Method for operating a vehicle energy system

Abstract

The invention relates to a method for operating a vehicle energy system (100) having a fuel cell system (104) which has a plurality of fuel cell stacks (106, 106'), and having at least one stored energy source (110) for storing electrical energy. The method comprises the steps of: acquiring information relating to the vehicle energy system (100); determining a time period (TH) to be predicted for an operating strategy of the vehicle energy system (100) on the basis of the acquired information; determining a predictive operating strategy for operating the vehicle energy system (100) on the basis of the determined time period (TH) to be predicted; and operating the vehicle energy system (100) according to the determined predictive operating strategy.

Description

[0001] R.417020

[0002] - 1 -

[0003] Description

[0004] title

[0005] Procedures for operating a

[0006]

[0007] State of the art

[0008] Hydrogen-based fuel cell systems are considered a mobility concept of the future, as they emit only water as exhaust gas and enable rapid refueling. Fuel cell systems require air and hydrogen for the chemical reaction. The only reaction product is water, which is released in varying proportions as a gas and a liquid. The water is produced at the catalyst layer on one side of the cathode and transported by a gas diffusion layer (GDL) towards the respective gas flow channels.

[0009] Such a fuel cell system can be integrated into or interact with a vehicle energy system. In addition to the fuel cell system, such a vehicle energy system includes an energy storage device, such as a battery, and can interact with a vehicle drive system, such as an electric motor.

[0010] An operating strategy for a fuel cell system can optimize several requirements simultaneously, for example minimizing hydrogen consumption and system aging at the same time.

[0011] Despite the advantages of the state-of-the-art methods for operating vehicle energy systems with a fuel cell system, these still offer potential for improvement. For example, predictive operating strategies exist for the optimal control of the power split of fuel cell systems (R.417020).

[0012] - 2 -

[0013] The process involves battery management, control between multiple stacks, and setpoint generation for their temperature, pressures, and stoichiometry. Predictive data is used in the control unit to optimize operation based on several objective functions. The prediction horizon lengths used in conventional predictive operating strategies are static and based, for example, on the predicted worst-case scenario or the limitations of the processing unit. Since sustainable decisions must be made early for operational planning, this duration must be at least as long as a necessary measure is required to comply with operating limits in the future. However, in normal operation, long prediction horizons offer only marginal benefits while placing a disproportionate burden on the control unit due to their calculation.

[0014] Disclosure of the invention

[0015] Within the scope of the present invention, a method for operating a vehicle energy system with a fuel cell system comprising multiple fuel cell stacks and at least one energy storage device for storing electrical energy, a vehicle energy system, a motor vehicle, and a computer program product are therefore proposed, which largely avoid the disadvantages of known methods for operating a vehicle energy system, a vehicle energy system, a motor vehicle, and a computer program product, and which, in particular, determine an ideal prediction horizon length for predictive fuel cell operating strategies depending on available information.

[0016] An inventive method for operating a vehicle energy system with a fuel cell system having multiple fuel cell stacks and with at least one energy storage device for storing electrical energy comprises the following steps, wherein individual or all steps can be repeated:

[0017] Acquiring information concerning the vehicle energy system, determining a predictable time duration for an operating strategy of the vehicle energy system based on the acquired information, R.417020

[0018] - 3 -

[0019] Determining a predictive operating strategy for operating the vehicle energy system based on the determined time duration to be predicted, and

[0020] Operating the vehicle's energy system according to the determined predictive operating strategy.

[0021] Under challenging conditions, the necessary prediction horizon can be provided to best meet operational requirements. This results, among other things, in minimal thermal power derating, maximum component protection by preventing, for example, temperature overshoots, and optimal warm-up time. In normal operating conditions, where the fuel cell stack is already well-conditioned, ambient temperature and pressure are within the normal range, etc., the prediction horizon can be reduced to utilize less computing power in the control unit without compromising performance. This also reduces energy consumption due to the lower computing power. Furthermore, it allows for the temporary execution of other processes that require high computing power (e.g., temporary diagnostics). This adjustment occurs automatically (adaptively), eliminating the need for an initial expert guess.This increases reliability and can reduce effort.

[0022] The information can include at least one and preferably several pieces of information selected from the following group, consisting of current and / or predicted values ​​of: ambient temperature, ambient humidity, ambient pressure, fuel cell stack conditions, battery state of charge, battery temperature, aging state of the fuel cell stack, aging state of the battery, aging state of the drive system, aging state of the components, trailer load, predicted future power demand of the electric motors, driving speed, and road gradient. Such inputs allow for reliable planning of the predicted duration for an operating strategy.

[0023] The time duration to be predicted can be determined using at least one characteristic curve, heuristically and / or by solving an optimization problem online (R.417020).

[0024] - 4 -

[0025] This allows the predictable duration for an operating strategy to be reliably and accurately determined.

[0026] The time duration to be predicted can be determined using an onboard control unit in a vehicle. This allows the prediction horizon length to be determined locally in the vehicle, thus avoiding or minimizing potential delays in the determination process. Alternatively, the time duration to be predicted can be determined using a cloud server and transmitted to the vehicle. Therefore, the determination of the time duration to be predicted using a control unit can alternatively be carried out via a cloud server and sent to the vehicle, provided good connectivity is available.

[0027] The process can also run when the vehicle is stationary and / or the fuel cell system is in standby mode. The horizon length planner can also be extended to include stationary vehicle or stationary / standby fuel cell system states in order to adaptively trigger or monitor functions that should run in these states, such as recovery functions, frost protection measures, and parking purge in the anode loop, and to predict wakeup intervals.

[0028] The time period to be predicted can be at least as long as the duration of a measure to be implemented by the operational strategy. This ensures that the prediction horizon length for the operational strategy is at least as long as a necessary measure is required to comply with operational limits in the future.

[0029] Furthermore, a vehicle energy system is proposed comprising a fuel cell system with multiple fuel cell stacks and at least one control unit. The control unit is configured to perform a method according to one of the embodiments described above or below. R.417020

[0030] - 5 -

[0031] The control unit can include a horizon length determination module and an operational management module. The horizon length determination module can be configured to determine a predictable operating time for a vehicle energy system strategy based on information about the vehicle energy system. The operational management module can be configured to determine a predictive operating strategy for operating the vehicle energy system based on the determined predictable operating time.

[0032] The control unit may also include a driving state estimation module. The driving state estimation module may be configured to supply predicted driving state data to the horizon length determination module, which in turn may be configured to consider the predicted driving state data for the duration to be predicted for an operating strategy of the vehicle energy system.

[0033] Furthermore, a motor vehicle is proposed that incorporates such a vehicle energy system.

[0034] Finally, a computer program product is proposed with program code means which, when the computer program product is executed on a computer, configure the computer to perform a method according to one of the embodiments described above or below.

[0035] The proposed computer program product could be, for example, a file to be downloaded from a server or a data carrier, such as a CD-ROM or a USB stick.

[0036] The advantages which have been described in detail with regard to the operating method for operating a fuel cell system according to the invention apply equally to the vehicle according to the invention and to the computer program product according to the invention.

[0037] Within the scope of the present invention, a fuel cell system can be understood to be a system comprising at least one fuel cell system.

[0038] - 6 -

[0039] The fuel cell system comprises at least one anode path comprising an anode, an anode gas supply line, and an anode gas return line; at least one cathode path comprising a cathode, a cathode gas supply line, and a cathode gas return line; at least one thermal system comprising a radiator; a sensor unit for acquiring data to determine the performance of the fuel cell system; a processing unit for determining the performance of the fuel cell system; and a control unit. The control unit is designed to regulate the operation of the fuel cell system or the fuel cell stack. A fuel cell stack comprises at least two fuel cells, preferably at least 10 fuel cells, and more preferably at least 100 fuel cells. A fuel cell consists of electrodes between which an electrolyte (ion conductor) is located. The electrodes are the aforementioned anode and cathode.The electrolyte can be a liquid, such as alkalis or acids, or molten alkali carbonate. In high-temperature fuel cells, a solid is used as the electrolyte, such as ion-conducting ceramic, which then forms a solid electrolyte. Membranes are also used. These are semipermeable membranes that are only permeable to one type of ion, e.g., protons. A membrane can also separate two different liquid electrolytes. The energy is supplied by a reaction of oxygen with the fuel. This is often hydrogen, but organic compounds such as methane or methanol are also used. Both reactants are continuously supplied via the electrodes. The fuel cell system can also include a housing, in which at least one fuel cell stack is contained.The fuel cell system may further include a control unit for activating a device for targeted adjustment of a water loading of a membrane of the fuel cell system, as well as a device for targeted adjustment of a water loading of a membrane of the fuel cell system.

[0040] The fuel cell system has several subsystems, such as the anode subsystem, which includes the anode path and one or more hydrogen tanks, the cathode subsystem, which includes the cathode path, an air compressor and a humidifier, and the elekt-R.417020

[0041] - 7 -

[0042] The electrical subsystem, which includes, among other things, the electrical components such as electrical connections, and the thermal system, which includes, among other things, heating, cooling system, coolant, coolant pump and fan or ventilator.

[0043] Within the scope of the present invention, a vehicle energy system can be understood to be a system comprising at least one fuel cell system and at least one energy storage device, such as a battery.

[0044] Within the scope of the present invention, a predictable time period can be understood as a mathematical formulation that uses variables to describe the length of a planning horizon of a predictive operating strategy.

[0045] Within the scope of the present invention, a horizon length determination module can be understood as a module that is configured to determine, based on information relating to the vehicle energy system, a predictable time duration for an operating strategy of the vehicle energy system.

[0046] Within the scope of the present invention, an operating control module can be understood as a module configured to determine a predictive operating strategy for operating the vehicle energy system based on the determined time duration to be predicted. The horizon length determination module and the operating control module interact with each other and / or transmit information to each other. The operating control module can be an economical model predictive (MPC) controller that optimizes the operation of the fuel cell stack using the degrees of freedom stack current, cathode pressure, and stoichiometry.

[0047] Within the scope of the present invention, a driving condition estimation module can be understood as a module configured to supply predicted driving condition data to the horizon length determination module. The predicted driving condition data can be future data relating to the vehicle, such as an expected speed, a

[0048] - 8 -

[0049] Expected power requirements of the fuel cell system and the like.

[0050] Brief description of the drawings

[0051] Further optional details and features of the invention will become apparent from the following description of preferred embodiments, which are shown schematically in the figures.

[0052] They show:

[0053] Figure 1 shows a schematic representation of a vehicle energy system according to an embodiment of the present invention in a vehicle,

[0054] Figure 2 shows a schematic representation of a result of two different time durations to be predicted, and

[0055] Figures 3A to 3D are exemplary visualizations of the result of two different time durations to be predicted.

[0056] Embodiments of the invention

[0057] Figure 1 shows a schematic representation of a vehicle energy system 100 according to an embodiment of the present invention. The vehicle energy system 100 is shown by way of example arranged in a vehicle 102. The vehicle 102 can be a passenger car or a truck, although other types of vehicles are conceivable in principle.

[0058] The vehicle energy system 100 comprises a fuel cell system 104 with several fuel cell stacks 106, 106'. Each fuel cell stack 106, 106' includes several fuel cells, which are not shown in detail for clarity. Each fuel cell stack 106, 106' is connected to further subsystems. Specifically, each fuel cell stack 106, 106' is connected to an electrical system, an air system, and a water-related system.

[0059] - 9 -

[0060] Each fuel cell stack 106, 106' is connected via the electrical subsystems to at least one electric motor 108 of the vehicle 102 and an energy storage device 110, such as a battery.

[0061] The vehicle energy system 100 also includes a control unit 112. The control unit 112 is designed to control the operation of the fuel cell system 100 or the fuel cell stacks 106, 106'. The control unit 112 can be implemented in a control unit 114 of the vehicle 102.

[0062] The control unit 112 receives information concerning the vehicle energy system 100. This information is provided by at least one information source (not shown in detail) located within and / or outside the vehicle energy system 100. The information includes at least one, and preferably several, pieces of information selected from the group consisting of current and / or predicted values ​​of: ambient temperature, ambient humidity, ambient pressure, fuel cell stack conditions, battery state of charge, battery temperature, aging state of the fuel cell stacks, aging state of the energy storage system, aging state of the drive system, aging state of the components, trailer load, predicted future power demand of the electric motors, vehicle speed, and road gradient.In particular, one or more sensors (not shown) in the vehicle 102 can acquire information about the current environmental and system state, or such information can be transmitted to the control unit 112. The information includes: current ambient temperature, current ambient humidity, current ambient pressure, current fuel cell stack conditions (such as fuel cell stack temperature, air pressure, air mass flows, estimated membrane humidity, and the like), current battery charge level, current battery temperature, current aging state of the fuel cell stacks, current aging state of the battery, current aging state of the drive system, current aging state of the components, and trailer load. R.417020.

[0063] - 10 -

[0064] The control unit 112 determines a predictable operating time for a vehicle energy system strategy based on the acquired information. The control unit 112 also includes a horizon length determination module 116 and an operational management module 118. The horizon length determination module 116 is configured to determine a predictable operating time for a vehicle energy system strategy based on information concerning the vehicle energy system.

[0065] The time duration to be predicted can be determined using at least one characteristic curve. In a characteristic curve-based approach, characteristic curves can be generated in advance for the combinatorics at the inputs, which output tailored values ​​for the time duration to be predicted. The characteristic curves can be generated by estimation or by solving an optimization problem offline.

[0066] Alternatively or additionally, the time duration to be predicted can be determined heuristically. This allows the use of specially designed functions that explicitly model the influence of operating states on the horizon length. An example could be:

[0067]

[0068] Here, TH is the time duration to be predicted, k are parameters, Cheat is the total heat capacity of the cooling system, and T is... env the ambient temperature, p-loss

[0069] r pred dj e predicted power loss, TNormai is a temperature at which the fuel cell system 104 is capable of full load, is T stack is the fuel cell stack temperature. The parameters ki are empirically determined beforehand to design the mathematical function for TH in such a way that a desired predictable time duration is output based on the current operating and environmental conditions. Methods for determination include expert estimation and mathematical curve fitting methods or parameter estimation methods such as "least-squares".

[0070] Alternatively or additionally, the time duration to be predicted can be determined online by solving an optimization problem. The optimization goal is R.417020.

[0071] - 11 -

[0072] The goal is to provide performance without violating boundary conditions. The optimization variable is the time duration TH to be predicted. The same system model used in the predictive operating strategy is employed for this purpose.

[0073] The determination of the predicted time duration TH using a control unit can alternatively also be determined using a cloud server and sent to the vehicle, provided there is good connectivity.

[0074] The horizon length planner can also be extended to include vehicle standstill or standstill or standby of the fuel cell system 104, in order to adaptively trigger or monitor functions that should run in these states, e.g. recovery functions, frost protection measures, parking purge in the anode loop, and to predict the wakeup intervals.

[0075] The operational management module 118 is configured to determine a predictive operating strategy for operating the vehicle's energy system based on the calculated duration to be predicted. The information mentioned above is also fed into the operational management module 118.

[0076] The control unit 112 also includes a driving state estimation module 120. The driving state estimation module 120 is configured to supply predicted driving state data to the horizon length determination module 116. The horizon length determination module 116 is configured to consider the predicted driving state data for determining the predicted duration for an operating strategy of the vehicle energy system. In this way, the horizon length determination module 116 can receive information regarding a predicted future power demand of the electric machines, as well as vehicle speed and road gradient. The driving state estimation module 120 is also configured to supply predicted driving state data to the operational management module 118.

[0077] The horizon length determination module 116 runs on the control unit together with the driving condition estimation module 120 and the operating management module 118. A driving condition estimation module 120 as used in predictive operating strategies not-R.417020

[0078] - 12 -

[0079] While agile, it allows for the prioritization of objective functions over an entire planning period, but is only optional for this invention. The operating strategy governs the vehicle, consisting of the fuel cell system and battery. The measured operating states are reported to the control unit.

[0080] An exemplary sequence of the inventive method for operating the vehicle energy system is described below in general form.

[0081] The aforementioned information concerning the vehicle energy system is collected. Based on this information, a predictable operating time for a vehicle energy system strategy is determined, for example, using characteristic maps as described above, heuristically, or by solving an optimization problem. Based on this predictable operating time, a predictive operating strategy for the vehicle energy system is then developed.

[0082] The vehicle's energy system is determined. Finally, the vehicle's energy system is operated according to the determined predictive operating strategy.

[0083] Figure 2 shows a schematic representation of the result of two different time durations to be predicted. In Figure 2, the upper part of the y-axis represents the electrical power demand P of the fuel cell stacks 106, 106', and the x-axis represents time t. Curve 122 shows the time course of the electrical power P of the fuel cell stacks 106, 106'.

[0084] In the lower part of Figure 2, the temperature T of the fuel cell stacks 106, 106' is plotted on the Y-axis and time t on the X-axis. Curve 124 shows the time course of the temperature T of the fuel cell stacks 106, 106' for a first time period TH1 to be predicted. Curve 126 shows the time course of the temperature T of the fuel cell stacks 106, 106' for a second time period TH2 to be predicted.

[0085] Initially, the fuel cell stack 106, 106' is cold, so the temperature T is low. After a certain time, a high electrical power output P is expected. Only at a sufficient temperature level can a fuel cell stack 106, 106' operate.

[0086] - 13 -

[0087] Retrieve rated power. For the first predicted time period TH1, the increase in power demand P lies outside the planning horizon; consequently, no rapid heating process is initiated. Therefore, the target power cannot be reached at planning horizon TH1, i.e., actual power < target power. Note that the actual power is not shown for clarity, but it is temporarily below the target power after the first predicted time period TH1. For the second predicted time period TH2, the increase in power demand P lies within the planning horizon, and all available heating measures can be implemented in time. Therefore, the target power can be met over the entire trajectory, i.e., actual power = target power.

[0088] Figures 3A to 3D show exemplary visualizations of the result of two different time durations to be predicted.

[0089] Figure 3A shows the electrical power demand P of a first fuel cell stack 106 and a second fuel cell stack 106' on the Y-axis and time t on the X-axis. Curve 128 represents the time course of the electrical power demand P. Curve 129 shows the time course of the 50% orientation of the electrical power demand P, illustrating a deviation from a uniform power distribution between the first fuel cell stack 106 and the second fuel cell stack 106'. Curve 130 shows the time course of the electrical power P of the first fuel cell stack 106, and curve 132 shows the time course of the electrical power P of the second fuel cell stack 106', each for a predicted time TH of 25 s.

[0090] Figure 3B shows the electrical power demand P of a first fuel cell stack 106 and a second fuel cell stack 106' on the Y-axis and time t on the X-axis. Curve 128 shows the time course of the electrical power demand P. Curve 130 shows the time course of the electrical power demand.

[0091] - 14 -

[0092] The curve P of the first fuel cell stack 106 and the curve 132 indicate the time course of the electrical power P of the second fuel cell stack 106' each for a predictable time period TH of 60s.

[0093] Figure 3C shows the temperature T of the first fuel cell stack 106 and the second fuel cell stack 106' on the Y-axis and time t on the X-axis. Curve 134 shows the temperature T of the first fuel cell stack 106 over time, and curve 136 shows the temperature T of the second fuel cell stack 106' over time, each for the predicted time TH of 25 s.

[0094] In Figure 3D, the Y-axis represents the temperature T of the first fuel cell stack 106 and the second fuel cell stack 106', and the X-axis represents time t. Curve 134 shows the time course of the temperature T of the first fuel cell stack 106, and curve 136 shows the time course of the temperature T of the second fuel cell stack 106', each for the predicted time period TH of 60 s.

[0095] Figures 3A to 3D provide an exemplary visualization of the result for two different horizon lengths TH. Figures 3A and 3B show an example power requirement P of a so-called multi-stack fuel cell system, i.e., a fuel cell system with more than one fuel cell stack, as indicated by curve 128, and how this power P is distributed between two single-stack fuel cell systems by the predictive operating strategy, as indicated by the stacked areas or curves 130 and 132. Figures 3C and 3D show the temperature profiles of the fuel cell stacks 106 and 106' of the individual fuel cell systems.

[0096] Figures 3A and 3C, as mentioned, depict the aforementioned profiles of electrical power P and temperature T for a horizon length of 25 s. Area 138 represents this period starting at t0 = 5 s. Not included in this horizon length is the electrical power requirement P of 180 kW for ei-R.417020.

[0097] - 15 -

[0098] At time t = 40s, the predictive operating strategy with power distribution and thermal system control cannot become proactive and cannot take suitable measures to meet the power requirement.

[0099] Figures 3B and 3D, as mentioned, depict the aforementioned profiles of electrical power P and temperature T for a horizon length of 60 s. Area 140 represents this period from t0 = 5 s. Since the 180 kW stage of the electrical power requirement P is included in this case, the operating strategy, due to the longer predicted time period compared to Figures 3A and 3C, can become proactive with power distribution and thermal system control and take appropriate measures to meet the power requirement.

Claims

R.417020 - 16 - Claims 1. Method for operating a vehicle energy system (100) with a fuel cell system (104) with multiple fuel cell stacks (106, 106') and with at least one energy storage device (110) for storing electrical energy, comprising the steps: Gathering information concerning the vehicle energy system (100), Determining a predictable time duration (TH) for an operating strategy of the vehicle energy system (100) based on the acquired information, Determining a predictive operating strategy for operating the vehicle energy system (100) based on the determined predictable time duration (TH), and Operating the vehicle energy system (100) according to the determined predictive operating strategy.

2. Method according to the preceding claim, wherein the information comprises at least one piece of information and preferably several pieces of information selected from the group consisting of current and / or predicted values ​​of: ambient temperature, ambient humidity, ambient pressure, fuel cell stack conditions, battery state of charge, battery temperature, aging state of the fuel cell stacks, aging state of the energy storage system, aging state of the drive system, aging state of the components, trailer load, predicted future power requirement of the electric machines, driving speed, road gradient.

3. A method according to any of the preceding claims, wherein the time duration (TH) to be predicted is determined by means of at least one characteristic map, heuristically and / or by online solving of an optimization problem. R.417020 - 17 - 4. Method according to one of the preceding claims, wherein the time duration (TH) to be predicted is determined by means of a control unit on board a vehicle or wherein the time duration (TH) to be predicted is determined by means of a cloud server and the determined time duration (TH) to be predicted is transmitted to a vehicle.

5. Method according to one of the preceding claims, wherein the method also takes place when a vehicle is stationary and / or the fuel cell system (104) is in standby mode.

6. Method according to one of the preceding claims, wherein the time period (TH) to be predicted is at least as long as the duration of a measure to be set by the operating strategy.

7. Vehicle energy system (100) comprising a fuel cell system (104) comprising multiple fuel cell stacks (106, 106') and comprising at least one control unit (112), wherein the control unit (112) is configured to perform a method according to one of the preceding claims.

8. Vehicle energy system (100) according to the preceding claim, wherein the control unit (112) comprises a horizon length determination module (116) and an operational management module (118), wherein the horizon length determination module (116) is configured to determine a predictable time period (TH) for an operating strategy of the vehicle energy system (100) based on information relating to the vehicle energy system (100), wherein the operational management module (118) is configured to determine a predictive operating strategy for operating the vehicle energy system (100) based on the determined predictable time period (TH).

9. Vehicle energy system (100) according to the preceding claim, wherein the control unit (112) further comprises a driving state estimation module (120), wherein the driving state estimation module (120) is configured to provide the horizon length determination module (116) with predicted driving state data. - 18 - to supply ten, wherein the horizon length determination module (116) is configured to take into account the predicted driving state data for determining the time duration (TH) to be predicted for an operating strategy of the vehicle energy system (100).

10. Vehicle (102) comprising a vehicle energy system (100) according to any one of claims 7 to 9.

11. Computer program product comprising program code means which, when the computer program product is executed on a computer, configure the computer to perform a method according to any one of claims 1 to 6.

Images