Skip to main content

Application of distributed control on smart structures

Final Report Summary - ARRAYCON (Application of distributed control on smart structures)

This project aimed at bridging the gap between industry and academia in the field of smart structures. The motivation comes mainly from the aerospace and automotive industries where the growing popularity of light composite structures introduces demand to attenuate very tightly packed weakly damped flexible modes. For a while now, in aerospace, lightweight is one of the key words. Manufacturing equipment needs to be light to guarantee quick response and reduce energy costs. In ground transportation, vehicle manufacturers also try to minimize the weight to reduce fuel consumption and carbon emissions. It is not just a matter of decreasing the weight but also preserving, or even increasing, the stiffness and dynamic characteristics of the mechanical structure. One of the major issues with lightweight structures is their extremely low material damping. To solve this problem, in actual cars, trim material is used adding additional mass. Also, to reduce low frequency noise and vibrations it is favorable to have a greater mass which conflicts with the lightweight requirement. Hence, instead of using passive means, the active damping concept is introduced, realized by adding actuators and sensors to counteract the vibrations by a controller. This active or smart structure concept is a multi-disciplinary one in which control, structural modelling, actuating and sensing, and new materials come together, covering applications in different sectors ranging from aerospace, automotive, civil and manufacturing applications to the domains of leisure sports and health-care, as well as the recently emerging Internet of Things. Scientific novelty lies in two aspects; first, introducing distributed control in the field of flexible structures and their vibrations and second, addressing the modeling and design of smart structures in a modular fashion for improved efficiency and scalability. Through a dedicated training program, potentials of spatially distributed control using arrays of actuators and sensors for smart structures are investigated in combination with aspects of system substructure modelling. Recent advances in sensors and actuators make large and dense actuator/sensor arrays a feasible and attractive concept. Challenges however arise in the modelling and integration of such a large number of sensor/actuators, hence the need for modular and efficient integration techniques. Another important ingredient of any active damping scheme is the control law. Relatively simple control laws, like direct velocity feedback, acceleration feedback, etc. have been applied successfully. By such control though, one is always restricted to controlling just a few modes by actuating at a few locations on the structure. However, in industrial structures (characterized by a high modal density) the application of such control strategies gives disappointing results. Therefore, there arises a need for other types of controllers to be incorporated into the industrial practice. Recent advances in MEMS sensors/actuators, and progress in computational power pave way to deployment of distributed control concepts, where a group of decision-making agents reach consensus over a network to address structural vibrations.
-Project objectives and a summary description of the work since its beginning-
The project objectives are achieved by a tandem between a high level academic partner, CTU in Prague, and an established industrial partner Siemens at Leuven. The primary objective of this Marie Curie EID is to provide a training program for two young researchers (ESRs) on active structures, to familiarize them with engineering tools commonly used in the industry and to bring them in contact with industrial research and development (R&D). The key focus was on providing a multi-disciplinary training in mathematical modelling, involving the structures, actuators, sensors and control designs. To develop entrepreneurial skills, the ESRs engaged in building a demonstrator that shall serve for experimental verification and facilitate dissemination of project results to broader audiences. The secondary objective of the proposed project is to explore potentials of distributed control for vibration damping and modular modelling, both applied to lightweight smart structures. In order to carefully plan the project execution, the consortium members successfully organized a kick-off meeting at the project’s beginning, at Siemens in Leuven in September 2013. There a suitable recruitment strategy was defined, training schedules and secondments planned, all in accordance with the original project proposal. The following steps were the recruitment of two ESRs; a total of 25 applicants were interested in the offered positions and from those 10 were shortlisted (based on eligibility and their CVs) for Skype/phone interviews in December 2013. The final two candidates were brought for a live interview in January 2014 and a contract was offered to both by the Siemens. Both students started the program in Spring 2014 and focused in their first year to follow the training program. The first secondment for both ESRs was at Siemens. During this time the ESRs pursued research and undertook various training. As a starting point for ESR1, the literature on plate theory and Finite Element Method (FEM) were reviewed. ESR1 mastered numerical tools Virtual lab and NX Nastran, which are of use in his continuing research. The mechanical coupling between plate and extra patches was studied. ESRs also attended specific training related to numerical optimal control and modal analysis. ESR2 performed research in state-space modelling based on the data from FEA solver, investigated widely-used vibration damping methods, e.g. velocity feedback, acceleration feedback, position feedback, integral force feedback, as well as LQG SISO control. In parallel, training related to data-driven modelling, numerical optimal control, embedded optimization, modal analysis, experimental NVH were provided for both ESRs. Additionally, ESRs attended the international conference on noise and vibration engineering ISMA2014 to get familiar with current research in the field.
Related to academic activities in September 2014, both ESRs started their secondment at CTU, mastering 3 compulsory courses: ‘Linear matrix inequalities in control’, (prof. Henrion), ‘Flexible structures control’, (dr. Schirer) ‘Cooperative control of multi-agent systems’, (dr. Hengster). During their secondment at CTU the ESRs pursued research in finite element models, PDE models and model order reductions. Also, control design was investigated and a GUI facilitating further tests was developed. Their progress was continuously tracked and steered through regular progress meetings with their supervisor and respective consultants dr. Hromcik and dr. Hengster. Regular communication regarding their progress was maintained with Siemens during that time via Skype. The ESRs further received additional training in technical documentation writing and presentation skills, both at their subsequent secondment at Siemens, as well as attended the ISMA International course on Modal Analysis and the Marie Curie Graduate School on Transferable skills both at KUL in 2015. Furthermore, the CISM-Marie Curie Graduate School on Smart Structures for Vibro-Acoustic Control in Undine, Italy was also attended, as well as the Matlab Seminar on Design Smart Systems with MATLAB and Simulink. The successful completion of the training program, (as originally proposed), is a major positive result for the project in view of the primary objectives set in the proposal. Starting from the second year larger attention was paid to the research aspects. As far as dissemination of results is concerned, two conference papers presenting early research results were submitted for the SPIE and ICEDyn conferences, the latter being accepted for presentation in June 2015. The research results subsequently grew in scope and number, being further presented as contributions to the IEEE Conference on Control Applications, part of IEEE Multi-Conference on Systems and Control in 2016, the 2016 IMAC34 - Model Validation and Uncertainty Quantification, A Conference and Exposition on Structural Dynamics as well as the 2017 IEEE Conference on Control Technology and Applications, to mention just a few. The developed mature form of these results is published as regular papers in top international journals; Smart Materials and Structure and IEEE Transactions on Control Systems Technology, both in 2017. The summary measure to evaluate the success of the research work carried out during the period is taken from the total number of publications made by the ESRs. In total, this work resulted in two accepted regular journal papers, twelve conference papers, five of which in international peer-reviewed conferences.
Regarding the academic aspects, a joint PhD agreement between CTU and KUL was reached, agreeing joint evaluation. Following the completion of compulsory courses and minimum credit requirements during the first two years, both ESRs passed their state exams at CTU and successfully concluded their 36 month presentations at KUL, which are precursor to the final dissertation defense. Dissertations were delivered to CTU by the project end date. To facilitate the unimpeded progress all project partners remained committed to see the project through extending both ESRs contracts, (Siemens and KUL). From a management perspective the ESRs progress was continuously tracked and steered through regular meetings with their supervisor and respective consultants while regular communication was exchanged among partners. Periodic management meetings also took place to allow for a close and dedicated assessment of project and ESRs progress; in April 2014, October 2014, June 2105, November 2016 and August 2017. Also in June 2015 the project Midterm review meeting took place at Siemens, offering an opportunity to present the progress to the EC evaluators and discuss the next steps. The project closing meeting took place in Prague in August 2017 where the final aspects were discussed and all project-related activities were wrapped-up, with a general opinion among the consortium members of a successful execution of a European Industrial Doctorate (EID) project.
--The expected final results and their potential impact and use--
The primary goal of the ARRAYCON EID project was to enhanced the ESRs capacity to independently conduct research and to provide them with capability to work and communicate across disciplines and sectors. The very essence of the EID programme is its strong inclination towards industry while maintaining high and stringent academic requirements on research work. The project’s output are two research engineers with very strong foundation in rigorous mathematical techniques and yet with heavy exposition to industrial R&D. During the training programme care was taken that a high scientific level is attained, while at the same time contacts with the industry, industrial processes, software tools used in industrial R&D environments, and contacts with customers were continuously fostered. In addition, special attention was paid to dissemination of results to academic, more application oriented and broader commercial audiences. Participation in the training programme perfectly positioned the ESRs professionally to further promote the developed methodologies in industry and enabled them to clearly show an added value for their future employers. Pursuing an academic career not only allowed exploiting synergies of complementary methods, but it also steered the ESRs’ research in a direction towards current industrial problems and community in the large.
Another important outcome of this project is a network in industry as well as academia that the ESRs built. This emerging network is directly related to the activities in ARRAYCON, although not limited solely to those, involving also established informal contacts. Immediate outcomes for individual researchers can be summarized as follows: solid interdisciplinary training in a multi-sectorial environment, training in state-of-the-art methodologies in academia and state-of the-use in industry, personal relationships in academic as well as in industrial networks, multi-cultural experience, stimulated creativity and entrepreneurial mind-set. From an industrial perspective the project can generate reputation for the company given the number of publication in the field of active structural damping, moreover it can build strong arguments for customers in terms of physical demonstrators and advanced simulations to display the developed capabilities, that could be used in future consultancy services by Siemens to third parties industrial customers; especially in the automotive, aerospace and heavy machinery industrial sectors.
The project website is another information source for wider public to access the project developments (https://dce.fel.cvut.cz/en/research/arraycon). Moreover, social media outlets such as Research Gate have also proven to be a great venue for dissemination of project results (https://www.researchgate.net/project/ARRAYCON-application-of-distributed-control-on-smart-structures). Furthermore, ARRAYCON was duly presented on all events organized as a part of a general outreach effort, e.g. during the company open days at Siemens, high school student competition Robosoute’z at the CTU, https://robosoutez.fel.cvut.cz/ and has been regularly included in the yearly leaflet published by the Control Engineering Department of the CTU.