Synthesizing controllers

Hi, my name’s Ferdie Reijnen and I’m in the Control Systems Technology group at the Department of Mechanical Engineering where my research focuses on developing methods to synthesize supervisory controllers for navigation locks and movable bridges. Synthesizing is a formal method that calculates the controller from the model of the system and the model of the control specifications. The advantage is that your controller adheres to the requirements imposed by design.

Proof of concept

The main challenge lies in how this method can be used to generate controllers that work for a family of locks, instead of a single lock. Furthermore, the complexity and the number of states that the locks and bridges may be in demands more efficient and powerful synthesizing algorithms. So we started by doing model-based case studies on existing locks and bridges, to become familiar with modeling these systems and their requirements. As a next step we want to implement the controllers to see how

they behave in the real systems. So far we have been able to synthesize controllers for a lock and a bridge, using simulation to validate the correctness of the synthesized controller. This delivered a proof of concept for the applicability of the method to the design of such controllers. 

Lower cost, faster development

So, for anyone working on supervisory control or logic coordination control for large complex systems, the synthesis method offers a formal way to calculate your controller, making it correct by construct, instead of manually coding the controller, which can be very error-prone. The use of formal methods will enhance the safety and correctness of the control systems of locks and bridges in The Netherlands. Besides, by using formal models and simulations, errors can be found early in the design process, reducing costs and shortening development times.