Algorithm 1. Model parameters learning adaptive control

Step1: Initialize the model parameter ${\mathit{M}}_0$, Plan length trajectory ${\mathit{y}}_{ref}$;

Step2: Obtain the initial pressure $u_0$ based on MPC;

Step3: Compare with length tolerance ${\epsilon }_t$ and ${\epsilon }_r$.

Step4: Obtain $u_i^m$ trough model parameters iterative learning law and MPC

Step5: Obtain $u_i^l$ based trough ILC

Step6: Obtain total input pressure $u_i$, real chamber length ${\mathit{y}}_i^t$ and model estimated length ${\mathit{y}}_i^p$ trough real system and estimated model, respectively. Then go to step 3.