Optimization of the computational load of a hypercube supercomputer onboard a mobile robot

Area of Science:

• Robotics
• Artificial Intelligence
• Computer Science

Background:

• Mobile robots require efficient processing for time-critical missions.
• Hypercube multiprocessors offer parallel processing capabilities but need effective management.

Purpose of the Study:

• To develop a combinatorial optimization methodology for efficient hypercube multiprocessor use in mobile robots.
• To implement this methodology using large-scale concurrent algorithms.

Main Methods:

• Developed a combinatorial optimization methodology.
• Implemented algorithms based on fast simulated annealing and nonlinear asynchronous neural networks.
• Utilized compact neuromorphic data structures to model constraints like precedence and task overlaps.

Main Results:

• Presented analytic expressions for single-neuron perturbation effects on system energy.
• Demonstrated the methodology's effectiveness on a robot-dynamics benchmark.
• Showcased efficient modeling of complex operational constraints.

Conclusions:

• The proposed methodology enables efficient hypercube multiprocessor utilization for intelligent mobile robots.
• Fast simulated annealing and neural networks provide viable algorithmic approaches.
• The approach is suitable for time-critical robotic missions.