abstract
- © 2019 IEEE.There are new algorithms such as artificial intelligence (AI) methodologies that have achieved accurate representation of experimental systems. On the other hand, undergraduate freshmen students must understand AI methodologies since the industry has developed several products based on those and some academic problems also can be solved using AI. If those students do not learn how to model real systems using AI, they will be losing the opportunity of applying this powerful tool for solving several real problems in their professional life. Since the AI model can be a representation for forecasting the performance of the real model, this model can help the design process and provide information during its operation. This paper proposes an engineering project to teach artificial intelligence algorithms using real systems that are non-linear. Since permanent magnets are used in several applications, they can be attractive for modeling those when they are interacting between them; hence, this paper shows the interaction among them when they are deployed as an electrical power source. Moreover, this source could be classified as a renewable energy source. The basic generation of electrical energy is based on changing the magnetic field. Although the operation principle is basic, the electrical source has a non-liner description that is extremely complex so AI could be applied to create a model that represents those non-linear relationships in a precise manner. The main goal of this work is to describe an undergraduate project that can be used for teaching how to model a real system using AI algorithms. The main characteristics and properties of the permanent magnets are studied for the comprehension of how magnets can be implemented. It is also examined the viability for the construction of an electric motor using only permanent magnets, based on the analysis of different designs and materials and finally an AI model is created.