The project aims at machine, vehicle and component manufacturers that build systems operating in highly variable contexts as well as the engineering software and service providers assisting them.
In this project, we want to help them increase performance by taking the contexts into account in their controllers. ‘Contexts’ refers to the task performed, the operating conditions, the loads or duty cycle, … all which change regularly, but not at every sample. Controllers are currently typically tuned conservatively, so that the performance is acceptable under all conditions but it is never optimal. It is therefore desirable to make controllers consider (be aware of and make choices based on) the current context, preferably without increasing controller complexity.
The project will deliver tools for automated context identification and grouping considering the control performance, as well as tools for development, tuning, and deployment of the resulting context aware controllers. As a result, the conservative controllers currently used in industry can be upgraded, leading to increased system performance such as reduced energy consumption, increased production capacity, comfort and accuracy. This will all be achieved faster and with lower operating costs by avoiding the need to tune the controllers manually.
The solution will rely on existing sensors, combined with data and models and will not need additional monitoring hardware or sensors. Also, since the context will be detected online and the parameters of the existing controller adapted accordingly, the complexity of the control algorithms will remain the same, meaning that the existing control platforms can be reused.
ConACon is an ICON-project (Interdisciplinair Coöperatief Onderzoek – Interdisciplinary Cooperative Research). We are looking for companies to join the consortium and work with us on the valorisation of the project.