Dynamic Temperature–Humidity Decoupling and Artificial Hummingbird Algorithm-Optimized Asymmetric Sliding-Mode Control for Medical Instrument Transport Boxes

Temperature–humidity coupling, unequal channel dynamics, and actuator saturation complicate environmental control in medical instrument transport boxes. This study combines dynamic feedforward decoupling with artificial hummingbird algorithm-optimized asymmetric sliding-mode control (AHA-ASMC). A four-channel first-order plus dead-time model and relative gain analysis support loop pairing and bidirectional compensation. Direction-dependent switching gains and a continuous boundary layer are tuned offline, and a Lyapunov analysis establishes ultimate boundedness under the stated disturbance conditions. In the ablation simulations, AHA-ASMC reduces temperature settling time from 1121 to 490 s and undershoot from 0.580 to 0.294 °C relative to symmetric sliding-mode control. Prototype experiments show cross-channel deviation reductions of 52.0% and 78.1%. Additional closed-loop calculations show that identified-model mismatch reduces humidity and temperature peak suppression from 89.80% to 85.64% and from 88.03% to 79.61%, respectively. Relative to PID, temperature and relative-humidity integral absolute errors decrease by 22.7% and 39.0% during joint regulation, although settling times increase by 183 and 79 s. Door-opening tests cover five duration and angle conditions with three repetitions per controller and condition. Temperature and relative-humidity recovery-time means decrease by 60.4–66.1% and 78.5–91.7%, respectively. The measured complete embedded task takes 400 ms within the 1 s update period.