arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2203.16633eess.SYcs.ROcs.SY
收藏

模型预测优化路径积分策略

Model Predictive Optimized Path Integral Strategies

Stanford University(斯坦福大学)

Dylan M. Asmar, Ransalu Senanayake, Shawn Manuel, Mykel J. Kochenderfer

摘要:

我们推广的模型预测路径积分控制(MPPI)的推导,允许一个单一的联合分布控制序列中的控制。这种改造允许将自适应重要性采样(AIS)算法实施到原始重要性采样步骤中,同时仍然保持MPPI的优点,例如与任意系统动态和代价函数一起工作。通过在每个控制步骤中集成AIS来优化建议分布的好处在模拟环境中得到了证明,包括控制轨道周围的多辆车。新算法比MPPI算法具有更高的采样效率,在更少的采样数下获得了更好的性能。这种性能差异随着动作空间的维度增加而增加。仿真结果表明,新算法可以作为一个随时算法,增加控制值在每次迭代依赖于一个大的样本集。

英文摘要:

We generalize the derivation of model predictive path integral control (MPPI) to allow for a single joint distribution across controls in the control sequence. This reformation allows for the implementation of adaptive importance sampling (AIS) algorithms into the original importance sampling step while still maintaining the benefits of MPPI such as working with arbitrary system dynamics and cost functions. The benefit of optimizing the proposal distribution by integrating AIS at each control step is demonstrated in simulated environments including controlling multiple cars around a track. The new algorithm is more sample efficient than MPPI, achieving better performance with fewer samples. This performance disparity grows as the dimension of the action space increases. Results from simulations suggest the new algorithm can be used as an anytime algorithm, increasing the value of control at each iteration versus relying on a large set of samples.

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。

↑