A system estimating battery health

Abstract

The present invention relates to a system (1) which estimates battery (2) health by means of machine learning algorithms under variable conditions that occur during vehicle use by using the resistance values that occur depending on time and usage in the cells of the battery (2) used in electric vehicles.

Description

[0001] A SYSTEM ESTIMATING BATTERY HEALTH

[0002] Technical Field

[0003] The present invention relates to a system which estimates battery health by means of machine learning algorithms under variable conditions that occur during vehicle use by using the resistance values that occur depending on time and usage in the cells of the battery used in electric vehicles.

[0004] Background of the Invention

[0005] Resistance increase is observed in electric vehicle batteries depending on usage and this resistance increase causes the battery life to shorten. Battery life varies under different usage profiles and conditions. In today’s applications, systems which enable the prediction of battery health depending on the resistances included in the battery, the profile of vehicle usage and the ambient conditions are used. Detection of battery health is provided by performing resistance measurements during discharge and recharging of the battery under fixed conditions specified in the said systems. However, these applications do not take varying user driving profiles and environmental conditions into consideration and also data is not analyzed in real time. Due to the fact that the data is not analyzed in real time together with varying environmental conditions, battery health may be measured incorrectly in real electric vehicle usage scenarios.

[0006] Therefore, there is need for a system which enables the detection of battery health on the vehicle in real usage scenarios by means of machine learning algorithms trained through data, which are previously generated with different usage scenarios and environmental conditions and resistance measurements of which are performed under different conditions (such as SOC, temperature), by using both historical driving data and instant operating data of the vehicle in order to determine the battery health accurately.

[0007] The Chinese patent document no. CN111487539A, an application included in the state of the art, discloses a cloud-based electric vehicle battery health detection system. The invention subject to the said Chinese patent document comprises a data monitoring system, a database system and a data visualization system. The data monitoring system uses a GPS positioning module, a current and voltage monitoring module, an internal resistance monitoring module and an operation attitude monitoring module to carry out data. For monitoring of an electric vehicle battery state and an electric vehicle operation condition, the database system uses a data transmission module, a data compression module and a data backup module to realize safe storage of monitoring data and the data visualization system uses a data analysis module. The system discloses that the relationship between the health condition of the electric vehicle battery and various factors can be obtained by monitoring the various parameters affecting the battery health, and the operation and driving condition of the electric vehicle. While this invention analyses the resistance variation under constant conditions of use by disclosing the relationship between the parameters affecting the battery health and the battery health, it remain incapable of predicting battery health from resistances under real and varying driving conditions.

[0008] In the Chinese patent document no. CN103293483A, an application included in the state of the art, lithium battery health condition estimation method comprises measurement of internal resistance of batteries and estimation of lithium battery health condition through a linear interpolation method according to values of the internal resistance. The lithium battery health condition estimation method based on internal resistance measurement can conveniently estimate the battery health condition in real time, calculation is fast, and estimation accuracy is high. However, current methods cannot meet the requirements of real and varying conditions of vehicle use on electric vehicles and cannot provide sufficient accuracy in terms of precision.

[0009] Summary of the Invention

[0010] An object of the present invention is to realize a system which analyses the resistances that occur depending on use in the cells of the battery used in electric vehicles by using machine learning algorithms under variable conditions of use, and estimates the battery health.

[0011] Detailed Description of the Invention

[0012] “A System Estimating Battery Health” realized to fulfd the objective of the present invention is shown in the figures attached, in which:

[0013] Figure 1 is a general view of an inventive battery.

[0014] The components illustrated in the figure are individually numbered, where the numbers refer to the following:

[0015] 1. System

[0016] 2. Battery

[0017] 3. Storage unit

[0018] 4. Control unit

[0019] An inventive system (1) for regularly estimating the health of the battery that stores the energy required to run electric vehicles comprises at least one battery (2) which stores the energy required to run an electric vehicle and comprises at least one cell that generates resistance during the storage and / or utilization of the energy; at least one storage unit (3) which performs resistance measurement in the battery cells under the condition that the vehicle is traveling at constant speed during operation of the electric vehicle and wherein the state of charge (SOC), temperature of the battery (2) at that moment; the state of health (SOH) of the battery (2) at the moment when the previous measurement condition is met; and the data of voltage, current during the past use of the cells (in a certain time interval, time elapsed between two measurement conditions, At) are kept under record historically; and at least one control unit (4) which is in communication with the battery (2) and the storage unit (3) and instantly estimates the health of cells by means of supervised machine learning algorithms (basically such as regression, decision trees, random forest, support vector machines, k-Nearest Neighbor or neural networks) created to estimate the health of a trained battery by using the information previously recorded in the storage unit (3) each time the vehicle moves in constant speed mode and using the data of battery health (SOH) obtained from battery tests previously aged under different ambient conditions and usage scenarios and cell resistances measured under different conditions (e.g. temperature, SOC); and records the results of the analysis historically in the storage unit (3).

[0020] The battery (2) included in the inventive system (1) is an energy storage unit for storing the energy required to run an electric vehicle and transmitting it to the electric vehicle during the operation of the vehicle. The battery (2) comprises at least one cell for storing the electrical energy. The battery (2) is in communication with the control unit (4) and is configured to ensure that the data related to the temperature, resistance and state of charge occurring in the cells during its operation are received by the control unit (4).

[0021] The storage unit (3) included in the inventive system (1) is in communication with the control unit (4) and is configured to be managed by the control unit (4). The storage unit (3) is configured to receive the analysis results together with the temperature, resistance, state of charge, historical usage voltage and current data of the battery (2) over the control unit (4) and then to record these.

[0022] The control unit (4) included in the inventive system (1) is in communication with the battery (2); performs resistance measurement in the battery cells under the condition that the vehicle is traveling at constant speed over the battery (2) and analyses it together with the state of charge (SOC), temperature of the battery (2) at that moment; the state of health (SOH) of the battery (2) at the moment when the previous measurement condition is met; and the data of voltage, current during the past use of the cells (in a certain time interval, time elapsed at the moment when two measurement conditions are met, At). The control unit (4) is configured to receive the resistance value occurring on the cell and the state of charge over the battery (2) by using the temperature data over the sensors located on the battery (not shown in the figures), the instantaneous current value provided to the cells during the vehicle operation, open and closed circuit voltages. The control unit (4) ensures that an estimation is made related to the state of health of the battery over the immediate data of the cells of the battery (2) as a result of the analysis and then records it on the storage unit (3). In one preferred embodiment of the invention, the control unit (4) estimates the battery health under the conditions where data is transmitted -i.e. under the conditions where the vehicle is traveling at constant speed and the resistance measurement is performed and then recorded- by means of machine learning algorithms created by using the measurements of temperature, state of charge, voltage, current, resistance recorded upon being measured under different conditions and at different times in the storage unit (3) by using machine learning algorithms and by using aging tests previously performed under different aging scenarios and environmental conditions and resistance measurement data under different conditions. The control unit (4) is configured to estimate the health condition of the battery (2) by means of machine learning algorithm previously created, by using the state of charge, resistance measurement data, temperature data of the vehicle instantly and the historical usage data (current, temperature, voltage profile) of the vehicle after the previous SOH estimation when it is triggered on a condition (when the vehicle is travelling at constant speed for a while and the resistances are measured). In one preferred embodiment of the invention, the control unit (4) ensures to estimate the battery (2) health by analysing the data received over the battery (2) upon triggering the vehicle, together with the historical data. The control unit (4) is configured to ensure that the estimated health data of the battery (2) is shared with an electronic computer (not shown in the figures) located on the vehicle and / or an electronic device owned by the driver.

[0023] Industrial Application of the Invention

[0024] In the inventive system (1), the battery (2) comprises at least one cell for storing the energy required to run an electric vehicle and enables the energy stored on the cell to be used during the operation of the electric vehicle. The storage unit (3) receives the state of charge, resistance measurement data, temperature data of the battery (2) cells during the operation of the vehicle, the voltage, current data of during the past use of the cells (in a certain time interval, time elapsed at the moment when two measurement conditions are met, At) on a predetermined condition and then records it together with time data. The control unit (4) instantly estimates the health of cells by means of supervised machine learning algorithms created to estimate the health of a trained battery by using the information previously recorded in the storage unit

[0025] (3) of the electric vehicle each time the vehicle moves in constant speed mode and using the data of battery health obtained from battery tests previously aged under different ambient conditions and usage scenarios and cell resistances measured under different conditions; and records it in the storage unit (3). The control unit

[0026] (4) comprises a model for estimating the battery (2) health by processing each recorded data by means of machine learning algorithms. The control unit (4) and the storage unit (3) ensures to provide an estimation of the battery (2) health by receiving the state of charge, resistance measurement data, temperature data and the data of voltage, current during the past use of the cells (in a certain time interval, the time elapsed between two measurement conditions, At) over the battery (2) instantly upon being triggered by the user in a condition where the vehicle is traveling at constant speed. The control unit (4) ensures the driver to be informed by transmitting the results of the estimation performed to the driver by means of an electronic computer located on the vehicle and / or an electronic device owned by the driver.

[0027] Within these basic concepts; it is possible to develop various embodiments of the inventive “A System Estimating Battery Health (1)”; the invention cannot be limited to examples disclosed herein and it is essentially according to claims.

Claims

CLAIMS1. A system (1) for regularly estimating the health of the battery that stores the energy required to run electric vehicles; comprising at least one battery (2) which stores the energy required to run an electric vehicle and comprises at least one cell that generates resistance during the storage and / or utilization of the energy; and characterized by at least one storage unit (3) which performs resistance measurement in the battery cells under the condition that the vehicle is traveling at constant speed during operation of the electric vehicle and wherein the state of charge (SOC), temperature of the battery (2) at that moment; the state of health (SOH) of the battery (2) at the moment when the previous measurement condition is met; and the data of voltage, current during the past use of the cells (in a certain time interval, time elapsed between two measurement conditions, At) are kept under record historically; and at least one control unit (4) which is in communication with the battery (2) and the storage unit (3) and instantly estimates the health of cells by means of supervised machine learning algorithms created to estimate the health of a trained battery by using the information previously recorded in the storage unit (3) each time the vehicle moves in constant speed mode and using the data of battery health obtained from battery tests previously aged under different ambient conditions and usage scenarios and cell resistances measured under different conditions; and records the results of the analysis historically in the storage unit (3).

2. A system (1) according to Claim 1; characterized by the battery (2) which is an energy storage unit for storing the energy required to run an electricvehicle and transmitting it to the electric vehicle during the operation of the vehicle.

3. A system (1) according to Claim 1 or 2; characterized by the battery (2) which comprises at least one cell for storing the electrical energy.

4. A system (1) according to any one of the preceding claims; characterized by the battery (2) which is in communication with the control unit (4) and is configured to ensure that the data related to the temperature, resistance and state of charge occurring in the cells during its operation are received by the control unit (4).

5. A system (1) according to any one of the preceding claims; characterized by the storage unit (3) which is in communication with the control unit (4) and is configured to be managed by the control unit (4).

6. A system (1) according to any one of the preceding claims; characterized by the storage unit (3) which is configured to receive the analysis results together with the temperature, resistance, state of charge, historical usage voltage and current data of the battery (2) over the control unit (4) and then to record these.

7. A system (1) according to any one of the preceding claims; characterized by the control unit (4) which is in communication with the battery (2); performs resistance measurement in the battery cells under the condition that the vehicle is traveling at constant speed over the battery (2) and analyses it together with the state of charge, temperature of the battery (2) at that moment; the state of health of the battery (2) at the moment when the previous measurement condition is met; and the data of voltage, current during the past use of the cells.

8. A system (1) according to any one of the preceding claims; characterized by the control unit (4) which is configured to receive the resistance value occurring on the cell and the state of charge over the battery (2) by using the temperature data over the sensors located on the battery (not shown in the figures), the instantaneous current value provided to the cells during the vehicle operation, open and closed circuit voltages.

9. A system (1) according to any one of the preceding claims; characterized by the control unit (4) which ensures that an estimation is made related to the state of health of the battery over the immediate data of the cells of the battery (2) as a result of the analysis and then records it on the storage unit (3).

10. A system (1) according to any one of the preceding claims; characterized by the control unit (4) which the control unit (4) estimates the battery health under the conditions where data is transmitted -i.e. under the conditions where the vehicle is traveling at constant speed and the resistance measurement is performed and then recorded- by means of machine learning algorithms created by using the measurements of temperature, state of charge, voltage, current, resistance recorded upon being measured under different conditions and at different times in the storage unit (3) by using machine learning algorithms and by using aging tests previously performed under different aging scenarios and environmental conditions and resistance measurement data under different conditions.

11. A system (1) according to Claim 10; characterized by the control unit (4) which is configured to estimate the health condition of the battery (2) by means of machine learning algorithm previously created, by using the state of charge, resistance measurement data, temperature data of the vehicle instantly and the historical usage data (current, temperature, voltage profile) of the vehicle after the previous SOH estimation when it is triggered on acondition (when the vehicle is travelling at constant speed for a while and the resistances are measured).

12. A system (1) according to any one of the preceding claims; characterized by the control unit (4) which is configured to estimate the battery (2) health by analysing the data received over the battery (2) upon triggering the vehicle, together with the historical data.

13. A system (1) according to any one of the preceding claims; characterized by the control unit (4) which is configured to ensure that the estimated health data of the battery (2) is shared with an electronic computer located on the vehicle and / or an electronic device owned by the driver.

Images