| Compared with the wheeled and tracked robots, legged robots had thecharacteristics of flexible movement, strong environmental adaptability anddiscontinuous ground support. But a lot of basic theories of legged robots are not perfect.In this thesis, a motion state estimation method of quadruped robot was researched. Avelocity estimation method of a quadruped robot based on the kinematic model wasproposed, and state perception of legged robot’s body was achieved.The three aspects of the tasks including theoretical algorithm proposed, thedynamic simulation environment built and verified and the experiment performed weremainly completed in this thesis. The main contents of the paper are as follows:Firstly, using the knowledge of robotics, the kinematic model for the mechanicalstructure of hydraulic actuated legged robot was established. Combining the measureresults of accelerometers, the algorithm was optimized and improved.Secondly, in order to verify the effectiveness of the algorithm, dynamic simulationenvironment was built using the Solidworks, ADMAS and Matlab software and acoordinated simulation model using the ADAMS model and Matlab/Simulink controlsystem.Using the way, we could control and motion the mechanical model. And then,the proposed algorithm was applied and verified on the simulation environment.Although the algorithm could be well validated in the simulation model, the actualprototype was still different from the simulation environment. In the fourth chapter, theexperiment was performed on the high-performance hydraulically actuated quadrupedbionic robot platform. Our prototype was elaborated in detail and the proposedalgorithm was applied on the QNX system development platform using C languagedevelopment tools. In this way, the speed of the legged robots could be calculated.The simulation and experiment results show the proposed method based on thekinematic model is effective and it can reflect the body’s movement in the absence ofthe speed sensors availably. |
PDF Full Text Request