Flatness-based control considering the limitations of the input range.

Abstract

Controlled object in the form of a partial system of the refrigerant system for the operation of an electrochemical energy converter (Σ p Regarding a method for operating (·)) in a model-based manner, particularly for control, the method is -Controlled object (Σ p (·)) Actuator state (m actuator (k)) is the actuator's control amount limit (u req,min (k) / u req.max (k)) should be taken into consideration when making the decision, - the determined state of the actuator (m actuator (k)) depends on the controlled object (Σ p It has the ability to activate (·)).

Description

[Technical Field]

[0001] The present invention is advantageously related to a method for controlling, and more particularly for controlling, a controlled object for the operation of an electrochemical energy converter, preferably in the form of a fuel cell system or an electrolytic cell, and more preferably in the form of a component system of a refrigerant system, on a model basis. Furthermore, the present invention relates to a corresponding computer program product, a corresponding controller, and a corresponding electrochemical energy converter. [Background technology]

[0002] In known flatness-based control, the limitations of the actuator are not systematically considered in the control concept. As a result, if the calculated manipulated variable is larger or smaller than the manipulated variable that the actuator can provide, the calculated variable is simply additionally limited when the series of calculations is complete. However, this leads to incorrect calculations of the state variables (storage in software) of the controlled model performed before this additional limitation, and the longer the actuator remains in the manipulated variable limitation state, the greater the error becomes over time (i.e., the state variables of the controlled model drift from the system state of the controlled object). Later, when the actuator moves away from the manipulated variable limitation again, the state variables of the controlled model that have drifted from that point negatively affect the control behavior. [Overview of the project]

[0003] The present invention intends to provide a method, particularly for controlling, a control target for operating an electrochemical energy converter, preferably in the form of a fuel cell system or an electrolytic cell, in a model-based manner, comprising the constituent elements of an independent claim relating to a method. In addition, the present invention intends to provide a corresponding computer program, a corresponding controller, and a corresponding energy converter, comprising the constituent elements of a parallel claim. The constituent elements and details described in relation to different embodiments and / or aspects of the present invention are of course applicable in relation to other embodiments and / or aspects, and vice versa, and as a result, the disclosures for individual embodiments and / or aspects are always, or may be, related to each other.

[0004] In a first aspect, the present invention is intended to provide a method (hereinafter also referred to as a control method) for controlling, in particular, a model-based operation of a control object for the operation of an electrochemical energy converter, preferably in the form of a fuel cell system or an electrolytic cell, and more preferably in the form of a subset of a refrigerant system.

[0005] This method has the following method steps / actions. -Controlled object Σ p (·) Actuator state m actuator (k) The limiting amount of the actuator u req,min (k) / u req.max (k) should be taken into consideration when making the decision. - The determined state of the actuator m actuator (k) The controlled object Σ p To activate (·).

[0006] In this way, a flatness-based control concept can be provided that systematically considers the limitations of the actuators in the control concept. This control concept can be applied to the components of the refrigerant circulation system of an electrolytic cell in an advantageous manner.

[0007] Actuators within the scope of the present invention may include, for example, pumps, valves, and the like.

[0008] This control method can be used to control various targets Σ in an advantageous manner. p (·) and / or various controlled models Σ m Can be used with (·).

[0009] A refrigerant system can be understood as any temperature control system that uses a cooling fluid or refrigerant to enable temperature control of an electrochemical energy converter, particularly in the form of a fuel cell stack or electrolytic cell in a fuel cell system.

[0010] Refrigerants can generally be called temperature regulators.

[0011] In many operating modes of fuel cell systems, such as during freeze-start and / or cold-start, a refrigerant system may be used to heat the fuel cell stack to the appropriate operating temperature. Similarly, in other operating modes of fuel cell systems, such as during standard operation or high-load operation, a refrigerant system may be used to cool the fuel cell stack to the appropriate operating temperature.

[0012] A refrigerant system within the scope of this disclosure may have a refrigerant pump for circulating the refrigerant. In addition, a refrigerant system may have a refrigerant bypass valve (for example, a three-way valve, a so-called mixing point of the refrigerant system) for directing a portion of the refrigerant to pass through a cooler (e.g., a radiator) and / or for directing the other portion to pass alongside the cooler.

[0013] Control target Σ of the refrigerant system p (·) may be, for example, a mixing point where the refrigerant inthe cooler conduit and a bypass cooling conduit are mixed.

[0014] Control target Σ of the refrigerant system p (·) may be, for example, a refrigerant passage through a fuel cell stack of a fuel cell system.

[0015] Control target Σ of the refrigerant system p (·) may be, for example, a refrigerant passage through a cooler.

[0016] In an advantageous embodiment, if the actuator (i.e., the manipulator) is exactly in a saturated state (i.e., the limiting mode), the limiting manipulated variable u lim (k) (i.e., the saturation value at each point of the manipulated variable), the control target model, and the inverse trajectory generator

[0017]

Equation

[0018] This approach allows for the systematic consideration of manipulated variable limits in a flatness-based control concept. As a result, if the actuator is saturated, or if it was saturated, it will not become a state variable of the controlled model that has drifted from that saturated state.

[0019] As described below, further advantages can be provided regarding specific application examples in fuel cell stacks (so-called fuel cell stacks). In the case of fuel cell stacks, the requirements are that the refrigerant intake temperature and the refrigerant temperature difference (i.e., the temperature difference between the refrigerant discharge temperature and the refrigerant intake temperature) must be controlled. Since the refrigerant temperature difference is not a flat output, the requirement to control the refrigerant temperature difference cannot be directly achieved with a control concept based on flatness. Therefore, in an advantageous embodiment, we propose selecting the refrigerant discharge temperature as the control variable instead of the refrigerant temperature difference, and calculating a set value signal for the refrigerant discharge temperature of the fuel cell stack based on a set value signal or limited set value signal for the refrigerant intake temperature and a set value signal for the refrigerant temperature difference.

[0020] Furthermore, the specific state of the actuator m actuator (k) Depends on the controlled object Σ p It may be considered to provide at least two modes for operating (·). That is, - Actuator requests manipulated amount u req (k) can be provided in unlimited quantities in an unrestricted (or free) mode m actuator,free and, - The actuator controls the requested operating amount u req (k) is limited by at least one controllable variable u req,min (k) / u req,max (k) Limited modes m that can be provided restrictively actuator,min / m actuator,max and, That is the case.

[0021] Below, Σ m,aug(·) represents the extended (dynamically nonlinear) controlled object model. The extended (dynamically nonlinear) controlled object model has the required compensation / modification term d req (k) (e.g., additional terms, multiplicative terms, etc.) is modeled Σ m By inserting it into one of the multiple equations of (·), the (dynamic nonlinear) controlled model Σ m This is obtained by extending (·). In this case, the requested compensation term / modification term d req (k) represents the intervention caused by the controller. In Figure 1a, block

[0022]

number

[0023]

number

[0024] However, the controlled object Σ p (·) and the controlled model Σ m (·) and the extended control target model Σ m,aug In general, (·) represents / explains a function represented by a dynamic nonlinear function. The symbol "(·)" is used to indicate that the nonlinear function depends on a number of input arguments, which are cited only when necessary for clarity to understand the concept.

[0025] In the following, the symbol Σ m,aug (·) represents a mathematical formulation of the (dynamic nonlinear) extended control target model, and the modeled control target output y m (k) (i.e., the value of the controlled variable y(k) calculated through the model) is calculated as a function of the manipulated variable u(k), and the calculated and / or measured (and therefore known) disturbance z mThis calculates (k) and then calculates the estimated deviation term / error term d(k), that is, y m =Σ m,aug (u,z m d) That is the case.

[0026] symbol

[0027]

number

[0028]

number

[0029]

number

[0030] Preferably, the extended control target model Σ m,aug (·) is a model equation that models the physical functional balance and controls Σ p (·) Controlled model Σ m This can be obtained by extending (·). Therefore, the controlled model Σ m For almost any of the model equations in (·), it is possible to estimate / calculate the deviation term / error term d(k) in an advantageous manner. The deviation term / error term d(k) is an internal model quantity that can be used in an advantageous manner with respect to the controlled model Σ m (·) can be expanded (for example, as an additional term, multiplicative term, etc.).

[0031] Furthermore, non-limited mode mactuator,free The following steps should be considered. -Setting value of the controlled variable y(k) y set (k) step, - Desired value y of the controlled variable y(k) des (k) and the controlled object Σ p (·) Extended control target model Σ m,aug Desired values ​​y up to the system order n of (·) des Time derivative of (k)

[0032]

number

[0033]

number

[0034]

number

[0035] In this case, the actuator is not in a saturated state. Guidance signal / setpoint signal y set (k) is the orbital generator Σ desEnter the (·) (for example, if the system order of the controlled model is n=1, a PT1 filter may be used as the orbital generator). Orbital generator Σ des The purpose of (·) is the so-called desired control quantity y des (k) (desired value of the controlled variable y(k)) and the time derivative

[0036]

number

[0037]

number

[0038]

number

[0039]

number

[0040]

number

[0041]

number

[0042]

number

[0043] Furthermore, the so-called limited set value y set,lim (k) is set to be equal to the set value y set (k). The limited set value y set,lim (k) is required by the switching logic in a switching device that performs switching between calculations in the non-limited mode and calculations in the limited mode.

[0044] Furthermore, it may be considered to perform the following steps in the limited mode m actuator,min / m actuator,max . - Step of determining the limited operation amount u lim (k) of the actuator, - Desired value y des (k) of the control quantity y(k), and the extended control object model Σ p (·) of the control object Σ m,aug Desired value y des (k) up to the system order n of the system Σ <​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​ A step of determining - the required operation amount u of the actuator req, (k) to be equal to the limited operation amount u lim (k) and setting it.

[0048] In this case, the actuator (i.e., the manipulator) is in a saturated state. Depending on whether the actuator is at the upper limit stop or the lower limit stop (i.e., in the upper saturation state or the lower saturation state), the operation amount at the upper limit stop (i.e., the maximum operation amount) u req,max (k), or the operation amount at the lower limit stop (i.e., the minimum operation amount) u req,min (k) corresponding values (i.e., their respective saturation values) are supplied into the model Σ lim (·) as the so-called limited operation amount u m,aug (k). Thereby, the inverse trajectory generator

[0049]

Number

[0050]

Number

[0051] In addition, the operation of the controlled object Σ p (·) is in the first mode m actuator,free and the second mode m actuator,min / m actuator,maxTo switch between the two states, it may be considered to provide a switching device (which can also be called a state automaton). On the one hand, the switching device sets a limiting set value y of the controlled variable y(k). set,lim (k) is the set value y of the controlled variable y(k) set (k) is compared with this, and depending on this comparison, the limited mode m actuator,min / m actuator,max Leaving the unrestricted mode actuator,free It can be used to determine whether it can be switched to. On the other hand, the switching device controls the required operating amount u of the actuator. req (k) Limit the controllable quantity u req,min (k) / u req.max (k) is compared with this, and depending on this comparison, the unrestricted mode m actuator,free Limited mode m actuator,min / m actuator,max It can be used to determine whether or not a replacement is necessary.

[0052] In that case, the switching device or state automaton will control the mode m of the operator. actuator (k) can be determined. A state automaton can handle multiple states: "actuator in lower saturated state", "actuator not saturated", and "actuator in upper saturated state". A state automaton can start in the state "actuator in lower saturated state". In this state, the mode m of the actuator actuator (k) to m actuator,min Set it to be equal to, in this case, m actuator,min This indicates that the actuator is currently in a lower saturation state. While the state automaton is in the state "actuator in lower saturation state", at any time step k, the limited set value y set,lim (k) is the set value y set The condition is checked to see if it is greater than (k). As long as this condition is not met, the state automaton remains in the state "actuator in lower saturated state". However, as soon as this condition is met, the state automaton switches to the state "actuator not saturated". In this state, the mode m of the actuator actuator (k) to m actuator,free Set it to be equal to, in this case mactuator,free This indicates that the actuator is not currently saturated. While the state automaton is in the state "operator not saturated", at any time step k, two conditions are checked: if the "requested" actuator u req (k) is its (current) maximum value u req,max If (k) is greater than or equal to (k), the state automaton switches to the state "operator in upper saturated state". req (k) is its (current) minimum value u req,min If (k) is less than or equal to (k), the state automaton switches to the state "actuator in lower saturated state". If neither of the two conditions is met, the state automaton remains in the state "actuator not saturated". In the state "actuator in upper saturated state", the mode m of the actuator actuator (k) to m actuator,max Set it to be equal to, in this case m actuator,max This indicates that the actuator is currently in an upper saturated state. While the state automaton is in the state "operator in upper saturated state", at any time step k, the limited set value y set,lim (k) is the set value y set The condition is checked to see if it is less than (k). As long as this condition is not met, the state automaton remains in the state "operator in upper saturated state". However, as soon as this condition is met, the state automaton immediately switches to the state "operator not saturated".

[0053] On the other hand, this method controls the Σ in the form of a fuel cell stack. p It is conceivable that (·) could be used, in particular for control, to operate it. In this case, advantageously, the refrigerant discharge temperature from the fuel cell stack can be used as the control variable y(k), in particular to control the refrigerant temperature difference by the fuel cell stack.

[0054] In addition, this method controls the morphology of the mixing point of the refrigerant system Σ pIt is conceivable that (·) may be used, in particular, for control purposes. In an advantageous embodiment, the temperature at which the refrigerant is introduced into the fuel cell stack may be used as the control variable y(k).

[0055] Furthermore, this method can be applied to the controlled object Σ in the form of a cooler (or generally a heat exchanger) in a refrigerant system. p It is conceivable that (·) may be used, in particular, for control purposes. In an advantageous embodiment, the refrigerant discharge temperature from the cooler may be used as the control quantity y(k).

[0056] Therefore, this method makes it possible to advantageously control the refrigerant system, and more advantageously the refrigerant system for the operation of an electrochemical energy converter, on a model basis.

[0057] Furthermore, it may be advantageous that this method can be used specifically for controlling a controlled object that has multiple input quantities and / or multiple manipulated quantities and / or multiple controlled quantities.

[0058] In another view, the present invention provides a computer program product that includes instructions to cause a computer to execute the method described above when the computer program product is executed by the computer. Using this computer program product, the same advantages as described above in relation to the method according to the present invention can be achieved. These advantages are again fully referenced here.

[0059] The corresponding controller provides another aspect of the present invention. A computer program in the form of code can be filed within the controller's memory unit, and the code implements the method as described above when the controller's arithmetic unit executes the code. Using this controller, the same advantages as those described above in relation to the method according to the present invention can be achieved. These advantages are again fully referenced here.

[0060] A corresponding electrochemical energy converter equipped with a corresponding controller provides another aspect of the present invention. Using this energy converter, the same advantages described above can be achieved in connection with the method according to the present invention. These advantages are again fully referenced here.

[0061] Next, the present invention, its further configurations, and its advantages will be described in more detail based on the drawings. The drawings are schematic in nature. [Brief explanation of the drawing]

[0062] [Figure 1a] This block diagram shows the basic calculations for systematically considering limitations on actuators (i.e., devices) that are not saturated (i.e., in "unlimited mode") in a flatness-based control concept. [Figure 1b] This block diagram shows the basic calculations for systematically considering the limitations on actuators that are saturated (i.e., in a "limited mode") in a flatness-based control concept. [Figure 2] This figure shows a state automaton that determines the mode mactuator(k) of the actuator used for switching between the calculation in Figure 1a (performed if the actuator is not saturated) and the calculation in Figure 1b (performed if the actuator is in an upper or lower saturated state). [Modes for carrying out the invention]

[0063] The "(·)" in the blocks in Figures 1a, 1b, and 2 are omitted to save space (e.g., Σ). p (·) is simplified to Σ p (To be written as follows).

[0064] Figures 1a, 1b, and 2 are used to illustrate, in particular a method for controlling (hereinafter sometimes referred to as a control method) a controlled object, preferably in the form of a subset system of a refrigerant system, for the operation of an electrochemical energy converter, for example, a fuel cell system or an electrolytic cell, on a model basis.

[0065] This method has the following method steps / actions. -Controlled object Σ p (·) Actuator state m actuator (k) is the actuator's control amount limit u req,min (k) / u req.max (k) should be taken into consideration when making the decision. - Determined actuator state m actuator (k) Depends on the controlled object (Σ p To activate (·)), It has.

[0066] This method provides a flatness-based control concept that systematically considers the limitations of the actuators within the control concept. This control concept is advantageous because it can be applied to the components of the refrigerant circulation system of an electrolytic cell.

[0067] Actuators within the scope of the present invention may include, for example, pumps, valves, and the like.

[0068] This control method applies to various control targets Σ p (·) and / or various controlled models Σ m It can be used in a favorable manner together with (·).

[0069] A refrigerant system can be understood as any temperature control system that uses a cooling fluid or refrigerant to enable temperature control of an electrochemical energy converter, particularly in the form of a fuel cell stack or electrolytic cell in a fuel cell system.

[0070] Refrigerants can generally be called temperature regulators.

[0071] In many operating modes of a fuel cell system, such as during freeze-start and / or cold-start, a refrigerant system may be used to heat the fuel cell stack to the appropriate operating temperature. Similarly, in other operating modes of a fuel cell system, such as during standard operation or high-load operation, a refrigerant system may be used to cool the fuel cell stack to the appropriate operating temperature.

[0072] A refrigerant system within the scope of this disclosure may have a refrigerant pump for circulating the refrigerant. In addition, a refrigerant system may have a refrigerant bypass valve (for example, a three-way valve, a so-called mixing point of the refrigerant system) for directing a portion of the refrigerant to pass through a cooler (e.g., a radiator) and / or for directing the other portion to pass alongside the cooler.

[0073] Controlled object of the refrigerant system Σ p (·) may be, for example, a mixing point where the refrigerant in the cooler passage mixes with the refrigerant in the bypass passage.

[0074] Controlled object of the refrigerant system Σ p (·) may be, for example, a refrigerant passage that penetrates the fuel cell stack of a fuel cell system.

[0075] Controlled object of the refrigerant system Σ p (·) may be, for example, a refrigerant passage through a cooler.

[0076] In a favorable configuration, if the actuator is in a saturated state (i.e., limited mode), then the limited controllable amount u lim (k) (i.e., the saturation value of the manipulated variable at each step), based on the model equations of the controlled object model and the orbit generator, a limiting set value y set,lim (k) can be calculated (see Figure 1b). Limited setting value y set,lim (k) can represent the limited control variable u lim (k) using (for example, the maximum value of the manipulated variable ureq,max (Using k), this refers to what setting value is achieved. In particular, a limited setting value y is preferred. set,lim (k) and the set value y set A switching device (see Figure 2) may be provided to determine when it is possible to switch from the limited mode to the unlimited mode based on a comparison with (k). Furthermore, the switching device may be provided in particular, preferably, with respect to the requested manipulated quantity u req (k) and its maximum value u req,max (k) or its minimum value u req,min This may be performed to determine when to switch from unrestricted mode to restricted mode based on a comparison with (k). Note that in unrestricted mode, a different calculation (see block diagram in Figure 1a) is performed than the calculation in restricted mode (see block diagram in Figure 1b).

[0077] This approach allows for the systematic consideration of manipulated variable limitations in flatness-based control concepts. As a result, if the actuator is saturated, or if it was saturated, it will not become a state variable of the controlled model that has drifted from that saturated state.

[0078] As described below, further advantages can be provided regarding specific application examples in fuel cell stacks (so-called fuel cell stacks). In the case of fuel cell stacks, the requirements are that the refrigerant intake temperature and the refrigerant temperature difference (i.e., the temperature difference between the refrigerant discharge temperature and the refrigerant intake temperature) must be controlled. Since the refrigerant temperature difference is not a flat output, the requirement to control the refrigerant temperature difference cannot be directly achieved with a control concept based on flatness. Therefore, we propose that, in an advantageous manner, the refrigerant discharge temperature is selected as the control variable instead of the refrigerant temperature difference, and that a set value signal for the refrigerant discharge temperature of the fuel cell stack is calculated based on a set value signal or limited set value signal for the refrigerant intake temperature and a set value signal for the refrigerant temperature difference.

[0079] As suggested by Figures 1a and 1b, the identified state m of the actuator actuator(k) Depends on the controlled object Σ p There can be at least two modes for operating (·). That is, - Actuator requests manipulated amount u req (k) can be provided in unlimited quantities in an unrestricted (or free) mode m actuator,free and, - The actuator controls the requested operating amount u req (k) is limited by at least one controllable variable u req,min (k) / u req,max (k) Limited modes m that can be provided restrictively actuator,min / m actuator,max and, It is possible to provide this.

[0080] If the actuator is not currently in a saturated state, the calculation shown in the block diagram of Figure 1a is performed. However, if the actuator is currently at the upper limit stop or lower limit stop (i.e., upper saturation or lower saturation), the calculation shown in the block diagram of Figure 1b is performed. The switching logic for switching between the calculations in Figure 1a and Figure 1b will be discussed later (see Figure 2). First, let's explain Figure 1.

[0081] Below, Σ m,aug (·) represents the extended (dynamically nonlinear) controlled object model. The extended (dynamically nonlinear) controlled object model has the required compensation / modification term d req (k) (e.g., additional terms, multiplicative terms, etc.) is modeled Σ m By inserting it into one of the multiple equations of (·), the (dynamic nonlinear) controlled model Σ m This is obtained by extending (·). In this case, the requested compensation term / modification term d req (k) represents the intervention caused by the controller. In Figure 1a, block

[0082]

number

[0083]

number

[0084] However, the controlled object Σ p (·) and the controlled model Σ m (·) and the extended control target model Σ m,aug In general, (·) represents / explains a function represented by a dynamic nonlinear function. The symbol "(·)" is used to indicate that the nonlinear function depends on a number of input arguments, which are cited only when necessary for clarity to understand the concept.

[0085] In the following, the symbol Σ m,aug (·) represents a mathematical formulation of the (dynamic nonlinear) extended control target model, and the modeled control target output y m (k) (i.e., the value of the controlled variable y(k) calculated through the model) is calculated as a function of the manipulated variable u(k), and the calculated and / or measured (and therefore known) disturbance z m This calculates (k) and then calculates the estimated deviation term / error term d(k), that is, y m =Σ m,aug (u,z m d) That is the case.

[0086] symbol

[0087]

number

[0088]

number

[0089]

number

[0090] Preferably, the extended control target model Σ m,aug (·) represents the controlled object Σ in the model equation that models the physical functional balance. p (·) Controlled model Σ m This is obtained by extending (·). Therefore, the controlled model Σ m For almost any of the model equations in (·), it is possible to estimate / calculate the deviation term / error term d(k) in an advantageous manner. The deviation term / error term d(k) is an internal model quantity of the controlled model Σ m (·) can be expanded in an advantageous manner (for example, as an additional term, a multiplicative term, etc.).

[0091] As suggested by Figure 1a, the unrestricted mode m actuator,free The following steps can then be taken. -Setting value of the controlled variable y(k) y set (k) step, - Desired value y of the controlled variable y(k) des (k) and the controlled object Σ p (·) Extended control target model Σ m,aug Desired values ​​y up to the system order n of (·) des (k) time derivative

[0092]

number

[0093]

number

[0094]

number

[0095] Next, we will explain Figure 1a (showing calculation in non-restrictive mode) in detail. Here, the actuator is not in a saturated state. Guidance signal / setpoint signal y set (k) is the orbital generator Σ des Enter the (·) (for example, if the system order of the controlled model is n=1, a PT1 filter may be used as the orbital generator). Orbital generator Σ des The purpose of (·) is the so-called desired control quantity y des (k) (desired value of the controlled variable y(k)) and the time derivative

[0096]

number

[0097]

number

[0098]

number

[0099]

number

[0100]

number

[0101]

number

[0102]

number

[0103] Furthermore, the so-called limited setting value y set,lim (k) is set to the value y set Set it to be equal to (k). Limited setting value y set,lim (k) is required by the switching logic within the switching device (see Figure 2) that switches between calculation in unrestricted mode (Figure 1a) and calculation in restricted mode (Figure 1b).

[0104] As suggested by Figure 1b, limited mode m actuator,min / m actuator,max The following steps can then be taken. -Limited operating amount of actuator ulim (k) step, - Desired value y of the controlled variable y(k) des (k) and the controlled object Σ p (·) Extended control target model Σ m,aug Desired values ​​y up to the system order n of (·) des (k) time derivative

[0105]

number

[0106]

number

[0107]

number

[0108] Next, we will explain Figure 1b (showing the calculation in limited mode) in detail. Here, the actuator is in a saturated state. Depending on whether the actuator is at the upper limit stop or the lower limit stop (i.e., whether it is in an upper saturated state or a lower saturated state), the manipulated amount at the upper limit stop (i.e., the maximum manipulated amount) u req,max(k) or the manipulated amount (i.e., minimum manipulated amount) u at the lower limit stop req,min The corresponding value of (k) (i.e., the respective saturation values) is the so-called limiting variable u lim (k) is the model Σ m,aug It supplies to the (·) inside. This creates the reverse orbit generator.

[0109]

number

[0110]

number

[0111] In addition, the controlled object Σ p The operation of (·) is in the first mode m actuator,free and the second mode m actuator,min / m actuator,max To switch between the two states, it may be considered to provide a switching device (which can also be called a state automaton). The switching device 10, on the one hand, sets a limited set value y of the controlled variable y(k). set,lim (k) is the set value y of the controlled variable y(k) set (k) is compared with this, and depending on this comparison, the limited mode m actuator,min / m actuator,max Leaving the unrestricted mode actuator,free It can be used to determine whether it can be switched to. On the other hand, the switching device 10 controls the requested operating amount u of the actuator. req (k) Limit the controllable quantity u req,min (k) / ureq.max (k) is compared with this, and depending on this comparison, the unrestricted mode m actuator,free Limited mode m actuator,min / m actuator,max It can be used to determine whether or not a replacement is necessary.

[0112] As shown in Figure 2, the switching device or state automaton uses the mode m of the actuator to switch between the calculation in Figure 1a and the calculation in Figure 1b. actuator (k) can be determined. The state automaton in Figure 2 can handle multiple states: "actuator in lower saturated state", "actuator not saturated", and "actuator in upper saturated state". The state automaton can start in the state "actuator in lower saturated state". In this state, the mode m of the actuator actuator (k) to m actuator,min Set it to be equal to, in this case, m actuator,min This indicates that the actuator is currently in a lower saturation state. While the state automaton is in the state "actuator in lower saturation state", at any time step k, the limited set value y set,lim (k) is the set value y set The condition is checked to see if it is greater than (k). As long as this condition is not met, the state automaton remains in the state "actuator in lower saturated state". However, as soon as this condition is met, the state automaton switches to the state "actuator not saturated". In this state, the mode m of the actuator actuator (k) to m actuator,free Set it to be equal to, in this case m actuator,free This indicates that the actuator is not currently saturated. While the state automaton is in the state "operator not saturated", at any time step k, two conditions are checked: if the "requested" actuator u req (k) is its (current) maximum value u req,max If (k) is greater than or equal to (k), the state automaton switches to the state "operator in upper saturated state". req (k) is its (current) minimum value u req,minIf (k) is less than or equal to (k), the state automaton switches to the state "actuator in lower saturated state". If neither of the two conditions is met, the state automaton remains in the state "actuator not saturated". In the state "actuator in upper saturated state", the mode m of the actuator actuator (k) to m actuator,max Set it to be equal to, in this case m actuator,max This indicates that the actuator is currently in an upper saturated state. While the state automaton is in the state "operator in upper saturated state", at any time step k, the limited set value y set,lim (k) is the set value y set The condition is checked to see if it is less than (k). As long as this condition is not met, the state automaton remains in the state "operator in upper saturated state". However, as soon as this condition is met, the state automaton immediately switches to the state "operator not saturated".

[0113] Mode m of the control unit actuator (k) is m actuator,min or m actuator,max If it is equal to (if the state of the state automaton in Figure 2 is "operator in lower saturated state" or "operator in upper saturated state"), the calculation is performed as shown in the block diagram in Figure 1b. In this case, m actuator (k) is m actuator,min In cases equivalent to this, the limiting variable u lim (k) is the minimum controllable amount u req,min Set it to be equal to (k). actuator (k) is m actuator,max In cases equivalent to this, the limiting variable u lim (k) is the maximum controllable amount u req,max Set it to be equal to (k). However, the mode of the actuator is m actuator (k) is m actuator,free If the result is equal to (i.e., the state of the state automaton in Figure 2 is in the "operator not in a saturated state"), the calculation is performed as shown in the block diagram in Figure 1a.

[0114] The corresponding computer program, the corresponding controller 100, and the corresponding energy converter (not shown in their entirety for convenience) also represent the aspects of the present invention.

[0115] The above description of embodiments is intended to illustrate the present invention solely within the framework of examples. Of course, the individual components of these embodiments may be combined with each other as they are technically meaningful, without departing from the scope of the present invention. [Explanation of symbols]

[0116] 10. Switching device 100 controllers Σ p (•) Controlled object Σ m,aug (•) Models subject to extended control m actuator (k) Status of the actuator m actuator,min / m actuator,max Limited Mode m actuator,free Unlimited Mode u lim (k) Limited manipulated variable u req (k) Requested operation amount u req,min (k) / u req.max (k) Limitation of controllable amount y(k) Controlled variable y des (k) Desired value of the controlled variable y set (k) Set value of the controlled variable y set,lim (k) Limiting setting of the controlled variable

Claims

1. Advantageously, the control target (Σ) is preferably in the form of a partial system of the refrigerant system for the operation of the electrochemical energy converter. ｐ In a method for operating (•) in a model-based manner, particularly for control, - The controlled object (Σ ｐ (・)) Actuator state (m ａｃｔｕａｔｏｒ (k)) the limit of the actuator's operating amount (u ｒｅｑ，ｍｉｎ (k) / u ｒｅｑ．ｍａｘ (k)) should be taken into consideration when making a decision. - The determined state of the actuator (m ａｃｔｕａｔｏｒ (k)) depends on the controlled object (Σ ｐ Activate (・) A method that has the following characteristics.

2. Depending on the specified state (m ａｃｔｕａｔｏｒ (k)) of the actuator, at least two modes for operating the controlled object (Σ ｐ (·)) are provided, that is, - The actuator is controlled by the requested amount (u ｒｅｑ (k)) provides unlimited access in an unrestricted mode (m ａｃｔｕａｔｏｒ，ｆｒｅｅ ), - The actuator controls the requested amount of operation (u ｒｅｑ (k)) at least one controllable variable limit (u ｒｅｑ，ｍｉｎ (k) / u ｒｅｑ，ｍａｘ (k)) Limited modes (m) that can be provided restrictively ａｃｔｕａｔｏｒ，ｍｉｎ／ I understand ａｃｔｕａｔｏｒ，ｍａｘ ), The method according to claim 1, characterized by providing a provision.

3. The aforementioned non-limiting mode (m ａｃｔｕａｔｏｒ，ｆｒｅｅ ) The following steps should be taken, namely, - Set value of the controlled variable (y(k)) (y ｓｅｔ (k)) step, - Desired value (y(k)) of the control variable (y(k)) ｄｅｓ (k)) and the controlled object (Σ ｐ Extended control target model of ( Σ ) ｍ，ａｕｇ The desired value (y) up to the system order (n) of (•) ｄｅｓ Time derivative of (k) [Math 1] and The set value (y(k)) of the controlled amount (y(k)) ｓｅｔ (k)) depends on, Trajectory generator (Σ ｄｅｓ Using (・)), The decision-making step, - The required amount of operation of the actuator (u ｒｅｑ (k)) The desired value (y(k)) of the controlled quantity (y(k)) ｄｅｓ (k) and the aforementioned desired value (y ｄｅｓ (k) the time derivative [Math 2] Depending on, The controlled object (Σ ｐ (・)) Inverse extension control target model [Math 3] Using The decision-making step, - The limited setting value (y(k)) of the control variable (y(k)) ｓｅｔ，ｌｉｍ (k)) the set value (y) determined by the control amount (y(k)) ｓｅｔ Steps to set it to be equal to (k), The method according to claim 2, characterized by carrying out the following.

4. The aforementioned limited mode (m ａｃｔｕａｔｏｒ，ｍｉｎ / m ａｃｔｕａｔｏｒ，ｍａｘ ) Perform the following steps, namely - The limiting amount of the actuator (u ｌｉｍ (k)) step, - Desired value (y(k)) of the control variable (y(k)) ｄｅｓ (k)) and the controlled object (Σ ｐ Extended control target model of ( Σ ) ｍ，ａｕｇ The desired value (y) up to the system order (n) of (•) ｄｅｓ Time derivative of (k) [Math 4] and The aforementioned limiting control amount (u ｌｉｍ (k)) depends on, The extended control target model (Σ ｍ，ａｕｇ Using (・)), The decision-making step, - The limited setting value (y(k)) of the control variable (y(k)) ｓｅｔ，ｌｉｍ (k)) The desired value (y(k)) of the controlled quantity (y(k)) ｄｅｓ (k) and the aforementioned desired value (y ｄｅｓ Time derivative of (k) [Math 5] Depending on, reverse orbit generator [Math 6] Using The decision-making step, - The required amount of operation of the actuator (u ｒｅｑ， (k)) the limiting control amount (u ｌｉｍ Steps to set it to be equal to (k), The method according to claim 2 or 3, characterized by carrying out the following.

5. The controlled object (Σ ｐ The operation of (・)) in the first mode (m ａｃｔｕａｔｏｒ，ｆｒｅｅ ) and the second mode (m ａｃｔｕａｔｏｒ，ｍｉｎ / m ａｃｔｕａｔｏｒ，ｍａｘ A switching device (10) is provided to switch between the two states. The switching device (10) sets a limited setting value (y(k)) of the controlled amount (y(k)). ｓｅｔ，ｌｉｍ (k)) set to the set value (y) of the controlled quantity (y(k)) ｓｅｔ (k)) is compared with the aforementioned limited mode (m) ａｃｔｕａｔｏｒ，ｍｉｎ / m ａｃｔｕａｔｏｒ，ｍａｘ ) away from the aforementioned non-limiting mode (m ａｃｔｕａｔｏｒ，ｆｒｅｅ Used to determine whether it can be replaced by ) and / or, the switching device (10) controls the requested amount of operation of the actuator (u ｒｅｑ (k)) the control amount limit (u ｒｅｑ，ｍｉｎ (k) / u ｒｅｑ．ｍａｘ (k)) is compared with the aforementioned non-limiting mode (m) ａｃｔｕａｔｏｒ，ｆｒｅｅ ) to the aforementioned limited mode (m ａｃｔｕａｔｏｒ，ｍｉｎ / m ａｃｔｕａｔｏｒ，ｍａｘ Used to determine whether or not a replacement must be made. The method according to any one of claims 2 to 4, characterized in that

6. The extended control target model (Σ ｍ，ａｕｇ (・)) is the control target (Σ ｐ (・)) The aforementioned controlled model (Σ ｍ The method according to any one of claims 1 to 5, characterized in that it is obtained by extending (•) with a model equation that models the physical functional balance.

7. The above method controls a fuel cell stack (Σ ｐ (•)) is used to operate, especially for control, and / or, the refrigerant discharge temperature from the fuel cell stack is used as the control variable (y(k)), In particular, it is used to control the refrigerant temperature difference by fuel cell stacks. A method according to any one of claims 1 to 6, characterized by the above.

8. The above method controls the form of the mixing point of the refrigerant system (Σ ｐ (•)) is used to operate, especially for control, and / or, use the refrigerant intake temperature into the fuel cell stack as the control variable (y(k)), A method according to any one of claims 1 to 7, characterized by the above.

9. The above method controls the form of a cooler in a refrigerant system (Σ ｐ (•)) is used to operate, especially for control, and / or, use the refrigerant discharge temperature from the cooler as the control variable (y(k)), A method according to any one of claims 1 to 8, characterized by the above.

10. The above method is used to control an object (Σ) that has multiple input quantities and / or multiple manipulated quantities and / or multiple output quantities and / or multiple controlled quantities. ｐ The method according to any one of claims 1 to 9, characterized in that it is used to operate, in particular to control.

11. A computer program product that includes an instruction to cause the computer to perform the method described in any one of claims 1 to 10 when the computer implements the computer program product.

12. A controller (100) having a memory unit in which code is filed and an arithmetic unit, wherein when the arithmetic unit implements the code, the method according to any one of claims 1 to 10 is implemented.

13. An electrochemical energy converter, particularly in the form of a fuel cell stack or electrolytic cell in a fuel cell system, having the controller (100) described in claim 12.

Images