This paper presents a simulation model to estimate the instantaneous power consumption of a mobile robot by taking into account both its mechanical and computational components. The simulation model is adaptable to be tuned based on the level of accuracy needed for estimating the power consumption for the robot and the simulation time penalty. This makes a multi-fidelity power estimation tool for the robot with the capability to run-time changing the fidelity according to environmental conditions and internal computational capabilities. Such multi-fidelity energy prediction is suitable for run-time predictive decision making in a wide range of usages such as training process in model-based Reinforcement Learning (RL) as well as decision making in Model Predictive Control (MPC). The experimental results show that the simulation accurately estimates energy consumption at different fidelity levels. Higher-fidelity models closely match real-world measurements, while lower-fidelity models trade some accuracy for faster predictions. Therefore, higher estimation precision comes at the cost of increased computation.

Digital Twins (DTs) are increasingly applied for optimization of operations in logistics, healthcare, smart cities, and beyond. However, implementing high-fidelity DTs remains challenging in computationally intensive domains such as Computational Fluid Dynamics (CFD). While simplified models can facilitate real-time operation, they often lack physical fidelity. This article presents an open-source scalable software framework along with a case study of CFD-based digital twining on stormwater management. The presented framework enables online execution of CFD-based models by containerizing and integrating them into OpenShift platform, providing a two-way communication channel for simulation parameters and results. The framework is capable of dynamically scaling computing resources to run computationally-intensive CFD-models. In the case study, we present a novel CFD simulation model of a bioretention cell intended to reduce runoff volumes of urban stormwater. The simulation model, implemented in OpenFOAM, is then integrated into the presented software framework to create the DT. The framework source code, simulation model and the DT are made publicly available to promote future research.

Energy management of mechanical and cyber parts in mobile robots consists of two processes operating concurrently at runtime. Both the two processes can significantly improve the robots' battery lifetime and further extend mission time. In each process, information on energy consumption of one of the two parts is captured and analyzed to manipulate various mechanical/computational actuators in a robot, such as motor speed and CPU voltage/frequency. In this article, we show that considering management of mechanical and computational segments separately does not necessarily result in an energy-optimal solution due to their co-dependence; as a consequence, a runtime co-management scheme is required. We propose a proactive energy optimization methodology in which dynamically trained internal models are utilized to predict the future energy consumption for the mechanical and computational parts of a mobile robot, and based on that, the optimal mechanical speed and CPU voltage/frequency are determined at runtime. The experimental results on a ground wheeled robot show up to 36.34% reduction in the overall energy consumption compared to the state-of-the-art methods.

Optimal control of electric vehicle (EV) batteries for maximal energy efficiency, safety and lifespan requires that the Battery Management System (BMS) has accurate real-time information on both the battery State-of-Charge (SoC) and its dynamics, i.e. long-term and short-term energy supply capacity, at all times. However, these quantities cannot be measured directly from the battery, and, in practice, only SoC estimation is typically carried out. In this article, we propose a novel parametric algebraic voltage model coupled to the well-known Manwell-McGowan dynamic Kinetic Battery Model (KiBaM), which is able to predict both battery SoC dynamics and its electrical response. Numerical simulations, based on laboratory measurements, are presented for prismatic Lithium-Titanate Oxide (LTO) battery cells. Such cells are prime candidates for modern heavy offroad EV applications.

Finding the optimal distribution of exerted effort by an athlete in competitive sports has been widely investigated in the fields of sport science, applied mathematics and optimal control. In this article, we propose a reinforcement learning-based solution to the optimal control problem in the running race application. Well-known mathematical model of Keller is used for numerically simulating the dynamics in runner's energy storage and motion. A feed-forward neural network is employed as the probabilistic controller model in continuous action space which transforms the current state (position, velocity and available energy) of the runner to the predicted optimal propulsive force that the runner should apply in the next time step. A logarithmic barrier reward function is designed to evaluate performance of simulated races as a continuous smooth function of runner's position and time. The neural network parameters, then, are identified by maximizing the expected reward using on-policy actor-critic policy-gradient RL algorithm. We trained the controller model for three race lengths: 400, 1500 and 10000 meters and found the force and velocity profiles that produce a near-optimal solution for the runner's problem. Results conform with Keller's theoretical findings with relative percent error of 0.59% and are comparable to real world records with relative percent error of 2.38%, while the same error for Keller's findings is 2.82%.

Understanding battery capacity degradation is instrumental for designing modern electric vehicles. In this paper, a Semi-Empirical Model for predicting the Capacity Loss of Lithium-ion batteries during Cycling and Calendar Aging is developed. In order to redict the Capacity Loss with a high accuracy, battery operation data from different test conditions and different Lithium-ion batteries chemistries were obtained from literature for parameter optimization (fitting). The obtained models were then compared to experimental data for validation. Our results show that the average error between the estimated Capacity Loss and measured Capacity Loss is less than 1.5% during Cycling Aging, and less than 2% during Calendar Aging. An electric mining dumper, with simulated duty cycle data, is considered as an application example.

This article describes ModelConductor-eXtended (MCX), which is an open-source software architecture for digital twins. The MCX framework facilitates co-execution of, and asynchronous data communication between, physical systems and their digital simulation models. MCX supports running FMUs (simulation models packaged according to the FMI specification) as well as machine learning models and customized models. We propose extensions to the previously published ModelConductor framework for higher performance and better scalability. The extensions include decoupling of the queue and the model computation module, utilization of a standard data transmission protocol and implementation of the facility to run time-consuming simulation models in a time synchronous manner. Additionally, three new validation case studies are presented. A performance evaluation shows that the extensions improve the average response time almost 4 times in three specific experiments.