MATLAB Implementations of Two Intelligent Optimization Algorithms: Biogeography-Based Optimization and Particle Swarm Optimization

Biogeography-Based Optimization (BBO) is an intelligent optimization algorithm inspired by species migration and habitat selection behaviors. The algorithm searches for optimal solutions by simulating species migration and mutation processes between different habitats. In MATLAB implementation, key steps involve habitat suitability assessment, migration operations, and mutation operations. The algorithm calculates the Habitat Suitability Index (HSI) for each habitat to determine migration probabilities, where high-suitability habitats are more likely to serve as emigration sources while low-suitability habitats act as immigration targets. The migration rate is typically implemented using a cosine-based probability function, while mutation introduces diversity through random habitat modifications. Particle Swarm Optimization (PSO) is a swarm intelligence optimization method simulating bird flock foraging behavior. In MATLAB implementation, each particle represents a potential solution with two key attributes: position and velocity. The algorithm updates particle states by tracking personal best (pbest) and global best (gbest) solutions. During each iteration, particle velocity and position are adjusted based on individual experience and collective knowledge, gradually converging toward the optimal solution. The velocity update equation incorporates inertia weight, cognitive component (personal best influence), and social component (global best influence), typically implemented using vectorized operations for efficiency. Both algorithms require careful consideration of several key aspects in MATLAB implementation: First, clear definition of objective functions and fitness calculation methods is essential. Second, proper parameter tuning is crucial - for BBO, parameters include habitat count, migration rates, and mutation probability; for PSO, key parameters are inertia weight, cognitive and social learning factors. Finally, effective termination conditions must be designed, such as maximum iteration counts or convergence thresholds. Comparatively, BBO demonstrates better performance for discrete optimization problems, while PSO excels in continuous space optimization scenarios. Both algorithms benefit from MATLAB's matrix operations and built-in optimization functions for efficient computation.