Monitoring and Estimating Sugarcane Losses during Harvesting Using Sound Analysis

Mechanical harvesting of sugarcane is done in two ways: green and burnt, and usually burnt harvest has between 25-50% less losses. When harvesting sugarcane, the sound of sugarcane pieces hitting the wall of the primary extractor hood can clearly be heard. Accordingly, it was decided to use the audio system to determine the relationship between these sounds and the losses of the primary extractor. To record sounds in the basic extractor, two models of full-directional and one-way capacitive microphone (cardioid) and Cool Record Edit Deluxe and Audacity software were used. To detect the wavelength of the sounds caused by the collision of different parts of sugarcane with the hood cap and extractor blades by throwing a large volume of straw along with 25 cm pieces of sugarcane billets, a sound record was set. A camera was also installed there to record the video of what was happening under the extractor compartment. The results showed that the one-way capacitive microphone installed in the upper part of the primary extractor housing received clearer sounds. Analyzing the recorded sounds and comparing them with the images obtained from the camera under the primary extractor revealed that the audio loss detection system detects the losses in the primary extractor with an accuracy of about 75 to 80%. The loss rate at 1200 rpm was about 1.5 times higher than the loss rate at 1100 rpm.