Workflow-based Fast Data-driven Predictive Control with Disturbance Observer in Cloud-edge Collaborative Architecture

Abstract

Data-driven predictive control (DPC) has been studied and used in various scenarios, since it could generate the predicted control sequence only relying on the historical input and output data. Recently, based on cloud computing, data-driven predictive cloud control system (DPCCS) has been proposed with the advantage of sufficient computational resources. However, the existing computation mode of DPCCS is centralized. This computation mode could not utilize fully the computing power of cloud computing, of which the structure is distributed. Thus, the computation delay could not been reduced and still affects the control quality. In this paper, a novel cloud-edge collaborative containerised workflowbased DPC system with disturbance observer (DOB) is proposed, to improve the computation efficiency and guarantee the control accuracy. First, a construction method for the DPC workflow is designed, to match the distributed processing environment of cloud computing. But the non-computation overheads of the workflow tasks, such as receiving and sending data, serialization and deserialization, are relatively high and affect the overall processing speed of workflow. Therefore, a cloud-edge collaborative control scheme with DOB is designed. The low-weight data could be truncated to reduce the non-computation overheads. Meanwhile, we design an edge DOB to estimate and compensate the uncertainty caused by the truncation operation in cloud workflow processing, and obtain the cloud-edge composite control variable. The uniformly ultimately bounded stability of the DOB is also proved. Third, to execute the workflow-based DPC controller and evaluate the proposed cloud-edge collaborative control scheme with DOB in the real cloud environment, we design and implement a practical workflow-based cloud control experimental system based on container technology. Finally, a series of evaluations show that, the computation times are decreased by 45.19% and 74.35% for two real-time control examples, respectively, and by at most 85.10% for a high-dimension control example.

Encrypted Data-driven Predictive Cloud Control with Disturbance Observer

“…To optimize the effort of subspace identification task, which is the basis of data-driven control, [13] decomposes the identification algorithm to interconnected containerized tasks through parallel computing. A further implementation of cloud-edge cooperative DPCC [5] uses workflow-based parallel cloud control and edge compensation.…”

“…Besides, the process and measurement noise may also impact the control effect. For reducing the uncertainty and disturbance existed in HE scheme and system dynamics, we adopt the solution in [5], which uses a cloud-edge cooperative control design with a data-driven DOB to estimate the uncertainty and disturbance brought by the cloud. The estimation result obtained by data-driven DOB could be added to the control input for compensation with a proper gain.…”

“…As a kind of model-free control approach, data-driven predictive control (DPC) [4] directly computes control sequences based on the input-output data of the system, which avoids the process of system modeling. Therefore, the combination of CCSs and DPC, i.e., data-driven predictive cloud control (DPCC) [5]- [7], takes advantage of data storage and computation in the cloud, as well as the model-free manner in control of complex systems, becoming a potential candidate in CCSs.…”

“…Hence, disturbance observer (DOB) [5], [9], [10] is used to guarantee the control accuracy under the uncertainty, including system noise and errors induced by HE scheme. The function of DOB is to estimate the effects performed on a system based on an auxiliary system.…”