A novel material identification method for different thermal conductivities materials based on the micro thermoelectric generator

Abstract

An artificial intelligence (AI) technical system is the foundation of the internet of things. While in the technical system, the material identification technology is the key point. Material identification technology can be applied to AI robots and smart skin. However, most of the traditional material identification technologies have limitations such as high cost and complicated operation, while the low-cost ones have the problem of low accuracy. To accurately, efficiently, and simply identify objects is a bottleneck that researchers urgently need to overcome. Based on the thermoelectric effect, a material identification theory is proposed and a prototype is fabricated in this paper. A series of experiments are carried out to test the performance of the prototype. According to the results, the prototype can precisely identify metal materials and nonmetal materials less than 3 s. There are many merits of the prototype, such as high identification accuracy, low cost, and low maintenance difficulty. Besides, considering complex surface of materials may decrease the identification accuracy, this paper proposes a kind of flexible thermal conductive material to improve identification accuracy. It is verified that the identification accuracy is improved with the flexible thermal conductive material.