Research On Remote Driving Control System Of Agricultural Hydraulic Chassis

Due to the complex terrain in hilly and mountainous areas,manned agricultural machinery has low safety and efficiency,making it difficult to develop completely unmanned agricultural machinery in a short period of time.This paper uses a small full hydraulic chassis crawler tractor as the control object,develops a remote driving control system based on a 4G network,and designs the BP-PID self-tuning controller to address the issue of easy chassis deviation in farmland,with the goal of increasing agricultural production in hilly and mountainous areas.On this basis,the prototype is tested,and the remote driving test and deviation correction test are designed to verify the function of the system.The main research is as follows:(1)The chassis hydraulic transmission system,composed of steering gear,a double pump and a hydraulic motor,is analyzed.The mathematical model of the system is established under ideal conditions,and the second-order transfer function between the input(voltageu)and output(angle θ)of the system is fitted by the MATLAB system identification tool.(2)Taking the Cortex-A7 embedded device I.MX6 ULL of the arm architecture as the main controller,taking the Linux operating system as the software platform,building a web video server based on the TCP protocol,and building a web remote driving interface with Bootstrap framework and Ajax technology to complete the development of the upper computer of the system.The lower computer of the system is constructed using the Cortex-M4 control core,relay,and steering gear as executive elements,the duplex pump,hydraulic motor,starter,and fuel injection nozzle as control objects,and the encoder and inertial measurement unit(IMU)as detection elements,the fitting precision is 90.1%.(3)Aiming at the problem that the chassis is easy to deviate in complex farmland environment,a skew correction strategy is designed.According to the nonlinear characteristics of the chassis drive system,a BP neural network algorithm is introduced based on the PID algorithm to design its deflection correction controller.Finally,Simulink is then used to simulate it and train its optimal weight and threshold.From the simulation results,it can be concluded that BP-PID has a faster convergence speed than a single PID controller.(4)In the remote driving test,the driving duration is 28 minutes,and the video display of the upper computer is smooth during driving.The minimum RTT delay from the upper computer to the lower computer is 170 ms,and the RTT delay of more than 1 s is 1310 ms,which occurs only once.The average RTT delay in the whole process is 222.75 ms.It is estimated that the overall average delay of the system is 406.95 ms.In the skew self-correction test,when driving at the speeds of 0.5 m/s,1 m/s,and 1.5 m/s,the average skew correction rates are 87.5%,81.2%,and 72.3%,respectively.