Fig. 3 | Performance analysis of microrobot navigation in various environments using reinforcement learning algorithms.

Reward trajectories for: a. Multi-output Tributary channel, b. Circuitous channel, c. Vascular channels with Dreamer V3 (blue) outperforming the hyperparameter-tuned state-of-the-art PPO (green) across simulation steps. d. Comparison of PPO and Dreamer algorithms in reaching targets across different channel types: Racetrack, tributary, SPA, Squares, Vascular, and Maze. e. Impact of different reward functions on the rate of target achievement. f. Effects of frame skipping on performance, presented in a logarithmic plot. g. Influence of training ratios on reward dynamics, highlighting consistent performance across various ratios in simulated environments.