Multi-Strategy Improved Connected Banking System Optimizer for Numerical Optimization and Real Problems

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Robotics and Control Systems

Background:

• Numerical optimization and Unmanned Aerial Vehicle (UAV) path planning are critical in various applications.
• Existing algorithms often struggle with complex constraints, exploration-exploitation balance, and initial population quality.
• Connected Banking System Optimizer (CBSO) provides a foundation but requires enhancements for superior performance.

Purpose of the Study:

• To introduce the Multi-Strategy Improved Connected Banking System Optimizer (MICBSO) for improved numerical optimization and 3D UAV path planning.
• To enhance the original CBSO with novel strategies for better search coverage, solution quality, and diversity.
• To validate MICBSO's effectiveness against state-of-the-art algorithms and in real-world UAV applications.

Main Methods:

• Developed MICBSO incorporating chaos-opposition learning for initialization, Gaussian perturbation-based multi-elite guidance, and a hybrid boundary control strategy.
• Evaluated MICBSO on the CEC2017 benchmark suite against 11 other algorithms across various dimensions.
• Applied MICBSO to four complex 3D UAV path planning scenarios to assess its performance in constrained environments.

Main Results:

• MICBSO demonstrated competitive convergence accuracy, stability, and robustness compared to existing algorithms.
• The algorithm successfully generated feasible and safe 3D UAV flight paths with reduced comprehensive costs.
• Experimental results confirmed MICBSO's superiority in handling complex optimization and path planning tasks.

Conclusions:

• MICBSO offers a significant improvement over the standard CBSO, providing an effective optimization framework.
• The proposed enhancements address key limitations in initialization, exploration-exploitation balance, and solution feasibility.
• MICBSO presents a promising solution for both benchmark optimization problems and constrained UAV path planning challenges.