Method for estimating a temperature of a rotor magnet of a motor

Abstract

A method for estimating a temperature of a magnet of a rotor, the rotor comprising a channel circulating a heat transfer fluid in contact with or near the magnet, between an inlet section of the channel and an outlet section of the channel, a temperature sensor being arranged in the channel, in particular in the vicinity of the outlet section of the channel, the method comprising: • a partial or total reduction, for a given duration, of a flow rate of the heat transfer fluid circulating in the channel, • a measurement, by the temperature sensor, of a temperature of a given heat transfer fluid which has stayed in the channel throughout the given duration, and • a determination of an estimated temperature of the magnet as being a sum of the measured temperature of the given heat transfer fluid and of a zero or non-zero constant temperature difference.

Description

Description Title of the invention: Method for estimating the temperature of a rotor magnet of a motor.

[0001] The invention relates to a method for estimating a temperature of a rotor magnet of a motor. The invention relates to a method for controlling an estimation of a temperature of a rotor magnet of a motor. The invention also relates to a device for estimating a temperature of a rotor magnet of a motor. The invention further relates to a motor vehicle equipped with such an estimation device. The invention further relates to a method for controlling a powertrain of such a motor vehicle. The invention also relates to a computer program implementing the estimation method or the control method mentioned. The invention finally relates to a recording medium on which such a program is recorded.

[0002] The invention relates to the field of rotor magnets of a motor, in particular the field of permanent magnets of a rotor of an electric vehicle motor, in particular of a motor vehicle or any other land or air vehicle.

[0003] It is essential to monitor the temperature of the permanent magnets in a rotor because too high a temperature rise is likely to irreversibly reduce the remanent induction of the magnet, thereby reducing the torque generated by the electric motor. To prevent the temperature of the magnets from rising too high, it is generally possible to arrange a circulation of heat transfer fluid near the magnets.

[0004] However, there is a strategy for protecting the powertrain, particularly when a control fault occurs in an electronic power converter, which consists of creating a short circuit between the stator windings. Such a strategy then generates a circulation of significant currents in the stator windings, creating significant magnetic induction fields facing the magnets, with the risk of demagnetizing them. Indeed, if the temperature of the magnets is high, then even a weak magnetic field will be enough to demagnetize the rotor magnets, thus permanently degrading the performance of the electric motor.

[0005] In other words, when the temperature of the rotor magnets is high, the implementation of a short circuit between the stator windings has irreversible consequences on the performance of the electric motor.

[0006] It is therefore necessary to know the temperature of the magnets as precisely as possible in order to take it into account in the powertrain management strategy.

[0007] There are solutions to measure or estimate the temperature of a magnet rotor. However, these solutions have drawbacks.

[0008] The aim of the invention is to provide a device and a method for estimating a temperature of a rotor magnet of a motor which overcomes the drawbacks mentioned above and improves the devices and methods for estimating a temperature of a rotor magnet of a motor known from the prior art. In particular, the invention makes it possible to produce a device and a method which are simple and reliable.

[0009] To this end, the invention relates to a method for estimating a temperature of a magnet of a rotor, the rotor comprising a channel implementing a circulation of a heat transfer fluid in contact with or near the magnet, an inlet section of the channel and an outlet section of the channel, a temperature sensor being arranged in the channel, in particular near the outlet section of the channel, the method comprising: • a step of partial or total reduction, for a given duration, of a flow rate of the heat transfer fluid circulating in the channel, • a measurement step by the temperature sensor of a temperature of a given heat transfer fluid having remained in the channel for the entire given duration, and • a step of determining an estimated temperature of the magnet as being a sum between the measured temperature of the given heat transfer fluid and a constant zero or non-zero temperature difference.

[0010] In one embodiment, the estimation method comprises a step of determining the given duration and the constant deviation comprising an implementation of a thermal model of an evolution of a temperature of the heat transfer fluid generated solely by a transfer of a thermal power from the magnet to the heat transfer fluid.

[0011] In one embodiment, the thermal model is: - a first model, built on a hypothesis of a constancy of a modeled temperature of the magnet over the given duration and of a homogeneity of a modeled temperature of the magnet over an entire length of the magnet, the length of the magnet being measured according to the direction of the channel, or - a second model, built on an assumption of homogeneity, at each instant of the given duration, of a modeled temperature of the magnet over its entire length, the modeled temperature of the magnet being variable over time. In addition, the first and second models each comprise a first variant in which, during the given duration, the flow rate of the heat transfer fluid is non-zero, and a second variant in which, during the given duration, the flow rate of the heat transfer fluid is zero.

[0012] In one embodiment, the thermal model provides a time evolution of a modeled temperature of a magnet and a time evolution of a modeled temperature of a heat transfer fluid circulating in contact with the magnet. Furthermore, the given duration is determined so that: - an increase in the modeled temperature of the magnet during the given duration is less than a first threshold, in particular the first threshold being equal to 6 degrees, or even 4 degrees and - at a simulation time located from the given duration, a difference between a modeled temperature of the magnet and a modeled temperature of the heat transfer fluid is substantially constant and less than a second threshold, for example less than 3 degrees. Furthermore, the constant deviation is equal to a difference, calculated at a simulation instant located from the given duration, between a modeled temperature of the magnet and a modeled temperature of the heat transfer fluid.

[0013] In one embodiment, the measuring step comprises a sub-step of resuming circulation of the heat transfer fluid at an intermediate flow rate, the intermediate flow rate being less than or equal to an initial flow rate of circulation of the heat transfer fluid before flow rate reduction.

[0014] The invention further relates to a device for estimating a temperature of a magnet of a rotor, a cooling circuit of the magnet comprising a set of conduits implementing a circulation of a cooling fluid in contact with the magnet, the device comprising a means for controlling a flow rate of a solenoid valve arranged on a conduit of the set of conduits, the conduit being located upstream of the magnet relative to a direction of circulation of heat transfer fluid near or in contact with the magnet, and the device comprising hardware and / or software elements implementing the method according to the invention, in particular hardware and / or software elements designed to implement the method according to the invention.

[0015] The invention further relates to a method for controlling a motor vehicle powertrain, comprising: • a detection step, depending on the conditions of use of the motor vehicle, of a need to estimate, at a given instant, a temperature of the magnet, or of a need to estimate, at a given frequency, a temperature of the magnet, then • a step of implementing, at a given time or at a given frequency, a method for estimating a current temperature of a magnet according to the invention, then • if the current temperature of the magnet is higher than a temperature threshold, a step of reduction of an available engine torque of the powertrain.

[0016] In one embodiment, the usage conditions include an engine torque controlled by a driver of the motor vehicle at the given time.

[0017] The invention further relates to a device for controlling a powertrain. of a motor vehicle, the device comprising hardware and / or software elements implementing the method according to the invention, in particular hardware and / or software elements designed to implement the method according to the invention.

[0018] The invention relates to a motor vehicle equipped with a control device according to the invention, or with an estimation device according to the invention.

[0019] The attached drawing represents, by way of example, an embodiment of a device for estimating a temperature of a motor rotor magnet according to the invention.

[0020] [Fig.l] is a first representation of a motor vehicle equipped with an embodiment of a device for estimating a temperature of a motor rotor magnet according to the invention.

[0021] [Fig.2] is a first representation of an embodiment of a device for estimating a temperature of a motor rotor magnet according to the invention.

[0022] [Fig.3] is a second representation of an embodiment of a device for estimating a temperature of a motor rotor magnet according to the invention.

[0023] [Fig.4] is a third representation of an embodiment of a device for estimating a temperature of a motor rotor magnet according to the invention.

[0024] [Fig.5] is a flowchart of an estimation method according to the invention.

[0025] [Fig.6] is a first representation of a first modeling of a magnet and a heat transfer fluid circulating near or in contact with the magnet.

[0026] [Fig.7] is a second representation of the first modeling of a magnet and a heat transfer fluid circulating near or in contact with the magnet.

[0027] [Fig.8] is a third representation of the first modeling of a magnet and a heat transfer fluid circulating near or in contact with the magnet.

[0028] [Fig.9] is a graph of the time evolution of the temperatures of a magnet and a heat transfer fluid circulating near or in contact with the magnet, the temperatures being modeled according to the first model.

[0029] [Fig.10] is a detailed view of a graph of the time evolution of the temperatures of a magnet and a heat transfer fluid circulating near or in contact with the magnet, the temperatures being modeled according to the first model.

[0030] [Fig.l 1] is a graph of the time evolution of a distance traveled by the heat transfer fluid, modeled according to the first model.

[0031] [Fig.12] is a graph of the evolution of a heat transfer fluid temperature, modeled according to the first model, as a function of the distance traveled by the fluid.

[0032] [Fig.13] is a graph of the time evolution of the temperatures of a magnet and a heat transfer fluid circulating near or in contact with the magnet, the temperatures being modeled according to the second model.

[0033] [Fig.14] is a graph of the evolution of the temperatures of the magnet and the heat transfer fluid, modeled according to the second model, as a function of the distance traveled by the heat transfer fluid.

[0034] [Fig.15] is a detailed view of a graph of the evolution of the temperatures of the magnet and the heat transfer fluid, modeled according to the second model, as a function of the distance traveled by the heat transfer fluid.

[0035] [Fig.16] is a graph of the time evolution of a distance traveled by the heat transfer fluid, modeled according to the second model.

[0036] [Fig.17] is a first representation of a second model of a magnet and a heat transfer fluid circulating near or in contact with the heat transfer fluid.

[0037] [Fig.18] is a second representation of the second modeling of a magnet and a heat transfer fluid circulating near or in contact with the heat transfer fluid.

[0038] [Fig.19] is a graph of the time evolution of the temperatures of a magnet and a heat transfer fluid circulating near or in contact with the magnet, the temperatures being modeled according to a third model.

[0039] [Fig.20] is a first representation of a third model of a magnet and a heat transfer fluid circulating near or in contact with the heat transfer fluid.

[0040] [Fig.21] is a second representation of the third modeling of a magnet and a heat transfer fluid circulating near or in contact with the heat transfer fluid.

[0041] [Fig.22] is a graph of the time evolution of the temperatures of a magnet and of each of the layers of a heat transfer fluid circulating near or in contact with the magnet, the temperatures being modeled according to a fourth model.

[0042] [Fig.23] is a first representation of a fourth model of a magnet and a heat transfer fluid circulating near or in contact with the heat transfer fluid.

[0043] [Fig.24] is a second representation of the fourth modeling of a magnet and a heat transfer fluid circulating near or in contact with the heat transfer fluid.

[0044] [Fig.25] is a graph of the time evolution of the temperatures of a magnet and a heat transfer fluid circulating near or in contact with the magnet, the temperatures being modeled according to a fourth model.

[0045] [Fig.26] is a graph of the time evolution of a difference between the temperatures of a magnet and a heat transfer fluid circulating nearby in contact with the magnet, the temperatures being modeled according to the fourth model.

[0046] [Fig.27] is a time graph of the control of a solenoid valve controlling a flow of a heat transfer fluid.

[0047] [Fig.28] is a flowchart of a control method according to the invention.

[0048] [Fig.29] is an abacus for calculating a Nusselt number in the case of a laminar fluid flow.

[0049] An example of a motor vehicle 100 equipped with an embodiment of a device 10 for estimating a temperature of a motor rotor magnet is described below with reference to FIGS. 1 to 29.

[0050] The motor vehicle 100 may be a vehicle of any type, for example a passenger vehicle or a utility vehicle or a public transport vehicle. The motor vehicle 100 may be an all-electric vehicle, or a hybrid vehicle.

[0051] More generally, the devices described in this document could equip electric or hybrid motors of all types, for example aircraft engines, agricultural machinery, boats, etc.

[0052] The motor vehicle 100 comprises a powertrain 10 comprising an electric motor 1. The electric motor 1 comprises a rotor 11 and a stator 12. Magnets 111 are arranged in the rotor, the permanent induction of the magnets 111 serving to generate a motor torque for the movement of the motor vehicle 100.

[0053] The rotor 11 described in this document is a magnet rotor. In an embodiment not described, the rotor could be a twisted magnet rotor, i.e. a rotor consisting of several sections angularly offset from each other, so that the magnets of the different sections are not aligned. Alternatively, the rotor 11 could be a wound rotor.

[0054] The powertrain further comprises a device 2 for controlling the powertrain according to the invention comprising a device 21 for estimating a temperature of the magnets 111 according to the invention.

[0055] In an advantageous embodiment, the powertrain control device 2 further comprises its own dynamic model 22 for estimating a temperature of a rotor magnet.

[0056] An embodiment of a dynamic model 22 is described briefly below. The principle of dynamic estimation of the temperature of the magnets of a machine, in particular of an electric motor, is based on a real-time calculation of the temperature of the magnets by solving equations relating to the thermal exchanges affecting the magnets.

[0057] Eddy current thermal losses occur in the magnet. These thermal losses are evaluated by solving Maxwell's equations of F electromagnetism, notably using the finite element method. Thermal loss calculations are performed for all operating points of the machine, each operating point being defined by a torque supplied by the machine and a rotational speed of the machine. The thermal loss values ​​calculated for each operating point of the machine are recorded in a memory, for example in the form of a matrix, in order to be used by the dynamic model 22. The thermal loss values ​​recorded in the matrix are fixed, that is to say they do not change during the operation of the machine.

[0058] In one embodiment, the dynamic model 22 is constructed on the assumption that the thermal power transferred by the magnet to the iron surrounding it is negligible (the magnet being inserted and glued into a housing made of iron). The thermal loss values ​​recorded in the matrix make it possible to estimate the temperature of the magnet in real time, as a function of the current operating point of the machine. This estimation can be carried out at a relatively high frequency, for example at a period of around ten seconds.

[0059] However, the accuracy of the dynamic model 22 may be insufficient when applied to machines whose performance is essential. In particular, the dynamic model 22 does not allow the temperature of the magnets to be known precisely. Indeed, there may be a difference of one or two tens of degrees between the estimate provided by the dynamic model 22 and the actual temperature of the magnet. However, for optimum operation of certain machines, in particular powertrains of electric vehicles, it is necessary to know precisely the temperature of the magnets in certain driving phases inducing a rise in temperature of the magnets. The estimation device 21 according to the invention can then advantageously be used to verify the temperature estimates provided by the dynamic model 22.Indeed, as long as the temperature of the magnet provided by the dynamic model 22 remains in a moderate temperature zone, the accuracy of the dynamic model 22 is sufficient. When the temperature of the magnet provided by the dynamic model 22 is in a high temperature zone, then the data from the estimation device 21 advantageously makes it possible to reliably measure the temperature of the magnet, in order to optimize the performance of the machine without risking damaging the magnets, and without it being necessary to apply significant margins of protection of the magnets.

[0060] Figures 2 to 4 schematically represent an embodiment of a cooling circuit 3 for the rotor magnets. The circuit 3 comprises a set of conduits 30 implementing a circulation loop for a cooling fluid, for example a dielectric fluid, circulating in contact with the magnets 111 of the rotor 11 for their cooling. A pump 31 circulates the cooling fluid in the set of conduits 30. For its cooling, the fluid passes through an exchanger 32. A solenoid valve 33 is integrated into the circuit 3: it makes it possible to control a partial or total reduction of a fluid flow rate in the circuit, in particular in channels 34 located near or in contact with the magnets 111. The fluid circulating in the set of conduits 30 is distributed in the channels 34 by F intermediate an inlet distributor 35 and an outlet distributor 36. The shape of the channels 34 is advantageously defined so that each magnet is close to or in contact with a heat transfer fluid circulating in the channel.

[0061] In the remainder of the document, the term "total reduction" of the flow rate corresponds to a stoppage (i.e. a total stoppage) of the flow of the heat transfer fluid.

[0062] The estimation device 21 comprises a means for controlling the flow rate of the fluid circulating near or in contact with the magnet. The means for controlling the flow rate is made - either by directly controlling the flow rate of the pump 31 and thus also modifying the flow rates of lubricating and cooling fluid of the stator 12, - either by including a solenoid valve 33 making it possible to modify only the flow of fluid cooling the magnets, while maintaining the flow of fluid for the other components in order to ensure the cooling and / or lubrication of the stator 12.

[0063] The solenoid valve 33 is arranged on a conduit of the assembly 30, the conduit being located upstream of the magnets 111 relative to a direction of circulation of heat transfer fluid near or in contact with the magnets.

[0064] The solenoid valve 33 is controlled by the estimation device 21 in order to implement the method for estimating a temperature of a rotor magnet according to the invention. The solenoid valve 33 makes it possible to partially or totally reduce the flow rate of a heat transfer fluid intended to cool a magnet. The solenoid valve advantageously makes it possible to partially or totally reduce the flow rate of heat transfer fluid in the rotor without modifying the flow rate of heat transfer fluid in other components of the motor vehicle. The solenoid valve 33 thus makes it possible to frequently take temperature measurements of the magnet without slowing down or stopping the flow rate of a heat transfer fluid which would be intended, for example, for the lubrication of bearings and / or the cooling of the stator.

[0065] For at least one magnet of the rotor, and preferably for each magnet of the rotor, the cooling circuit previously described allows circulation of a heat transfer fluid in contact with or near the magnet 111, between an inlet section 341 of a channel 34 running along the magnet and an outlet section 342 of the channel 34. A temperature sensor 4 is arranged in the channel, in particular near the outlet section 342 of the channel 34.

[0066] In one embodiment, the temperature sensor 4 can be arranged just after a junction point of at least two channels, so as to be in contact with a heat transfer fluid coming from the at least two channels.

[0067] The estimation device 21 according to the invention comprises means for implementing an estimation method according to the invention, in particular a processing unit 210, comprising a microprocessor 211, a memory 212 and communication interfaces. communication 213. The microprocessor 211 mainly comprises the following modules which cooperate with each other: - a module 2111 for determining a given duration of partial or total reduction of a flow rate of the heat transfer fluid circulating in a channel, - a module 2112 for partial or total reduction, for the given duration, of the flow rate of the heat transfer fluid circulating in the channel, this module being able to cooperate with the solenoid valve 33, - a module 2113 for measuring the temperature of a heat transfer fluid by the temperature sensor 4, this module being able to cooperate with the temperature sensor 4, and - a module 2114 for determining an estimated temperature of the magnet.

[0068] The control device 2 of the powertrain advantageously comprises means for determining conditions of use of the motor vehicle 100 which may generate a risk of heating of the rotor magnets likely to damage the magnets. The conditions of use include in particular an engine torque controlled by a driver of the motor vehicle.

[0069] For this, the control device 2 can use information from an on-board data network of the motor vehicle.

[0070] The control device 2 comprises means for implementing a control method according to the invention, in particular a processing unit 20, comprising a microprocessor 201, a memory 202 and communication interfaces 203. The microprocessor 201 mainly comprises the following modules which cooperate with each other: - a 2011 module for detecting a need to estimate the temperature of a magnet, - a 2012 module for implementing an estimation method according to the invention, - a 2013 module for reducing the engine torque available from the powertrain.

[0071] The motor vehicle 100, in particular the control device 2, and the estimation device 21, preferably comprises all the hardware and / or software elements configured so as to implement the methods defined in the subject of the invention or the methods described below.

[0072] A mode of execution of an estimation method is described below with reference to [Fig.5]. The estimation method according to the invention comprises steps E1 to E4 which are executed successively.

[0073] In a first step El, a given duration D_reduc of partial or total reduction of a flow rate of the heat transfer fluid circulating in a channel is determined, and of a constant difference EC between an estimated temperature of the magnet T_estimated and a measured temperature T_mes of the heat transfer fluid circulating near the magnet.

[0074] The given duration D_reduc is a duration during which, in step E2 of reduction, we will totally or partially reduce a circulation of heat transfer fluid near the magnet.

[0075] The determination of the given duration D_reduc and the constant deviation EC includes an implementation of a thermal model Mi l, M12, M21, M22, M23 of an evolution of a temperature of the heat transfer fluid generated solely by a transfer of a thermal power from the magnet to the heat transfer fluid.

[0076] In an alternative or complementary embodiment, the given duration D_reduc was determined upstream of the implementation of the powertrain control device, using a series of experiments, then the duration D_reduc was recorded, for example in a memory of a powertrain control device. Advantageously, the experiments also made it possible to adjust the models M1, M2, M21, M22, M23.

[0077] Figures 6 to 8 illustrate a geometry of a magnet and a heat transfer fluid channel intended for cooling the magnet. The calculations described in the rest of the document take into account the following dimensions: the width of the channel 111 is called l_Fluid and the length of the channel is called L_Canal. In the example illustrated by [Fig.7], the width l_Fluid is equal to 30 millimeters, and the fluid height is 2 millimeters.

[0078] In the rest of the document, it is assumed that the flow of the heat transfer fluid in a channel is laminar. Therefore, the Nusselt number is constant. [Fig.29] represents a table for calculating the Nusselt number as a function of the geometry of a section through which the heat transfer fluid flows.

[0079] The Nusselt number is a dimensionless number used to characterize heat transfer between a fluid and a wall, called convective heat transfer. It represents the ratio of convective heat transfer to conductive heat transfer across an interface. The Nusselt number is equal to the dimensionless temperature gradient at the surface, and it provides a measure of the convective heat transfer occurring at the surface.

[0080] By definition, the Nusselt number is: - Ml where: h is a heat exchange coefficient calculated according to the following formula; h —X and expressed in W . m -2 . K -1 n Magnet-Fluid ~ D Hydrmditlue k is the thermal conductivity of the fluid expressed in W. m -1 . K -1 X. is the conductivity of the fluid, expressed in W .irr 1 .K -1 D the hydraulic diameter D Hydraulic expressed in meters.

[0081] By definition, the hydraulic diameter is equal to the product by 4 of the quotient between the surface and the perimeter of the fluid flow section, i.e. for a rectangular section of dimensions a and b, the hydraulic diameter is:

[0082] In the rest of the document, the following notations are used: P Fluid is the density of the fluid, expressed in kg . m -3 , Cpiuide est L a specific heat of the fluid, expressed in J. kg -1 . K -1 This is the section of the canal expressed in m 2 . JCaria l

[0083] Furthermore, in the rest of the document, reference is made to a heat exchange coefficient between the magnet and the heat transfer fluid.

[0084] A thermal model is a mathematical model used to characterize the temporal evolution of the temperature of a magnet and the temperature of a heat transfer fluid located near or in contact with the magnet. Based on different modeling hypotheses, different methods of implementing a thermal model are proposed.

[0085] In one embodiment, the thermal model is - a first model, built on a hypothesis of a constancy of a temperature of the magnet over the given duration D_reduc of reduction of the flow rate of heat transfer fluid and of a homogeneity of a temperature of the magnet over an entire length of the magnet, the length of the magnet being measured according to the direction of the channel, or - a second model, constructed on an assumption of homogeneity, at each instant of the given duration D_reduc of reduction of the heat transfer fluid flow rate, of a temperature of the magnet over its entire length, the temperature of the magnet being variable over time, the first and second models each comprising a first variant in which, during the given duration D_reduc, the heat transfer fluid flow rate is non-zero, and a second variant in which, during the given duration D reduc, the heat transfer fluid flow rate is zero.

[0086] In other words, the first model is based on the assumption that the magnet is a source of temperature, and the second model is based on the assumption that the magnet is a component in which thermal power is dissipated, this thermal power coming from the eddy currents circulating in the magnet.

[0087] A first hypothesis concerns a level of reduction in the flow rate of heat transfer fluid, the level being able to be total or partial. We thus define two families of models, - a first family Ml of models Mil, M12 relates to a case of partial reduction of the flow rate of the heat transfer fluid, - a second family M2 of models M21, M22, M23, relates to a case of total reduction of the flow rate of the heat transfer fluid.

[0088] Each of the thermal models Mi l, M12, M21, M22, M23 provides - a temporal evolution Evol_TMA of a modeled temperature of a TMA magnet, and - a temporal evolution Evol_TMF of a modeled temperature of a heat transfer fluid TMF circulating in contact with the magnet.

[0089] As for the first family of Ml models, two model variants are treated: - a first variant Mi l in which we assume that the magnet is isothermal: the temperature of the magnet is assumed to be homogeneous over its entire length and constant over time, - a second variant M12 in which we assume that the temperature of the magnet is homogeneous over its entire length but varies over time.

[0090] In other words, in the first and second variants M11, M12, the heat exchanges by conduction from one area of ​​the magnet to another area of ​​the magnet take place instantaneously, the thermal conductivity of the magnet being high enough to allow this assumption to be made. In addition, in the second variant M12, the temperature of the magnet is variable over time. The mass of the magnet and the specific heat of the magnet, therefore the heat capacity of the magnet, are taken into account.

[0091] The first variant Ml 1 of the thermal model in the case of a partial reduction in the flow rate of the heat transfer fluid is described with reference to figures 9 to 12.

[0092] We consider a volume element of heat transfer fluid of thickness dx and section equal to the section of the Scanal channel.

[0093] We calculate a thermal power transferred by the magnet via the contact surface with the heat transfer fluid, noted P exchanged by the Magnet with the fluid, selon a first formula:

[0094] In addition, the thermal power transferred by the magnet via the contact surface with the heat transfer fluid is calculated using a second formula: Magnet Exchange WHERE is the width of the channel.

[0095] Thus, we obtain the following equation:

[0096] Simplifying by dx, we obtain:

[0097] A resolution of this equation by the finite difference method with a time increment At, allows us to obtain the following equation:

[0098] Thus, according to the mathematical model Mi l, the evolution of the temperature of the heat transfer fluid is described by the following graphs: - a first graph Glll ([Fig.9]) shows a temporal evolution of the temperature of the heat transfer fluid between an instant T0 of start of reduction of the flow rate, and an instant Tmax equal to 150 seconds, - a second graph G112 ([Fig.10]) details a temporal evolution of the temperature of the heat transfer fluid between the instant T0 of start of reduction of the flow rate, and an instant Tmax equal to 100 seconds, - a third graph G113 ([Fig.11]) shows the distance traveled by the heat transfer fluid in the channel as a function of time, - a fourth graph G114 ([Fig.12]) details an evolution of the temperature of the heat transfer fluid as a function of a distance traveled in the channel.

[0099] Graphs G111 and G112 show that the heat transfer fluid reaches approximately the temperature of the magnet after 60 seconds.

[0100] In summary, according to the first thermal model Ml 1, the elementary volume of fluid of thickness dx heats up as it progresses in the channel. Its heating rate is determined by the exchange coefficient which is constant, due to the laminar flow inducing a constant Nusselt number.

[0101] According to the thermal model Ml 1, the flow rate of the heat transfer fluid therefore does not affect its heating rate. On the other hand, a high flow rate would cause the fluid element of thickness dx in contact with the magnet to quickly exit the channel, not leaving time for this fluid element to heat up to the point where its temperature approaches closely that of the magnet. The flow rate must therefore be reduced to give the fluid element the time needed to reach the temperature of the magnet.

[0102] In the embodiment of the magnet and the heat transfer fluid channel considered, for a heat transfer fluid flow rate of 0.007 L / min, the channel crossing time is 62 seconds, and at the end of this time the temperature reached by the fluid element is equal to the temperature of the magnet, i.e. 150°C.

[0103] Furthermore, it takes only 37 seconds for the heat transfer fluid to reach a temperature 3° Celsius below the temperature of the magnet. The distance covered is then 7 centimeters, i.e. significantly less than the length of the magnet. The heat transfer fluid flow rate could therefore be increased so that the distance covered in 37 seconds is equal to the length of the magnet, i.e. equal to 12 centimeters in the embodiment of the magnet and the heat transfer fluid channel described.

[0104] In other words, when the heat transfer fluid flow rate varies by a factor of two, the variation in heat transfer fluid temperature recorded is limited to a few degrees. It is therefore not necessary to precisely determine a heat transfer fluid flow rate value, which makes the magnet temperature measurement very robust.

[0105] With reference to Figures 5 to 7 and 13 to 16, the second variant M12 of the thermal model with flow reduction is described below, in which it is assumed that the temperature of the magnet is homogeneous along its entire length but varies over time.

[0106] The thermal model used was constructed on the basis of the following assumptions: - In the M12 model, and unlike the Mi l model, the mass of the magnet and the specific heat of the magnet are taken into account. The temperature of the magnet depends on the different heat exchanges with the magnet. - Thus, according to this model, a thermal power called "iron losses" (Hysteresis and eddy current) dissipates in the magnet and a thermal power is transferred from the magnet to the heat transfer fluid, which is colder than the magnet.

[0107] The respective temperature variations of the magnet and the heat transfer fluid circulating in the channel are analyzed in order to determine a necessary duration for reducing the fluid flow rate. - so that a thermal equilibrium temperature between the magnet and the heat transfer fluid is reached, or - failing that, that a temperature difference DIFF between the magnet and the heat transfer fluid is constant, which makes it possible to deduce the temperature of the magnet from that of the heat transfer fluid.

[0108] Compared to the first modeling method Mi l, the second modeling method M12 advantageously allows for taking into account a more realistic evolution of temperatures and for taking into account physical phenomena at stake, in particular the cooling and heating of magnets.

[0109] We consider a volume element of heat transfer fluid of thickness dx and section equal to the section of the Scanal channel.

[0110] We calculate a thermal power transferred by the magnet via the contact surface with the heat transfer fluid, noted P échangée par ï Magnet with the fluid, according to a first formula: [YES] The thermal power dissipated in the magnet P Thermal J is distributed homogeneously in the volume of the magnet. In an elementary volume of magnet of thickness dx, the power dissipated in the magnet is:

[0112] For the magnet, the heat exchange balance is written: where $ Magnet is the cross-section of the magnet and P / U . ■rrUlt n U. is the density of the magnet.

[0113] By simplifying by dx, and by formatting we obtain the following equation:

[0114] By applying a resolution method called finite differences, with a time dt, we obtain the following equation:

[0115] For the fluid, the heat exchange balance is written in the same way as we saw in the case of the isothermal magnet. The temperature variation of the fluid volume of length dx is dictated by the thermal power received by this volume. We write:

[0117] P exchanged by the magnet with the fluid corresponding to the thermal power given off by the magnet by the contact surface with the fluid. This surface is worth dx x width of the channel.

[0118] Heat transfer is determined by a product of a temperature difference between the magnet and heat transfer fluid and an exchange coefficient hA iimnt _ liquid-

[0119] Combining the two previous equations, we obtain:

[0120] Simplifying by dx, we obtain the equation

[0121] Solving this equation by the finite difference method with a time increment gives: By treating the heat exchange coefficient and its relationship with the Nusselt number using the same method as in the Mil model and considering the same embodiment of the magnet while taking into account a power dissipated in the magnet of 20 Watts, the resolution of the equation seen previously makes it possible to obtain the graphs in figures 13 to 16.

[0122] Thus, according to the mathematical model Ml 2, the evolution of the temperature of the heat transfer fluid is described by the following graphs: - a first graph G121 ([Fig.13]) shows a temporal evolution of the temperature of the magnet and the heat transfer fluid between an instant T0 of start of reduction of the flow rate, and an instant Tmax equal to 90 seconds, - a second graph G122 ([Fig.14]) shows a temporal evolution of the temperature of the magnet and the heat transfer fluid as a function of a distance traveled by the heat transfer fluid, - a third graph G123 ([Fig.15]) details a temporal evolution of the temperature of the magnet and the heat transfer fluid as a function of a distance traveled by the heat transfer fluid, - a fourth graph G124 ([Fig.16]) describes a distance traveled by the heat transfer fluid in the channel as a function of time.

[0123] On graph G123, we observe that the elementary volume of fluid of thickness dx, heats up as it progresses in the channel. Its heating rate is determined by the exchange coefficient which is constant, due to the laminar flow inducing a constant Nusselt number.

[0124] As with the mathematical model Mi l, in the model Ml 2, the flow rate of the heat transfer fluid in the channel does not influence the heating rate of the fluid, however it is necessary to determine a flow rate low enough to allow the fluid element dx to reach the temperature of the magnet.

[0125] In the practical case considered here, for a heat transfer fluid flow rate equal to 0.007 liters per minute, the time taken for the fluid to cross the channel is 62 seconds. In the first half of the simulation, a drop in the temperature of the magnet is observed due to a transfer of thermal energy from the magnet to the heat transfer fluid. Gradually, the temperature of the heat transfer fluid increases until it reaches the temperature of the magnet and both temperatures increase very slightly as a function of time. However, the thermal power dissipated in the magnet due to iron losses in the magnet continues and therefore the balance is such that the magnet heats up.

[0126] We see that we are approaching a kind of asymptote of difference between the temperature of the magnet and that of the fluid at a distance of 10 centimeters. Beyond this distance, the temperature of the magnet begins to rise, but the difference between the temperature of the magnet and that of the fluid is only 2 degrees. During this time, the increase in the temperature of the magnet does not exceed 2°C.

[0127] The flow rate can be increased slightly and thus the distance at which the heat transfer fluid temperature closest to that of the magnet can be increased by one or two centimeters, without the magnet temperature beginning to rise. Alternatively, the situation illustrated in Figures 13 to 16 can be applied, with the increase in magnet temperature being only 2°C.

[0128] We can therefore see that we have several means of adjusting the flow rate to be able to measure the temperature of the magnet.

[0129] A second family M2 of models M21, M22, M23 is described, relating to a case of total reduction of the heat transfer fluid flow rate. In other words, during the duration D_reduc the heat transfer fluid does not circulate in the channel.

[0130] Apart from the implementation of the estimation method, the heat transfer fluid circulates in the channel at the nominal speed and crosses the channel in a few seconds. The temperature variation of the fluid between the channel inlet and the channel outlet is relatively small. When the fluid suddenly stops circulating, all the fluid contained in the channel is at approximately a fairly low, uniform temperature. For the duration D_reduc, a heat transfer takes place from the magnet to the heat transfer fluid, which creates an increase in the temperature of the fluid until it approaches that of the magnet asymptotically.

[0131] The models of the second family are divided into two subsets, - a first subset, comprising the third and fourth models M21 and M22, in which the magnet is considered to be isothermal, and - a second subset comprising the fifth model M23, in which the magnet is considered not to be isothermal.

[0132] The third model M21 is described with reference to figures 17 to 19. In the third model M21 it is assumed that the heat transfer fluid is monobloc, that is to say that the temperature of the fluid is homogeneous throughout the thickness eF of the channel.

[0133] We are interested in the variation of the temperature of the heat transfer fluid only, that of the magnet, T_Aimant being constant. We determine a duration D_reduc necessary for thermal equilibrium to be reached, that is to say to achieve a stabilization of the temperature of the fluid in the channel located under the magnet. Alternatively, we can determine a duration D_reduc necessary for a temperature difference between the magnet and the fluid to be constant.

[0134] Thermal power is transferred by conduction from the magnet to the fluid, with the thermal power transfer being determined by the temperature gradient between the magnet and the fluid as well as the thermal conductivity of the fluid. The transferred thermal power is calculated using the following formula:

[0135] The equation governing the fluid temperature as a function of the thermal power transfer from the magnet to the fluid is written:

[0136] By defining a diffusivity ô by the following formula

[0137] we obtain the following equality: This first degree differential equation is solved literally to obtain the following relationship between fluid temperature and magnet temperature: where T Aimant is constant and T initial pi is the initial temperature of the heat transfer fluid at the moment when it comes to a standstill in the channel.

[0138] We can also solve this equation by the finite difference method, with a time increment zlf, we obtain: [Fig.19] is a graph of the evolution of the temperature of the heat transfer fluid as a function of time. It illustrates the fact that the temperature of the heat transfer fluid reaches that of the magnet after approximately 90s to 100s. As a note, this duration can be significantly reduced if we take into account the thermal power transferred by the iron of the rotor to the fluid present in the channel.

[0139] The fourth thermal model M22 is described with reference to Figures 20 to 22. In the fourth model M22 it is assumed that the heat transfer fluid is a superposition of ten layers of equal thickness, each having a homogeneous temperature that varies over time. The heat transfer from the magnet to the fluid is done by pure thermal conduction in the fluid, from the hottest zone to the coldest zone.

[0140] Such modeling of the heat transfer fluid in several layers concerns cases where the flow rate of the heat transfer fluid is zero: we then block the flow rate of heat transfer fluid and wait for the temperature of the heat transfer fluid to become substantially equal to the temperature of the magnet within a "reasonable" time.

[0141] This fluid modeling is illustrated in Figure 20. The thickness of a fluid layer is called e F and the channel thickness is equal to 10 times the thickness of a layer e F.

[0142] As in the M21 model, in the M22 model it is assumed that the heat transfer fluid is heated only by the magnet. In other words, the heat transfer fluid is considered to be isolated from the rotor iron, in contact with which it circulates. The temperature of the fluid is considered homogeneous in each layer and it evolves in the time.

[0143] We consider the first layer of fluid, that is, the layer of fluid closest to the magnet, and we express the equations of thermal power transfer by conduction from the magnet to the first layer of fluid.

[0144] The thermal balance of the first layer of fluid is expressed as follows:

[0145] Simplifying, we obtain:

[0146] By defining a diffusivity ô by the following formula x > fluid the resolution by the finite difference method with an increment u — nc ^Fluid Fluide temporal At gives: The thermal balance of fluid layer i is expressed as follows: layer i receives thermal power from layer i - 1 and transfers thermal power to layer i +1, the thermal power received from layer i - 1 being calculated as follows, thermal power transferred to layer i + 1 calculated as follows

[0147] The total power received by layer i is expressed as follows: P Thermal received by layer i ~ ^Thermal received from layer il ~ P Thermal transferred to layer i+1

[0148] Either,

[0149] The temperature variation of layer i is linked to the thermal power received by layer i, which allows us to obtain the following equation: By defining a diffusivity ô by the following formula § — ApIuide we obtain the following equation: P Fluid / ' Fluid dT Fluid & > s / T Fluid M (*) - 2 '^Fluid ld)+ T Fluid i+ 1(f) dt \ ej.

[0150] By applying a finite difference resolution method with a time increment dt we obtain

[0151] T Fluide i (t + dt) = T Fl

[0152] For the tenth layer we obtain,

[0153] After successive resolutions of the equations of the different layers by the finite difference method, we obtain for each layer a curve of the time evolution of the temperature of a heat transfer fluid circulating in each layer, as represented by graph G221 of [Fig.22],

[0154] The duration D_reduc of stopping the circulation of the fluid is determined by an instant when all the layers of fluid have reached the same temperature, which corresponds to a duration between 65 seconds and 75 seconds. As a note, this duration can be significantly reduced if we take into account the thermal power transferred by the iron of the rotor to the fluid present in the channel.

[0155] We see that the decomposition into several layers provides a better understanding of the phenomena involved. However, the duration D_reduc necessary to reach the same limit temperature in all the layers of heat transfer fluid is substantially the same as the duration obtained with the thermal model M21, that is to say with a model considering the fluid contained in the channel as a single block. (At least, for the reasonable channel thicknesses considered).

[0156] The fifth thermal model M23, described with reference to figures 23 to 26 differs from other models in that it takes into account a non-isothermal magnet in the context of a total reduction in heat transfer fluid flow rate.

[0157] The fifth thermal model M23 is therefore defined from the following hypotheses: - the temperature of the magnet is uniform over its entire length, thermal exchanges by conduction from one area of ​​the magnet to another area of ​​the magnet take place instantly due to the sufficiently high thermal conductivity of the magnet, - the temperature of the magnet varies over time; for this, the mass of the magnet and the specific heat of the magnet (therefore the heat capacity) are taken into account, - the fluid in the channel is considered as a block of material of homogeneous temperature and variable in time.

[0158] The temperature of the magnet depends on the various heat exchanges that the magnet undergoes. A first thermal power called "iron losses" dissipates in the magnet and a second thermal power escapes from the magnet into the heat transfer fluid, which is colder than the magnet.

[0159] We are interested in the variation of the temperature of the heat transfer fluid only, and in the variation of the temperature of the magnet. We determine a duration D_reduc necessary for thermal equilibrium to be reached, that is to say to achieve a stabilization of the temperature of the fluid and the magnet. Alternatively, we can determine a duration D_reduc necessary for a temperature difference EC between the magnet and the fluid to be constant.

[0160] Thus, when we measure this temperature, we can deduce that of the magnet using a basic calculation, which will allow us to know the temperature of the magnet precisely. This method has the advantage of reflecting the reality of the evolution of temperatures and of taking into account the physical phenomena involved, in particular the cooling and heating of the magnets.

[0161] Referring to Figures 23 and 24, the heat transfer fluid is considered in its entirety as a block of thickness eF, width l_Fluid and length L_Channel. It is assumed that the fluid is heated only by its contact surface with the magnet and that the part of the fluid in contact with the iron is thermally insulated.

[0162] The heat balance of the heat transfer fluid is calculated using the following formula:

[0163] Simplifying, we obtain,

[0164] By defining a diffusivity ô by the following formula

[0165] The resolution by the finite difference method with a time increment At can be written:

[0166] The thermal balance of the magnet is calculated using the following formula:

[0168] AI Magnet being the mass of the magnet and C Aimant being the specific heat of the magnet.

[0169] We then obtain

[0170] Solving this equation using the finite difference method with a time increment At allows us to obtain: T Magnetized + At) = T Magnet^J) We obtain a curve of the time evolution of the temperature of the magnet and the temperature of the heat transfer fluid circulating in the channel, as represented by graph G231 in [Fig.25].

[0171] Figures 25 and 26 illustrate the application of the M23 thermal model under the following conditions; - the length of the channel and the magnet is equal to 12 centimeters, and - the power dissipated in the magnet is 20 Watt.

[0172] Graph G231 shows that reaching an asymptote between the respective temperature curves of the magnet and the fluid requires a total reduction time D_reduc of the heat transfer fluid flow rate of approximately 90 seconds. After reaching the asymptote, the temperature difference between the magnet and the fluid is approximately 5°C. During the period of total reduction of the heat transfer fluid flow rate, the magnet temperature increased by 12°C. The total reduction of the heat transfer fluid flow rate has a significant impact on the cooling of the magnet. However, if we take into account the thermal power transferred by the rotor iron to the fluid present in the channel, then we can estimate that this duration is significantly reduced. This reduces the harmful consequences of stopping the fluid flow on the magnet temperature. We can consider that we have an upper limit temperature value and this value can constitute a safety margin. The magnet will cool down fairly quickly when the nominal flow rate is restored.

[0173] In summary, the five thermal models M11, M12, M21, M22, M23 previously described are mathematical models taking into account physical, dimensional and physical property quantities of materials.

[0174] The models were established on the assumption that the fluid heating was only done by the transfer of thermal power from the magnet to the fluid. However, in reality, the fluid is also heated by the other walls of the channel made of the rotor iron. As a result, the fluid temperature rises faster than in the described models, and the maximum fluid temperature is reached earlier than the D_reduc duration calculated by the models, which therefore allows the magnet temperature to be acquired even more quickly.

[0175] The use of thermal models shows that it is preferable to reduce the flow rate rather than interrupt it completely. The limit temperature is reached more quickly and the magnet temperature rises reasonably. The heat transfer fluid flow rate can be very low, for example 7 milliliters per second. The reduction time D_reduc depends on the characteristics of the magnet and the heat transfer fluid.

[0176] In one embodiment, the given duration D_reduc is determined from - of a respective temporal evolution Evol_TMA of a modeled temperature of a magnet TM A and - of a temporal evolution Evol_TMF of a modeled temperature of a heat transfer fluid TMF.

[0177] In particular, the given duration D_reduc is determined so that: - an increase in the modeled temperature of the TMA magnet during the given duration D_reduc is less than a first threshold SI, the first threshold SI being able to be equal to 6 degrees, or even 4 degrees, and - at a simulation time located from the given duration D_reduc, a difference DIFF between a modeled temperature of the magnet TMA and a modeled temperature of the heat transfer fluid TMF is substantially constant and less than a second threshold S2, for example less than 3 degrees.

[0178] Additionally, in one embodiment, the constant deviation EC is equal to a difference, calculated at a simulation time located from the given duration D_reduc, between a modeled temperature of the magnet TMA and a modeled temperature of the heat transfer fluid TMF.

[0179] Once the duration D_reduc of reduction of the fluid flow rate, the constant difference EC and the applied fluid flow rate have been determined, we move on to step E2 of partial or total reduction of the flow rate of heat transfer fluid circulating in a channel of the rotor.

[0180] In step E2, the flow rate of heat transfer fluid is partially or totally reduced for the duration D_reduc. To do this, the solenoid valve 33 is controlled to change its state from a first state passing to a second blocked state, and this for the duration D_reduc, as illustrated by [Fig.27],

[0181] Then we move on to step E3 of measuring the temperature of the heat transfer fluid using the temperature sensor 4.

[0182] The measurement step E3 may comprise, before the temperature measurement, a sub-step E31 of resuming circulation of the heat transfer fluid in which circulation of the heat transfer fluid is implemented at an intermediate flow rate, the intermediate flow rate being lower than an initial flow rate of circulation of the heat transfer fluid before flow rate reduction. In other words, just before measuring the temperature of the fluid at the outlet of the channel, the flow rate of the fluid is significantly increased to ensure that the temperature of a heat transfer fluid that has remained in the channel for the entire given duration of flow rate reduction is measured. The sub-step E31 is more particularly necessary if the flow rate of heat transfer fluid has been completely interrupted for the given duration D_reduc.

[0183] Following the measurement of the temperature T_mes of the heat transfer fluid, in step E4 an estimated temperature T_estimated of the magnet 111 is determined as being a sum between the measured temperature T_mes of the heat transfer fluid and the constant temperature difference EC zero or non-zero calculated in step EL

[0184] The method for controlling a motor vehicle powertrain according to the invention comprises three steps E11, E12, E13 which are executed successively.

[0185] In a first step El 1, a need is detected to estimate a temperature of a magnet 111 of the rotor.

[0186] As previously described, the powertrain includes its own dynamic model 22, which makes it possible to estimate a magnet temperature with an accuracy of the order of one or two tens of degrees.

[0187] In step E1 1, it is determined whether the accuracy of the dynamic model 22 is sufficient to ensure proper operation of the powertrain, or whether it is necessary to measure the temperature of the magnet more precisely using the estimation device 21 according to the invention to verify the temperature estimations. provided by the dynamic 22 model, and to optimally control the powertrain.

[0188] In one embodiment, the detection of a need to use the estimation device 21 according to the invention comprises a comparison, with a temperature threshold, of the temperature estimated of the magnet by the dynamic model 22. Thus, when the temperature estimated of the magnet by the dynamic model 22 exceeds a temperature threshold, then a need to measure the temperature of the magnet more precisely using the estimation device 21 is determined.

[0189] In addition, the detection of a need to use the estimation device 21 according to the invention may be linked to conditions of use of the motor vehicle. For example, the detection may relate to a high engine torque controlled at a given time by a driver of the vehicle. Alternatively, the detection could relate to maintaining a high travel speed for a given duration. Advantageously, the detection relates to driving parameters or configurations likely to generate heating of the magnet 111.

[0190] A need to measure the temperature of the magnet punctually may be detected, and / or a need to measure the temperature of the magnet periodically may be detected.

[0191] Then we move on to a step of implementing, at the given instant or at the given frequency, the method of estimating a current temperature of a magnet according to the invention.

[0192] As soon as the estimated current temperature is higher than a given temperature threshold, the available torque of the powertrain is reduced. Thus, the motor vehicle will temporarily operate with degraded performance, and in the event of a powertrain control fault, the currents flowing in the stator windings will be moderated. This will prevent demagnetization of the rotor magnets.

[0193] Finally, the invention makes it possible to estimate a temperature of a rotor magnet from a single temperature measurement of a cooling fluid at the outlet of the channel intended for cooling the magnet. The estimation method according to the invention does not require knowing the temperature of the fluid at the inlet of the channel, nor the flow rate of heat transfer fluid in the channel.

[0194] The invention is based on a very significant, or even total, reduction in the flow rate of fluid circulating in the channels under the magnets, for a given duration. Thus, the temperature of the heat transfer fluid at the outlet of the channel reaches that of the magnets. After measuring the temperature, we return to the nominal fluid flow rate necessary for cooling the rotor and the magnets in particular.

[0195] Thanks to the thermal models implemented in the process, a duration of partial or total reduction of the heat transfer fluid flow rate is determined such that: - at the end of the reduction time, the temperature of the fluid at the outlet of the channel has reached a value very close to that of the magnet, - the magnet temperature has not increased unacceptably during the reduction time.

[0196] The estimation method according to the invention improves the powertrain control method. Improving the accuracy of estimating the temperature of the rotor magnets makes it possible to reduce the torque of the powertrain appropriately, and thus to deprive the user of the maximum performance of the vehicle as little as possible. In addition, the control method according to the invention makes it possible to avoid damage to the rotor magnets due to short-circuiting of the stator windings during a powertrain control fault.

Claims

Claims

1. Method for estimating a temperature of a magnet (111) of a rotor (II), the rotor (11) comprising a channel (34) implementing a circulation of a heat transfer fluid in contact with or near the magnet (III), between an inlet section of the channel (341) and an outlet section of the channel (342), a temperature sensor (4) being arranged in the channel (34), in particular near the outlet section of the channel (342), characterized in that it comprises: • a step (E2) of partial or total reduction, for a given duration (D_reduc), of a flow rate of the heat transfer fluid circulating in the channel (34), • a step (E3) of measuring by the temperature sensor (4) a temperature (T_mes) of a given heat transfer fluid having remained in the channel (34) throughout the given duration (D_reduc), and • a step (E4) of determining an estimated temperature (T_estimated) of the magnet (111) as being a sum between the measured temperature (T_mes) of the given heat transfer fluid and a constant temperature difference (EC) which is zero or non-zero.

2. Estimation method according to the preceding claim, characterized in that it comprises a step (El) of determining the given duration (D_reduc) and the constant difference (EC) comprising an implementation of a thermal model (M11, M12, M21, M22, M23) of an evolution of a temperature of the heat transfer fluid generated solely by a transfer of a thermal power from the magnet to the heat transfer fluid.

3. Estimation method according to the preceding claim, characterized in that the thermal model is a first model, constructed on an assumption of a constancy of a modeled temperature of the magnet over the given duration and of a homogeneity of a modeled temperature of the magnet over an entire length of the magnet, the length of the magnet being measured along the direction of the channel, or a second model, constructed on an assumption of a ho- homogeneity, at each instant of the given duration, of a modeled temperature of the magnet over its entire length, the modeled temperature of the magnet being variable over time, the first and second models each comprising a first variant in which, during the given duration (D_reduc), the flow rate of the heat transfer fluid is non-zero, and a second variant in which, during the given duration (D_reduc), the flow rate of the heat transfer fluid is zero.

4. Estimation method according to one of claims 2 or 3, characterized in that the thermal model (M11, M12, M21, M22, M23) provides a temporal evolution of a modeled temperature of a magnet (TMA) and a temporal evolution of a modeled temperature of a heat transfer fluid (TMF) circulating in contact with the magnet, in that the given duration (D_reduc) is determined so that - an increase in the modeled temperature of the magnet (TMA) during the given duration (D_reduc) is less than a first threshold (SI), in particular the first threshold (SI) being equal to 6 degrees, or even 4 degrees and - at a simulation time located from the given duration (D_reduc), a difference (DIFF) between a modeled temperature of the magnet (TMA) and a modeled temperature of the heat transfer fluid (TMF) is substantially constant and less than a second threshold (S2), for example less than 3 degrees, and in that the constant difference (EC) is equal to a difference, calculated at a simulation time located from the given duration (D_reduc), between a modeled temperature of the magnet (TMA) and a modeled temperature of the heat transfer fluid (TMF).

5. Estimation method according to one of the preceding claims, characterized in that the measurement step (E3) comprises a sub-step of resuming (E31) a circulation of the heat transfer fluid at an intermediate flow rate, the intermediate flow rate being less than or equal to an initial flow rate of circulation of the heat transfer fluid before flow rate reduction.

6. Device (21) for estimating a temperature of a magnet of a rotor, a cooling circuit of the magnet comprising a set of conduits (30) implementing a circulation of a cooling fluid in contact with the magnet (111), the device comprising a means for controlling a flow rate of a solenoid valve (33) arranged on a conduit of the set of conduits (30), the conduit being located upstream of the magnet relative to a direction of circulation of heat transfer fluid near or in contact with the magnet, and the device comprising hardware and / or software elements (210, 211, 212, 213, 2111, 2112, 2113) implementing the method according to one of the preceding claims, in particular hardware elements (210, 211, 212, 213) and / or software designed to implement the method according to one of the preceding claims.

7. Method for controlling a motor vehicle powertrain, characterized in that it comprises • a step (El 1) of detecting, as a function of the conditions of use of the motor vehicle, a need to estimate, at a given instant, a temperature of the magnet, or a need to estimate, at a given frequency, a temperature of the magnet, then • a step (E12) of implementing, at the given instant or at the given frequency, a method for estimating a current temperature of a magnet according to one of claims 1 to 5, then • if the current temperature of the magnet is higher than a temperature threshold, a step (El 3) of reduction of an available engine torque of the powertrain.

8. Control method according to the preceding claim, characterized in that the conditions of use comprise an engine torque controlled by a driver of the motor vehicle at the given time.

9. Control device (2) for a motor vehicle powertrain (100), the device comprising hardware and / or software elements (20, 21, 22, 201, 202, 203, 2011, 2012, 2013) implementing the method according to one of claims 7 or 8, in particular hardware elements (20, 21, 22, 201, 202, 203) and / or software designed to implement the method according to one of claims 7 or 8.

10. Motor vehicle (100) equipped with a control device (2) according to the preceding claim, or with an estimation device (21) according to claim 6.

Images