Patent Application
20200210541
Lithium batteries are used in many modern devices, including electric vehicles, computers, and cell phones. One attractive aspect of lithium-ion batteries is that they may be fast charged at a quicker rate than other rechargeable batteries. Fast charging does, however, have disadvantages. For example, fast charging can cause an accelerated capacity fading, resulting in the possibility of triggering a safety issue. During fast charging, lithium-ions tend to plate on the negative active material surface instead of intercalating into the material. Once lithium ions are plated, the lithium-ion battery degrades in several ways, including but not limited to creating an electrical pathway between the active material and the electrolyte through a solid electrolyte interface (SEI), exposing electrons to the electrolyte.
To minimize lithium metal plating, battery cells have been subject to extensive lithium plating tests to determine the maximum region and continuous charging current limits as a function of a state of charge (SOC) and temperature. However, practical models that are usable in real time and provide accurate results are not available by systems and processes of the prior art.
The present technology, roughly described, utilizes a multiple particle reduced order model to accurately predict lithium plating potential in real time during the life of a lithium battery cell. The battery model can be based on several observations and assumptions, such as for example that cell voltage protection with a single particle reduced order model is accurate for a low or pulsing electrical load when a lithium concentration and potential gradient inside a cell is negligible. In the current multi-particle reduced order modeling system, only the current density and the potential distributions are solved iteratively. This is based on a premise that the electrical field and the charge transfer action processes occur at a smaller timescale than the diffusion timescale.
Once the current distribution is solved, lithium concentration distribution is solved without involving any iterative process. By solving the lithium concentration distribution as a separate step after the iteratively determined current density and potential distributions, the computation time required by the model to generate an output is dramatically reduced by avoiding solving multiple partial derivative equations iteratively. The accuracy of potential distribution within a cell is significantly improved compared to a single particle base model. Based on the potential distribution information provided by the output of the model, lithium plating potential can be determined and actions can be taken, such as modified charging techniques and rates, to minimize future lithium plating.
In embodiments, a method is disclosed for modeling a battery cell to detect lithium ion plating potential that may lead to battery cell degradation. The method may include setting a lithium ion concentration for a model battery by a battery management system on a battery-powered system. The battery model can provide a model for the battery cell on the battery-powered system. A temperature of the battery cell on the battery-powered system can be predicted in the model battery, and the temperature can be set as the modeled battery cell temperature. Material properties for the model battery can be set based at least in part on the modeled battery temperature. The potential distribution and current density for the model battery can be iteratively determined by the battery management system. A lithium plating potential for the model battery can then be calculated by the battery management system based at least in part on the potential distribution.
In embodiments, a non-transitory computer readable storage medium includes a program, the program being executable by a processor to perform a method for modeling a battery cell to detect battery cell degradation. The method may include setting a lithium ion concentration for a model battery by a battery management system on a battery-powered system. The battery model can provide a model for the battery cell on the battery-powered system. A temperature of the battery cell on the battery-powered system can be detected, and the model battery temperature can be set as the battery cell temperature. Material properties for the model battery can be set based at least in part on the modeled battery temperature. The potential distribution and current density for the model battery can be iteratively determined by the battery management system. A lithium plating potential for the model battery can then be calculated by the battery management system based at least in part on the potential distribution.
In embodiments, a system for modeling a battery cell to detect battery cell degradation includes one or more processors, memory, and one or more modules stored in memory and executable by the one or more processors. When executed, the modules may set a lithium ion concentration for a modeled battery by a battery management system on a battery powered system, the battery model providing a model for a battery cell on the battery powered system, detect a temperature of the battery cell on battery powered system and setting the modeled battery temperature as the battery cell temperature, set material properties for the modeled battery based at least in part on the modeled battery temperature, iteratively determine potential distribution and current density for the modeled battery by the battery management system, and calculate a lithium plating potential for the modeled battery by the battery management system based at least in part on the potential distribution.
The present technology, roughly described, utilizes a multiple particle reduced order model to accurately predict lithium plating potential in real time during the life of a lithium battery cell. The battery model can be based on several observations and assumptions, such as for example that cell voltage protection with a single particle reduced order model is accurate for a low or pulsing electrical load when a lithium concentration and potential gradient inside a cell is negligible. Under a continuous electrical load such as during charging, however, the single particle model prediction will start to deviate from the measurement. This is due to the model being forced to use the average current density in the calculation.
In a full order model, the current density distribution, potential distribution such as the electrode potential and electrolyte potential, and the lithium concentration distribution are mutually dependent. Because the model is highly nonlinear, the model solution needs to be solved iteratively. In the current multi-particle reduced order modeling system, only the current density and the potential distributions are solved iteratively. This is based on a premise that the electrical field and the charge transfer action processes occur at a smaller timescale than the diffusion timescale.
Once the current distribution is solved, lithium concentration distribution is solved without involving any iterative process. By solving the lithium concentration distribution as a separate step after the iteratively determined current density and potential distributions, the computation time required by the model to generate an output is dramatically reduced by avoiding solving multiple partial derivative equations iteratively. The accuracy of potential distribution within a cell is significantly improved compared to a single particle base model. Based on the potential distribution information provided by the output of the model, lithium plating potential can be determined and actions can be taken, such as modified charging techniques and rates, to minimize future lithium plating.
The modeling technique of the present technology provides advantages over other modeling techniques and that it provides accurate results and can be implemented in real time, for example on a battery-powered system such as an electric vehicle, computer, mobile phone, or other device. Real-time applications of a physics-based model by prior systems are limited due to the high computational cost. In a lithium-ion battery cell model, many particles are considered to represent an electrode to capture current density and potential distribution inside the battery cell. The process of modeling is computationally intensive as it involves iteratively solving many partial differential equations. To reduce computation time for real-time application, a common model reduction scheme is to consider a single particle to represent an electrode. In some and set of solving multiple partial differential equations at each discrete time step, only a single partial differential equation needs to be solved with a single particle model. With this approach, however, the accuracy is poor because it cannot capture spatial dependent current density distribution. Any reliance on such a model to detect and avoid lithium plating will lead to erroneous results.
The technical problem addressed by the present technology relates to identifying degradation in batteries by modeling a battery cell. In some prior solutions degradation in batteries, such as lithium plating, is determined by modeling the battery. To provide an accurate model, a battery is modeled using multiple particles to represent each electrode. Though the typical multiple particle electrode model can provide accurate results, it requires large computational resources, cannot provide results in real time, and is not practical for use in consumer systems. Other models represent electrodes as a single particle rather than multiple particles, and require much less computational cost. A single particle electrode model, however, has the disadvantage of not providing very accurate results, which can lead to incorrect lithium plating detection and prediction.
The present technology provides a technical solution to the technical problem of modeling a battery cell in real time so that the model can be used by a battery powered system with the battery being modeled. The battery cell model of the present technology provides a multi-particle reduced order model that iteratively determines the current density and potential distribution, and then determines a lithium plating potential as a separate non-iterative step after the iterative process is done. By determining the lithium plating potential as a separate step after the iterative process, a very large computational cost is avoided, which provides a more efficient computational process for implementing the lithium battery model. Further, by providing a model that addresses multiple particle electrodes rather than representing each electrode as a single particle, the model is much more accurate than models representing electrodes as single particles, providing a much more reliable lithium plating potential determination.
The elements illustrated in
The data processing discussed herein is also discussed in a manner and organization intended to be exemplary, and it not intended to be limiting. For example, although an exemplary process is described in which data is retrieved from a battery 116 and processed by battery management system 112, the data may be retrieved by, processed in whole or in part, and transmitted in raw or processed form between different machines, servers and systems, modules and sub-modules, whether or not illustrated in
Battery-powered system 110 may implement a system or product that utilizes a battery. Examples of a battery-powered system 110 include an electronic vehicle, mobile phone, computer, or some other device that utilizes a battery. Battery-powered system 110 includes battery management system 112, charge control 114, battery 116, and load 118. Battery-powered system 110 may receive a charge for battery 116 from battery charging source 120. The charge provided by source 120 may be received by charge control 114, which may then apply the charge the battery 116. In some instances, charge control 114 may communicate with battery management system 112 regarding how to apply a charge to better 116. For example, battery management system 112 may specify to charge control 114 a C-rate at which battery 116 may be charged, including the voltage and current at which to charge the battery 116. Battery management system may determine the voltage and current at which battery 116 should be charged based on a default voltage and current or customize voltage and current based on battery conditions detected or determined to exist by battery modeling. Load 118 may include one or more loads internal to or external to battery-powered system 110 to which battery 116 is to provide power. More details for battery 116 are discussed with respect to
BMS 112 may be implemented as hardware and/or software that controls and measures batter 114, and controls charging of battery 114 on system 110. BMS may include logic, modules, and components to provide a multiple particle reduce order model of battery 116. The battery model may be used to determine lithium plating potential in real time such that lithium plating in battery 116 can be detected and steps may be taken to reduce any such plating in the future. More detail for BMS 112 are discussed with respect to
Battery charging source 120 may include any suitable source of charge for charging a battery 114. In some instances, in the case of a system 110 implemented as an electronic vehicle, battery charging source 120 may be a dealership, charging pump, or a power outlet commonly found in a home, business or other building. When system 110 is implemented as a phone or computer, a suitable battery charging source 120 may include a mobile charging pack, car charger, or power outlet found in a home, business or other building.
Battery modeling 530 may model a battery 116 of a battery-powered system. The battery modeling may utilize a multi-particle reduce order model to provide accurate modeling for the battery within a system in real time. The battery model may receive inputs of applied electrical load and ambient temperature, and may output cell voltage, temperature, electric potential distribution including electrode potential lithium plating potential, and the concentration distribution inside the battery cell. The ambient temperature may be measured and provided, or in some instances may be predicted and then provided to the model. The prediction can involve, in some instances, thermal energy balancing techniques. Battery modeling 530 may iteratively determine a current density and potential distribution, and then use that information to determine the lithium plating potential. Battery modeling 530 may also communicate with charge manager 510 to indicate that lithium plating exists within the battery 116. In response, charge manager 510 may adjust a charging process of battery 116 to set a voltage and current during charge to minimize or eliminate further lithium plating. More detail for battery modeling 530 is discussed with respect to
The elements of BMS 112 may be implemented as software modules stored in memory and executed by one or more processors, hardware components, or a combination of these. Further, the elements listed and BMS 112 are exemplary, and more or fewer elements may be implemented to perform the functionality described herein.
A battery may be modeled using a reduced order model at step 730. The model may implement a multiple particle reduce order model, which saves considerable computational resources by iterating a current density and potential distribution iteratively, while determining a lithium plating potential as a separate step after the iterator process is complete. More detail for modeling a battery using a reduced order model is discussed with respect to the method of
A determination is made as whether a lithium plating potential that indicates the presence of lithium plating is detected at step 740. In some instances, a lithium plating potential having a value of less than zero indicates that lithium plating has occurred. If the lithium plating potential indicates the presence of lithium plating, a modified charging protocol is applied to a battery in order to reduce lithium plating at step 750. In some instances, a charging process to reduce lithium plating may involve applying a much lower charging rates to the battery, such as C/50. If, at step 740, no lithium plating is detected based on the lithium plating potential, a typical charging protocol may be applied at step 760.
A prescribed electrical load and ambient temperature are applied to the load of the battery model at step 830. The load is determined by an actual load 118 applied to actual battery 116 in the system of
The current density and potential distribution for the battery are iteratively determined at step 840. For each time step, the current density distribution and potential distributions, including electrode potential and the electrolyte potential, are determined in an iterative manner. Iteratively determining the current density and potential distributions are discussed in more detail with respect to the method of
A lithium-ion plating potential is calculated at step 850. In some instances, a lithium-ion plating material is determined after the iterative calculations are complete. Lithium-ion plating potential can be estimated as a function of one or more of the electrode potential φs, electrolyte potential φe, current i, and a resistance of a solid electrolyte interphase (SEI) film Rfilm formed within the battery cell. In some instances, lithium-ion plating potential can be determined as follows:
φLi=φs−φe−iRfilm.
A cell voltage based on the current distribution may then be determined at step 860. A lithium-ion distribution in electrolyte and particles can be determined based on the current distribution at step 870. A thermal energy balance equation for this model battery cell can be solved at step 880, and steps 820-880 can be repeated until any user conditions are met, if any.
The components shown in
Mass storage device 1030, which may be implemented with a magnetic disk drive, an optical disk drive, a flash drive, or other device, is a non-volatile storage device for storing data and instructions for use by processor unit 1010. Mass storage device 1030 can store the system software for implementing embodiments of the present invention for purposes of loading that software into main memory 1020.
Portable storage device 1040 operates in conjunction with a portable non-volatile storage medium, such as a floppy disk, compact disk or Digital video disc, USB drive, memory card or stick, or other portable or removable memory, to input and output data and code to and from the computer system 1000 of
Input devices 1060 provide a portion of a user interface. Input devices 1060 may include an alpha-numeric keypad, such as a keyboard, for inputting alpha-numeric and other information, a pointing device such as a mouse, a trackball, stylus, cursor direction keys, microphone, touch-screen, accelerometer, and other input devices. Additionally, the system 1000 as shown in
Display system 1070 may include a liquid crystal display (LCD) or other suitable display device. Display system 1070 receives textual and graphical information and processes the information for output to the display device. Display system 1070 may also receive input as a touch-screen.
Peripherals 1080 may include any type of computer support device to add additional functionality to the computer system. For example, peripheral device(s) 1080 may include a modem or a router, printer, and other device.
The system of 1000 may also include, in some implementations, antennas, radio transmitters and radio receivers 1090. The antennas and radios may be implemented in devices such as smart phones, tablets, and other devices that may communicate wirelessly. The one or more antennas may operate at one or more radio frequencies suitable to send and receive data over cellular networks, Wi-Fi networks, commercial device networks such as a Bluetooth device, and other radio frequency networks. The devices may include one or more radio transmitters and receivers for processing signals sent and received using the antennas.
The components contained in the computer system 1000 of
The foregoing detailed description of the technology herein has been presented for purposes of illustration and description. It is not intended to be exhaustive or to limit the technology to the precise form disclosed. Many modifications and variations are possible in light of the above teaching. The described embodiments were chosen to best explain the principles of the technology and its practical application to thereby enable others skilled in the art to best utilize the technology in various embodiments and with various modifications as are suited to the particular use contemplated. It is intended that the scope of the technology be defined by the claims appended hereto.
