Automatic generation of control software for mechatronic systems

Project members:prof.dr. R. Babuška (Robert), dr. G.A.D. Lopes (Gabriel), B. De Schutter (Bart)
Keywords:Robotics and mechatronics, Intelligent control, Model-based control, Learning and adaptive control, Discrete-event systems
Sponsored by:SenterNovem

Design of a mechatronic system consists of many domains: mechanics, electronics, embedded hardware and software, and control. The design begins with the functional description of the mechatronic system. In traditional design approach the design continues with the mechanical design followed by the electronics design, and then the embedded hardware and software co-design is completed. Finally the controller which will maintain the global behavior of the mechatronic system is designed and the control code is embedded into the system whose body is almost completed before control design phase.

What we want to achieve is a new design methodology for the design of mechatronic systems where the four domains: mechanics, electronics, embedded hardware and software, and control designs are initiated right after the functional description of the system specifications (see Figure 1). In this approach, we will need an integration framework to bring these domains together so that all these four domains will interact with each other during the whole design process of the mechatronic system.

Figure 1: Parallel design approach (adopted from: National Instruments).

We will mainly focus on the integration of the control design and the electronics design including the embedded hardware and software design, basically the control board on which the generated control software will be validated to meet the system specifications, considering the real-time embedded systems programming issues.

The first thing to be done is to find a way to design controllers using the qualitative behaviors of the plant to be controlled and the performance requirements for the closed loop control system. We will investigate which qualitative parameters related to the plant and the performance requirements are necessary to design the controller.

After having the qualitative behaviors, some quantitative parameters will be obtained considering the timing issues for real-time implementation of the control algorithm on the actual hardware which is the control board. There will be interaction with the control design and electronics design domains in this sense of having the flexibility of choosing the microprocessor's processing speed to meet the sampling time constraint, and choosing the A/D, D/A converters to meet the resolution needs, choosing the ROM, RAM size such that all the control code and the data to be processed simultaneously can be fit in.

© Copyright Delft Center for Systems and Control, Delft University of Technology, 2017.