多旋翼飞行器的多传感器信息融合算法研究

发布时间：2018-01-01 16:06

本文关键词：多旋翼飞行器的多传感器信息融合算法研究　出处：《中北大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着微小型处理器、传感器、电机以及姿态解算方法和控制算法的发展,多旋翼飞行器的研究和开发受到广大科研人员的关注。如何得到相对准确的飞行器的飞行数据,从而更好的控制飞行器的飞行是其核心内容之一,而这也直接影响着多旋翼飞行器的系统性能及推广应用。论文分析和研究了多旋翼飞行器飞行状态的测量所需要的传感器即陀螺仪、加速度计、磁力计、GPS、气压计和激光测距仪的原理及其相互间的关系,从而根据其可能产生的误差进行了相应的误差建模。依据传感器输出特性,对多旋翼飞行器的姿态角、高度、位置和速度信息进行了解算。针对单一传感器不能准确给出飞行器当前飞行状态的数据,对多个相关传感器的数学解算建立了信息融合算法模型。建立了基于同源传感器的集中式经典卡尔曼滤波模型;对本文的多传感器进行了分散式的联邦卡尔曼滤波过程建模,分别设计了其姿态、高度、位置信息的子滤波器以及主滤波器;通过对比分析提出了改进的联邦卡尔曼滤波算法,构建了改进的联邦卡尔曼滤波框架并对其关键性作用的主滤波进行了改进建模,实现了低成本传感器的测量精度的提高。对本文研究的集中式经典卡尔曼滤波、分散式联邦卡尔曼滤波以及提出的改进联邦卡尔曼滤波效果分别进行误差仿真试验对比,对滤波效果进行分析讨论,通过仿真结果表明改进的联邦卡尔曼滤波算法有效提高了飞行器飞行状态测量精度,验证了改进算法的有效性,更好地提高了飞行器的飞行稳定性和可靠性。
[Abstract]:With the development of micro processor, sensor, motor and attitude calculation method and control algorithm, the research and development of multi rotor aircraft has drawn much attention. How to get a relatively accurate aircraft flight data, so as to better control of aircraft flight is one of its core content, which directly affects the performance of the system and the application of multi rotor aircraft. The analysis of sensor and multi rotor aircraft flight measurements are needed: gyroscope, accelerometer, magnetometer, GPS principle, barometer and laser rangefinder and the relationship between them, and the corresponding error modeling according to the error of the possible. According to the output characteristics of the sensor, the height of multi rotor aircraft attitude angle, position and velocity are calculated. For a single sensor can give flying Is the flight data of multiple sensors related to the mathematical model established information fusion algorithm. A centralized classic Calman filter model of homologous sensor based on multi sensor; this paper studied the dispersion modeling process of the federal Calman filter, respectively design the attitude, height, position information filter and the main filter; through the comparative analysis of proposed federal Calman filter to improve the construction of the federal Calman filtering framework improved and the main filter for its key role to improve the modeling accuracy and low cost, realize the measurement of sensor is improved. In this paper the centralized classic Calman filter, improve the filtering effect of the federal Calman decentralized federal Calman filter and proposed are compared to the error simulation test, the filtering effect is analyzed. The simulation results show that the improved federated Calman filter algorithm improves the measurement accuracy of the flight state effectively, verifies the effectiveness of the improved algorithm, and improves the flight stability and reliability of the aircraft.

【学位授予单位】：中北大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：V249;TP212

【参考文献】