This study focuses on the critical aspect of robust state estimation for the safe navigation of an Autonomous Vehicle (AV). Existing literature primarily employs two prevalent techniques for state estimation, namely filtering-based and graph-based approaches. Factor Graph (FG) is a graph-based approach, constructed using Values and Factors for Maximum Aposteriori (MAP) estimation, that offers a modular architecture that facilitates the integration of inputs from diverse sensors. However, most FG-based architectures in current use require explicit knowledge of sensor parameters and are designed for single setups. To address these limitations, this research introduces a novel plug-and-play FG-based state estimator capable of operating without predefined sensor parameters. This estimator is suitable for deployment in multiple sensor setups, offering convenience and providing comprehensive state estimation at a high frequency, including mean and covariances. The proposed algorithm undergoes rigorous validation using various sensor setups on two different vehicles: a quadricycle and a shuttle bus. The algorithm provides accurate and robust state estimation across diverse scenarios, even when faced with degraded Global Navigation Satellite System (GNSS) measurements or complete outages. These findings highlight the efficacy and reliability of the algorithm in real-world AV applications.
This paper presents a framework to jointly optimize the design and control of an electric race car equipped with a multiple-gear transmission (MGT), specifically accounting for discrete gearshift dynamics. We formulate the problem as a mixed-integer optimal control problem (MIOCP), and deal with its complexity by combining convex optimization and Pontryagin's Minimum Principle (PMP) in a computationally efficient iterative algorithm satisfying necessary conditions for optimality upon convergence. Finally, we leverage our framework to compute the achievable lap time of a race car equipped with a fixed-gear transmission (FGT), a continuously variable transmission (CVT) and an MGT with 2 to 4 speeds, revealing that an MGT can strike the best trade-off in terms of electric motor control, and transmission weight and efficiency, ultimately yielding the overall best lap time.
In general, electric motor design procedures for automotive applications go through expensive trial-and-error processes or use simplified models that linearly stretch the efficiency map. In this paper, we explore the possibility of efficiently optimizing the motor design directly, using high-fidelity simulation software and derivative-free optimization solvers. In particular, we proportionally scale an already existing electric motor design in axial and radial direction, as well as the sizes of the magnets and slots separately, in commercial motor design software. We encapsulate this motor model in a vehicle model together with the transmission, simulate a candidate design on a drive cycle, and find an optimum through a Bayesian optimization solver. We showcase our framework on a small city car, and observe an energy consumption reduction of 0.13% with respect to a completely proportional scaling method, with a motor that is equipped with relatively shorter but wider magnets and slots. In the extended version of this paper, we include a comparison with the linear models, and add experiments on different drive cycles and vehicle types.
In this paper, a novel high-performance active switched quasi-Z-Source inverter (HP-AS-qZSI) dual-source for fuel cell hybrid electric vehicle (FC-HEV) is proposed. In order to eliminate extra dc-dc converters, dual-energy sources based on FC and lithium-ion capacitors (LiCs) are integrated into the Z-source network (ZSN). By adding an anti-parallel power switch, the proposed topology enables to deal with the uncontrollable and distorted dc-link voltages in FCEV applications-based broad-range of loads over the traditional AS-qZSI. The modeling and the operation modes analysis are firstly presented. Real-time simulation based on Opal-RT is then implemented to validate the operation and performance of the proposed topology. As a result, it provides a higher average efficiency (3.06%) and lower component size and volume of passive elements for the EV system. Furthermore, this topology also indicates improved aging performance indexes of high specific-energy sources under the studied Artemis-long driving cycle, compared to the hybrid energy storage system conventional two-stage inverter.
An energy management strategy for mild hybrids that prevents battery overheating is introduced in this digest. Energy management strategy design for mild hybrids requires particular care to prevent overheating of the battery pack as they typically do not have an active cooling system. To tackle this issue, we extend the well-known equivalent consumption minimization strategy approach to develop a real-time capable fuel-optimal controller that is sensitive to the battery’s thermal dynamics and that can enforce constraints on its temperature. The rationale for our formulation is developed using Pontryagin’s minimum principle from optimal control theory. The same principle is also used to design an off-line numerical procedure for the energy management strategy’s calibration. The effectiveness of the procedure is corroborated by numerical experiments on two different drive cycles, whose results are also compared with the solution obtained with a dynamic programming algorithm. Several peculiar aspects of our solution procedure, such as the method used to incorporate state constraints and the approximate boundary value problem solution method using a particle swarm optimization algorithm, are also detailed and discussed. The proposed controller is computationally light-weight and can be readily extended to on-line control provided that a suitable co-state selection procedure is employed, based on the data collected by using our calibration method on a large number of driving missions.
Electrified vehicles (EVs) are one of the promising technologies for promoting the clean energy revolution. The hybrid energy storage system (HESS), which has multiple energy storage components, requires an energy management strategy (EMS) to reasonably allocate the overall power demand to sub-components. In this paper, a new predictive EMS is proposed to allocate the overall demanded current for the HESS of an EV. More specifically, an end-to-end prediction method is proposed using recurrent neural networks to forecast the bus current demand. Under power demand prediction, a rule-based EMS is developed to allocate the loads between the supercapacitor and Li-ion battery via multi-objective optimization. The proposed EMS is validated with respect to prediction accuracy and other metrics provided by the 2022 IEEE VTS motor challenge. And simulation results demonstrate the superior performance of the proposed algorithm, compared to other conventional methods.
Brake blending design is a complex task in the development of electric vehicles (EV) and requires coordinated control on electric powertrain and friction brake system. The presented study discusses the architecture, control strategy, and functional validation of the brake blending as applied to an all-wheel drive EV equipped with four in-wheel motors (IWMs) and the decoupled electro-hydraulic brake system. The main focus is on the use of distributed and shared X-in-the-loop (XIL) test environment as a relevant development methodology enabling a wide spectrum of validation procedures. The paper introduces XIL-based experiments, where the brake blending operation has been evaluated for several test scenarios as the service braking and the Worldwide harmonized Light Duty Test Cycle (WLTC).
This paper presents a methodology that leverages learning techniques and robust control theory to design an adaptive controller for a wide class of linear dynamical dissipative vehicle systems. In particular, learning techniques such as neural networks are used as adaptive learning blocks in the feedback loop with the system under control to update the controller parameters. In order to guarantee the stability of the closed-loop system, a library of parametrized controller blocks that satisfy either the strictly negative imaginary property (SNI), in the case of the negative imaginary system (NI), or the strictly positive real property (SPR) in the case of a positive real system (PR), is developed. The parameters in these controllers are learned using a chosen learning block. The main advantage of including a learning block is to continuously improve performance in the presence of any uncertainty in the environment and the changes in the system's dynamics. This is achieved by allowing the learning block to update the controller parameters based on a defined cost function. Simulation flights testing a quad-copter system are given to illustrate our approach.
This paper explores the possibility of implementing distributed optimization through peer-to-peer energy (P2P) trading platform while keeping the temperature of batteries in electric vehicles below the safe threshold. While most thermal management schemes consider a battery as a point to simplify the computation, this work proposes to model and regulate the battery's temperature profile over its cells. This more granular access to the battery's temperature allows for more effective management as the temperature distribution over cells is not homogeneous. As a result, a linear discrete-time model is derived for the cell-level temperature vector and incorporated into the P2P energy trading platform. A multi-period P2P policy is developed to maximize the EV's owner benefits while preventing battery cells' temperature from rising too high. Numerical results with a set of 15 agents which are a part of IEEE 33-bus radial distribution test feeder are presented to demonstrate the performance of the proposed algorithm on P2P trading and battery temperatures.
To reduce the time-to-market of electric vehicles, fast and accurate energetic simulations are needed. This paper aims to propose a fast computational dynamic model that allows a good compromise between accuracy and computation time while respecting the dynamics of the system. Its accuracy and computation time are evaluated compared to conventional static and dynamic models. The results show that the proposed dynamic model estimates the same energy consumption as the traditional dynamic model for a computation time 85 times faster. The computation time of the static model is four times faster than the proposed model, but the accuracy is reduced.
The paper proposes an energy efficiency analysis for an Ultra-Fast Charging Station (UFCS). The UFCS, realized at the DIETI Lab of the University of Naples Federico II, is equipped with two charging slots of 160 kW rated power, a grid-tied power converter of 50 kW and a Battery Energy Storage System (BESS). It is operated in Grid to Vehicle (G2V) mode using a peak-shaving strategy. The infrastructure uses two DC/DC power converters based on H-bridge, a medium frequency transformer, and diode rectifier architecture. The infrastructure is able to operate the two converters as a single or twin EV charging configuration, respectively. This paper focuses on the operating behavior of the power converters during the phases of EV fast charging. Analyses of energy efficiency have been carried out taking into account the switching and thermal behavior of the power converters. The validation of the analyses is performed through comparison with the numerical results obtained in PLECS environment. The proposed methodology is a viable tool to give support in designing UFCSs in terms of components and subsystems sizing.
The suspension mechanism plays an important role in the dynamic behaviour of the vehicle. However, the equivalent suspension and damping rates, which neglect the kinematic effects of the suspension mechanism, are mainly employed in the conventional vehicle model. This paper presents a Jacobian approach to determine the equivalent suspension and damping rates based on the kinematic characteristics of planar double-wishbone suspension mechanism. Introducing the Jacobian method of a planar mechanism, the kinestatic relations of a double-wishbone suspension mechanism are described using instantaneous screws. Based on the kinestatic relations, it determines the equivalent suspension and damping rates. The dynamic performances of the conventional quarter-vehicle model are derived using Lagrange's equations. Finally, the dynamic performances of the conventional quarter-vehicle model are analysed and compared with the kineto-dynamic model and the simulation software Adams, respectively.
Battery electric vehicles are able to mitigate against peaks in energy generation and consumption of smart homes, if a bidirectional vehicle-to-home charging concept is established. Additionally, battery electric vehicles can be utilized as energy storage to prevent blackouts, to reduce energy costs and to increase self-sufficiency of these homes. Here, we show an ad-hoc machine learning approach to use vehicle-to-home to optimize renewable energy generation and consumption within a smart home grid. This solution is compared to state-of-the art optimization charging.
Wireless Power Transfer (WPT) will play a key role in the future of electric vehicles (EVs). WPT is usually applied for battery charger, and they operate in a unidirectional fashion. Bidirectional WPT(BD-WPT) can be integrated with the grid which could create opportunities to wirelessly feed power to the grid. A dual active bridge (DAB) configuration along with LCC resonant network is used for the BD-WPT system. Finding switching frequency and angle phase shift are two key factors that have a significant impact on the overall performance of the system. This paper presents a new computational methodology for tuning resonant switching frequency and phase shift for a DAB. The performance of the system in the grid to a vehicle (G2V) and vehicle to grid(V2G) operation for 3000 Watts is analyzed. Simulation results show power is injected into the DC grid, thus validating the feasibility of the proposed system.
In the last years, railway industry experts have been focused on the implementation of the smart grid concept and using renewable energy sources in electric railway systems (ERSs). In this paper, a model of future 9 kV MVDC ERS is proposed using MATLAB software integrating distributed energy sources and EV charging infrastructures. An algorithm is modified to accept real data from a physical railway system and simulate a digital twin (DT) based model. In this context, Rome-Florence high-speed railway line is considered as a real case study. An example modeling on the integration of wind turbines (WT), photovoltaic (PV), and EV charging infrastructures as auxiliary power supply to MVDC railway microgrid is presented with a modified power management system considering regenerative braking energy of trains.
Expending manufacturing capacity and development of high-energy batteries greatly stimulate the growth and applications of electric vehicles (EVs). However, battery diagnostics and prognostics related to capacity degradation (referred as state of health, SOH) and safety issues (referred as state of safety, SOS) in real-world applications is still a big deal. Due to the uncertainties in materials and manufacturing, dynamic operation conditions as well as a lack of plentiful, high-quality on-road data, accurate diagnosis of battery performance for “real EVs” is very challenging. Considering the difficulty in accurately predicting battery behaviors in real-world applications, brand-new control area networks (CAN) and cloud-based solution could have considerable benefits. An AI-powered cloud-based framework integrating longitudinal electronic health records with real-world data enables continuous battery performance evaluation for EVs. This offers opportunities for combining data generation with data-driven approaches to predict the behavior of complex, time-varying electrochemical systems. It is hoped that this paper will be of reference value to the EV and battery industries for ameliorating some of the hurdles for battery diagnostics and prognostics under realistic EV conditions.
This work describes the optimization of paths and velocity profiles for the usage in semi-autonomous vehicles. The focus of the optimization lies on an energy efficient motion while considering the driver’s longitudinal demands. The trade-off between travel time and energy consumption of the vehicle can be directly chosen by the driver. With help of a nonlinear model predictive control framework the global optimum of the velocity optimization with nonlinear constraints based on the vehicle’s physical limits is found. The online path planner together with the online velocity planner provide the input to the underlying vehicle’s control modules, as a basis for automated driving. The effectiveness of the proposed approach is shown via hardware-in-the-loop tests and real-world driving tests.
The implementation and management of Electric Vehicle (EV) charging points in parking spaces (whether in multifamily condominiums or in private company parking lots) presents a challenge, since the available contracted power may be insufficient for the building devices needs and for EV chargers to work simultaneously. The available power is limited to a fixed value generally determined by the expected peak power consumption of the building. An increase in the contracted power leads to unwanted costs and, sometimes, this is not possible without a complete rework of the electrical installation at even higher costs. The available power is not fully used all the time, since not all the buildings devices are always working simultaneously. These spare power can be monopolized by a mesh network of chargers, taking advantage of the full contracted power without increasing costs. In this work we present a manager of a mesh network of chargers that can distribute the available power to an array of chargers based on several conditions, e.g. load balancing, priority of chargers or charging.
Tire vertical force is an important factor in the vehicle system because the tire vertical force directly affects the longitudinal force such as the driving force and the braking force, and the lateral force caused by steering motion. Therefore, estimating the tire vertical force is an essential issue. In this study, we propose a tire vertical force observer(TVFOB) using the measurable acceleration sensor and the suspension deformation sensor in a vehicle. The observer set a tire radius as a state, and the derivative of the tire radius is assumed as stochastic white Gaussian noise. A quarter car model is used to verify the performance of the algorithm The proposed algorithm is evaluated through the simulation by using Matlab Simulink and Carmaker.
Knowledge of the environmental impact at different steps of the battery life cycle is essential due to the environmental and geopolitical tensions surrounding the electric vehicles. Because of the diversity of data on this subject, it is difficult to deduce which battery chemistry has the least impact on the environment. In this paper, a method to determine the environmental impact of batteries for raw material extraction, batteries production, transport, end-of-life, and recycling is proposed. The analysis shows that the lead-acid battery has a lower global warming potential than lithium batteries for the same energy.
