Design of a Distance Measuring System for Sugarcane Collection and Transfer Vehicle Based on Mutation Numerical Feature Extraction and Processing

Abstract

Highlights Maintaining the appropriate distance between two vehicles can effectively reduce the loss of sugarcane harvesting. The data processing method of numerical mutation feature extraction combined with sliding mean filtering can effectively filter out the interference from the sugarcane field environment on the distance measurement. The distance measurement system has been successfully constructed and applied to practical work, and the average relative error of distance measurement is 0.05%. Abstract. To improve the automation of the whole stalk sugarcane combine harvester and reduce the work intensity of the driver of the sugarcane collection and transfer vehicle, a sugarcane field collection and transfer vehicle that follows a ranging system was independently designed in this article. Then, a data processing method that extracts the numerical mutation features and combines the sliding mean filtering of the fluctuation characteristics of the ranging data in rugged and hilly land was proposed. First, frequency modulated continuous wave (FMCW) radar was used to measure the distance between two vehicles. The effects of the travel speed and shaking angle on the ranging accuracy were analyzed according to the working environment. The results showed that they have significant effects on the ranging results. Second, the abnormal data were identified and replaced using numerical mutation features, and then sliding mean filtering was applied to the replaced overall data. Finally, the system was implemented in an experiment in the field, and the results showed that the ranging accuracy of the system was significantly improved. The average relative error was reduced from 0.12% to 0.05%, and the ranging stability was good. Keywords: FMCW radar, Numerical feature extraction, Ranging system, Sliding mean filtering, Sugarcane collection and transportation.