源管理大数据回波信号去噪,预防信号失真现象发生;最后,按照原始定位波形特征,去除回波信号波形伪峰值点,挑选适当的指数函数进行sinc函数波形拟合修正,保证资源管理大数据挖掘结果的准确性。结果仿真实验表明,此方法可以大幅提升资源管理大数据挖掘精度,且耗时较短。结论实现资源管理大数据挖掘,具备较高的应用价值。

Objective To realize the accurate mining of big data in resource management and ensure the authenticity and effectiveness of the mining results,a method for resource management big data mining based on chaotic time series was proposed.Methods Firstly,the equivalent multi-dimensional state space was reconstructed by phase space,and the correlation function of prediction point was found by using chaotic attractor in phase space to realize accurate judgment of the location of the target area;secondly,the echo signal of resource management big data was decomposed by empirical mode decomposition.According to the correlation between echo signals,soft threshold and rough punishment were combined to denoise the echo signals of big data in resource management to prevent signal distortion;finally,according to the original location waveform characteristics,the pseudo peak point of echo signal waveform was removed,and the appropriate exponential function was selected to fit and modify sinc function waveform,so as to ensure the accuracy of the mining results of big data in resource management.Results The simulation results showed that the method greatly improved the precision of big data mining in resource management with a short time.Conclusion It has a high application value to realize big data mining of resource management.