Data clustering plays a critical role in organising similar data objects into meaningful groups, thereby reducing complexity and enhancing knowledge discovery without prior class information. In this study, we introduce a novel clustering framework based on NOVIRA-driven TRIVEX-DE, designed to improve clustering performance through intelligent adaptive optimisation. The proposed approach strategically partitions candidate solutions into optimised communities by regulating the NOVIRA evaluation mechanism, ensuring robust convergence toward globally optimal cluster structures. Unlike conventional clustering methods that often suffer from premature convergence or sensitivity to initialisation, the proposed model emphasises global exploration while preserving adaptive exploitation capabilities. A comprehensive comparative analysis is performed against several established clustering techniques, including Self-Adaptive BFO, K-means, Particle Swarm Optimisation clustering, and ACO clustering. Experimental findings demonstrate that the proposed algorithm efficiently handles datasets with varying cluster sizes, densities, and dimensional complexities while achieving superior clustering accuracy and stability. Furthermore, TRIVEX introduces an advanced optimisation paradigm for addressing NP-hard clustering problems by integrating adaptive evolutionary search with intelligent solution refinement. To validate the robustness and competitiveness of the proposed method, a benchmark suite of 25 CEC-2005 real-parameter single-objective optimisation functions is employed, where the algorithm exhibits strong performance compared with existing evolutionary approaches. Additionally, extensive experiments conducted on five real-world datasets, including Iris, Glass, Breast Cancer, Wine, and Vowel, confirm that the proposed NOVIRA-based TRIVEX approach significantly enhances clustering quality, optimisation efficiency, and overall computational performance compared to conventional evolutionary clustering algorithms.