Method and system for estimating internal resistance of a battery of a vehicle in real time

Abstract

Disclosed herein is method and system for estimating internal resistance of a battery of a vehicle in real time. The method comprises receiving battery related data during charging and discharging of the battery, wherein the battery related data is extracted from latest diagnostic load collective data of the vehicle. The method further comprises processing the battery related data to generate at least one charge-discharge cycle for an unseen period of time. Based on the generated at least one charge-discharge cycle, the method comprises estimating capacity of the battery by using a battery capacity estimator model. Lastly, based on estimated capacity of the battery, the method comprises estimating the internal resistance of the battery, by using a battery resistance estimator model.

Description

InternalInternal Ref.: 2024PF01617METHOD AND SYSTEM FOR ESTIMATING INTERNAL RESISTANCE OF A BATTERY OF A VEHICLE IN REAL TIMETECHNICAL FIELD

[0001] The present invention generally relates to the field of automobiles, and more5 particularly relates to a method and system for estimating internal resistance of a battery of a vehicle in real time.BACKGROUND

[0002] The following description includes information that may be useful in understanding the present invention. It is not an admission that any of the information provided herein is10 prior art or relevant to the presently claimed invention, or that any publication specifically or implicitly referenced is prior art.

[0003] Modem battery technology aims to make batteries more efficient and have a longer life. A key factor in the design of battery packs is internal resistance. Internal resistance is a natural property of the battery that slows down the flow of electric current and is made up of the resistance found in the electrolyte, electrodes, and connections inside the battery. In single battery cells, this resistance decides how much energy is lost as heat when the battery charges and discharges. For larger battery packs or systems, like those in electric vehicles (EVs), understanding and managing the internal resistance becomes very important, as it has direct impact on the range, efficiency and thermal management20 of the EV.

[0004] Furthermore, as the battery ages with respect to time, the internal resistance of the battery increases which affects its overall performance. Due to continuous increase in internal resistance various issues may arise. For example, increased internal resistance causes significant voltage drop resulting in energy losses there by decreasing the energy efficiency. It can cause the voltage across the terminals to sag there by leading to unstable operation of load connected devices. It degrades the capacity of the battery. One of the major attributes related to the internal resistance is the temperature sensitiveness and reaction of the battery pack to this change in temperature. When internal resistance of the battery increases, the energy loss of the battery also increases.30 This energy loss is transformed as heat energy which needs to be removed from the battery.

[0005] With aged batteries, the rate of energy loss will be higher than the rate of cooling power in-built in the battery pack. The main reason for this rate to increase is capacity degradation of the battery which is again caused by the increase in the internalInternalInternal Ref.: 2024PF01617 resistance of the battery pack. During charging or discharging of the aged battery packs, if the designed cooling power is not sufficient and if the cell temperature increases beyond a certain threshold, the battery aging will accelerate Thus, excessive heat generation due to high internal resistance is not merely an efficiency concern but also5 a safety issue. Batteries operating at higher temperatures are at an increased risk of thermal runaway, a condition where the battery becomes self-heating, leading to potential fires or explosions. Estimating the internal resistance in advance is, therefore, crucial to ensure the safe operation of the battery pack and safety of human life.

[0006] Few of the existing technologies available in the art which deals which the measurement of internal resistance are JP2005347166A that discloses calculation of internal resistance of a secondary battery based on measured values of the voltage and current, and CN110988723 A that discloses battery internal resistance prediction and fault early warning method based on LSTM.SUMMARY15

[0007] The present disclosure overcomes one or more shortcomings of the prior art and provides additional advantages. Embodiments and aspects of the disclosure described in detail herein are considered a part of the claimed disclosure.

[0008] In one non-limiting embodiment of the present disclosure, a method for estimating internal resistance of a battery of a vehicle in real time, is disclosed. The method20 comprises receiving battery related data during charging and discharging of the battery. The battery related data is extracted from latest diagnostic load collective data of the vehicle. The method further comprises processing the battery related data to generate at least one charge-discharge cycle for an unseen period of time. Based on the generated at least one charge-discharge cycle, the method further comprises estimating capacity of the battery by using a battery capacity estimator model. Moving ahead, based on estimated capacity of the battery, the method further comprises estimating the internal resistance of the battery, by using a battery resistance estimator model.

[0009] In another non-limiting embodiment of the present disclosure, wherein for processing the battery related data to generate the at least one charge -discharge cycles, the method30 further comprising generating a charge-discharge profile for at least one parameter of the battery related data, and wherein the charge-discharge profile mimics performance behaviour of the at least one parameter in real time.InternalInternal Ref.: 2024PF01617

[0010] In yet another non-limiting embodiment of the present disclosure, wherein the battery capacity estimator model is built using the steps comprising collecting a data set relating to vehicle performance. The data set comprises one or more battery related parameters as predictor variables and the capacity of the battery as a response variable, wherein the5 one or more battery related parameters comprise at least one of: voltage, current and temperature during charge or discharge of the battery. The steps further comprise splitting the data set to obtain a training dataset and test dataset, training a first regression model using the training dataset for learning a relationship between the predictor and the response variables, and evaluating performance of the first regression10 model using the test dataset after the training.

[0011] In yet another non-limiting embodiment of the present disclosure, wherein for estimating the capacity of the battery by using the battery capacity estimator model, the method further comprises estimating the capacity of the battery for the at least one charge-discharge cycle using the trained first regression model.

[0012] In yet another non-limiting embodiment of the present disclosure, wherein the internal resistance estimator model is built using the steps comprising collecting a data set relating to vehicle performance, the data set includes the capacity of the battery as predictor variable and the internal resistance of the battery as a response variable. The steps further comprise training a second regression model using the training dataset for20 learning a relationship between the predictor and the response variables, and evaluating performance of the second regression model using the test dataset after the training.

[0013] In yet another non-limiting embodiment of the present disclosure, wherein estimating the internal resistance of the battery, by using a battery resistance estimator model, the method further comprises estimating the internal resistance of the battery for the at least one charge-discharge cycle using the trained second regression model.

[0014] In yet another non-limiting embodiment of the present disclosure, the method further comprising determining if the internal resistance of the battery is greater than a predefined threshold, and generating a warning to driver of the vehicle if the internal resistance is greater than a predefined threshold.30

[0015] In yet another embodiment of the present disclosure, a system for estimating internal resistance of a battery of a vehicle in real time, is disclosed. The system comprises a memory and a processor which is electronically coupled to the memory. The memory stores diagnostic load collective data of the vehicle. The processor receives battery related data during charging and discharging of the battery, wherein the battery relatedInternalInternal Ref.: 2024PF01617 data is extracted from latest diagnostic load collective data of the vehicle. The processor processes the battery related data to generate at least one charge-discharge cycle, wherein the at least one charge-discharge cycle is generated for an unseen period of time. Further, based on the generated the at least one charge-discharge cycle, the5 processor estimates capacity of the battery by using a battery capacity estimator model. Lastly, based on the estimated capacity of the battery, the processor estimates the internal resistance of the battery, by using a battery resistance estimator model.

[0016] In yet another embodiment of the present disclosure, wherein to process the battery related data to generate the at least one charge-discharge cycles, the processor is10 configured to generate a charge-discharge profile for at least one parameter of the battery related data. The profile mimics performance behaviour of the at least one parameter in real time.

[0017] In yet another embodiment of the present disclosure, wherein to build the battery capacity estimator model, the processor collects data set relating to battery performance, the data set comprises one or more battery related parameters as predictor variables and the capacity of the battery as a response variable. The one or more battery related parameters comprises at least one of: voltage, current and temperature during charge or discharge of the battery. The processor further splits the data set to obtain a training dataset and test dataset and trains a first regression model using the training20 dataset for learning a relationship between the predictor and the response variables. Lastly, the processor evaluates performance of the second regression model using the test dataset after the training.

[0018] In yet another embodiment of the present disclosure, wherein to estimate the capacity of the battery by using a battery capacity estimator model, the processor estimates the capacity of the battery for the at least one charge-discharge cycle using the trained first regression model.

[0019] In yet another embodiment of the present disclosure, wherein to build the internal resistance estimator model, the processor collects a data set relating to vehicle performance, the data set includes the capacity of the battery as predictor variable and30 the internal resistance of the battery as a response variable. The processor splits the data set to obtain a training dataset and a test dataset and trains a second regression model using the training dataset for learning a relationship between the predictor and the response variables. Lastly, the processor evaluates performance of the second regression model using the test dataset after the training.InternalInternal Ref.: 2024PF01617

[0020] In one non-limiting embodiment of the present disclosure, a computer-readable storage medium having program instructions stored thereon, is disclosed. The program instructions, when executed by a processor, cause the processor to perform the step of receiving battery related data during charging and discharging of the battery. The5 battery related data is extracted from latest diagnostic load collective data of the vehicle. The program instructions further cause the processor to perform the step of processing the battery related data to generate at least one charge-discharge cycle for an unseen period of time. Based on the generated at least one charge-discharge cycle, the program instructions further cause the processor to perform the step of estimating capacity of10 the battery by using a battery capacity estimator model. Moving ahead, based on estimated capacity of the battery, the program instructions further cause the processor to perform the step of estimating the internal resistance of the battery, by using a battery resistance estimator model.

[0021] The foregoing summary is illustrative only and is not intended to be in any way limiting. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features will become apparent by reference to the drawings and the following detailed description.BRIEF DESCRIPTION OF DRAWINGS

[0022] The features, nature, and advantages of the present disclosure will become more20 apparent from the detailed description set forth below when taken in conjunction with the drawings in which like reference characters identify correspondingly throughout. Some embodiments of system and / or methods in accordance with embodiments of the present subject matter are now described, by way of example only, and with reference to the accompanying Figs., in which:

[0023] FIG. 1 depicts an exemplary environment for estimating internal resistance of a battery of a vehicle in real time, in accordance with embodiments of the present disclosure.

[0024] FIG. 2 depicts an exemplary block diagram illustrating a system for estimating internal resistance of a battery of a vehicle in real time, in accordance with embodiments of the present disclosure.30

[0025] FIG. 3 represents an exemplary block diagram illustrating building of a regressor model, in accordance with embodiments of the present disclosure.

[0026] FIG. 4 represents an exemplary diagram illustrating estimation of battery capacity using a pre-trained Battery Capacity Estimator Model, in accordance with embodiments of the present disclosure.InternalInternal Ref.: 2024PF01617

[0027] FIG. 5 represents an exemplary diagram illustrating estimation of internal resistance of a battery using pre-trained Internal Resistance Estimator Model, in accordance with embodiments of the present disclosure.

[0028] FIG. 6 represents an exemplary diagram illustrating generation of a warning or alert,5 in accordance with embodiments of the present disclosure.

[0029] FIG. 7 represents flowchart of an exemplary method for estimating internal resistance of a battery of a vehicle in real time, in accordance with embodiments of the present disclosure.

[0030] FIG. 8 represents a scatter plot illustrating the relationship between predictor variables10 and response variable, in accordance with embodiments of the present disclosure.

[0031] FIG. 9 represents a scatter plot illustrating predicted internal resistance of the battery, in accordance with embodiments of the present disclosure.

[0032] FIG. 10 represents a scatter plot illustrating estimated internal resistance of the battery, in accordance with embodiments of the present disclosure.

[0033] It should be appreciated by those skilled in the art that any block diagrams herein represent conceptual views of illustrative systems embodying the principles of the present subject matter. Similarly, it will be appreciated that any flow charts, flow diagrams, state transition diagrams, pseudo code, and the like represent various processes which may be substantially represented in a computer readable medium and20 executed by a computer or processor, whether or not such computer or processor is explicitly shown.DETAILED DESCRIPTION

[0034] The foregoing has broadly outlined the features and technical advantages of the present disclosure in order that the detailed description of the disclosure that follows may be better understood. It should be appreciated by those skilled in the art that the conception and specific embodiment disclosed may be readily utilized as a basis for modifying or designing other structures for carrying out the same purposes of the present disclosure.

[0035] The novel features which are believed to be characteristic of the disclosure, both as to its organization and method of operation, together with further objects and advantages30 will be better understood from the following description when considered in connection with the accompanying figures. It is to be expressly understood, however, that each of the figures is provided for the purpose of illustration and description only and is not intended as a definition of the limits of the present disclosure.InternalInternal Ref.: 2024PF01617

[0036] High internal resistance in a battery pack can significantly impact its efficiency. As electric current flows through the battery during charging and discharging, energy is lost primarily as heat, a direct consequence of the internal resistance. In practical terms, this means a greater portion of the battery’s stored energy is wasted, reducing its5 effective capacity. For EVs, this translates to reduced mileage per charge and for energy storage systems, a lower overall energy output.

[0037] The resistance of a battery pack depends on the internal resistance of each cell and also on the configuration of the battery cells (series or parallel). The overall performance of a battery pack depends on balancing the internal resistances of all its cells. High internal10 resistance in a pack can make it less efficient, reduce its range, and create too much heat in EVs, which can be a threat to human life and shorten battery’s life. Therefore, an early prediction of the internal resistance of battery / battery packs is needed for safe operation of the battery.

[0038] To overcome the above-mentioned challenges, the present disclosure provides technique(s) that enables real time future estimation / prediction of internal resistance of the battery of a vehicle. The internal resistance of the battery is estimated accurately and effectively utilizing the capacity of the battery. One of the advantages of the future estimation / prediction of the internal resistance is early prognosis of battery failure. Another advantage is the safety of the battery and occupants from unforeseen battery20 degradation failures, such as thermal runaway. A detailed explanation of the proposed technique(s) is disclosed in the forthcoming paragraphs.

[0039] FIG. 1 depicts an exemplary environment 100 for estimating internal resistance of a battery of a vehicle 101 in real time, in accordance with embodiments of the present disclosure. The exemplary environment 100 particularly depicts the vehicle 101 that may incorporate a system 102. The vehicle 101 may be an electric vehicle. The battery may be a Lithium-ion battery, Nickel-Metal Hydride battery, Lead-Acid battery, ultracapacitor battery, Lithium-Sulphur battery or any other battery that can perform operations of the vehicle 101.

[0040] In some implementations, the system 102 may comprise, a processor 103, a memory30 104, and two Artificial Intelligence (Al) models, namely, Battery Capacity Estimator Model 105 and Internal Resistance Estimator Model 106. The system 102 may include other components (not shown in this fig.) to implement desired functions of the system 102. The models 105 and 106 may be any suitable supervised machine learning modelInternalInternal Ref.: 2024PF01617 such as but not limited thereto a regression model that is trained to perform desired functions of the present invention.

[0041] In an exemplary implementation, the processor 103 may receive battery related data during charging and discharging of the battery. The battery related data is extracted5 from latest diagnostic load collective data of the vehicle which is stored in the memory 104. The processor 103 may process the battery related data and may generate at least one charge-discharge cycle to estimate battery degradation. For the generated at least one charge-discharge cycle, the processor 103 then estimates capacity of the battery by using the Battery Capacity Estimator Model 105. Further, based on estimated capacity10 of the battery, the processor 103 may estimate the internal resistance of the battery, by using the Internal Resistance Estimator Model 106. A detailed explanation of the system 102 is provided in the forthcoming paragraphs in conjunction with FIG.s 2-6, 8-10.

[0042] FIG. 2 depicts an exemplary block diagram illustrating details of a system 200 (which is system 102 of FIG.1) for estimating internal resistance of a battery of a vehicle in real time, in accordance with embodiments of the present disclosure. In some implementations, the system 200 may further comprise one or more sensors 201, a processor 202, a memory 203, a Battery Capacity Estimator Model 204, Internal Resistance Estimator Model 205 and a Warning Generation Unit 206.20

[0043] In one implementation, the processor 202 may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, state machines, logic circuitries, and / or any devices that manipulate signals based on operational instructions. Among other capabilities, the processor 202 may be configured to fetch and execute computer-readable instructions and other information stored in the memory 203.

[0044] In some implementations, the memory 203 may include one or more non-transitory computer-readable storage media, such as RAM, ROM, EEPROM, EPROM, one or more memory devices, flash memory devices, etc., and combinations thereof. In an embodiment, data / information such as Load Collective data 203a may be stored within30 the memory 203 in the form of various data structures. The memory 203 may also store other data 203b such as temporary data and temporary files, generated by the processor 202 or other any other parts of the system 200 including the Battery Capacity Estimator Model 204, the Internal Resistance Estimator Model 205, the Warning generation Unit 206 for performing the various functions of the present invention. The Load CollectiveInternalInternal Ref.: 2024PF01617 data 203a stores vehicle’s performance and refers to the stress patterns experienced by vehicle components during their lifetime. In an example, without limitation, the Load Collective data 203a may comprise stress patterns relating to battery parameters during charge and discharge, speed distribution, drive and acceleration torque, load, rotational5 speed of a drive machine, pitches, gear changes, and braking operations with associated vehicle velocity and braking pressure, etc.

[0045] In some implementations, the Battery Capacity Estimator Model 204 may comprise an input unit 204a and an estimation unit 204b. In some implementations, the Internal Resistance Estimator Model 205 may comprise an input unit 205a and an estimation10 unit 205b. In some implementations, the processor 202 may be operatively coupled to the one or more sensors 201, the memory 203, the Battery Capacity Estimator Model 204 and the Internal Resistance Estimator Model 205 that cooperate to estimate the internal resistance of the battery of the vehicle in real time.

[0046] In the illustrated figure, both the models, i.e., the Battery Capacity Estimator Model204 and the Internal Resistance Estimator Model 205 are shown to reside outside the processor 202 and may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, state machines, logic circuitries, and / or any devices that manipulate signals based on operational instructions. However, one of ordinary skill will appreciate that in other20 implementations, the Battery Capacity Estimator Model 204 and the Internal Resistance Estimator Model 205 may also form a part of the processor 202 and may be implemented through software or hardware or a suitable combination of software and hardware as per the implementation requirements of the present disclosure. In said implementation, the processor 202 may perform all the functions carried out by the different units / components of the Battery Capacity Estimator Model 204 and the Internal Resistance Estimator Model 205, respectively.

[0047] In one non-limiting example, the Battery Capacity Estimator Model 204 and the Internal Resistance Estimator Model 205 are trained Al models that use supervised learning algorithms for prediction / estimation of capacity of the battery. In one example,30 but not limited thereto, the Battery Capacity Estimator Model 204 is a linear regression model. In one example, but not limited thereto, the Internal Resistance Estimator Model205 is a non-linear regression model. However, a person of ordinary skill will appreciate that any other suitable Al model that serves the purpose of the presentInternalInternal Ref.: 2024PF01617 disclosure may be utilized for prediction / estimation of capacity of the battery and the internal resistance.

[0048] In the illustrated figure, the Warning Generation Unit 206 is shown to reside outside the processor 202 and may be implemented as one or more microprocessors,5 microcomputers, microcontrollers, digital signal processors, central processing units, state machines, logic circuitries, and / or any devices that manipulate signals based on operational instructions. However, one of ordinary skill will appreciate that in other implementations, the Warning Generation Unit 206 may also form a part of the processor 202 and may be implemented through software or hardware or a suitable10 combination of software and hardware as per the implementation requirements of the present disclosure.

[0049] In an exemplary scenario, while the vehicle 101 is operation, one or more sensors 201 installed on the vehicle 101 may sense the one or more parameters related to vehicle’s performance. In an exemplary embodiment, the processor 202 may receive the one or more parameters related to vehicle’s performance from the one or more sensors 201 that are coupled to or otherwise included within the vehicle 101 and may store the one or more parameters in the memory 203. The non-limiting examples of the one or more sensors 201 may include temperature sensors, pressure sensors, motion sensors, voltage sensors, current sensors, internal battery sensors, external battery sensors, energy20 management sensors and / or other sensors, etc.

[0050] The one or more parameters related to vehicle’s performance are stored in the memory 203 as the Load Collective data 203a. The Load Collective data 203a is stored in the form of diagnostic sessions which periodically get logged into the system 200. In an example without limitation, the Load Collective data 203a is stored in the memory 203 in the form of “bins” or “classes” that describes total amount of time or count the vehicle has spent in a particular category of a parameter related to the vehicle’s performance. The processor 202 may extract the Load Collective data 203a to obtain battery related data during charging and discharging of the battery. In an example without limitation, the processor 202 may extract the battery related data from latest diagnostic load30 collective data of the vehicle.

[0051] In the next step, the processor 202 may process the battery related data to generate at least one charge-discharge cycle for an unseen period of time. In a non-limiting example, the unseen period of time is indicative of a time in future for which at least one charge-discharge is generated to predict the capacity of the battery and hence theInternalInternal Ref.: 2024PF01617 internal resistance of the battery. In an exemplary embodiment, to generate the at least one charge -discharge cycle, the processor 202 may generate a charge-discharge profile for at least one parameter of the battery related data. The charge-discharge profile mimics performance behavior of the at least one parameter of the battery related data5 in real time. In other words, for generating the charge-discharge cycles for the unseen period of time, it is assumed that vehicle usage will be similar to its logged historical performance. For example, if a vehicle user uses heavy charge and discharge cycles, then the charge-discharge cycles will also contain similar charging- discharging profiles leading to faster degradation of the battery. The charge -discharge cycles may also be10 referred to as Synthetic Charging Cycles that may be used to evaluate the degradation caused by particular working conditions of the battery and to analyze how the degradation may cause loss of performance of the vehicle.

[0052] Tables A-D shown below represents one such non-limiting examples for generation of the at least charge discharge cycle for the unseen period of time, in accordance with embodiments of the present disclosure. Table A represents actual data indicative of a frequency of occurrence of charging events in certain start and end State of Charge (SOC) ranges.Table A:20

[0053] In the illustrated example as above, the SOC is considered as an example of the least one parameter of the battery related data and represents a total of 221 events. However, it is appreciated that the actual data may be produced for any other parameter of the battery related data as per the implementation requirement of present disclosure.InternalInternal Ref.: 2024PF01617

[0054] Table B represents percentage (%) value in each category or class of the start and end SOC. The processor 202 may calculate the percentage value in each category by dividing the individual frequency of occurrence with the total events. For example, for start SOC >90 and end SOC >90 the frequency of occurrence is 10 (bold and underlined5 in Table A) and therefore the percentage value in table of Table B would be 5 [(10 / 221)* 100 ~ 5% (approx.)], which is indicated as bold and underlined in Table B).Table B:10

[0055] In one example without limitation, let’s assume, at certain instance, it is required to generate ‘n’= 200 charge- discharge cycles. Thus, the ‘n’ charge- discharge cycles will be generated by multiplying percentage values with ‘n’. For example, if percentage value between start SOC 60-80 (mid value 70) and end SOC 80-90 (mid value 85) (as shown in Table A) is 2% (bold and underlined in Table B), number of the charge-15 discharge cycles with these SOC ranges will be 4 (2% of 200) which is represented inTable C in bold.Table C:InternalInternal Ref.: 2024PF01617

[0056] In another example, without limitation, if percentage value between start SOC 40-60 (mid value 50) and end SOC 60-80 (mid value 70) is 1%, number of the chargedischarge cycles with these SOC ranges will be 2 (1% of 200, see Table B) which is represented in Table D in bold.5 Table D:

[0057] The following is a non- limiting example of a pseudo code for generation of chargedischarge cycles: function: synthetic cycles generation10 input: Load collective signals output: C = {ci, C2, C3, C4 cn} - list of synthetic cyclesSet number of synthetic cycles: n for each parameter P do compute percentage distribution of each class multiply ‘n’ with percentage values for each cycle ‘n’ do repeat category values according to percentage create cycle list CRandomly allocate values to cycles20 end for each end for each

[0058] It may be noted that the examples provided in Tables A-D are purely illustrative in nature and the same should not be considered as limiting examples to implement the present disclosure.

[0059] After, the at least one charge-discharge cycle is generated, the processor 202 may estimate capacity of the battery by using the Battery Capacity Estimator Model 204, based on the generated at least one charge-discharge cycle. As previously noted, the Battery Capacity Estimator Model 204 may be a pre-trained linear regressor model.30 During the development / building phase of the Battery Capacity Estimator Model 204,InternalInternal Ref.: 2024PF01617 the model is trained using one or more sensor data relating to vehicle’s performance received from the one or more sensors 201 installed on the vehicle.

[0060] FIG. 3 represents one such exemplary block diagram illustrating building of a regressor model, in accordance with embodiments of the present disclosure. To build the5 regressor model such as Battery Capacity Estimator Model 204, the processor may collect a data set relating to vehicle performance, namely measurement data 301. The measurement data comprises one or more battery related parameters as predictor variables (X) and the capacity of the battery as a response variable (Y). In an example, without limitation, the one or more battery related parameters comprises at least one of:10 voltage, current and temperature, etc, during charge or discharge of the battery. The processor 202 may further split the data set to obtain a training dataset 301 and test dataset 303. The processor 202 may then train a regression model 304 (such the Battery Capacity Estimator Model 204) using the training dataset to learn a relationship between the predictor and the response variables. The linear regression model 304 assumes a linear functional form:Y = B0 + Bl*Xl + B2*X2 + B3*X3 + ... BN*XN.

[0061] Where B0 801 is the intercept (value of Y when all X values are zero) and Bl 802 is the slope (representing the relationship between X and Y) of a best fit line as represented in FIG. 8. The goal of linear regression is to find the line that best fits the20 data. This allows to make predictions about the capacity of the battery (Y) based on values of the one or more battery related parameters (X). In simple terms, linear regression uses a straight line to describe the relationship between a predictor variable (X) and a response variable (Y).

[0062] Once the regression model 304 is trained, the processor may use the test dataset 303 to evaluate performance of the regression model 304. Thereafter, the trained regressor model 304 (i.e. the pre-trained Battery Capacity Estimator Model 204) can be used to make predictions for the capacity of the battery (Y) for new values of the one or more battery related parameters (X) as illustrated in FIG 4.

[0063] As illustrated in FIG. 4, the processor may estimate the capacity of the battery by using30 the pre-trained Battery Capacity Estimator Model 402 (which is the pre-trained Battery Capacity Estimator Model 204 of FIG. 2). In particular, the processor 202 may feed the battery related data 401a and the generated charge-discharge cycles 401b to the Input Unit 204a of the pre-trained Battery Capacity Estimator Model 402. Thereafter, the Estimation Unit 204b of the pre-trained Battery Capacity Estimator Model 402 mayInternalInternal Ref.: 2024PF01617 estimate the capacity of the battery for the at least one charge-discharge cycle based on the battery related data 401a. FIG. 9 represents a scatter plot illustrating the prediction or estimation of the capacity of the battery, in accordance with embodiments of the present disclosure. From the scatter plot, it may be observed that predicted values 9015 (on Y axis) are very close to the actual values 902 (on X axis) for the observations in the data set and the straight line represents that the model has perfectly predicted the capacity of the battery.

[0064] Once the capacity of the battery is estimated using the pre-trained Battery Capacity Estimator Model 402, the estimated capacity of the battery is fed to the pre-trained10 Internal Resistance Estimator Model 205. The processor 202 may then use the pretrained Internal Resistance Estimator Model 205 to estimate the internal resistance of the battery, by using a battery resistance estimator model.

[0065] As previously noted, the Internal Resistance Estimator Model 205 may be a pre-trained non-linear regressor model. During the development / building phase of the Internal Resistance Estimator Model 205, the model is trained using one or more sensor data relating to vehicle’s performance received from the one or more sensors 201 installed on the vehicle.

[0066] FIG. 3 represents one such exemplary block diagram illustrating building of a regressor model, in accordance with embodiments of the present disclosure. To build the20 regressor model such as the Internal Resistance Estimator Model 205, the processor may collect a data set relating to vehicle performance, namely a measurement data 301. In this case, the measurement data 301 comprises the capacity of the battery as predictor variable (X) and the internal resistance of the battery as a response variable (Y). The processor 202 may further split the data set to obtain a training dataset 301 and test dataset 303. The processor 202 may then train a regression model 304 (such as the Internal Resistance Estimator Model 205) using the training dataset to learn a relationship between the predictor and the response variables. The regression model 304 assumes a non- linear functional form:Resistance r = -1.30 + 2.62 * battery capacity -1.66 * battery capacity2+30 0.34482444 * battery capacity3

[0067] The goal of the non- linear regression is to find the curve that best fits the data. This allows to make predictions about the internal resistance of the battery (Y) based on values of the estimated capacity of the battery (X). In simple terms, linear regressionInternalInternal Ref.: 2024PF01617 uses a straight line to describe the relationship between a predictor variable (X) and a response variable (Y).

[0068] Once the regression model 304 is trained, the processor may use the test dataset 303 to evaluate performance of the regression model 304. Thereafter, the trained regression5 model 304 (i.e. the pre-trained Internal Resistance Estimator Model 205) can be used to make predictions for the internal resistance for new values of the capacity of the battery (X) as illustrated in FIG 5.

[0069] As illustrated in FIG. 5, the processor may estimate the internal resistance of the battery by using the pre-trained Internal Resistance Estimator Model 502 (which is the pre¬10 trained Internal Resistance Estimator Model 205 of FIG. 2). In particular, the processor 202 may feed the estimated capacity of the battery to the Input Unit 205a of the pretrained Internal Resistance Estimator Model 502. Thereafter, the Estimation Unit 205b of the pre-trained Internal Resistance Estimator Model 502 may estimate the internal resistance of the battery based on the estimated capacity of the battery 501. FIG. 10 represents a scatter plot illustrating the prediction or estimation of internal resistance 1001 (on Y axis) of the battery based on the predicted battery capacity 1002 (on X axis), in accordance with embodiments of the present disclosure.

[0070] Importance of battery capacity estimation for predicting internal resistance:It is well known that the internal resistance of the battery increases with battery aging.20 When battery starts aging, there are certain electrochemical reactions inside the battery which leads to unwanted solid electrolyte interphase (SEI) layer growth or dendrite growth. As a result, the internal resistance increases. This will reduce the battery capacity. Here below are few exemplary scenarios without limitation that indicates the relation of voltage with battery capacity• Example Scenario 1 : New Battery Let’s say voltage is 3.2V, Capacity is 10 Ah, then current rating or nominal operating current is 1C. This means the battery will run for 1 hour duration by giving voltage of 3.2V and current of 10 ampere.30 • Example Scenario 2: Old battery Since the battery is aged, internal resistance increases as battery capacity decreases.InternalInternal Ref.: 2024PF01617 Voltage remains 3.2V, but the capacity of aged battery will be 8Ah. For example: if the system still requires 10 amperes current, then the current rating will be 1.25C. This means the battery will run for 48mins duration by giving voltage of 3.2V5 and current of 10 ampere.Therefore, battery capacity is required, and the internal resistance cannot be merely forecasted using voltage and current tables.

[0071] FIG. 6 represents an exemplary flow diagram illustrating generation of a warning or alert, in accordance with embodiments of the present disclosure.10

[0072] As previously described, the processor 202 utilizes pre-trained Battery Capacity Estimator Model 602 to estimate battery capacity for the synthetic charging cycles 604 based on battery related data 601 as an input. Based on the estimated battery capacity, the processor 202 utilizes pre-trained Internal Resistance Estimator Model 603 to estimate the internal resistance 605 of the battery. Once the internal resistance 605 is estimated, the processor 202 may determine at 606 if the internal resistance 605 of the battery is greater than a predefined threshold. In an example without limitation, the predefined threshold for the internal resistance may be set by a battery manufacturer based on the type and model of the battery. If the estimated internal resistance 605 is greater than a predefined threshold, the processor may generate and send a warning or20 alert 607 to driver of the vehicle. Based on said warning or alert 607, the driver may take necessary actions such as visiting to a workshop or replacing the battery. If the estimated internal resistance 605 is less than a predefined threshold, the processor may keep estimating the battery capacity based on the synthetic charging cycles 604 and may estimate the internal resistance 605 based on the battery capacity.

[0073] FIG. 7 represents flowchart of an exemplary method for estimating internal resistance of a battery of a vehicle in real time, in accordance with embodiments of the present disclosure. The order in which the method 700 is described is not intended to be construed as a limitation, and any number of the described method blocks may be combined in any order to implement the method. Additionally, individual blocks may30 be deleted from the methods without departing from the spirit and scope of the subject matter described. Furthermore, the method can be implemented in any suitable hardware, software, firmware, or combination thereof. However, for ease of explanation, in the embodiments described below, the method 700 may be consideredInternalInternal Ref.: 2024PF01617 to be implemented by the respective components and / or by the processor 202 and / or the Al models 204, 205 of FIG. 2.

[0074] At step 701, the method may include receiving battery related data during charging and discharging of the battery. The battery related data is extracted from latest diagnostic5 load collective data of the vehicle. In one implementation, the processor 202 may receive the battery related data the during charging and discharging of the battery. In another implementation, the processor 202 may extract the battery related data from latest diagnostic load collective data of the vehicle.

[0075] At step 702, the method may include processing the battery related data to generate at10 least one charge-discharge cycle for an unseen period of time. In one implementation, the processor 202 may process the battery related data to generate at least one chargedischarge cycle for an unseen period of time.

[0076] At step 703, the method may include estimating capacity of the battery by using a battery capacity estimator model, based on the generated at least one charge-discharge cycle. In one implementation, the processor 202 may estimate the capacity of the battery by using a battery capacity estimator model based on the generated at least one chargedischarge cycle.

[0077] At step 704, the method may include estimating the internal resistance of the battery, by using a battery resistance estimator model, based on estimated capacity of the20 battery. In one implementation, the processor 202 may estimate the internal resistance of the battery, by using a battery resistance estimator model, based on estimated capacity of the battery.

[0078] Additionally, the method may include determining if the internal resistance of the battery is greater than a predefined threshold. Further, the method may include generating a warning to driver of the vehicle if the internal resistance is greater than a predefined threshold.

[0079] The order in which the method 700 is described is not intended to be construed as a limitation, and any number of the described method blocks may be combined in any order to implement the method. Additionally, individual blocks may be deleted from30 the methods without departing from the spirit and scope of the subject matter described.

[0080] The illustrated steps are set out to explain the exemplary embodiments shown, and it should be anticipated that ongoing technological development will change the manner in which particular functions are performed. These examples are presented herein for purposes of illustration, and not limitation. Further, the boundaries of the functionalInternalInternal Ref.: 2024PF01617 building blocks have been arbitrarily defined herein for the convenience of the description. Alternative boundaries can be defined so long as the specified functions and relationships thereof are appropriately performed.

[0081] Alternatives (including equivalents, extensions, variations, deviations, etc., of those5 described herein) will be apparent to persons skilled in the relevant art(s) based on the teachings contained herein. Such alternatives fall within the scope and spirit of the disclosed embodiments.

[0082] Advantages of the embodiment of the present disclosure are illustrated herein-As previously indicated, aspects of the present disclosure enable prediction / estimation 10 internal resistance of a battery of a vehicle in real time for an unseen period of time.The internal resistance of the battery is estimated or predicted based on the battery capacity quite accurately and effectively and lays a foundation for early warning of battery faults. One of the advantages of the real time future estimation / prediction of the internal resistance is early prognosis of battery failure. Another advantage is the safety15 of the battery and occupants from unforeseen battery degradation failures, such as thermal runaway.

Claims

InternalInternal Ref.: 2024PF01617WE CLAIM:

1. A method for estimating internal resistance of a battery of a vehicle in real time, comprising: receiving battery related data during charging and discharging of the battery,5 wherein the battery related data is extracted from latest diagnostic load collective data of the vehicle; processing the battery related data to generate at least one charge-discharge cycle for an unseen period of time; based on the generated at least one charge-discharge cycle, estimating capacity10 of the battery by using a battery capacity estimator model; and based on estimated capacity of the battery, estimating the internal resistance of the battery, by using a battery resistance estimator model.

2. The method of claim 1, wherein processing the battery related data to generate the at least one charge -discharge cycles, comprising: generating a charge-discharge profile for at least one parameter of the battery related data, and wherein the charge-discharge profile mimics performance behaviour of the at least one parameter in real time.

3. The method of claim 1, wherein the battery capacity estimator model is built using the steps comprising:20 collecting a data set relating to vehicle performance, the data set comprises one or more battery related parameters as predictor variables and the capacity of the battery as a response variable, wherein the one or more battery related parameters comprises at least one of: voltage, current and temperature during charge or discharge of the battery; splitting the data set to obtain a training dataset and test dataset; training a first regression model using the training dataset for learning a relationship between the predictor and the response variables; and evaluating performance of the first regression model using the test dataset after the training.

4. The method of claim 1, wherein the internal resistance estimator model is built using30 the steps comprising: collecting a data set relating to vehicle performance, the data set includes the capacity of the battery as predictor variable and the internal resistance of the battery as a response variable;InternalInternal Ref.: 2024PF01617 splitting the data set to obtain a training dataset and test dataset; training a second regression model using the training dataset for learning a relationship between the predictor and the response variables; and evaluating performance of the second regression model using the test dataset5 after the training.

5. The method of claims in any one of preceding claims, wherein the method comprises estimating the internal resistance of the battery for the at least one chargedischarge cycle using the trained second regression model; and estimating the capacity of the battery for the at least one charge-discharge cycle10 using the trained first regression model.

6. A system for estimating internal resistance of a battery of a vehicle in real time, comprises: a memory configured to store diagnostic load collective data of the vehicle; and a processor electronically coupled to the memory, wherein the processor is configured to: receive battery related data during charging and discharging of the battery, wherein the battery related data is extracted from latest diagnostic load collective data of the vehicle; process the battery related data to generate at least one charge-discharge cycle,20 wherein the at least one charge-discharge cycle is generated for an unseen period of time; based on the generated the at least one charge-discharge cycle, estimate capacity of the battery by using a battery capacity estimator model; and based on the estimated capacity of the battery, estimate the internal resistance of the battery, by using a battery resistance estimator model.

7. The system of claim 6, wherein to process the battery related data to generate the at least one charge -discharge cycles, the processor is configured to: generate a charge-discharge profile for at least one parameter of the battery related data, and wherein the profile mimics performance behaviour of the at least one30 parameter in real time.

8. The system of claim 6, wherein to build the battery capacity estimator model, the processor is configured to: collect data set relating to battery performance, the data set comprises one or more battery related parameters as predictor variables and the capacity of the battery asInternalInternal Ref.: 2024PF01617 a response variable, wherein the one or more battery related parameters comprises at least one of: voltage, current and temperature during charge or discharge of the battery; split the data set to obtain a training dataset and test dataset; train a first regression model using the training dataset for learning a5 relationship between the predictor and the response variables; and evaluate performance of the second regression model using the test dataset after the training.

9. The system of claim 6, wherein to build the internal resistance estimator model, the processor is configured to: collect a data set relating to vehicle performance, the data set includes the capacity of the battery as predictor variable and the internal resistance of the battery as a response variable; split the data set to obtain a training dataset and a test dataset; train a second regression model using the training dataset for learning a15 relationship between the predictor and the response variables; and evaluate performance of the second regression model using the test dataset after the training.

10. The system of claims 6 and 9, wherein the processor is configured to: estimate the capacity of the battery for the at least one charge-discharge cycle using the trained first regression model; and estimate the internal resistance of the battery for the at least one chargedischarge cycle using the trained second regression model.25

Images