Towards Applying Fuzzy Systems in Intelligent Agent-based CPS: A Case Study


Karaduman B., TEZEL B. T., CHALLENGER M.

2021 6th International Conference on Computer Science and Engineering (UBMK), Ankara, Turkey, 15 September 2021, pp.735-740

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/ubmk52708.2021.9558970
  • City: Ankara
  • Country: Turkey
  • Page Numbers: pp.735-740
  • Dokuz Eylül University Affiliated: Yes

Abstract

Cyber-physical Systems (CPS) are complex systems that interact with the physical world where instant external changes exist. To tackle these systems, in the literature, software agents are considered as an option and their enhancement with fuzzy logic can also be employed. Agent-oriented approaches and particularly Belief-Desire-Intention (BDI) architecture is used in different application areas. In all these domains, autonomy and proactive behaviour as well as the human-like reasoning are important aspects to enhance the intelligence of a system. These features can also be applied in CPS where the systems are complex. In this study, it is aimed to enhance a CPS using fuzzy logic for intelligent agents so that intelligent decision making can be applied to eliminate inaccurate calculations in the CPS. To evaluate the proposed fuzzy-BDI approach, we used a CPU fan control system. First, the system is tested assuming that the temperature sensor is ideal without including any sampling noise. We applied traditional BDI and fuzzy-BDI separately. Then, we tested data sampling process considering ∓ 5 error rate. We concluded that fuzzy-BDI approach for both ideal and noisy data showed considerable improvement.