3.4.2 Stability Analysis of the Behavior State Machine
In this section, we validate the effectiveness of the hysteresis execution strategy by recording and comparing the orientation during the navigation process. The state transitions of the planner can be observed directly through changes in orientation.
The experiment was conducted in Scenario 3, and the comparison of orientation changes over time is represented in Fig. 7. Owing to the substantial disturbances in Scenario 3, frequent transitions occur between the planner states. The figure clearly illustrates that the orientation, represented by the blue line, undergoes dramatic changes due to the absence of a hysteresis strategy. This is particularly noticeable within the time frames of 33≤\(t\)≤94s and 121≤\(t\)≤185s. These periods correspond to complex environments that the robot navigates through.
With the integration of the hysteresis strategy, the changes in orientation become smoother. The zoomed view in figures (a) and (b) further illustrates this point. This indicates a more stable state transition when equipped with a hysteresis strategy.
The comparative test in Scenario 3 thus validates the effectiveness of the hysteresis strategy in preventing additional state transitions caused by environmental noise and enhancing system stability.