arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2003.01190cs.ROcs.LG
收藏

动力学系统辨识信息轨迹的对抗生成

Adversarial Generation of Informative Trajectories for Dynamics System Identification

University of Edinburgh(爱丁堡大学)

Marija Jegorova, Joshua Smith, Michael Mistry, Timothy Hospedales

摘要:

动态系统辨识方法通常严重依赖于进化和基于梯度的优化技术,以产生最佳的激励轨迹,用于确定机器人平台的物理参数。目前的优化技术往往产生单一的轨迹。这是昂贵的,并且对于较长的轨迹是棘手的,因此限制了它们用于系统辨识的功效。我们建议通过使用多个较短的循环轨迹来解决这个问题,这些轨迹可以并行生成,然后组合在一起以实现与较长轨迹相同的效果。至关重要的是,我们展示了如何通过使用基于生成对抗网络(GAN)的架构来提高数据集的生成速度和质量,从而进一步扩展这种方法,以生成有效和多样化的激励轨迹的大型数据库。据我们所知,这是第一个机器人工作,探索系统辨识与多个循环轨迹,并开发基于GAN的技术,用于标量产生激励轨迹,在控制参数和惯性参数空间都是不同的。我们表明,我们的方法大大加快了轨迹优化,同时提供更准确的系统辨识比传统的方法。

英文摘要:

Dynamic System Identification approaches usually heavily rely on the evolutionary and gradient-based optimisation techniques to produce optimal excitation trajectories for determining the physical parameters of robot platforms. Current optimisation techniques tend to generate single trajectories. This is expensive, and intractable for longer trajectories, thus limiting their efficacy for system identification. We propose to tackle this issue by using multiple shorter cyclic trajectories, which can be generated in parallel, and subsequently combined together to achieve the same effect as a longer trajectory. Crucially, we show how to scale this approach even further by increasing the generation speed and quality of the dataset through the use of generative adversarial network (GAN) based architectures to produce a large databases of valid and diverse excitation trajectories. To the best of our knowledge, this is the first robotics work to explore system identification with multiple cyclic trajectories and to develop GAN-based techniques for scaleably producing excitation trajectories that are diverse in both control parameter and inertial parameter spaces. We show that our approach dramatically accelerates trajectory optimisation, while simultaneously providing more accurate system identification than the conventional approach.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑