KR102438358B1 - Status Diagnosis and Disability Prediction System for Moving Objects Battery and thereof Method - Google Patents

Abstract

The present invention provides a mobile battery state diagnosis and failure prediction system and a method thereof. The system comprises: a battery measurement module which measures battery cells, modules, and packs individually; a battery management module for a battery pack which manages a current state of a battery embedded in a mobile body; a cloud server which generates battery state diagnosis data and failure prediction data by analysis using bigdata through a database build through AI learning of measurement data of current state data of the battery by the battery management module and a learning result database build through an AI learning process by the measurement data measured by the battery measurement module; a battery state inquiry module which performs preventive maintenance on the battery of which failure is predicted by inquiring the battery state based on the battery state diagnosis data and the failure prediction data generated by the analysis in the cloud server; and a mobile user terminal which receives the battery state diagnosis data and the failure prediction data analyzed from the cloud server. Therefore, the state of the mobile battery is analyzed in more detail, thereby accurately predicting the time of failure of the mobile battery. In addition, the battery state inquiry module executes the preventive maintenance for the failure in advance according to the battery state of the mobile body input from the cloud server, and transmits the battery state to the cloud server via the battery measurement module to continuously and accurately predict the failure of the battery.

Description

Status Diagnosis and Disability Prediction System for Moving Objects Battery and its Method}

The present invention provides a system and method for diagnosing and predicting the state of a mobile battery that analyzes and diagnoses the battery state of a mobile body (all moving objects using batteries, such as electric vehicles, electric motorcycles, robots, drones, etc.) in real time, predicts a failure, and notifies the user is an invention about

More specifically, it transmits the state data about the battery status (current, voltage, temperature, impedance) of the BMM (Battery Management Module) of the moving object to a remote cloud server, and the cloud server performs AI-based big data analysis in real time. It relates to a system for diagnosing and diagnosing the state of a battery in real time, predicting a battery failure and notifying a user (text message, KakaoTalk, vehicle navigation) to a user (text message, Kakaotalk, vehicle navigation) for diagnosing the state of a mobile battery and predicting a failure, and a method therefor.

According to the system and method for diagnosing the state of mobile batteries and predicting failures, the AI learning-based big data analysis of the cloud server is based on the battery current (A), voltage (V), temperature (T), and impedance (I) measurement data. (t) Time function data according to the function is analyzed by big data based on AI learning to diagnose the state of the mobile battery. In particular, the impedance (I) is the resistance component (R) of the impedance component through AI-based big data analysis. , inductive component (L), and capacity component (C) can be classified in detail to analyze the battery condition in more detail to accurately predict the time of occurrence of a failure in the battery of a moving object. In addition, according to the system and method for diagnosing the state of a mobile battery and predicting a failure according to the present invention, the battery state inquiry module performs preventive maintenance on failures in advance according to the battery state of the mobile object input from the cloud server and checks the battery state. By transmitting it to the cloud server via the battery measurement module, it is possible to continuously accurately predict the failure of the battery.

As the demand and utilization of mobile devices equipped with batteries, including electric vehicles, are rapidly increasing, damage caused by operation and shutdown of mobile devices is increasing due to battery failures that cannot be predicted at an unspecified time. There is an irreversible social loss of human and material due to fire caused by battery fluid leakage and battery cell short circuit.

As a method for measuring the state of a battery according to the existing prior art, a current measuring device (Reg. Patent 10-1680297), an impedance measuring method (Patent Publication No. 10-2019-0072790), and a life prediction device (Registration of Patent 10-1937591) ), life measurement circuit (Registration Patent 10-2103089), Remaining Life Real-time Estimation Device (Registration Patent 10-1615139), Usage Pattern Analysis (Registration Patent 10-1726483), Big Data-based Battery Degradation Estimation Device (Registration Patent 10) -1946163), a battery information management system and management method for electric vehicles (Patent Publication No. 10-2020-0080353), and a battery management system for personal mobility (Registration Patent 10-2346826). In general, in the structure of the battery of a mobile body, a plurality of battery cells constitute one battery module suitable for the battery capacity, and several battery modules constitute one battery pack. ) is built-in to check the state of the battery and control the SOC (State of Charge) so that the battery can operate in an optimal state.

In the battery management module of the battery pack, predetermined values such as current (I), voltage (V), temperature (T), SOC (State of Charge), SOH (State of Heath), and DoD (Depth of Discharge) are stored. Self-monitoring control to control voltage (charge voltage, over-discharge voltage), current control (over-current, charge current, discharge current), output limit (output control by voltage and current control), temperature control (cooling system control), cell balancing , SOC control and cell protection (prevention of overcharging and charging from regenerative braking) protect the battery, but there is a limit to predicting battery failure.

Therefore, based on the data measured from the battery management module of the mobile battery pack, it is transmitted to the cloud server in real time and analyzed big data based on AI to diagnose the state of the mobile battery and predict a failure. It is being demanded.

: Patent Publication No. 10-2020-0080353 (2020. 07. 07) : Registered Patent Publication No. 10-2346826 (2020.01.23)

According to the present invention, in order to solve the problems of the prior art, the data (current (A), voltage (V), temperature (T), impedance (I)) measured from the battery management module of the mobile battery pack is transferred to the cloud server. A mobile battery that can receive and convert the measured data into time function data by reflecting the time (t) function, and analyze big data based on AI-based power consumption and battery state change data according to time change to diagnose the battery condition The purpose of this is to provide a system and method for diagnosing the state of

In addition, according to the present invention, the components of the impedance (I) (resistance component (R), inductive component (L), capacity component (C)) required for more accurate battery condition diagnosis analysis and failure prediction diagnosis are Because it is impossible to directly measure in the management module, the changed state data of the impedance (I) according to the change in temperature (T) and time (t) according to the load change of power consumption can be analyzed in advance through AI learning to analyze the results of big data. The purpose of this is to provide a system and method for diagnosing the state of a mobile battery and predicting a failure by estimating the component data of the impedance (I) based on the convergence to a more accurate prediction value, enabling accurate analysis and prediction of failure.

In addition, according to the present invention, by measuring the battery state from a sufficiently large number of various types of battery cells, battery modules, and battery packs in advance, it is converted into a time function value, and AI learning-based big data is analyzed based on the time function change data. The purpose of this is to create a state diagnosis and failure prediction DB of the battery, and to receive new state data generated from the battery management module of the mobile battery pack from the cloud server to provide state diagnosis and failure prediction of the mobile battery.

In addition, according to the present invention, according to the result of analyzing and diagnosing the battery state in the cloud server of the present invention, for the battery whose failure is predicted, the failure prediction condition is notified to the user terminal of the mobile body through text message, Kakao Talk, and vehicle navigation. An object of the present invention is to provide a system and method for diagnosing the state of a mobile battery and predicting a failure, which can prevent various problems caused by battery failure of the mobile body in advance.

According to the present invention, the battery condition inquiry module retrieves more detailed analysis data about the mobile battery from the cloud server and performs preventive maintenance on the failure of the battery in advance, and then measures the measured data measured by the battery measurement module for the repaired battery. Mobile battery that can operate safely by improving the reliability of battery condition diagnosis and failure prediction by continuously updating the AI learning result DB (component separation learning result DB, state analysis learning result DB) through AI learning by sending it to the cloud server The purpose of this is to provide a system and method for diagnosing the state of

A system for diagnosing the state of a mobile battery and predicting a failure according to the present invention. A battery measurement module for individually measuring battery cells/modules/packs, a battery management module for a battery pack that manages the current state of a battery built into a moving body, and an AI learning process based on measurement data measured from the battery measurement module A cloud server that generates the battery condition diagnosis data and failure prediction data by analyzing the data of the current state data of the battery of the battery management module by building the learning result DB by the AI learning of the measurement data and using the big data that the DB is built by; , a battery condition inquiry module that performs preventive maintenance on a battery whose failure is predicted by inquiring the state of the battery according to the condition diagnosis data and failure prediction data of the battery generated by analysis in the cloud server and the battery analyzed from the cloud server It characterized in that it comprises a user terminal of the mobile body for receiving the state diagnosis data and the failure prediction data.

The battery measurement module of the present invention includes a battery measurement module for extracting measurement data by individually measuring battery cells/modules/packs to generate initial AI learning result data, and a battery cell/module actually measured from the battery measurement module. / It is characterized in that it includes a measurement data DB that stores the measurement data of the pack.

The battery management module of the present invention is characterized in that it extracts current (A), voltage (V), temperature (T), and impedance (I) data, which are current battery state data, and transmits it to a cloud server.

In the cloud server of the present invention, current (A), voltage (V), temperature (T), impedance (I), resistance component of a battery cell/module/pack extracted by actually measuring from the battery measuring module of the battery measuring module (R), inductive component (L), capacity component (C), etc., through the AI learning process based on the measurement data to generate the component separation learning result data for the battery, and through the AI learning process using the measurement data to the battery The state analysis learning result data is generated, and battery state data such as current (A), voltage (V), temperature (T), and impedance (I) of the current battery extracted from the battery management module are separated from the component learning result. By referring to the data, the impedance I(t) of the battery is divided into component data for each component such as resistive component R(t), inductive component L(t), and capacitive component C(t) data, and each separated component data After converting the time function change data into time function change data, the converted time function change data is generated as the state diagnosis and failure prediction result data of the battery with reference to the state analysis learning result data, and the state diagnosis and failure of the battery are predicted and the mobile terminal It is characterized by notifying

The cloud server of the present invention includes a time function conversion module that converts the measurement data actually measured by the battery measurement module into time function data, and converts the state data of the battery received from the battery management module into time function data; An AI component separation learning module that generates component separation learning result data through a continuous AI learning process with respect to the time function data of the measurement data generated and input from the time function conversion module, and impedance among the time function data for the state data of the battery An AI component separation module that separates the component data of the impedance by referring to the component separation learning result data of the AI component separation learning module generated through AI learning, and the time for the measurement data of the battery measurement module from the time function conversion module Receive function data, extract time function change data that varies during power consumption Pm, temperature T _n , and time ts, receive time function data for state data of the battery management module from the time function conversion module, and separate the AI component A time function change data extraction module that receives component data separated from impedance among the time function data of the state data of the battery management module from the module and extracts time function change data that varies during power consumption Pm, temperature T _n , and time ts; 10 steps through continuous AI learning process to generate initial learning result data for estimating battery condition diagnosis and failure prediction based on time function change data for measurement data extracted from the time function change data extraction module AI state analysis learning module that generates the subdivided state analysis learning result data, and the state data representing the current battery state of the battery management module is changed during power consumption Pm and temperature T _n and time ts in the time function change data extraction module AI battery state that determines the state diagnosis and failure prediction state of the battery by referring to the state analysis learning result data generated and stored through continuous AI learning in the initial AI state analysis learning module for the average value of the converted 7 time function change data It is characterized in that it includes a diagnosis and failure prediction module.

The cloud server of the present invention, the battery current (A), voltage (V), temperature (T), impedance (I), resistive component (R), inductive component (L) , A(t), V(t), T(t), I(t), R(t), L(t), C(t) It is converted into the same time function data, and the current (A), voltage (V), temperature (T), and impedance (I) of the battery that is extracted and input from the battery management module is the current state data of the battery as a time function. a time function conversion module for converting time function data such as A(t), V(t), T(t), I(t) according to ; Time of measurement data such as A(t), V(t), T(t), I(t), R(t), L(t), C(t) generated and input from the time function conversion module Function data I(t)_PmTn, R(t)_PmTn, L(t)_PmTn, C(t) through continuous AI learning process based on power consumption (Pm), temperature (Tn), time (ts) data an AI component separation learning module for generating component separation learning result data for a battery that is _PmTn data; AI from the AI component separation learning module 320 with respect to the impedance I(t) among time function data such as A(t), V(t), T(t), I(t) for the state data of the battery Impedance (I(t)) component with reference to component separation learning data such as I(t)_PmTn, R(t)_PmTn, L(t)_PmTn, C(t)_PmTn generated and extracted through the learning process an AI component separation module that separates and extracts component data such as R(t), L(t), and C(t) in detail for each component; Time function data such as A(t), V(t), T(t), I(t), R(t), L(t), C(t) for the measured data of the battery measurement module It receives input from the function conversion module, converts it into time function change data that changes during power consumption Pm, temperature T _n , and time ts, and extracts it, and A(t), V(t), T( I(t), R(t), which receive time function data such as t) and I(t) from the time function conversion module and are separated from the impedance I(t) among the time function data of the state data of the battery management module; ΔA(ts)_PmTn, ΔV(ts)_PmT, ΔT(ts) that changes during power consumption Pm and temperature T _n and time ts by receiving component data such as L(t) and C(t) from the AI component separation module a time function change data extraction module for extracting time function change data such as _PmTn, ΔI(ts)_PmTn, ΔR(ts)_PmTn, ΔL(ts)_PmTn, and ΔC(ts)_PmTn; Time functions such as A(t), V(t), T(t), I(t), R(t), L(t), C(t) of the measurement data extracted from the time function change data extraction module ΔA(ts)_PmTn, ΔV(ts)_PmT, ΔT(ts)_PmTn, ΔI(ts)_PmTn, ΔR(ts)_PmTn, ΔL(ts) for data power consumption Pm and temperature T _n and time ts Based on the time function change data such as _PmTn and ΔC(ts)_PmTn, to estimate the battery state diagnosis and failure prediction that can know the internal state of the battery, the initial AI state analysis learning subdivided into 10 stages through the continuous AI learning process AI state analysis learning module for generating result data; Time such as A(t), V(t), T(t), I(t), R(t), L(t), C(t) of the state data indicating the current battery state of the battery management module ΔA(ts)_PmTn, ΔV(ts) _{_PmT} , ΔT(ts)_PmTn, ΔI(ts)_PmTn, ΔR(ts)_PmTn, ΔL(ts) for function data )_PmTn, ΔC(ts)_PmTn, about the average value of the seven time function change data extracted from the time function change data extraction module, which is generated and stored through continuous AI learning in the initial AI state analysis learning module. an AI battery condition diagnosis and failure prediction module for determining the condition diagnosis and failure prediction state of the battery with reference to the data; The AI battery condition diagnosis and failure prediction module receives the diagnosis result data, which is the diagnosis result data of the battery, and the fault prediction data, which is the current state grade (ranks 0 to 3) of the battery, and transfers the diagnosis result data to a mobile device. a diagnosis result notification module that notifies the user terminal 500 of the through text message, KakaoTalk, or vehicle navigation; and input the diagnosis result data according to the time function change data, which is the current state grade (0 to 3) of the battery, which is the state diagnosis data and the failure prediction data, which are the diagnosis result data from the AI battery state diagnosis and failure prediction module. a battery condition inquiry module for receiving and maintaining a battery of a moving object in need of maintenance, and transmitting result data on the battery for which the maintenance has been completed to the battery measurement module of the battery measurement module; It is characterized in that it includes.

According to the present invention, the equation converted from the time function data to the time function change data during the power consumption Pm, the temperature T _n and the time ts, the current (A) change amount is ΔA(ts)_PmTn = A(t _n )-A( t _n-1 ), the amount of change in voltage (V) is ΔV(ts)_PmTn = V(t _n )-V(t _n-1 ), and the amount of change in temperature (T) is ΔT(ts)_PmTn = T(t _n )- T(t _n-1 ), impedance (I) change amount is ΔI(ts)_PmTn = I(t _n )-I(t _n-1 ), resistance component (R) change amount is ΔR(ts)_PmTn = R(t _n )-R(t _n-1 ), inductive component (L) change amount is ΔL(ts)_PmTn = L(t _n )-L(t _n-1 ), capacitance component (C) change amount is ΔC(ts)_PmTn = C(t _n )-C(t _n-1 ).

In the method for diagnosing the state of a mobile battery and predicting a failure according to the present invention, the current ( A), voltage (V), temperature (T), impedance (I), resistive component (R), inductive component (L), capacitive component (C) data of measurement data (A, V, T, I, R, Extracting L, C), storing it in the measurement data DB, and transmitting it to the time function conversion module (S100), and the measurement data (A, V, T, I, R) measured from the battery measurement module in the time function conversion module , L, C) as a function of time (t) as A(t), V(t), T(t), I(t), R(t), L(t), C(t) In order to generate AI learning result data through AI learning from the step (S200) of converting into time function data and storing it in the time function data DB, and the time function data for the measurement data of the battery generated in the time function conversion module, the battery A step (S300) consisting of an AI component separation learning step (S400) of separating each component by analyzing the components of in detail, an AI condition analysis learning step (S500) for analyzing the state of the battery (S300), and an AI component separation learning step (S400 receives the impedance I(t) data that can represent the components of the battery among the time function data for the measurement data of the battery to the AI component separation learning module in order to analyze the components of the battery in detail and separate them for each component It is a step to separate the impedance I(t) data for each component by performing the AI component separation learning process to build a component separation learning result data DB, and the AI state analysis learning step (S500) is, in order to analyze the state of the battery, A(t), V(t), T(t), I(t), R(t), L(t), C(t) The data is transmitted to the AI state analysis learning module and the AI state analysis learning process is carried out to build a state analysis learning result data DB, and the current (A) which is the current battery state data of the battery pack from the battery management module of the battery pack. ), voltage (V), temperature (T), impedance (I) data is extracted and transmitted to the time function conversion module of the cloud server (S600), and the time function conversion module extracts the data from the battery management module inside the battery It receives the current (A), voltage (V), temperature (T), and impedance (I) data, which are state data, and A(t), V(t), which are time function data converted according to the time function (t), A step (S700) of generating T(t), I(t) data, storing it in the time function data DB, and transmitting it to the AI component separation (R, L, C) module (S700) and AI component (R, L, C) separation Time from the time function conversion module for the state data of the battery management module in the module In order to receive the impedance I(t) data among the time function data converted according to the function (t), and to separate the input impedance I(t) data into more detailed component data, the AI component separation learning in the step (S400) By referring to the component separation learning result data such as I(t)_PmTn, R(t)_PmTn, L(t)_PmTn, C(t)_PmTn extracted from the AI component separation learning module through the AI component separation learning process of step After separating R(t), L(t), C(t) component data for the impedance I(t) component, it is stored in the component separation DB, and the separated component data for the impedance I(t) component is stored A step of providing an input to the time function change data extraction module (S800), and the time function data A(t), V( t), T(t), and I(t) are directly input from the time function conversion module for battery status diagnosis and failure prediction, I(t), R(t), L(t), C(t), which are component data separated into components of , ΔA(ts)_PmTn, ΔV(ts)_PmTn, ΔT(ts)_PmTn, ΔI(ts)_PmTn, ΔR(ts)_PmTn, ΔL(ts)_PmTn with changes in temperature (T) and time (t) , ΔC(ts)_PmTn extracting time function change data, storing it in a time function change data DB, and inputting the time function change data to an AI battery condition diagnosis and failure prediction module (S900) and AI battery In the state diagnosis and failure prediction module, ΔA(ts)_PmTn, ΔV(ts)_PmTn, ΔT(ts)_PmTn, ΔI(ts)_PmTn, which are time function change data for the internal battery state data extracted from the battery management module , ΔR(ts)_PmTn, ΔL(ts)_PmTn, ΔC(ts)_PmTn The state diagnosis and failure prediction state of the battery by the average value of the seven time function change data with reference to the AI state-learned state analysis learning result data generated in the AI state analysis learning step of the step (S500) A step (S1000) of generating the diagnosed state diagnosis data and the failure prediction data, storing the battery state diagnosis and failure prediction result DB, and transmitting the diagnosis result notification module and the battery state inquiry module (S1000) and the AI battery state in the battery state inquiry module By receiving the diagnosis result data according to the time function change data, which is the current status grade (0 to 3) of the battery status diagnosis data, which is the diagnosis result data from the diagnosis and fault prediction module, and the fault prediction data, if the moving object that needs maintenance For the battery of to the cloud server (S1100) and when the state diagnosis result, which is the state diagnosis data and the failure prediction data of the battery in the diagnosis result notification module, is predicted to fail, in real time to a user terminal that operates a mobile body equipped with a battery. It is characterized in that it comprises a step (S1200) of transmitting through a text message, KakaoTalk, or vehicle navigation.

In the AI component separation learning step (S400) according to the present invention, the time function data A(t), V(t), T(t), I( As the result data through continuous AI learning based on power consumption (Pm), temperature (Tn), and time (ts) data by receiving t), R(t), L(t), C(t) data, I (t)_PmTn, R(t)_PmTn, L(t)_PmTn, C(t)_PmTn, etc. are generated, stored in the component separation learning result DB, and input to the AI component separation module. characterized.

In the AI state analysis learning step (S500) according to the present invention, the time function data A(t), V(t), T(t), I which are the time function data generated from the time function conversion module in the time function change data extraction module (t), R(t), L(t), C(t) data are input and the time function change data that is changed according to the change of power consumption (Pm), temperature (Tn), time (ts) data is extracted. Storing the time function change data DB, and transmitting the extracted time function change data to the AI state analysis learning module (S510), and A(t) extracted from the time function change data extraction module in the AI state analysis learning module , V(t), T(t), I(t), R(t), L(t), C(t) time function change during time ts in the state of power consumption Pm and temperature T _n Based on the data inside the battery: ΔA(ts)_PmTn, ΔV(ts)_PmT, ΔT(ts)_PmTn, ΔI(ts)_PmTn, ΔR(ts)_PmTn, ΔL(ts)_PmTn, ΔC(ts)_PmTn In order to generate basic learning result data for estimating battery state diagnosis and failure prediction that can know the state, state analysis learning results to build a learning DB by generating learning result data subdivided into 10 steps through a continuous AI learning process It is stored in the DB and transmitted to the AI battery state diagnosis and failure prediction module (S520).

The system and method for diagnosing the state of a mobile battery and predicting a failure according to the present invention, the data (current (A), voltage (V), temperature (T), impedance (I)) measured from the battery management module of the mobile battery pack The cloud server receives the data, reflects the time (t) function, converts the measured data into time function data, and analyzes the battery state change data according to power consumption and time change based on AI big data to diagnose the battery state in detail. There is an effect that can be done.

In addition, the system and method for diagnosing the state of a mobile battery and predicting a failure according to the present invention provide components of impedance (I) (resistance component (R), inductive component (L), capacity component) necessary for more accurate battery condition analysis and diagnosis. Since (C)) cannot be directly measured by the battery management module of the mobile battery pack, the state data of the change in impedance (I) according to the change in temperature (T) and time (t) according to the load change of power consumption is recorded in advance. By estimating the component data of impedance (I) based on the analysis result of big data through AI learning, it converges to an accurate predicted value, which has the effect of diagnosing the state of the mobile battery and predicting the failure by more accurate analysis and prediction.

In addition, the system and method for diagnosing the state of a mobile battery and predicting a failure according to the present invention measure the state of a battery from a sufficiently large number of various types of battery cells, battery modules, and battery packs in advance, convert it into time function data, and perform a time function Based on the change data, it analyzes AI learning-based big data to create a battery state analysis learning result DB, and receives new state data generated from the battery management module of the mobile battery pack from the cloud server to diagnose the state of the AI battery and diagnose the failure. The prediction module has the effect of diagnosing the state of the mobile battery and predicting the failure.

In addition, the system and method for predicting failure of a mobile battery according to the present invention provide a battery for which the failure is predicted according to the result of analyzing and diagnosing the battery state in the cloud server of the present invention. By notifying the battery status diagnosis and failure prediction status to the user terminal, it is possible to prevent various problems caused by the battery failure of the mobile body in advance, as well as the battery status inquiry module provides a more detailed analysis of the mobile battery from the cloud server After performing preventive maintenance for battery failure by inquiring data in advance, the measured data measured by the battery measurement module is transmitted to the cloud server for the repaired battery, and the AI learning result DB (component separation learning result) continues through AI learning DB, status analysis learning result DB) is updated to increase the reliability of battery condition diagnosis and failure prediction, and preventive maintenance can be performed before the failure of the mobile battery occurs, thereby enabling safe mobile operation.

1 is a block diagram of a system for diagnosing the state of a mobile battery and predicting a failure according to the present invention.
2 is a flowchart for generating initial learning data in the system for diagnosing the state of a mobile battery and predicting a failure according to the present invention.
3 is a flowchart for diagnosing the state of the battery and predicting the failure in the system for diagnosing the state of the mobile battery and predicting the failure according to the present invention.
4 is a flowchart illustrating a process of transmitting and receiving initial learning data for generating initial learning data in the system for diagnosing the state of a mobile battery and predicting a failure according to the present invention.
5 is a flowchart illustrating a data transmission/reception process for diagnosing the state of a battery and predicting a failure in the system for diagnosing and predicting the state of a mobile battery according to the present invention.

Since the description of the present invention is merely an embodiment for structural or functional description, the scope of the present invention should not be construed as being limited by the embodiment described in the text. That is, since the embodiment is capable of various changes and may have various forms, it should be understood that the scope of the present invention includes equivalents capable of realizing the technical idea.

Elements having the same function in all the drawings below are omitted from repetitive descriptions using the same reference numerals, and the following terms are defined in consideration of the functions in the present invention, which are consistent with the technical spirit of the present invention. It specifies that the concept and its own and common meaning should be interpreted.

On the other hand, the meaning of the terms described in the present invention should be understood as follows.

Terms such as “first” and “second” are for distinguishing one component from another, and the scope of rights should not be limited by these terms. For example, a first component may be termed a second component, and similarly, a second component may also be termed a first component.

When a component is referred to as being “connected” to another component, it may be directly connected to the other component, but it should be understood that other components may exist in between. On the other hand, when it is mentioned that a certain element is "directly connected" to another element, it should be understood that the other element does not exist in the middle. On the other hand, other expressions describing the relationship between elements, that is, "between" and "between" or "neighboring to" and "directly adjacent to", etc., should be interpreted similarly.

The singular expression is to be understood as including the plural expression unless the context clearly dictates otherwise, and terms such as "comprises" or "have" refer to the specified feature, number, step, action, component, part or these It is intended to indicate that a combination exists, and it should be understood that it does not preclude the possibility of the existence or addition of one or more other features or numbers, steps, operations, components, parts, or combinations thereof.

Identifiers (eg, a, b, c, etc.) in each step are used for convenience of description, and the identification code does not describe the order of each step, and each step clearly indicates a specific order in context. Unless otherwise specified, it may occur in a different order from the specified order. That is, each step may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

All terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs, unless otherwise defined. Terms defined in a commonly used dictionary should be interpreted as having the meaning consistent with the context of the related art, and cannot be interpreted as having an ideal or excessively formal meaning unless explicitly defined in the present invention.

In addition, in describing the present invention, if it is determined that a detailed description of a related known function or configuration may obscure the gist of the present invention, the detailed description thereof will be omitted.

Hereinafter, a preferred embodiment according to the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram of a system for diagnosing the state of a mobile battery and predicting a failure according to the present invention.

As shown, the state diagnosis and failure prediction system 10 of the mobile battery according to the present invention includes a battery actual measurement module 100, a battery management module 200, a cloud server 300, and a battery. It includes a state inquiry module 400, and a terminal 500 of a mobile user.

The system for diagnosing the state of a mobile battery and predicting a failure according to the present invention (10) is a system in which the reliability of battery state analysis diagnosis and failure prediction through AI learning is improved as data continuously increases, and the battery actual measurement module (100) Based on the measurement data of and the state data of the battery management module 200, the battery state inquiry module 400 and the mobile user terminal 500 for receiving the battery diagnosis result for the battery by the cloud server 300 are integrated. It is a platform system connected to

According to the present invention, in order for the state diagnosis and failure prediction system 10 of the mobile battery to operate normally, the mobile battery is actually measured by the battery measurement module 110 of the battery measurement module 100 to generate initial learning data in advance. It is absolutely necessary to generate measurement data and conduct AI learning, and at this time, the AI learning is performed in 10 steps in terms of the battery's good, bad, and bad progress.

Specifically, in order for the state diagnosis and failure prediction system 10 of the mobile battery according to the present invention to operate normally, in order to generate initial learning data in advance in advance, the battery measurement module 110 of the battery measurement module 100 Components such as current (A), voltage (V), temperature (T), impedance (I), resistive component (R), inductive component (L), and capacity component (C) measured from this battery cell/module/pack The measured data for the data is converted into time function data through the time function conversion module 310 and stored in the time function data DB 315, and the time function data generated by the time function conversion module 310 is separated into AI components Component separation learning in which the impedance I(t) for the battery is separated into each component such as a resistive component R(t), an inductive component L(t), and a capacitive component C(t) by the AI learning process of the learning module 320 It is generated as the result data and stored in the component separation learning result DB 325 to build the result DB according to the AI learning process.

In addition, the time function data is extracted as time function change data through the time function change data extraction module 340 and stored in the time function change data DB 345, and the extracted time function change data is an AI state analysis learning module ( 350) is generated as state analysis learning result data for the battery subdivided into 10 stages by the AI learning process, and stored in the state analysis learning result DB 355 to build the result DB according to the AI learning process.

A system for diagnosing a state of a mobile battery and predicting a failure according to the present invention includes: a battery measuring module 100 for individually measuring battery cells/modules/packs; a battery management module 200 of a battery pack that manages the current state of a battery built into the mobile body; Based on the measured data measured from the battery measurement module 100, the DB is constructed through AI learning of the current state data of the battery of the battery management module 200 and the construction of the learning result DB by the AI learning process. a cloud server 300 for generating battery condition diagnosis data and failure prediction data by analysis through big data; a battery condition inquiry module 400 for performing preventive maintenance on the battery predicted to fail by inquiring the state of the battery according to the condition diagnosis data and failure prediction data of the battery generated by analysis in the cloud server 300; and a mobile user terminal 500 for receiving the battery state diagnosis data and failure prediction data analyzed from the cloud server 300 .

Hereinafter, the state diagnosis and failure prediction system 10 of the mobile battery according to the present invention will be described in more detail with reference to FIG. 1 .

The battery measurement module 100 includes a battery measurement module 110 that extracts measurement data by individually measuring battery cells/modules/packs in order to generate initial AI learning result data, and from the battery measurement module 110 . and a measurement data DB 120 that stores measurement data of battery cells/modules/packs extracted by actual measurement.

The battery measurement module 110 of the battery measurement module 100 uses an EIS (Electrochemical Impedance Spectronization) measurement method, which is a method for separating and measuring the data on the impedance (I) of the battery by component, or the resistance component of the battery Measurement data is generated by measuring data for (R), inductive component (L), and capacity component (C) using individual measuring instruments.

When measuring the battery cell, battery module, and battery pack of the battery, the battery's current (A), voltage (V), After generating temperature (T), impedance (I), resistive component (R), inductive component (L), and capacitive component (C) data, the generated measurement data is stored as a backup in the measurement data DB 120 , Measurement data of current (A), voltage (V), temperature (T), impedance (I), resistive component (R), inductive component (L), capacitive component (C) data is a time function of the cloud server 300 It is transmitted to the conversion module 310 .

The battery management module 200 is a management module for managing a battery built into a moving body, and is current (A), voltage (V), and temperature (T), which are current battery state data extracted from the battery management module 200 . , and transmits the impedance (I) data to the time function conversion module 310 of the cloud server 300 .

Hereinafter, the cloud server 300 will be described.

Cloud server 300 according to the present invention, current (A), voltage (V), temperature (T), impedance of the battery cell / module / pack measured from the battery measurement module 110 of the battery measurement module 100 In order to generate component separation learning result data for the battery through the AI learning process based on measurement data such as (I), resistive component (R), inductive component (L), capacity component (C), time function conversion module ( 310), the time function data DB 315, the AI component separation learning module 320 and the component separation learning result DB 325 are included.

In addition, in order to generate the state analysis learning result data for the battery through the AI learning process using the measured data, the time function conversion module 310, the time function data DB 315, and the time function change data extraction module 340 , a time function change data DB 345 , an AI state analysis learning module 350 , and a state analysis learning result DB 355 .

In addition, the cloud server 300 according to the present invention, battery state data such as current (A), voltage (V), temperature (T), impedance (I) of the current battery extracted from the battery management module 200 The impedance (I(t)) of the battery in the AI component separation module 330 with reference to the component separation learning result data generated through the AI learning process of the AI component separation learning module 320 using the resistance component R(t) , separated into component data for each component, such as the derived component L(t), and the capacity component C(t) data, and each component data separated by the AI component separation module 330 is separated by the time function change data extraction module 340 ) in the time function change data, the converted time function change data refers to the state analysis learning result data generated through the AI learning process in the AI state analysis learning module 350 to diagnose the condition of the battery and predict the failure. In order to generate data, the time function conversion module 310, the time function data DB 315, the AI component separation module 330, the time function change data extraction module 340, and the time function change data DB 345 , and an AI battery state diagnosis and failure prediction module 360 , and a battery state diagnosis and failure prediction result DB 365 .

Hereinafter, the cloud server 300 according to the present invention will be described in more detail.

The time function conversion module 310, the current (A), voltage (V), temperature (T), impedance (I), resistive component (R), inductive component (L) measured by the battery measurement module 110 ), A(t), V(t), T(t), I(t), R(t), L(t), C( t) After being converted into time function data for constructing component separation learning result data through AI learning, which is data, it is stored in the time function data DB 315 .

In addition, the time function conversion module 310, the time from the current state data for the current (A), voltage (V), temperature (T), impedance (I) of the battery received from the battery management module 200 The A(t), V(t), T(t), I(t) data converted according to the function is converted into time function data for battery condition diagnosis and failure prediction, and then stored in the time function data DB (315). Save.

The time function data DB 315 is a time function for constructing the component separation learning result data through AI learning, in which the measured data measured from the battery measurement module 100 is converted into a time function in the time function conversion module 310 Data of A(t), V(t), T(t), I(t), R(t), L(t), C(t) are stored, and also from the battery management module 200 A(t), V(t), T(t), I (t) The data is stored.

In order to build a battery component separation learning result data DB through an AI learning process according to power consumption (Pm), temperature (Tn), and time (t) from the measured data of the battery, the measured data of time (t) The time function data generated from the time function conversion module 310 reflecting the function is provided as an input to the AI component separation learning module 320 so that AI can be learned.

In addition, in order to build a state analysis learning result data DB of the battery through the AI learning process from the measured data of the battery, the time function generated from the time function conversion module 310 reflecting the function of time (t) in the measured data The data is provided as an input to the AI state analysis learning module 350 so that AI can be learned via the time function change data extraction module 340 in which the data is extracted as the amount of change according to power consumption, temperature, and time fluctuations.

AI component separation learning module 320, time function data A(t), V(t), T(t), I(t), R of the measurement data generated and input from the time function conversion module 310 (t), L(t), C(t) are result data through continuous AI learning based on power consumption (Pm), temperature (Tn), and time (t) data, I(t)_PmTn, R( Component separation learning result data for the battery, which is t)_PmTn, L(t)_PmTn, C(t)_PmTn data, is generated and stored in the component separation learning result DB 325 and is input to the AI component separation module 330 .

The component separation learning result DB 325 is a component separation for the battery generated through continuous AI learning based on power consumption (Pm), temperature (Tn), and time (t) data from the AI component separation learning module 320 . The learning result data is input and a DB for the AI-learned component separation learning result data is built.

AI component separation module 330, the impedance I ( I(t)_PmTn, R(t)_PmTn, L(t) generated and extracted through AI learning from the AI component separation learning module 320 in order to separate component data divided for each component in more detail with respect to t) )_PmTn, C(t)_PmTn, with reference to the component separation learning result data, after separating the R(t), L(t), C(t) component data for the impedance (I(t)) component It is stored in the component separation DB 335 and provided as an input to the time function change data extraction module 340 .

That is, the time function data for the impedance I(t) among the battery state data is the resistance component R extracted from the AI component separation learning module 320 based on the power consumption (Pm), temperature (Tn), and time (t) data. (t)_PmTn, inductive component L(t)_PmTn, capacitive component C(t)_PmTn, and impedance I(t)_PmTn data, such as component separation learning data, are separated into specific component data and stored in component separation DB (335). is saved

In the component separation DB 335, I(t), R(t), L(t), C(t) component data from which the components are separated from the AI component separation module 320 are input and stored.

Time function change data extraction module 340, time function data A(t), V(t), T for the measurement data of the battery measurement module 100 converted into a time function in the time function conversion module 310 (t), I(t), R(t), L(t), C(t) data received from the time function conversion module 310 to construct state analysis learning result data through AI learning and power consumption (P), temperature (T), time (t) by extracting the time function change data changed according to the change is stored in the time function change data DB (345).

In addition, the time function change data extraction module 340, A(t), V(t), which are time function data for the state data of the battery management module 200 converted into a time function in the time function conversion module 310 ), T(t), I(t), time function data A(t), V(t), T(t), I( t) Data is directly input, and I(t), R( t), L(t), C(t) data are input from the AI component separation module 330 for battery status diagnosis and failure prediction, and power consumption (P), temperature (T), and time (t) changes The time function change data changed according to the time function change data is extracted and stored in the time function change data DB 345 .

That is, the time function change data extraction module 340 converts and extracts the time function change data changed from the time function data during the time ts in the power consumption Pm and the temperature T _n state.

The time function change data of the time function change data extraction module 340 converted with respect to the time function data is converted and extracted according to the following equation, respectively.

Current (A) change: ΔA(ts)_PmTn = A(t _n )-A(t _n-1 )

Voltage (V) change: ΔV(ts)_PmTn = V(t _n )-V(t _n-1 )

Temperature (T) change: ΔT(ts)_PmTn = T(t _n )-T(t _n-1 )

Impedance (I) change amount: ΔI(ts)_PmTn = I(t _n )-I(t _n-1 )

Resistance component (R) change: ΔR(ts)_PmTn = R(t _n )-R(t _n-1 )

Inductive component (L) change: ΔL(ts)_PmTn = L(t _n )-L(t _n-1 )

Change in capacitance component (C): ΔC(ts)_PmTn = C(t _n )-C(t _n-1 )

According to the present invention, in order to diagnose and analyze the current state inside the mobile battery, data of changed state of the battery according to changes in power consumption (P), temperature (T), and time (t) and AI component separation learning that is learned and stored Impedance among the state data extracted from the battery management module 200 of the mobile body to analyze the result data, i.e., the resistance component (R), the inductive component (L), the capacitance component (C), and the impedance (I) data and their correlations with each other. (I) is separated for each component from the AI component separation module 320 as a resistive component (R), an inductive component (L), and a capacity component (C), and the input component data is analyzed as AI component separation learning result data, and time Create function change data.

The time function change data DB 345 stores time function change data extracted from the time function change data extraction module 340, and is converted from the measured data measured by the battery measurement module 100 among the stored time function change data. The time function change data is provided as an input for AI learning in the AI state analysis learning module 350 .

In addition, the time function change data converted from the internal state data of the battery management module among the time function change data stored in the time function change data DB 345 is AI battery state diagnosis in order to generate result data for state diagnosis and failure prediction. and the failure prediction module 360 .

AI state analysis learning module 350, the power consumption Pm and temperature T _n extracted from the time function change data extraction module 340, A(t), V(t), T(t) changed during time ts in the state , ΔA(ts)_PmTn, ΔV(ts)_PmT, ΔT(ts)_PmTn, ΔI(ts), which are time function change data for , I(t), R(t), L(t), C(t) data Based on _PmTn, ΔR(ts)_PmTn, ΔL(ts)_PmTn, ΔC(ts)_PmTn data, continuous Through the AI learning process, the learning result data subdivided into 10 steps is generated and stored in the state analysis learning result DB 355 to build a DB for the state analysis learning result data.

The state analysis learning result DB 355 converts the change data changed during the time (ts) in the state of power consumption (Pm) and temperature (T _n ) from the AI state analysis learning module 350 into 10 steps through the AI learning process. The state analysis learning result data generated by segmentation is input, and a DB for the state analysis learning data learned by AI is built.

The AI battery state diagnosis and failure prediction module 360 is ΔA, which is seven time function change data converted from the internal state data of the battery with reference to the state analysis learning result data learned by AI in the initial AI state analysis learning module 350 (ts)_PmTn, ΔV(ts)_PmTn, ΔT(ts)_PmTn, ΔI(ts)_PmTn, ΔR(ts)_PmTn, ΔL(ts)_PmTn, ΔC(ts)_PmTn based on the average value of battery status diagnosis and the failure prediction status are diagnosed as follows.

1) If the average value of each of the 7 time function change data is 70% or less of the initial value (new), 5 or more specific values are 30% or less, and the battery status is 2 or less, the battery failure prediction grade 0 Stop using the moving object and recommend immediate maintenance.

2) If the average value of each of the seven time function change data is 70% or less of the initial value (new), 3 to 4 specific values are 30% or less, and the battery status is 3 stages or less, the battery failure expected grade 1st Maintenance is recommended within 1 week.

3) If the average value of each of the time function change data is 70% or less of the initial value (new), the two specific values are 30% or less, and the battery status is 4 stages or less, maintenance within 2 weeks as the 2nd priority of the battery failure expected grade recommend

4) If the average value of each of the time function change data is 70% or less of the initial value (new), one specific value is 30% or less, and the battery status is 6 stages or less, maintenance is performed within 4 weeks as the 3rd priority for battery failure recommend

The above diagnosis result is a diagnosis determined based on the state analysis learning result data learned by AI from the initial AI state analysis learning module 350, and as the system use continues, more detailed diagnosis section and diagnosis through AI learning Since the history is generated, the accuracy of diagnosis is improved.

In addition, by transmitting the measured data measured through the battery measuring module 100 to the cloud server 300 for the actual battery preventively maintained by the battery state inquiry module 400, the accuracy of the battery state diagnosis and failure prediction data is improved. can be raised

The battery state diagnosis and failure prediction result DB 365 is a diagnosed battery state grade (0th to 3rd order) that can check the state of the battery from the internal state data of the battery in the AI battery state diagnosis and failure prediction module 360 ), the state diagnosis and failure prediction result data of the time function change data is stored.

The diagnosis result notification module 370 is the result data diagnosed by the AI battery state diagnosis and failure prediction module 360, that is, the current state grade (0 to 3) of the battery, which is the state diagnosis data and the failure prediction data of the battery. The diagnosis result data is received, and the diagnosis result data is notified to the user terminal 500 of the mobile body through a text message, KakaoTalk, or vehicle navigation.

The battery state inquiry module 400 is the result data diagnosed from the AI battery state diagnosis and failure prediction module 360, that is, the current state grade of the battery, which is the state diagnosis data and the failure prediction data of the battery (ranks 0 to 3). By receiving the diagnosis result data according to the time function change data, the maintenance is performed on the battery of the mobile body that needs maintenance, and the result data for the battery that has been maintained is transmitted to the battery measurement module 100 and the battery measurement module 110 ) transmits the measured data extracted by actually measuring the maintenance-completed battery to the cloud server 300 again.

The user terminal 500 of the mobile body is a user terminal using a mobile device that operates a mobile device equipped with a battery. received through

2 is a flowchart for generating initial learning defect data in the system for diagnosing the state of a mobile battery and predicting a failure according to the present invention, and FIG. and failure prediction, and FIG. 4 is a flowchart showing the transmission/reception process of initial learning data for generating initial learning data in the state diagnosis and failure prediction system of the mobile battery according to the present invention, and FIG. It is a flowchart showing the data transmission/reception processing process for battery condition analysis diagnosis and failure prediction in the mobile battery condition diagnosis and failure prediction system.

The method for generating the initial learning data of the mobile battery according to the present invention shown in FIGS. 2 and 4 is described with reference to the block diagram of the state diagnosis and failure prediction system of the mobile battery according to the present invention in FIG. As follows.

The method for generating initial learning data according to the present invention is to measure a battery to extract measured data, to generate time function data converted from the extracted measured data according to a time function, and from the time function data, impedance data of the battery For R, L, C, component data, component separation learning result data, which is component data, is created through AI learning and built into a DB. Construct state analysis learning data.

More specifically, the method for generating the initial training data is as follows.

Step (S100): individually measured and extracted while artificially changing the load (lm) and temperature (t) in the battery cell, the battery module, and the battery pack by the battery measurement module 110 of the battery measuring module 100 Current (A), voltage (V), temperature (T), impedance (I), resistive component (R), inductive component (L), capacitive component (C) measurement data (A, V, T, I) , R, L, C) is extracted and stored in the measurement data DB 120 and transmitted to the time function conversion module 310 (S100);

Step (S200): the time function conversion module 310 converts the measured data (A, V, T, I, R, L, C) measured from the battery measurement module 110 to a time function according to a time (t) function Extracting data A(t), V(t), T(t), I(t), R(t), L(t), C(t) data and storing it in the time function data DB 315 Step (S200) and;

Step (S300): generating AI learning result data through AI learning from the time function data generated by the time function conversion module 310 (S300); is made of

The step (S300) is divided into an AI component separation learning step (S400) in which the components of the battery are analyzed in detail and separated for each component, and an AI state analysis learning step (S500) for analyzing the state of the battery.

That is, the step (S300) can be roughly divided into two steps as follows.

One step is the AI component separation learning step (S400), and in order to analyze the components of the battery in detail and separate them for each component, the impedance I(t) data that can represent the components of the battery among the time function data is learned by AI component separation It is a step of constructing a component separation learning result data DB by transmitting to the module 320 and performing AI component separation learning to separate the impedance I(t) data for each component.

The other step is the AI state analysis learning step (S500), which can represent the state of the battery among time function data in order to analyze the state of the battery A(t), V(t), T(t), I( t), R(t), L(t), C(t) data is transmitted to the AI state analysis learning module 350 to conduct AI state analysis learning, and to build a state analysis learning result data DB.

Here, as described above, the reason for transmitting the time function data to the AI component separation learning module 320 and the AI state analysis learning module 350, that is, through two learning paths, is to diagnose the state of the mobile battery according to the present invention and In order for the failure prediction system 10 to operate normally, it is to construct the initial AI learning result data in advance.

In order to generate the initial learning result data according to the present invention, R ( t), L(t), C(t) data, which are component separation learning result data, and at the same time, state diagnosis data and failure generated from the measurement data through the AI learning process of the AI state analysis learning module 350 Time function data is transmitted through two paths in order to construct the state analysis learning result data of the mobile battery, which is the data on the good, bad, and bad progress of the battery, which is the predictive data.

That is, the result data of R, L, C, and component data, which are R, L, C, and component data through AI component separation learning for the impedance data of the battery, is created and built as a DB, and at the same time, AI state analysis is performed on the state data of the battery among the time function data. Through learning, the battery condition analysis learning result data is created and built as a DB to build the initial learning result data DB.

The AI component separation learning step (S400) will be described in more detail as follows.

(S400) step: AI component separation learning module 320 is the time function data generated by the time function conversion module 310 A(t), V(t), T(t), I(t), R (t), L(t), C(t) data as input, and as the result data through continuous AI learning based on power consumption (Pm), temperature (Tn), and time (ts) data, I(t)_PmTn , R(t)_PmTn, L(t)_PmTn, C(t)_PmTn, etc. generating component separation learning result data, storing it in the component separation learning result DB 325, and inputting it into the AI component separation module 330 ( S400).

The AI state analysis learning step (S500) will be described in more detail as follows.

(S510) step: time function change data extraction module 340 is the time function data generated by the time function conversion module 310 A(t), V(t), T(t), I(t), Time function change data by taking R(t), L(t), C(t) data as input storing in the DB 345 and transmitting the extracted time function change data to the AI state analysis learning module 350 (S510);

(S520) step: AI state analysis learning module 350, A(t), V(t), T(t), I(t), R(t) extracted from the time function change data extraction module 340 ), L(t), and C(t) data, ΔA(ts)_PmTn, ΔV(ts)_PmT, ΔT(ts)_PmTn, which are time function change data that fluctuated during time ts in the state of power consumption Pm and temperature T _n Based on , ΔI(ts)_PmTn, ΔR(ts)_PmTn, ΔL(ts)_PmTn, ΔC(ts)_PmTn data, reference learning result data for estimating battery status diagnosis and failure prediction to know the internal state of the battery In order to generate a learning result data subdivided into 10 steps through a continuous AI learning process to generate Transmitting to (S520); characterized in that it includes a.

With respect to the time function data for the measured data in the step (S510), the change data changed during the power consumption (Pm), temperature (Tn), and time (ts) are each time function change by the following conversion function equation Convert to data.

Current (A) change: ΔA(ts)_PmTn = A(t _n )-A(t _n-1 )

Voltage (V) change: ΔV(ts)_PmTn = V(t _n )-V(t _n-1 )

Temperature (T) change: ΔT(ts)_PmTn = T(t _n )-T(t _n-1 )

Impedance (I) change amount: ΔI(ts)_PmTn = I(t _n )-I(t _n-1 )

Resistance component (R) change: ΔR(ts)_PmTn = R(t _n )-R(t _n-1 )

Inductive component (L) change: ΔL(ts)_PmTn = L(t _n )-L(t _n-1 )

Change in capacitance component (C): ΔC(ts)_PmTn = C(t _n )-C(t _n-1 )

According to the present invention, in order to diagnose and analyze the current state inside the mobile battery, the change in the state of the battery according to the change in power consumption (P), temperature (T), and time (t) is measured by the resistive component (R), the inductive component ( L), in order to analyze from the correlation between the capacitance component (C) and the impedance (I), the impedance (I) of the state data extracted from the battery management module 200 of the mobile body is the resistive component (R) and the inductive component (L) , and the capacity component (C) is separated for each component from the AI component separation module 320, and the input component data is analyzed to generate the time function change data as described above.

The method for diagnosing the state of a mobile battery and predicting a failure according to the present invention shown in FIGS. 3 and 5 is described with reference to the block diagram of the system for diagnosing the state of the mobile battery and predicting a failure according to the present invention in FIG. As follows.

In the method for diagnosing the state of a mobile battery and predicting a failure according to the present invention, current (A), voltage (V), temperature ( T), impedance (I) data is extracted, and the extracted state data is converted into A(t), V(t), T(t), I(t) data, which are time function data according to a time function, and converted Component separation learning generated through AI learning in the AI component separation learning module 320, which is the AI component separation learning step (S400), in order to separate the impedance I(t) into more detailed component data among the time function data. R(t), L(t) for the impedance (I(t)) component referring to the result data I(t)_PmTn, R(t)_PmTn, L(t)_PmTn, C(t)_PmTn, etc. ) and C(t) are separated into component data and stored in the component separation DB 335 , and input is provided to the time function change data extraction module 340 .

After converting each component data separated and input by the AI component separation module 330 into time function change data in the time function change data extraction module 340, an AI state analysis learning module using the converted time function change data ( 350), by referring to the state analysis learning result data generated through the AI learning process, the battery state diagnosis and failure prediction result data is generated.

In more detail, a method for diagnosing the state of a mobile battery and predicting a failure is as follows.

Step (S600): The battery management module 200 converts the time function of the cloud server 300 by extracting current (A), voltage (V), temperature (T), and impedance (I) data that are current battery state data Transmitting to the module 310 (S600) and;

Step (S700): the time function conversion module 310 receives the current (A), voltage (V), temperature (T), and impedance (I) data, which are state data of the battery management module 200, A(t), V(t), T(t), I(t) data, which are converted time function data according to function change, are generated and stored in the time function data DB 320 and AI component separation (R, L) , C) transmitting to the module 330 (S700) and;

(S800) step: AI component (R, L, C) separation module 330, the time converted from the time function conversion module 310 to the state data of the battery management module 200 according to the time function (t) In order to receive the impedance I(t) data from the function data, and to separate the input impedance I(t) data into more detailed component data, AI component separation is performed through the AI component separation learning step (S400). R(t) for the impedance I(t) component with reference to data such as I(t)_PmTn, R(t)_PmTn, L(t)_PmTn, C(t)_PmTn extracted from the learning module 320, L(t) and C(t) are separated into component data, stored in the component separation DB 335, and provided as input to the time function change data extraction module 340 (S800);

Step (S900): The time function change data extraction module 340, the time function data A(t), V( From t), T(t), I(t), the impedance I(t) is directly input from the time function conversion module 310 for battery condition diagnosis and failure prediction, and among the time function data of the state data converted into a time function. ), I(t), R(t), L(t), C(t), which are component data separated into each component, are input from the AI component separation module 330 for battery condition diagnosis and failure prediction , power consumption (P), temperature (T), time (t) change according to the change data extracted and stored in the time function change data DB (345), AI battery status diagnosis and failure prediction module (360) providing an input (S900);

That is, the time function change data extraction module 340 extracts the change data changed from the time function data during the time ts in the power consumption Pm and the temperature T _n state. Each is converted and extracted according to the formula.

Current (A) change: ΔA(ts)_PmTn = A(t _n )-A(t _n-1 )

Voltage (V) change: ΔV(ts)_PmTn = V(t _n )-V(t _n-1 )

Temperature (T) change: ΔT(ts)_PmTn = T(t _n )-T(t _n-1 )

Impedance (I) change amount: ΔI(ts)_PmTn = I(t _n )-I(t _n-1 )

Resistance component (R) change: ΔR(ts)_PmTn = R(t _n )-R(t _n-1 )

Inductive component (L) change: ΔL(ts)_PmTn = L(t _n )-L(t _n-1 )

Change in capacitance component (C): ΔC(ts)_PmTn = C(t _n )-C(t _n-1 )

Step (S1000): AI battery state diagnosis and failure prediction module 360, ΔA(ts)_PmTn, ΔV(ts), which is converted time function change data according to battery internal state data extracted from the battery management module 200 )_PmTn, ΔT(ts)_PmTn, ΔI(ts)_PmTn, ΔR(ts)_PmTn, ΔL(ts)_PmTn, ΔC(ts)_PmTn AI state analysis learning module 350 which is the AI state analysis learning step (S500) ) by referring to the AI state-learned state analysis learning result data generated from the state diagnosis data and failure prediction data for diagnosing the state diagnosis and failure prediction state for the battery by the average value of the seven time function change data, and storing the result of battery state diagnosis and failure prediction in the DB 365, and transmitting the result to the diagnosis result notification module 370 and the battery state inquiry module 400 (S1000).

Here, the time function change data ΔA(ts)_PmTn, ΔV(ts)_PmTn, ΔT(ts)_PmTn, ΔI(ts)_PmTn, ΔR(ts)_PmTn, ΔL(ts)_PmTn, ΔC(ts)_PmTn data and ΔC(ts)_PmTn data The method of determining the AI battery status diagnosis and failure prediction status for the battery by the average value of the same 7 data is as follows.

1) If the average value of each of the 7 time function change data is 70% or less of the initial value (new), 5 or more specific values are 30% or less, and the battery status is 2 or less, the battery failure prediction grade 0 Stop using the moving object and recommend immediate maintenance.

2) If the average value of each of the seven time function change data is 70% or less of the initial value (new), 3 to 4 specific values are 30% or less, and the battery status is 3 stages or less, the battery failure expected grade 1st Maintenance is recommended within 1 week.

3) If the average value of each of the time function change data is 70% or less of the initial value (new), the two specific values are 30% or less, and the battery status is 4 stages or less, maintenance within 2 weeks as the 2nd priority of the battery failure expected grade recommend

4) If the average value of each of the time function change data is 70% or less of the initial value (new), one specific value is 30% or less, and the battery status is 6 stages or less, maintenance is performed within 4 weeks as the 3rd priority for battery failure recommend

Step (S1100): The battery state inquiry module 400 performs the diagnosis result data from the AI battery state diagnosis and failure prediction module 360, that is, the state diagnosis data and the failure prediction data of the battery. 3) of the time function change data, the diagnosis result data according to the time function change data Transmitting to the module 110, and transmitting the measured data extracted from the actual measurement by the battery measurement module 110 of the battery measurement module 100 to the cloud server 300 with respect to the battery that has completed the transmitted maintenance (S1100) ; and

Step (S1200): the diagnosis result notification module 370, when the state diagnosis result, which is the state diagnosis data and the failure prediction data of the battery, is predicted to be a failure, the user terminal 500 for operating the mobile body equipped with the battery in real time It includes a step (S1200) of transmitting through a text message, KakaoTalk, or vehicle navigation.

100: battery measurement module
110: battery measurement module 120: measurement data DB
200: battery management module
300: cloud server
310: time function conversion module 315: time function data DB
320: AI component separation learning module 325: component separation learning result DB
330: AI component separation module 335: component separation DB
340: time function change data extraction module 345: time function change data DB
350: AI state analysis learning module 355: state analysis learning result DB
360: AI battery status diagnosis and failure prediction module
365: Battery status diagnosis and failure prediction result DB
370: diagnosis result notification module
400: battery status inquiry module
500: mobile user terminal

Claims (12)

delete delete delete a battery measuring module for individually measuring battery cells/modules/packs;
a battery management module of the battery pack for managing the current state of the battery built into the mobile body;
Based on the measured data measured from the battery measurement module, the learning result DB is constructed by the AI learning process, and the current state data of the battery of the battery management module is analyzed through the big data in which the DB is built through the AI learning of the measured data. a cloud server for generating battery condition diagnosis data and failure prediction data;
a battery condition inquiry module for performing preventive maintenance on the battery whose failure is predicted by inquiring the condition of the battery based on the condition diagnosis data and failure prediction data of the battery generated by analysis in the cloud server; and
and a mobile user terminal for receiving the battery state diagnosis data and failure prediction data analyzed from the cloud server,
The battery measurement module includes a battery measurement module that extracts measurement data by individually measuring battery cells/modules/packs to generate initial AI learning result data, and battery cells/modules/packs actually measured from the battery measurement module. including a measurement data DB that stores the measurement data of
The battery management module extracts current (A), voltage (V), temperature (T), and impedance (I) data, which are current battery state data, and transmits it to the cloud server,
The cloud server, the current (A), voltage (V), temperature (T), impedance (I), resistive component (R) of the battery cell / module / pack that is actually measured and extracted from the battery measurement module of the battery measurement module , induction component (L), capacity component (C), etc. through the AI learning process to generate component separation learning result data for the battery, and state analysis of the battery through the AI learning process using the measured data Generate learning result data, and refer to the component separation learning result data for battery state data such as current (A), voltage (V), temperature (T), and impedance (I) of the current battery extracted from the battery management module Thus, the impedance I(t) of the battery is divided into component data for each component such as resistive component R(t), inductive component L(t), and capacitive component C(t) data, and the separated component data is used as a time function After converting to change data, the converted time function change data is generated as the result data of condition diagnosis and failure prediction of the battery with reference to the condition analysis learning result data, and the condition diagnosis and failure of the battery are predicted and notified to the mobile terminal. A system for diagnosing the state of a mobile battery and predicting a failure, characterized in that it gives. a battery measuring module for individually measuring battery cells/modules/packs;
a battery management module of the battery pack for managing the current state of the battery built into the mobile body;
Based on the measured data measured from the battery measurement module, the learning result DB is constructed by the AI learning process, and the current state data of the battery of the battery management module is analyzed through the big data in which the DB is built through the AI learning of the measured data. a cloud server for generating battery condition diagnosis data and failure prediction data;
a battery condition inquiry module for performing preventive maintenance on the battery whose failure is predicted by inquiring the condition of the battery based on the condition diagnosis data and failure prediction data of the battery generated by analysis in the cloud server; and
and a mobile user terminal for receiving the battery state diagnosis data and failure prediction data analyzed from the cloud server,
The battery measurement module includes a battery measurement module that extracts measurement data by individually measuring battery cells/modules/packs to generate initial AI learning result data, and battery cells/modules/packs actually measured from the battery measurement module. including a measurement data DB that stores the measurement data of
The battery management module extracts current (A), voltage (V), temperature (T), and impedance (I) data, which are current battery state data, and transmits it to the cloud server,
The cloud server includes a time function conversion module that converts the measured data actually measured by the battery measurement module into time function data and converts the state data of the battery pack received from the battery management module into time function data, and the time function conversion An AI component separation learning module that generates component separation learning result data through a continuous AI learning process for time function data of measurement data generated and inputted from the module, and AI for impedance among time function data for the state data of the battery An AI component separation module that separates the component data of impedance by referring to the component separation learning result data of the AI component separation learning module generated through learning, and the time function data for the measurement data of the battery measurement module from the time function conversion module It receives input and extracts time function change data that varies during power consumption Pm, temperature T _n , and time ts, receives time function data for state data of the battery management module from the time function conversion module, and manages the battery from the AI component separation module A time function change data extraction module that receives component data separated from impedance among the time function data of the module's state data and extracts time function change data that varies during power consumption Pm, temperature T _n , and time ts, and the time function change State analysis subdivided into 10 stages through continuous AI learning process to generate initial learning result data for estimating battery condition diagnosis and failure prediction based on time function change data for measured data extracted and input from the data extraction module AI state analysis learning module for generating learning result data, and state data representing the current battery state of the battery management module is a time function change data extraction module, the power consumption Pm, temperature T _n , and converted seven AI battery state diagnosis and failure to determine the state diagnosis and failure prediction state of the battery by referring to the state analysis learning result data generated and stored through continuous AI learning in the initial AI state analysis learning module for the average value of the time function change data A system for diagnosing the state of a mobile battery and predicting a failure, comprising a prediction module. a battery measuring module for individually measuring battery cells/modules/packs;
a battery management module of the battery pack for managing the current state of the battery built into the mobile body;
Based on the measured data measured from the battery measurement module, the learning result DB is constructed by the AI learning process, and the current state data of the battery of the battery management module is analyzed through the big data in which the DB is built through the AI learning of the measured data. a cloud server for generating battery condition diagnosis data and failure prediction data;
a battery condition inquiry module for performing preventive maintenance on the battery whose failure is predicted by inquiring the condition of the battery based on the condition diagnosis data and failure prediction data of the battery generated by analysis in the cloud server; and
and a mobile user terminal for receiving the battery state diagnosis data and failure prediction data analyzed from the cloud server,
The battery measurement module includes a battery measurement module that extracts measurement data by individually measuring battery cells/modules/packs to generate initial AI learning result data, and battery cells/modules/packs actually measured from the battery measurement module. including a measurement data DB that stores the measurement data of
The battery management module extracts current (A), voltage (V), temperature (T), and impedance (I) data, which are current battery state data, and transmits it to the cloud server,
The cloud server, the current (A), voltage (V), temperature (T), impedance (I), resistive component (R), inductive component (L), capacity component of the battery actually measured and input by the battery measurement module Time functions such as A(t), V(t), T(t), I(t), R(t), L(t), C(t) according to the time function of the measured data such as (C) Converted to data and extracted from the battery management module, the current (A), voltage (V), temperature (T), and current state data such as impedance (I) of the battery are converted to A( A time function conversion module for converting time function data such as t), V(t), T(t), I(t), and A(t), V(t) generated and inputted from the time function conversion module Time function data of measurement data such as , T(t), I(t), R(t), L(t), C(t) are converted to power consumption (Pm), temperature (Tn), time (ts) data AI component separation learning module that generates component separation learning result data for batteries that are I(t)_PmTn, R(t)_PmTn, L(t)_PmTn, C(t)_PmTn data through a continuous AI learning process based on And, the AI component separation learning module 320 with respect to the impedance I(t) among time function data such as A(t), V(t), T(t), I(t) for the state data of the battery Impedance (I(t)) with reference to component separation learning data such as I(t)_PmTn, R(t)_PmTn, L(t)_PmTn, C(t)_PmTn generated and extracted through the AI learning process from An AI component separation module that separates and extracts component data such as R(t), L(t), C(t) in detail for each component for each component, and A( Time function data such as t), V(t), T(t), I(t), R(t), L(t), C(t) are input from the time function conversion module, and the power consumption Pm and temperature A(t), V(t) for the state data of the battery management module by converting and extracting time function change data that varies during T _n and time ts ), T(t), I(t) received from the time function conversion module, and I(t), R separated from the impedance I(t) among the time function data of the state data of the battery management module ΔA(ts)_PmTn, ΔV(ts)_PmT, ΔA(ts)_PmTn, ΔV(ts)_PmT, which receive component data such as (t), L(t), C(t) from the AI component separation module, and change during power consumption Pm and temperature T _n and time ts A time function change data extraction module for extracting time function change data such as ΔT(ts)_PmTn, ΔI(ts)_PmTn, ΔR(ts)_PmTn, ΔL(ts)_PmTn, ΔC(ts)_PmTn, and time function change data Consumption of time function data such as A(t), V(t), T(t), I(t), R(t), L(t), C(t) of the measurement data extracted from the extraction module ΔA(ts)_PmTn, ΔV(ts) _{_PmT} , ΔT(ts)_PmTn, ΔI(ts)_PmTn, ΔR(ts)_PmTn, ΔL(ts)_PmTn, ΔC(ts), ΔA(ts)_PmTn, ΔV(ts)_PmT, ΔT(ts)_PmTn, Based on time function change data such as ts)_PmTn, in order to estimate the battery condition diagnosis and failure prediction that can know the internal state of the battery, the initial AI condition analysis learning result data is generated, which is subdivided into 10 steps through the continuous AI learning process. AI state analysis learning module, and A(t), V(t), T(t), I(t), R(t), L(t) of state data indicating the current battery state of the battery management module For time function data such as , C(t), ΔA(ts)_PmTn, ΔV(ts) _{_PmT} , ΔT(ts)_PmTn, ΔI(ts)_PmTn, ΔR Continuous AI learning in the initial AI state analysis learning module about the average value of 7 time function change data extracted from the time function change data extraction module converted as (ts)_PmTn, ΔL(ts)_PmTn, ΔC(ts)_PmTn State diagnosis of the battery by referring to the state analysis learning result data created and stored through and an AI battery state diagnosis and failure prediction module for determining the failure prediction state, and the current state diagnosis data of the battery, which is the result data diagnosed by the AI battery state diagnosis and failure prediction module, and the current state grade (0 to 3) of the battery, which is the failure prediction data. a diagnosis result notification module that receives the diagnosis result data of the order) and notifies the diagnosis result data to the user terminal 500 of the mobile body through a text message, KakaoTalk, or vehicle navigation, and the AI battery status diagnosis and failure prediction module It receives the diagnosis result data according to the time function change data, which is the current status grade (0 to 3) of the battery, which is the fault prediction data, and the status diagnosis data of the battery, which is the result data, and performs maintenance on the battery of the moving object in need of maintenance. A system for diagnosing and predicting failure of a mobile battery, comprising a battery state inquiry module for transmitting result data on the battery that has been repaired to the battery measurement module of the battery measurement module. 7. The method of claim 6,
The equation converted from the time function data to the time function change data during the power consumption Pm, the temperature T _n and the time ts is,
Current (A) change amount: ΔA(ts)_PmTn = A(t _n )-A(t _n-1 ),
Voltage (V) change: ΔV(ts)_PmTn = V(t _n )-V(t _n-1 ),
Temperature (T) change amount: ΔT(ts)_PmTn = T(t _n )-T(t _n-1 ),
Impedance (I) change amount: ΔI(ts)_PmTn = I(t _n )-I(t _n-1 ),
Resistance component (R) change: ΔR(ts)_PmTn = R(t _n )-R(t _n-1 ),
Inductive component (L) change amount: ΔL(ts)_PmTn = L(t _n )-L(t _n-1 ),
Change in capacitance component (C): ΔC(ts)_PmTn = C(t _n )-C(t _n-1 ),
A system for diagnosing the state of a mobile battery and predicting a failure, characterized in that Current (A), voltage (V), temperature (T), impedance (A), voltage (V), temperature (T), impedance ( Measured data (A, V, T, I, R, L, C), which are I), resistance component (R), inductive component (L), and capacity component (C) data, are extracted and stored in the measurement data DB and stored in a time function Transmitting to the conversion module (S100);
The time function conversion module converts the measured data (A, V, T, I, R, L, C) measured from the battery measurement module to A(t), V(t), T( t), converting into time function data such as I(t), R(t), L(t), and C(t) and storing the converted time function data in a time function data DB (S200);
AI component separation learning step (S400) of analyzing the components of the battery in detail and separating each component in order to generate AI learning result data through AI learning from the time function data for the measurement data of the battery generated in the time function conversion module ) and a step (S300) consisting of an AI state analysis learning step (S500) for analyzing the state of the battery;
AI component separation learning step (S400, in order to analyze the components of the battery in detail and separate them for each component, the impedance I(t) data that can represent the component of the battery among the time function data for the measurement data of the battery as the AI component It is transmitted to the separation learning module and proceeds with the AI component separation learning process to separate the impedance I(t) data for each component to build a component separation learning result data DB,
AI state analysis learning step (S500), in order to analyze the state of the battery, A(t), V(t), T(t), I (t), R(t), L(t), C(t) data is transmitted to the AI state analysis learning module and proceeds with the AI state analysis learning process to build a state analysis learning result data DB,
extracting current (A), voltage (V), temperature (T), and impedance (I) data as the current state data of the battery from the battery management module and transmitting the data to the time function conversion module of the cloud server (S600);
The time function conversion module receives the current (A), voltage (V), temperature (T), and impedance (I) data, which are state data extracted from the battery management module, and converts the time function according to the time function (t). A step of generating A(t), V(t), T(t), I(t) data that is data, storing it in the time function data DB, and transmitting it to the AI component separation (R, L, C) module (S700) ;
In the AI component separation (R, L, C) module, the impedance I(t) data is transmitted among the time function data converted according to the time function (t) from the time function conversion module for the state data of the battery management module, In order to separate the input impedance I(t) data into more detailed component data, I(t)_PmTn extracted from the AI component separation learning module through the AI component separation learning process of the AI component separation learning step (S400) , R(t)_PmTn, L(t)_PmTn, C(t)_PmTn, R(t), L(t), C(t) for the impedance I(t) component with reference to the component separation learning result data After separating into each component data, storing it in the component separation DB, and providing the separated component data for the impedance I(t) component as input to the time function change data extraction module (S800);
The state of the battery from the time function data A(t), V(t), T(t), I(t) converted into the time function (t) of the battery state data of the battery management module in the time function change data extraction module I(t), R(t), which are component data that are directly input from the time function conversion module for diagnosis and failure prediction, and separated into each component from the impedance I(t) among the time function data of the state data converted into a time function ), L(t), C(t) are input from the AI component separation module for battery condition diagnosis and failure prediction, Time function change data such as (ts)_PmTn, ΔV(ts)_PmTn, ΔT(ts)_PmTn, ΔI(ts)_PmTn, ΔR(ts)_PmTn, ΔL(ts)_PmTn, ΔC(ts)_PmTn are extracted storing the function change data in a DB and providing the time function change data as input to an AI battery state diagnosis and failure prediction module (S900);
In the AI battery state diagnosis and failure prediction module, ΔA(ts)_PmTn, ΔV(ts)_PmTn, ΔT(ts)_PmTn, ΔI(ts) which are time function change data for the internal battery state data extracted from the battery management module )_PmTn, ΔR(ts)_PmTn, ΔL(ts)_PmTn, ΔC(ts)_PmTn data, referring to the AI state learning state analysis learning result data generated in the AI state analysis learning step (S500) above, 7 Generating state diagnosis data and failure prediction data for diagnosing the state diagnosis and failure prediction state of the battery according to the average value of the time function change data, and storing the state diagnosis data and the failure prediction data in the battery state diagnosis and failure prediction result DB (S1000);
The battery condition inquiry module is a diagnosis result according to the time function change data, which is the current condition grade (0 to 3) of the battery condition diagnosis data, which is the result data diagnosed from the AI battery condition diagnosis and failure prediction module, and the failure prediction data. If data is input, maintenance is performed on the battery of a moving object that needs maintenance, and the result data of the repaired battery is transmitted to the battery measurement module, and for the transferred battery, the battery measurement module of the battery measurement module transmitting the measured data of the battery actually measured and extracted to the cloud server (S1100); and
The diagnosis result notification module transmits the state diagnosis data, which is the state diagnosis data and the failure prediction data, of the battery through a text message, Kakao Talk, or vehicle navigation to a user terminal operating a mobile body equipped with a battery in real time when a failure is predicted. (S1200) further comprising,
In the AI component separation learning step (S400), the time function data A(t), V(t), T(t), I(t), R generated from the time function transformation module in the AI component separation learning module (t), L(t), C(t) data as input, and as the result data through continuous AI learning based on power consumption (Pm), temperature (Tn), and time (ts) data, I(t)_PmTn , R(t)_PmTn, L(t)_PmTn, C(t)_PmTn, etc. are generated, stored in the component separation learning result DB, and input to the AI component separation module,
In the AI state analysis learning step (S500), the time function data A(t), V(t), T(t), I(t), which are time function data generated from the time function conversion module in the time function change data extraction module, Time function change data by taking R(t), L(t), C(t) data as input A step (S510) of storing in the DB and transmitting the extracted time function change data to the AI state analysis learning module, A(t), V(t) extracted from the time function change data extraction module in the AI state analysis learning module ), T(t), I(t), R(t), L(t), C(t) data, _ΔA ( ts)_PmTn, ΔV(ts)_PmT, ΔT(ts)_PmTn, ΔI(ts)_PmTn, ΔR(ts)_PmTn, ΔL(ts)_PmTn, ΔC(ts)_PmTn In order to generate basic learning result data for estimating battery condition diagnosis and failure prediction, the learning result data subdivided into 10 steps through the continuous AI learning process is generated and stored in the state analysis learning result DB to build the learning DB. , A method for diagnosing the state of a mobile battery and predicting a failure, comprising the step of transmitting to the AI battery state diagnosis and failure prediction module (S520).
delete delete delete 9. The method of claim 8,
The equation converted from the time function data to the time function change data during the power consumption Pm, the temperature T _n and the time ts is,
Current (A) change amount: ΔA(ts)_PmTn = A(t _n )-A(t _n-1 ),
Voltage (V) change: ΔV(ts)_PmTn = V(t _n )-V(t _n-1 ),
Temperature (T) change amount: ΔT(ts)_PmTn = T(t _n )-T(t _n-1 ),
Impedance (I) change amount: ΔI(ts)_PmTn = I(t _n )-I(t _n-1 ),
Resistance component (R) change: ΔR(ts)_PmTn = R(t _n )-R(t _n-1 ),
Inductive component (L) change amount: ΔL(ts)_PmTn = L(t _n )-L(t _n-1 ),
Change in capacitance component (C): ΔC(ts)_PmTn = C(t _n )-C(t _n-1 ),
A method for diagnosing the state of a mobile battery and predicting a failure, characterized in that

Images