Aiming at the problems of training load perception lag, extensive intensity regulation and insufficient teaching feedback in college basketball classroom, this paper constructs a framework of training load control and teaching effect optimization for intelligent monitoring. This study fuses inertial sensing, heart rate monitoring, task tags and individual baseline information, designs an intelligent identification algorithm for training load, and establishes a dynamic classroom control model on this basis to realize the collaborative analysis of students’ load status, recovery trend and teaching task matching. Experimental results show that the Accuracy, Recall and Macro-F1 of the proposed algorithm in classroom scenes reach 92.84%, 93.12%and 92.61%respectively, and the average response time of a single window is 128 ms, which has good real-time recognition ability. After the application of the model, the classroom load overload rate decreased to 8.9%, the technical action standard rate increased to 87.9%, and the proportion of effective participation time reached 84.1%. The research shows that intelligent monitoring can connect load identification, classroom adjustment and effect evaluation as a closed-loop process, which provides a feasible path for scientific addition and reduction and accurate optimization in college basketball teaching.
