The prevalent research on the path planning problem in mobile node-based data collection techniques only considers the simple situation involving a single mobile node or path endpoint located at the centre of the communication. This paper considers additional situations involving both of the above two factors and abstracts from these scenarios to formulate a hybrid optimisation problem. This optimisation problem has the characteristics of high dimensionality and a large search space. To solve this problem, the following modifications were made. First, k sub-paths based on the k-SPLITOUR algorithm were obtained. Second, a method to eliminate path intersections was designed to optimise the discrete components. Finally, a hybrid glowworm swarm optimisation (HGSO) algorithm was proposed to optimise the positions of access points along the communication circle to optimise the continuous components. The global convergence analysis of the proposed HGSO algorithm is given. Simulations and comparisons with other algorithms verified that the proposed strategy can solve the path planning problem in data collection utilising multi-mobile nodes effectively.