If you would like more information about the profile please click
. You will be directed to a new page where you will need
to provide some information about yourself/your business cooperation needs and submit
your expression of interest in the profile . A Network Partner will then contact
you to discuss your interest in greater detail.
A German university developed a procedure that improves the computing power of controllers in mechatronic systems. The new procedure upgrades the model predictive control (MPC) by reducing the computational effort significantly. The university offers license agreements to industrial partners.
For a fast and precise control of mechatronic systems usually non-linear PID (proportional–integral–derivative controller) controllers are used. The more nonlinearities a system contains, the more complex the control task becomes. A more advanced control concept is model predictive control (MPC), which resolves a number of disadvantages of the classic control concepts. These include the explicit adherence to state and input constraints and the intuitive controller design. The MPC also involves solving an optimal control problem that is computationally expensive. Since the solving must be done between two sampling steps, processes with high dynamics can push the process to its limits.
A German university developed an enhancement of MCP, a model predictive trajectory set control (MPTSC). It reduces the computational effort significantly, without losing the model predictive character. Essentially, MPTSC is based on the discretization of the input domain and can thereby forego gradient-based optimization algorithms in particular. Therefore, any cost terms and prediction models can be used. Adaptation of the discretization ensures a high approximation quality of the classical MPC.
The control system can be used in all machines and systems, especially if the control problem is somewhat more complicated and fast control is required (mechatronic systems, robots, etc.). For simple applications, classical control methods are sufficient.
The university offers mechanical engineering companies a license agreement. The partner companies could come, for example, from the fields of pneumatics or robotics. They should be interested in implementing the technology.
Conventional model predictive regulations are computationally intensive. The calculation process and the movement of the machines therefore often do not conform. The invention presented here improves the computing power of the controllers and is real-time capable for high sampling frequencies. MPTSC can be used for a wide range of mechatronic systems as there are no restrictions on the cost terms and prediction model. The adjustment of the control is intuitive, so that it can be carried out by the user himself.
Proof of function
A German and and an international patent have been applied for.
The university is offering a license agreement to engineering companies working in sectors, such as, pneumatics or robotics.
First and foremost, a license agreement is sought. Within the cooperation it is possible to adapt the invention to the requirements of the licensee (i.e. specifically to the control engineering problem). However, it is also possible, especially if the licensee already has experience in the area of model predictive regulations, for the licensee to take over the complete implementation himself.