| The quadruped robot has always been a key research topic in the direction of robots.In this paper,a parallel quadruped robot based on the 3PUPS mechanism is designed.It can adapt to various complex terrains and efficiently complete specified tasks.In this paper,the method of parameter identification is used to calculate the dynamic model,so that it can complete the dynamic calculation of the mechanism faster and more accurately.In order to make the quadruped robot walking more stable,this paper proposes a virtual model gait control algorithm based on central mode control.First,the dynamic model is linearized,the inertial parameters of the mechanism are extracted,and then the minimum inertial parameter set is found.The collected arrival data is input into the dynamic model and singular value decomposition is performed to obtain the minimum inertial parameter set of the dynamic model.In order to accurately identify the parameters,the motion trajectory is expressed by the robot's motion parameters,and then the trajectory of the Fourier series is used to describe the trajectory.Optimize the influence of the ill-conditionedness of the matrix on parameter identification.Finally,the weighted least square method is used to solve the statically indeterminate equations composed of the dynamic model to obtain the inertia parameters of the mechanism.The controller of a quadruped mobile robot is divided into two parts.First,the network topology is composed of four coupled Kuramoto nonlinear oscillators,and the mapping relationship between the oscillator network and the actual output of the joint is constructed,and then the actual output is converted into the actual driving torque through a spring damper.Secondly,through the virtual model controller,calculate the foot-end virtual force required for the posture adjustment of the quadruped mobile robot,and convert the foot-end virtual force into the output required by each joint according to the virtual working principle. |
PDF Full Text Request