SPEA2 is an improved version of the Strength Pareto Evolutionary Algorithm (SPEA) designed for multiobjective optimization. It introduces a fine-grained fitness assignment strategy, a density estimation technique, and an enhanced archive truncation method. The algorithm compares SPEA2 with SPEA, PESA, and NSGA-II on various test problems, showing promising results. SPEA2 uses a fixed archive size and incorporates density information to improve diversity preservation and boundary solution preservation. The fitness assignment in SPEA2 considers both the number of solutions an individual dominates and is dominated by, as well as density information. The environmental selection process ensures the archive maintains a fixed size, with a truncation method that preserves boundary solutions. The algorithm's performance is evaluated on a range of test problems, including continuous and combinatorial ones. Results show that SPEA2 outperforms SPEA on all problems, and performs similarly to NSGA-II. It also shows advantages over PESA in higher-dimensional objective spaces. The study concludes that SPEA2 is an effective multiobjective evolutionary algorithm with improved fitness assignment and archive management.SPEA2 is an improved version of the Strength Pareto Evolutionary Algorithm (SPEA) designed for multiobjective optimization. It introduces a fine-grained fitness assignment strategy, a density estimation technique, and an enhanced archive truncation method. The algorithm compares SPEA2 with SPEA, PESA, and NSGA-II on various test problems, showing promising results. SPEA2 uses a fixed archive size and incorporates density information to improve diversity preservation and boundary solution preservation. The fitness assignment in SPEA2 considers both the number of solutions an individual dominates and is dominated by, as well as density information. The environmental selection process ensures the archive maintains a fixed size, with a truncation method that preserves boundary solutions. The algorithm's performance is evaluated on a range of test problems, including continuous and combinatorial ones. Results show that SPEA2 outperforms SPEA on all problems, and performs similarly to NSGA-II. It also shows advantages over PESA in higher-dimensional objective spaces. The study concludes that SPEA2 is an effective multiobjective evolutionary algorithm with improved fitness assignment and archive management.