Learning Trajectories for Real- Time Optimal Control of Quadrotors
Learning Trajectories for Real- Time Optimal Control of Quadrotors
复制标题
DOI：
10.1109/iros.2018.8593536
复制
发表时间：
2018-10
期刊：
影响因子：
--
通讯作者：
Gao Tang;Weidong Sun;Kris K. Hauser
中科院分区：
文献类型：
--
作者：
Gao Tang;Weidong Sun;Kris K. Hauser
非线性最优控制问题因其非凸性而难以有效求解。提出了一种将机器学习预测最优轨迹与二次优化求精相结合的轨迹优化方法。首先，离线计算最优轨迹库并用于训练神经网络。在线，神经网络预测新的初始状态和代价函数的轨迹，并通过稀疏二次规划求解器进一步优化该预测。我们将这种方法应用于室内四旋翼的飞行到目标的运动问题。实验证明，该技术在几毫秒内就能计算出接近最佳的轨迹，并产生比现有方法更准确的灵活运动。
Nonlinear optimal control problems are challenging to solve efficiently due to non-convexity. This paper introduces a trajectory optimization approach that achieves realtime performance by combining machine learning to predict optimal trajectories with refinement by quadratic optimization. First, a library of optimal trajectories is calculated offline and used to train a neural network. Online, the neural network predicts a trajectory for a novel initial state and cost function, and this prediction is further optimized by a sparse quadratic programming solver. We apply this approach to a fly-to-target movement problem for an indoor quadrotor. Experiments demonstrate that the technique calculates near-optimal trajectories in a few milliseconds, and generates agile movement that can be tracked more accurately than existing methods.