Method for optimising a criterion or a combination of criteria relating to a motor vehicle

Abstract

The invention relates to a method for optimising a criterion or a combination of criteria relating to a motor vehicle (2) comprising a fuel or hydrogen tank, a battery and / or supercapacitors (30), a combustion engine (M) or a fuel cell (P), an electric machine (ME), a plurality of devices, each characterised by at least one state variable, and a computer (4) that is configured to manage the drive train of the motor vehicle (2) along a predetermined route and capable of controlling the combustion engine (M) or fuel cell (P), the electric machine (ME), and / or the devices, by transmitting a set of control setpoints. The invention also relates to a computer (4) and to a computer program product for implementing such a method, as well as to a motor vehicle (2) comprising such a computer (4).

Description

DESCRIPTION METHOD FOR OPTIMIZING A CRITERION OR A COMBINATION OF CRITERIA RELATING TO A MOTOR VEHICLE

[0001] The present invention relates to the optimization of a criterion or a combination of criteria relating to a motor vehicle, and more particularly to a method for optimizing a criterion or a combination of criteria of a motor vehicle on a predefined route. The invention aims in particular to generate optimized control instructions for controlling the vehicle on said route to be traveled by minimizing the criterion or the combination of criteria. The criterion to be optimized is typically a pollution level, an energy consumption, a state of health of the vehicle or a duration of a predetermined journey of the vehicle. The combination of criteria to be optimized consists of a combination of at least two criteria chosen from the group consisting of: a pollution level, an energy consumption, a state of health and a duration of a predetermined journey.For the purposes of the present invention, the term "vehicle energy consumption" means a fuel consumption of an internal combustion engine, a hydrogen or any other fuel consumption of a fuel cell, an electrical energy consumption of an electrical energy storage system (battery, super-capacitors, etc.), a number of restarts of a generator set in a series hybrid vehicle (also called a "range extender"), or a combined consumption of several of these elements. For the purposes of the present invention, the term "vehicle health" means a percentage that varies from 100% to 0% and that makes it possible to characterize the level of deterioration of an electric battery and / or one or more super-capacitors and / or a fuel cell of the vehicle, taking into account replacement and maintenance costs.

[0002] In a motor vehicle, it is known to optimize the energy consumption of the powertrain on a given or planned journey. Such optimization can be carried out on fuel, on electrical energy, on hydrogen or other fuel consumption (in the case of a vehicle equipped with a fuel cell) or on two or three of these criteria at the same time, by acting appropriately on the control instructions.

[0003] As is well known, optimization can be performed using the principle known as the Pontryagin Maximum Principle (PMP). This method consists of minimizing the Hamiltonian function (or Hamiltonian) from the criterion to be optimized, for example the quantity of fuel or hydrogen or the electrical energy consumed, and the description of the dynamics of the system. The dynamics of the system is defined from the derivatives of the different state variables of the vehicle (vehicle speed, state of charge of the battery, and / or super-capacitors, temperatures, etc.) and different control instructions (torque instructions to be applied to the vehicle wheels for the thermal engine or for the electric machine(s), torque instructions to be applied for the engine(s) of the generator set, and / or power or current instructions for the fuel cell, and / or catalyst heating instructions, and / or cooling circuit control instructions, etc.). The Hamiltonian function is minimized in order to determine the control instructions allowing the minimum consumption of fuel or hydrogen, and / or electrical energy to be obtained.

[0004] Each input or setpoint is notably dependent on the state of certain variables. For example, the torque setpoint for the electric machine to be applied to the vehicle wheels is dependent on the vehicle speed and the battery charge state, the cooling circuit control setpoint is dependent on the real-time temperature in the cooling circuit, the catalyst heating setpoint is dependent on the real-time temperature in the catalyst.

[0005] The Hamiltonian thus determined is then minimized. In other words, the control setpoint values for which the Hamiltonian value is the smallest are selected and applied to the vehicle. The control setpoint values are thus determined in real time based on the current state of the system.

[0006] In many real cases, additional “gradient constraints” must also be considered on the signals of control instructions (the problem is then called “Optimal control under gradient constraints”).

[0007] Known solutions consist of ignoring unacceptable control values (i.e., deriving from an excessive gradient step) outside an applicable control setpoint domain, before computing the optimal control solution by minimizing the Hamiltonian in a reduced control setpoint domain.

[0008] However, in case of strong and severe gradient constraints in the control values, the use of the Pontryagin Maximum Principle (PMP) algorithm can lead to a blocking point, because it remains fixed at a fixed local minimum, and therefore does not constitute a global minimum. Indeed, strict limitations in the variations of the control values can cause the control signals not to move sufficiently at each time step, preventing them from reaching better optimal points, sometimes too far from the current values. The algorithm then does not allow to reach the most optimal control values (global minimum of the Hamiltonian), while respecting the imposed gradient constraints.

[0009] There is therefore a need to have an optimal control method allowing the optimization of a criterion or a combination of criteria relating to a motor vehicle by implementing a PMP model, while making it possible to achieve the most optimal control values (global minimum of the Hamiltonian) in all circumstances, and while respecting the imposed gradient constraints.

[0010] To this end and according to a first aspect, the invention relates to a method, implemented in a computer embedded in a motor vehicle, for optimizing a pollution level of the vehicle, the vehicle comprising a fuel or hydrogen or other fuel tank, an electric battery and / or super-capacitors capable of supplying electrical energy, a heat engine powered by the fuel tank or a fuel cell powered by the hydrogen or other fuel tank, at least one electric machine powered by electrical energy supplied by the battery and / or the super-capacitors, at least one device relating to the heat engine or the fuel cell, at least one device relating to the electric machine, and at least one device relating to the electric battery or to the super-capacitors, the computer being configured to manage the powertrain of the motor vehicle over a predetermined route and being capable of controlling the heat engine or the fuel cell, the electric machine and / or the devices by issuing a set of control instructions, the heat engine or the fuel cell, the electric machine, and the devices each being characterized by at least one state variable, each state variable making it possible to describe the operating state of the device that it characterizes, the entire heat engine or the fuel cell, the electric machine, and the devices being represented by a system of state equations modeling the dynamics of the vehicle, said state equations being functions at least of the instantaneous control instruction values and the state variables,the pollution level to be optimized being represented by a criterion equation which is a function of at least the instantaneous control setpoint values and the state variables, at least one of these control setpoints being subject to at least one gradient constraint, said method being implemented for a duration divided into constant sampling times, the method comprising, at each sampling time, the steps of:, - a) determination of a domain of applicable control instructions comprising a set of values for each control instruction, - b) calculation, for each control instruction and in the determined domain of applicable control instructions, of all possible values of the state derivatives, given by the state equations, - c) calculation, for each control setpoint and in the determined range of applicable control sets, of all possible values for the pollution level to be optimized, given by the criterion equation, - d) calculation, in the determined domain of applicable control instructions, of all possible values of a Hamiltonian function of said instructions, using the pollution level values to be optimized determined by the criterion equation, and the values of the state equations, for said instructions, - e) determination of the value of the control instructions, in the determined applicable control instruction domain, for which said Hamiltonian function is the lowest, - f) calculation of a differential variation, over the sampling duration, between said determined value of each setpoint and the setpoint value applied at the previous sampling instant, - g) if, for all the control instructions, said differential variation is less than or equal to a predetermined maximum limit gradient value of the instruction, taking into account said determined value of the instruction as the optimal value, at the current sampling time, for the pollution level of the vehicle to be optimized, - h) otherwise, - h1) determination of a reduced control setpoint domain comprising a set of reduced values for each control setpoint, the gradient of each setpoint value being bounded in absolute value by said predetermined maximum setpoint gradient limit value, - h2) calculation, for each control setpoint and in the determined reduced control setpoint domain, of all possible values of the state derivatives, given by the state equations, - h3) calculation, for each control setpoint and in the determined reduced control setpoint domain, of all possible values for the pollution level to be optimized, given by the criterion equation, - h4) calculation, in the determined reduced control setpoint domain, of all possible values of a Hamiltonian function of said setpoints, using the pollution level values to be optimized determined by the criterion equation, and the values of the state equations, for said setpoints, - h5) determination of the value of the control instructions, in the determined reduced control instruction domain, for which said Hamiltonian function is the smallest, the minimization of the Hamiltonian function being carried out by minimizing a Lagrangian function, the Lagrangian function being determined from the previously determined Hamiltonian function and gradient constraint equations multiplied by so-called “Karush, Kuhn & Tucker” parameters relating to these constraints, in the determined reduced control setpoint domain, - h6) displacement, in the determined reduced control setpoint domain, of the Lagrangian function calculated for said control setpoints, by modifying said “Karush, Kuhn a Tucker” parameters, such that said determined values of the control setpoints for which said Hamiltonian function is the lowest correspond to a saturation value of the setpoint for which the differential variation of the setpoint is equal to the predetermined maximum limit gradient value of the setpoint; and taking into account said saturation value of the setpoint as the optimal value, at the current sampling instant, for the pollution level of the vehicle to be optimized.

[0011] According to a second aspect, the invention also relates to a method, implemented in an on-board computer within a motor vehicle, for optimizing the energy consumption of the vehicle, the vehicle comprising a fuel or hydrogen or other fuel tank, an electric battery and / or supercapacitors capable of supplying electrical energy, a heat engine powered by the fuel tank or a fuel cell powered by the hydrogen or other fuel tank, at least one electric machine powered by electrical energy supplied by the battery and / or the supercapacitors, at least one device relating to the heat engine or the fuel cell, at least one device relating to the electric machine, and at least one device relating to the electric battery or the supercapacitors,the computer being configured to manage the powertrain of the motor vehicle over a predetermined route and being capable of controlling the heat engine or the fuel cell, the electric machine and / or the devices by issuing a set of control instructions, the heat engine or the fuel cell, the electric machine, and the devices each being characterized by at least one state variable, each state variable making it possible to describe the operating state of the device that it, characterizes, the whole of the thermal engine or the fuel cell, the electric machine, and the devices being represented by a system of state equations modeling the dynamics of the vehicle, said state equations being functions at least of the instantaneous control setpoint values and the state variables, the energy consumption to be optimized being represented by a criterion equation which is a function at least of the instantaneous control setpoint values and the state variables, at least one of these control setpoints being subject to at least one gradient constraint, said method being implemented for a duration divided into constant sampling instants, the method comprising, at each sampling instant, the steps of: - i) determining a domain of applicable control instructions comprising a set of values for each control instruction, - j) calculation, for each control setpoint and in the determined domain of applicable control sets, of all possible values of the state derivatives, given by the state equations, - k) calculation, for each control setpoint and in the determined range of applicable control setpoints, of all possible values for the energy consumption to be optimized, given by the criterion equation, - I) calculation, in the determined domain of applicable control instructions, of all possible values of a Hamiltonian function of said instructions, using the energy consumption values to be optimized determined by the criterion equation, and the values of the state equations, for said instructions, - m) determination of the value of the control instructions, in the determined applicable control instruction domain, for which said Hamiltonian function is the lowest, - n) calculation of a differential variation, over the sampling duration, between said determined value of each setpoint and the setpoint value applied at the previous sampling instant, - o) if, for all control instructions, said differential variation is less than or equal to a maximum limit value of predetermined setpoint gradient, taking into account said determined setpoint value as the optimal value, at the current sampling time, for the energy consumption of the vehicle to be optimized, - p) otherwise, - p1) determination of a reduced control setpoint domain comprising a set of reduced values for each control setpoint, the gradient of each setpoint value being bounded in absolute value by said predetermined maximum setpoint gradient limit value, - p2) calculation, for each control setpoint and in the determined reduced control setpoint domain, of all possible values of the state derivatives, given by the state equations, - p3) calculation, for each control setpoint and in the determined reduced control setpoint domain, of all possible values for the energy consumption of the vehicle to be optimized, given by the criterion equation, - p4) calculation, in the determined reduced control setpoint domain, of all possible values of a Hamiltonian function of said setpoints, using the energy consumption values to be optimized determined by the criterion equation, and the values of the state equations, for said setpoints, - p5) determination of the value of the control instructions, in the determined reduced control instruction domain, for which said Hamiltonian function is the smallest, the minimization of the Hamiltonian function being carried out by minimizing a Lagrangian function, the Lagrangian function being determined from the previously determined Hamiltonian function and the gradient constraint equations multiplied by so-called “Karush, Kuhn & Tucker” parameters relating to these constraints, in the determined reduced control instruction domain, - p6) displacement, in the determined reduced control setpoint domain, of the Lagrangian function calculated for said control setpoints, by modifying said parameters of “Karush, Kuhn a Tucker”, such that said determined values of the control setpoints for which said Hamiltonian function is the lowest correspond to a setpoint saturation value for which the differential variation of the setpoint is equal to the predetermined maximum setpoint gradient limit value; and taking into account said setpoint saturation value as the optimal value, at the current sampling instant, for the energy consumption of the vehicle to be optimized.

[0012] According to a third aspect, the invention also relates to a method, implemented in an on-board computer within a motor vehicle, for optimizing a state of health of the vehicle, the vehicle comprising a fuel or hydrogen or other fuel tank, an electric battery and / or supercapacitors capable of supplying electrical energy, a heat engine powered by the fuel tank or a fuel cell powered by the hydrogen or other fuel tank, at least one electric machine powered by electrical energy supplied by the battery and / or the supercapacitors, at least one device relating to the heat engine or the fuel cell, at least one device relating to the electric machine, and at least one device relating to the electric battery or the supercapacitors,the computer being configured to manage the powertrain of the motor vehicle over a predetermined route and being capable of controlling the heat engine or the fuel cell, the electric machine and / or the devices by issuing a set of control instructions, the heat engine or the fuel cell, the electric machine, and the devices each being characterized by at least one state variable, each state variable making it possible to describe the operating state of the device that it characterizes, the entire heat engine or the fuel cell, the electric machine, and the devices being represented by a system of state equations modeling the dynamics of the vehicle, said state equations being functions at least of the instantaneous control instruction values and the state variables,the state of health to be optimized being represented by a criterion equation which is a function of at least the instantaneous control setpoint values and the state variables, at least one of these control setpoints being subject to at least, a gradient constraint, said method being implemented for a duration divided into constant sampling times, the method comprising, at each sampling time, the steps of: - q) determination of a domain of applicable control instructions comprising a set of values for each control instruction, - r) calculation, for each control instruction and in the determined domain of applicable control instructions, of all possible values of the state derivatives, given by the state equations, - s) calculation, for each control setpoint and in the determined domain of applicable control sets, of all possible values for the health state to be optimized, given by the criterion equation, - t) calculation, in the determined domain of applicable control instructions, of all possible values of a Hamiltonian function of said instructions, using the health state values to be optimized determined by the criterion equation, and the values of the state equations, for said instructions, - u) determination of the value of the control instructions, in the determined applicable control instruction domain, for which said Hamiltonian function is the lowest, - v) calculation of a differential variation, over the sampling duration, between said determined value of each setpoint and the setpoint value applied at the previous sampling instant, - w) if, for all the control instructions, said differential variation is less than or equal to a predetermined maximum limit gradient value of the instruction, taking into account said determined value of the instruction as the optimal value, at the current sampling time, for the state of health of the vehicle to be optimized, - x) otherwise, - x1) determination of a reduced control setpoint domain comprising a set of reduced values for each control setpoint, the gradient of each setpoint value being bounded in absolute value by said predetermined maximum setpoint gradient limit value, - x2) calculation, for each control setpoint and in the determined reduced control setpoint domain, of all possible values of the state derivatives, given by the state equations, - x3) calculation, for each control instruction and in the determined reduced control instruction domain, of all possible values for the health status of the vehicle to be optimized, given by the criterion equation, - x4) calculation, in the determined reduced control setpoint domain, of all possible values of a Hamiltonian function of said setpoints, using the health state values to be optimized determined by the criterion equation, and the values of the state equations, for said setpoints, - x5) determination of the value of the control instructions, in the determined reduced control instruction domain, for which said Hamiltonian function is the smallest, the minimization of the Hamiltonian function being carried out by minimizing a Lagrangian function, the Lagrangian function being determined from the previously determined Hamiltonian function and the gradient constraint equations multiplied by so-called “Karush, Kuhn & Tucker” parameters relating to these constraints, in the determined reduced control instruction domain, - x6) displacement, in the determined reduced control setpoint domain, of the Lagrangian function calculated for said control setpoints, by modifying said “Karush, Kuhn a Tucker” parameters, such that said determined values of the control setpoints for which said Hamiltonian function is the lowest correspond to a saturation value of the setpoint for which the differential variation of the setpoint is equal to the predetermined maximum limit gradient value of the setpoint; and taking into account said saturation value of the setpoint as the optimal value, at the current sampling instant, the state of health of the vehicle to be optimized.

[0013] According to a fourth aspect, the invention also relates to a method, implemented in an on-board computer within a motor vehicle, for optimizing a duration of a predetermined journey of the vehicle, the vehicle comprising a fuel or hydrogen or other fuel tank, an electric battery and / or super-capacitors capable of supplying electrical energy, a heat engine powered by the fuel tank or a fuel cell powered by the hydrogen or other fuel tank, at least one electric machine powered by electrical energy supplied by the battery and / or the super-capacitors, at least one device relating to the heat engine or the fuel cell, at least one device relating to the electric machine, and at least one device relating to the electric battery or the super-capacitors,the computer being configured to manage the powertrain of the motor vehicle over a predetermined route and being capable of controlling the heat engine or the fuel cell, the electric machine and / or the devices by issuing a set of control instructions, the heat engine or the fuel cell, the electric machine, and the devices each being characterized by at least one state variable, each state variable making it possible to describe the operating state of the device that it characterizes, the entire heat engine or the fuel cell, the electric machine, and the devices being represented by a system of state equations modeling the dynamics of the vehicle, said state equations being functions at least of the instantaneous control instruction values and the state variables,the duration of the predetermined journey to be optimized being represented by a criterion equation which is a function of at least the instantaneous control setpoint values and the state variables, at least one of these control setpoints being subject to at least one gradient constraint, said method being implemented for a duration divided into constant sampling instants, the method comprising, at each sampling instant, the steps of:, - y) determination of a domain of applicable control instructions comprising a set of values for each control instruction, - z) calculation, for each control setpoint and in the determined domain of applicable control sets, of all possible values of the state derivatives, given by the state equations, - aa) calculation, for each control instruction and in the determined domain of applicable control instructions, of all possible values for the duration of the predetermined journey to be optimized, given by the criterion equation, - bb) calculation, in the determined domain of applicable control instructions, of all possible values of a Hamiltonian function of said instructions, using the values of the duration of the journey to be optimized determined by the criterion equation, and the values of the state equations, for said instructions, - cc) determination of the value of the control instructions, in the determined applicable control instruction domain, for which said Hamiltonian function is the lowest, - dd) calculation of a differential variation, over the sampling time, between said determined value of each setpoint and the setpoint value applied at the previous sampling time, - ee) if, for all the control instructions, said differential variation is less than or equal to a predetermined maximum limit gradient value of the instruction, taking into account said determined value of the instruction as the optimal value, at the current sampling instant, for the duration of the predetermined path to be optimized, - ff) otherwise, - ff 1) determination of a reduced control setpoint domain comprising a set of reduced values for each control setpoint, the gradient of each setpoint value being bounded in absolute value by said predetermined maximum setpoint gradient limit value, - ff2) calculation, for each control setpoint and in the determined reduced control setpoint domain, of all possible values of the state derivatives, given by the state equations, - ff3) calculation, for each control setpoint and in the determined reduced control setpoint domain, of all possible values for the duration of the predetermined journey to be optimized, given by the criterion equation, - ff4) calculation, in the determined reduced control setpoint domain, of all possible values of a Hamiltonian function of said setpoints, using the values of the duration of the journey to be optimized determined by the criterion equation, and the values of the state equations, for said setpoints, - ff5) determination of the value of the control instructions, in the determined reduced control instruction domain, for which said Hamiltonian function is the smallest, the minimization of the Hamiltonian function being carried out by minimizing a Lagrangian function, the Lagrangian function being determined from the previously determined Hamiltonian function and the gradient constraint equations multiplied by so-called “Karush, Kuhn & Tucker” parameters relating to these constraints, in the determined reduced control instruction domain, - ff6) displacement, in the determined reduced control setpoint domain, of the Lagrangian function calculated for said control setpoints, by modifying said “Karush, Kuhn a Tucker” parameters, such that said determined values of the control setpoints for which said Hamiltonian function is the smallest correspond to a saturation value of the setpoint for which the differential variation of the setpoint is equal to the predetermined maximum limit gradient value of the setpoint; and taking into account said saturation value of the setpoint as the optimal value, at the current sampling instant, for the duration of the predetermined path to be optimized.

[0014] According to a fifth aspect, the invention also relates to a method, implemented in a computer embedded in a motor vehicle, for optimizing a combination of at least two criteria relating to the vehicle and chosen from the group consisting of: a pollution level, energy consumption, a state of health and a duration of a journey. predetermined, the vehicle comprising a fuel or hydrogen or other fuel tank, an electric battery and / or supercapacitors capable of supplying electrical energy, a heat engine powered by the fuel tank or a fuel cell powered by the hydrogen or other fuel tank, at least one electric machine powered by electrical energy supplied by the battery and / or the supercapacitors, at least one device relating to the heat engine or the fuel cell, at least one device relating to the electric machine, and at least one device relating to the electric battery or the supercapacitors, the computer being configured to manage the powertrain of the motor vehicle on said predetermined route and being capable of controlling the heat engine or the fuel cell, the electric machine and / or the devices by issuing a set of control instructions,the heat engine or the fuel cell, the electric machine, and the devices each being characterized by at least one state variable, each state variable making it possible to describe the operating state of the device that it characterizes, the whole of the heat engine or the fuel cell, the electric machine, and the devices being represented by a system of state equations modeling the dynamics of the vehicle, said state equations being functions at least of the instantaneous control setpoint values and the state variables, the combination of criteria to be optimized being represented by a criterion equation that is a function at least of the instantaneous control setpoint values and the state variables, at least one of these control setpoints being subject to at least one gradient constraint, said method being implemented for a duration divided into constant sampling instants, the method comprising,at each sampling time, the steps of:, - gg) determination of a domain of applicable control instructions comprising a set of values for each control instruction, - hh) calculation, for each control setpoint and in the determined domain of applicable control sets, of all possible values of the state derivatives, given by the state equations, - ii) calculation, for each control setpoint and in the determined domain of applicable control sets, of all possible values for the combination of criteria to be optimized, given by the criterion equation, - jj) calculation, in the determined domain of applicable control instructions, of all possible values of a Hamiltonian function of said instructions, using the values of the combination of criteria to be optimized determined by the criterion equation, and the values of the state equations, for said instructions, - kk) determination of the value of the control instructions, in the determined applicable control instruction domain, for which said Hamiltonian function is the lowest, - U) calculation of a differential variation, over the sampling duration, between said determined value of each setpoint and the setpoint value applied at the previous sampling instant, - mm) if, for all the control instructions, said differential variation is less than or equal to a predetermined maximum limit gradient value of the instruction, taking into account said determined value of the instruction as the optimal value, at the current sampling instant, for the combination of criteria to be optimized, - nn) otherwise, - oo1) determination of a reduced control setpoint domain comprising a set of reduced values for each control setpoint, the gradient of each setpoint value being bounded in absolute value by said predetermined maximum setpoint gradient limit value, - oo2) calculation, for each control setpoint and in the determined reduced control setpoint domain, of all possible values of the state derivatives, given by the state equations, - oo3) calculation, for each control setpoint and in the determined reduced control setpoint domain, of all possible values for the combination of criteria to be optimized, given by the criterion equation, - oo4) calculation, in the determined reduced control setpoint domain, of all possible values of a Hamiltonian function of said setpoints, using the values of the combination of criteria to be optimized determined by the criterion equation, and the values of the state equations, for said setpoints, - oo5) determination of the value of the control instructions, in the determined reduced control instruction domain, for which said Hamiltonian function is the smallest, the minimization of the Hamiltonian function being carried out by minimizing a Lagrangian function, the Lagrangian function being determined from the previously determined Hamiltonian function and the gradient constraint equations multiplied by so-called “Karush, Kuhn & Tucker” parameters relating to these constraints, in the determined reduced control instruction domain, - 006) displacement, in the determined reduced control setpoint domain, of the Lagrangian function calculated for said control setpoints, by modifying said “Karush, Kuhn a Tucker” parameters, such that said determined values of the control setpoints for which said Hamiltonian function is the lowest correspond to a saturation value of the setpoint for which the differential variation of the setpoint is equal to the predetermined maximum limit gradient value of the setpoint; and taking into account said saturation value of the setpoint as the optimal value, at the current sampling instant, for the combination of criteria to be optimized.

[0015] In each of the methods described above, the steps are looped iteratively at each new sampling time. In addition, each of the methods is implemented for a predefined / given route for the vehicle. Such a route is for example predicted by an “electronic information horizon” type system (or “eHorizon” in English, which is conventionally based on the ADASIS data format standard – from the English “Advanced Driver-Assistance Systems Interface Specifications” – for predictive driving assistance systems, or on any other type of device), which is connected to the vehicle’s computer. In a manner known per se, such a "eHorizon" type system can manage both static data relating to the road infrastructure (such as, for example, the nature of the roads, intersections, the regulatory speed limits applied, etc.), and dynamic data (average speed of vehicles located on the road, traffic density, dynamic data relating to road infrastructure elements, etc.). Such an "eHorizon" type system is capable of receiving this data, decoding it (via a decoder), reconstructing it (via a data reconstructor), and transmitting it to the vehicle's computer, and can implement vehicle route prediction algorithms using the concept of "most probable route or path" ("Most Probable Path" in English).

[0016] Each of the methods according to the invention makes it possible to define control instructions in order to provide the power and acceleration requested by the driver during operation of the vehicle on the given or planned route, while minimizing the pollution level, energy consumption and / or the health status of the vehicle. This assumes that there is at least one degree of freedom (or several solutions) making it possible to provide the power and acceleration requested by the driver in various forms, with several possible combined values of engine torque (between recharge or boost) or battery power (with 1 or N modules activated) or generator set starts, etc., the optimal control (with the PMP model) here making it possible to choose the best combination minimizing the desired criterion(s).Furthermore, by taking into account the Lagrangian function including the so-called “Karush, Kuhn & Tucker” parameters associated with the gradient constraints, as well as the displacement (translation) of this Lagrangian function by modifying these “Karush, Kuhn & Tucker” parameters, each of the methods makes it possible to achieve the most optimal control values (global minimum of the Hamiltonian) in all circumstances, respecting the imposed gradient constraints. Indeed, unlike the methods of the prior art, and by virtue of the aforementioned characteristics, each of the methods according to the invention prevents the optimization from remaining blocked on a fixed local minimum and guarantees reaching the global minimum of the Hamiltonian. Each of the methods according to the invention is therefore more precise and / or more stable from the point of view of the control strategy. optimization, compared to the methods of the prior art, and consequently constitutes a method of optimizing a criterion or a combination of criteria relating to a motor vehicle, by implementing an adapted and saturated PMP model.

[0017] Preferably, the step of moving, in the determined reduced control setpoint domain, the Lagrangian function calculated for said control setpoints, by modifying said “Karush, Kuhn & Tucker” parameters, is implemented via an analytical or numerical calculation method (in particular iterative).

[0018] The invention also relates to a computer for managing the powertrain of a motor vehicle over a predetermined route, the vehicle comprising, in addition to the computer, a fuel or hydrogen or other fuel tank, an electric battery and / or supercapacitors capable of supplying electrical energy, a heat engine powered by the fuel tank or a fuel cell powered by the hydrogen or other fuel tank, at least one electric machine powered by electrical energy supplied by the battery and / or the supercapacitors, at least one device relating to the heat engine or the fuel cell, at least one device relating to the electric machine, and at least one device relating to the electric battery or the supercapacitors, the computer being capable of controlling the heat engine or the fuel cell,the electric machine and / or the devices by the emission of a set of control instructions, the heat engine or the fuel cell, the electric machine, and the devices each being characterized by at least one state variable, each state variable making it possible to describe the operating state of the device that it characterizes, the whole of the heat engine or the fuel cell, the electric machine, and the devices being represented by a system of state equations modeling the dynamics of the vehicle, said state equations being functions at least of the instantaneous control instruction values and the state variables, at least one of these control instructions being subject to at least one gradient constraint, the computer being such that it is configured to implement the steps of one of the methods as described previously.,

[0019] The invention also relates to a motor vehicle comprising a fuel or hydrogen or other fuel tank, an electric battery and / or supercapacitors capable of supplying electrical energy, a heat engine powered by the fuel tank or a fuel cell powered by the hydrogen or other fuel tank, at least one electric machine powered by electrical energy supplied by the battery and / or the supercapacitors, at least one device relating to the heat engine or the fuel cell, at least one device relating to the electric machine, at least one device relating to the electric battery or the supercapacitors, and a computer for managing the powertrain of a motor vehicle on a predetermined route, the computer being capable of controlling the heat engine or the fuel cell, the electric machine and / or the devices by issuing a set of control instructions,the heat engine or the fuel cell, the electric machine, and the devices each being characterized by at least one state variable, each state variable making it possible to describe the operating state of the device that it characterizes, the whole of the heat engine or the fuel cell, the electric machine, and the devices being represented by a system of state equations modeling the dynamics of the vehicle, said state equations being functions at least of the instantaneous control setpoint values and the state variables, at least one of these control setpoints being subject to at least one gradient constraint, the vehicle further comprising a computer for managing the powertrain as described previously.,

[0020] The invention also relates to a computer program product which is remarkable in that it comprises a set of program code instructions which, when executed by one or more processors, configure the processor(s) to implement one of the methods as described above.

[0021] Embodiments of the present invention will be described below, by way of non-limiting examples, with reference to the appended figures in which: - [Fig.1] schematically illustrates a vehicle according to the invention, the vehicle being equipped with an on-board computer; and - [Fig.2] is a flowchart representing a method for optimizing a criterion or a combination of criteria relating to the vehicle, implemented by the computer of figure 1, according to the present invention.

[0022] Referring to Figure 2, the present invention relates to a method, implemented in a computer 4 embedded within a motor vehicle 2 (visible in Figure 1), for optimizing a criterion or a combination of criteria relating to the motor vehicle 2. The motor vehicle 2 is for example (but not limited to) a hybrid vehicle. In addition to the computer 4, such a hybrid vehicle 2 also conventionally comprises a thermal engine M (in the case of a hybrid vehicle 2 of the “thermal-electric” type), often called an “ICE” engine for “Internal Combustion Engine” in English, or a fuel cell P operating for example on hydrogen (in the case of a hybrid vehicle 2 of the “hydrogen-electric” type).Such a hybrid vehicle 2 further comprises at least one electric machine ME, often called an “EMA machine” for “Electrical Machine” in English, a fuel tank or a hydrogen or other fuel tank (not shown in FIG. 1) and an electrical power supply battery 30 (or supercapacitors in a variant not shown).

[0023] The motor vehicle 2 also comprises at least one device relating to the heat engine M and in particular a cooling device 10 for the heat engine M and a catalyst 20, or at least one device relating to the fuel cell P (such a device not being shown in FIG. 1). The cooling device 10 makes it possible to reduce the temperature of the heat engine M during its use. The cooling device 10 comprises in particular a coolant. The catalyst 20, connected to the heat engine M by the exhaust system, is capable of reducing the quantity of polluting products in the exhaust gases emitted by the heat engine M before discharging them outside the vehicle. The catalyst 20 also comprises a heating device capable of increasing the temperature in the catalyst 20, in order to carry out the depollution of the exhaust gases.The catalyst heating device 20 must be supplied with electrical energy to operate. The motor vehicle 2 also comprises at least one device relating to the machine. electric ME, in particular a set of voltage converters (not shown in FIG. 1) making it possible to convert the voltage between the battery 30 and the electric machine ME. The motor vehicle 2 also comprises at least one device relating to the electric battery 30 (or to the supercapacitors), such a device not being shown in FIG. 1 for reasons of clarity.

[0024] The heat engine M is in particular capable of being powered from the fuel supplied by the fuel tank, and the fuel cell P is capable of being powered by the hydrogen tank or other fuel. The heat engine M also comprises a system for exhausting the exhaust gases emitted during the combustion of the air and fuel mixture in the heat engine M. The electric machine ME is capable of being powered by the electrical energy supplied by the battery 30.

[0025] The vehicle 2 may include other devices relating to the heat engine M or to the fuel cell P, and other devices relating to the electric machine ME.

[0026] The term "system" refers to all the elements mounted in the vehicle 2, capable of consuming or producing electrical energy, fuel or hydrogen or other fuel. For example, the system comprises all the devices described previously: the heat engine M or the fuel cell P, the electric machine ME, the cooling device 10, the catalyst 20 and the battery 30.

[0027] Each device is characterized by at least one state variable, making it possible to describe the operating state of the device. For example, the cooling device 10 is characterized by a coolant temperature. For example again, the catalyst 20 is characterized by an internal temperature value. For example still, the battery 30 is characterized by a state of charge variable, the heat engine M is characterized by a rotation speed, the fuel cell P is characterized by a temperature or pressures prevailing in the hydrogen and oxygen supply circuits, etc.

[0028] The criterion or combination of criteria to be optimized via the method according to the present invention is for example an energy consumption of the vehicle 2, a pollution level of the vehicle 2, a state of health of the vehicle 2, or a combination of at least two criteria relating to the vehicle 2 and chosen from the group consisting of: a pollution level, an energy consumption, a state of health and a duration of a predetermined journey. The criterion or the combination of criteria to be optimized is represented by a criterion equation g (u,q). The criterion equation g (u,q) is a function at least of the instantaneous control setpoint values u and the state variables q. More preferably, the criterion equation is also a function of the disturbances and / or setpoints w applied to the vehicle 2, and is then written g (u,q,w). Each control setpoint is subject to at least one gradient constraint, as will be detailed later.

[0029] The system is represented by a system of state equations f (u,q) modeling the dynamics of the vehicle 2. The state equations f (u,q) are functions at least of the instantaneous control setpoint values u and the state variables q. More preferably, the state equations are also functions of the disturbances and / or the setpoints w applied to the vehicle 2, and are then written f (u,q,w).

[0030] The computer 4 is for example part of a data processing unit storing a computer application or program capable of cooperating with the computer 4 (the data processing unit and the computer application or program not being shown in FIG. 1 for reasons of clarity). Alternatively, the computer application or program is stored directly in the computer 4. The computer 4 is connected to the heat engine or fuel cell P, to the electric machine ME, to the electric battery 30, as well as to all of the devices described previously including in particular the cooling device 10 and the catalyst 20.

[0031] The computer 4 is capable of controlling each device to which it is connected, by issuing a control instruction, as a function of the value(s) of variables relating to this device.

[0032] Thus, for example, the instruction sent to the thermal engine M designates the value of the torque to be applied to the thermal engine M and is in particular a function of the speed of the vehicle and the power demand of the driver of the vehicle. For example again, the instruction sent to the fuel cell fuel P designates the power that the fuel cell P is asked to provide and is in particular a function of the state of charge of the battery 30 and / or the supercapacitors, as well as the driver's power demand. For example, the instruction sent to the electric machine ME designates the torque to be applied to the electric machine ME and is in particular a function of the state of charge of the battery 30 and the driver's power demand.

[0033] The setpoint sent to the catalyst 20 relates to the temperature in the catalyst 20 and is a function of the temperature measured in the catalyst 20. The setpoint sent to the cooling device 10 relates to the temperature of the coolant and is dependent on the measured temperature of the coolant.

[0034] The computer 4 is also configured to determine the domain of applicable control setpoints comprising a set of values for each control setpoint. The computer 4 is also configured to implement the principle of the PMP method, in other words the Pontryagin Maximum Principle method, by determining the Hamiltonian function H (q, u*, X) from the different setpoint values of the domain of applicable control setpoints. The notation u* is introduced here, which represents the optimal control. Adjoint states X (also called “adjoint parameters”, “Lagrange parameters”, “adjoint vectors”, or even “co-state vectors”) are also introduced. These adjoint states are associated with the equations of state which represent the conditions of the dynamic behavior of the physical system, and will allow the complete resolution of the optimization problem.

[0035] The computer 4 comprises a processor capable of implementing a set of instructions making it possible to carry out these functions.

[0036] With reference to Figure 2, an embodiment of the method for optimizing a criterion or a combination of criteria of the vehicle 2 according to the invention will now be described, implemented by a computer 4 as described previously. In order to simplify the description, the criterion to be optimized is for example the energy consumption of the vehicle 2. Of course, any other criterion or combination of criterion(s) relating to the vehicle 2 can be optimized in the same way using the method illustrated in figure 2, and in particular a pollution level of the vehicle 2, a state of health of the vehicle 2, a duration of a predetermined journey, or even a combination of at least two criteria relating to the vehicle 2 and chosen from the group consisting of: a pollution level, energy consumption, a state of health and a duration of a predetermined journey.

[0037] In order to simplify the description, the different variables considered are the following: the temperature of the coolant of the cooling device 10 and the state of charge of the battery 30 or the supercapacitors. The different control setpoints considered are for example: the torque of the thermal engine M and the torque of the electric machine ME. Other parameters could be considered, concerning the at least one device relating to the electric machine ME OR parameters concerning the fuel cell P or the at least one device relating to the fuel cell P, or even parameters concerning the at least one device relating to the battery 30 or the supercapacitors.

[0038] The method is implemented for a duration divided into constant sampling instants, and comprises steps that are looped iteratively at each new sampling instant. The sampling duration between two consecutive sampling instants depends on the dynamics of the system and can be chosen, for example, between 10 ms and 500 ms, typically equal to 100 ms.

[0039] The method comprises an initial step 22 during which the computer 4 determines a domain of applicable control instructions u comprising a set of values for each control instruction u to be applied to the variables.

[0040] The method then comprises a following step 24 during which the computer 4 calculates, for each control setpoint and in the determined domain of applicable control setpoints u, all the possible values of the state derivatives, using at least the state equations f (u,q,w) describing the system.

[0041] The method comprises a following step 26 during which the computer 4 calculates, for each control instruction and in the determined domain of applicable control instructions u, all the possible values for the criterion or the combination of criteria to be optimized, using the criterion equation g (u,q,w). In a variant not shown, the calculation steps 24, 26 can be reversed.

[0042] The method then comprises a following step 28 during which the computer 4 calculates, in the determined domain of applicable control setpoints u, all the possible values of a Hamiltonian function H (u,q,X,w) of said setpoints, using the values of the criterion or of the combination of criteria to be optimized determined by the criterion equation g (u,q,w), and the values of the state equations, for said setpoints. To do this, the computer 4 uses the principle of the PMP method, in other words the Pontryagin Maximum Principle method.

[0043] The Hamiltonian function H (u,q,X,w) is then expressed as: H (u, q, Å, w) = g (u, q, w) + T . f(u,q,w) with A T the transpose of the adjoint vector X.

[0044] It should be noted that, when the journey time is part of the criterion(s) to be optimized, the Hamiltonian function H (u,q,X,w) is then expressed as: H(u,q,To,w) = g(u,q,w)+a+ T . f (u, q, w) with a a strictly positive number; a then being a weighting coefficient allowing the duration of the journey to be weighted.

[0045] The computer 4 then determines, during a following step 31, the set of so-called “optimal” control instructions u*, by minimizing, for the control instructions, the Hamiltonian function calculated during the previous step 28. In the embodiment considered, the computer 4 thus determines a first optimal “thermal” instruction intended to control the thermal engine M, and a second optimal “electrical” instruction intended to control the electric machine ME.

[0046] During a following step 32, the calculator 4 calculates a differential variation, over the sampling duration, between the value of each setpoint determined during step 31 and the setpoint value applied during the previous sampling instant. This calculation of a differential variation over a sampling instant thus corresponds to a gradient calculation.

[0047] If, for all the control instructions, the differential variation calculated during the previous step 32 is less than or equal to a predetermined maximum limit gradient value of the instruction, the computer 4 takes into account, during a following step 34, the value of the instruction determined during the step 31 as the optimal value, at the current sampling instant, for the criterion(s) to be optimized.

[0048] If, for at least one of the control setpoints, the differential variation calculated during the previous step 32 is greater than the predetermined maximum limit gradient value, the computer 4 determines, during a following step 36, a reduced control setpoint domain u re d comprising a set of reduced values for each control setpoint. In this reduced control setpoint domain Ured, the gradient of each setpoint value is bounded in absolute value by the predetermined maximum setpoint gradient limit value.

[0049] During a step 38 following this step 36, the computer 4 calculates, for each control instruction and in the reduced control instruction domain u re d determined, all possible values of the state derivatives, using at least the state equations f (u,q,w) describing the system.

[0050] During a step 40 following this step 38, the computer 4 calculates, for each control instruction and in the reduced control instruction domain u re d determined, all possible values for the criterion or combination of criteria to be optimized, using the criterion equation g (u,q,w). In a variant not shown, the calculation steps 38, 40 can be reversed.

[0051] During a step 42 following this step 40, the computer 4 calculates, in the reduced control setpoint domain u re d determined, all possible values of a Hamiltonian function H (u,q,X,w) of said instructions, using the values of the criterion or combination of criteria to be optimized determined by the criterion equation g (u,q,w), and the values of the state equations, for said instructions. To do this, the calculator 4 uses the principle of the PMP method, in other words the Pontryagin Maximum Principle method.

[0052] The computer 4 then determines, during a following step 43, the set of so-called “optimal” control instructions u*, by minimizing, for the control instructions, the Hamiltonian function calculated during the previous step 42. This minimization of the Hamiltonian function is carried out by minimizing a Lagrangian function, the Lagrangian function being determined from the previously determined Hamiltonian function and the gradient constraint equations multiplied by so-called “Karush, Kuhn & Tucker” parameters relating to these constraints, in the reduced control instruction domain u red determined. The Lagrangian is thus determined according to the “Karush, Kuhn & Tucker” parameters, themselves defined according to the setpoint gradient constraints and calculated in such a way that the optimal control setpoints minimize the Lagrangian function. Thus, computer 4 calculates the “Karush, Kuhn & Tucker” parameters so that the derivative of the Lagrangian function is zero at the points of the optimal control setpoints obtained.

[0053] During a step 44 following this step 43, the computer 4 moves, in the reduced control setpoint domain u red determined, the Lagrangian function calculated for the control setpoints, by modifying said “Karush, Kuhn & Tucker” parameters, such that said determined values of the control setpoints for which the Hamiltonian function is the lowest correspond to a saturation value of the setpoint for which the differential variation of the setpoint is equal to the predetermined maximum limit gradient value of the setpoint.

[0054] The Hamiltonian function H (u,q,X,w) is then expressed as: H(u,q,To,w) = g(u,q,w)+ T . f (u, q, w) + pl. with To Tthe transpose of the adjoint vector X; p1 and p2 the “Karush, Kuhn & Tucker” parameters, with p1 or p2 non-zero, associated with the setpoint gradient constraint equations; h1 and h2 equations respecting in particular the predetermined maximum limit value of the setpoint gradient, and representing upper and lower saturations which are never active at the same time.

[0055] It should be noted that, when the journey time is part of the criterion(s) to be optimized, the Hamiltonian function H (u,q,X,w) is then expressed as: with a a strictly positive number; a then being a weighting coefficient allowing the duration of the journey to be weighted.

[0056] Preferably, this step 44 of moving the Hamiltonian function, and in particular of calculating the “Karush, Kuhn & Tucker” parameters p1, p2 is implemented via an analytical or numerical calculation method (in particular iterative) known to those skilled in the art. The resulting value of p1, p2 can then be used to obtain a transversal condition which makes the link between the time derivative of the adjoint state X and the gradient of the Hamiltonian function H, which now also depends on the “Karush, Kuhn & Tucker” parameters p1, p2. This makes it possible to obtain the correct evolution of the adjoint state Δ(t) for the next sampling instants.

[0057] During a step 46 following this step 44, the calculator 4 takes into account this saturation value of the setpoint as the optimal value, at the current sampling time, for the criterion or the combination of criteria to be optimized.

[0058] The process can be repeated while the vehicle is in use.

[0059] The method thus makes it possible to obtain control setpoint values for which the pollution level, and / or the energy consumption, and / or the state of health of the vehicle 2, and / or the journey time are minimal.

[0060] Furthermore, by taking into account the Lagrangian function and the so-called “Karush, Kuhn & Tucker” parameters in the minimization of the Hamiltonian function, as well as the displacement (translation) of the Lagrangian function by modifying these “Karush, Kuhn & Tucker” parameters, the method makes it possible to achieve the most optimal control values (global minimum of the Hamiltonian) in all circumstances, with respect to imposed setpoint gradient constraints. The method is therefore more precise and / or more stable from the point of view of the optimization control strategy, compared to the methods of the prior art, and constitutes by consequently a method of optimizing a criterion or a combination of criteria relating to a motor vehicle, by implementing an adapted and saturated PMP model.

Claims

CLAIMS 1. Method, implemented in a computer (4) embedded in a motor vehicle (2), for optimizing a criterion relating to the vehicle (2) and chosen from the group consisting of: a pollution level, an energy consumption, a state of health and a duration of a predetermined journey, the vehicle (2) comprising a fuel or hydrogen or other fuel tank, an electric battery and / or supercapacitors (30) capable of supplying electrical energy, a heat engine (M) powered by the fuel tank or a fuel cell (P) powered by the hydrogen or other fuel tank, at least one electric machine (ME) powered by electrical energy supplied by the battery and / or the supercapacitors (30), at least one device (10, 20) relating to the heat engine (M) or to the fuel cell (P), at least one device (30) relating to the electric machine (ME),and at least one device relating to the electric battery or to the super-capacitors (30), the computer (4) being configured to manage the powertrain of the motor vehicle (2) on a predetermined route and being able to control the heat engine (M) or the fuel cell (P), the electric machine (ME) and / or the devices (10, 20, 30) by issuing a set of control instructions, the heat engine (M) or the fuel cell (P), the electric machine (ME), and the devices (10, 20, 30) each being characterized by at least one state variable, each state variable making it possible to describe the operating state of the device that it characterizes, the whole of the heat engine (M) or the fuel cell (P), the electric machine (ME), and the devices (10, 20, 30) being represented by a system of state equations modeling the dynamics of the vehicle (2),said state equations being functions at least of the instantaneous control setpoint values and the state variables, the criterion to be optimized being represented by a criterion equation which is a function at least of the instantaneous control setpoint values and the state variables, at least one of these control setpoints being subject to at least one gradient constraint, said method being implemented for one, duration divided into constant sampling instants, the method being characterized in that it comprises, at each sampling instant, the steps of: - a) determination (22) of a domain of applicable control instructions comprising a set of values for each control instruction, - b) calculation (24), for each control instruction and in the determined domain of applicable control instructions, of the possible values of the state derivatives, given by the state equations, - c) calculation (26), for each control instruction and in the determined domain of applicable control instructions, of the possible values for the criterion to be optimized, given by the criterion equation, - d) calculation (28), in the determined domain of applicable control instructions, of the possible values of a Hamiltonian function of said instructions, using the criterion values to be optimized determined by the criterion equation, and the values of the state equations, for said instructions, - e) determination (31) of the value of the control instructions, in the determined applicable control instruction domain, for which said Hamiltonian function is the lowest, - f) calculation (32) of a differential variation, over the sampling duration, between said determined value of each setpoint and the setpoint value applied at the previous sampling instant, - g) if, for all the control instructions, said differential variation is less than or equal to a predetermined maximum limit gradient value of the instruction, taking into account (34) said determined value of the instruction as the optimal value, at the current sampling time, for the criterion of the vehicle (2) to be optimized, - h) otherwise, - h1) determination (36) of a reduced control setpoint domain comprising a set of reduced values for each control setpoint, the gradient of each setpoint value being bounded in absolute value by said predetermined maximum setpoint gradient limit value, - h2) calculation (38), for each control instruction and in the determined reduced control instruction domain, of the possible values of the state derivatives, given by the state equations, - h3) calculation (40), for each control setpoint and in the determined reduced control setpoint domain, of the possible values for the criterion to be optimized, given by the criterion equation, - h4) calculation (42), in the determined reduced control setpoint domain, of the possible values of a Hamiltonian function of said setpoints, using the criterion values to be optimized determined by the criterion equation, and the values of the state equations, for said setpoints, - h5) determination (43) of the value of the control instructions, in the determined reduced control instruction domain, for which said Hamiltonian function is the smallest, the minimization of the Hamiltonian function being carried out by minimizing a Lagrangian function, the Lagrangian function being determined from the previously determined Hamiltonian function and the gradient constraint equations multiplied by so-called “Karush, Kuhn & Tucker” parameters relating to these constraints, in the determined reduced control instruction domain, - h6) displacement (44), in the determined reduced control setpoint domain, of the Lagrangian function calculated for said control setpoints, by modifying said “Karush, Kuhn a Tucker” parameters, such that said determined values of the control setpoints for which said Hamiltonian function is the lowest correspond to a saturation value of the setpoint for which the differential variation of the setpoint is equal to the predetermined maximum limit gradient value of the setpoint; and taking into account (46) said saturation value of the setpoint as the optimal value, at the current sampling instant, for the criterion (2) to be optimized.

2. Method, implemented in a computer (4) embedded in a motor vehicle (2), according to the preceding claim, for optimizing a combination of at least two criteria relating to the vehicle (2) and chosen from the group consisting of: a pollution level, an energy consumption, a state of health and a duration of a predetermined journey, the vehicle (2) comprising a fuel or hydrogen or other fuel tank, an electric battery and / or super-capacitors (30) capable of supplying electrical energy, a heat engine (M) powered by the fuel tank or a fuel cell (P) powered by the hydrogen or other fuel tank, at least one electric machine (ME) powered by electrical energy supplied by the battery and / or the super-capacitors (30), at least one device (10, 20) relating to the heat engine (M) or to the fuel cell (P), at least one device (30) relating to the electric machine (ME),and at least one device relating to the electric battery or to the super-capacitors (30), the computer (4) being configured to manage the powertrain of the motor vehicle (2) on said predetermined path and being able to control the heat engine (M) or the fuel cell (P), the electric machine (ME) and / or the devices (10, 20, 30) by issuing a set of control instructions, the heat engine (M) or the fuel cell (P), the electric machine (ME), and the devices (10, 20, 30) each being characterized by at least one state variable, each state variable making it possible to describe the operating state of the device that it characterizes, the whole of the heat engine (M) or the fuel cell (P), the electric machine (ME), and the devices (10, 20, 30) being represented by a system of state equations modeling the dynamics of the vehicle (2),said state equations being functions at least of the instantaneous control setpoint values and the state variables, the combination of criteria to be optimized being represented by a criterion equation which is a function at least of the instantaneous control setpoint values and the state variables, at least one of these control setpoints being subject to at least one gradient constraint, said method being implemented for a duration divided into constant sampling instants, the method being characterized in that it comprises, at each sampling instant, the steps of:, - gg) determination (22) of a domain of applicable control instructions comprising a set of values for each control instruction, - hh) calculation (24), for each control instruction and in the determined domain of applicable control instructions, of the possible values of the state derivatives, given by the state equations, - ii) calculation (26), for each control instruction and in the determined domain of applicable control instructions, of the possible values for the combination of criteria to be optimized, given by the criterion equation, - jj) calculation (28), in the determined domain of applicable control instructions, of the possible values of a Hamiltonian function of said instructions, using the values of the combination of criteria to be optimized determined by the criterion equation, and the values of the state equations, for said instructions, - kk) determination (31) of the value of the control instructions, in the determined applicable control instruction domain, for which said Hamiltonian function is the lowest, - U) calculation (32) of a differential variation, over the sampling duration, between said determined value of each setpoint and the setpoint value applied at the previous sampling instant, - mm) if, for all the control instructions, said differential variation is less than or equal to a predetermined maximum limit gradient value of the instruction, taking into account (34) said determined value of the instruction as the optimal value, at the current sampling instant, for the combination of criteria to be optimized, - nn) otherwise, - oo1) determination (36) of a reduced control setpoint domain comprising a set of reduced values for each control setpoint, the gradient of each setpoint value being limited in absolute value by said predetermined maximum setpoint gradient limit value, - oo2) calculation (38), for each control instruction and in the determined reduced control instruction domain, of the possible values of the state derivatives, given by the state equations, - oo3) calculation (40), for each control setpoint and in the determined reduced control setpoint domain, of the possible values for the combination of criteria to be optimized, given by the criterion equation, - oo4) calculation (42), in the determined reduced control setpoint domain, of the possible values of a Hamiltonian function of said setpoints, using the values of the combination of criteria to be optimized determined by the criterion equation, and the values of the state equations, for said setpoints, - oo5) determination (43) of the value of the control instructions, in the determined reduced control instruction domain, for which said Hamiltonian function is the smallest, the minimization of the Hamiltonian function being carried out by minimizing a Lagrangian function, the Lagrangian function being determined from the previously determined Hamiltonian function and the gradient constraint equations multiplied by so-called “Karush, Kuhn & Tucker” parameters relating to these constraints, in the determined reduced control instruction domain, - 006) displacement (44), in the determined reduced control setpoint domain, of the Lagrangian function calculated for said control setpoints, by modifying said “Karush, Kuhn a Tucker” parameters, such that said determined values of the control setpoints for which said Hamiltonian function is the lowest correspond to a saturation value of the setpoint for which the differential variation of the setpoint is equal to the predetermined maximum limit gradient value of the setpoint; and taking into account (46) said saturation value of the setpoint as the optimal value, at the current sampling instant, for the combination of criteria to be optimized.

3. Method according to any one of claims 1 to 2, in which the step of moving (44), in the determined reduced control setpoint domain, the Lagrangian function calculated for said setpoints control, by modifying the said parameters of “Karush, Kuhn & Tucker”, is implemented via an analytical or numerical calculation method.

4. Computer (4) for managing the powertrain of a motor vehicle (2) over a predetermined route, the vehicle (2) comprising, in addition to the computer (4), a fuel or hydrogen or other fuel tank, an electric battery and / or supercapacitors (30) capable of supplying electrical energy, a heat engine (M) powered by the fuel tank or a fuel cell (P) powered by the hydrogen or other fuel tank, at least one electric machine (ME) powered by electrical energy supplied by the battery and / or the supercapacitors (30), at least one device (10, 20) relating to the heat engine (M) or the fuel cell (P), at least one device (30) relating to the electric machine (ME), and at least one device relating to the electric battery or the supercapacitors (30), the computer (4) being capable of controlling the heat engine (M) or the fuel cell (P),the electric machine (ME) and / or the devices (10, 20, 30) by issuing a set of control instructions, the heat engine (M) or the fuel cell (P), the electric machine (ME), and the devices (10, 20, 30) each being characterized by at least one state variable, each state variable making it possible to describe the operating state of the device that it characterizes, the whole of the heat engine (M) or the fuel cell (P), the electric machine (ME), and the devices (10, 20, 30) being represented by a system of state equations modeling the dynamics of the vehicle (2), said state equations being functions at least of the instantaneous control instruction values and the state variables, at least one of these control instructions being subject to at least one gradient constraint, the computer (4) being characterized in that it is configured to implement the steps of the method according to any one of claims 1 to 3., 5. Motor vehicle (2) comprising a fuel or hydrogen or other fuel tank, an electric battery and / or supercapacitors (30) capable of supplying electrical energy, a heat engine (M) powered by the fuel tank or a fuel cell (P) powered by the hydrogen or other fuel tank, at least one electric machine (ME) powered by electrical energy supplied by the battery and / or the super-capacitors (30), at least one device (10, 20) relating to the heat engine (M) or to the fuel cell (P), at least one device (30) relating to the electric machine (ME), at least one device relating to the electric battery or to the super-capacitors (30), and a computer (4) for managing the powertrain of a motor vehicle (2) on a predetermined route, the computer (4) being capable of controlling the heat engine (M) or the fuel cell (P), the electric machine (ME) and / or the devices (10, 20, 30) by issuing a set of control instructions, the heat engine (M) or the fuel cell (P), the electric machine (ME), and the devices (10, 20, 30) each being characterized by at least one state variable, each state variable making it possible to describe the operating state of the device that it characterizes, the entire heat engine (M) or fuel cell (P), electric machine (ME),and devices (10, 20, 30) being represented by a system of state equations modeling the dynamics of the vehicle (2), said state equations being functions at least of the instantaneous control setpoint values and of the state variables, at least one of these control setpoints being subject to at least one gradient constraint, the vehicle (2) further comprising a computer (4) for managing the traction chain according to claim 4., 6. Computer program product characterized in that it comprises a set of program code instructions which, when executed by one or more processors, configure the processor(s) to implement a method according to any one of claims 1 to 3.

Images