【摘要】 为解决电能结算中存在的电量数据缺失问题，准确恢复缺失电量数据，本文将压缩感知(CS)技术引入缺失电量数据恢复中，提出了兼具普适性及自适应性的稀疏字典构建方法，可适用于不同类型的用户用电数据特征的提取，并且克服了用户用电数据特征变化对缺失数据恢复准确度的影响；基于自适应稀疏字典和缺失电量数据，提出并构建了快速数据恢复算法。最后，通过某供电局的实测电量数据对算法进行测试。结果表明连续缺失1～3个数据点时，恢复准确度在90%以上的占比高于90%,能够高准确度地实现缺失电量数据的恢复。本文为缺失电量数据的恢复提供了新方法，在电能结算方面具有重要意义。更多还原

【Abstract】 In order to solve the problem of missing electricity data in electricity settlement and accurately recover the missing electricity data, compressed sensing technology is introduced into the recovery of missing electricity data. A sparse dictionary construction method that combines universality and adaptability is proposed, which can be applied to the extraction of electricity data features for different types of users, and which overcomes the impact of changes in electricity data features on the accuracy of missing data recovery; A fast data recovery algorithm was proposed and constructed based on adaptive sparse dictionary and missing electricity data. Finally, the algorithm was tested using the measured electricity consumption data from a certain power supply bureau. The results show that when 1～3 consecutive data points are missing, the proportion of recovery accuracy above 90% is higher than 90%, which can achieve high accuracy in recovering missing electricity data. This article provides a new method for recovering missing electricity data, which is of great significance to electricity settlement.更多还原