Session: Turbine Modeling & Technology

Paper Number: 167650

167650 - Hardware Implementation of Floating Offshore Wind Turbine Control Software 

Abstract: 

Floating Offshore Wind Turbines (FOWTs) require advanced control strategies to optimise energy production, extend component lifespan, and reduce operational costs. However, a critical gap exists between research-orientated simulation controllers and real-world turbine hardware controllers. While open-source simulation controllers enable testing and tuning in virtual environments, they lack the safety features required for direct deployment. Conversely, turbine manufacturers provide proprietary hardware controllers with limited transparency, restricting wind farm operators' ability to optimise control parameters. This gap hinders innovation and slows down the transition from research to practical application.

This work presents a methodology for bridging this gap by implementing new control laws into a Programmable Logic Controller (PLC) using Model-Based Systems Engineering (MBSE) principles. We adapted an advanced C++ control library originally developed for the OpenFAST simulator and integrated it into a hardware environment using Docker technology. This approach ensures seamless continuity between simulation-based controller development and real-world deployment.

The software is a C++ control code initially designed for OpenFAST, a simulation tool that does not impose real-time constraints on controllers. Within OpenFAST, the control algorithms operate in a time-agnostic manner, meaning the simulation results are independent of the execution time. It has been developed for performance assessment of control strategies. The library has been adapted to real-time execution. For hardware implementation, a PLC running CodeSys executes real-time scheduled tasks while interfacing with a Docker container housing the control algorithms. The Docker container, cross-compiled for the PLC’s 32-bit architecture, utilises a Transmission Control Protocol (TCP) client-server setup for communication. This unified software-hardware approach minimises discrepancies between simulation and deployment.

The PLC executes four key tasks: (1) ensuring the control Docker container is running, (2) acquiring sensor measurements, (3) transmitting control-relevant data to the control container, and (4) receiving and forwarding actuator commands. This setup guarantees real-time execution and synchronisation of control inputs and outputs. The control development workflow is integrated with GitLab Continuous Integration (CI), enabling automatic documentation updates and version-tagged Docker container releases for streamlined deployment.

We validated this methodology using the open-source IEA 15 MW wind turbine model with the UMaine floating platform. Controller tuning aimed to maximise power production while minimising Damage Equivalent Loads (DEL) through iterative batch simulations. Real-time testing workflows were developed for both software and hardware stages to ensure the PLC's capability to execute control algorithms in real time. Three validation tests were performed:

Software Test: A Python-based replayer script emulated sensor and actuator signals using precomputed OpenFAST results. The replayer periodically transmitted measurements to the CodeSys layer, which relayed them to the Dockerised controller. Controller outputs were compared to the reference simulation to assess accuracy.

In Hardware Test: The SIL test was replicated on the PLC, with Python scripts running inside Docker containers. Execution time variations due to hardware constraints were analysed.

Hardware with communication test: The control container was deployed on the PLC, and the replayer container transmitted measurements directly to the CodeSys layer, which interfaced with the controller in real time.

Results demonstrated a high degree of consistency between pre-simulated and hardware controller outputs across all tests, validating the effectiveness of the hardware integration. Results present minor discrepancies that are under investigation to achieve perfect alignment between simulation and real-world execution.

This methodology streamlines the development and deployment of FOWT control strategies by ensuring continuity from simulation to hardware. By providing an open and flexible platform for control law implementation, it enhances the efficiency, reliability, and maintainability of floating wind turbines. In the long term, this approach could significantly accelerate innovation in FOWT control and contribute to reducing the Levelised Cost of Energy (LCOE) for offshore wind farms.

Presenting Author: Simon Puech D-ICE engineering

Presenting Author Biography: Simon Puech specializes in control systems for Floating Offshore Wind Turbines (FOWTs) and real-time decision-support solutions for maritime applications. He develops advanced control strategies, integrating C++ algorithms into industrial hardware to optimize power production and reduce structural loads. With expertise in OpenFAST, CodeSys, Docker, and real-time Linux, he bridges numerical simulation with industrial deployment. Passionate about innovation, Simon ensures offshore wind systems are efficient, scalable, and reliable for the future of sustainable energy.

Authors:

Simon Puech D-ICE engineering
Mattéo Capaldo One Tech
Domenico Di Domenico One Tech
Sofien Kerkeni D-ICE engineering