This paper aims to overcome the poor effect and slow speed of traditional cleaning methods for clogged nozzles of the fused deposition modelling (FDM) 3D printers. For this purpose, a high-temperature melting experiment was carried out on thermoplastics, and a thermal analysis was performed by fitting the differential scanning calorimetry (DSC) curve was fitted by cubic spline interpolation. On this basis, an intelligent, rapid nozzle cleaning control system was developed considering the physical cleaning method. The system was applied to clean the clogged nozzle of an actual FDM printer. The results show that the system can complete the cleaning task rapidly without damaging the nozzle. The control system works stably, automatedly and conveniently, providing a guarantee for the effective and timely operation of FDM printers. The research findings shed new light on the application and promotion of FDM 3D printers.