Cloud-Edge-End Collaboration for the Finishing Rolling System: An Accelerated Data-driven Cooperative Distributed Predictive Control Approach

In this brief, an accelerated data-driven cooperative distributed predictive control approach with the cloud-edge-end collaboration implemented is proposed for the finishing rolling system(FRS) of the hot strip mill process. Specifically, in the end device, the operating data of the rolling mill stands and looper systems are collected and uploaded to the cloud server, which contributes to satisfying the interconnected information communication requirements for optimization problem-solving. Then, the control optimization problem is quickly solved in the cloud server, where the cooperative distributed solving method is applied, and an accelerated iteration strategy is adopted. Moreover, a regularized optimization technique is used to suppress the interference of disturbance and noise. Thus, the edge device can set the input of the bottom-level control to realize the control of the finishing rolling system with the optimal control sequences sent from the cloud server. Finally, simulations on a real finishing rolling system are conducted to illustrate the effectiveness of the proposed method.