This paper involves the adoption of a multipurpose and easy-to-use Model Predictive Controller (MPC) written in Python and its interface with an industrial Programmable Logic Controller (PLC) and the related PID-based control system. As hardware, a full PLC Siemens CPU 1516F-3 PN/DP is used and operated as an intermediate layer between the MPC and the process to be controlled. The MPC solution operates as a client by reading and writing tag values to the Open Platform Communication (OPC) server. As interface software, an OPCConnector package is built in Python using the opcua library; the OPC server configuration is performed using the TIA portal. The research was conducted in collaboration with the CPCLAB, leveraging its expertise and resources in MPC development and testing. This research advances industrial automation by integrating real and simulated environments, allowing for comparative analysis and providing practical insights on integrating MPC with real PLCs. Key contributions include identifying constraints, offering optimization guidelines, and assessing simulation accuracy in relation to real-world applications.
Integrated Testing of Model Predictive Control on Industrial PLC
Di Capaci, Riccardo Bacci;Pannocchia, Gabriele
2025-01-01
Abstract
This paper involves the adoption of a multipurpose and easy-to-use Model Predictive Controller (MPC) written in Python and its interface with an industrial Programmable Logic Controller (PLC) and the related PID-based control system. As hardware, a full PLC Siemens CPU 1516F-3 PN/DP is used and operated as an intermediate layer between the MPC and the process to be controlled. The MPC solution operates as a client by reading and writing tag values to the Open Platform Communication (OPC) server. As interface software, an OPCConnector package is built in Python using the opcua library; the OPC server configuration is performed using the TIA portal. The research was conducted in collaboration with the CPCLAB, leveraging its expertise and resources in MPC development and testing. This research advances industrial automation by integrating real and simulated environments, allowing for comparative analysis and providing practical insights on integrating MPC with real PLCs. Key contributions include identifying constraints, offering optimization guidelines, and assessing simulation accuracy in relation to real-world applications.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
