Search | Korea Science

Yoon, Jae Ung;Lee, Ju Hong
- Smart Media Journal
- /
- v.11 no.9
- /
- pp.9-20
- /
- 2022
Deep reinforcement learning(RL) is an end-to-end data-driven control method that is widely used in the autonomous driving domain. However, conventional RL approaches have difficulties in applying it to autonomous driving tasks due to problems such as inefficiency, instability, and uncertainty. These issues play an important role in the autonomous driving domain. Although recent studies have attempted to solve these problems, they are computationally expensive and rely on special assumptions. In this paper, we propose a new algorithm MCDT that considers inefficiency, instability, and uncertainty by introducing a method called uncertainty sequence modeling to autonomous driving domain. The sequence modeling method, which views reinforcement learning as a decision making generation problem to obtain high rewards, avoids the disadvantages of exiting studies and guarantees efficiency, stability and also considers safety by integrating uncertainty estimation techniques. The proposed method was tested in the OpenAI Gym CarRacing environment, and the experimental results show that the MCDT algorithm provides efficient, stable and safe performance compared to the existing reinforcement learning method.
PDF KSCI

Jee, Jung-Eun;Lee, Sang-Ji;Shin, Yong-Tae
- Journal of Advanced Navigation Technology
- /
- v.15 no.5
- /
- pp.687-695
- /
- 2011
The defect caused by the software industry that is the source of knowledge-centric, information-centric and technology-centric affects an engine which operate a major role for operation of aircraft. Therefore, we should minimize the danger from the defect by strengthening the stability of aviation through the stability analysis of software. In this paper, we examine the laws and systems about the aircraft defects and software safety and propose the enhancement and the enactment of the law or measures to strengthen aviation safety. We should the existing law or system as items, such as the revision related to the safety analysis, standards of quality assurance including safety, application of quality assurance that you must attach the safety analysis report, assessment of detailed instructions of certification authorities. In addition, we should enact the new law and system as items such as the mandatory software evaluation and certification, continuous assessment based on the software life cycle, mandatory introduction of a standardized development methodology, strengthening of advanced workforce system. We can expect the improvement of software quality and an enhanced aviation safety by improving existing laws or systems and enacting new laws or systems.
https://doi.org/10.12673/jant.2011.15.5.687 인용 PDF KSCI