U.S. Army Combat Capabilities Development Command Ground Vehicle System Center

Electric motor and power electronics-based inverter are the major components in industrial and automotive electric drives. In this paper, we present a model-based fault diagnostics system developed using a machine learning technology for detecting and locating multiple classes of faults in an electric drive. Power electronics inverter can be considered to be the weakest link in such a system from hardware failure point of view; hence, this work is focused on detecting faults and finding which switches in the inverter cause the faults. A simulation model has been developed based on the theoretical foundations of electric drives to simulate the normal condition, all single-switch and post-short-circuit faults. A machine learning algorithm has been developed to automatically select a set of representative operating points in the (torque, speed) domain, which in turn is sent to the simulated electric drive model to generate signals for the training of a diagnostic neural network, fault diagnostic neural network (FDNN). We validated the capability of the FDNN on data generated by an experimental bench setup. Our research demonstrates that with a robust machine learning approach, a diagnostic system can be trained based on a simulated electric drive model, which can lead to a correct classification of faults over a wide operating domain.

The estimation of temperature inside a battery cell requires accurate information about the cooling conditions even when the battery surface temperature is measured. This paper presents a model-based approach for estimating temperature distribution inside a cylindrical battery under unknown convective cooling conditions. A reduced-order thermal model using a polynomial approximation of the temperature profile inside the battery is used. A dual Kalman filter (DKF), a combination of a Kalman filter and an extended Kalman filter, is then applied for the identification of the convection coefficient and the estimation of the battery core temperature. The thermal properties are modeled by volume averaged lumped-values under the assumption of a homogeneous and isotropic volume. The model is parameterized and validated using experimental data from a 2.3 Ah 26650 lithium-iron-phosphate battery cell with a forced-air convective cooling during hybrid electric vehicle drive cycles. Experimental results show that the proposed DKF-based estimation method can provide an accurate prediction of the core temperature under unknown cooling conditions by measuring battery current and voltage along with surface and ambient temperatures.

In physics-based engineering modeling, the two primary sources of model uncertainty, which account for the differences between computer models and physical experiments, are parameter uncertainty and model discrepancy. Distinguishing the effects of the two sources of uncertainty can be challenging. For situations in which identifiability cannot be achieved using only a single response, we propose to improve identifiability by using multiple responses that share a mutual dependence on a common set of calibration parameters. To that end, we extend the single response modular Bayesian approach for calculating posterior distributions of the calibration parameters and the discrepancy function to multiple responses. Using an engineering example, we demonstrate that including multiple responses can improve identifiability (as measured by posterior standard deviations) by an amount that ranges from minimal to substantial, depending on the characteristics of the specific responses that are combined.

The expansion of battery material during lithium intercalation is a concern for the cycle life and performance of lithium ion batteries. In this paper, electrode expansion is quantified from in situ neutron images taken during cycling of pouch cells with lithium iron phosphate positive and graphite negative electrodes. Apart from confirming the overall expansion as a function of state of charge and the correlation with graphite transitions that have been observed in previous dilatometer experiments we show the spatial distribution of the expansion along the individual electrodes of the pouch cell. The experiments were performed on two cells with different electrode areas during low and high c-rate operation. The measurements show how charging straightened the cell layers that were slightly curved by handling of the pouch cell during setup of the experiment. Subsequent high charging rate, that exceeded the suggested operating voltage limits, was shown to have a strong influence on the observed expansion. Specifically, during high-rate cycling, the battery showed a much larger and irreversible expansion of around 1.5% which was correlated with a 4% loss in capacity over 21 cycles.

One of the existing challenges toward the electrification of military vehicles is the selection of the most suitable energy storage device. Moreover, a single energy storage technology might not provide the most benefit out of powertrain electrification. In this paper, a generalized framework for the simultaneous selection of the optimal energy storage device, in the form of a standalone or hybrid solution, and online energy management is presented. This paper investigates the cooperation of energy-dense Li-ion batteries and power-dense supercapacitors to assist engine operation in a series hybrid electric military truck. Pontryagin’s minimum principle is adopted as the energy management strategy in a forward-looking vehicle simulator, in which the optimal design and control parameters are found using particle swarm optimization. Simulation results show that adopting a hybrid energy storage system reduces fuel consumption by 13% compared to the case of battery-only hybridized powertrain.

This paper shows how neutron radiography can be used for in situ quantification of the lithium concentration across battery electrodes, a critical physical system state. The change in lithium concentration between the charged and discharged states of the battery causes a change in number of detected neutrons after passing through the battery. Electrode swelling is also observed during battery charging. The experimental setup and the observations from testing a pouch cell with LFP cathode and graphite anode are reported here. The bulk Li concentration across the electrode and folds of the pouch cell is quantified at various states of charge. To interpret the measurements, the optics of the neutron beam (geometric unsharpness) and detector resolution are considered in order to quantify the lithium concentration from the images due to the thinness of the electrode layers. The experimental methodology provides a basis for comprehensive in situ metrology of bulk lithium concentration.

The automobile brake-by-wire (BBW) system, which is also called the electromechanical brake system, has become a promising vehicle braking control scheme that enables many new driver interfaces and enhanced performances without a mechanical or hydraulic backup. In this paper, we survey BBW control systems with focuses on fault tolerance design and vehicle braking control schemes. At first, the system architecture of BBW systems is described. Fault tolerance design is then discussed to meet the high requirements of reliability and safety of BBW systems. A widely used braking model and several braking control schemes are investigated. Although previous work focused on antilock and antislip braking controls on a single wheel basis, we present a whole-vehicle control scheme to enhance vehicle stability and safety. Simulations based on the whole-vehicle braking model validate a proposed fuzzy logic control scheme in the lateral and yaw stability controls of vehicles.

Although the battery surface temperature is commonly measured, the core temperature of a cell may be much higher hence more critical than the surface temperature. The core temperature of a battery, though usually unmeasured in commercial applications, can be estimated by an observer, based on a lumped-parameter battery thermal model and the measurement of the current and the surface temperature. Even with a closed loop observer based on the measured surface temperature, the accuracy of the core temperature estimation depends on the model parameters. For such purpose, an online parameterization methodology and an adaptive observer are designed for a cylindrical battery. The single cell thermal model is then scaled up to create a battery cluster model to investigate the temperature pattern of the cluster. The modeled thermal interconnections between cells include cell to cell heat conduction and convection to the surrounding coolant flow. An observability analysis is performed on the cluster before designing a closed loop observer for the pack. Based on the analysis, guidelines for determining the minimum number of required sensors and their exact locations are derived that guarantee the observability of all temperature states.

In this paper the parameterization of an integrated electro-thermal model for an A123 26650 LiFePO4 battery is presented. The electrical dynamics of the cell are described by an equivalent circuit model. The resistances and capacitances of the equivalent circuit model are identified at different temperatures and SOC’s, for charging and discharging. Functions are chosen to charac-terize the fitted parameters. A two-state thermal model is used to capture the core and surface temperatures of the battery. The electrical model is coupled with the thermal model through heat generation. Parameters of the thermal model are identified us-ing a least squares algorithm. The electro-thermal model is then validated against voltage and surface temperature measurements from a realistic drive cycle experiment. 1

A multi-dimensional model was applied to investigate the influence of combustion regimes on heat transfer losses in internal combustion engines. By utilizing improved turbulence and heat transfer sub-models, the combustion and heat transfer characteristics of the engine were satisfactorily reproduced for operation under conventional diesel combustion, homogeneous charge compression ignition, and reactivity controlled compression ignition regimes. The results indicated that the total heat transfer losses of conventional diesel combustion are the largest among the three combustion regimes due to the direct interaction of the high-temperature flame with the piston wall, while the heat transfer losses of reactivity controlled compression ignition are the lowest and nearly are independent of combustion phasing because of the avoidance of high-temperature regions adjacent to the cylinder walls. Compared to conventional diesel combustion, homogeneous charge compression ignition shows more potential for the reduction of exhaust energy and improvement of fuel efficiency. In reactivity controlled compression ignition combustion, the reduction of heat transfer and exhaust losses outweigh its increase in combustion losses and offer the opportunity for further improvement of fuel efficiency. Furthermore, by evaluating the widely used Woschni and Chang et al.’s empirical heat transfer correlations, it was found that both correlations considerably overestimate the heat transfer rate for the reactivity controlled compression ignition regime. It is necessary to improve empirical heat transfer models to take account of the flow and combustion characteristics under various combustion modes.

The engine cooling system for a typical class 3 pickup truck with a medium duty diesel engine was modeled with a commercial code, GT-Cool, in order to explore the benefit of a controllable electric pump on the cooling performance and the pump operation. As the first step, the cooling system model with a conventional mechanical coolant pump was validated with experimental data. After the model validation, the mechanical pump submodel was replaced with the electric pump submodel, and then the potential benefit of the electric pump on fuel economy was investigated with the simulation. Based on coolant flow analysis, a modified thermostat hysteresis was proposed to reduce the recirculating flow and the electric pump effort. It was also demonstrated that the radiator size could be reduced without any cooling performance penalty by replacing the mechanical pump with the electric pump. The predicted results indicate that the cooling system with the electric pump can dramatically reduce the pump power consumption during the FTP 74 driving schedule and that the radiator can be downsized by more than 27% of the original size, under the grade load condition.

We summarize and numerically compare two approaches for modeling and simulating the dynamics of dry granular matter. The first one, the discrete-element method via penalty (DEM-P), is commonly used in the soft matter physics and geomechanics communities; it can be traced back to the work of Cundall and Strack [P. Cundall, Proc. Symp. ISRM, Nancy, France 1, 129 (1971); P. Cundall and O. Strack, Geotechnique 29, 47 (1979)GTNQA80016-850510.1680/geot.1979.29.1.47]. The second approach, the discrete-element method via complementarity (DEM-C), considers the grains perfectly rigid and enforces nonpenetration via complementarity conditions; it is commonly used in robotics and computer graphics applications and had two strong promoters in Moreau and Jean [J. J. Moreau, in Nonsmooth Mechanics and Applications, edited by J. J. Moreau and P. D. Panagiotopoulos (Springer, Berlin, 1988), pp. 1-82; J. J. Moreau and M. Jean, Proceedings of the Third Biennial Joint Conference on Engineering Systems and Analysis, Montpellier, France, 1996, pp. 201-208]. The DEM-P and DEM-C are manifestly unlike each other: They use different (i) approaches to model the frictional contact problem, (ii) sets of model parameters to capture the physics of interest, and (iii) classes of numerical methods to solve the differential equations that govern the dynamics of the granular material. Herein, we report numerical results for five experiments: shock wave propagation, cone penetration, direct shear, triaxial loading, and hopper flow, which we use to compare the DEM-P and DEM-C solutions. This exercise helps us reach two conclusions. First, both the DEM-P and DEM-C are predictive, i.e., they predict well the macroscale emergent behavior by capturing the dynamics at the microscale. Second, there are classes of problems for which one of the methods has an advantage. Unlike the DEM-P, the DEM-C cannot capture shock-wave propagation through granular media. However, the DEM-C is proficient at handling arbitrary grain geometries and solves, at large integration step sizes, smaller problems, i.e., containing thousands of elements, very effectively. The DEM-P vs DEM-C comparison is carried out using a public-domain, open-source software package; the models used are available online.

This article considers autonomous systems whose behaviors seek to optimize an objective function. This goes beyond standard applications of condition-based maintenance, which seeks to detect faults or failures in nonoptimizing systems. Normal agents optimize a known accepted objective function, whereas abnormal or misbehaving agents may optimize a renegade objective that does not conform to the accepted one. We provide a unified framework for anomaly detection and correction in optimizing autonomous systems described by differential equations using inverse reinforcement learning (RL). We first define several types of anomalies and false alarms, including noise anomaly, objective function anomaly, intention (control gain) anomaly, abnormal behaviors, noise-anomaly false alarms, and objective false alarms. We then propose model-free inverse RL algorithms to reconstruct the objective functions and intentions for given system behaviors. The inverse RL procedure for anomaly detection and correction has the training phase, detection phase, and correction phase. First, inverse RL in the training phase infers the objective function and intention of the normal behavior system using offline stored data. Second, in the detection phase, inverse RL infers the objective function and intention for online observed test system behaviors using online observation data. They are then compared with that of the nominal system to identify anomalies. Third, correction is executed for the anomalous system to learn the normal objective and intention. Simulations and experiments on a quadrotor unmanned aerial vehicle (UAV) verify the proposed methods.

Two approaches are commonly used for handling frictional contact within the framework of the discrete element method (DEM). One relies on the complementarity method (CM) to enforce a nonpenetration condition and the Coulomb dry-friction model at the interface between two bodies in mutual contact. The second approach, called the penalty method (PM), invokes an elasticity argument to produce a frictional contact force that factors in the local deformation and relative motion of the bodies in contact. We give a brief presentation of a DEM-PM contact model that includes multi-time-step tangential contact displacement history. We show that its implementation in an open-source simulation capability called Chrono is capable of accurately reproducing results from physical tests typical of the field of geomechanics, i.e., direct shear tests on a monodisperse material. Keeping track of the tangential contact displacement history emerges as a key element of the model. We show that identical simulations using contact models that include either no tangential contact displacement history or only single-time-step tangential contact displacement history are unable to accurately model the direct shear test.

This article develops a stochastic programming framework for multiagent systems, where task decomposition, assignment, and scheduling problems are simultaneously optimized. The framework can be applied to heterogeneous mobile robot teams with distributed subtasks. Examples include pandemic robotic service coordination, explore and rescue, and delivery systems with heterogeneous vehicles. Owing to their inherent flexibility and robustness, multiagent systems are applied in a growing range of real-world problems that involve heterogeneous tasks and uncertain information. Most previous works assume one fixed way to decompose a task into roles that can later be assigned to the agents. This assumption is not valid for a complex task where the roles can vary and multiple decomposition structures exist. Meanwhile, it is unclear how uncertainties in task requirements and agent capabilities can be systematically quantified and optimized under a multiagent system setting. A representation for complex tasks is proposed: agent capabilities are represented as a vector of random distributions, and task requirements are verified by a generalizable binary function. The conditional value at risk is chosen as a metric in the objective function to generate robust plans. An efficient algorithm is described to solve the model, and the whole framework is evaluated in two different practical test cases: capture-the-flag and robotic service coordination during a pandemic (e.g., COVID-19). Results demonstrate that the framework is generalizable, is scalable up to 140 agents and 40 tasks for the example test cases, and provides low-cost plans that ensure a high probability of success.

Hybrid and electric vehicle (HEV/EV) technology is reasonably mature at this time, with a few million vehicles around in the world, and there is a significant amount of literature in the public domain on this subject. However, there is not enough literature on the application of this technology for off-road vehicles, including construction equipment, other industrial utility vehicles, and nonautomotive applications, such as a locomotive, ships, or airborne vehicles. With this in mind, the author presents here the topic and its current status. In addition, the author discusses the issue related to the decision-making process before the above technology is introduced for any HEV/EV application so that one is assured that the technology will bring benefit if applied for a particular purpose.

Omnidirectional vehicles have been widely applied in several areas, but most of them are designed for the case of motion on flat, smooth terrain, and are not feasible for outdoor usage. This paper presents the design and development of an omnidirectional mobile robot that possesses high mobility in rough terrain. The omnidirectional robot consists of a main body with four sets of mobility modules, called an active split offset caster (ASOC). The ASOC module has independently driven dual wheels that produce arbitrary planar translational velocity, enabling the robot to achieve its omnidirectional motion. Each module is connected to the main body via a parallel link with shock absorbers, allowing the robot to conform to uneven terrain. In this paper, the design and development of the ASOC‐driven omnidirectional mobile robot for rough terrain are described. A control scheme that considers the kinematics of the omnidirectional mobile robot is presented. The mobility of the robot is also evaluated experimentally based on a metric called the ASOC mobility index. The mobility evaluation test clarifies a design tradeoff between terrain adaptability and omnidirectional mobility due to the shock absorbers. In addition, an odometry improvement technique that can reduce position estimation error due to wheel slippage is proposed. Experimental odometry tests confirmed that the proposed technique is able to improve the odometry accuracy for sharp‐turning maneuvers.

Large-scale metal additive manufacturing (AM) provides a unique solution to rapidly develop prototype components with net-shape or near-net shape geometries. Specifically, additive friction stir deposition (AFSD) is a solid-state method for large-scale metal AM that produces near-net shape depositions capable of high deposition rates. As AFSD is utilized for a broader range of applications, there is a need to understand deposition strategies for larger and more complex geometries. In particular, components with larger surface areas will require overlapping deposition passes within a single layer. In this study, the AFSD process was used to create depositions utilizing multiple passes with a varying deposition path overlap width. The effects of overlapping parallel pass depositions on the mechanical and microstructural properties of aluminum alloy 7075 were examined. The grain size and microstructural features of the deposited material were analyzed to evaluate material mixing and plastic flow in the observed overlap regions. Additionally, hardness and tensile experiments were conducted to observe the relationship between the overlap width and as-deposited material behavior. In this study, an ideal overlap width was found that produced acceptable as-deposited material properties.

Computer-generated animated and enhanced movies and training simulations are now commonplace because of an exponential increase in cost-effective computer power and software robustness. Today, pilots and astronauts routinely train and practice normal and emergency scenarios much as they would in a real vehicle in an operational environment. As entertainment-seeking consumers, we regularly find ourselves viewing movies, TV, and games containing imaginary dinosaurs, aircraft dogfights, monsters, and voyages to other planets and galaxies.

Spiking artificial neurons emulate the voltage spikes of biological neurons and constitute the building blocks of a new class of energy efficient, neuromorphic computing systems. Antiferromagnetic materials can, in theory, be used to construct spiking artificial neurons. When configured as a neuron, the magnetization in antiferromagnetic materials has an effective inertia that gives them intrinsic characteristics that closely resemble biological neurons, in contrast with conventional artificial spiking neurons. It is shown here that antiferromagnetic neurons have a spike duration on the order of picoseconds, a power consumption of about 10−3 pJ per synaptic operation, and built-in features that directly resemble biological neurons, including response latency, refraction, and inhibition. It is also demonstrated that antiferromagnetic neurons interconnected into physical neural networks can perform unidirectional data processing even for passive symmetrical interconnects. The flexibility of antiferromagnetic neurons is illustrated by simulations of simple neuromorphic circuits realizing Boolean logic gates and controllable memory loops.