基于多传感器数据融合的播深检测系统研究

Research on Sowing Depth Detection System Based on Multi-sensor Data Fusion

目前,播种深度的实时检测存在检测精度不高、可靠性差及数据处理不足的问题。为此,在分析了播种单体四连杆运动机理后,提出了一种基于多传感器数据融合的播种深度检测方法。该方法采用角度传感器、激光传感器、超声波传感器作为播种深度检测,利用播种单体四连杆仿形上下平动特性,测量播种深度变化,并利用卡尔曼滤波与加权平均相结合的方式进行多传感器数据融合。试验与软件仿真验证表明:系统稳定可靠,且检测精度在±0.04cm以内,说明多传感器数据融合后的估计值相较单独传感器检测数值更能准确反应播种深度变化,为后期进行播种深度控制提供数据支撑。

The real-time detection of sowing depth is the basis of realizing sowing depth control.Due to the influence of soil block and stubble on no tillage planter,the real-time detection of sowing depth becomes more difficult.Due to the problems of low detection accuracy and insufficient data processing in sowing depth detection,after analyzing the motion mechanism of four connecting rods of sowing monomer,a sowing depth detection method based on multi-sensor data fusion is proposed.This method uses angle sensor,laser sensor and ultrasonic sensor as sowing depth detection method,The variation of sowing depth was measured by using the up-down translational characteristics of four connecting rods of sowing monomer.The combination of kalman filter and weighted average is used for data fusion.Compared with the detection effect of single sensor,the fusion effect is improved.Through software simulation and actual test,the system is stable and reliable,and the detection accuracy is within±0.04cm.