Falsification of Learning-Based Controllers through Multi-Fidelity Bayesian Optimization
Falsification of Learning-Based Controllers through Multi-Fidelity Bayesian Optimization
复制标题
通过多保真贝叶斯优化伪造基于学习的控制器
DOI：
--
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Ali Baheri
中科院分区：
文献类型：
--
作者：
Zahra Shahrooei;Mykel J. Kochenderfer;Ali Baheri
基于仿真的证伪是一种实用的测试方法，可提高系统满足安全要求的信心。由于全保真模拟对计算要求很高，因此我们研究了具有不同保真度水平的模拟器的使用。作为第一步，我们表示的整体安全规格的环境参数和结构作为一个优化问题的安全规格。我们提出了一个多保真度伪造框架，使用贝叶斯优化，这是能够确定在哪个级别的保真度，我们应该进行安全评估，除了从环境中找到可能的情况下，导致系统失败。这种方法使我们能够以具有成本效益的方式在来自低保真度模拟器的廉价，不准确的信息和来自高保真模拟器的昂贵，准确的信息之间自动切换。我们在不同环境下的仿真实验表明，多保真度贝叶斯优化具有与单保真度贝叶斯优化相当的证伪性能，但成本低得多。
Simulation-based falsification is a practical testing method to increase confidence that the system will meet safety requirements. Because full-fidelity simulations can be computationally demanding, we investigate the use of simulators with different levels of fidelity. As a first step, we express the overall safety specification in terms of environment parameters and structure this safety specification as an optimization problem. We propose a multi-fidelity falsification framework using Bayesian optimization, which is able to determine at which level of fidelity we should conduct a safety evaluation in addition to finding possible instances from the environment that cause the system to fail. This method allows us to automatically switch between inexpensive, inaccurate information from a low-fidelity simulator and expensive, accurate information from a high-fidelity simulator in a cost-effective way. Our experiments on various environments in simulation demonstrate that multi-fidelity Bayesian optimization has falsification performance comparable to single-fidelity Bayesian optimization but with much lower cost.