Predication of Rock Cutting Force of Conical Pick Based on RBF Neural Network

Abstract

The mechanized level of coal mining work is developing rapidly, the roadheader and shearer are the most important equipments. The conical pick acts on coal or rock directly, and its design and selection play an important role in the operational reliability of these equipments. However, the cutting force is the basis to design and select conical pick [1, . At home and abroad, many researchers attempted to establish the theoretical models for calculating the peak cutting force of conical picks. Evans used the maximum tensile criterion to set up the theoretical model for conical pick, and the model was widely used in the design of roadheader, shearer and so on in Europe and America [. Goktan et al. developed the presented model based on Evans theoretical model, and they given the modified model [. Okan et al. adopted the finite element method and the discrete element method to simulate the rock or coal cutting process to predicate the cutting force [. As mentioned above, the theoretical model is easy to calculate the cutting force, but the results have big difference with the experimental result due to simplify the relations between the cutting force and parameters. The numerical method is difficult to predicate the cutting force accurately, because the influence of rock or coal constitutive model, element failure criterion and boundary condition on cutting process. Therefore, the RBF Neural Network for predicating the cutting force was established base on a large amount experimental data in this paper. The aim is to predicate the cutting force accurately and reliably, and provide some basis to design and select the concial pick.