Model-Based Control of Mechatronic Systems - Bridging Between Advanced Methods and Industrial Applications

Nowadays, every product manufacturing industry is continuously driven to develop innovative and sustainable products that differentiate through outstanding performance, customer satisfaction and appealing brand values. Intelligent, mechatronic systems are instrumental to accommodate this increasing demand for product quality, flexibility and customization. Since its intelligent nature is essentially the raison d’ëtre of a mechatronic system, it is crucial to have a seamless integration between the development cycles of the multi-physics hardware and the embedded controls software. In the current industrial practice, this connection is mainly realized by sharing closed, black-box simulation models of the physical design and the control software between departments, enabling the virtual verification of functional requirements long before a first physical prototype is built. This doctoral research demonstrates how model-based methodologies for control and estimation can extend the usage of system models beyond this simulation-based verification of requirements. On the basis of four industrially relevant applications, this thesis analyzes the capabilities of a model-based paradigm for control and estimation to accelerate controls development in the early stage of the design process and to expand the usage of system models into the product’s operational phase. In a first contribution, this dissertation studies how Model Predictive Control (MPC) techniques can accelerate the development process of a hydraulic active suspension system. A second contribution comprises an innovative estimation algorithm for the inertial properties of a road vehicle, which can be employed in an adaptive model-based chassis control system. Combining a control architecture based on Internal Model Control (IMC) with an identification procedure based on Experimental Modal Analysis, a third contribution proposes a vibration control system for a multi-axis hydraulic shaker platform, that is used in the aerospace industry for testing satellites. A final contribution demonstrates the value of MPC methodologies in the control architecture of autonomously driving road vehicles. By positioning model-based control in the industrial context of Model-based Systems Engineering (MBSE), this doctoral research has created a clear interest from several industrial partners in model-based control and the underlying state-of-the-art methodologies. This transfer of state-of-the-art academic methodologies into state-of-the-use industry practice is fully in line with the objectives of the IWT-Baekeland mandate that has supported this doctoral research.1. Introduction 2. Model Predictive Control 3. Model-based control of a hydraulic active suspension 4. Estimation of vehicle inertial parameters for chassis control systems 5. Model-based control of a multi-axis shaker platform 6. Autonomous driving with balanced performance attributes 7. Conclusions and suggestions for future researchnrpages: 184status: publishe