Share this

Design and Implementation of Path Navigation System for Electromagnetically Induced Agricultural Spraying Robot

2026-04-06 06:32:18 · · #1
Abstract: To meet the need for automated pesticide spraying operations in greenhouses, an electromagnetically guided agricultural spraying robot path navigation system was designed. A signal generator system based on an ARM7 microcontroller was designed and implemented using digital waveform synthesis technology. A robot position detection sensor and a magnetic marker detection sensor based on a Hall chip were developed. The navigation control of the spraying robot was successfully implemented using the signal generator and sensors . Keywords: robot; navigation system; spraying 1 Introduction There are various navigation and positioning methods for robots, including machine vision navigation, GPS navigation, sensor navigation, and electromagnetic navigation. Visual navigation primarily involves robots acquiring environmental information via CCD cameras and planning a path to move to a predetermined target without human intervention. A major drawback of this method is insufficient information from the CCD camera when the robot turns, necessitating additional sensors. GPS navigation, a satellite-based navigation and positioning system, is suitable for wireless users, offering high positioning accuracy and all-weather operation, but suffers from poor interference resistance. Sensor navigation encompasses various forms, such as CCD visual sensors, ultrasonic sensors, infrared sensors, and magnetic field sensors. Multi-sensor fusion technology is employed in sensor navigation, making it widely applicable; however, the detection range and accuracy of sensors are not ideal in certain special environments. Electromagnetic navigation, developed in the United States in the 1950s, rapidly developed and became widely used in flexible manufacturing systems in the 1970s. The Shenyang Institute of Automation, Chinese Academy of Sciences, has produced multiple generations of mobile robots based on electromagnetic navigation. The BRAIN Institute in Japan has applied this navigation method to the automation of agricultural spraying machinery and developed related products. The pesticide spraying robot designed by this author based on electromagnetic navigation technology is small, flexible, and highly reliable, meeting the requirements of greenhouse plant protection operations. This paper describes the composition of the pesticide spraying robot, elaborates on the design of the electromagnetic navigation system, focuses on the design and implementation of the signal generator system and sensor system, and finally presents the relevant experimental results of the robot. 2 Composition of the robot system Figure 1 shows the robot system used in the experiment. The robot consists of a four-wheel drive unit, a control unit, a sensor conditioning unit, and a sprayer. The front and rear axles are each connected to the vehicle body by a bearing, which limits the angle between the front and rear axles and the vehicle body to within ±45°. Each wheel is driven individually by a DC motor and is encapsulated in a motor housing. The robot is powered by a battery. The robot uses the differential speed of the left and right wheels to achieve steering control. Eight electromagnetic sensors are installed at the front and rear of the four wheels to obtain the deviation signal between the robot and the guide line. Two degree sensors are installed on the bearings at the connection between the wheel axles and the vehicle body to obtain the angular deviation signal between the front and rear axles and the vehicle body. Two magnetic marker sensors are placed at the front and rear of the vehicle to detect the fixed position during movement. The block diagram of the robot control system is shown in Figure 2. [align=center] [/align] 3. Robot Navigation System 3.1 Navigation Working Principle According to Figures 2 and 3, the robot path navigation system consists of a robot controller, a guidance signal generator, a guide line, magnetic markers, sensors, and a remote controller. Their relationship is as follows: the signal generator provides the guidance line signal used by the robot for movement; based on the actual greenhouse environment, the guide line is pre-laid in the field to make work plans; the robot uses position sensors to detect the guide line signal, obtain path information, and achieve tracking; and uses marker sensors to detect magnetic markers (ferromagnetic materials) at corresponding positions to perform spraying operations; the remote controller can remotely control and monitor all processes of the robot. The guide line (see 5 in Figure 3) carries an alternating current of a certain frequency, which is drawn from the signal generator (see 6 in Figure 3), providing the path and direction information for the robot's movement. There are two types of sensors: one is the position detection sensor (see 1, 2, 4, 8 in Figure 3), which mainly picks up the magnetic field signal of the guide line to obtain the robot's position information; the other is the magnetic marker detection sensor (see 3 in Figure 3), which is mainly used to detect magnetic materials at predetermined positions (see 9 in Figure 3) to perform specific operations such as spraying. Our designed navigation system uses eight position detection sensors: four sensors (sets 8 and 2) in front of the robot's wheels when moving forward, and four sensors (sets 1 and 4) behind the wheels when moving backward. The robot uses these sensors to obtain path signals, which are then processed by the controller to control its movement. 3.2 Path Guidance Signal System 3.2.1 Guidance Signal Generator The designed signal generator is a system implemented using an ARM7 series microcontroller (LPC2106, manufactured by Philips) as its core. It has functions such as signal generation, power amplification, 16-channel switching, LED display, and wireless data transmission. The signal generator uses digital waveform synthesis technology to generate frequency-controllable sine waves, triangular waves, and other waveforms. The working principle of the signal generation unit of this system is shown in Figure 4. The signal generation principle is as follows: the control pulse is a controllable pulse sequence issued by the LPC2106 controller, which is generated by the timer/counter in the controller. After the control pulse is input to two counters (8 bits in total), the counters start counting pulses, and the output terminals sequentially output signals 0x00, 0x01, 0x02... These signals become the address signals of the EPROM. The EPROM stores discretized data of sine waves or other waveforms, which are then output to the D/A converter to be converted into sine waves. After filtering by a second-order filter, a smooth sine wave is obtained. In the experiment, it was found that the sine wave driving capability was insufficient, and the signal attenuated when the induced line was long. We added a power amplifier circuit, which enabled the signal generator to work stably. In addition, the remote control can set the frequency, time the signal, switch channels, and monitor the status of the signal generator. It also has functions such as walking control, spray control, and status monitoring for the robot. 3.2.2 Induction Signal Sensor (Position Sensor) When an N-turn coil with a cross-sectional area of ​​S is placed next to an infinitely long conductor carrying an alternating current J, the induced electromotive force s generated at both ends of the coil circuit is: where θ is the angle between the normal to the coil cross-section and the direction of the magnetic field; μ<sub>o</sub> is the permeability of free space, a constant; r is the distance between the sensor and the induction line; and is the rate of change of current. It can be seen that when a coil L is placed around a conductor carrying an alternating current, the magnitude of the induced electromotive force at both ends of L is related to the rate of change of current, the number of turns N of the coil, the cross-sectional area S of the coil, the angle θ between the coil and the conductor, and the distance r between the coil and the conductor. The structural principle diagram of the position detection sensor is shown in Figure 5(a). It consists of an inductor and a capacitor C. Their combination can detect the magnetic field around a conductor carrying an alternating current, induce an alternating voltage signal u, and then convert it into a DC voltage signal through a signal conditioning circuit. Figure 5(b) is the working principle diagram of the magnetic marker detection sensor. As shown in the figure, the robot path recognition involves detecting the distance deviation between the sensor and the induction line. Two sensors are placed on either side of the robot, with the guide line below them. Ideally, when the guide line is between the two sensors, the sensor output voltages are the same, and their difference is zero. Conversely, if the guide line is biased towards either side, the closer sensor will have a larger output voltage, resulting in a non-zero difference between the two sensors, either positive or negative, creating a deviation. Using this information, adjusting the speed of the left and right wheels can guide the robot back to the ideal position, thus achieving path navigation along the guide line. Figure 6 shows the actual characteristic curve of our designed position detection sensor, illustrating the relationship between detection distance and output voltage. Considering factors such as the effective voltage signal output, we ultimately chose a sensor with L=11.3mH and C=1μF. The sensor design also involved the selection of inductor core material. The data in Figure 6 was measured when the alternating current in the conductor was 0.18 A. In practical applications, the effective detection distance of the sensor must be above 15cm (determined by the working environment) to meet the requirements of the working environment. The voltage signal detected by the sensor designed at 15cm is approximately 12mV, which must be amplified before use. The output voltage u is amplified to 3V by a voltage amplifier and then input to the robot controller via an AD converter. 4. Ferromagnetic Detection Sensor: In practical applications, robots need to perform specific operations at certain locations during movement. Ferromagnetic objects are placed at corresponding locations, and the robot's magnetic detection sensor detects these objects. The robot can then perform related operations based on the detected ferromagnetic information. The working principle of the magnetic detection sensor is that when the spraying robot passes a location with a ferromagnetic object, the sensor outputs a voltage pulse signal that changes from 1 to 0 or from 0 to 1. This signal is transmitted to the robot controller. The controller records this pulse signal; it records n times for each n ferromagnetic object, and finally performs the corresponding action based on the number of pulses n. The designed sensor consists of two Hall effect chips, an instrumentation amplifier, and a trigger. The principle block diagram is shown in Figure 7. The UGN3503 is a linear Hall effect sensor manufactured by Al-Iegro MicroSystems. When there is no magnetic field, the sensor outputs half the supply voltage; when a magnetic field is present, the output voltage is greater than half the supply voltage when facing the magnetic field, and less than half the supply voltage when facing away from the magnetic field. If a magnetic mark detection sensor made from a single UGN3503 chip is used in the design, the sensor circuit will be more complex and suffer from instability issues such as temperature drift. However, if two UGN3503 chips are used, facing the magnetic field in opposite directions respectively, and their outputs are connected to the differential input of an instrumentation amplifier, twice the output signal of a single UGN3503 chip can be obtained, enhancing the detection of magnetic fields and improving drift resistance. The characteristic curve in Figure 8 reflects the relationship between the sensor reference voltage and the detection distance. In previous research and designs, the main problem with the sensor was its short detection distance, requiring calibration every two weeks. The main reason for the sensor failure is that the output voltage ∞ of the instrumentation amplifier is calibrated at 0V (obtained by adjusting potentiometer VR) when there is no magnetic field, and the reference voltage (see Figure 7) is 2V. With changes in temperature and time, ∞ will drift to about +3V above the reference voltage, or much lower than the reference voltage, such as -3V (which is beyond the amplification range of the instrumentation amplifier), thus causing the sensor to fail. The improved design optimizes the reference voltage by directly grounding it to 0V; and adjusts the output voltage of the instrumentation amplifier to about -2V under static conditions. The improved sensor has an effective detection distance of more than 20cm, which meets the usage requirements. The characteristic curve of the improved sensor is shown in Figure 8 when U0 = -2.0V. 5 Indoor Experiment We conducted a walking experiment of the spraying robot indoors. Using the spraying robot shown in Figure 1, the current in the induction line was operating at a frequency of 1.5kHz, the position sensor was 15cm above the ground, and the magnetic marker sensor was 20cm above the ground. Two sets of tests were conducted: walking in a straight line and turning. The minimum turning radius of the robot is 600cm, which is determined by the mechanical structure of the robot; the robot can follow the guide line when its walking speed is in the range of 0.15-0.52m/s; when the robot walks in a straight line at a speed of 0.45m/s, the maximum distance h from the midpoint of its front wheel to the guide line is about 2cm; when walking at a 45° angle, the maximum distance h from the midpoint of its front wheel to the guide line is about 30cm; at the same time, the magnetic marker sensor can detect ferromagnetic objects within a predetermined height and can assist the robot in performing predetermined operations. 6 Conclusion The path navigation system designed in this paper enables the application and development of electromagnetic navigation in the intelligentization of agricultural machinery. The experiment shows that this navigation method can assist the spraying robot in completing all walking functions, and this navigation method also provides a reference for the intelligentization research of other agricultural machinery. References [1] Cho SI, Lee JH. Chung S O. Autonomous speed sprayer using DGPS and fuzzy control (I) [C]. Journal of the Korean Society for Agricultural Machinery, 1997, 22(4): 487-496. [2] Ouyang Zhengzhu, He Kezhong. Application of GPS in intelligent mobile robots [J]. Microcomputer Information, 2001, 15(11): 56-58. [3] Wang Zhiwen, Guo Ge. Current status and prospects of mobile robot navigation technology [J]. Robot, 2003, 25(5): 470-474. [4] Iida M. Field Automation in Japanese Orchard [EB/OL]. http://pars.ifas.un.edu/Tedmieal%20Session/O1.1017. Miehi&JP/BRAIN%20by%20Miehi.pdf, 2001. [5] Mao Zhenlong. Magnetic field measurement [M]. Beijing: Atomic Energy Press, 1985, 45-55. [6] Zhang B, Chen z Q. The development of a new type of magnetic field sensor used for detecting magnetic marks [A]. Proceedings of the World Engenders' Convention [C], Beijing, China: China Science and Technology Press, 2004, pp. 180-183. About the authors: Yang Shisheng (1980-), male, Master's student. Research field: electromechanical control technology. Zhang Bin (1964-), male, Professor, Doctoral supervisor. Research field: electromechanical control technology, robotics. Yu Xufeng (1978-), male, Master's student. Research field: electromechanical control technology.
Read next

CATDOLL 131CM Amber Silicone Doll

Height: 131 Silicone Weight: 28kg Shoulder Width: 32cm Bust/Waist/Hip: 67/56/72cm Oral Depth: N/A Vaginal Depth: 3-15cm...

Articles 2026-02-22
CATDOLL 128CM Diana Silicone Doll

CATDOLL 128CM Diana Silicone Doll

Articles
2026-02-22
CATDOLL 115CM Alice TPE

CATDOLL 115CM Alice TPE

Articles
2026-02-22
CATDOLL 148CM Qing Silicone Doll

CATDOLL 148CM Qing Silicone Doll

Articles
2026-02-22