One of the major problems to design and implement a control/supervision system for a process lies in the need to establish an ad-hoc system for each process installation. On the other side, an open challenge related to the deployment of Intelligent Decision Support Systems (IDSSs) is the interoperability of the different methods used, in order to allow interaction and reuse of different data mining methods and the use of methods based on a model or an expert. Thus, this paper proposes the use of visual workflows, to enable the automation of the design task and the implementation of Intelligent Process Control Systems (IPCSs). The framework will allow the user to specify the design and control of a concrete process as well as the required data-driven and expert models using a graphical workflow environment. The framework is based on a three-layer architecture: first, a comprehensive data science flow description layer (dataflow layer) to produce/discover data-driven models from process data; second, a flowchart of the different components of the process (process-design flow layer) to obtain a simulation model from the design. Finally, the on-line IPCS (process control workflow layer), where the different data-driven models, expert-based models and intelligent reasoning methods interoperate to control and supervise the process. Thus, the resulting system can automatically generate both simulation models of the process and programming code to control and supervise the process, using workflows designed for each particular installation. The case study is focused on the supervision of a Wastewater Treatment Plant (WWTP) located in the Barcelona region.

The authors acknowledge the partial support of this work by the Industrial Doctorate Programme (2017-DI-006) and the Research Consolidated Groups/Centres Grant (2017 SGR 574) from the Catalan Agency of University and Research Grants Management (AGAUR), from Catalan Government.

Peer Reviewed