Research On Active Following Control Of The Pseudo-driven Wheel Of A Three-wheeled Planetary Rover

The terrains on the surface of planets and other astronomical bodies such as Mars and the Moon are mostly rough and deformable,which thus present considerable challenges to the trajectory tracking control of wheeled mobile robots(WMRs).Increasing the number of driving wheels can reduce the load on each wheel,which in turn helps to decrease wheel sinkage and slip,thus improving the WMR tractive capability.However,it causes the issue of multi-wheel redundant drive control and high energy consumption of the system.When the planetary rover moves on rough terrain,there is some internal force between the wheels resulting from the different motion states of the wheels,which reduces the movement efficiency of the rover.Therefore,the coordinated control of the driving wheels is the key to the performance of planetary rovers' motion and optimization energy consumption.The main problem to be solved in this paper is to reduce the internal force confrontation between the wheels and to ensure that the planetary rover has a higher moving efficiency.First,a new wheel is defined from the perspective of control: Pseudo-driven Wheel(PDW).In a three-wheeled rover,the kinematic constraint equations of the PDW were derived to reduce the dimensionality of the kinematic model.For reducing the internal force confrontation between the wheels,researching the principle of active following control of the velocity-tracking force by adaptively distributing the driving velocity and steering angle of a PDW.Designing an active following control algorithm with the active disturbance rejection control(ADRC)technology,which takes the generation velocity of the PDW as the control variable and takes its force state as the perception variable.The Routh-Hurwitz stability criterion(Routh-Hurwitz)is used to prove the stability of the closed-loop control.In order to enhance the following ability of the PDW to body on uneven terrain,an active following control strategy based on the degradation by degrees-offreedom is proposed.To prove the effectiveness of the active disturbance rejection control algorithm and the control strategy,a simulation platform of planetary rover motion control was built using ADAMS-SIMULINK.Then,for achieving the PDW control efficiently and effectively,the characteristics of the kinematics model control which providing a sufficient driving ability,and the characteristics of the wheel-terrain mechanics control which decreasing the internal force confrontation between the wheels are combined to design an online neural network PID self-tuning control system based on a speed feedforward.In the system,the kinematics model of the rover is used for a speed feedforward,and the online extreme learning machine PID(OS-ELMPID)is to deal with the influence of dynamic disturbance.The established simulation platform of motion control proves that this control system has a better performance in unknown environments.Finally,some path-following experiments of three-wheeled planetary rovers is carried out to analyze the practicability of the proposed kinematics constraints and the velocity-tracking force method of the PDW.It proves that active following control of the OS-ELM-PID control based on a velocity feedforward has important application value.