Study On Mechanical Design And Balance Control Of Stair Climbing Mechanism With Star Wheel

With the advent of the age of information and the age of robots,the task type and the use scenarios become more diverse,so the robot is not limited to the single demand that meets the specific environment.As the most mature and most widely used wheeled structure,more robots are used by robots to give the ability to exercise in complex terrain.In the structure complexity,load capacity and speed performance,it has unique advantages in the structured environment.But in semi-structured and unstructured environments,traditional wheel structures are difficult to meet the needs of use.At the same time,the robot in the process of the work,with the increase of the complexity of the task and the traffic environment,need to be aware of the environment to ensure the stability and safety of the robot,the amount of the system is increased while the dependence on environmental perception increases.In this paper,the design of the wheel climbing robot and the design work of the star wheel climbing robot and the proposed and verification work of the balanced control method are carried out from the research of the weak drive mechanism design,the robot model establishment,the balancing algorithm and the simulation experiment.The main research of this paper is as follows:The design and prototype of the star wheel climb staircase robot are completed.Based on the principle of star wheel,differential principle,and inverted pendulum principle,the innovation mechanism is designed and applied to the robot,so that the motion mode can be adapted to the self-adjustment in the environment of both the flat and the steps,and the intelligent switching between the high speed and high torque.The mechanism of the robot can adapt to the environment changes,the robot has the ability of intelligent choice motion mode and power output mode,the flexibility of the motor body also makes the robot is sensitive to the change of the environment,so in the process of the process(especially the steps of the step),there is no need for accurate detection and modeling of the environmental information,reducing the dependence of the robot on environmental detection and the complexity of the system.The dynamic modeling strategy of migration item equivalent interference is proposed.According to the characteristics of the system,the model is migrated to the target model in the process of the robot dynamics model,and the difference is equivalent to interference.The two input four output system equivalent becomes an input and two output system with bounded interference,and the complexity of the robot model is reduced to a certain extent.In this paper,the decoupling sliding mode algorithm of radial basis function(RBF)network interference compensation is proposed based on the equivalent model.In this paper,the model is used to understand the model,and the equivalent interference is compensated by the RBF neural network,and the method of the sliding mode algorithm is used to construct the balance control algorithm of the robot.At the same time,because there is no direct participation in the system of rotation,the characteristics of the environmental effect in the process of the robot are also reduced the dependence on environmental detection.Complete the robot’s prototype simulation and physical experiment.Through Matlab and Matlab/Adams,the physical experiment of the simulation and robot is verified in the aspects of the feasibility of the mechanism,the accuracy of the model and the validity of the balanced control algorithm,and the superiority of the balanced algorithm in the ability of convergence speed and anti-interference ability,also proves the ability of the robot to have good traffic ability and motion mode in the ground and the way of the flight.At the same time,the experiment of different sections of different sections(the tunnel,slope and upper and lower step)of the situation under the condition of the scene detection also shows the feasibility of reducing the dependence on environmental detection through the flexible characteristics of the underdriven mechanism and the RBF neural network adaptive interference.