Enhanced kernel search optimizer with sand cat behavior and multi-strategy fusion for economic emission dispatch

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Metaheuristic Computing

Background:

• Kernel Search Optimization (KSO) faces accuracy loss in high-dimensional spaces.
• KSO's mapping approximation may fail to find true optimal solutions in complex scenarios.

Purpose of the Study:

• To introduce an enhanced optimization algorithm, CSTKSO, addressing KSO's limitations.
• To improve population initialization, solution perturbation, and search strategies in optimization.

Main Methods:

• Utilizes chaotic mapping for population initialization to prevent premature convergence.
• Incorporates Adaptive t-distribution mutation with dynamic parameter adjustment for exploration-exploitation balance.
• Employs sand cat behavior-inspired mechanisms for robust search optimization.

Main Results:

• CSTKSO demonstrated superior performance against original KSO and other algorithms on 50 IEEE CEC benchmark functions.
• The algorithm achieved better results in a practical economic emission scheduling problem.
• CSTKSO effectively addresses KSO's shortcomings while retaining its strengths.

Conclusions:

• CSTKSO offers a more robust and efficient optimization tool for complex, high-dimensional problems.
• The enhanced algorithm shows significant potential for real-world applications.
• CSTKSO represents a notable advancement in metaheuristic optimization techniques.