Abstract: Datamissing phenomena may happen in engineering testing due to the mismatch of component parts or breakdown of the testing system. In this paper, a parameter model is proposed to recover the missingdata. Through analyzing the correct data, the structure, the best order and the specific parameters of the model can be determined, and be further tested and optimized, so that the best model can be obtained. Finally, as an example, an AR model is applied to resume the missing data in the testing of an actual flight. The results show that the method is feasible and effective.

Key words: vibration and wave, missingdata, parameter model, ARMA model, AR model, data recovery

摘要： 工程实验测试中常会因测试系统部件选取不合适或发生故障等原因而造成数据缺失现象。研究利用参数模型方法对其进行恢复。通过对被测信号中正确数据的分析，确定出模型结构、阶次、具体参数，并进行检验和优化，得出最佳模型。针对实际的空测数据进行了仿真和应用，证明该方法是有效的。

关键词: 振动与波, 缺失数据, 参数模型, ARMA模型, AR模型, 数据恢复