Balancing strategy between exercise training intensity and injury risk under multi-objective optimization framework

There is a challenge in designing training programmes to balance exercise intensity and the prevention of injury, especially when using exercises like Chinese Tai Chi for both physical conditioning and injury prevention. The research proposes a multi-objective optimization framework which aims to systematically identify optimal training strategies by maximizing training effectiveness while simultaneously minimizing injury risk. Balanced Intensity Optimization for Training with Injury Control using Non-dominated Sorting Genetic Algorithm-II (BIOTIC-NSGA), a multi-objective optimization method that aims at maximizing training efficacy and reducing the time spent on injury risk by intelligent training schedule adaptation. The framework was validated using a Tai Chi exercise dataset collected from 16 participants, comprising biomechanical and neuromuscular variables including muscle strength, joint torque, joint angle, activity level, and injury risk to identify personalized training intensities and recovery windows. BIOTIC-NSGA employs the NSGA-II evolutionary algorithm with constraint-dominance logic and repair mapping projection to ensure feasibility in the solution space. The optimization targets minimizing injury probability while maximizing performance and load balance, providing Pareto-optimal scheduling and intensity strategies. Experimental evaluation against benchmark methods using Precision and F1-score metrics demonstrates that BIOTIC-NSGA achieves 9.14% higher Precision and 10.13% higher F1-score, indicating improved training adaptation and injury-risk optimization. The proposed framework provides a data-driven decision-support tool for coaches to optimize Tai Chi training by systematically balancing performance improvement and injury prevention, while its applicability to other exercise domains will be investigated in future work.