Fully Tensorized GPU-accelerated Multi-population Evolutionary Algorithm for Constrained Multiobjective Optimization Problems

2025-09-24

AI Summary

For time-sensitive constrained multi-objective optimization problems (CMOPs), existing constrained multi-objective evolutionary algorithms (CMOEAs) suffer from low computational efficiency, slow convergence, and poor GPU parallelization capability. This paper proposes a fully tensorized, GPU-accelerated multi-population evolutionary algorithm. Leveraging decomposition-based multi-population coevolution, it establishes an end-to-end tensorized and pipelined framework encompassing population evolution, constraint handling, fitness evaluation, and environmental selection. Critical GPU performance bottlenecks—including memory bandwidth limitations and thread scheduling overhead—are systematically identified and mitigated. Empirical evaluation on standard benchmarks and real-world weapon-target assignment problems demonstrates that the proposed algorithm significantly outperforms state-of-the-art CMOEAs under strict time constraints, achieving both rapid convergence and high-quality Pareto fronts.

Technology Category

• Constraint Satisfaction and Optimization: Constraint Optimization
• Search and Optimization: Evolutionary Computation
• Machine Learning: Hardware-aware ML

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Real world constrained multiobjective optimization problems (CMOPs) are prevalent and often come with stringent time-sensitive requirements. However, most contemporary constrained multiobjective evolutionary algorithms (CMOEAs) suffer from a number of drawbacks, including complex designs, low computational efficiency, and long convergence times, which are particularly pronounced when addressing time-sensitive CMOPs. Although research on accelerating evolutionary algorithms using GPU parallelism has advanced, existing CMOEAs still face significant limitations within GPU frameworks. To overcome these challenges, this paper proposes a GPU-accelerated multi-population evolutionary algorithm, termed GMPEA. We first systematically analyze the performance bottlenecks of representative CMOEAs when implemented in a GPU environment. To address the trade-off between computational speed and solution performance, GMPEA introduces a decomposition-based multi-population approach that is fully parallelized across its entire workflow. We conducted comparative experiments on various benchmark tests and real world applications: the Weapon Target Assignment Problems. The results demonstrate that GMPEA achieves competitive performance even without time constraints, while its computational speed significantly surpasses that of the compared algorithms. More critically, under a strict time limit, the performance of GMPEA drastically outperforms its counterparts. This work provides compelling evidence of GMPEA's superiority in solving time-sensitive CMOPs.

Problem

• Addressing time-sensitive constrained multiobjective optimization problems with computational efficiency
• Overcoming complex designs and long convergence times in existing evolutionary algorithms
• Developing fully parallelized GPU-accelerated algorithms for real-world constrained optimization

Innovation

• Fully parallelized GPU-accelerated multi-population evolutionary algorithm
• Decomposition-based approach to balance speed and performance
• Systematic GPU performance bottleneck analysis for algorithm optimization