Intelligent and Active Suspension Controller Development
Maximum driving comfort through the best possible vibration decoupling and optimum design of the dynamic wheel loads; this is what developers of active suspension systems are striving for. However, the development of active control strategies is a great challenge due to conflicting objectives, physical limitations and the complexity of the modelling.
The use of artificial intelligence (AI) in this area enables enormous advantages. Intelligent controllers synthesize the complex formulations and performance requirements. The controllers are trained in such a way that they not only learn how to improve driving comfort, but also achieve a high level of driving safety.
Extensive expertise
Our team of experts in the fields of vehicle dynamics and AI development has extensive experience in controller development. Our broad expertise in the development of mathematical and simulation models, ranging from simple vehicle models to complex multi-body systems, is a key factor. In this context, we investigate the performance with various tools such as Matlab/Simulink and ADAMS. Thanks to our test infrastructure, our simulation models can be validated quickly and realistically.
Our deep understanding in these areas enables us to evaluate the AI-based deep reinforcement learning algorithms that are created daily in the community. We identify the few, which are applicable for use in controller development and apply them in the vehicle context.
In this process, the strengths and weaknesses of the controller are analysed, taking into account factors such as robustness, real-time implementation and learning time. Subsequently, the performance of the controller is validated by using the controller in both the simulation and test bench environments, and then testing and further developing the controller in the overall vehicle context.
Do you have further questions or are you interested in a cooperation? Feel free to contact us at any time, we look forward to receiving your message.