A Positioning Algorithm based on Vehicle-vehicle Collaborative Location for Auxiliary Transport Vehicles in Underground Mines

Aiming at the problem of narrow space and serious multipath effect in the underground tunnel of mines, the automatic driving vehicle will receive signals propagated by multiple paths, which leads to the large positioning error of the automatic driving vehicle, the Channel-slam algorithm based on the vehicle-vehicle collaborative positioning is proposed. The algorithm uses the vehicle-vehicle sending position information to each other and the multipath signal received by the vehicle to improve the positioning accuracy of the target vehicle. Firstly, the propagation of multipath signals and vehicle-vehicle communication scenarios are modelled, and then the Bayesian model of the positioning system is established by taking advantage of the characteristics of information exchange and multipath effect of downhole vehicles. The factor graph is used to estimate the position of the two target vehicles, to achieve the precise positioning of downhole vehicles. Through simulation experiments, compared with the traditional single-vehicle localisation algorithm, the vehicle-vehicle collaborative localisation algorithm proposed in this paper has faster particle convergence, smoother and more stable target tracking trajectories, and improved positioning accuracy by 48.6%.