Investigation of the Existence of Thermal Insulations in Wall Systems of Building Envelopes Using Uwb Technique

Abstract

Abstract-Hybrid pattern recognition is used to predict the types of insulation materials used inside wall systems of building envelopes. The hybrid pattern recognition features vector is built using the characteristics of UWB signals. UWB signals can penetrate objects, resulting in scattered signals based on the object's dielectric properties. The object's dielectric properties and structure have a signature within the scattered signals. This paper demonstrates that proper hybrid pattern recognition can be used to experimentally detect the existence and the type of insulation material inside wall systems with a high success rate.