Probably Approximately Correct Nonlinear Model Predictive Control (PAC-NMPC)
Probably Approximately Correct Nonlinear Model Predictive Control (PAC-NMPC)
复制标题
DOI：
10.1109/lra.2023.3315209
复制
发表时间：
2022-10
影响因子：
5.2
通讯作者：
A. Polevoy;Marin Kobilarov;Joseph L. Moore
中科院分区：
文献类型：
--
作者：
A. Polevoy;Marin Kobilarov;Joseph L. Moore
随机非线性模型预测控制(SNMPC)方法通常对系统动力学做出限制性假设，并依赖于近似来表征潜在不确定性分布的演化。因此，它们往往无法捕获更复杂的分布(例如，非高斯或多峰)，并且无法提供准确的性能保证。在这封信中，我们提出了一种基于抽样的SNMPC方法，该方法利用最近导出的样本复杂性界来证明反馈策略的性能，而不需要对系统动态或潜在的不确定性分布做出假设。通过将我们的方法并行化，我们能够在仿真和硬件上使用1/10比例的拉力轿车和24英寸翼展固定翼无人机(UAV)来演示具有统计安全性保证的实时滚动时间SNMPC。
Approaches for stochastic nonlinear model predictive control (SNMPC) typically make restrictive assumptions about the system dynamics and rely on approximations to characterize the evolution of the underlying uncertainty distributions. For this reason, they are often unable to capture more complex distributions (e.g., non-Gaussian or multi-modal) and cannot provide accurate guarantees of performance. In this letter, we present a sampling-based SNMPC approach that leverages recently derived sample complexity bounds to certify the performance of a feedback policy without making assumptions about the system dynamics or underlying uncertainty distributions. By parallelizing our approach, we are able to demonstrate real-time receding-horizon SNMPC with statistical safety guarantees in simulation and on hardware using a 1/10th scale rally car and a 24-inch wingspan fixed-wing unmanned aerial vehicle (UAV).