D. Freitas

Wheel Slip Control of EVs Based on Sliding Mode Technique With Conditional Integrators

This paper presents a new control system, based on field programmable gate array technology, targeting the powertrain control of multi-motor electric vehicles (EVs). The control chip builds around a reusable intellectual property core named propulsion control unit, which features motor control functions with field-orientation methods, and energy loss minimization of induction motors. In order to improve the EV safety, the control system was extended with a wheel slip controller based on the sliding mode framework. The robustness to parametric and modeling uncertainties is the main attraction in this design, thanks to a simple connection that was found between the driving torque request and the model uncertainty. To overcome the chattering issue, which arrives from the discontinuous nature of the sliding control, the conditional integrator approach was employed, enabling a smooth transition to a Proportional+Integral control law, with anti-windup, when the tire slip is close to the setpoint. The controller asymptotic stability and robustness was analytically investigated through the Lyapunov method. Experimental results, obtained with a multi-motor EV prototype under low grip conditions, demonstrate a good slip regulation and robustness to disturbances.

Torque blending and wheel slip control in EVs with in-wheel motors

Among the many opportunities offered by electric vehicles (EVs), the design of power trains based on in-wheel electric motors represents, from the vehicle dynamics point of view, a very attractive prospect, mainly due to the torque-vectoring capabilities. However, this distributed propulsion also poses some practical challenges, owing to the constraints arising from motor installation in a confined space, to the increased unsprung mass weight and to the integration of the electric motor with the friction brakes. This last issue is the main theme of this work, which, in particular, focuses on the design of the anti-lock braking system (ABS). The proposed structure for the ABS is composed of a tyre slip controller, a wheel torque allocator and a braking supervisor. To address the slip regulation problem, an adaptive controller is devised, offering robustness to uncertainties in the tyre–road friction and featuring a gain-scheduling mechanism based on the vehicle velocity. Further, an optimisation framework is employed in the torque allocator to determine the optimal split between electric and friction brake torque based on energy performance metrics, actuator constraints and different actuators bandwidth. Finally, based on the EV working condition, the priorities of this allocation scheme are adapted by the braking supervisor unit. Simulation results obtained with the CarSim vehicle model, demonstrate the effectiveness of the overall approach.

Application of MRI and biomedical engineering in speech production study.

Speech production has always been a subject of interest both at the morphological and acoustic levels. This knowledge is useful for a better understanding of all the involved mechanisms and for the construction of articulatory models. Magnetic resonance imaging (MRI) is a powerful technique that allows the study of the whole vocal tract, with good soft tissue contrast and resolution, and permits the calculation of area functions towards a better understanding of this mechanism. Thus, our aim is to demonstrate the value and application of MRI in speech production study and its relationship with engineering, namely with biomedical engineering. After vocal tract contours extraction, data were processed for 3D reconstruction culminating in model construction of some of the sounds of European Portuguese. MRI provides useful morphological data about the position and shape of the different speech articulators, and the biomedical engineering computational tools for its analysis.

Enhancing the accessibility of mathematics for blind people: The AudioMath project

“How can a blind person surpass the difficulty in reading an on-line document’s mathematical expressions? Why wasn’t this completely solved yet? Is it not necessary? Is not easy?” – These questions are only the top of the iceberg of a big problem with accessibility in the Internet. This concerns technical, scientific or even simple documents presented on-line that involve mathematical expressions. Addressing these issues the authors developed the AudioMath [1] project at LSS. It can be connected to a text-to-speech engine (TTS), providing speech rendering of the W3C’s MathML [2, 3] coded mathematical expressions. The paper intends to present the project methodology as well as the results already obtained. With AudioMath we intend to increase the accessibility of, not only, e-learning websites that use MathML, but also general websites. Therefore, AudioMath is an accessibility tool that can bring great benefits for visual impaired persons, but not only.

A new online identification methodology for flux and parameters estimation of vector controlled induction motors

A new online identification methodology for estimation of the rotor flux components and the main electrical parameters of vector controlled induction motors is presented in this paper. The induction motor model is referred to the rotor reference frame for estimation of rotor flux and rotor parameters, and referred to the stator reference frame to estimate stator parameters. The stator parameters estimation is achieved by a prediction error method based on a model structure described by a linear regression that is independent of rotor speed and rotor parameters. The rotor flux components and rotor parameters are estimated by a reduced order extended Kalman filter, using a 4th-order state-space model structure where the state equation is described by matrices that are diagonal and independent of rotor speed as well as stator parameters. Both methods work in a boot-strap manner.

Optimal sizing and energy management of hybrid storage systems

This paper targets the development of hybrid energy storage systems (ESS), based on the batteries-supercapacitors blending. In particular, we will develop two methodologies for the combined sizing/energy management of hybrid ESS. The first method assumes that the division of power between the sources is performed through low/high pass filters, which allow us to evaluate, in a simple and rapid way, the trade-offs and economic gains due to the hybridization. The second approach relies on a nonlinear optimization problem, and seeks the minimization of the installation and electrical charging costs of the sources. Simulation results reveal the existence of a threshold in the EV range, from which the introduction of the supercapacitors is less beneficial in economic terms.

Using Statistical Deformable Models to Reconstruct Vocal Tract Shape from Magnetic Resonance Images

The mechanisms involved in speech production are complex and have thus been subject to growing attention by the scientific community. It has been demonstrated that magnetic resonance imaging (MRI) is a powerful means in the understanding of the morphology of the vocal tract. Over the last few years, statistical deformable models have been successfully used to identify and characterize bones and organs in medical images and point distribution models (PDMs) have gained particular relevance. In this work, the suitability of these models has been studied to characterize and further reconstruct the shape of the vocal tract in the articulation of Portuguese European (EP) speech sounds, one of the most spoken languages worldwide, with the aid of MR images. Therefore, a PDM has been built from a set of MR images acquired during the artificially sustained articulation of 25 EP speech sounds. Following this, the capacity of this statistical model to characterize the shape deformation of the vocal tract during the production of sounds was analysed. Next, the model was used to reconstruct five EP oral vowels and the EP fricative consonants. As far as a study on speech production is concerned, this study is considered to be the first approach to characterize and reconstruct the vocal tract shape from MR images by using PDMs. In addition, the findings achieved permit one to conclude that this modelling technique compels an enhanced understanding of the dynamic speech events involved in sustained articulations based on MRI, which are of particular interest for speech rehabilitation and simulation.

Adaptive-Robust Friction Compensation in a Hybrid Brake-by-Wire Actuator

This work focuses on the development of a pressure-loop controller for a hybrid brake-by-wire system, composed of a hydraulic link and an electro-mechanical actuator. Towards this goal, we will start by constructing a reduced model that is capable of capturing the fundamental dynamics of the actuator, which is particularly useful for control design purposes. Motivated by the large friction disturbances that affect the system, we also investigate linear-in-the-parameter models suitable for (online) model-based friction compensation. More specifically, results from the theory of function approximation, together with optimization techniques, are explored to approximate the Stribeck friction model through a linear-in-the-parameter model. This new linear-in-the-parameter model is then employed in the design of a control law for tracking the braking pressure of the hybrid brake-by-wire. The main features of this controller are the robustness to parametric uncertainties, thanks to the inclusion of a switching-σ adaptive mechanism, and the attenuation of non-parametric disturbances with a continuous sliding mode action. The stability and robustness properties of the closed-loop system are investigated with the help of the Lyapunov method. Finally, experimental tests demonstrate the effectiveness of the proposed approach and its ability to handle disturbances.

A control allocation approach to manage multiple energy sources in EVs

This article is concerned with the design of an energy management system (EMS) for the hybridization of multiple energy sources (ESs) in electric vehicles, focusing in a particular configuration composed by batteries and supercapacitors (SCs). As a first design step, we investigated an (non-causal) optimal power allocation, targeting the minimization of the energy losses over a complete driving cycle. Albeit the solution obtained with this formulation demands the advance knowledge of the vehicle driving cycle, it also provides a useful benchmark solution to assess the performance of causal EMSs. A more practical EMS is then derived, based on the control allocation (CA) concept. This approach, typically employed in redundant control systems, enable us to address the various objectives and constraints that appear in EMS design problem, such as the DC bus voltage regulation, SC state of charge tracking, minimization of power losses, current and state of charge limits, etc. Simulation results show the effectiveness of the proposed CA based EMS, yielding performances very close to the optimal non-causal power allocation.

Optimal Linear Parameterization for On-Line Estimation of Tire-Road Friction

Abstract In this article, we propose a linear parameterization (LP) for representing the friction in the tire-road interface, suitable for on-line identification. This model was obtained by employing function approximation techniques, which results in an optimization problem to minimize the fitting error between the LP and the nonlinear Burckhardt model. Compared with others LPs proposed in the literature, the optimal LP features a reduced number of parameters and good approximation capabilities. Moreover, the proposed model can be identified though linear identification techniques, simplifying the on-line peak friction estimation. Simulation results obtained with a vehicle simulator demonstrates the effectiveness of the proposed parameterization.