Robot Trajectory Planning in Time-Space for Moving Obstacle Avoidance

The patent extends the traditional 2D configuration space by adding a time dimension, creating a 3D state space (x, y, t). This dimensional extension allows the robot to plan trajectories that avoid moving obstacles by utilizing the time dimension, transforming a complex dynamic avoidance problem into a pathfinding problem in an expanded space where obstacles appear as volumetric regions to be avoided.

The motion planning algorithm is made dynamic by incorporating real-time obstacle position updates and recalculating trajectories accordingly. The system continuously monitors obstacle movements and adjusts the robot's path in real-time, transforming a static planning approach into a dynamic adaptive system that can respond to changing environmental conditions.

The system performs preliminary motion planning in the expanded state space before execution, pre-calculating safe trajectories that maintain required distances from obstacles. By planning the entire path in advance with safety constraints built-in, the robot can execute movements more efficiently without continuous reactive adjustments, maintaining safety while improving speed.

The algorithm dynamically adjusts the safety distance parameter based on obstacle characteristics, robot velocity, and environmental context. Rather than maintaining a fixed predetermined distance, the system modifies the safety margin parameter in real-time, allowing closer approaches when safe and enforcing larger buffers when risk increases, thus optimizing both safety and speed.