New Optimization Approach Using Clustering-Based Parallel Genetic Algorithm
This work addresses computational bottlenecks in optimization for researchers and practitioners, but it is incremental as it combines existing methods like SLM and genetic algorithms with parallelization.
The paper tackles the high computational cost of the Subdivision Labeling Method (SLM) for global optimization in high-dimensional spaces by proposing a Clustering-Based Parallel Genetic Algorithm (CBPGA). The results show that SLMCBPGA improves speed and efficiency, as demonstrated through numerical examples.
In many global Optimization Problems, it is required to evaluate a global point (min or max) in large space that calculation effort is very high. In this paper is presented new approach for optimization problem with subdivision labeling method (SLM) but in this method for higher dimensional has high calculation effort. Clustering-Based Parallel Genetic Algorithm (CBPGA) in optimization problems is one of the solutions of this problem. That the initial population is crossing points and subdividing in each step is according to mutation. After labeling all of crossing points, selecting is according to polytope that has complete label. In this method we propose an algorithm, based on parallelization scheme using master-slave. SLM algorithm is implemented by CBPGA and compared the experimental results. The numerical examples and numerical results show that SLMCBPGA is improved speed up and efficiency.