Daniel J. Auger

Electric vehicle battery model identification and state of charge estimation in real world driving cycles

This paper describes a study demonstrating a new method of state-of-charge (SoC) estimation for batteries in real-world electric vehicle applications. This method combines realtime model identification with an adaptive neuro-fuzzy inference system (ANFIS). In the study, investigations were carried down on a small-scale battery pack. An equivalent circuit network model of the pack was developed and validated using pulse-discharge experiments. The pack was then subjected to demands representing realistic WLTP and UDDS driving cycles obtained from a model of a representative electric vehicle, scaled match the size of the battery pack. A fast system identification technique was then used to estimate battery parameter values. One of these, open circuit voltage, was selected as suitable for SoC estimation, and this was used as the input to an ANFIS system which estimated the SoC. The results were verified by comparison to a theoretical Coulomb-counting method, and the new method was judged to be effective. The case study used a small 7.2 V NiMH battery pack, but the method described is applicable to packs of any size or chemistry.

Accuracy Versus Simplicity in Online Battery Model Identification

This paper presents a framework for battery modeling in online, real-time applications where accuracy is important but speed is the key. The framework allows users to select model structures with the smallest number of parameters that is consistent with the accuracy requirements of the target application. The tradeoff between accuracy and speed in a battery model identification process is explored using different model structures and parameter-fitting algorithms. Pareto optimal sets are obtained, allowing a designer to select an appropriate compromise between accuracy and speed. In order to get a clearer understanding of the battery model identification problem, “identification surfaces” are presented. As an outcome of the battery identification surfaces, a new analytical solution is derived for battery model identification using a closed-form formula to obtain a battery’s ohmic resistance and open circuit voltage from measurement data. This analytical solution is used as a benchmark for comparison of other fitting algorithms and it is also used in its own right in a practical scenario for state-of-charge estimation. A simulation study is performed to demonstrate the effectiveness of the proposed framework and the simulation results are verified by conducting experimental tests on a small NiMH battery pack.

Lithium–Sulfur Cell Equivalent Circuit Network Model Parameterization and Sensitivity Analysis

Compared to lithium-ion batteries, lithium–sulfur (Li-S) batteries potentially offer greater specific energy density, a wider temperature range of operation, and safety benefits, making them a promising technology for energy storage systems especially in automotive and aerospace applications. Unlike lithium-ion batteries, there is not a mature discipline of equivalent circuit network (ECN) modelling for Li-S. In this study, ECN modelling is addressed using formal ‘system identification’ techniques. A Li-S cells performance is studied in the presence of different charge/discharge rates and temperature levels using precise experimental test equipment. Various ECN model structures are explored, considering the tradeoffs between accuracy and speed. It was concluded that a ‘2RC’ model is generally a good compromise, giving good accuracy and speed. Model parameterization is repeated at various state-of-charge (SOC) and temperature levels, and the effects of these variables on Li-S cells ohmic resistance and total capacity are demonstrated. The results demonstrate that Li-S cells ohmic resistance has a highly nonlinear relationship with SOC with a break-point around 75% SOC that distinguishes it from other types of battery. Finally, an ECN model is proposed which uses SOC and temperature as inputs. A sensitivity analysis is performed to investigate the effect of SOC estimation error on the models accuracy. In this analysis, the battery models accuracy is evaluated at various SOC and temperature levels. The results demonstrate that the Li-S cell model has the most sensitivity to SOC estimation error around the break-point (around 75% SOC) whereas in the middle SOC range, from 20% to 70%, it has the least sensitivity.

Low-cost programmable battery dischargers and application in battery model identification

This paper describes a study where a low-cost programmable battery discharger was built from basic electronic components, the popular MATLAB programming environment, and an low-cost Arduino microcontroller board. After its components and their function are explained in detail, a case study is performed to evaluate the dischargers performance. The setup is principally suitable for any type of battery cell or small packs. Here a 7.2 V NiMH battery pack including six cells is used. Consecutive discharge current pulses are applied and the terminal voltage is measured as the output. With the measured data, battery model identification is performed using a simple equivalent circuit model containing the open circuit voltage and the internal resistance. The identification results are then tested by repeating similar tests. Consistent results demonstrate accuracy of the identified battery parameters, which also confirms the quality of the measurement. Furthermore, it is demonstrated that the identification method is fast enough to be used in real-time applications.

Fuzzy Logic Control for energy saving in Autonomous Electric Vehicles

Limited battery capacity and excessive battery dimensions have been two major limiting factors in the rapid advancement of electric vehicles. An alternative to increasing battery capacities is to use better: intelligent control techniques which save energy on-board while preserving the performance that will extend the range with the same or even smaller battery capacity and dimensions. In this paper, we present a Type-2 Fuzzy Logic Controller (Type-2 FLC) as the speed controller, acting as the Driver Model Controller (DMC) in Autonomous Electric Vehicles (AEV). The DMC is implemented using realtime control hardware and tested on a scaled down version of a back to back connected brushless DC motor setup where the actual vehicle dynamics are modelled with a Hardware-In-the-Loop (HIL) system. Using the minimization of the Integral Absolute Error (IAE) has been the control design criteria and the performance is compared against Type-1 Fuzzy Logic and Proportional Integral Derivative DMCs. Particle swarm optimization is used in the control design. Comparisons on energy consumption and maximum power demand have been carried out using HIL system for NEDC and ARTEMIS drive cycles. Experimental results show that Type-2 FLC saves energy by a substantial amount while simultaneously achieving the best IAE of the control strategies tested.

Lithium–Sulfur Battery State-of-Charge Observability Analysis and Estimation

Lithium–Sulfur (Li–S) battery technology is considered for an application in an electric-vehicle energy storage system in this study. A new type of Li-S cell is tested by applying load current and measuring cells terminal voltage in order to parameterize an equivalent circuit network model. Having the cells model, the possibility of state-of-charge (SOC) estimation is assessed by performing an observability analysis. The results demonstrate that the Li–S cell model is not fully observable because of the particular shape of cells open-circuit voltage curve. This feature distinguishes Li-S batteries from many other types of battery, e.g., Li-ion and NiMH. As a consequence, a Li-S cells SOC cannot be estimated using existing methods in the literature and special considerations are needed. To solve this problem, a new framework is proposed consisting of online battery parameter identification in conjunction with an estimator that is trained to use the identified parameters to predict SOC. The identification part is based on the well-known prediction-error minimization algorithm; and the SOC estimator part is an adaptive neuro-fuzzy inference system in combination with coulomb counting. Using the proposed method, a Li-S cells SOC is estimated with a mean error of 4% and maximum error of 7% in a realistic driving scenario.

State of Charge and State of Health Estimation Over the Battery Lifespan

The battery management system (BMS) plays a critical role in battery packs especially for the lithium-ion battery chemistry. Protecting the cells from overcharge and overdischarge, controlling the temperature at the desired level, prolonging the life of the battery pack, guaranteeing the safety and indicating the available power and energy of the battery are the key functionalities of a BMS. In this chapter, two important concepts of a BMS are discussed: (i) battery state-of-charge (SoC) and (ii) battery state-of-health (SoH). Battery SoC and SoH are variables which should be determined precisely in order to use the battery optimally and safely. Batteries are time-varying systems that behave very differently at various states. In other words, the internal states of a battery tell us what we should expect from it. Depending on the battery chemistry, various techniques have been developed in the literature for SoC and SoH estimation. This covers a wide range from simple integration of current over time (i.e. coulomb counting) to advanced estimation techniques such as Kalman filter. In this study, almost all the existing battery SoC and SoH estimation approaches are reviewed and proper references are cited for further studies in each category.

Analysis of autopilot disengagements occurring during autonomous vehicle testing

In present-day highly-automated vehicles, there are occasions when the driving system disengages and the human driver is required to take-over. This is of great importance to a vehicle U+02BC s safety and ride comfort. In the U.S state of California, the Autonomous Vehicle Testing Regulations require every manufacturer testing autonomous vehicles on public roads to submit an annual report summarizing the disengagements of the technology experienced during testing. On 1 January 2016, seven manufacturers submitted their first disengagement reports: Bosch, Delphi, Google, Nissan, Mercedes-Benz, Volkswagen, and Tesla Motors. This work analyses the data from these disengagement reports with the aim of gaining abetter understanding of the situations in which a driver is required to takeover, as this is potentially useful in improving the Society of Automotive Engineers U+0028 SAE U+0029 Level 2 and Level 3 automation technologies. Disengagement events from testing are classified into different groups based on attributes and the causes of disengagement are investigated and compared in detail. The mechanisms and time taken for take-over transition occurred in disengagements are studied. Finally, recommendations for OEMs, manufacturers, and government organizations are also discussed.

Modular Battery Cell Model for Thermal Management Modelling

Internal policies of the major car markets are urging for a cut in oil imports, leading to powertrain electrification. Due to their high weight-to-power ratio, Lithium-ion batteries, especially Lithium-Nickel-Manganese-Cobalt Oxide 21700 cylindrical cells, are rapidly becoming the most diffused electric powertrain energy storage devices. These devices need to be operated in a tight temperature range to prevent major power drops and also for safety reasons. It is therefore essential to provide an accurate but computationally inexpensive battery model. Current models are either too simplistic and not applicable for thermal management design purposes or too computationally expensive and impractical for heat exchange modelling purposes. This work was focused on a computationally convenient system-level-modelling-oriented battery cell model. Starting from a 1D model obtained from manufacturer’s data, experiments were carried out on real cells, a more sophisticated 3D model for cell characterization was implemented and then a lighter 1D model obtained from it was proposed. The outcome is a novel thermal model of batteries, with a reasonable computational cost, developed on the purpose of thermal management design. This represents an advancement in battery thermal management design, as no such model is currently available in literature.

A hardware-in-the-loop test rig for development of electric vehicle battery identification and state estimation algorithms

This paper describes a hardware-in-the-loop (HIL) test rig for the test and development of electric vehicle battery parameterisation and state-estimation algorithms in the presence of realistic real-world duty cycles. The rig includes two electric machines, a battery pack, a real-time simulator, a thermal chamber and a PC for human-machine interface. Other parts of a vehicle powertrain system are modelled and used in the real-time simulator. A generic framework has been developed for real-time battery measurement, model identification and state estimation. Measurements are used to extract parameters of an equivalent circuit network model. Outputs of the identification unit are then used by an estimation unit trained to find the relationship between the battery parameters and state-of-charge. The results demonstrate that even with a high noise level in measured data, the proposed identification and estimation algorithms are able to work well in real-time.