MOTOR VEHICLE COMPRISING A TEMPERATURE SENSOR COUPLED WITH A TEMPERATURE ESTIMATION MODEL, METHOD AND PROGRAM BASED ON SUCH A VEHICLE

Abstract

The invention relates to a motor vehicle comprising a battery (BT) and a management system, the battery (BT) comprising: - preferably a housing and / or at least one terminal tab; - battery modules (M) equipped with battery cells (CB); - at least one busbar (B) arranged between two rows of battery cells (CB); - at least one temperature sensor (C) disposed on said busbar (B) and / or the housing and / or said terminal tab; - at least one model for evaluating the temperature of points on the battery; - a means for calculating the actual temperature of the points from the corresponding measurement corrected in part or in full by the data from said evaluation model. The invention also relates to a method and a program based on such a vehicle. Figure 1

Description

Title of the invention: MOTOR VEHICLE COMPRISING A TEMPERATURE SENSOR COUPLED WITH A TEMPERATURE ESTIMATION MODEL, METHOD AND PROGRAM BASED ON SUCH A VEHICLE

[0001] The invention relates to the field of battery management systems for diagnosing and monitoring the condition of a traction battery in a motor vehicle. More particularly, the invention relates to temperature measurements of battery cells.

[0002] Measuring the temperature of certain batteries, such as those with prismatic cells, is difficult. Indeed, it is not feasible to place a temperature sensor inside the cell casing in an industrial setting. Therefore, one solution proposed by the applicant is to place a temperature sensor on the cell casing, on a cell tab, or on a bus bar of the module. The temperature measured at these locations is either used directly or a temperature margin is used to approximate the cell core temperature. Ultimately, this temperature value is used as input for cell temperature control and for cell diagnostics.

[0003] Most current solutions do not consider the temperature of the cell casing which helps to calculate the heat dissipation of the refrigerant.

[0004] The objective of a battery management system is to control the core temperature of the cell, more specifically the stack (or "jellyroll") located inside the casing, where the electrochemical reactions take place. The use of existing solutions is problematic because: - The measurement location is not on the stack itself (ex-situ measurement); - The measurement location includes several materials with thermal properties different from those of the stack. This can lead to a significant measurement error compared to a measurement on the stack itself; - The definition of the margin added to the measurement is complex and insufficient to accurately represent all battery operating scenarios. Therefore, the measurement error is reduced compared to using the measurement directly, but it can still lead to a significant error; - The existing problems recur if we want to take into account the temperature gradient (minimum and maximum temperatures) of the stack. Taking the gradient into account is necessary because the cell is cooled on one side, and certain factors limiting factors depend on the minimum temperature, while other limiting factors depend on the maximum temperature; - Because the temperature inside the cell is not well known, it is difficult to determine the heat dissipated by the cooling system. This makes it more difficult to control the cooling system to provide the appropriate cooling power at the right time.

[0005] The consequence of a measurement error is: - a loss of performance when temperature becomes the most limiting factor on battery usage (in cold and hot temperature scenarios and at maximum power, during fast charging); - a loss of accuracy in battery diagnosis because it is an important input information; - A loss of performance in the control of the cooling system because the current solution may provide delayed or inaccurate information about the heat dissipated by the coolant. The cooling system may not deliver the correct amount of cooling power at the right time.

[0006] An objective of the present invention is to propose a solution to improve performance in temperature control by limiting the problems of accuracy of temperature measurements of battery cells.

[0007] To achieve this objective, the invention proposes a motor vehicle comprising a traction battery and a battery management system connected to the traction battery, the traction battery comprising: - preferably a housing and / or at least a terminal tab; - battery modules equipped with battery cells; - at least one busbar arranged between two rows of battery cells; - at least one temperature sensor placed on said bus bar and / or the housing and / or said terminal tab; - at least one temperature evaluation model for at least one point on the battery; - a means of calculating the actual temperature of said point of the battery from the corresponding temperature measurement corrected in part or in full by the data of said evaluation model.

[0008] Advantageously, the invention allows a simplified thermal model of the cell to be coupled with ex-situ measurements using an estimation method such as the Kalman filter. The simplified thermal model is composed of several thermal nodes of the stack to take into account the thermal gradient of the cell core and the heat dissipated by the cooling system while keeping the solution integrated into the car's computer.

[0009] Preferably, the calculation means uses a Kalman filter.

[0010] This improves the accuracy of temperature calculation.

[0011] Preferably, the calculation method implements a Kalman filter estimation phase defined by the formula: ~ x^ + ^k (zk~^kx^ki ) °where X^ are the states of the estimation phase; this estimation being composed of the states of the prediction phase and a term correcting this prediction; this correction being based on a weighted difference between a model Ck X^^ and a measure zk.

[0012] This makes it possible to improve the accuracy of the estimation phase, and to better calibrate the Kalman filter.

[0013] Preferably, the calculation method implements a Kalman gain Kk defined by the formula: ] Ck with Pj^ ] : the prediction covariance described by the formula: k ^~~sT~ P^ = W-Ku AtT+ Wk^: 1“ model uncertainties; and Sk: the covariance, itself defined by the formula q — r> pf T . v with Vk; the uncertainties of the measurement. ^k~^k rkk-l^k +vk

[0014] This makes it possible to improve the accuracy of the calculation of the kalman gain, and therefore of the temperature.

[0015] Preferably, said model is based on one of the following sets of thermal nodes at said points: - two thermal nodes representing the two tabs of the cell; - three thermal nodes representing the stacking of the cell; - two thermal nodes representing the casing on the large faces of the cell; and - a thermal node representing the housing on the underside of the cell.

[0016] This makes it possible to determine the temperatures at different specific points of the cell or battery.

[0017] Another object of the invention relates to a method for estimating the temperature of at least one point of the traction battery of a motor vehicle according to the invention, comprising the following steps: - a step to evaluate the temperature at said point using at least one evaluation model; - a step of calculating the actual temperature of said point of the battery from the corresponding temperature measurement corrected in part or in full by the data of said evaluation model.

[0018] Preferably, in the calculation step, a Kalman filter is used.

[0019] Preferably, in the calculation step, an estimation phase of the Kalman filter is implemented, defined by the formula: ~ +^k ^kk-i ) °ù X^ are the states of the estimation phase; this estimation being composed of the states of the prediction phase Xj^ and a term correcting this prediction; this correction being based on a weighted difference between a model Ck and a measure ^k-

[0020] Preferably, in the calculation step, a Kalman gain Kk defined by the formula is implemented: Pk^ckTavec Pldk-l: the prediction covariance described by the formula: K k = —sT~ PlM^I-Cm ) Pm \ ' + W'tav“ : the Certainties of the model; and Sk : the covariance, itself defined by the formula S = Ci P + V with • the uncertainties of the measurement. kk kkk- 1 kk

[0021] The invention further relates to a computer program comprising program code instructions for executing the steps of the estimation process according to the invention, when said program is running on a computer.

[0022] The invention will be further detailed by describing non-limiting embodiments, and based on the accompanying figures in which: - [Fig.1] schematically illustrates a plan view of a battery module comprising cells and an omnibus bar equipped with a temperature sensor, as part of a preferred variant of the invention; - [Fig.2] schematically illustrates a simplified electro-thermal model of the cell within the framework of a process and a system according to a preferred embodiment of the invention; - [Fig.3] schematically illustrates external views of a battery with the different thermal nodes whose temperature is determined by the model; - [Fig.4] schematically illustrates a view of the stacking of a cell with thermal nodes of the stacking.

[0023] In a module M, one or more temperature sensors C are placed on the bus bar B to estimate the temperature of the cell CB. In [Fig. 1], these sensors C are shown in the middle of the battery module M.

[0024] This invention proposes coupling at least one of these temperature measurements to an electro-thermal model of at least one CB cell in order to improve the cell temperature estimates. The estimation method proposed in this invention is the discrete Kalman filter, but other estimation methods could be applied.

[0025] The prediction phase at step k of the Kalman filter is defined by the formula: = A^X^-y^ + Bk Uk where are the states of the prediction phase, Ak and Bk are the matrix of the state space representation of the electro-thermal model, and uk are respectively the states of the previous estimation phase and the inputs of the estimator.

[0026] These elements will be described later in the section on the electrothermal model.

[0027] The estimation phase of the Kalman filter is defined by the formula: xkk = + ^k (zk~^k ) °where X^ are the states of the estimation phase; this estimation being composed of the states of the prediction phase Xj^ and a term correcting this prediction; this correction being based on a weighted difference between a model Ck X#^ and a measure zk-

[0028] In this invention, the proposal is to base part or all of the correction on the location where a temperature sensor is placed on the module M or the cell CB. These temperature sensors C can be placed on the bus bar of the module as shown in Figure 1, or with a measurement on the tab or the cell housing. Therefore, Ck Xj^ is a thermal model relating the thermal nodes of the cell to the temperature at the measurement location. The invention described below will take the positive tab temperature as the temperature measurement location for the estimation.

[0029] The weighted coefficient applied to the difference is defined by the Kalman gain Kk where: Pkk ! Ck with P^ । : the prediction covariance described by the formula: - Ak ( I - KkA CkA ) PAk + W / vec Wk: the uncertainties of the model; and Sk: the covariance, itself defined by the formula $ =CPCT -i- V with Vk ; the uncertainties of the measurement. kkk\K~ ï kk

[0030] It is interesting to note that the weight of the correction is based on the thermal model and its uncertainties. Therefore, if we have confidence in the measurement, the estimator will correct the model while respecting the physics described in the thermal model and taking into account the model uncertainties of each thermal node.

[0031] Having described the estimation method, let us explain how to obtain the remaining elements of the Kalman filter (A^ BkCk...') with the description of a simplified electrothermal model of the cell. This model is described in the diagram in [Fig.2].

[0032] In this figure, NT is a thermal node, therefore representing a temperature deduced by the model calculation (here in °C); ET is an inlet temperature (here in °C) obtained by measurement or by another function in the battery management system; CT is a thermal conductance between two thermal nodes NT (W / K); GC is a heat generator (in W) associated with a thermal node NT.

[0033] This model is composed of 8 thermal nodes NT, two representing the two tabs of the cell (NT1, NT2), three representing the stacking of the cell (NT3, NT4, NT5), two representing the housing on the large faces of the cell (NT6, NT7), and one representing the housing on the lower face of the cell (NT8).

[0034] The position of each of these thermal nodes NT on a CB cell and the temperature measurement for this example can be represented as illustrated in figures 3 and 4.

[0035] The location of these thermal nodes allows the minimum, maximum, and average stack temperatures to be calculated using NT1, NT2, and NT3. This information is useful for feeding the battery management system estimators to indicate the state of charge and available power. The heat dissipated by the coolant can be evaluated using NT8. This is useful for informing the cooling system of the heat that will soon reach the cooler in order to adjust its response to the cooling requirement.

[0036] The invention further relates to a method and a pre-check program implementing the elements discussed above. The program can be loaded into the memory of a vehicle battery management system or into that of a specific computer.

[0037] Alternatively, within the scope of the invention, one may consider: - a variant with measurement of the case temperature, and a measurement of the bus bar temperature; - a variant with several temperature measurements; - a variant with a slightly different model: with more or less heat exchange between the thermal nodes; more or fewer thermal nodes on the faces of the casing; more or fewer thermal nodes on the tabs; more or fewer thermal nodes on the stacks; more thermal nodes for the electrolyte; more thermal nodes for the internal ambient air inside the casing; more thermal nodes for one or both of the current collectors; - a variant with other estimation techniques (Luenberger for example); - a variant with a different method to calculate the heat generated by the stacks: addition of entropy; use of a voltage or overvoltage model.

Claims

Demands

1. Motor vehicle comprising a traction battery (BT) and a battery management system connected to the traction battery (BT), the traction battery (BT) comprising: - preferably a housing and / or at least one terminal tab (L); - battery modules (M) equipped with battery cells (CB); - at least one bus bar (B) arranged between two rows of battery cells (CB); - at least one temperature sensor (C) disposed on said bus bar (B) and / or the housing and / or said terminal tab (L); - at least one temperature evaluation model of at least one point on the battery (NT1, NT2, NT3, NT4, NT5, NT6, NT7, NT8); - a means of calculating the actual temperature of said point of the battery (NT1, NT2, NT3, NT4, NT5, NT6, NT7, NT8) from the corresponding temperature measurement corrected in part or in full by the data of said evaluation model.

2. Motor vehicle according to claim 1, characterized in that the calculation means uses a Kalman filter.

3. A motor vehicle according to claim 2, characterized in that the calculation means implements a Kalman filter estimation phase defined by the formula: + Kk ( h - ck x^ ) where are the states of the estimation phase; this estimation being composed of the states of the prediction phase X^^ and a term correcting this prediction; this correction being based on a weighted difference between a model QX^ and a measure %k-

4. A motor vehicle according to any one of claims 1 to 3, characterized in that the calculation means implements a Kalman gain Kk defined by the formula: P^, C? with j: the prediction covariance described by the formula: = Ak(I-KkA Ck^ P^ AkT+Wk^ uncertainties of the model; and Sk: the covariance, itself defined by the formula v _ fpc T aV with V: the uncertainties of the measurement. ^k ^kik-ï ^k ' ' k

5. Motor vehicle according to any one of claims 1 to 4, characterized in that said model is based on one of the following sets of thermal nodes at said points: - two thermal nodes (NT1, NT2) representing the two tabs of the cell; - three thermal nodes (NT3, NT4, NT5) representing the stacking of the cell (NT3, NT4, NT5); - two thermal nodes (NT6, NT7) representing the housing on the large faces of the cell; and - one thermal node (NT8) representing the housing on the lower face of the cell.

6. A method for estimating the temperature of at least one point (NT1, NT2, NT3, NT4, NT5, NT6, NT7, NT8) of the traction battery of a motor vehicle according to any one of claims 1 to 5, comprising the following steps: - a step of evaluating the temperature of said point (NT1, NT2, NT3, NT4, NT5, NT6, NT7, NT8) by means of at least one evaluation model; - a step of calculating the actual temperature of said point of the battery (NT1, NT2, NT3, NT4, NT5, NT6, NT7, NT8) from the corresponding temperature measurement corrected in part or in whole by the data of said evaluation model.

7. Estimation method according to claim 6, characterized in that in the calculation step, a Kalman filter is used.

8. An estimation method according to claim 7, characterized in that, in the calculation step, an estimation phase of the Kalman filter is implemented, defined by the formula: + Kk(zk-Ck x^) where x* are the states of the estimation phase; this estimation being composed of the states of the prediction phase and a term correcting this prediction; this correction being based on a weighted difference between a model Ck and a measurement Zk-

9. An estimation method according to any one of claims 6 to 8, characterized in that in the calculation step, a Kalman gain Kk defined by the formula is implemented: P^ cj with P^ J: the prediction covariance described by the formula: = A^IK^ + the model uncertainties; and: the covariance, itself defined by the formula p T +y with: the measurement uncertainties.

10. Computer program comprising program code instructions for performing the steps of the estimation process according to any one of claims 6 to 9, when said program is running on a computer.

Images