Identification of baled materials through capacitive sensing and data driven modelling

Recycle and reuse of waste materials are important measures in achieving circular economy, reducing resource waste, and protecting environment. However, current recycling rate is low and a key issue causing low recycling rate is the uncertainty in the quality of baled materials. In this study, a new method based on a capacitive sensor and a data driven model is proposed for identifying baled materials. A novel capacitive sensor with satisfactory sensitivity and sensitivity distribution is designed for this purpose using finite element method. The transmitter and receiver units as well as advanced signal conditioning circuit are developed. To achieve automated identification of the baled materials based on the sensor outputs, the support vector machine (SVM) algorithm is used. To verify the proposed method, experiments were carried out to measure different baled materials. Experimental results suggest that the proposed method is able to successfully identify these baled materials with satisfactory accuracy.