Objective The multi-objective particle swarm algorithm shows great potential in solving complex multi-objective optimization problems due to its simple operation and fast convergence. However how to effectively maintain the diversity of solutions while ensuring the fast convergence of the algorithm has always been a key challenge in this field. Methods To address this challenge this study innovatively proposes a multi-objective particle swarm optimization algorithm that combines the Cubic-Sine hybrid mapping and the Centre strategy CSMOPSO . With its unique design this algorithm achieves a comprehensive and significant improvement in performance. In the initialization stage of the CSMOPSO algorithm through a carefully designed chaotic parameter selection mechanism the algorithm can intelligently decide whether to use the Cubic or Sine chaotic mapping to generate the initial population. This strategy not only ensures the diversity and uniform distribution of the population but also lays a solid foundation for the algorithm to conduct an efficient search in a vast solution space significantly enhancing its global search ability. Furthermore the CSMOPSO algorithm introduces a cone domain to store non-dominated solutions and uses the Centre strategy to optimize the management of the external archive. Under the combined action of these innovative measures the quality of nondominated solutions in the archive is significantly improved and thus the convergence and diversity of the algorithm are balanced. Results By comparing CSMOPSO with other multi-objective particle swarm algorithms and multi-objective evolutionary algorithms on a series of benchmark test problems such as the ZDT and UF series the experimental results show that the CSMOPSO algorithm exhibits excellent performance in terms of both diversity and convergence speed and outperforms other algorithms on most test functions. Conclusion The CSMOPSO algorithm successfully achieves a good balance between convergence and diversity significantly improving its overall performance and demonstrating strong competitiveness in solving multi-objective optimization problems.