发表机构
University of California, Berkeley; Boston University; University of Maryland, College Park(加州大学伯克利分校; 波士顿大学; 马里兰大学帕克分校)
机构由 AI 辅助整理,请以论文原文为准。AI 中文总结
针对无目标时间序列数据的情况,提出为非线性系统综合参数以满足连续时间STL规范的方法,利用梯度优化与可达性验证在高维参数空间学习,在三个系统上验证了方法的有效性与可扩展性。
AI 中文摘要
信号时序逻辑(STL)越来越多地用于描述最优控制和学习方法的可解释目标与约束,尤其在无目标时间序列数据时。本文提出为非线性系统综合参数,使其对不确定初始条件稳健满足连续时间STL规范。为此,利用基于梯度的优化和基于集合的可达性验证,在高维参数空间高效学习并为优化参数提供可证明的满足保证。在三个至多18维参数的系统上证明了方法的有效性和可扩展性。
英文摘要
Signal Temporal Logic (STL) is increasingly used to describe interpretable objectives and constraints for optimal control and learning methods, especially when no target time series data is available. In this work, we propose to synthesize parameters for nonlinear systems that robustly satisfy continuous-time STL specifications for uncertain initial conditions. To this end, we use gradient-based optimization along with set-based reachability verification to efficiently learn in high-dimensional parameter spaces while providing provable satisfaction guarantees for the optimized parameters. We demonstrate the effectiveness and scalability of our method on three systems with up to 18 parameter dimensions.