GENERATION OF A PREDICTION OF INTERNAL SHORT CIRCUITS

Abstract

Techniques are provided for generating a prediction of an internal short circuit in a battery cell.According to one embodiment, the techniques include receiving feature measurements from a plurality of cells, determining a snapshot moving average of the feature measurement of each of the plurality of cells based on a corresponding measurement window, determining a health indicator of each of the plurality of cells based on the respective snapshot moving averages, classifying a plurality of health indicators of the respective plurality of cells, wherein the plurality of health indicators are classified according to size, determining a short-circuit threshold based on a second-rank health indicator, and generating a prediction of a short circuit in a cell corresponding to the first-rank health indicator after determining that a first-rank health indicator exceeds the short-circuit threshold.

Description

INTRODUCTION

[0001] The present disclosure relates to battery cell prognosis and, more particularly, to the prediction of an internal short circuit of a battery cell.

[0002] An electric vehicle or hybrid electric vehicle uses a battery pack to power the motor and auxiliary electrical systems. The battery pack contains multiple interconnected cells that can develop internal short circuits. An internal short circuit can bypass or reduce a cell's designed resistance and cause a sudden increase in current through a shorted electrical path, which can lead to reduced performance or cell failure. SUMMARY

[0003] According to an exemplary embodiment, a method for predicting a circuit state of a power system of a vehicle is provided.The method includes receiving feature measurements of a plurality of cells of the power system, determining a snapshot moving average of the feature measurement of each of the plurality of cells based on a corresponding measurement window, determining a health indicator of each of the plurality of cells based on the respective snapshot moving averages, ranking a plurality of health indicators of the respective plurality of cells, the plurality of health indicators being ranked by size, determining a short circuit threshold based on a second rank health indicator, and generating a prediction of a short circuit in a cell corresponding to the first rank health indicator upon determining that a first rank health indicator exceeds the short circuit threshold.

[0004] In addition to one or more of the features described herein, the feature measurements include a voltage, current, temperature, pressure measurement, or gas venting value; and the health indicators include a voltage drop, an estimated internal short circuit resistance, a solid electrolyte interface thickness, a state of charge, a cell capacity, or a lithium plating index.

[0005] In addition to one or more of the features described herein, a size of the measurement window is based on an expected short-circuit resistance value of the majority of cells.

[0006] In addition to one or more of the features described herein, a size of the measurement window is constant, wherein data of the measurement window is input to a parallel, staggered data buffer, and wherein the data buffer is processed at time intervals based on an expected short-circuit resistance value of the plurality of cells.

[0007] In addition to one or more of the features described herein, the feature measurements are determined during a relaxation period of the plurality of cells.

[0008] In addition to one or more of the features described here, the feature measurements are determined from cells balanced between cells and the feature measurements are reversed to unbalanced between cells via voltage-state-of-charge conversions.

[0009] In addition to one or more of the features described herein, the plurality of health indicators are ranked in ascending or descending order, wherein the first-ranked health indicator indicates a potential short circuit of a cell corresponding to the first-ranked health indicator and wherein the second-ranked health indicator indicates an absence of a short circuit of a cell corresponding to the second-ranked health indicator.

[0010] According to another exemplary embodiment, a system for predicting a circuit state of a power system of a vehicle is provided. The system includes a processor and memory or storage comprising an algorithm or computer instructions that, when executed by the processor, perform an operation including receiving feature measurements of a plurality of cells of the power system, determining a snapshot moving average of the feature measurement of each of the plurality of cells based on a corresponding measurement window, determining a health indicator of each of the plurality of cells based on the respective snapshot moving averages, ranking a plurality of health indicators of the respective plurality of cells, wherein the plurality of health indicators are ranked by size,determining a short circuit threshold based on a second-rank health indicator and generating a prediction of a short circuit in a cell corresponding to the first-rank health indicator after determining that a first-rank health indicator exceeds the short circuit threshold.

[0011] In addition to one or more of the features described herein, the feature measurements include a voltage, current, temperature, pressure measurement, or gas venting value; and the health indicators include a voltage drop, an estimated internal short circuit resistance, a solid electrolyte interface thickness, a state of charge, a cell capacity, or a lithium plating index.

[0012] In addition to one or more of the features described herein, a size of the measurement window is based on an expected short-circuit resistance value of the majority of cells.

[0013] In addition to one or more of the features described herein, a size of the measurement window is constant, wherein data of the measurement window is input to a parallel, staggered data buffer, and wherein the data buffer is processed at time intervals based on an expected short-circuit resistance value of the plurality of cells.

[0014] In addition to one or more of the features described herein, the feature measurements are determined during a relaxation period of the plurality of cells.

[0015] In addition to one or more of the features described herein, the feature measurements are determined from cells balanced between cells, and the feature measurements are reversed to unbalanced between cells via voltage-state-of-charge conversions.

[0016] In addition to one or more of the features described herein, the plurality of health indicators are ranked in ascending or descending order, wherein the first-ranked health indicator indicates a potential short circuit of a cell corresponding to the first-ranked health indicator, and wherein the second-ranked health indicator indicates an absence of a short circuit of a cell corresponding to the second-ranked health indicator.

[0017] According to yet another exemplary embodiment, a computer-readable storage medium having computer-readable program code embodied therewith for predicting a circuit state of a power system of a vehicle is provided. The computer-readable program code is executable by one or more computer processors to perform an operation including receiving feature measurements of a plurality of cells of the power system, determining a snapshot moving average of the feature measurement of each of the plurality of cells based on a corresponding measurement window, determining a health indicator of each of the plurality of cells based on the respective snapshot moving averages, ranking a plurality of health indicators of the respective plurality of cells, wherein the plurality of health indicators are ranked by size,determining a short circuit threshold based on a second-rank health indicator and generating a prediction of a short circuit in a cell corresponding to the first-rank health indicator after determining that a first-rank health indicator exceeds the short circuit threshold.

[0018] In addition to one or more of the features described herein, the feature measurements include a voltage, current, temperature, pressure measurement, or gas venting value; and the health indicators include a voltage drop, an estimated internal short circuit resistance, a solid electrolyte interface thickness, a state of charge, a cell capacity, or a lithium plating index.

[0019] In addition to one or more of the features described herein, a size of the measurement window is based on an expected short-circuit resistance value of the majority of cells.

[0020] In addition to one or more of the features described herein, a size of the measurement window is constant, wherein data of the measurement window is input to a parallel, staggered data buffer, and wherein the data buffer is processed at time intervals based on an expected short-circuit resistance value of the plurality of cells.

[0021] In addition to one or more of the features described herein, the feature measurements are determined from cells balanced between cells, and the feature measurements are reversed to unbalanced between cells via voltage-state-of-charge conversions.

[0022] In addition to one or more of the features described herein, the plurality of health indicators are ranked in ascending or descending order, wherein the first-ranked health indicator indicates a potential short circuit of a cell corresponding to the first-ranked health indicator, and wherein the second-ranked health indicator indicates an absence of a short circuit of a cell corresponding to the second-ranked health indicator.

[0023] The above features and advantages and other features and advantages of the disclosure will be readily apparent from the following detailed description when taken in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Further features, advantages and details appear only as examples in the following detailed description, which refers to the drawings; they show: Fig. 1 a vehicle according to an embodiment; Fig. 2 a computing environment according to an embodiment; Fig. 3 shows a flowchart of a method for determining a health indicator according to an embodiment; and Fig. 4 is a flowchart of a method for predicting an internal short circuit based on a health indicator according to an embodiment. DETAILED DESCRIPTION

[0025] The following description is merely exemplary in nature and is not intended to limit the present disclosure, its application, or uses. It should be understood that corresponding reference characters designate like or corresponding parts and features throughout the drawings. As used herein, the term "module" may refer to one or more algorithms, instruction sets, software applications, or other computer-readable program code executable by a processor to perform the functions, operations, or processes described herein.

[0026] Embodiments of the present disclosure improve battery systems by predicting internal short circuits in battery cells and taking preventative measures to mitigate potential damage caused by the internal short circuits. According to one embodiment, a prediction module uses a snapshot of moving averages of characteristic measurements (e.g., a voltage, a current, a temperature, a pressure measurement, a gas vent value, or the like) of cells to determine health indicators (e.g., a voltage drop, an estimated resistance value of an internal short circuit, a solid electrolyte interface (SEI) thickness, a state of charge (SOC), a cell capacity, or the like) of the cells. The prediction module may rank the health indicators and determine a short circuit threshold based on the ranking.If a cell's health indicator exceeds the short-circuit threshold, the prognostic module can predict the presence or development of a short circuit in the cell. The prognostic module can then notify a vehicle user of the short circuit or control the vehicle's operation to mitigate the short circuit's impact on the cell.

[0027] One advantage of the disclosed embodiments is the improvement of signal-to-noise ratios for determining cell health indicators in electrically noisy environments, which can aid in diagnosing or predicting cell problems. Furthermore, embodiments of the present disclosure can prevent or mitigate the impact of using a shorted cell, which can cause reduced cell performance or cell failure.

[0028] Fig.1 shows a vehicle 100 according to one embodiment. The vehicle 100 includes a body 102 that can support a power system 104, a sensor system 106, a propulsion system 120, a controller 140, and other systems of the vehicle 100 described herein.

[0029] According to one embodiment, vehicle 100 is an internal combustion engine (ICE) vehicle, an electric vehicle (EV), or a hybrid electric vehicle (HEV). According to the illustrative embodiment, vehicle 100 is an HEV powered in part by power system 104, which includes a plurality of interconnected battery cells. Power system 104 may be electrically coupled to at least one electric motor assembly of propulsion system 120. According to one embodiment, power system 104 is electrically coupled to a direct current (DC) converter module 108 (e.g., a DC-to-DC converter) and to an inverter module 110 (e.g., a traction power conversion unit). The inverter module 110 may include a plurality of inverters that convert DC signals from the power system 104 into three-phase alternating current (AC) signals for driving the electric motors of the propulsion system 120.Additionally, the power system 104 may be electrically coupled to vehicle electronic systems such as audio systems, display systems, navigation systems, temperature control systems, or the like.

[0030] The sensor system 106 includes a plurality of sensors disposed on or integrated with various components of the system 100. According to one embodiment, the sensor system 106 is communicatively coupled to the power system and to the controller 140 to transmit measurements from the power system 104 to the controller 140. The sensor system 106 may include a current sensor, a voltage sensor, a temperature sensor, or the like.

[0031] The propulsion system 120 may include an ICE system 122 and at least one electric motor assembly (e.g., a first electric motor 124 and a second electric motor 126). Each component of the propulsion system 120 may be configured to drive at least one of the wheels 130 of the vehicle 100 via a transmission system coupled to a front axle shaft or a rear axle shaft coupled to a front or rear set of wheels 130, respectively.

[0032] According to one embodiment, the controller 140 is configured to analyze outputs of the power system 104 and generate a prediction of an internal short circuit of the power system 104. The controller is in Fig. 2. Techniques used to analyze the outputs of the performance system 104 are described in Fig. 3. Techniques used to generate the short circuit prediction are discussed in Fig. 4 discussed.

[0033] Fig. 2 illustrates a computing environment 200 according to one embodiment. According to the illustrated embodiment, the computing environment 200 includes a controller 140, a network 230, and a cell 240.

[0034] According to one embodiment, controller 140 includes a processor 202 that receives instructions and data from memory 204 or storage 208 via a bus 222. Not all components of controller 140 are shown. Generally, controller 140 is under the control of an operating system (OS) capable of performing or supporting the functions or processes disclosed herein. Processor 202 is a programmable logic device that performs instruction, logic, and mathematical processing and may represent one or more CPUs. The processor may execute one or more algorithms, instruction sets, or applications in memory 204 or storage 208 to perform the functions or processes described herein.

[0035] Memory 204 and storage 208 may represent hard disk drives, solid-state drives, flash memory devices, optical media, and the like. Storage 208 may also include structured storage (e.g., a database). Additionally, memory 204 and storage 208 may be considered to include storage physically located elsewhere. For example, memory 204 and storage 208 may be physically located in another computer communicatively coupled to controller 140 via bus 222 or network 230.

[0036] The controller 140 may be connected to other computers (e.g., controllers, distributed databases, servers, or web hosts) or to the cell 240 via a network interface 220 and the network 230. Examples of the network 230 include a controller area network (CAN), a transmission control protocol (TCP) bus, electrical buses, physical transmission cables, optical transmission fibers, wireless transmission media, routers, firewalls, switches, gateway computers, edge servers, a local area network, a wide area network, a wireless network, or the like. The network interface 220 may be any type of network communication device that enables the controller 140 to communicate with computers and other components of the computing environment 200 via the network 230.

[0037] According to the illustrated embodiment, memory 204 includes a prediction module 206. Prediction module 206 represents one or more algorithms, instruction sets, software applications, or other computer-readable program code executable by processor 202 to perform the functions, operations, or processes described herein.

[0038] According to one embodiment, the prognostic module 206 receives feature measurements of the power system 104 from the sensor system 106 coupled to the power system 104. Examples of feature measurements include a voltage, a current, a temperature, a resistance, a pressure, a gas vent value, or the like. The prognostic module 206 may store the feature measurements as cell data 212 in the storage memory 208.

[0039] According to the illustrated embodiment, cell 240 is one of the cells included in power system 104. The circuit configuration shown in cell 240 represents an example of a battery cell equivalent circuit model that is common for the cells of power system 104. According to the illustrated embodiment, cell 240 includes a voltage source 242 connected to a first end 244 of first internal resistor 246. According to one embodiment, voltage source 242 carries an open circuit voltage (V OC ). A second end 248 of the first internal resistor 246 is connected to a measuring point 250, to a load (not shown), and to a first connection point 252 of a short-circuit resistor 254. A second connection point 256 of the short-circuit resistor is connected to the voltage source 242 and to the load.

[0040] The sensor system may use the measurement point 250 as a connection for measuring a characteristic (e.g., a voltage drop or a resistance) corresponding to the short-circuit resistance 254. According to one embodiment, a short circuit developing in the cell 240 reduces the resistance of the short-circuit resistance 254. As the resistance of the short-circuit resistance 254 approaches 0 ohms, a larger excess current may flow in the electrical path containing the short-circuit resistance 254.

[0041] According to one embodiment, the storage memory 208 V OC 2SOC data 210 such as an open circuit voltage state of charge table (V OC-SOC table), an open circuit voltage-state of charge map or an open circuit voltage-state of charge graph, which can be used to convert a voltage or current at the measurement point 250 into an SOC value and vice versa. The prediction module 206 can use the feature measurements of the cell 240 or the V OC 2Use SOC data 210 to determine a health indicator of the power system 104 and predict an internal short circuit based on the health indicator. This process is described in Fig. 3 and Fig. 4 further discussed.

[0042] Fig. 3 illustrates a flowchart of a method 300 for determining a health indicator according to one embodiment. The method 300 begins in block 302.

[0043] In block 304, the prognostic module 206 receives characteristic measurements of a plurality of cells. As previously discussed, the sensor system 106 may be coupled to the power system 104 to measure the characteristics of the cells of the power system 104. Examples of the characteristic measurements include: a voltage, a current, a temperature, a pressure, a gas vent value, or the like. According to one embodiment, the sensor system 106 measures the characteristics at the measurement point 250.

[0044] At block 306, the prediction module 206 determines a snapshot moving average of the feature measurement value of each of the plurality of cells based on a corresponding measurement window. According to one embodiment, the measurements of the sensor system 106 include electrical noise that may affect a measurement output signal at different times. For example, a voltage signal may have an average measurement value of 5 V, but sampling the signal at a first time (e.g., at 0.01 seconds) may yield a measurement value of 4.9 V, while sampling the signal at a second time (e.g., at 0.02 seconds) may yield a measurement value of 5.1 V.

[0045] Thus, the prediction module 206 may form a moving average of the feature measurement value over a snapshot time period (i.e., a measurement window), which may be centered at periodic time intervals, to accurately determine a feature measurement value of a cell. For example, assuming that the feature measurement value is a voltage output of the cell 240 (represented as V(i), where i represents an iterative count corresponding to a time range), and assuming that the time range is 5 seconds, and assuming that the measurement window size extends over 1 second, after a period of the first 5 seconds of signal output, the prediction module 206 would determine V(1) as the moving average of V(i) from 4.5 seconds to 5.5 seconds. In the next time range, the prediction module 206 would determine V(2) as the moving average of V(i) from 9.5 seconds to 10.5 seconds, and so on.

[0046] According to one embodiment, the prediction module 206 determines the size of the measurement window to apply to the feature measurement based on an expected short-circuit resistance value of the short-circuit resistor 254. As previously discussed, a larger excess of current may flow in the electrical path containing the short-circuit resistor 254 as the resistance value of the short-circuit resistor 254 approaches 0 ohms. Thus, the prediction module 206 may use smaller measurement window durations to more quickly identify the presence or development of the short circuit as the resistance value of the short-circuit resistor 254 decreases.

[0047] According to another embodiment, the prediction module 206 is configured to use predetermined measurement window sizes according to the following calibration guidelines. Short-circuit resistance (R S ) R S < 10 Ohm 10 ≤ R S < 100 Ohm 100 ≤ R S ≤ 1000 Ohm Measurement window size range 0 - 200 - seconds 200 + - 500 - 500 + - 7200 seconds recommended measuring window size 10 seconds 200 seconds 600 seconds

[0048] According to yet another embodiment, the measurement window size remains constant, and the feature measurements covered by the measurement window are input into a parallel, offset data buffer. In this way, the prediction module 206 can identify low short-circuit resistance values ​​even with large measurement window sizes (e.g., 600 seconds), as long as the buffers are offset in time ranges corresponding to the measurement window size ranges of an expected short-circuit resistance value (e.g., 0-200 seconds for R S < 10 Ohm).

[0049] At block 308, the prognostic module 206 determines a health indicator for each of the plurality of cells based on the respective snapshot moving average values. Examples of the health indicator include a voltage drop, an estimated internal short circuit resistance, a solid electrolyte interface (SEI) thickness, a state of charge, a cell capacity, a lithium plating index, or the like.

[0050] According to one embodiment, the prediction module 206 determines the health indicator as a difference between two snapshot moving averages of consecutive time periods. For example, continuing the example from block 306, a voltage drop health indicator may be determined as V(2)-V(1).

[0051] According to one embodiment, changes in feature measurements may be difficult to detect when a state of charge (SOC) of cell 240 reaches 35%-45%. Thus, the prognostic module 206 may determine a difference in health indicators by first pausing external charging or discharging of cell 240. Without external charging or discharging, cell 240 enters a relaxation period during which chemical reactions, ion migration, polarization effects, and the like of cell 240 cause compensating effects until cell 240 reaches equilibrium. During the relaxation period, the prognostic module 206 may use ongoing feature measurements to determine the health indicator. The cell's relaxation time may also be used as a health indicator to predict a short circuit of cell 240.

[0052] According to one embodiment, the prediction module 206 may calculate the health indicator using the V OC2SOC data 210. For example, after determining a snapshot moving average of a feature measurement, the prediction module 206 may determine the SOC of the cell 240 from a V OC 2SOC table, from a V OC 2SOC card or from a V OC Identify 2SOC graphs using V(1) or V(2) as an index.

[0053] According to one embodiment, the prognostic module 206 determines unbalanced feature measurements to determine the health indicators when the power system 104 is undergoing cell balancing. To remove an influence of current discharge on feature measurements (e.g., a voltage or SOC) of the balanced cell, the prognostic module 206 may invert balanced feature measurements into unbalanced feature measurements by bypassing a resistance value of a balanced cell. The prognostic module 206 may then measure the unbalanced features via the sensor system 106. Afterward, the prognostic module 206 may compare the unbalanced features to the balanced features between cells using the VOC2SOC conversions.

[0054] For example, after discharging the cell, the prediction module 206 may determine a balanced SOC of the cell 240 as follows: SOC B (i) = V OC 2SOC (VB (i) + p), (1) where SOC B (i) represents the balanced SOC of the cell 240 in an i-th time range, V OC 2SOC a conversion function based on the V OC 2SOC data 210 represents V B (i) represents a measured cell-to-cell balanced voltage of cell 240 and p represents a polarization correction (e.g., a voltage drop across the first internal resistor 246).

[0055] Thereafter, the prediction module 206 may use the balanced SOC of the cell 240 to determine the unbalanced SOC as follows: SOC(i) = SOC B (i) - ΔSOC B (i), (2) where SOC(i) represents the unbalanced SOC of cell 240, SOC B (i) is determined in equation 1 and Δ SOC B (i) represents a change in the balanced SOC of the cell 240 in an i-th time range. According to one embodiment, Δ SOC B(i) be determined as the measured balanced amperes per hour (“A(i) / h”) of cell 240 divided by the capacity of cell 240.

[0056] Finally, the prediction module 206 may use the unbalanced SOC of the cell 240 to determine the unbalanced feature measurement value of the cell 240 as follows: V(i) = V OC 2SOC (SOC(i)) - p, (3) where V(i) represents the unbalanced voltage reading of cell 240, SOC(i) is determined in equation 2, V OC 2SOC an implementation function based on the V OC 2SOC data 210 and p represents a polarization correction.

[0057] The forecast module 206 may then use V(i) to determine health indicators as previously described.

[0058] According to another embodiment, the prediction module 206, as in Fig.4, set a short-circuit threshold to account for the voltage balanced between cells. The method 300 ends in block 310.

[0059] Fig. 4 illustrates a flowchart of a method 400 for predicting an internal short circuit based on a health indicator, according to one embodiment. The method 400 begins in block 402.

[0060] In block 404, the prediction module 206 ranks a plurality of health indicators of a respective plurality of cells, wherein the plurality of health indicators are ranked by size. The health indicators may be ranked in ascending or descending order. As previously described, examples of the health indicator include a voltage drop, an estimated internal short-circuit resistance value, an SEI thickness, a state of charge, a cell capacity, a lithium plating index, or the like.

[0061] According to one embodiment, an indicator of rank n represents the health indicator with the nth largest magnitude in the ranking and corresponds to the nth cell of the plurality of cells. For example, the first-rank health indicator may represent a largest voltage drop corresponding to a first cell; and a second-rank indicator may represent a second-largest voltage drop corresponding to a second cell.

[0062] At block 406, the prediction module 206 determines a short circuit threshold based on a second-rank health indicator. According to one embodiment, the prediction module 206 assumes that the second-rank health indicator indicates an absence of an internal short circuit of the second cell, while the first-rank health indicator may indicate an internal short circuit of the first cell. According to another embodiment, the first-rank and second-rank health indicators may simultaneously indicate internal short circuits of the first and second cells, while a third-rank health indicator indicates an absence of an internal short circuit of a third cell.

[0063] According to yet another embodiment, the prediction module 206 may determine whether an nth-rank health indicator deviates from an average of the smaller-sized health indicator sizes. If the nth-rank health indicator deviates from this average, the prediction module 206 determines that the nth-rank health indicator may indicate an internal short circuit of the nth-rank cell, and the smaller-sized health indicators do not indicate an internal short circuit of the corresponding cells. Under these circumstances, the prediction module 206 may determine the short circuit threshold based on the nth-rank health indicator.

[0064] According to one embodiment, the prediction module 206 determines the short-circuit threshold based on demonstrated empirical evidence to minimize false alarms. For example, the short-circuit threshold may be set to 1.6 * the second-rank health indicator (or the rank (n + 1) indicator), which may reduce false positives in the short-circuit predictions of cell 240.

[0065] As previously discussed, according to one embodiment, the prediction module 206 may set a short-circuit threshold to account for cases where the majority of cells are subject to cell balancing. In such cases, the prediction module 206 may adjust the threshold by decreasing the short-circuit threshold by a difference between the unbalanced voltage (V(i)) and a balanced voltage (V B (i)).

[0066] At block 408, the prediction module 206 determines whether a first-rank health indicator exceeds the short-circuit threshold. Upon determining that the first-rank health indicator does not exceed the short-circuit threshold, the method 400 proceeds to block 404, where it is executed as described above. Conversely, upon determining that the first-rank health indicator exceeds the short-circuit threshold, the method 400 proceeds to block 410.

[0067] In block 410, the prediction module 206 generates a prediction of a short circuit in a cell corresponding to the first-rank health indicator. If the first-rank health indicator does not exceed the short circuit threshold, the prediction module 206 determines that no short circuit is present in the first cell. Conversely, if the first-rank health indicator exceeds the short circuit threshold, the prediction module 206 determines that a short circuit is present or developing in the first cell. According to one embodiment, the prediction module 206 may also predict degraded capacity of cell 240 if a voltage-based health indicator of cell 240 is greater than the voltage-based health indicators of other cells of the power system 104.

[0068] After generating the prediction of a short circuit in the first cell, the prognostic module 206 may take action to mitigate the effects of the short circuit or prevent the short circuit from developing further. For example, the prognostic module 206 may generate and send a warning about the short circuit to a user of the vehicle 100 or to a repair facility so that the affected cell can be repaired or replaced. Furthermore, the prognostic module 206 may control the operation of the vehicle 100 to mitigate the impact of the short circuit on the cell (e.g., to prevent the cell from overheating). At block 412, the method 400 ends.

[0069] The terms "a" and "an" do not imply a limitation on the quantity, but rather denote the presence of at least one of the mentioned objects. Unless the context clearly indicates otherwise, the term "or" means "and / or." Reference throughout the specification to "an aspect" means that a particular element described in connection with the aspect (e.g., feature, structure, step, or property) is included in at least one aspect described herein and may or may not be included in other aspects. It is also understood that the described elements according to the various aspects may be combined in any suitable manner.

[0070] When an element, such as a layer, film, region, or substrate, is referred to herein as being "on" another element, it may be directly on top of the other element or may also have intervening elements present. In contrast, when an element is referred to as being "directly on" another element, no intervening elements are present.

[0071] Unless otherwise specified herein, all examination standards are the most recent examination standards in effect as of the filing date of this application or the filing date of the earliest priority application in which the examination standard appears, if priority is claimed.

[0072] Unless otherwise defined, technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs.

[0073] Although the above disclosure has been described with reference to exemplary embodiments, those skilled in the art will understand that various changes may be made and equivalents may be substituted for elements thereof without departing from the scope thereof. Furthermore, many changes may be made to adapt a particular situation or material to the teachings of the disclosure without departing from its essential scope. Thus, the present disclosure is not intended to be limited to the particular embodiments disclosed, but is intended to include all embodiments falling within its scope. legend

[0074] In the drawings, N stands for No and Y stands for Yes.

Claims

[1] A system for predicting a circuit state of a power system of a vehicle, comprising: a processor; and Memory or storage comprising an algorithm or computer instructions which, when executed by the processor, perform an operation which includes: Receiving feature measurements of a plurality of cells of the power system; determining a snapshot moving average of the feature measurement value of each of the plurality of cells based on a corresponding measurement window; determining a health indicator of each of the plurality of cells based on the respective snapshot moving average values; Classifying a plurality of health indicators of the respective plurality of cells, wherein the plurality of health indicators are classified according to size; Determining a short-circuit threshold based on a second-rank health indicator; and Generating a prediction of a short circuit in a cell corresponding to the first-rank health indicator after determining that a first-rank health indicator exceeds the short circuit threshold. [2] The system of claim 1, wherein the feature measurements include a voltage, a current, a temperature, a pressure measurement, or a gas vent value; and wherein the plurality of health indicators include a voltage drop, an estimated internal short circuit resistance, a solid electrolyte interface thickness, a state of charge, a cell capacity, or a lithium plating index. [3] The system of claim 1, wherein a size of the measurement window is based on an expected short-circuit resistance value of the plurality of cells. [4] The system of claim 1, wherein a size of the measurement window is constant, wherein data of the measurement window is input to a parallel, staggered data buffer, and wherein the data buffer is processed at time intervals based on an expected short-circuit resistance value of the plurality of cells. [5] The system of claim 1, wherein the feature measurements are determined during a relaxation period of the plurality of cells. [6] The system of claim 1, wherein the feature measurements are determined from cells balanced between cells, and wherein the feature measurements are inverted to unbalanced values between cells via voltage-state-of-charge conversions. [7] The system of claim 1, wherein the plurality of health indicators are ranked in ascending or descending order, wherein the first rank health indicator indicates a potential short circuit of a cell corresponding to the first rank health indicator, and wherein the second rank health indicator indicates an absence of a short circuit of a cell corresponding to the second rank health indicator. [8] A computer-readable storage medium having computer-readable program code embodied therewith for predicting a circuit state of a power system of a vehicle, the computer-readable program code being executable by one or more computer processors to perform an operation comprising: Receiving feature measurements of a plurality of cells of the power system; determining a snapshot moving average of the feature measurement value of each of the plurality of cells based on a corresponding measurement window; determining a health indicator of each of the plurality of cells based on the respective snapshot moving average values; Classifying a plurality of health indicators of the respective plurality of cells, wherein the plurality of health indicators are classified according to size; Determining a short-circuit threshold based on a second-rank health indicator; and Generating a prediction of a short circuit in a cell corresponding to the first-rank health indicator after determining that a first-rank health indicator exceeds the short circuit threshold. [9] The computer-readable storage medium of claim 8, wherein the feature measurements include a voltage, a current, a temperature, a pressure measurement, or a gas vent value; and wherein the plurality of health indicators include a voltage drop, an estimated internal short circuit resistance, a solid electrolyte interface thickness, a state of charge, a cell capacity, or a lithium plating index. [10] The computer-readable storage medium of claim 8, wherein the plurality of health indicators are ranked in ascending or descending order, wherein the first-rank health indicator indicates a potential short circuit of a cell corresponding to the first-rank health indicator, and wherein the second-rank health indicator indicates an absence of a short circuit of a cell corresponding to the second-rank health indicator.

Images