Periodic Reporting for period 3 - APROCS (Automated Linear Parameter-Varying Modeling and Control Synthesis for Nonlinear Complex Systems)
Periodo di rendicontazione: 2020-09-01 al 2022-02-28
Linear Parameter-Varying (LPV) systems are flexible mathematical models capable of representing Nonlinear (NL)/Time- Varying (TV) dynamical behaviors of complex physical systems (e.g. wafer scanners, car engines, chemical reactors), often encountered in engineering, via a linear structure. The LPV framework provides computationally efficient and robust approaches to synthesize digital controllers that can ensure desired operation of such systems - making it attractive to (i) high-tech mechatronic, (ii) automotive and (iii) chemical-process applications. Such a framework is important to meet with the increasing operational demands of systems in these industrial sectors and to realize future technological targets. However, recent studies have shown that, to fully exploit the potential of the LPV framework, a number of limiting factors of the underlying theory ask a for serious innovation, as currently it is not understood how to (1) automate exact and low-complexity LPV modeling of real-world applications and how to refine uncertain aspects of these models efficiently by the help of measured data, (2) incorporate control objectives directly into modeling and to develop model reduction approaches for control, and (3) how to see modeling & control synthesis as a unified, closed-loop system synthesis approach directly oriented for the underlying NL/TV system. Furthermore, due to the increasingly cyber-physical nature of applications, (4) control synthesis is needed in a plug & play fashion, where if sub-systems are modified or exchanged, then the control design and the model of the whole system are only incrementally updated. This project aims to surmount Challenges (1)-(4) by establishing an innovative revolution of the LPV framework supported by a software suite and extensive empirical studies on real-world industrial applications; with a potential to ensure a leading role of technological innovation of the EU in the high-impact industrial sectors (i)-(iii).
The work performed so far resulted on important results on four fronts:
1. A significant progress has been made on the automatic conversion of given first-principles models of engineering systems (which are often highly complex and nonlinear) to low complexity LPV models that can be readily used for analysis of the system with computationally cheap convex methods and also for controller design. This significantly lowers the required expertise form the side of control engineers to apply the LPV framework in industrial application. The resulting methods automatically minimise complexity and conservativeness of the resulting LPV models to make further analysis and design based on them more sharp. Further research aims at preservation of important structural properties in the resulting LPV models to ensure feasibility of the follow up LPV analysis or controller design and to provide physical interpretabiltiy of the LPV models.
2. By establishing a novel frequency domain understanding of LPV systems, the research paved the way for using industrial experience in controller shaping for linear time invariant systems to be used through the LPV framework for nonlinear systems, ensuring wider application of current industrial methods in controller design. Current research aims to exploit these results by frequency domain tuning of LPV controllers directly from measured application data, characterisation of performance shaping in optimal-gain controller synthesis, and incorporation of shaping objectives in model conversion and data-driven modelling.
3. As a major accomplishment of the research, a novel way has been found to give hard-core guarantees of stability and performance of synthesised LPV controllers for nonlinear plants and understand the boundaries of their application. As a consequence it became possible to use "linear" control design and the connected vast industrial experience to design directly controllers for nonleianr plants with stability and performance guarantees for the first time. This also allows to have a unified concept of performance shaping for highly complex mechatronic systems with non-linear behaviour.
4. The resulting methodologies have been successfully tested on laboratory scale applications and current planning involves their industrial tryouts. Next to this, an open-source toolbox will released soon to share the archived results with the public.
1. A significant progress has been made on the automatic conversion of given first-principles models of engineering systems (which are often highly complex and nonlinear) to low complexity LPV models that can be readily used for analysis of the system with computationally cheap convex methods and also for controller design. This significantly lowers the required expertise form the side of control engineers to apply the LPV framework in industrial application. The resulting methods automatically minimise complexity and conservativeness of the resulting LPV models to make further analysis and design based on them more sharp. Further research aims at preservation of important structural properties in the resulting LPV models to ensure feasibility of the follow up LPV analysis or controller design and to provide physical interpretabiltiy of the LPV models.
2. By establishing a novel frequency domain understanding of LPV systems, the research paved the way for using industrial experience in controller shaping for linear time invariant systems to be used through the LPV framework for nonlinear systems, ensuring wider application of current industrial methods in controller design. Current research aims to exploit these results by frequency domain tuning of LPV controllers directly from measured application data, characterisation of performance shaping in optimal-gain controller synthesis, and incorporation of shaping objectives in model conversion and data-driven modelling.
3. As a major accomplishment of the research, a novel way has been found to give hard-core guarantees of stability and performance of synthesised LPV controllers for nonlinear plants and understand the boundaries of their application. As a consequence it became possible to use "linear" control design and the connected vast industrial experience to design directly controllers for nonleianr plants with stability and performance guarantees for the first time. This also allows to have a unified concept of performance shaping for highly complex mechatronic systems with non-linear behaviour.
4. The resulting methodologies have been successfully tested on laboratory scale applications and current planning involves their industrial tryouts. Next to this, an open-source toolbox will released soon to share the archived results with the public.
The project aims to revolutionize the currently available state-of- the-art solutions and achieve the long chased dream of using simple linear design tools to synthesize reliable, robust, high-performance nonlinear controllers directly for complex physical systems. This new 2.0 version of the existing LPV framework, focuses on developing an automated toolchain of methods from modeling to control synthesis with a major emphasis on the achieved controlled (i.e. in case of feedback control, the closed-loop) behaviour, i.e. a Direct Controlled System Synthesis (DCSS) with stability and performance guarantees using the lessons learned in real-world applications and addressing the current industrial needs.The research has the potential to lead to off-shelf solutions for present and future technological problems in the high-tech, automotive and process domains.