Computational load reduction of the agent guidance problem using Mixed Integer Programming

Area of Science:

• Robotics and Artificial Intelligence
• Computational Geometry

Background:

• Agent-guidance systems face challenges with real-time obstacle avoidance due to computational demands.
• Existing methods often struggle to balance on-board processing limitations with complex calculations for trajectory planning.

Purpose of the Study:

• To propose a novel deferred decision-based technique for efficient obstacle avoidance in agent guidance.
• To reduce the computational burden in real-time applications by optimizing obstacle processing.

Main Methods:

• Implementing a deferred decision-making strategy to create dynamic obstacle clusters.
• Considering spatial and temporal relevance of obstacles during the planning process.
• Pruning irrelevant areas to decrease computational complexity.

Main Results:

• The proposed method significantly reduces the number of required computational operations.
• Processing times are comparable to the lower-bound of relaxed problem forms.
• The technique dynamically scales computational effort, increasing it only when necessary.

Conclusions:

• Deferred decision-making offers an effective approach to optimize real-time obstacle avoidance.
• This problem modeling improvement enhances agent guidance efficiency without compromising performance.
• The strategy provides a practical solution for resource-constrained real-time applications.