基于频率评估系统的ETL优化与小区性能评估研究

Research of ETL processing optimization and cell performance estimation based on frequency evaluation system

针对频率评估系统中的ETL处理过程效率较低的问题,在分治算法和贪心算法思想的基础上,提出了一种优化算法——DC-Greedy算法。通过在评估系统中对小区数据进行处理后,基于主成分分析法,研究了单个小区综合性能的评估策略。ETL处理时间和小区综合性能评估值的仿真结果表明,对比一般情况下ETL的穷举处理过程,经过优化后ETL处理效率明显提高;对小区评估策略的评价结果表明该评估方法可以完整地反映小区的综合性能,评估结果图能较直观地呈现小区综合性能水平,达到了预期研究目标。

In frequency evaluation system, due to extract transform load （ETL） processing is inefficient, an optimization algo- rithm DC-Greedy algorithm, which is based on divide and conquer algorithm and greedy algorithm, is proposed to solve the problem in this paper. After processing of cell data in frequency evaluation system, the strategy of evaluating single cell compre- hensive performance is studied that based on principal component analysis （PCA）. The simulation of ETL＇ s processing time and the evaluation of cell＇ s overall performance states： the optimized processing efficiency of ETL is clearly improved, compared to previous ETL processing with exhaustion method. The result of evaluation strategy shows that the evaluation method can reflect completely a cell＇ s overall performance. The evaluation graph of cell performance can present more intuitive cell comprehensive level and shows the desired goal is achieved.