Renewable Energy Sources (RES) have drawn significant attention in the past years to make the transition towards low carbon emissions. On the one hand, the intermittent nature of RES, resulting in variable power generation, hinders their high-level penetration in the power system. On the other hand, RES can aid not only to supply much more eco-friendly energy but also it allows the power system to enhance its stability by ancillary service provision. This article reviews the challenges related to the most intermittent RES utilised in Belgium, that is, wind energy and solar energy. Additionally, wind speed and solar irradiance variations, which are the cause of wind and solar intermittency, are studied. Then, recent techniques to forecast their changes, and approaches to accommodate or mitigate their impacts on the power system, are discussed. Finally, the latest statistics and future situation of RES in the Belgian power system are evaluated.

Accurate knowledge of the induction machine rotor temperature is important for both condition monitoring and motor performance. Lumped-parameter thermal networks allow to flexibly model the temperature distribution with low computational effort. However, their accuracy highly depends on the accuracy of their thermal parameters. Moreover, these parameters are determined offline and held constant during operation. In this article, a rotor temperature virtual sensing strategy is proposed, combining a lumped-parameter thermal network with stator windings temperature measurements. The latter measurement data can be easily collected and together with the thermal network serve as input to a dual Kalman filter to estimate both the rotor temperature evolutions and the thermal model parameters. The advantage of the presented approach is that the thermal model does not need exact identification for the considered motor. One can start from a thermal model identified from intrusive rotor measurement data on a comparable motor that is subsequently transferred to the motor to be monitored. This results in a fast and convenient way to estimate the rotor temperature in real-time. Experimental results on a 4-pole 5.5 kW induction machine demonstrate the validity of the presented strategy to accurately monitor the rotor temperature within an error range of approximately ±6 °C.

Solenoid valves (SV) are essential components of industrial systems and therefore widely used. As they suffer from high failure rates in the field, fault prognosis of these assets plays a major role for improving their maintenance and reliability. In this work, Bayesian convolutional neural networks are used to predict the remaining useful life (RUL) of SV, by training them on the valve's current signatures. Predictive performance is further improved upon by using salient physical features obtained from an electromechanical model as the network's training input. Results show that our designed network architecture produces well-calibrated uncertainty estimations of the RUL predictive distributions, which is an important concern in prognostic decision-making.

Despite the use of state-of-the-art thermal modeling tools in the design stage, the measured thermal behavior of prototype electrical machines can differ significantly from the modeled ones. This paper shows how a thermal model, based on the finite-element method, of an electric machine can be improved using inverse modeling techniques. In a thorough study, a forward high fidelity finite-element thermal model of a 4-kW axial flux permanent magnet (PM) machine is introduced and improved using inverse modeling techniques via noncollocated thermal sensors. Parametric model order reduction of the high fidelity finite-element thermal model based on the moment matching method is performed to make the recovery of the actual thermal parameters characterizing the thermal behavior of the axial flux PM machine tractable. Furthermore, the same reduced order model is used to identify the different power loss components in the machine. Experimental results confirm that the presented two-stage approach is capable of identifying the thermal parameters and losses with high accuracy.

This article presents an application of neural network-based Model Predictive Control (MPC) to improve the wind turbine control system’s performance in providing frequency control ancillary services to the grid. A closed-loop Hammerstein structure is used to approximate the behavior of a 5MW floating offshore wind turbine with a Permanent Magnet Synchronous Generator (PMSG). The multilayer perceptron neural networks estimate the aerodynamic behavior of the nonlinear steady-state part, and the linear AutoRegressive with Exogenous input (ARX) is applied to identify the linear time-invariant dynamic part. Using the specific structure of the Cascade Hammerstein design simplifies the online linearization at each operating point. The proposed algorithm evades the necessity of nonlinear optimization and uses quadratic programming to obtain control actions. Eventually, the proposed control design provides a fast and stable response to the grid frequency variations with optimal pitch and torque cooperation. The performance of the MPC is compared with the gain-scheduled proportional-integral (PI) controller. Results demonstrate the effectiveness of the designed control system in providing Frequency Containment Reserve (FCR) and frequency regulation in the future of power systems.

Dynamic models of mechatronic systems are abundantly used in the context of motion control and design of complex servo applications. In practice, these systems are often plagued by unknown interactions, which make the physics-based relations of the system dynamics only partially known. This article presents a neural network augmented physics (NNAP) model as a combination of physics-inspired and neural layers. The neural layers are inserted in the model to compensate for the unmodeled interactions, without requiring direct measurements of these unknown phenomena. In contrast to traditional approaches, both the neural network and physical parameters are simultaneously optimized, solely by using state and control input measurements. The methodology is applied on experimental data of a slider–crank setup, for which the state-dependent load interactions are unknown. The NNAP model proves to be a stable and accurate modeling formalism for dynamic systems that ab initio can only be partially described by physical laws. Moreover, the results show that a recurrent implementation of the NNAP model enables improved robustness and accuracy of the system state predictions, compared to its feedforward counterpart. Besides capturing the system dynamics, the NNAP model provides a means to gain new insights by extracting the neural network from the converged NNAP model. In this way, we discovered accurate representations of the unknown spring force interaction and friction phenomena acting on the slider mechanism.