Research On Electrical System Control Technology Of Agricultural Unmanned Vehicle

With the continuous development of electrical machinery technology and the increasing labor cost,the automation of agricultural machinery and equipment has become one of the important contents in the field of agricultural robot research.Although there are some large-scale unmanned automatic operation equipment,such as unmanned tractor and high-speed self-walking rice transplanter,these equipment will not be effective for small areas of terraced fields and hilly areas in southwest and northwest of China.In order to solve this problem,it is necessary to design a small electronic controlled automatic agricultural machine,that is,a two-wheeled agricultural unmanned vehicle,which can meet the needs of small areas of farmland cultivation,and can improve operational efficiency and reduce labor consumption.According to the structural characteristics of the two-wheeled farming machine and the functional requirements of unmanned operation,the existing equipment was transformed automatically and the actuator in the electrical system was studied.According to the structure analysis of mobile platform of agricultural unmanned vehicle,the mathematical model of motion system is established.Considering the uncertainty of motion model and influence of complex farming environment,the straight path tracking controller is designed based on RBF neural network model.When the vehicle is cultivating on straight line,the system nonlinear function can be approximated on line effectively so that high precision straight path tracking is realized.When the agricultural unmanned vehicle drives to the corner of the field,it needs to make automatic steering action.Therefore,a dual-input steering controller is designed.The control algorithm is a compound control combining single neural network and fuzzy control,which ultimately enables the vehicle to achieve the purpose of differential steering.On the basis of motion control and combined with function modules of positioning and navigation,path planning and remote operation,the integrated experiment of whole-vehicle tillage was carried out for the developed experimental prototype.In the experiment,the unmanned agricultural vehicle could repeatedly take the "snake" route back and forth in the field according to the planned path,and independently complete the tillage task.Experiment data shows that the lateral error of agricultural unmanned vehicle is within ± 10 cm when driving straight and the error of steering angle in corners within ± 5°,which meets the controlling demands.