Abstract: Aiming at the problem that quick access recorder (QAR) data do not satisfy the data compatibility due to the systematical error of the measurement sensor, this paper transforms the parameter estimation problem of the systematical error into the function optimization problem on the solution space with butterfly optimization algorithm and the least squares criterion. In order to improve the performance of butterfly optimization algorithm, chaotic mapping is used to initialize the population and dynamically adjust the conversion probability. This method is applied to the correction and reconstruction of flight attitude angle and angular velocity of QAR data. The experimental results show that this method is simple and feasible, with high data reconstruction accuracy, and meets the requirements of data compatibility.

Key words: data reconstruction, quick access recorder(QAR) data, butterfly optimization algorithm, data compatibility; parameter estimation, least squares criterion