Evaluation of Obstacle Avoidance Performance for Autonomous Navigation Based on Combinatorial Empowerment

Autonomous navigation obstacle avoidance technology can ensure that unmanned systems avoid collisions with obstacles or other objects during task execution, especially in dynamic and complex environments. This ability is particularly important, but there is currently no unified standard to evaluate the obstacle avoidance performance of autonomous navigation systems. Therefore, this article proposes a method for evaluating the obstacle avoidance performance of autonomous navigation based on combination weighting. Through in-depth research on the relevant knowledge of autonomous navigation systems and the influencing factors of autonomous navigation obstacle avoidance performance, a reasonable evaluation index system is established. The five basic indicators of time consumption ratio, path consumption ratio, energy consumption ratio, continuity, and safety are applied to quantitatively calculate the obstacle avoidance ability of autonomous navigation systems from different dimensions. Using entropy weight method and analytic hierarchy process to combine and weight indicators, and overcoming the subjectivity of subjective weighting by introducing objective weighting methods. A performance evaluation model for autonomous navigation obstacle avoidance has been constructed to scientifically evaluate the obstacle avoidance level of autonomous navigation systems. And experimental verification was conducted, and the experimental results proved that the method proposed in this paper can scientifically and reasonably evaluate the obstacle avoidance performance of autonomous navigation systems, which is of great significance for the research and development of autonomous navigation.