Due to changes in the working temperature in a shearer transmission system, it is easy to cause the transmission gear to work abnormally. In this paper, the first gear pair in a cutting transmission system of a permanent magnet semi-direct drive shearer was taken as the research object. The coupling effect of time-varying meshing stiffness, meshing damping, bearing clearance and gear backlash under thermal deformation were fully considered. Then, a three degree of freedom nonlinear dynamic model of the gear pair was established, and a dimensionless analysis was performed. Finally, the Runge-Kutta method was used in the numerical calculations. The motion characteristics of the system were analysed through the gear’s bifurcation diagram of the temperature at different frequencies and the bifurcation diagram of the frequency at different temperatures. The meshing state was analysed at different frequencies and temperatures. For the chaotic motion generated in the system, a periodic resonance excitation was applied to control unstable motion. This study has revealed various specific effects of the temperature on the dynamic characteristics of the system. In addition, the periodic excitation method can effectively control the nonlinear motion in the system and realise the control of the chaos under the temperature effect.

The accurate stable measurement of position for shearer is a key technology in realizing an automated fully-mechanized coal mining face. Using as a base of SINS/UWB integrated positioning system, the signal interference and node damage of UWB often cause invalid integrated positioning system, because of the bad working environment of shearer in the coal mine. Consequently, positioning accuracy of SINS/UWB integrated system decreases quickly, due to the drift error of SINS. Aiming at two invalid styles include partial anchor node failure and all anchor nodes failure of UWB, this paper proposes a stable SINS/UWB integrated positioning system of shearer using the multi-model intelligent switching method based on tightly- coupled integrated model and decision tree fault-tolerant model. First of all, the tightly-coupled integrated model and decision tree fault-tolerant model are built using measured distance and the final position of UWB respectively. Secondly, the multi-model intelligent switching algorithm can be established based on the shearer Markoff movement state and residual threshold judgment of Kalman Filter. Finally, the experimental results show that the multi-model intelligent switching fault-tolerant positioning algorithm can effectively overcome increased positioning error of tightly-coupled model and decision tree model based on these two invalid styles. The maximum position error with the positioning model proposed in this paper is 0.93 m in 327 s, and the positioning accuracy with multi-model intelligent switching model is much greater than that with tightly-coupled integration and the decision tree fault tolerant models.