Detection of production discontinuities in the extrusion process based on current signature by means of a discrete wavelet transform

This paper presents a new method for identifying production discontinuities during soybean grain extrusion using current signature and discrete wavelet analysis. Conducted experimental studies using a single-screw extruder showed that the observation of frequency changes in the value of the current spectrum using discrete wavelet analysis can provide valuable diagnostic information that can help prevent production stoppages and possible malfunctions in the functioning of extruder working elements, for instance during the process of lapping of interfacing elements, which can help improve the efficiency and productivity of the machine. Studies have shown that for optimal control of the extrusion process, a time interval of about 5 s is sufficient for the controller to take appropriate decision-making functions regarding the selection of soybean dosing parameters. The significance of the conducted research has practical implications and can be utilized by scientists involved in the extrusion process for efficient monitoring.