Guillaume Crevecoeur

Multi-color magnetic nanoparticle imaging using magnetorelaxometry

Magnetorelaxometry (MRX) is a well-known measurement technique which allows the retrieval of magnetic nanoparticle (MNP) characteristics such as size distribution and clustering behavior. This technique also enables the non-invasive reconstruction of the spatial MNP distribution by solving an inverse problem, referred to as MRX imaging. Although MRX allows the imaging of a broad range of MNP types, little research has been done on imaging different MNP types simultaneously. Biomedical applications can benefit significantly from a measurement technique that allows the separation of the resulting measurement signal into its components originating from different MNP types. In this paper, we present a theoretical procedure and experimental validation to show the feasibility of MRX imaging in reconstructing multiple MNP types simultaneously. Because each particle type has its own characteristic MRX signal, it is possible to take this a priori information into account while solving the inverse problem. This way each particle types signal can be separated and its spatial distribution reconstructed. By assigning a unique color code and intensity to each particle types signal, an image can be obtained in which each spatial distribution is depicted in the resulting color and with the intensity measuring the amount of particles of that type, hence the name multi-color MNP imaging. The theoretical procedure is validated by reconstructing six phantoms, with different spatial arrangements of multiple MNP types, using MRX imaging. It is observed that MRX imaging easily allows up to four particle types to be separated simultaneously, meaning their quantitative spatial distributions can be obtained.

Experimental ex-vivo validation of PMMA-based bone cements loaded with magnetic nanoparticles enabling hyperthermia of metastatic bone tumors

Percutaneous vertebroplasty comprises the injection of Polymethylmethacrylate (PMMA) bone cement into vertebrae and can be used for the treatment of compression fractures of vertebrae. Metastatic bone tumors can cause such compression fractures but are not treated when injecting PMMA-based bone cement. Hyperthermia of tumors can on the other hand be attained by placing magnetic nanoparticles (MNPs) in an alternating magnetic field (AMF). Loading the PMMA-based bone cement with MNPs could both serve vertebra stabilization and metastatic bone tumor hyperthermia when subjecting this PMMA-MNP to an AMF. A dedicated pancake coil is designed with a self-inductance of 10 μH in series with a capacitance of 0.1 μF that acts as resonant inductor-capacitor circuit to generate the AMF. The thermal rise is appraised in beef vertebra placed at 10 cm from the AMF generating circuit using optical temperatures sensors, i.e. in the center of the PMMA-MNP bone cement, which is located in the vicinity of metastatic bone tumo...

Data-Driven Optimal Power Flow Management in an Electric Dual-Drive Topology for Vehicle Electrification

The limited operating range on a single charge can be seen as an important detriment to contemporary vehicular technology, necessitating regular charging of the battery pack. Due to the high load variability during driving, incorporating two different electric motors in the drive can provide significant improvements in terms of energy consumption. A data-driven approach towards optimal power flow management in such configuration is proposed. Computationally expensive dynamic models are translated into an equivalent power flow-based representation, taking into account peak start-up losses. Optimal synchronization of both machines is then assessed over a given drive cycle, providing an optimal actuation policy for all embodied subsystems. Modifications to a standard dynamic programming formulation are introduced, reducing the computation time by a factor 125. The dual-drive topology furthermore offers the capability of cutting down energy consumption by 19.9%. Notable range extensions can thus be achieved by intelligently formulating and tackling the power flow management problem in a dual-drive topology.

Adaptive state space representations enabling reliable and robust decision-making in asynchronous drives for mechatronic applications

In contemporary mechatronic applications decision-making is often based on information about the underlying model governing the dynamical evolution, in order to ensure optimal operation with respect to a prioritized objective. Modeling errors stemming from parameter uncertainty or varying operational conditions result in inevitable deviations from the theoretical estimate and consequently in suboptimal operation. Intelligent systems need to be equipped with inherent means to compensate for these a priori unknown discrepancies, hereby guaranteeing a robust operation in uncertain environments. In this manuscript, advanced filtering techniques are applied to assess both an optimal model representation and state estimates. An appropriate interconnection between both model and state estimation is determined. The proposed methodology is demonstrated for an electric drive, embodying a DC-source, a voltage source inverter (VSI) and an asynchronous machine, as the presence of discrete switching sequences and physical constraints introduces additional challenges. Results prove that the error on the state estimates can be improved by 92.7–97.2%, outperforming the classical estimation techniques, while the relative model mismatch is scaled down to 0.03%, even in highly demanding scenarios. The introduced strategy thus enables high-fidelity virtual sensing and reliable decision-making procedures for advanced asynchronous drives when modeling errors can be anticipated.

Polynomial chaos explicit solution of the optimal control problem in Model Predictive Control

A difficulty still hindering the widespread application of Model Predictive Control (MPC) methodologies, remains the computational burden that is related to solving the associated Optimal Control (OC) problem for every control period. In contrast to numerous approximation techniques that pursue acceleration of the online optimization procedure, relatively few work has been devoted towards shifting the optimization effort to a precomputational phase, especially for nonlinear system dynamics. Recently, interest revived in the theory of general Polynomial Chaos (gPC) in order to appraise the influence of variable parameters on dynamic system behaviour and proved to yield reliable results. This article establishes an explicit solution of the multi-parametric Nonlinear Problem (mp-NLP) based on the theoretical framework of gPC, which enabled a polynomial approximated nonlinear feedback law formulation. This resulted in real-time computations allowing for real-time MPC, with corresponding control frequencies up to 2 kHz,

Convex Mapping Formulations Enabling Optimal Power Split and Design of the Electric Drivetrain in All-Electric Vehicles

All-electric drivetrains have been identified as a promising alternative to contemporary hybrid vehicle technology. Extending their operational range is key and can be achieved by means of design procedures based on high-fidelity models capturing the dynamical behavior of the electric drivetrain. This paper proposes a dedicated power split embodying a dual electric drive and a model-based strategy to design the drivetrain. Advancements are required in model-based design that can cope with the complexity of the computationally expensive and high-dimensional parametric design problems. We propose a nested optimization approach wherein parameter exploration is attained using an evolutionary algorithm and the optimal power flows are determined by abstracting the high-fidelity behavioral models into appropriate convex loss mappings. This allows for an accelerated design procedure based on convex optimization without compromising accuracy. We size an electric drivetrain for maximal range extension, consisting of a battery stack, buck–boost converter, inverter and mechanically coupled induction motors subjected to variable load conditions. A tractable convex formulation is obtained and optimization time is reduced by 99.3% compared to the traditional approach without convexification. Optimal control of the incorporated power split increases the operational range by 0.7% compared to the isolated operation of a single motor. The proposed methodology thus paves the way for extensive designs of drivetrains and complex mechatronic systems in a general context.