This paper presents a preliminary study on multi-swarm sharing scenario for particle swarm optimization, MSSPSO, to deal with uncertain-dimension factor space optimization problems. The proposed MSSPSO can provide more wide capability for unknown solution space exploration. In this paper, the MSSPSO is applied to UAV path planning problem. Based on characteristic number of different paths, it has to use different number of control point to produce varied flight paths. In order to explore suitable solution within suitable characteristic number interval, MSSPSO is employed to explore better solution and the variable-length crossover concept is used to share information among different dimension swarms. The simulation is show that MSSPSO has the ability to explore suitable solution and determine suitable characteristic for flight path. On the other hand, swarm crossover helps swarm to avoid falling local optimal position; swarm manager is applied to enhance computing efficiency and prune the helpless swarms.