烟叶采收机仿垄行走控制系统设计与试验

Design and experiment of moving along ridge control system for tobacco picking machine

为实现烟叶采收机的无人驾驶,设计了一种烟叶采收机仿垄行走控制系统,该系统采用光电测距传感器检测左右两侧车轮到烟垄的距离。该文分析一阶滞后低通滤波的灵敏度、平滑度以及相位滞,同时修正滤波的控制参数,以期达到较好的拟合效果,从而精确剔除杂草、土块等无效的尖锐信号,最后根据优化处理后的数据调整动力系统的电机转速,实现烟叶采收机仿垄行走。一阶滞后低通滤波经动态修正后无论从灵敏度还是平滑度看,效果均较好,相位滞后也得到明显修正。田间试验表明,滤波后的数据标准差为3.8,能够得到较为准确的垄形信息;跑偏角度低于30°时,角度修正较好,成功率为95%,侧移修正成功率为89%;跑偏角度低于20°时,成功率为100%,侧移修正成功率为98.3%。该行走控制系统可准确控制烟叶采收机仿垄行走,为无人烟叶采收机行进控制方法的深入研究提供了参考。

For the realization of the tobacco leaf picking machine in unmanned, a ridge moving control system of tobacco picking machine is designed, this system adopts photoelectric ranging sensor detecting the distance of both sides of the wheels to the tobacco ridge. In this paper, the sensitivity, smoothness and phase delay of a first order lag low-pass filter are analyzed, filtering control parameters are corrected. This paper expects to achieve a better fitting effect, and eliminate weeds, blocks and other invalid sharp signals accurately, and finally, adjust the motor speed of power system according to the optimized data, realize the ridge moving of tobacco picking machine. The simulation results show that after dynamic correction, both in terms of sensitivity and smoothness, the effect of the first order lag low-pass filtering is very good, and the phase lag has been significantly modified. The field test shows that, after filtering the data standard deviation is 3.8, it can get accurate ridge information. When deviation angle less than 30°, angle correction is better, the success rate was95%, and lateral shift correction success rate is 89%, When deviation angle is less than 20°, the success rate was 95%, and lateral shift correction success rate is 98.3%. The moving control system can accurately control the tobacco picking machine moving along the ridge, providing moving control method for unmanned tobacco picking machine.