Data Processing Program for Combinatorial Optimization

The patent changes the parameter of state variable transitions by determining multiple state variables for simultaneous change based on energy function differences. This parameter change enables the system to escape local minima by making larger jumps in the state space, thereby reducing solution search time while maintaining solution optimality through energy-based selection criteria

The patent introduces dynamic adaptation by adjusting the number of state variables to change based on the calculated energy function differences. When energy differences indicate promising directions, the system dynamically increases the number of state variables to change, creating a dynamic search strategy that balances exploration and exploitation to avoid local solutions

The patent applies partial action by determining a specific number of state variables to change based on energy function differences, rather than changing all state variables or using a fixed number. This partial change approach achieves sufficient exploration to escape local solutions while limiting the calculation complexity to only the necessary subset of state variables

The patent uses parameter changes by calculating energy function differences to dynamically determine the number of state variables to change. This parameter-based approach provides a systematic way to balance search efficiency and calculation complexity, changing only enough variables to escape local minima without unnecessary computational overhead