Research On Micro-Agriculture Machine's Remote Monitoring And Visual Navigation Technology

The traditional Micro-Agriculture Machine need to be operated by operator in the field when it is used under the environment of greenhouse and orchard. There are some problems such as poor passing through the crop, low security, bad working conditions and tending to fatigue. In order to overcome these drawbacks, it is necessary to develop one kind of Micro- Agriculture Machine which has the ability of certain autonomous navigation. This paper is supported by the"863"project (2008AA100903-7), taking the research work as follows:Combining the technology of remote monitoring and visual navigation, we designed a set of autonomous navigation system based on the technology of remote monitoring and visual navigation; this system contains two parts: remote monitoring system and visual navigation system.1. In the areas of designing remote monitoring and control system:(1) We studied the wireless transmission protocol of control signals, the system achived successfully real-time transmission of wireless control signal using the wireless module; the signal delays about 10ms and the experiment shows that it has favorable real-time characteristic. The image signal also required the excellent real-time characteristic. Then we studied the IEEE802.11b and set up local wireless network based on this protocol successfully, achieving wireless real-time transmission for two-way video signals. The experiment shows that the transmission is stable;(2) We developed the upper machine monitoring software based on VC++6.0. The operator can monitor the movement of Micro-Agriculture Machine and send the control signals via this software, greatly reduced the operational strength;(3) In the lower computer system, we designed the MCU driving circuit and corresponding control circuit, researched the hydraulic system suited by the actual working conditions and ultimately drived the hydraulic system by corresponding relay driving circuit through the lower computer system successfully. Finally the system achieved the functions of turning left, turning right, pausing, forward and movement of former working department of Micro-Agriculture Machine.2. In the research field of Vision-based navigation system: (1) Through studying and comparing several methods of camera calibration, we calibrated the single camera using the method of"Tisa two-step", getting the internal and external camera parameters. The result of the experiment shows that the calibration error is validated within the allowable range and this calibration method is effective and reliable;(2) There is simple pretreatment of road images on the monitoring PC, then detected the edge of the road using the edge detecting operator, and then fit the edge through Hough transform, extracting the real-time navigation path through the method of centerline; On this basis, we carried out the research work of getting real-time lateral deviation and heading angle using Hough transform. There are some theoretical proofs which can show this method can achieve the purpose of extracting real-time navigation parameter;(3) According to the operator's practical experience, we designed the fuzzy control rules and fuzzy controller which meet the objective conditions, and then we simulate the fuzzy controller in the simulation environment of MATLAB-SIMULINK. The result shows that the fuzzy controller meets the basic request in theory.