一种无人直升机飞行力学模型辨识方法研究

研究了一种无人直升机飞行力学模型辨识方法.将状态子空间辨识法和误差预报辨识法这两种不同机理的辨识方法相结合,用于无人直升机飞行力学模型的辨识.通过仿真计算,成功地辨识得到了悬停状态下算例无人直升机的高阶飞行力学模型.结果表明:提出的辨识方法具备状态子空间辨识法和误差预报辨识法各自的优点,不会出现经典辨识算法中寻优过程中出现的局部极小现象以及迭代带来的收敛性问题.

Study on a Method for Unmanned Helicopter Flight Dynamics Model Identification

A method is presented for unmanned helicopter flight dynamics model identification. In this method, subspace state space system identification method and prediction error method are combined to compensate the shortcomings of each single one. By simulation, the flight dynamics model of unmanned helicopter example is successfully identified. The results show that the method presented by this paper takes the advantage of two distinguished system identification methods and is always convergent and numerically stable.