FR3134356A1 - DE-RATE CONTROL SYSTEM FOR A MOTOR VEHICLE BATTERY, METHOD AND VEHICLE BASED ON SUCH A SYSTEM - Google Patents

Abstract

The invention relates to a derating control system for a battery (B) based on machine learning model data for evaluating the maximum power with which the temperature (Ti) of the components remains at a maximum temperature in continuous operation for a significantly long continuous time, depending on the ambient temperature (Te), and the speed of the vehicle. Derating is implemented by applying a temperature margin. Data from a law of temperature increase of battery components as a function of brief accelerations and data on driving habits of the vehicle, make it possible to indicate whether short acceleration power is available, and in the affirmatively authorize said brief acceleration during derating. The invention also relates to a method and a vehicle (V) based on such a system. Figure 3

Description

DE-RATE CONTROL SYSTEM FOR A MOTOR VEHICLE BATTERY, METHOD AND VEHICLE BASED ON SUCH A SYSTEM

The invention relates to the field of motor vehicle traction battery battery pack systems. The invention relates more particularly to the control of derating triggering (or “derating” in English) of the electronic components of the battery pack.

Overheating presents itself as a constraint on the operation of the electronic components in the battery pack. For example, busbars, harnesses, relays, and fuses malfunction above their maximum temperatures.

To protect them, derating, namely the reduction of power, is requested by a battery management system (or “BMS” for “Battery Management System” in English, or even Electric Vehicle Control Unit, abbreviated “eVCU "). Derating helps prevent component failure.

The prior art solution for derating electronic components is based on “time-power” curves of all relevant components C1, C2. These curves indicate how long a power can persist before having to drop to a lower level.

The “time – power” curve is based on the temperature measurements of component C1, C2 while ensuring the same cooling conditions as in the real application. The ambient temperature is fixed at a higher value in the actual application. The time for a certain power is determined when the maximum temperature of the component is reached.

A disadvantage of the solution of the prior art is the risk of loss of control of the vehicle during its use, because the derating law of the prior art LD risks indicating a surprising drop in power. Indeed, it may happen that the power of the vehicle drops surprisingly, if the power in use is equal to the power limit.

In addition, the derating law of the prior art LD may make high power unavailable for occasional accelerations that may be necessary for driving safety. Indeed, it may happen that the higher power desired for overtaking or on the acceleration path is missing.

Thus, a first objective of the invention is to propose a solution for controlling the triggering and the derating process in a more intelligent manner to provide the user with a better driving experience.

A second objective is to control the derating level in a more intelligent way to provide the user with better comfort and better safety while driving.

To achieve this objective, the invention proposes a derating control system for a motor vehicle traction battery, the control system comprising:
- data from a first automatic learning model for thermal simulation of battery components,
- means for evaluating the temperature of the components in a future horizon on the basis of data from said first model;
- data from a second machine learning model for evaluating maximum power with which the temperature of the components remains at a maximum temperature in continuous operation for a significantly long continuous time, as a function of the ambient temperature, and the speed of the vehicle,
- a derating means for reducing, when a threshold component temperature is reached in the future, the power requested from the battery towards said maximum power, for a reduction time,
the derating means applying a temperature margin to determine the threshold temperature lower than said maximum temperature,
- data from a law of temperature increase of battery components as a function of brief accelerations and data on driving habits of the vehicle, the brief acceleration being carried out for a short time significantly less than the continuous time,
- a means for indicating whether brief acceleration power is available, on the basis of said law and second model data,
the derating means authorizing said brief acceleration during derating if said brief acceleration power is available.

Advantageously, the invention is based on the prediction of the temperature of the components by automatic learning. In addition, the invention avoids the surprising drop in power during driving. In addition, the invention allows punctual acceleration in a limited time despite derating.

According to a variant, the derating means determines the temperature margin by predictive control on the basis of the vehicle's driving habits data. This allows the use of a specific user and vehicle specific margin.

According to a variant, the derating means is configured to determine a sufficiently long power reduction time for user comfort. This limits sudden drops in power and improves the user's driving comfort.

According to a variant, the control system includes a means of assisting in the selection of a brief acceleration intensity, based on data on the vehicle's driving habits. This allows the user to be given the hand for such an assisted selection.

According to a variant, the derating means is maintained until the temperature of the battery becomes lower than 50°C, preferably lower than 40°C. This makes it possible to avoid too many repetitions of the derating operation, for better driving comfort.

According to one variant, the brief acceleration lasts less than 60s, preferably less than 30s. This makes it possible to avoid impacting the temperature of the components and to meet the needs for occasional short accelerations.

According to a variant, the system further comprises means of signaling overheating by a light placed on the dashboard, and/or means of signaling short acceleration power available by a light placed on the dashboard. This helps inform the user accordingly.

The invention further relates to a derating control method for a motor vehicle traction battery, the control method comprising:
- a step of producing a first automatic learning model for thermal simulation of the battery components,
- means for evaluating the temperature of the components in a future horizon on the basis of data from said first model;
- a step of producing a second automatic learning model for evaluating maximum power with which the temperature of the components remains at a maximum temperature in continuous operation for a significantly long continuous time, as a function of the ambient temperature and the vehicle speed,
- a derating step to reduce, when a threshold component temperature is reached in the future, the power requested from the battery towards said maximum power, for a reduction time,
the derating step being carried out by applying a temperature margin to determine the threshold temperature lower than said maximum temperature,
- a step of producing a law for increasing the temperature of battery components as a function of brief accelerations and vehicle driving habits data, the brief acceleration being carried out for a short time significantly less than the continuous time ,
- a step to indicate whether brief acceleration power is available, on the basis of said law and models,
the derating step authorizing said brief acceleration during derating if said brief acceleration power is available.

Another object of the invention relates to a computer program comprising program code instructions for executing the steps of the derating control method according to the invention, when said program operates on a computer.

The invention further relates to a motor vehicle comprising a derating control system according to the invention.

The invention will be further detailed by the description of non-limiting embodiments, and on the basis of the appended figures illustrating variants of the invention, in which:
- schematically illustrates a derating curve according to the prior art as a function of time t and power P;
- schematically illustrates the realization of the machine learning models used in a preferred variant of the invention; And
- schematically illustrates the implementation of machine learning models in the context of the preferred variant of the invention.

Overheating presents itself as a constraint on the operation of electronic components C1, C2 in the battery pack.

The current derating solution is based on “time-power” curves of all relevant components C1, C2. The curves of components C1, C2 are provided by suppliers who measure the temperature evolution of their components with the same cooling conditions and under a defined maximum ambient temperature.

For a certain power, component C1, C2 reaches the maximum temperature it can sustain after a certain time. This pair of “power – time” values corresponds to a point on the curve.

There is a power with which component C1, C2 can operate for an indefinitely long time at a defined cooling condition. There is also a power limited by other factors (electrical, flammability, etc.) where operation is prohibited regardless of temperature. All the curves of components C1, C2 form a common zone where the operation of all components C1, C2 is permitted.

To have a safety margin, the derating law is defined step by step in a common area. The derating curve LD indicates how long a power P can persist before having to drop to a lower level, shown by the .

This derating law has two disadvantages. The first is control precision.

Derating can be triggered before component C1, C2 reaches its maximum temperature, for example when the ambient temperature is less severe and the component heats up from ambient temperature.

On the other hand, derating can be activated too late. For example, after operating for a certain time, the temperature of component C1, C2 becomes higher than the initial temperature in the tests. Then, the time to reach the maximum temperature of the component becomes shorter for the same power.

The co-pending application FR2100225 makes it possible to predict the temperature of component C1, C2 by automatic learning, which makes it possible to achieve a reaction-like control, and thus resolves this disadvantage.

The second disadvantage is the risk of losing control of the vehicle during use in two cases. It may happen that the vehicle's power drops sharply and surprisingly, because the derating is done in stages as shown in .

It may also happen that a desired high point power is absent, for example for an overtaking event or when entering the highway in the acceleration lane, because the derate limitation line is a rigid control which does not allow overtaking.

The invention relates in particular to the second aforementioned disadvantage concerning the control law taking advantage of it, preferably the co-pending application FR2100225.

The invention proposes to have a margin at the maximum temperature of the components C1, C2. In other words, derating takes place as soon as the temperature of components C1, C2 reaches a temperature lower than the maximum temperature.

The means implemented in the invention are part of a battery controller C, which can itself be part of a battery management system (generally called BMS)

At the start of derating, the temperature increases despite the reduction in power. After a certain level of decrease, the temperature of component C1, C2 starts to drop. The primary objective of the margin is to allow a slow reduction in power which, in combination with an overheating indicator on the dashboard, gives the user a gentler warning of the unavailability of power.

The margin is naturally component specific, because each component C1, C2 has its own thermal characteristics. The margin is also obviously specific use because the power used and the thermal conditions of the vehicle strongly influence the thermal evolution of components C1, C2.

To achieve an intuitive reduction rate for the user, this invention leads to a second integration of the same means described by the co-pending application FR2100225, where the calculated temperature gradient is taken as a constant value for a certain future time (future horizon ), designated by Δt1. In other words, for each moment, the temperature at future Δt1 must be calculated assuming that the thermal conditions at this moment persist.

In the case of a continuous increase in temperature, this second integration gives a higher temperature. As soon as the temperature at Δt1 reaches the maximum temperature of component C1, C2, derating is triggered.

After triggering, the power must be lowered to the power which allows operation over an indefinitely long time, for example t _C which can reach 3 hours, or even more. This power must be maintained for a certain time. We can speak of maximum power for this purpose.

The co-pending demand model FR2100225 makes it possible to find this power for each component C1, C2 as a function of the thermal conditions: namely the ambient temperature and the speed of the vehicle. The process is described as follows:
- For each combination of ambient temperature and vehicle speed, for example 20°C + 50 km/h, the co-pending demand model FR2100225 can find the power with which the component temperature remains at its maximum value for “unlimited” time.
- With input data like ambient temperature and vehicle speed, and output data like maximum power found in step 1, a new machine learning M2 model is developed.
- This modeling (M2 model) must be repeated for each component C1, C2, etc., and the M2 model which gives the lowest power is used for prediction as a specific vehicle model.

Component suppliers normally provide this power but only as a value by fixing the thermal conditions. With the co-pending demand model FR2100225, this power is a value depending on the thermal conditions which vary continuously during driving.

The difference between the current power and the power for continuous operation, ΔP, and the time for this decrease, denoted by Δt2, give the rate of decrease in power. Preferably, Δt2 is between 5 and 10 seconds. It is understandable and intuitive for the user that in a case with a lower ambient temperature Te and a higher vehicle speed, the power difference, ΔP, is lower. The reduction in power is therefore less severe. Otherwise, the user experiences a faster reduction in power.

The reduction time, Δt2 (or reduction time), is defined by the comfort requirement or by engineering experience during vehicle development. This is not a perfectly objective value, and is a compromise between the power available for the general case and the user's sensation of the reduction in power during derating.

A slower decrease implies an earlier derate trigger, and thus lower power available in general. The tests can be carried out at the pack level during thermal tests or at the vehicle level during the summer mission, because the reduction is greatest in extreme heat. The objective is to find the shortest Δt2, only to alert the user in combination with a warning light on the dashboard.

The determination of the time Δt1 is linked to Δt2, because during the first part of Δt2, the temperature of the components C1, C2 increases further. The longer Δt2 is, the higher the temperature increase during derating, thus Δt1 is also longer to avoid overheating of the components.

After determining Δt2, the value of Δt1 is to be found by the co-pending demand model FR2100225 by simulating the temperature evolution during derating (simulation model M1), and by observing the maximum temperature during the derating phase . Δt1 is then determined ensuring the absence of overheating.

After dropping down to the power that allows continuous operation for such a long time, it is necessary to maintain this power level for a certain time, allowing the entire battery pack to be cooled to the temperature where no derating begins, i.e. 40°C, the upper battery temperature limit. As the battery has the largest thermal capacity in the battery pack, in fact much larger than all other electronic components, derating up to 40°C sufficiently avoids frequent repetition of derating triggering of electronic components C1, C2.

Δt1 and Δt2 do not provide sufficient control: the second objective of this margin is to allow a certain level of acceleration for a certain time in certain specific cases. On the acceleration lane or when overtaking, the user needs more power when derating can be triggered at the same time. To respect this user need and, at the same time, protect the electronic components C1, C2, it is necessary to have a greater margin to allow this high power for a limited period.

The need for acceleration depends on a specific user. Users who often travel on motorways need greater acceleration than users who travel only in city centers, for example. This invention proposes to learn the user's habit through a statistical study for example: the on-board computer records the acceleration and its duration, then transfers this data to a mobile application, where the results of analysis are made available to the user: the acceleration as a function of time, the maximum and average level of acceleration, the maximum and average level of the acceleration duration, the maximum and average acceleration index which is the multiplication of the acceleration and its duration. The client has the right to choose a percentage index to cover his needs: 100% to cover all acceleration events and 50% to cover the average index for example.

The second law is constructed by the increase in temperature, ΔT, during point acceleration. The acceleration index implies a specific increase in temperature which is also recorded by the on-board computer. By choosing the acceleration index as a percentage, the temperature increase coverage is also chosen at the same level. The acceleration index is preferably provided to the user because it is more intuitive and understandable than the temperature increase.

The final derating law is a combination of the law based on the temperature variation ΔT and the law based on the future horizon Δt1: derating is to be triggered as soon as the temperature predicted for Δt1 later reaches the maximum temperature of the components – ΔT, thus the derating is triggered earlier than the law on the future horizon Δt1 in itself.

The following points should be noted for this acceleration index:

The availability of power for one-off acceleration implies lower power in general, because derating has to be triggered earlier. The greater the index coverage, the lower the power for general cases.

Preferably only occasional accelerations of the vehicle are recorded, typically less than t _A =30s. Due to this short duration, the environmental impact (ambient temperature and vehicle speed) on the occasional temperature increase is negligible. Therefore, the temperature increase is only related to acceleration.

To obtain high power when derating is already activated, the user must push the Pd accelerator pedal all the way, as confirmation of their need for power. This acceleration is only present for a limited time. As soon as the law of the future horizon Δt1 without consideration of the temperature reduction ΔT, is applicable, the power switches to decrease without respecting the position of the accelerator pedal Pd. Indeed, in this case, the temperature threshold to trigger derating is higher. Preferably, Δt1 is between 60 and 120 seconds. Acceleration in this condition can be informed to the user by a warning light on the dashboard to alert the short presence of high power.

A statistical study on acceleration habit is feasible because most occasional acceleration events do not affect derating, so the vehicle operates normally. Furthermore, the thermal characteristics of electronic components do not depend significantly on temperature, but rather on the battery. The data accumulated in winter and summer are equivalent.

In summary, this invention proposes two laws of control of derating triggering preferably based on the prediction of the temperature of the components by the automatic learning proposed in co-pendant FR2100225. The law based on time prevents the user from being surprised by the reduction in vehicle power. The law based on temperature increase ensures the availability of desired high power for a limited time.

Claims (10)

Derating control system for a traction battery (B) of a motor vehicle (V), the control system comprising:
- data from a first automatic learning model (M1) for thermal simulation of the battery components,
- means for evaluating the temperature of the components (C1, C2) in a future horizon on the basis of data from said first model;
- data from a second machine learning model (M2) for evaluating maximum power with which the temperature (Ti) of the components (C1, C2) remains at a maximum temperature in continuous operation for a continuous time (t _C ) significantly long, reaching three hours or more, depending on the ambient temperature (Te), and the speed of the vehicle,
- a derating means for reducing, when a threshold component temperature is reached in the future, the power requested from the battery (B) towards said maximum power, for a reduction time,
the derating means applying a temperature margin to determine the threshold temperature lower than said maximum temperature,
- data from a law of temperature increase of battery components (C1, C2) as a function of brief accelerations and data on driving habits of the vehicle, the brief acceleration being carried out for a short time (t _A ) less than 60s significantly less than the continuous time (t _C ),
- a means for indicating whether brief acceleration power is available, on the basis of said law and second model data,
the derating means authorizing said brief acceleration during derating if said brief acceleration power is available. Control system according to claim 1, characterized in that the derating means determines the temperature margin by predictive control on the basis of vehicle driving habit data. Control system according to any one of claims 1 to 2, characterized in that the derating means is configured to determine a power reduction time sufficiently long for user comfort, the value of the reduction time being a compromise between available power and a user sensation of reduced power during derating. Control system according to any one of claims 1 to 3, comprising means for assisting in the selection of a brief acceleration intensity, on the basis of data on the vehicle's driving habits. Control system according to any one of claims 1 to 4, characterized in that the derating means is maintained until the temperature (Ti) of the battery becomes less than 50°C, preferably less than 40° vs. Control system according to any one of claims 1 to 5, characterized in that the brief acceleration lasts less than 30s. Control system according to any one of claims 1 to 6, further comprising means of signaling overheating by a light placed on the dashboard, and/or means of signaling brief acceleration power available by a light placed on the dashboard. Derating control method for a motor vehicle traction battery (V), the control method comprising:
- a step of producing a first automatic learning model (M1) for thermal simulation of the components (C1, C2) of the battery (B),
- means for evaluating the temperature of the components (C1, C2) in a future horizon on the basis of data from said first model;
- a step of producing a second automatic learning model (M2) for evaluating maximum power with which the temperature of the components (C1, C2) remains at a maximum temperature in continuous operation for a continuous time (t _C ) significantly long, reaching three hours or more, depending on the ambient temperature (Te), and the speed of the vehicle,
- a derating step to reduce, when a threshold component temperature is reached in the future, the power requested from the battery towards said maximum power, for a reduction time,
the derating step being carried out by applying a temperature margin to determine the threshold temperature lower than said maximum temperature,
- a step of producing a law for increasing the temperature of battery components (C1, C2) as a function of brief accelerations and vehicle driving habits data, the brief acceleration being carried out for a short time (t _A ) of less than 60s significantly less than the continuous time (t _C ),
- a step to indicate whether brief acceleration power is available, on the basis of said law and models,
the derating step authorizing said brief acceleration during derating if said brief acceleration power is available. A computer program comprising program code instructions for executing the steps of the derating control method of claim 8, when said program is operating on a computer. Motor vehicle (V) comprising a derating control system according to any one of claims 1 to 7.