Abstract: The kinematics research was considered to the premise of the drill boom positioning control. Aiming at the problem that the Pieper criterion was not satisfied by the drilling boom structure of the 7-dof drilling jumbo, and the effective kinematic model could not be obtained by the analytical method and the numerical method. An inverse kinematics solution method of the drilling boom based on the radial basis function (RBF) neural network was proposed. Firstly, based on the construction motion constraints of the jumbo drill boom, the improved D-H method was used to establish a forward kinematics matrix model of the drilling boom. Through the forward kinematics matrix was used to collect the drill boom motion samples, and build a threelayer RBF neural network inverse kinematics model to train the samples by using the MATLAB platform. Then, the RBF method and the numerical method were used to predict and verify by taking the tunnel face with 63 holes as an example.Finally, the feasibility of the drill boom construction was verified by the ADAMS-Simulink co-simulation. The research results show that the maximum positioning prediction error of the drilling boom obtained by simulation is 0.62%, and the maximum error of the X and Y directions of the hole position and orientation predicted by the RBF method is 2.588mm and 2.336mm respectively. The prediction accuracy is better than that of the numerical method. The proposed control method can make the positioning error of the end of the drilling boom within the allowable range of drilling construction, improve the drilling positioning accuracy of the rock drilling jumbo, and avoid tunnel section overexcavation.
Key words: drilling jumbo; drill boomstructure; drill boom positioning control; end positioning error; kinematic analysis; radial basis function(RBF) neural network