http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
적응형 정책학습을 이용한 다양한 환경에서의 강인한 주행성능 개선
이교운(Kyowoon Lee),김성운(Seongun Kim),한지연(Jiyeon Han),최환일(Hwanil Choi),Daiki Matsunaga,최재식(Jaesik Choi) 대한전기학회 2021 대한전기학회 학술대회 논문집 Vol.2021 No.10
The optimal robot navigation policies vary according to the environment. This paper addresses a challenge to adaptively learn and find an optimal policy for the various environment settings for a mobile robot without extensively training for each environment. To solve this problem, we apply policy transfer with strategy optimization. We first train a family of policies for various environmental settings, and find the best policy with strategy optimization. We evaluate our method on the simulation with the various environment settings and show that the proposed algorithm is adaptive to the various environments.