Design and test of real-time monitoring system for maize entrainment loss based on piezoelectric signal classification

Abstract

This study describes a novel entrainment loss monitoring system. The features of the collision signals of different maize materials were extracted in the time and frequency domains, and a kernel loss identification model was constructed using machine learning algorithms. The results show that the random forest model performed optimally, with classification accuracies of 96.83% and 94.85% on the training and test sets, respectively. An entrainment loss monitoring system was developed, and the optimal sensor location was determined. The accuracy of kernel classification was 91.09% and that of cob classification was 86.69% for the kernel/cob mixture. Field validation tests showed that the proposed entrainment loss monitoring system outperformed the traditional monitoring system. The method enables the precise monitoring of entrainment losses during maize harvesting, enabling intelligent maize production.