四旋翼或多旋翼飞行器环境适应能力强、机械结构简单、操控灵活、飞行稳定,应用广泛。其核心算法包含姿态融合算法和飞行控制算法。为估算飞行器最优姿态,提出一种融合闭环控制和时间序列预测法的加速度传感器数据和陀螺仪数据融合的算法。该算法采用时间序列预测法估算加速度传感器、陀螺仪误差,积分控制器修正陀螺仪漂移误差。通过理论分析、MATLAB仿真、四旋翼飞行器实验平台验证,该姿态融合算法对于估算飞行器姿态具有较快的收敛速度和较高的精度,其角度误差、角速度误差极小。 Quadrotor or multi-rotor aircraft environment adaptability, mechanical structure is simple, flexible control, flight stability, wide range of applications. The core algorithms include attitude fusion algorithm and flight control algorithm. In order to estimate the optimal attitude of the aircraft, an algorithm for the fusion of accelerometer data and gyroscope data based on closed-loop control and time series prediction is proposed. The algorithm estimates the acceleration sensor and gyroscope error with time series prediction and the integral controller corrects the gyroscope drift error. Through theoretical analysis, MATLAB simulation and validation of the experimental platform of quadrotor, the attitude fusion algorithm has faster convergence rate and higher accuracy for estimating attitude of aircraft, and has very little angle error and angular velocity error.