ELECTRICAL SYSTEM AND METHOD AND DEVICE FOR DETERMINING A VALUE CHARACTER OF A CONTROLLED VARIABLE

Description

[0001] The present invention relates to a method for determining a value profile of a control variable of an electrical system and a corresponding device. Furthermore, the invention relates to an electrical system. The electrical system can, in particular, be an electric drive or comprise an electric drive. State of the art

[0002] The document DE 10 2016 220536 A1 discloses a method for limiting the temperature in an electrical machine, wherein after a predetermined continuous temperature is exceeded, the retrieval of higher drive power is permitted for a short time and it is proposed to determine the temperature ϑ in the electrical machine and, after a predetermined continuous temperature ϑ is exceeded, to allow a further temperature increase in a time period calculated by a controller and, after this time period has elapsed, to reduce the power and / or the torque and / or at least one phase current of the electrical machine.

[0003] The publications US 10 151 301 B2 and WO 2019 / 219214 A1 disclose methods for determining the value profile of a controlled variable. Since losses occur during the operation of an electrical system, such as an electric drive, components of the electrical system, such as the power modules of an inverter, or a rotor or stator of a machine, heat up. To prevent overheating and the resulting damage to temperature-sensitive components, the resulting temperatures are monitored using sensors and models.

[0004] Derating or reduction procedures are used and implemented as self-protection measures. This means that the operating state of the electrical system components is modified or reduced to prevent critical temperatures from being exceeded.

[0005] To select a suitable operating strategy for the electric drive, it is advantageous if the vehicle control unit (VCU) (e.g., in the vehicle) receives information from an electric drive control unit regarding the power that can be demanded from the electric drive in a future time interval, such as the next 1, 5, or 10 seconds, so that the protective measures are not yet activated. In contrast to range prediction or model-predictive control, this is a prediction of the performance of the electric drive.

[0006] Various forward modeling strategies exist for this purpose. For example, based on the current operating state, it is possible to calculate whether a derating is to be expected for a future time interval. However, this only provides the higher-level control unit with information about whether the current operating state can be maintained or not. A potentially necessary adjustment of the operating state cannot be deduced.

[0007] Furthermore, based on the current thermal state and assuming maximum load during the future time interval, it is possible to calculate which temperatures would be reached with this load. The corresponding limited load state is determined from a given derating characteristic. However, the iterative nature of derating—that is, the feedback of the current temperature to the currently permissible load—is not taken into account. The prediction is therefore usually too conservative.

[0008] Forward modeling can be repeated iteratively with different load cases until a limit temperature is met, but this is laborious. If the derating procedure intervenes, the load case can be iterated consistently with the derating conditions, but this is also very laborious.

[0009] A model predictive approach is known from Wallscheid et al., "Derating of Automotive Drive Systems Using Model Predictive Control", 4th Symposium on Predictive Control of Electrical Drives and Power Electronics, 2017. A prerequisite is that the driving strategy and derating are controlled jointly, which is often not the case because the electric drive is not responsible for the operating or driving strategy and the higher-level VCU, in turn, is not responsible for its own protection.

[0010] Finally, artificial intelligence can be used, but this requires complex data generation and recording, as the predicted loads only serve as additional information in a driving strategy and are often not actually driven. Therefore, complex special profiles would have to be run to obtain the training data. Disclosure of the invention

[0011] The invention provides a method and a device for determining a value profile of a control variable of an electrical system, as well as an electrical system having the features of the independent patent claims.

[0012] Preferred embodiments are the subject of the respective subclaims.

[0013] According to a first aspect, the invention therefore relates to a method for determining a value curve of a controlled variable of an electrical system in a predetermined time interval. The electrical system is controlled by the controlled variable. Due to the system behavior of the electrical system, an output variable is established in the electrical system as a function of the controlled variable. Depending on an actual value of the output variable at an initial time of the time interval and depending on the control variable assuming the value zero within the time interval, a value of the output variable is determined at an end time of the predetermined time interval. A first difference between a first predetermined value of the output variable at the end time and the determined value of the output variable at the end time is determined.Furthermore, at least a second difference between at least one second predetermined value of the output variable at the end time and the determined value of the output variable at the end time is determined. A value profile of the controlled variable in the time interval is predetermined, wherein the value profile depends on an end value at the end time and is uniquely determined by specifying the end value. A first permissible value of the controlled variable is determined such that, when the electrical system is controlled, the value of the output variable at the end time is equal to the determined first difference, depending on the controlled variable assuming the predetermined value profile within the time interval, wherein the end value is equal to the first permissible value.At least one second permissible value of the control variable is determined in such a way that, when the electrical system is controlled, the value of the output variable at the end time is equal to the determined second difference, depending on the control variable assuming the specified value curve within the time interval, wherein the end value is equal to the second permissible value. Based on the first permissible value of the control variable for the first specified value of the output variable and based on the at least one second permissible value of the control variable for the at least one second specified value of the output variable, a curve of the permissible values ​​of the control variable is determined as a function of the output variable. An intersection point of the determined curve and a specified derating function is determined, wherein the derating function describes a maximum permissible control variable in a derating method as a function of the output variable.The value curve of the control variable within the time interval is determined as the value curve for which the final value of the value curve assumes the value of the control variable at the intersection point.

[0014] According to a second aspect, the invention relates to a device for determining a value profile of a controlled variable of an electrical system in a predetermined time interval, wherein the electrical system can be controlled with the controlled variable and wherein, due to a system behavior of the electrical system as a function of the controlled variable, an output variable is established in the electrical system. The device comprises a computing device which determines a value of the output variable at an end time of the predetermined time interval, as a function of an actual value of the output variable at a start time of the time interval, and as a function of the controlled variable assuming the value zero within the time interval. Furthermore, the computing device determines a first difference between a first predetermined value of the output variable at the end time and the determined value of the output variable at the end time.The computing device further determines at least a second difference between at least one second predetermined value of the output variable at the end time and the determined value of the output variable at the end time. The computing device specifies a value profile of the control variable in the time interval, wherein the value profile depends on an end value at the end time and is uniquely determined by specifying the end value. The computing device determines a first permissible value of the control variable such that, when the electrical system is controlled, the value of the output variable at the end time is equal to the determined first difference, depending on the control variable assuming the predetermined value profile within the time interval, wherein the end value is equal to the first permissible value.The computing device determines at least one second permissible value of the controlled variable such that, when the electrical system is controlled, the value of the output variable at the end time is equal to the determined second difference, depending on the control variable assuming the predetermined value profile within the time interval, wherein the end value is equal to the second permissible value. The computing device determines a curve of the permissible values ​​of the controlled variable as a function of the output variable, based on the first permissible value of the controlled variable for the first predetermined value of the output variable and based on the at least one second permissible value of the controlled variable for the at least one second predetermined value of the output variable. The computing device determines an intersection point of the determined curve and a predetermined derating function.The derating function describes a maximum permissible control variable in a derating process as a function of the output variable. The computing device determines the value curve of the control variable within the time interval as the value curve for which the final value of the value curve assumes the value of the control variable at the intersection point. Furthermore, the device comprises an interface configured to output the determined value curve of the control variable for controlling the electrical system.

[0015] According to a third aspect, the invention relates to an electrical system comprising a device according to the invention for determining a value profile of a control variable of the electrical system. The electrical system further comprises a control device configured to control the electrical system within the time interval depending on the determined value profile of the control variable. Advantages of the invention

[0016] According to the invention, a value curve of a control variable is predicted, which can be used for the specified time interval to control the electrical system. The starting point of the time interval corresponds to the current point in time or a starting point of the control process. The value curve of the control variable is calculated in such a way that a derating process is prevented from starting or that a limitation due to the active intervention of the derating occurs.

[0017] The method determines the maximum possible load case that the electrical system can provide in the time interval without conflicting with derating. The method allows for the correct consideration of any derating curves through suitable and efficient backward modeling of the just-possible load state. The value curve of the control variable should thus be able to be used for control purposes as determined, even taking the derating behavior into account, throughout the entire time interval. This prevents unforeseen changes in the control variables due to derating behavior.

[0018] By determining the value of the output variable at the end time, where the control variable assumes the value zero within the time interval, a decay process is described in the no-load case, ie under the assumption that no further load is applied during the time interval.

[0019] Determining the first difference or second difference corresponds to a respective reserve, such as a temperature reserve, which exists compared to the no-load case and thus remains for the application of a load within the time interval.

[0020] The specified value progression of the control variable can be a constant step. In this case, the value of the control variable is set to a fixed value at the initial time and remains at the set value throughout the time interval. A load step can be described by a Heaviside function with a scalable prefactor. This makes the calculation quick and easy.

[0021] The specified value progression of the control variable can also follow a specified function. For example, the value of the control variable can increase linearly or quadratically. The value progression is thus linear or parabolic.

[0022] According to one embodiment of the method, the value progression can also depend on the starting value of the control variable at the beginning of the time interval. Accordingly, the determined value progression of the control variable within the time interval can also depend on a starting value of the control variable at the beginning of the time interval.

[0023] According to one embodiment of the method, the determined value profile of the control variable is provided to a higher-level system as input information for the operating or driving strategy. In particular, the invention relates to a method for controlling the electrical system, wherein the determined value profile of the control variable is used to control the electrical system.

[0024] According to one embodiment of the method, the electrical system is controlled such that the controlled variable follows the specific value profile within the time interval. However, it can also be provided to control the controlled variable such that, for every point in time within the time interval, it is less than or equal to the corresponding value of the specific value profile. The value profile thus specifies a maximum permissible profile, which may not be exceeded but may not be undercut.

[0025] The method can be implemented without iterations. It is fast, highly computationally efficient, and accurate, making it suitable for simultaneously considering multiple components to be protected, such as in the drive train on the microcontroller of an inverter, without significantly increasing computational load or memory requirements. The effort required for model creation and parameterization is minimal.

[0026] The at least one control variable (or regulated variable) can be, for example, a current or voltage. Furthermore, the control variable can also be a torque, for example, if the control variable comes from the driving strategy of the higher-level VCU, which is used to control the electrical system.

[0027] According to one embodiment of the method, the output variable to be protected in the electrical system is a temperature, a pressure or the like.

[0028] According to a further embodiment of the method, the curve of the permissible values ​​of the control variable is determined by interpolating the first permissible value of the control variable for the first predetermined value of the output variable and the at least one second permissible value of the control variable for the at least one second predetermined value of the output variable.

[0029] According to a further embodiment of the method, the dependence of the output variable on the controlled variable is described by a linear or linearizable model. According to the invention, the method is applicable to all electrical systems that have state variables or output variables to be limited, which can be described with a linear or linearizable model and for which a derating method, for example in the form of a curve (derating curve), is specified. According to the further embodiment, the curve can describe a relationship between the load or excitation variable and the output variable to be protected. The linear or linearizable model can be a description of a time-invariant behavior (linear time invariant, LTI) or a linear parameter variant behavior (linear parameter variant, LPV).

[0030] According to a further embodiment of the method, the value progression is abrupt, whereby for a given value of the output variable the permissible value u allowed the taxable amount is calculated using the following formula: u allowed = Δ y Φ t 1 , where Δy the determined difference and Φ ( t 1 ) the value of a step response assigned to the linear or linearizable model at the end time t 1 is.

[0031] According to a further embodiment of the method, the control variable is smaller than a threshold value specified due to system limitations of the electrical system. The value progression of the control variable within the time interval is determined as the value progression for which the final value of the value progression assumes the specified threshold value if the first permissible value of the control variable and the at least one second permissible value of the control variable are greater than the specified threshold value.

[0032] According to a further embodiment of the method, the electrical system is an electric drive, wherein the at least one output variable comprises a temperature of at least one component of the electrical system.

[0033] According to a further embodiment of the method, a phase current or torque of the electric drive is controlled as a function of the determined value profile of the control variable.

[0034] According to a further embodiment of the device, the computing device is designed to determine the curve of the permissible values ​​of the control variable by interpolating the first permissible value of the control variable for the first predetermined value of the output variable and the at least one second permissible value of the control variable for the at least one second predetermined value of the output variable. Short description of the drawings

[0035] They show: Figure 1 shows a schematic block diagram of a device for determining a value profile of a controlled variable of an electrical system according to an embodiment of the invention; Figure 2 shows a schematic block diagram of an electric drive with a device for determining a value profile of a controlled variable of the electrical system according to an embodiment of the invention; Figure 3 shows a linear or linearizable model for describing the behavior of components of the electrical machine; Figure 4 shows a schematic block diagram for explaining the linear model; Figure 5 shows a representation of a scalable load step; Figure 6 shows a schematic illustration of a derating curve; Figure 7 shows a flowchart of a method for determining a value profile of a controlled variable of an electrical system according to an embodiment of the invention.Figure 8 shows an explanation of the determination of the control variable by calculating an intersection point of a derating curve and a determined curve; Figure 9 shows a further flowchart of a method for determining a value curve of a control variable of an electrical system according to an embodiment of the invention; Figure 10 shows an exemplary curve of a control variable and an output variable up to the start time of the time interval; Figure 11 shows an exemplary curve of a control variable and an output variable with a predicted curve in the specified time interval; Figure 12 shows an exemplary curve of a control variable and an output variable with a predicted curve in the specified time interval for different threshold values; Figure 13 shows an exemplary curve of an output variable as a function of time;Figure 14 illustrates the determination of the control variable by calculating the intersection point of a derating curve and a determined curve for the case of a phase current; and Figure 15 illustrates an example of a temperature and phase current curve as a function of time.

[0036] In all figures, identical or functionally equivalent elements and devices are provided with the same reference numerals. The numbering of process steps serves the purpose of clarity and is generally not intended to imply a specific chronological order. In particular, several process steps can be performed simultaneously. Description of the embodiments

[0037] Figur 1 shows a schematic block diagram of a device 1 for determining a value curve of a control variable of an electrical system 2 in a predetermined time interval.

[0038] Electrical system 2 may be, for example, an electric drive. Electrical system 2 is operated using a derating method. Electrical system 2 can be controlled with a control variable, whereby an output variable is established in electrical system 2 based on the system behavior of electrical system 2 as a function of the control variable.

[0039] The device 1 comprises a computing device 11, which is coupled to the electrical system 2 and receives instantaneous values ​​of at least one output variable from it. The at least one output variable depends on an operating state of at least one component of the electrical system 2 and includes, for example, a temperature of components of the electrical system 2. The components of the electrical system can be, for example, temperature-sensitive semiconductor components such as IGBTs. The computing device 11 can comprise one or more microprocessors and memory.

[0040] The computing device 11 determines a value profile of a control variable for the specified time interval, such as 1 second, 5 seconds, or 10 seconds. The control variable can be, for example, a phase current for an electric drive, a voltage, a torque derived therefrom, or the like. A starting time t 0 The time interval can correspond to a current point in time or a point in time in the future.

[0041] The computing device 11 calculates the dependence of the output variable on the control variable using a linear or linearizable model.

[0042] The computing device 11 determines a value of the output variable at an end time t 1 of the specified time interval, depending on an actual value of the output variable at the initial time t 0 of the time interval, and depending on the control variable taking the value zero within the time interval.

[0043] Furthermore, the computing device 11 determines a first difference of a first predetermined value of the output variable at the end time t 1 and the determined value of the output variable at the end time t 1 . Furthermore, the computing device 11 determines at least a second difference of at least a second predetermined value of the output variable at the end time t 1 and the determined value of the output variable at the end time t 1 . The computing device 11 specifies a value progression of the control variable in the time interval, wherein the value progression extends from a final value to the final time t 1 and is uniquely determined by specifying the final value. The computing device 11 determines a first permissible value of the control variable such that when the electrical system 2 is activated, the value of the output variable at the end time t 1 is equal to the determined first difference, depending on the control variable assuming the specified value curve within the time interval, wherein the final value is equal to the first permissible value. The computing device 11 determines at least a second permissible value of the control variable such that when the electrical system 2 is actuated, the value of the output variable at the end time t 1 is equal to the determined second difference, depending on the control variable assuming the specified value curve within the time interval, wherein the final value is equal to the second permissible value. The computing device 11 determines a curve of the permissible values ​​of the control variable as a function of the output variable, based on the first permissible value of the control variable for the first specified value of the output variable and based on the at least one second permissible value of the control variable for the at least one second specified value of the output variable. The computing device 11 determines an intersection point of the determined curve and a specified derating function. The derating function describes a maximum permissible control variable in a derating method as a function of the output variable.The computing device 11 determines the value curve of the control variable within the time interval as the value curve for which the final value of the value curve assumes the value of the control variable at the intersection point.

[0044] The determined temporal value profile of the control variable is output via a wireless or wired interface 12 to a control unit of the electrical machine 2, which controls the electrical machine 2 or components of the electrical machine 2 taking into account the determined value profile of the control variable.

[0045] Figur 2 shows a schematic block diagram of an electric drive 2 as an electrical system 2 with a device 1 described above. The electric drive 2 comprises a DC battery 21, which is coupled to an inverter 22, which in turn supplies an electric machine 23 that provides a torque Tq. The device 1, which outputs the temporal value profile of the control variable to a control device (VCU) 24, is integrated into the inverter 22. The control device 24 controls the inverter 22 of the electric machine 2 taking into account the temporal value profile of the control variable. The determined value profile of the control variable can be used directly for control. Alternatively, the determined value profile of the control variable can specify a maximum profile that must not be exceeded.

[0046] Figur 3 shows a linear or linearizable model 31 for describing a behavior of components of the electrical machine 2. The linear model 31 can, for example, describe an LTI or LPV system and is used to calculate the temporal behavior of the relevant output variable y(t) as a function of time, for example a temperature T. The calculation is based on a previous system state x 0 and a tax variable u. The tax variable u can also be referred to as input variable, excitation or load. The linear or linearizable model 31 calculates under different load conditions u(t) taking into account a system dynamic 32 a current system state x and predicts the value of the output variable y.

[0047] Figur 4 shows a schematic block diagram to explain a linear model 41, such as the linear model 31 described above. The linear model 41 is divided into two parts, namely a first model 42 and a second model 43.

[0048] The first model 42 is used to calculate the behavior of the output variable for the load-free (homogeneous) case with initial conditions, ie a transient decay behavior due to the load history, i.e. under the assumption that the electrical system 2 is not loaded with respect to at least one control variable in the time interval. Using the first model 42, the value of the output variable at the end time t 1 of the specified time interval depending on an actual value of the output variable at the initial time t 0 of the time interval, depending on the control variable taking the value zero within the time interval.

[0049] The second model 43 is used to calculate a step response Φ( t ), ie the behavior of the output variable when loaded with an adjustable load within the time interval and without preload (initial condition = 0). The second model 43 thus calculates a load behavior of the at least one output variable y without preload of the electrical system 2 and under the assumption that an adjustable load of the electrical system 2 is present in the time interval, in particular a unit load step H ( t ) .

[0050] Figur 5 shows a representation of an exemplary scalable load step H ( t ) as a function of time t. The tax variable (burden) u jumps from a value 0 to the start time t = 0of the given time interval to a constant value Λ, is therefore described by a Heaviside function with a scalable prefactor A.

[0051] Figur 6 shows a schematic illustration of a derating curve D to describe the derating process, ie a control relationship between the output variable y (state) to be limited (protected) and the control variable to be reduced if necessary u (Load quantity, e.g. power loss or phase current). Here, a function dependent on the control variable u f(u) applied, whereby in the simplest case f(u) = u applies.

[0052] As soon as the output variable y to be protected (e.g. the temperature of a component of the electrical system 2) reaches a lower threshold y low reached, the value of the function decreases f(u) from the maximum value f(u max ), until it reaches an upper threshold y up the output variable y takes the value 0. In the simplest case, the behavior in this range is linear. Between the threshold values y low and y up This results in derating (down regulation), ie the load is reduced.

[0053] Figur 7 shows a flowchart of a method for determining a value curve of a control variable of an electrical system 2, which is operated as described above using a predetermined derating method.

[0054] In a first method step S71, depending on an actual value of the output variable y at an initial time of the time interval (ie on a current system state x 0 ) and depending on the control variable taking the value zero within the time interval, a value of the output variable y at the end time t 1 of the specified time interval. This calculates a reserve available without load with respect to at least one output variable y, e.g. a temperature reserve y resv .

[0055] In a second process step S72, a permissible value u allowed the taxable amount u calculated, ie a permissible maximum amplitude u allowed for a step function as input signal, with which a limit is set during the period y limit the output variable y is just reached.

[0056] If the determined permissible tax amount u allowed for the value of the output variable y limit , at which derating starts is greater than the upper limit possible due to other system limitations u max ( u allowed > u max ), then the prediction can be u pred = u max be used.

[0057] Otherwise, steps S71 and S72 are repeated to determine a working characteristic curve that describes the relationship between the output variable achieved during the period y limit,n and the associated load u allowed,n describes. Here, n = 2, ie the calculation is carried out for two limits for y. Alternatively, the calculation can be carried out for several values, ie n > 2, where, for example, polynomial interpolation or spline interpolation can be used. The operating characteristic curve can be linearized or quadratic, for example. The operating characteristic curve is a curve of the permissible values ​​of the control variable.

[0058] Any special cases that may occur will be dealt with appropriately, for example if u allowed > u max is or the size y resv < is 0.

[0059] Due to the assumption / precondition of a linear or linearizable dynamic system behavior, it is possible to determine exactly the input excitation for which a relevant output variable is exactly achieved in an analytical backward calculation, instead of inferring the corresponding output variable with a freely selectable level of input excitation (power loss) in a forward model.

[0060] The dynamic system behavior can be described by a differential equation (DGL), where the output variable y(t) the solution of the ODE for a known control variable u(t) The ODE has the general form: a n y n + ⋯ a 2 y ¨ + a 1 y ˙ + a 0 y = b m u m + ⋯ + b u ¨ + b 1 u ˙ + b 0 u .

[0061] The solution y(t) consists of a homogeneous solution y H (t) and a particulate solution y P (t) : y t = y H t + y p t

[0062] The starting point is a target value for y (Δ t pred ) = y limit after a given (prediction) period Δ t pred .

[0063] In a process step S74, the boundary condition is used that the load during the period of a scaled step function u ( t ) = u·H ( t ), where H a Heaviside function with jump at t 0 designated.

[0064] The particular solution can therefore be described using the step response Φ(t): y p t = u Φ t .

[0065] The homogeneous solution can be described analytically for a known prediction period by suitable discretization, so that with known initial conditions (in the form of temporal derivatives from the history, or states in a state space description) a decay value in the prediction period y H (Δ t pred ) is calculable. In this context, the above equation can be rewritten as: y limit = y H Δt pred + u ⋅ Φ Δt pred .

[0066] In a process step S73, this equation can be solved for the unknown height of the load jump u: y limit − y h Δt pred Φ Δt pred = u = u allowed .

[0067] The method is directly applicable when the input variable can be described as a single-variable system, such as the power loss, which is relevant for the output temperature. An extension to a multivariable system with a linear system of equations is possible.

[0068] An intersection point of the curve or operating characteristic and the derating function D provides the predicted possible load u allowed . The value u pred is translated into a value that can be used for the driving strategy, such as a torque tq pred as input variable for the control device 24.

[0069] Figur 8 shows an explanation of the determination of the control variable by calculating an intersection point of a derating function D and the determined curve K.The derating function shown top left D has the in Figur 6 The curve shown in the upper right corner K runs linearly, whereby permissible values u allowed,1 and u allowed,2 the taxable amount u respective thresholds y limit,1 and y limit,2 the output variable y An intersection point of the two curves D, K gives the value u pred the taxable amount u. A corresponding value progression in the time interval corresponds to a jump at the initial time t 0 to the determined value u pred the taxable amount u, which is assumed throughout the entire time interval.

[0070] Figur 9 shows a further flowchart of a method for determining a value profile of a control variable of an electrical system 2 according to an embodiment of the invention. For this purpose, in a first method step S91, a derating functionD In a second process step S93, a working characteristic curve K In a third process step S92, the operating characteristic curve K and the derating characteristic D the value u pred calculated, as described in more detail above. In a fourth method step S94, a transformation into an output variable for the driving strategy takes place, such as a torque tq pred .

[0071] Figur 10 shows an example of a curve of an output variable y ( Figur 10 above) and a control variable u ( Figur 10 below) until the start time t 0 the prediction, that is, up to the initial time t 0 of the time interval. The end time t 1 of the time interval is also shown. For the output variable, a limit value y limit specified.

[0072] Figur 11 shows an example of an output variable y( Figur 11 above) and a control variable u ( Figur 11 below) with a value curve of the control variable u in the specified time interval determined using the method according to the invention.

[0073] Figur 12 shows exemplary curves 121, 122 of an output variable y ( Figur 12 above) and curves 123, 124 of a control variable u ( Figur 12 below) with a value curve of the control variable determined using the method according to the invention u in the specified time interval for different limit values y limit,1 and y limit,2 .

[0074] Figur 13 shows an example of an output variable y as a function of time t. Shown are the contributions due to the decay behavior 133 and due to a load behavior 132. The two contributions result in the overall behavior 131. The temperature reserve T res is used optimally.

[0075] Figur 14 shows an explanation of the determination of the control variable by calculating an intersection point of a derating function 141 and a determined curve 142 for the case of a phase current Iph. Also shown are the maximum phase current Iph max and the lower and upper temperature thresholds, T low and T high .

[0076] Figur 15 shows an example of a temperature T ( Figur 15 above) with thresholds T limit,1 and T limit,2 , according to the thresholds T low and T high for the derating function 141 in Figur 14 .

[0077] In Figur 15 Below are the predicted or desired phase current 153 and the maximum permissible phase current 151. The actual behavior 152 of the phase current results when, from the start time t 0 the determined value curve is set and from the time t 1 the derating takes effect and the actual phase current is below the determined value. The temperature curve in Figur 15 corresponds to the current curve 152.

Claims

1. Method for determining a value profile of a control variable (u) of an electrical system (2) at a predefined time interval, wherein the electrical system has an output variable (y) which is to be limited, which can be described with a linear or linearizable model and for which a derating function (D) is predefined, and wherein the time interval describes a decay process in the load-free case, wherein the electrical system (2) is controlled with the control variable (u), and wherein, due to a system behaviour of the electrical system (2), the output variable (y) in the electrical system (2) is set on the basis of the control variable (u), with the steps of: determining a value of the output variable (y) at an end time (t1) of the predefined time interval on the basis of an actual value of the output variable (y) at a start time (t0) of the time interval and on the basis of the fact that the control variable (u) assumes the value of zero within the time interval and the time interval thus describes a decay process in the load-free case; determining a first difference between a first predefined value (ylimit,1) of the output variable (y) at the end time (t1) and the determined value of the output variable (y) at the end time (t1); determining at least one second difference between at least one second predefined value (ylimit,2) of the output variable (y) at the end time (t1) and the determined value of the output variable (y) at the end time (t1); predefining a value profile of the control variable (u) in the time interval, wherein the value profile depends on a final value at the end time (t1) and is clearly determined by specifying the final value; determining a first permissible value (uallowed,1) of the control variable (u) such that, when controlling the electrical system (2), the value of the output variable (y) at the end time (t1) is equal to the determined first difference on the basis of the fact that the control variable (u) assumes the predefined value profile within the time interval, wherein the final value is equal to the first permissible value (uallowed,1), determining at least one second permissible value (uallowed,2) of the control variable (u) such that, when controlling the electrical system (2), the value of the output variable (y) at the end time (t1) is equal to the determined second difference on the basis of the fact that the control variable (u) assumes the predefined value profile within the time interval, wherein the final value is equal to the second permissible value (uallowed,2); determining a curve (K) of the permissible values of the control variable (u) on the basis of the output variable (y) using the first permissible value (uallowed,1) of the control variable (u) for the first predefined value (ylimit,1) of the output variable (y) and using the at least one second permissible value (uallowed,2) of the control variable (u) for the at least one second predefined value (ylimit,2) of the output variable (y); determining a point of intersection of the determined curve (K) and the predefined derating function (D), wherein the derating function (D) describes a maximum permissible control variable (umax) in a derating method on the basis of the output variable (y); and determining the value profile of the control variable (u) within the time interval as that value profile for which the final value of the value profile assumes the value (upred) of the control variable (u) at the point of intersection.

2. Method according to Claim 1, wherein the curve of the permissible values of the control variable (u) is determined by interpolating the first permissible value (uallowed,1) of the control variable (u) for the first predefined value (ylimit,1) of the output variable (y) and the at least one second permissible value (ual-lowed,2) of the control variable (u) for the at least one second predefined value (ylimit,2) of the output variable (y).

3. Method according to Claim 1 or 2, wherein the dependence of the output variable (y) on the control variable (u) is described by a linear or linearizable model.

4. Method according to Claim 3, wherein the value profile is erratic, and wherein, for a predefined value of the output variable (y), the permissible value ual-lowed of the control variable (u) is calculated by means of the following formula: u allowed = Δ y Φ t 1 , where Δy is the determined difference and Φ(t1) is the value of a step response assigned to the linear or linearizable model at the end time t1.

5. Method according to one of the preceding claims, wherein the control variable (u) is less than a threshold value predefined on the basis of system limitations of the electrical system (2), and wherein the value profile of the control variable (u) within the time interval is determined as that value profile for which the final value of the value profile assumes the predefined threshold value if the first permissible value (uallowed,1) of the control variable (u) and the at least one second permissible value (uallowed,2) of the control variable (u) is greater than the predefined threshold value.

6. Method according to one of the preceding claims, wherein the electrical system (2) is an electric drive, and wherein the at least one output variable (y) comprises a temperature of at least one component of the electrical system (2).

7. Method according to Claim 6, wherein a phase current or torque of the electric drive is controlled on the basis of the determined value profile of the control variable (u).

8. Apparatus (1) for determining a value profile of a control variable (u) of an electrical system (2) in a predefined time interval, wherein the electrical system has an output variable (y) which is to be limited, which can be described with a linear or linearizable model and for which a derating function (D) is predefined, and wherein the time interval describes a decay process in the load-free case, wherein the electrical system (2) can be controlled with the control variable (u), and wherein, due to a system behaviour of the electrical system (2), the output variable (y) in the electrical system (2) is set on the basis of the control variable (u), having: a computing device (11) which is designed - to determine a value of the output variable (y) at an end time (t1) of the predefined time interval on the basis of an actual value of the output variable (y) at a start time (t0) of the time interval and on the basis of the fact that the control variable (u) assumes the value of zero within the time interval and the time interval thus describes a decay process in the load-free case, - to determine a first difference between a first predefined value (ylimit,1) of the output variable (y) at the end time (t1) and the determined value of the output variable (y) at the end time (t1), - to determine at least one second difference between at least one second predefined value (ylimit,2) of the output variable (y) at the end time (t1) and the determined value of the output variable (y) at the end time (t1), - to predefine a value profile of the control variable (u) in the time interval, wherein the value profile depends on a final value at the end time (t1) and is clearly determined by specifying the final value, - to determine a first permissible value (uallowed,1) of the control variable (u) such that, when controlling the electrical system (2), the value of the output variable (y) at the end time (t1) is equal to the determined first difference on the basis of the fact that the control variable (u) assumes the predefined value profile within the time interval, wherein the final value is equal to the first permissible value (uallowed,1); - to determine at least one second permissible value (uallowed,2) of the control variable (u) such that, when controlling the electrical system (2), the value of the output variable (y) at the end time (t1) is equal to the determined second difference on the basis of the fact that the control variable (u) assumes the predefined value profile within the time interval, wherein the final value is equal to the second permissible value (uallowed,2); - to determine a curve (K) of the permissible values of the control variable (u) on the basis of the output variable (y) using the first permissible value (ual-lowed,1) of the control variable (u) for the first predefined value (ylimit,1) of the output variable (y) and using the at least one second permissible value (ual-lowed,2) of the control variable (u) for the at least one second predefined value (ylimit,2) of the output variable (y); - to determine a point of intersection of the determined curve (K) and the predefined derating function (D), wherein the derating function (D) describes a maximum permissible control variable (umax) in a derating method on the basis of the output variable (y); and - to determine the value profile of the control variable (u) within the time interval as that value profile for which the final value of the value profile assumes the value (upred) of the control variable (u) at the point of intersection; and an interface (12) which is designed to output the determined value profile of the control variable (u) for controlling the electrical system (2).

9. Apparatus (1) according to Claim 8, wherein the computing device (11) is designed to determine the curve of the permissible values of the control variable (u) by interpolating the first permissible value (uallowed,1) of the control variable (u) for the first predefined value (ylimit,1) of the output variable (y) and the at least one second permissible value (uallowed,2) of the control variable (u) for the at least one second predefined value (ylimit,2) of the output variable (y).

10. Electrical system (2) having: an apparatus (1) for determining a value profile of a control variable (u) of the electrical system (2) according to Claim 8 or 9; and a control device (24) which is designed to control the electrical system (2) within the time interval on the basis of the determined value profile of the control variable (u).