Reinforcement Learning Models for Embedded System Optimization in EV and Smart Grid Education

Reinforcement Learning (RL) is a robust technique that can help educators and students optimize challenging and complex decision-making processes, particularly in evolving technologies such as Electric Vehicles (EVs) and Smart Grid education, technologies that are crucially advancing energy efficiencies, sustainable transportation and intelligent power management systems. Current optimization methods for embedded systems do not have reasoning in uncertain situations, such as in dynamic environments, and often use static algorithms that depend on rules and procedures that are difficult to use in real-time (energy) management. Current models struggle with high computational complexity in their training phases and do not scale well in educational simulations. To accommodate this complexity with a process and facilitate these complexities, it presents a new framework, RLESE (Reinforcement Learning Embedded System Educator), and describe how it can integrate deep reinforcement learning algorithms and use them to enhance embedded systems simulations. RLESE allows the student educator to vary the operating conditions dynamically in different operational environments of the EV and Smart Grid systems, and supports the student's energy management and reasoning in the real-time context. A RLESE framework is a conceptual model that can be used in educational settings, and provides a simulated and experiential manner for students and researchers to tangibly investigate energy optimization solutions in EV and Smart Grids. Using continuous learning techniques and interactions from the environment in a framework for education provides an environment that can gradually grow in its performance capabilities over time without predefined rules for the system. The data collected shows that RLESE performed with greater energy efficiency, and quicker response time, with greater agility than typical teacher-centred energy learning strategies. RLESE can be deployed for educators and in embedded systems applications in energy systems.

Перевод пока недоступен