Method of determining the state of power of a traction battery of an electric vehicle, corresponding system, vehicle and computer program product

Abstract

It is disclosed a method for determining the state of power of a battery (B) of an electric vehicle (V). The state of charge (SOC), the temperature (T), the voltage (V) provided by the battery and the current (I) provided by the battery are sensed. A first predicted power limit Formula (I) is computed (2), as a function of the state of charge, the temperature, the provided voltage and the provided current. The first predicted power limit is computed for a first time interval having a duration higher than 1 second, and is equal to the power that, if continuously provided or absorbed by the battery over the first time interval, brings the voltage (V Cell ) of the battery respectively to a minimum (V min ) or maximum limit value. A second predicted power limit Formula (II) is computed (2), as a function of the state of charge, the temperature, the provided voltage and the provided current. The second predicted power limit is computed for a second time interval having a duration lower than 1 second, and is equal to the power that, if continuously provided or absorbed by the battery over the second time interval, brings the voltage of the battery respectively to the minimum or maximum limit value. The power (P Bat ) provided or absorbed instantaneously by the battery is sensed, or computed as a function of the provided voltage and the provided current. In response to (C2) the power provided or absorbed instantaneously by the battery being lower than the first predicted power limit, the value of a real power limit Formula (I) is set (41) to the present value of the first predicted power limit Formula (I). In response to (C1) the power provided or absorbed instantaneously by the battery reaching the first predicted power limit, the value of the real power limit is set (42) to the minimum value selected between: (a) the instantaneous value of the first predicted power limit at the time (t 1 ) when the power provided or absorbed instantaneously by the battery reaches the first predicted power limit, and (b) the present value of the second predicted power limit.

Description

[0001] “Method of determining the state of power of a traction battery of an electric vehicle, corresponding system, vehicle and computer program product”

[0002] * * * *

[0003] TEXT OF THE DESCRIPTION

[0004] Field of the invention

[0005] The present invention relates to a method and a system for determining the State Of Power (SOP) of a traction battery of a Battery Electric Vehicle (BEV). The invention also relates to a corresponding electric vehicle and a corresponding computer program product.

[0006] The state of power of a battery is defined as the (peak) power that the battery can continuously provide to the power system of the vehicle in a certain time interval. The same definition applies to the power that the battery can continuously receive from the power system of the vehicle, when the electric motor works as a generator during regenerative braking. The state of power is a function of the state of charge (SOC), the temperature, and the present conditions of use (current and voltage) of the battery.

[0007] Prior art

[0008] The known algorithms for estimating the state of power of a high- voltage battery (here also referred to as “SOP algorithms”) aim to compute a set of power limits that define the operating conditions of the battery itself, given the state of charge (SOC), the temperature, and the present conditions of use (measured voltage and current) of the battery. Typically, three different power limits are computed (valid for both battery discharge and battery charging), namely a short-term limit, a medium-term limit, and a long-term limit, which correspond to the power values that can be continuously provided (or absorbed) by the battery over three different time intervals. For example, the short-term, medium-term, and long-term limits can be related to intervals of 2 seconds, 10 seconds, and 30 seconds, respectively. Such SOP algorithms are usually executed by a control unit associated with the battery and / or the Battery Management System (BMS).

[0009] Typically, battery cell manufacturers provide a datasheet specifying a set of maps showing the current limit values that can be provided or absorbed by each battery cell. Each map is associated to a time interval, and is a function of the state of charge and the temperature of the battery. In this context, “limit value of current over a certain time interval” means the current that can be continuously provided or absorbed by the battery cell for that time interval, before reaching respectively the minimum or maximum admissible voltage. These maps are obtained experimentally and, for each pair of points (SOC, T) of the map - where SOC is the state of charge and T is the temperature - the limit value of current is determined by placing the battery in a stationary condition, in which the provided or absorbed current I is zero ( / = 0) and the voltage \ / across the cell is equal to the open circuit voltage VOC (V = VOC(SOC, T)\ Furthermore, the current limit values indicated in such maps are usually saturated (i.e., limited to a maximum value) at respective current values that can actually be withstood by the battery components (e.g., electrical cables, connectors). This is because, in a certain range of SOC and temperature values (in particular, high SOC and T), the current that in a given time interval brings the single cell to the minimum or maximum voltage could assume very high values, which some battery components (or external components) cannot withstand.

[0010] Even if obtained experimentally (where current saturation is not active), such maps are not entirely suitable for describing the real conditions of use of traction batteries in electric vehicles, because at a certain time the current I is not zero and the voltage \ / is not equal to the open circuit voltage. Nevertheless, it is possible to compute the current limit value using a so- called Equivalent Circuit Model (ECM); in this case, depending on the accuracy of the model, the current limit value may correspond to the operating conditions of the battery. In electric vehicles, current limit values must then be converted into power limit values in order to maximize the performance of the powertrain (PWT) of the vehicle. The conversion between current limit values and power limit values can be implemented in different ways and therefore the use of maps alone or their integration with a modelling computation is a design choice.

[0011] In any case, the power limits thus computed define safety margins for the operation of the high-voltage battery (i.e., a margin with respect to the minimum voltage or maximum voltage that are specified in the battery cell datasheet, respectively if in the discharge or charging phase). These margins are more conservative the longer the time interval to which the power limit is associated (i.e., the short, medium or long term mentioned above). When the power provided by the battery reaches the limit value, the voltage of the battery remains constant, and this implies that the value of the current provided by the battery decreases over time, in turn determining a decrease in the provided power. In terms of driving experience, this translates into the fact that during a vehicle launch maneuver or more generally during an acceleration with the accelerator pedal fully pressed, if the power provided by the battery reaches the limit value defined by the state of power, a power peak is reached which is immediately followed by a decay of the provided power, as the power limit value (which is cyclically computed by a vehicle control unit) also starts to decrease. In particular, such decay of the provided power occurs when the state of charge of the battery is below a certain threshold, and the power limit provided by the SOP algorithm becomes lower than the maximum traction power that would be accepted by the powertrain control module (i.e. , when the propulsion is “battery-limited”, that is limited by the performance of the battery rather than by the performance of the electric traction system). For example, in some vehicles the powertrain performance is limited by the state of charge of the battery when the state of charge (SOC) decreases below 80%. Furthermore, the difference between the maximum power provided considering the limits established by the SOP algorithm and the power that can actually be provided (linked to the minimum or maximum admissible voltage, considering a zero margin) represents a margin that is never used (and which is greater the longer the time interval over which the power limit is computed, i.e. 2 s, 10 s or 30 s). The decay of the power provided during the acceleration maneuver, and the fact that a part of the deliverable power is in fact never used, are aspects considered unsatisfactory in particular in the case of high-performance electric vehicles.

[0012] Therefore, there is a need to develop an improved method for determining the state of power of a traction battery of an electric vehicle, which allows the computation of power limits that are closer to the real physical limits of the battery, and which therefore allows for better performance of the electric vehicle, in particular a better and more constant power delivery. Object of the invention

[0013] The object of the present invention is to provide such an improved method for determining the state of power of a traction battery of an electric vehicle, as well as a corresponding system, a corresponding electric vehicle and a corresponding computer program product. At least part of the method can be implemented by means of an electronic processor.

[0014] The precise computation of the state of power of the battery can be used advantageously in high-performance vehicles to maximize the power delivery at certain times (for example, during the execution of launch or overtaking maneuvers) and / or to avoid power “gaps” due to the progressive reduction of the power delivery when the state of charge (SOC) of the battery is lower than a certain threshold value. The method according to the invention therefore allows to exploit more deeply the physical limits of the traction battery, in particular during the acceleration of the vehicle, thus improving the performance of the vehicle and the driving experience, i.e. , providing higher and / or more constant acceleration (without power “gaps”), regardless of the state of charge (SOC) of the battery.

[0015] Summary of the invention

[0016] According to an aspect of the present disclosure, the invention relates to a method of determining the state of power of a traction battery of an electric vehicle. The method comprises the following steps: i) sensing the state of charge of the battery, the temperature of the battery, the voltage provided by the battery and the current provided by the battery; ii) computing, as a function of the state of charge, the temperature, the provided voltage and the provided current, at least one first predicted power limit, the at least one first predicted power limit being computed for at least one first time interval having a duration higher than 1 second, wherein the at least one first predicted power limit is equal to the power that, if continuously provided or absorbed by the battery over the at least one first time interval, brings the voltage of the battery respectively to a minimum or maximum limit value;

[0017] (iii) computing, as a function of the state of charge, the temperature, the provided voltage and the provided current, a second predicted power limit, the second predicted power limit being computed for a second time interval having a duration lower than 1 second, wherein the second predicted power limit is equal to the power that, if continuously provided or absorbed by the battery over the second time interval, brings the voltage of the battery respectively to the minimum or maximum limit value; iv) sensing, or computing as a function of the provided voltage and the provided current, the power provided or absorbed instantaneously by the battery; v) in response to the power provided or absorbed instantaneously by the battery being lower than the at least one first predicted power limit, setting the present value of at least one respective real power limit to the present value of the at least one first predicted power limit; and vi) in response to the power provided or absorbed instantaneously by the battery reaching the at least one first predicted power limit, setting the present value of the at least one respective real power limit to the minimum value selected between: (a) the instantaneous value of the at least one first predicted power limit at the time when the power provided or absorbed instantaneously by the battery reaches the at least one first predicted power limit, and (b) the present value of said second predicted power limit.

[0018] As will be shown in greater detail in the description that follows, the main idea underlying the present invention is to compute, in addition to one or more “conventional” power limits, an “instantaneous” power limit, and allow the battery to provide a power greater than that dictated by the conventional limit, in compliance with the instantaneous limit.

[0019] According to a further aspect of the present disclosure, the invention relates to a system for determining the state of power of a high-voltage traction battery of an electric vehicle. The system comprises a plurality of sensors configured to sense the state of charge of the battery, the temperature of the battery, the voltage provided by the battery and the current provided by the battery, and an electronic processing unit configured to receive data from the plurality of sensors and to carry out the method according to one or more embodiments of the invention.

[0020] According to a further aspect of the present disclosure, the invention relates to an electric vehicle comprising a high-voltage traction battery and a system according to one or more embodiments of the invention.

[0021] According to a further aspect of the present disclosure, the invention relates to a computer program product loadable in the memory of an electronic processing unit and comprising instructions which, when the program is executed by the electronic processing unit, cause the electronic processing unit to carry out the method according to one or more embodiments.

[0022] Detailed description of the invention

[0023] Further features and advantages of the invention will be apparent from the following description with reference to the attached drawings, provided purely by way of non-limiting example, in which:

[0024] - Figure 1 is a circuit diagram representing an equivalent circuit model (ECM) of a battery cell;

[0025] - Figure 2 is a block diagram illustrating some operations carried out by an algorithm for computing the power limits of a high-voltage battery, according to one or more embodiments;

[0026] - Figure 3 is a diagram illustrating the trend over time of some signals and physical quantities during a certain time interval of use of an electric vehicle, in particular: the medium-term computed power limit, the instantaneous computed power limit, the real power limit modified according to the method of the present invention, the power provided by the battery, the pressure on the accelerator pedal, and the voltage of the battery cell;

[0027] - Figure 4 is a block diagram illustrating the operation of a logic for defining the real power limits of the battery according to one or more embodiments;

[0028] - Figure 5 is a block diagram illustrating the operation of a logic for defining the real power limits of the battery according to one or more embodiments, comprising the algorithm of Figure 2 and the logic of Figure 4;

[0029] - Figure 6 is a diagram illustrating the trend over time of the same signals and physical quantities of Figure 3, during a certain time interval of use of an electric vehicle in which the driver’s request for maximum power is briefly interrupted; - Figure 7 is a block diagram illustrating the operation of a logic for defining the real power limits of the battery according to one or more embodiments, comprising the logic of Figure 4 and a reset logic; and

[0030] - Figure 8 is a diagram illustrating the trend over time of the same signals and physical quantities of Figures 3 and 6, during a certain time interval of use of an electric vehicle in which the driver’s request for maximum power is maintained for a short time interval.

[0031] In the figures annexed hereto, corresponding parts or elements are indicated with the same reference numbers.

[0032] As mentioned, one or more embodiments relate to an improved method for determining the state of power (SOP) of a traction battery of an electric vehicle, i.e. , for determining its real power limits. In particular, the method is based on (i) an online computation SOP algorithm that uses an equivalent circuit model (ECM) to compute the predicted power limits and optionally on the use of current limitation maps drawn up by the battery cell supplier, and on (ii) a logic that determines the real power limit as a function of the current power request by the driver and the computed predicted power limits.

[0033] A possible equivalent circuit model of a high-voltage battery cell B (mounted on board a vehicle V) is shown in the circuit diagram of Figure 1 , and a power limit computation algorithm 2 using such a model is shown in Figure 2. The ECM model comprises a voltage generator G providing an open circuit voltage VOC(SOC, T), a series resistance Rs(SOC, T) through which the current l(t) provided by the cell flows (by convention, the current l(t) is positive during the discharge phase of the cell), an RC circuit comprising a resistance R-i(SOC, T) and a capacitance Ci(SOC, T) coupled together in parallel and overall in series with the resistance Rs, where the RC circuit is subjected to a voltage drop Vi(t) and provides an output voltage V(t) between the second terminal of the RC circuit and the reference terminal (ground). As can be seen from the terminology adopted, the values of VOC, Rs, Ri and Ci are dependent on the state of charge SOC and the temperature T (these dependencies are not reported in Figure 1 for simplicity). Given the ECM model illustrated in Figure 1 , the prediction problem that the SOP power limit computation algorithm intends to solve is to determine a constant power value PTsop(output value from the diagram in Figure 2) that cell B can continuously provide over a certain time interval T such that the cell output voltage V(t) changes from the present value V(to) at time to to a certain limit value Vum, i.e. , the power value PTsuch that the following equality (1 ) is verified:

[0034] In the remainder of this disclosure, reference will be made mainly to the computation of the battery power limits in the case in which the battery provides current for vehicle traction, as this application case appears to be of greater interest in terms of improving vehicle performance. On the other hand, the same method can be used to compute the power limits during battery regeneration.

[0035] In one or more embodiments, this prediction problem (computation of PTsop) is solved in two successive phases.

[0036] In the first part of the first phase of algorithm 2 (block 20 of the diagram of Figure 2), a constant current IPCMis computed, which solves the same prediction problem defined by equality (1 ). The solution of this first step is obtained by solving the equality (1 ) with the ECM model of Figure 1 , and it can be shown that the current value IPCM(i.e., the current that, if provided continuously for the entire time interval T, brings the voltage from V(to) to Vum) is given by the following equation (2): where Cceii is the cell capacity expressed in Ah, and the initial voltage Vi(to) on Ri and Ci is given by the following equation (3):

[0037] It will be noted that the dependence of the values Rs, Ri and Ci on the state of charge SOC and on the temperature T, derivable from the maps drawn up by the battery cell supplier, is omitted for simplicity in the equations presented here.

[0038] Once the current value IPCMhas been computed using equations (1 ), (2) and (3) with the on-line SOP algorithm, in a second part of the first phase of algorithm 2 the current limit values indicated in the maps drawn up by the battery cell supplier can optionally be considered (block 22 of the diagram in Figure 2), in order to select as the real limit value of the discharge current ITthe most conservative value between the computed value IPCMand the value / TMapgiven by the map, as a function of SOC and T. Therefore, once the limit value / TMaphas been determined as a function of the supplier’s maps, the lower value between IPCMand / TMapis selected as the real limit value IT(block 24 of the diagram in Figure 2). In this way, the constraints on the battery current and voltage values imposed by the manufacturer are respected, which must not be exceeded as this could activate the battery protection mechanisms and open the battery contactors.

[0039] In the second phase of algorithm 2 (blocks 26 and 28 of the diagram in Figure 2), a constant power value PTis computed such that the energy provided by the cell in the time interval T is equal to the energy obtained by driving the cell with the current ITfor the time period T, i.e. the following equation (4) is solved: (4)

[0040] The term is substantially equal to the average output voltage predicted in the time interval T in the case in which the current ITis applied, and the average voltage can be approximated by the following formula (5): where P(t0) = V(t0)>RS(IT ~ f(t0)) is the predicted value of the cell output voltage at the beginning of the time interval T (i.e., at time to), if the current ITis already applied. Therefore, P(t0+ T) = VLim. Once the average output voltage is computed with equation (5) in block 26 of Figure 2, the power limit PTsopis determined by multiplying the average voltage by the current IT(block 28 of Figure 2).

[0041] Therefore, Figure 2 illustrates an algorithm for computing the power limit PTsopthat relies on the ECM model of Figure 1 and equations (1 ), (2) and (3) to compute a first current limit value IPCM, compares this value with the cell manufacturer-imposed limit value / TMapand selects the smaller of the two for conservative reasons, and determines the power limit PTsopby multiplying the average voltage given by equation (5) and the current value If

[0042] Algorithm 2 of Figure 2 can be used to compute the power limit PTsop over a time interval T of arbitrary duration. An innovative aspect of the present invention is to compute an additional power limit value compared to the conventional ones, i.e. , not only the short-, medium- and long-term limit values (e.g., for Tshort= 2 s, tmed= 10 s, zlong= 30 s), but also a limit value here referred to as “instantaneous” limit value, which is computed over a time interval Tinstwhose duration is substantially comparable with the computation time of the parameter itself, or rather with the refresh time of algorithm 2. For example, considering that in some applications the execution of the algorithm for computing the power limits requires about 10 ms, the instantaneous limit value can be computed over a time interval Tinstof duration equal to 20 ms, 50 ms or 100 ms. Since the algorithm in Figure 2 also uses the value maps drawn up by the supplier of the battery supplier B, the computation of the power limit value PTsopfor Tinstwill be based on the value maps drawn up by the supplier of the battery supplier B for this new time interval as well. It should also be noted that equation (2) reduces to the following equation (6) when the time interval T tends to zero: and therefore the power limit value PTsopfor Tinstis approximated by the following equation (7): (7)

[0043] Since the instantaneous power limit value is computed in order to provide the driver with the possibility of fully exploiting the power that can be provided by the battery, without perceiving drops in power and without “saving” energy, a further part of the method according to the invention is implemented by a logic that computes, instant by instant, a real power limit that connects the “conventional” limits (short, medium and long term) to the instantaneous one, computing a real limit value that allows to increase, compared to conventional systems, the power provided for a certain limited period of time.

[0044] As an example, to better understand how this logic works, which here is called “PLP logic” (“Power Limit Performance Logic”), reference can be made to Figure 3, which illustrates the trend over time of some signals or physical quantities, in the case of a maneuver in which the driver of the vehicle presses the accelerator fully down for a rather long period of time (for example, to perform a launch or overtaking maneuver) when the state of charge SOC of the battery is low, and therefore the power provided to the engine is limited by the state of the battery. In particular, the signals or physical quantities in Figure 3 are the following: instantaneous power limit Ptnst (eg-. computed with algorithm 2 for Tinst= 100 ms), medium-term power limit P^epd(e.g., computed with algorithm 2 for zmed= 10 s), real power limit P^Lepddetermined by the PLP logic as a function of the limits P$°spand P^ed’ power Peafactually provided by the high-voltage battery, percentage of accelerator pedal pressure GP, cell voltage Vceii, and minimum allowed cell voltage Vmin. In this example, reference is made to the medium-term power limit Pmed, but a similar operation can also be obtained for the short- and long-term power limits.

[0045] Essentially, the PLP logic manages the power limits computed with algorithm 2 of Figure 2 (i.e. , the instantaneous limit P$°spand at least one of the short-, medium- and long-term limits Pf°prt, Pmedar|d Pio°nPg)toperform the maneuver requested by the driver at constant power for the entire duration of the time interval, and subsequently at decreasing power, which progressively decreases following the instantaneous power limit P$°sp. In the following, we will talk about “predicted power limit” to indicate each of the power limits computed by the SOP algorithm of Figure 2, collectively indicated as PTsop(with T = inst, short, med, long) and “real power limit” to indicate the power limit determined by the PLP logic, indicated as PPLP. Essentially, as shown in Figure 3:

[0046] - before time ti, the driver fully presses the accelerator pedal GP; as a result, the power Peafprovided by the battery increases until it reaches one of the predicted power limits (in this example, the medium-term limit Pmea indicated by the dotted line); until time ti, i.e. until the power actually provided reaches the predicted power limit, the real power limit P^Lepdcoincides with the predicted power limit, i.e. the PLP logic maintains P^Lepd= pSOP.

[0047] - at time ti, when the power Peafreaches the predicted limit P^edand the driver continues to keep the accelerator pedal fully pressed, a conventional algorithm would limit the real power to the value P^epdwhich, however, being recomputed cyclically, gradually decreases between ti and t2 because energy continues to be drawn from the battery; instead, the PLP logic of the present invention “fixes” the real limit P^Lepdto the value of the limit P™pdat the time ti, and therefore allows the vehicle to draw a constant power Peaffrom the battery throughout the time interval between ti and t2, such constant power being higher than the limit power P^edand in particular being equal to the limit power P^edcomputed at the time ti;

[0048] - by maintaining P^Lepd= Pmed(tl') starting from the time ti, the real power limit P^Lepd(and therefore the provided power Peaf) intersects the predicted instantaneous power limit P$°spafter a time interval approximately equal to the duration of the “average” interval, for example 10 s, as indicated at the time tz at this moment, the cell voltage Vceii reaches the limit allowed by the manufacturer (in this case the minimum allowed value, since it is a battery discharge phase);

[0049] - if the driver continues to keep the accelerator pedal fully pressed even beyond the time t2, the PLP logic sets the real power limit P^Lepdso that it follows the decreasing trend of the predicted instantaneous power limit Ptnst (which represents the upper limit of the real limit value), thus limiting the power Peafprovided by the battery; in this condition, the cell voltage Vceii is maintained at the minimum value Vmin and there is a gradual reduction in the current provided by the cell, thus ensuring the battery operates within the safe operating area;

[0050] - at time ts, when the driver releases the accelerator pedal and the power demand is reduced, the provided power Peafdecreases sharply and the real power limit P^Lepdcan be reset, i.e. , the PLP logic sets the real power limit P^ed so that it follows again the trend of the medium -term predicted power limit P^epd(as before time ti).

[0051] Therefore, in summary, the idea behind the present invention is to compute, in addition to the “conventional” SOP limits, a fourth “instantaneous” SOP limit and allow the battery to provide more power than the conventional limit, while respecting the instantaneous limit.

[0052] In various embodiments, the PLP logic may further implement a hysteresis to compare the value of the power Peafprovided by the battery and the predicted power limit PTsop(refer again to times ti and ts in Figure 3), this being the short, medium or long term limit. Figure 4 is a block diagram illustrating the operation of the PLP logic, in accordance with what was described with reference to the diagram in Figure 3. In particular, PLP logic 4 receives as input the values of the three predicted power limits Ps °ort’pmedar|dPion9computed as shown in Figure 2 and collectively indicated as PTsopin Figure 4. Furthermore, PLP logic 4 receives as input the value of the predicted instantaneous power limit P$°sp(also computed as shown in Figure 2), the measured value of the power provided by the battery Peaf, and a signal TAU that can assume one of three values (e.g., TAU = {short, med, long}) to indicate that the vehicle is currently using the short-term, mediumterm or long-term limit. In particular, the choice of one of the three “conventional” limits (i.e., the value of the signal TAU) usually depends on the driving mode currently set on the vehicle (including the “turtle mode”, i.e., the mode that is automatically set when the traction battery charge level is very low) or on driving modes selectable by the driver, such as the launch mode (which allows acceleration from a standstill by making the most of the potential of the propulsion system) or the sub-modes of the “race” mode available on high-performance vehicles. In an initial state 41 , the PLP logic is inactive and the real power limit PPLPis set to follow the variations of one of the three predicted power limits PTsop(selected depending on the value of the signal TAU). When the power provided by the battery P6afreaches the predicted power limit PTsopminus a certain threshold Thrs (see again Figure 3) at some time ti, i.e. when \PBat| > \PTsop- Thrs\, condition C1 is verified and the PLP logic 4 goes into an active state 42 where the real power limit PPLPis set to the minimum value between the value of the predicted power limit PTsopat time ti (denoted by P = PTsop@t1) and the present value of the predicted instantaneous power limit P$°sp, i.e. PPLP= min{|P|, \Pt °sp|}. When the power provided by the battery Peafgoes below the predicted power limit PTsopminus the threshold Thrs and minus a hysteresis margin Hys (see again Figure 3) at a certain time ts, i.e. when condition C2 is verified and the PLP logic 4 returns to state 41 where the real power limit PPLPfollows the variations of the predicted power limit PTsop.

[0053] As illustrated in the block diagram of Figure 5, a system 5 for computing the real power limits Pp£prt, 'stherefore composed of a first part 2 as illustrated in Figure 2, which computes the four predicted power limits Ps °ort’ Pmea, Pio°ngar|d Ptnst using equations (1 ) to (5), and a second part 4 as illustrated in Figure 4, which compares the power Peafwith at least one of the three “conventional” power limits Ps °ort’ Pmea,pio°nPgar|d consequently determines at least one of the three real power limits PPhort’ Pmeaand P^ng using the instantaneous power limit P$°spas an upper limit that cannot be exceeded, in order to maximize the vehicle performance.

[0054] In some preferred embodiments, the PLP logic illustrated in Figure 4 includes an additional feature to manage any brief interruptions in power demand due to external disturbances (e.g., a pothole in the road surface) and / or the driver’s driving style. This feature makes the PLP logic more robust against such possible short interruptions of power demand (i.e. , short releases of the accelerator pedal followed by a new full press), as illustrated in the timing diagram of Figure 6, which illustrates the same signals and physical quantities as in Figure 3, in case in the interval between ti and t2 (i.e., when the PLP logic is “pushing” the real limit PPLPabove the respective predicted limit PTsopwhile respecting the instantaneous limit P$°sp) the driver briefly releases the accelerator pedal GP (see the time period between ti’ and ti” in Figure 6) and therefore the real provided power Peafgoes below the predicted power limit PTsopminus the threshold Thrs and minus the hysteresis margin Hys. If the logic described above was applied, the real limit PPLPwould be reset to the value of the predicted limit PTsopat the time ti’, when the power Peafgoes below PTsop- Thrs - Hys (condition C2 of Figure 3). According to the additional feature described here, however, the reset of the real limit PPLPis not performed immediately when condition C2 occurs, but after a specific time interval Tcai, which can be calibrated if necessary, so that if the reduction in the power request from the battery Peaflasts less than this interval Tcai, the value of the real power limit PPLPis not reset but kept constant, and the vehicle’s response is quicker in the event of a subsequent pressure on the accelerator pedal. The block diagram of the PLP logic 4 including this preferred feature is shown in Figure 7, which is substantially similar to Figure 4 but shows that, when condition C2 is satisfied (i.e., when \PBat the PLP logic does not immediately return to state 41 , but goes into a reset state 71 where essentially a timer is activated, and then goes to state 41 if the timer reaches the value Tcai (condition C3), otherwise it returns to state 42. In this second case (i.e., if the PLP logic returns from state 71 to state 42 without going through state 41 , and therefore without resetting the real power limit PPLPto follow the changes in the predicted power limit PTsop), the time interval t2-ti may also be longer than the value of T (e.g., longer than 10 seconds if the medium-term limit is being used), but this does not compromise the functioning of the battery because it is the constant respect of the instantaneous limit P$°spthat guarantees remaining in a region of correct battery usage.

[0055] The timing diagram in Figure 8 illustrates the same signals and physical quantities as in Figure 3 and Figure 6, in the case where the driver’s maximum power request (i.e. , full-throttle pressure on the accelerator pedal) is exhausted before the provided power PBat intersects the predicted instantaneous power limit In this case, starting from the time f / when the power Peafgoes below the predicted power limit PTsop(minus the threshold Thrs and the hysteresis margin Hys), the real power limit PPLPis maintained at the previous value (i.e. at the value P = PTsop@t1), and at the end of the time interval Tcai it is reset, i.e. it is re-attached to the variations of the predicted power limit PTsop.

[0056] Of course, notwithstanding the principle of the invention, the construction details and the embodiments may vary widely with respect to what is described and illustrated purely by way of example, without thereby departing from the scope of the present invention, as defined in the attached claims.

Claims

CLAIMS1. A method (5) of determining the state of power of a high- voltage traction battery (B) of an electric vehicle (V), the method comprising: i) sensing the state of charge (SOC) of the battery, the temperature ( T) of the battery, the voltage (V) provided by the battery and the current ( / ) provided by the battery;(ii) computing (2), as a function of said state of charge (SOC), said temperature (T), said provided voltage (V) and said provided current ( / ), at least one first predicted power limitsaidatleastonefirst predicted power limit being computed for at least one first time interval having a duration higher than 1 second, wherein said at least one first predicted power limit is equal to the power that, if continuously provided or absorbed by the battery (B) over said at least one first time interval, brings the voltage ( Vceii) of the battery respectively to a minimum ( Vmin) or maximum limit value; iii) computing (2), as a function of said state of charge (SOC), said temperature (T), said provided voltage ( V) and said provided current ( / ), a second predicted power limit (P^f ), said second predicted power limit being computed for a second time interval having a duration lower than 1 second, wherein said second predicted power limit is equal to the power that, if continuously provided or absorbed by the battery (B) over said second time interval, brings the voltage ( Vceii) of the battery respectively to said minimum ( Vmin) or maximum limit value; iv) sensing, or computing as a function of said provided voltage (V) and said provided current ( / ), the power fPeaf) provided or absorbed instantaneously by the battery (B); v) in response to (C2) the power (Peaf) provided or absorbed instantaneously by the battery (B) being lower than said at least one first predicted power limit (Ps °orf Pmea, Pio°nPg)’ setting (41 ) the value of at least one respective real power limit (Pshort, Pmedthe present value of said at least one first predicted power limit (Ps °ort’ Pmea , Ptm’ g )’ar|d vi) in response to (C1) the power (Peaf) provided or absorbed instantaneously by the battery (B) reaching said at least one first predicted power limit (Ps °orf Pmea , Pio°nPg)’ setting (42) the value of said at least onerespective real power limit (Pshort’ Pmed* PPong)tothe minimum value selected between: (a) the instantaneous value of said at least one first predicted power limit (Ps °orf Pmed, Pio°nPg)atthe time (ti) when the power (pea?) provided or absorbed instantaneously by the battery (B) reaches said at least one first predicted power limit (Ps °orf Pmed, Pio°nPg)’ar|d (b) the present value of said second predicted power limit2. The method of claim 1 , wherein the step of computing (2) at least one first predicted power limit (Ps °orf Pmea , Pio°ng) comprises:- determining (20), using equations of an equivalent circuit model (G, Rs, Ri , Ci) of said battery (B), at least one first model-derived current limit value (I^CM), wherein said at least one first model-derived current limit value is equal to the current that, if continuously provided or absorbed by the battery (B) over the at least one first time interval, brings the voltage ( Vceii) of the battery respectively to said minimum ( Vmin) or maximum limit value;- computing (26, 28) the at least one first predicted power limit (Pshort, Pmed ’ Pi °ng)asthe constant power value which would cause the energy provided by said battery (B) over the at least one first time interval being equal to the energy obtained by driving said battery (B) with the respective at least one first model-derived current limit value (I^CM) over the respective at least one first time interval; and wherein the step of computing (2) a second predicted power limit (Pt °sp) comprises:- determining (20), using said equations of said equivalent circuit model (G, Rs, Ri, Ci) of said battery (B), a second model-derived current limit value (I^CM), wherein said second model-derived current limit value is equal to the current that, if continuously provided or absorbed by the battery (B) over the second time interval, brings the voltage ( Vceii) of the battery respectively to said minimum ( Vmin) or maximum limit value;- computing (26, 28) the second predicted power limit (P?°£ ) as the constant power value which would cause the energy provided by said battery (B) over the second time interval being equal to the energy obtained by driving said battery (B) with the second model-derived current limit value (I^CM) over the second time interval.

3. The method of claim 2, wherein the step of computing (2) at least one first predicted power limit (Ps °orf Pmed > Ptm’ g ) comprises:- determining (22), using a map of the safe operating area of said battery (B), at least one first map-derived current limit valuewherein said at least one first map-derived current limit value is equal to the maximum current that can be continuously provided or absorbed by the battery (B) over the at least one first time interval as a function of the state of charge (SOC) and the temperature (T) of the battery;- determining (24) at least one first real current limit value ( / T) as the minimum value between said at least one first model-derived current limit value (I^CM) and said at least one first map-derived current limit value- computing (26, 28) the at least one first predicted power limit (Pshort, Pmed’ Pi °ng)asthe constant power value which would cause the energy provided by said battery (B) over the at least time interval being equal to the energy obtained by driving said battery (B) with the respective at least one first real current limit value ( / T) over the respective at least one first time interval; and wherein the step of computing (2) a second predicted power limit (Pi °sPt ) comprises:- determining (22), using a map of the safe operating area of said battery (B), a second map-derived current limit valuewherein said second map-derived current limit value is equal to the maximum current that can be continuously provided or absorbed by the battery (B) over the second time interval as a function of the state of charge (SOC) and the temperature (7} of the battery;- determining (24) a second real current limit value ( / T) as the minimum value between said second model-derived current limit value (I^CM) and said second map-derived current limit value- computing (26, 28) the second predicted power limit (P?°£ ) as the constant power value which would cause the energy provided by said battery (B) over the second time interval being equal to the energy obtained by driving said battery (B) with the second real current limit value ( / T) over the second time interval.

4. The method of any of the previous claim, wherein, in step (vi), the value of said at least one respective real power limit (Pshort> Pmed, PPoLng) is set (42) to the minimum value selected between (a) and (b) in response to the power (Peaf) provided or absorbed instantaneously by the battery (B)reaching said at least one first predicted power limit (Ps°ort’ Pmed,piong) minus a certain threshold margin (Thrs).

5. The method of any of the previous claim, wherein, in step (v), the value of said at least one respective real power limit (P^ort, Pmed, PPong) is set (41 ) to the present value of said at least one first predicted power limit (Pshort, Pmed’ Pi °ng) 'nresponse to the power (Peaf) provided or absorbed instantaneously by the battery (B) being lower than said at least one first predicted power limit (Ps°orf Pmea, Pio°nPg) minus a certain threshold margin (Thrs) and / or minus a certain hysteresis margin (Hys).

6. The method of any of the of the previous claims, wherein, in step (v), the value of the at least one respective real power limit (P^ort, Pmed’ PPoLng) 's set(41 ) to the present value of said at least one first predicted power limit (Ps°orf Pmed> Pio°nPg)wi^acertain delay (71 , Tcai) with respect to the time (ti’) when the power (PBat) provided or absorbed instantaneously by the battery (B) goes below said at least one first predicted power limit / pSOP pSOP pSOP \\rshort’rmed ’rlong )-7. A system for determining the state of power of a high-voltage traction battery (B) of an electric vehicle (V), the system comprising:- a plurality of sensors configured to sense the state of charge (SOC) of the battery, the temperature (7) of the battery, the voltage (V) provided by the battery and the current (I) provided by the battery (B), and- an electronic processing unit configured to receive data from said plurality of sensors, and further configured to carry out the method of any of the of the previous claims.

8. An electric vehicle (V) comprising a high-voltage traction battery (B) and a system according to claim 7.

9. A computer program product loadable in the memory of an electronic processing unit and comprising instructions which, when the computer program product is executed by the electronic processing unit, cause the electronic processing unit to carry out the method of any of claims 1 to 6.

Images