Hot Box Optimization(HBO)

This algorithm, which is a temperature-based stochastic optimization method inspired by simulated annealing

This MATLAB code implements the Hot Box Optimization (HBO) algorithm, which is a temperature-based stochastic optimization method inspired by Simulated Annealing. It includes adaptive cooling and a hybrid local search to enhance convergence speed and accuracy.Key Features:

• Objective Functions: Supports Sphere, Rastrigin, and Rosenbrock functions.
• Population-Based Approach: Uses a population of solutions to explore the search space.
• Temperature-Controlled Search: Starts with a high temperature and cools down adaptively.
• Adaptive Cooling Rate: Adjusts every 100 iterations to balance exploration and exploitation.
• Hybrid Local Search: Gradient descent-inspired refinement every 200 iterations.
• Convergence Tracking: Stores the best fitness value at each iteration and plots a convergence graph.

Main Steps:

1. Initialize Population: Randomly generate solutions within given bounds.
2. Evaluate Fitness: Compute the objective function values.
3. Hot Box Optimization Loop:

• Reduce temperature over iterations.
• Apply perturbation-based search to explore new solutions.
• Accept better solutions or probabilistically accept worse ones.
• Perform local refinement (gradient-inspired).

1. Store and Plot Convergence: Track and visualize fitness evolution.

Output:

• The optimal solution found.
• Best fitness value achieved.
• Convergence graph showing how the solution improves over iterations.