Multi-objective excavation trajectory optimization for unmanned electric shovels based on pseudospectral method

With the proposal of intelligent mines, unmanned electric shovels have become a research hotspot in recent years. In the field of autonomous mining, optimal excavation trajectory planning is a key issue since it has a considerable influence on production efficiency and energy consumption. In this paper, a multi-objective trajectory optimization framework based on pseudospectral method is proposed for excavation trajectory planning in autonomous mining scenarios. First, the machinery-ore coupling dynamics of electric shovel is modeled based on the Lagrange method and a multi-objective optimization model is established. Then, the trajectory optimization model is considered as a continuous Optimal Control Problem (OCP) with multiple constraints, a Radau pseudospectral method is developed to discretize the constraints, states and control variables of the dynamics model at the Legendre-Gauss-Radau collocation points, and the shovel dynamics and objective function are converted to algebraic forms. Finally, the associated Non-Linear Programming (NLP) is solved to obtain the optimal excavation trajectory and optimal control variables. In addition, the mapping relationship between the co-states of the OCP and KKT multipliers of the NLP is derived to assess the optimality of the solutions. The results confirm the effectiveness of applying the proposed framework to produce optimal excavation trajectories for unmanned electric shovels by performing a number of simulation and experimental studies. Moreover, the proposed framework tends to be more capable in terms of excavation time and energy consumption compared with other common approaches.

Automated summary

In this paper, a multi-objective trajectory optimization framework based on the pseudospectral method is proposed for excavation trajectory planning in autonomous mining scenarios. The framework models the machinery-ore coupling dynamics of electric shovels and establishes a multi-objective optimization model. By discretizing the constraints, states, and control variables using the Radau pseudospectral method, the framework solves the associated Non-Linear Programming (NLP) to obtain the optimal excavation trajectory and control variables. Experimental studies confirm the effectiveness of the proposed framework in generating optimal excavation trajectories for unmanned electric shovels.