To solve the problems of low efficiency and poor dynamic performance in the obstacle avoidance process of mobile robot

a robot obstacle avoidance method based on improved starling murmuration optimizer (ISMO) and rolling window (RW) was proposed. Firstly

two strategies were introduced to improve the SMO algorithm: Taylor optimal neighborhood strategy was introduced to reduce the possibility of the algorithm falling into local optimum

the dynamic opposition strategy was introduced to balance the search and development stages of the algorithm. Secondly

considering the problem of obstacles with unknown motion status

the RW algorithm was introduced for dynamic planning to improve the dynamic obstacle avoidance ability of the robot. Finally

simulation and experiment were carried out

the simulation results show that the ISMO-RW algorithm has shorter path length and running time in both static and dynamic scenarios

and the real vehicle experiment verifies the practical performance of ISMO-RW algorithm.