Reinforcement learning (RL) is widely applied to various engineering fields. Especially, RL has shown successful performance for control problems, such as vehicles, robotics, and active structural control system. However, little research on application of RL to optimal structural design has conducted to date. In this study, the possibility of application of RL to structural design of reinforced concrete (RC) beam was investigated. The example of RC beam structural design problem introduced in previous study was used for comparative study. Deep q-network (DQN) is a famous RL algorithm presenting good performance in the discrete action space and thus it was used in this study. The action of DQN agent is required to represent design variables of RC beam. However, the number of design variables of RC beam is too many to represent by the action of conventional DQN. To solve this problem, multi-agent DQN was used in this study. For more effective reinforcement learning process, DDQN (Double Q-Learning) that is an advanced version of a conventional DQN was employed. The multi-agent of DDQN was trained for optimal structural design of RC beam to satisfy American Concrete Institute (318) without any hand-labeled dataset. Five agents of DDQN provides actions for beam with, beam depth, main rebar size, number of main rebar, and shear stirrup size, respectively. Five agents of DDQN were trained for 10,000 episodes and the performance of the multi-agent of DDQN was evaluated with 100 test design cases. This study shows that the multi-agent DDQN algorithm can provide successfully structural design results of RC beam.

Microbial proteases are more economical than plant- and animal-derived proteases due to their ease of production and high activity. This study aimed to optimize the production of proteases from fermentative food-derived microorganisms. Five strains with proteolytic activity among 50 Bacillus sp. were first screened. Two strains with high protease activity were identified: Bacillus amyloliquefaciens SRCM 102139 and Bacillus subtilis SRCM 104999. SRCM 102139 strain and SRCM 104999 strain had the highest protease activity in 0.8% glucose and 0.3% yeast extract, and in 0.8% starch and 0.1% soy peptone, respectively. The production of protease for two strains was optimized by the Central Composite Design (CCD) under response surface methodology. The optimal conditions for protease production in SRCM 102139 were 0.5% and 0.347%, pH 6.0, for carbon (glucose) and nitrogen (yeast extract springer 0202) sources, respectively, with a predicted value of 0.929 U/mL. Additionally, the optimal conditions for protease production in SRCM 104999 were 0.5% and 0.5%, pH 6.7, for carbon (starch) and nitrogen (soy peptone HSP-349) sources, respectively, with a predicted value of 0.431 U/mL. The actual protease activities of SRCM 102139 and SRCM 104999 under the established conditions were 0.926 U/mL and 0.428 U/mL, respectively, closely matching the predicted values.

Dongchimi is one of the most representative types of watery kimchi in Korea. Although the scope of application of dongchimi has been expanded, research on the quality of dongchimi is insufficient compared to that of baechu kimchi. This study aimed to find the optimal manufacturing condition of high-acidity dongchimi and investigate the effects of potassium citrate (P) and sugar (S) on quality. Different dongchimi samples (Control, S, P) were prepared and stored at 15oC for 15 days. Among the various sugars, sucrose made the highest acidity of dongchimi. When potassium citrate was added as a pH buffer, the acidity was 2.2 times higher than the control group. In the PS group, where sucrose and potassium citrate were added, the acidity and organic acid content were 3-fold and 3.7-fold higher, respectively, than those in the control group. As a result of the analysis of the free sugar content, the metabolism of lactic acid bacteria was expected to be relatively more activated in the PS group. Therefore, this study could be utilized in the dongchimi manufacturing industry or as primary data for other dongchimi research.