This action might not be possible to undo. Are you sure you want to continue?
By: Prof. Francis Assadian and Dr Sajjad Fekriasl Automotive Mechatronics Centre, Cranfield University
During the last two decades, there have been substantial advances in the theory and application of robust multivariable feedback control system design. The reason for a need of such robust algorithms arises from several inherent uncertainty sources such as various operational conditions, process changes, sensor noises and unmeasured exogenous disturbances. While robust control systems have been successfully employed to tackle a wide range of engineering applications, including aerospace systems, automotive industry has not received satisfactory advantages of these modern control techniques. However, it is interesting to recall that, most of the control software designs and requirements capturing in the automotive engineering domain have been adopted from the aerospace industry. One of the main reasons in this course, perhaps, is the fact that the process of automotive systems, unlike the aerospace industry, is in the state of flux, being made not "standardised" as of yet. It is turned out that there has been a continuous increase in the gap between the control theory and the practical control strategies utilised in the existing production vehicles. This gap indeed results in significant missed opportunities in addressing several fundamental functionalities such as fuel economy, emissions, driveability, unification of control architecture and integration of the Automotive Mechatronics units on-board vehicle and most importantly control software calibrations. In this article, we shall briefly discuss how we have endeavoured to bridge this gap by employing robust feedback control systems design in the Automotive Mechatronics domain. Automotive Mechatronics in Research A wide range of modern automotive products are currently designed by the integration of mechanical components and electronic hardware into one packaging unit leading towards true mechatronic solutions, such as HEV energy management systems, active chassis systems, next-generation HEVs etc (Figure 1). On the other hand, there are various
practical challenging problems of automotive systems including calibration efforts, time and cost to production, reliability and diagnostics, control system robustness and performance issues and hardware constraints. Existing methodologies, however, are no longer liable to meet such requirements for increasingly complex new vehicle and therefore a variety of innovative mechatronics-based design methodologies are desperately required. Mechatronics applications are the best and ultimate solutions to the challenging automotive industries requirements as they offer versatile potentials with regards to functionality, cost, space requirements and quality. The key objective of the Automotive Mechatronics is thus to pursue, in both research and development, a harmonised approach to the design of mechatronic systems for the automotive systems.
Figure 1. Automotive Mechatronics Applications Automotive Mechatronics Centre at Cranfield 1 University , led by Prof. Francis Assadian, contributes to advance mechatronics applications with wide range of innovative and cost-effective research and development technologies. To this end, we aim to build strong links with industrial partners in our research collaborations. Examples which have been successfully investigated in our centre include complex automotive applications such as advanced control systems for torque management and driveability improvement of hybrid electric vehicles (HEVs) which is the main topic of this article. The Technology In our EPSRC funded research project on multivariable controls, the aim was to design
For more information on Automotive Mechatronics Centre at Cranﬁeld University, please see: http://www.cranﬁeld.ac.uk/soe/departments/automotive/automechatronics/index.jsp
an electric Crankshaft Integrated Motor Controller (CIMG). 1 For more information on Automotive Mechatronics Centre at Cranﬁeld University. sufficiently realistic clutch models. whilst the low-frequency engine output torque responses are actually delivered by the engine itself but. as a result. at highfrequencies modes (rapid torque requests). Here. The Structure of the HEV for the application of Torque Management. This complexity results in many challenges for both automotive OEMs and suppliers. Our project was aimed to deliver a setup with reduced fuel consumption and CO2 emissions. an ICE speed controller. pumping and friction torques). using so-called mixed-mu synthesis. the developed robust multivariable controller fully achieves our requirement from the HEV driveability viewpoint by delivering sufficiently fast total torque response. electric motor effectively compensates the engine output torque lags. Recall that due to different bandwidths at two ICE/CIMG control channels.uk/soe/departments/automotive/automechatronics/index. HIL architecture implemented based on the dSPACE platforms. As illustrated. the controller has made the CIMG to help bring the total torque rapidly to the requested torque. This was motivated due to the fact that existing hybrid powertrain control methods are based on off-line (sub-optimal) algorithms. Figure 2.ac. more expertise in the mechatronics area particularly at the OEM side. It also turns out that intensive calibration efforts are required. multivariable torque controller and their associated bumpless anti-windup controls that were designed and tested in HIL (Figure 3).jsp . use of advanced robust control techniques and more refined integration approaches are worth mentioning.cranﬁeld. This is indeed a challenging highly-coupled multivariable control problem that single PID loops cannot cope with. We have developed a Simulink package for the HEV energy management application (Figure 2). model uncertainties are ignored and torque estimation errors in the feedback are not considered. please see: http://www. drastically reduce the need for manual drivability calibration efforts. referred to as ‘torque filling’. the HEV torque management application is a complex frequency-weighted problem which can be solved by the robust MIMO designs and. a better definition of roles of suppliers and automotive OEMs. In words. This includes an empirical diesel engine model. in which driveability is an afterthought. we intent to show only some typical results of the multivariable robust control design. management of a Hybrid Electric Vehicle (HEV) (Figure 4). Figure 3. applied successfully to the case study of torque Figure 4. adaptive torque estimation algorithms (for both ICE and CIMG output torques). more efficient development processes. over specific drive cycles including NEDC. Among these challenges. by meeting the increasingly stringent emissions standards and enhanced reliability and diagnostics. together with saturated actuators and torque loss models (such as ICE ancillary. Robust multivariable torque control results designed and tested at Cranfield University.and develop a pragmatic advanced modelbased (dynamical) controller to the torque management of HEV that is a challenging application due to complexity of the HEV dynamics.
It is our hope that the results of this research work could motivate automotive industry to pay sufficient attention towards the advanced robust control designs and their potential capabilities in addressing a broad range of complex applications in the automotive control systems by improving vehicle performance and driveability while still continuing to reduce costs and meeting new emission standards. The Automotive Mechatronics Centre. 3) To address the knowledge gap in automotive mechatronics through short courses as well as establishment of an MSc program in Automotive Mechatronics in 2013. aimed at both transient and steadystate. In order to disseminate our latest achievements. Cranfield University. Top: ICE torque. We also recently held a workshop entitled "Applied Robust Control: From Theory to Automotive Industrial Applications" at the 50th IEEE Conference on Decision and Control and European Control Conference. 2) To carry out long term fundamental research in the automotive green technology area.Yellow: Requested torque. through governmental support and long term industrial projects. vehicle Dynamics and integrated chassis Control.uk/soe/departments/automotive/automechatronics/index. designed for both speed-control and torque-control modes.ac. to mechatronics modelling. from process and methodology. Besides. Cranfield University Tel: +44 (0) 1234 754657 Email: f. Figure 5. The Automotive Mechatronics Centre at Cranfield University was established in 2009 to help address some of the applied advanced control issues discussed above. Bottom: Total Torque). To meet increasingly new challenges in the pioneered automotive research. advanced automotive control and energy management strategies. we are now in the process of testing these torque estimations and are planning to write an invention disclosure in due course. we have co-authored a number of publications to overview some of the theoretical and practical aspects of the applied robust control. Francis Assadian Head of Department of Automotive Engineering & Director of Automotive Mechatronics Centre School of Engineering. Furthermore. please see: http://www. both academia and industry. have been very satisfactory and promising.ac.cranﬁeld. It has been our intention to highlight that the missed opportunities within the automotive applications could have been addressed by utilising advanced control techniques. Middle: CIMG Torque. Contact: Prof.jsp . To present operational and practical issues of automotive mechatronics. design and development. The Summary In collaboration with our industrial partners. Magenta: Estimated torque. our current research successes are in the design and development of robust controls for a challenging application of Hybrid Electric Vehicle torque management. Cranfield Automotive Mechatronics Centre Cranfield University is one of the leading institutes of higher education in the UK in the fields of design and engineering of automotive technology. our research activities are mainly focused on vehicle electrifications (novel electrical and control architectures). Orlando. USA (December 2011) at the above field which has received very promising feedback and attention from wider researchers and engineers.assadian@cranfield. Florida. the results of our adaptive engine brake torque estimations. our overarching goals of this centre are listed as follows: 1) To address the immediate needs and gap in the mechatronics and advanced control system design in a coordinated pragmatic approach through short term projects with industry.uk 1 For more information on Automotive Mechatronics Centre at Cranﬁeld University.