Method and control device for controlling a drive motor of a muscle-powered vehicle

Abstract

A method for controlling (S4) a drive motor (4) of a human-powered vehicle (2) is described. The method comprises determining (S1) a nominal setpoint value of a rotational speed of a shaft (6) connected to the drive motor (4) and determining (S2) at least one adaptable parameter. The method comprises determining (S3) a threshold value for an absolute value of the setpoint value of the rotational speed as a function of a specific temperature of an element (8) of the vehicle (2) and as a function of the specific parameter. The method further comprises determining (S4) a setpoint value of the rotational speed of the shaft (6), wherein the specific nominal setpoint value is determined as the setpoint value if an absolute value of the nominal setpoint value is smaller than the threshold value for an absolute value of the setpoint value, and otherwise the setpoint value is determined such that an absolute value of the setpoint value corresponds to the specific threshold value.The method further comprises controlling (S5) the drive motor (4) as a function of the determined target speed. A control device (10) configured to carry out such a method is also described. Furthermore, a drive train (12) having such a control device (10) and a muscle-powered vehicle (2) having such a drive train (12) are described.

Description

[0001] The present invention relates to a method for controlling a drive motor of a human-powered vehicle. Furthermore, the present invention relates to a control device configured to carry out this method. Furthermore, the present invention relates to a drive train having such a control device and further to a vehicle having such a drive train.

[0002] Human-powered vehicles have drive motors to support the muscular power of a driver. A vehicle designed as a pedelec has an electric machine as the drive motor. Such a drive motor provides a specific speed to drive the vehicle. The higher the provided speed, the greater the power loss of the drive motor. If the speed is too high, this can lead to overheating of the drive motor and thus to reduced efficiency or thermal destruction of the drive motor. For this reason, such drive motors are controlled using derating functions that limit the power of the drive motor to prevent such overheating. DE 10 2022 200 541 A1, for example, describes how a derating function is modeled depending on several parameters. For this purpose, an expected thermal load for future points in time is determined.The drive motor is then controlled using the derating function modulated in this way.

[0003] The object of the invention is to provide an improved method for controlling a drive motor of a human-powered vehicle. This object is achieved by a method having the features of the independent claim.

[0004] In a first aspect, the present invention relates to a method for controlling a drive motor of a human-powered vehicle. The human-powered vehicle can be, for example, a bicycle, an e-bike, a pedelec, or a cargo bike. The drive motor can be, for example, an electric machine.

[0005] The method for controlling the drive motor can be a method for operating the drive motor. The drive motor can be speed-controlled and, alternatively or additionally, speed-regulated. The drive motor can be part of a drive train of the vehicle. The drive motor can relieve or assist a driver of the vehicle when driving or pushing the vehicle by providing drive force. The method can be a computer-implemented method. A control device as part of the vehicle can be configured to carry out steps of the method. The vehicle can have a crank and pedals for absorbing muscle power as drive force.

[0006] The method comprises determining a nominal setpoint for a rotational speed of a shaft connected to the drive motor. Alternatively or in addition to the step of determining the nominal setpoint, the method can comprise a step of reading in the nominal setpoint. In this case, for example, the nominal setpoint can be determined using a different method and read in the reading-in step as the first step of this method. The nominal setpoint can be determined independently of a temperature. The nominal setpoint can be a desired setpoint for the rotational speed of the shaft. In this determining step, one or more nominal setpoints for a rotational speed of the shaft can be determined, for example as a temporal progression of the nominal setpoint for the rotational speed of the shaft.When determining the nominal setpoint of a shaft speed, a single value or, alternatively, a time profile of the nominal setpoint of the shaft speed can be determined. The nominal setpoint can be referred to as the nominal value of the setpoint of a speed. The shaft can be connected directly or indirectly to the drive motor.

[0007] The method further comprises determining at least one adaptable parameter. The method can comprise determining exactly one or alternatively several adaptable parameters. The at least one adaptable parameter can be variable, for example during operation of the vehicle. The at least one adaptable parameter can be estimated during the determination. For example, the at least one adaptable parameter can describe a relationship between a power loss profile of the drive motor and a temperature profile of an element of the vehicle. For example, the at least one adaptable parameter can be determined on a vehicle-specific basis, i.e. for example for different product variants and alternatively or additionally production variants of the vehicle.Alternatively or additionally, the at least one adaptable parameter can be determined situation-specifically, for example for different operating conditions. Different operating conditions can be, for example, a different vehicle speed, a different wind speed, and alternatively or additionally, different vehicle pollution. A value of the at least one adaptable parameter can be unknown a priori. Initially, a predefined standard value can be used as the adaptable parameter. The at least one adaptable parameter may, for example, not be directly measurable. The at least one adaptable parameter cannot, for example, be determined before operation of the drive motor as part of product development or production.The at least one adaptable parameter can be determined, for example, during operation of the drive motor, for example, during the execution of the method for controlling the drive motor. The at least one adaptable parameter may not be determinable before the method is carried out.

[0008] The method comprises determining a threshold value for an absolute value of the target value of the rotational speed as a function of a specific temperature of the element of the vehicle and as a function of the at least one specific parameter. The determination can be a determination of one or more threshold values ​​for an absolute value of the target value of the rotational speed. In this case, for example, a threshold value for a specific point in time or, for example, a temporal progression of threshold values ​​can be determined. If multiple threshold values ​​are determined, the threshold values ​​for the absolute values ​​of target values ​​of the rotational speed can be determined. The determined temperature can be a predetermined temperature, such as, for example, a temperature progression of the element of the vehicle, such as a part of the drive motor of the vehicle.The determined temperature can be less than or equal to a maximum temperature that may prevail at the element without risking overheating of the drive motor. The threshold value should then be determined, for example, in such a way that this maximum temperature at the element is never exceeded. For example, the determination should be carried out in such a way that the following applies at all times: ϑ≤ϑmax, where ϑ is the specific temperature and ϑ max the maximum temperature.

[0009] The method further comprises determining a target value for the rotational speed of the shaft. In this case, the determined nominal target value is determined as the target value if an amount of the nominal target value is less than the threshold value for an amount of the target value. In an alternative embodiment, the determined nominal target value can be determined as the target value if the amount of the nominal target value is less than or equal to the threshold value for an amount of the target value. In this case, the determination can comprise comparing the nominal target value with the threshold value. When determining the target value, one target value can be determined; alternatively, several target values ​​of the shaft can be determined, for example as a time profile of the target value of the rotational speed of the shaft. In this case, for example, several determined nominal target values ​​can then be compared with several determined threshold values ​​for amounts of the target value, for example as time profiles.Otherwise, for example, if the nominal setpoint value is greater than or equal to the threshold value for the setpoint value, the setpoint is determined such that the setpoint value corresponds to the specified threshold value and the sign of the setpoint value corresponds to the sign of the specified nominal setpoint value. This can also be performed for individual values ​​or for a temporal progression of setpoints.

[0010] The method further comprises controlling the drive motor as a function of the determined target speed value. If, for example, multiple target speed values ​​are determined, for example, as a temporal progression of the target speed value, the drive motor can be controlled as a function of the multiple determined target speed values, such as the determined temporal progression of the target speed value. The control can be a control and, alternatively or additionally, a regulation of the drive motor. In this case, the speed at the shaft can, for example, be set close to the determined target speed value.

[0011] Such a method for controlling the drive motor can be used to implement an adaptive derating function. This allows drive power, provided here, for example, by controlling the drive motor using the speed, to be limited in order to prevent overheating of the drive motor. The method can thus be used for model-based derating of the drive motor. By determining the adaptable parameter, parameters of the model for model-based derating, such as the at least one adaptable parameter, can be determined and adapted during operation of the vehicle. In conventional derating methods, the greatly varying cooling capacity of the vehicle's drive motor during operation, for example due to different vehicle speeds, is often not taken into account.Failure to consider this variability often impairs the performance of conventional derating methods, as, for example, more or less cooling capacity is actually available than previously assumed when determining the parameters for the derating method. Alternatively, determining the parameters for conventional derating methods for different product variants and different operating parameters is very complex in development. For example, derating must be performed for different vehicles and for different vehicle speeds with different cooling capacities and thus with different parameters.

[0012] However, with the method described here, at least one adaptable parameter for derating can be determined situationally, and thus the method can be used to react situationally to changing cooling performance. This allows derating to be carried out more precisely. This makes it possible to achieve consistent and thus improved performance. The number of parameters to be parameterized can also be reduced, which must be laboriously parameterized before operation in order to be able to map different production variants or operating conditions with the derating method. The at least one adaptable parameter can still be undetermined before the vehicle is operated and can therefore, for example, be deliberately left open. For example, a default value can initially be used for the adaptable parameter, which is non-vehicle-specific and alternatively or additionally predefined non-situation-specific.This means that the development of the process described here can be simpler and less complex, faster and therefore more cost-effective.

[0013] By controlling the drive motor depending on the specific speed setpoint, the power of the drive motor can be effectively limited and reduced. Derating can prevent unwanted power losses and, alternatively or additionally, unwanted wear of the drive motor by preventing overheating of the drive motor. Normally, i.e. when the magnitude of the nominal setpoint is less than the threshold for the magnitude of the setpoint, the nominal setpoint can be passed on and used as a specific setpoint to control the drive motor. At the same time, a speed and thus a power of the drive motor can be automatically restricted and limited if the magnitude of the nominal setpoint is not less than the threshold and, for example, overheating would threaten if the nominal setpoint were used as the speed setpoint for control.The threshold value is then used to control the drive motor as the setpoint, with the direction of rotation of the shaft defined by the nominal setpoint being used for control. The method can therefore provide both nominal operation and derating operation. The method can implement a situational limitation of the drive motor's power. Situational can mean that in every situation where the nominal setpoint can be used to control the drive motor, it should also be used, and the drive motor's power should therefore not be limited or restricted. The method can therefore be combined with other methods and implementations for speed control; for example, another method can independently provide a nominal speed setpoint.

[0014] Using the method presented here, this nominal setpoint can then be read in, for example, and either used to control the setpoint or to limit the power of the drive motor. This method can therefore be integrated into several different control devices for controlling different drive motors of different vehicles. This can reduce development effort. The method therefore provides an easily partitionable method for controlling the drive motor, as it does not place any special requirements on, for example, higher-level methods for speed control, for example for determining a nominal setpoint. The nominal setpoint can be determined independently of the threshold value for the absolute value of the speed setpoint. The nominal setpoint can be determined using a variety of possible methods.Regardless of the specific method used to determine the nominal setpoint, the method for controlling the drive motor can be used for derating. This derating method can be combined with a variety of different methods for determining the nominal setpoint. This can further reduce development effort. This allows the actual logic for controlling the drive motor during normal operation, such as determining the nominal setpoint, to be implemented independently of the described method for controlling the drive motor for derating, i.e., for physically limiting the drive motor's power. The specific nominal setpoint used to control the drive motor therefore does not automatically prevent, for example, the maximum temperature of the element from being exceeded.However, comparing with the threshold value, performed using this method, leads to such a power limitation. Thus, development effort can be further reduced because the nominal setpoint can be determined independently.

[0015] According to a further embodiment, the method can be characterized in that the nominal setpoint can be determined as a function of a target speed. The nominal setpoint can be proportional to the target speed. Alternatively or additionally, the nominal setpoint can converge toward the target speed or a value proportional thereto. Converging here can mean an asymptotic convergence and, alternatively or additionally, a tendency toward convergence between the nominal setpoint and the target speed. The target speed can be a speed of the crank or the driven wheel of the vehicle.

[0016] The target speed can be specified by the driver. For example, a user input for the target speed or a value proportional to the target speed can be read in. User input can be read in via a user interface, such as a switch or a display on the vehicle. The user can, for example, enter how fast the vehicle should travel and thus how fast a wheel of the vehicle should turn. This way, the user can be assisted as required when pushing or pedaling, for example in Eco Mode or Speed ​​Mode. For example, the driver can specify that the driver wants push assistance or a certain vehicle speed. The target speed can then be determined using a table, for example. The target speed can be determined independently of a speed measurement. Alternatively, the target speed can be a measured speed on the vehicle.

[0017] Thus, the method can be used to determine a nominal setpoint, or alternatively, multiple nominal setpoints, such as a temporal progression of the nominal setpoint, depending on the situation, depending on the target speed. The method can thus be used to synchronize or control the bicycle's speeds, for example, by converging the nominal setpoint toward the target speed or toward a value proportional to the target speed.

[0018] According to a further embodiment, the method can be characterized in that the nominal setpoint is determined as a function of a measured speed of a crank of the vehicle as the target speed. The speed of the crank can be referred to as cadence. The speed of the vehicle's crank can be measured, for example, using a speed sensor on the crank. The measured speed can, for example, be a speed applied by the driver of the vehicle, for example, by pedaling.

[0019] For example, the method can implement pedal assist to support the driving force applied by the driver to the vehicle's pedals using muscle power. This method can increase comfort because the nominal speed setpoint, for example, converges to the cadence applied by the driver, thus preventing the pedals from hitting the driver's feet due to the pedal assist.

[0020] According to a further embodiment, the method can be characterized in that the nominal setpoint is determined as a function of a measured rotational speed of a driven wheel of the vehicle as the target rotational speed. The driven wheel can be a rear wheel of the vehicle configured as a pedelec. A sensor can be configured to measure the rotational speed of the driven wheel. The driven wheel can be mechanically operatively connected to the shaft.

[0021] For example, the method can implement push assist to support the driving force applied by the driver to the vehicle's handlebars using muscle power. This method can increase comfort because the nominal speed setpoint, for example, converges to the speed of the driven wheel applied by the driver, thus preventing the handlebars from hitting the driver's hands due to the push assist.

[0022] According to a further embodiment, the method can be characterized in that the determination of the nominal setpoint can be carried out depending on a gear ratio of the vehicle. The gear ratio can be a ratio between the target speed of the vehicle's crank and the nominal setpoint speed. Alternatively or additionally, the gear ratio is the ratio between the target speed of the driven wheel and the nominal setpoint. The gear ratio can define a relationship between the shaft speed and the target speed of the vehicle, such as the cadence or the speed of the driven wheel. The gear ratio can be part of a proportional relationship between the nominal setpoint and the target speed. The following formulas show the relationship between the nominal setpoint speed and the target speed: ωref,nom=kpedl ωpedl ωref,nom→kpedl ωpedl ωref,nom=kwhl ωwhl ωref,nom→kwhl ωwhl

[0023] Where ω ref,nom the nominal speed setpoint, k pedl is the ratio between the target speed of the vehicle's crank and the nominal speed setpoint. k whl is the ratio between the target speed of the driven wheel and the nominal speed setpoint. ω pedl is the target speed at the crank and ω whl is the target speed at the driven wheel. According to the first and third equations above, there is a proportional relationship between the target speed and the nominal setpoint. According to the second and fourth formulas, the nominal setpoint converges to the target speed.

[0024] The method can thus be used for different vehicles with different gear ratios, and for different gear ratios within a single vehicle, for example, due to a gearshift. A nominal target value can then be determined depending on the situation.

[0025] According to a further embodiment, the method can be characterized in that the determination of the nominal setpoint can be carried out recursively. The determination of the nominal setpoint can be carried out statefully. For example, a first nominal setpoint can be determined for a first point in time and a second nominal setpoint can be determined for a second point in time later than the first. The determination of the second nominal setpoint can be dependent on the determined first nominal setpoint. In this case, provided that the nominal setpoint has not previously been used as the setpoint, nominal setpoints can be determined as the method continues to run depending on specific speed setpoints. These nominal setpoints can thus be overwritten, for example with specific threshold values ​​with amounts of the speed setpoint.

[0026] According to a further embodiment, the method can be characterized in that determining the threshold value can include determining the power loss of the drive motor using a first mapping. The first mapping can map the power loss of the drive motor, environmental parameters, and the at least one adaptable parameter to the temperature. For example, the first mapping can be an operator, such as a differential operator. The relationship between the temperature, the power loss, the environmental parameters, and the at least one adaptable parameter can be represented by the following formula: ϑ=Φ[P,ψ,μ]

[0027] Here, ϑ is the temperature, Φ is the first mapping designed as an operator, ψ are the environmental parameters and µ is the at least one adaptable parameter. The environmental parameters can include, for example, the ambient temperature, air humidity, vehicle speed, temperature values ​​and, alternatively or additionally, values ​​of time derivatives of the temperature at a specific point in time. These environmental parameters can influence the relationship between a temporal progression of the power loss and the temperature. These environmental parameters can be known during and before operation of the vehicle and thus during and before carrying out the method. For example, these environmental parameters can be determined during production or product development. For example, the environmental parameters can be measured directly with a sensor during or before carrying out the method.

[0028] The first mapping, designed as an operator, can map temporal curves of the power, the ambient parameters, and the at least one adaptable parameter to a temporal curve of the temperature. To determine the power loss, a permissible target temperature curve is selected, for example, expressed by the following equation: ϑlim≤ϑmax

[0029] For example, at any given time a limited temperature ν lim be less than or equal to the maximum permissible temperature ϑ max For example, it can also apply for any point in time that the permissible target curve ϑ lim equal to the maximum permissible temperature ϑ max The previously mentioned specific temperature of the element, on which the determination of the threshold value for an amount of the setpoint of the speed depends, can, for example, be the permissible setpoint curve of the temperature ϑ lim be.

[0030] Furthermore, determining the threshold value may include determining the threshold value using a second mapping, which can map the speed and drive motor parameters to the power loss of the drive motor. The second mapping may be configured as a function, as represented by the following equation: P=γ(ω,ξ)

[0031] P is the power loss of the drive motor, γ is the second mapping, designed as a function, which can map the speed ω and drive motor parameter ξ to the power loss. The function γ can represent a power loss map between the power loss, the speed ω, and the drive motor parameters ξ. The drive motor parameters ξ can, for example, include the temperature of the drive motor and, alternatively or additionally, a torque of a shaft connected to the drive motor. These drive motor parameters can affect the power, in this case power loss, of the drive motor in addition to the speed. The second mapping, designed as a function, for example, can be strictly monotonically increasing with the speed, as can be expressed as follows: ∂γ∂|ω|>0

[0032] This means that the power increases as the speed increases.

[0033] Advantageously, the method can thus use physical models and relationships in the form of the first and second mappings to control the drive motor. This makes it possible to achieve a high level of accuracy when determining the speed setpoint. Derating can thus be implemented effectively and efficiently, so that the method for controlling the drive motor can achieve a convenient derating function with comfortable and high driving comfort. Furthermore, this method can be highly reused. Due to the physical models formed by the first and second mappings, the method can be combined with various other methods for controlling the drive motor and determining the nominal setpoint. The method can therefore also be combined with alternative applications with other steps for determining a nominal setpoint.This minimizes further development effort. Furthermore, calibration of this method is particularly simple because only a few parameters based on physical models are required to execute the method, such as environmental parameters and drive motor parameters. The at least one adaptable parameter does not need to be initially calibrated because the adaptable parameter is determined during implementation of the method. This reduces the development effort of the method. Simpler and better calibration of the parameters can also lead to a better coordinated transition between normal operation and derating mode. This can also increase driving comfort. Furthermore, parameterization can be simplified and accelerated, for example on a test bench.Especially compared to previous derating methods, which are not based on physical-mathematical models with physical parameterization, this can result in significantly fewer parameters needing to be parameterized on a test bench. This can then lead to fewer malfunctions in the field with the method presented here due to more suitable calibration for fewer parameters.

[0034] According to a further embodiment, the method can be characterized in that determining the power loss can comprise inverting the first mapping, wherein the inverted first mapping can be evaluated using a specific function. For this purpose, for example, the first mapping designed as an operator can be approximately left-invertible with respect to the power loss. This can mean that the approximately left-inverse Φ̂ -1of Φ with respect to the power loss exists such that the following equation applies: Φ^−1[Φ[P,ψ,μ],ψ,μ]=P

[0035] This should be at least approximately fulfilled. Thus, the curve of the power loss P lim to achieve the temperature curve ϑ lim can be estimated using the following equation: Plim=Φ^−1[ϑlim,ψ,μ]

[0036] The operator Φ̂ -1 Include interpolation functions. Interpolation functions can include characteristic curves and, alternatively or additionally, characteristic maps. Interpolation functions can include regressors. Regressors can be polynomials. Regressors can be radial basis functions. Interpolation functions can include machine learning models. Machine learning models can include neural networks. Machine learning models can include Gaussian processes. Machine learning models can include support vector machines.

[0037] The rewritten equation can be written using an explicit expression for Φ̂ -1 be evaluated. For this purpose, the underlying mathematical model, i.e., the first mapping that maps the power loss of the drive motor and loss parameters to the temperature, can be invertible without solving a differential equation for the power loss. For example, the mathematical model can be differentially flat. An explicit representation can be used for the power loss according to the equation just defined: Plim=α(δ−τ0ϑlim(0),δ−τ1ϑlim(1),…,δ−τnϑlim(n),δ−τψ,ψ,μ) where α can be a function, the notation ϑlim(i) the i-th time derivative of ϑ lim For example, ϑlim(0)=ϑlim δ can be a temporal shift operator. For example, the notation (δ τ f)(t) = f (t-τ). τ iwith i = 0, ..., n and τ ψ , can be dead times. This makes it particularly easy to evaluate the equation for the power loss, since, for example,

[0038] Solving a differential equation can be omitted. The dead times τ i with i = 0, ..., n and alternatively or additionally τ ψ can be 0 according to the explicit representation above. This can have the advantage, for example, that no prediction is required for the corresponding quantities. ϑ lim can be chosen as a constant function, which further simplifies the evaluation of the equation described above, since all derivatives of ϑlim(i) with order i>0 are identically equal to 0. The function α can contain interpolation functions. Interpolation functions can contain characteristic curves and, alternatively or additionally, characteristic maps. Interpolation functions can contain regressors. Regressors can be polynomials. Regressors can be radial basis functions. Interpolation functions can include machine learning models. Machine learning models can include neural networks. Machine learning models can include Gaussian processes. Machine learning models can include support vector machines.

[0039] Means for determining values ​​of ψ may be available. Means for prediction, and thus, for example, forecasting values ​​of parameters of ψ, may be available. Determining values ​​of parameters of ψ, i.e., the environmental parameters ψ, can be done using sensors.

[0040] Determining values ​​of ψ can involve evaluating mathematical models. Mathematical models can involve machine learning models. Predicting values ​​of parameters of ψ can involve constantly extrapolating determined values ​​of parameters of ψ into the future. Predicting values ​​of parameters of ψ can involve regressors. Regressors can be polynomials. Regressors can be radial basis functions. Predicting values ​​of parameters ψ can involve numerical simulation. Predicting values ​​of parameters ψ can involve evaluating mathematical models. Mathematical models can involve machine learning models. Predicting values ​​of parameters ψ can be done using state observers. State observers can include linear or non-linear Kalman filters or particle filters.State observers, linear Kalman filters, nonlinear Kalman filters, and, alternatively or additionally, particle filters can incorporate mathematical models. Mathematical models can be machine learning models.

[0041] Machine learning models can include neural networks. Machine learning models can include Gaussian processes. Machine learning models can include support vector machines.

[0042] Values ​​of parameters ψ can be received via a signal interface. Values ​​of parameters ψ can be determined by data processing units. Data processing units can be integrated into the drive motor. Data processing units can be integrated into the drive train and alternatively or additionally into the vehicle and mechanically connected to it. Alternatively, data processing units can be geographically independent of the vehicle. Data processing units can be, for example, a PC, a server, a central computer of the vehicle, a mobile phone or a smartwatch. Values ​​of parameters ψ can be determined based on map data and alternatively or additionally based on weather information, for example by determining ambient temperature, air humidity and alternatively or additionally based on map and weather data.For this purpose, the vehicle's coordinates can be determined, for example by means of satellite navigation or by means of localization based on neighboring radio transmission systems and, alternatively or additionally, radio reception systems.

[0043] Furthermore, determining the threshold may involve inverting the second mapping. From the monotonicity condition defined above regarding the function y, the threshold can be determined using the following equation: |ω|≤γ−1(Plim,ξ) where γ -1 a left inverse of γ with respect to the speed. Thus, the speed can be determined such that the temperature never exceeds the maximum permissible temperature. In an example case, the temperature ϑ lim equal to the maximum permissible temperature ϑ max In a preferred representation, the threshold value |ω ref | lim can be determined using the following equation: |ωref|lim=γ−1(Plim,ξ)

[0044] The function γ -1can contain interpolation functions. Interpolation functions can contain characteristic curves or characteristic maps. Interpolation functions can contain regressors. Regressors can be polynomials. Regressors can be radial basis functions. Means can be present for determining values ​​of parameters ξ, drive motor parameters. Means can be present for predicting values ​​of parameters ξ. Determinations of values ​​of parameters ξ can be carried out using sensors. Determinations of values ​​of parameters ξ can include the evaluation of mathematical models. Mathematical models can include machine learning models. Predictions of values ​​of parameters ξ can include constant extrapolation of determined values ​​of parameters ξ into the future. Predictions of values ​​of parameters ξ can include regressors. Regressors can be polynomials. Regressors can be radial basis functions.Predictions of values ​​of parameters ξ can include numerical simulations. Predictions of values ​​of parameters ξ can include the evaluation of mathematical models. Mathematical models can include machine learning models. Predictions of values ​​of parameters ξ can be made using state observers. State observers can include linear or nonlinear Kalman filters or particle filters. State observers, linear Kalman filters, nonlinear Kalman filters, and alternatively or additionally particle filters can include mathematical models. Mathematical models can be machine learning models. Machine learning models can include neural networks. Machine learning models can include Gaussian processes. Machine learning models can include support vector machines. Values ​​of parameters ξ can be received via a signal interface. Values ​​of parameters ξ can be determined by data processing units.Data processing units can be integrated into the drive motor. Data processing units can be integrated into the powertrain. Data processing units can be mechanically connected to the vehicle. Data processing units can be located independently of the vehicle. Data processing units can be a PC, server, the vehicle's central computer, a mobile phone, or a smartwatch.

[0045] Accordingly, the target values ​​of the speed can be determined using the following relationship: |ωref,nom|=|ωref|lim

[0046] In this case, amounts of the nominal setpoint |ω ref,nom | with threshold values ​​for the setpoint speed |ω ref | lim If the inequality just described applies, the setpoint values ​​of the speed ω ref the nominal setpoints are used, i.e.: ωref=ωref,nom.

[0047] Otherwise, the following relationship applies: ωref={|ωref|lim,falls ωref,nom≥0,−|ωref|lim,falls ωref,nom<0

[0048] According to a further embodiment, at least one temperature value of the element can be determined. Determining the temperature value can comprise measuring the temperature value with a sensor. Determining the temperature value can comprise reading in a temperature value. When determining at least one temperature value, exactly one or more temperature values ​​can be determined. Furthermore, at least one power loss of the drive motor can be determined. Exactly one or more values ​​for the power loss can be determined. Determining the power loss can comprise measuring the power loss. For example, a speed and a torque can be measured with sensors, and the power loss can be determined depending thereon. Determining the power loss can be performed using the second mapping as described above.Furthermore, the determination of the at least one adaptable parameter can be carried out as a function of the specific temperature value and the specific power loss. The determination of the at least one adaptable parameter can also be carried out as a function of specific environmental parameters. For example, different sets of environmental parameters can be determined for different specific temperature values ​​and different power losses. The determination of the at least one adaptable parameter can be carried out as a function of one or more sets consisting of a specific temperature value, a specific power loss, and, for example, an additional specific environmental parameter. The determination of at least one temperature value and of at least one power loss can be carried out using mathematical models. Mathematical models can contain interpolation functions.Interpolation functions can contain characteristic curves or characteristic maps. Interpolation functions can contain regressors. Regressors can be polynomials. Regressors can be radial basis functions. Interpolation functions can include machine learning models. Machine learning models can include neural networks. Machine learning models can include Gaussian processes. Machine learning models can include support vector machines.

[0049] With such a method, the at least one adaptable parameter can be determined as a function of a specific power loss and a specific temperature value. In this way, the adaptable parameter can be determined in a situation-specific manner. For example, the adaptable parameter can map how the power loss introduced at a first location in the vehicle leads to an increased temperature at a second location in the vehicle. The increased temperature can be the specific value of the temperature of the element. Different situational heat transfer, for example due to a specific vehicle or a specific vehicle speed, can thus be mapped in the adaptable parameter and used to control the drive motor for derating.It can also prevent the temperature from exceeding a certain threshold, as the temperature is determined, for example, by measuring it while the procedure is being carried out. This can ensure that the surface temperature of certain parts of the vehicle does not exceed a certain temperature. This can increase safety and comfort for the vehicle's user.

[0050] According to a further embodiment, the first mapping can map the power loss of the drive motor, environmental parameters, and the at least one adaptable parameter to the temperature. Furthermore, determining the at least one adaptable parameter can include determining an estimated value of the temperature of the element based on the first mapping and as a function of the determined power loss, certain environmental parameters, and an initial value of the adaptable parameter. ϑ̂ iFor example, the determined estimates of the temperature of the element based on the first image Φ and depending on the determined power loss P, on certain environmental parameters ψ k and an initial adaptable parameter µ and can be determined using the following formula: ϑ^i(μ)=Φ[(Pk)k−1,…,i,(ψk)k=1,…,i,μ].

[0051] This involves determining a set of estimated values ​​ϑ̂ i (µ) with sets of certain power loss (P k ) k=1,...,i , certain environmental parameters (ψ k ) k=1,...,i and the initial value of the adaptable parameter µ.

[0052] Determining the at least one adaptable parameter may comprise minimizing a metric between the determined value of the temperature ϑ i and the determined estimate of the temperature ϑ̂ iThe metric d can be a mathematical distance measure and an estimated value of the adaptable parameter can be determined using the following equation: μ^=arg minμd((ϑi)i=1,…,L,(ϑ^i(μ))i=1,…,L) with d a metric on the vector space ℝ L with ℝ the set of real numbers and µ̂ is, for example, an estimated value of at least one adaptable parameter.

[0053] In one embodiment, the metric d is a sum of squared errors: d((xi)i=1,…,L,(yi)i=1,…,L)=∑i=1L(xi−yi)2 with x i , y i ∈ ℝ, i = 1,..., L. In an alternative embodiment, the metric d is an exponentially decaying weighted sum of squared errors: d((xi)i=1,…,L,(yi)i=1,…,L)=∑i=1Lq(i−1)(xi−yi)2 with q a real-valued parameter from the open interval (0,1).

[0054] With such a method, the at least one adaptable parameter can be determined by estimating the temperature of the element and comparing this estimated temperature with, for example, a measured temperature, e.g., the determined temperature value. The comparison can be comprised in minimizing the metric. For example, these steps for determining the at least one adaptable parameter can be performed for multiple sets of specific temperatures, power losses, and environmental parameters. These multiple sets can represent a batch of data. Thus, analogous to supervised learning of a neural network, the determination of the at least one adaptable parameter can be performed using such a batch of data.

[0055] According to a further embodiment, the determination of the at least one adaptable parameter can be carried out based on at least one previously determined adaptable parameter. The determination of the at least one adaptable parameter can be carried out successively. For example, the determination of the at least one adaptable parameter can be carried out cyclically. For example, the determination of the at least one adaptable parameter can lead to an improved adaptable parameter. For example, estimated values ​​for the at least one adaptable parameter µ are determined by successive approximation, i.e., for example, adaptation. Successive approximation can be that new estimated values ​​µ̂ k for µ, with k ∈ ℤ an index and ℤ the set of integers, based on N ∈ ℕ past estimates (µ̂ k-i ) i=1,...,Nwith ℕ as the set of natural numbers for µ, for example, can be successively improved. Thus, µ̂ k be an improved and more accurate estimate for µ than previous estimates µ̂ k-i , i ≥ 1, for µ. The specific adaptable parameter can be adapted to changing product properties and / or operating conditions over time. For example, a specific estimated value µ̂ k reflect changed product properties and / or operating conditions more accurately than previous estimates µ̂ k-i , i ≥ 1, for µ. For example, the determination of the adaptable parameter depends on exactly one previously determined adaptable parameter. In such a case, the resource requirement for storing past values ​​can be particularly low. For example, to determine the at least one parameter, only one sample ϑ k the temperature ϑ, only one sample P kof power P and only one sample ψ k the environmental parameters. A sample can be a specific value. In such a case, the resource requirements for determining the samples and / or their storage can be particularly low. For example, the determination of the at least one adaptable parameter, such as a successive approximation, can be carried out using a gradient method. A gradient method is a calculation rule of the form: μ^k=μ^k−1−λk∂e∂μ(ϑk,Pk,ψk.μ^k−1) with λ k ∈ ℝ a learning rate, ∂(⋅)∂μ the gradient of the expression (.) with respect to the second parameter µ, e a model error of the operator Φ, i.e. a difference between an initial value of the operator Φ and the temperature ϑ, e(ϑ,P,ψ,μ)=Φ[P,ψ,μ]−ϑ

[0056] A gradient method has the advantage that it requires very few computational operations.

[0057] Thus, determining the adaptable parameter is a gradual improvement of the estimated adaptable parameter. To determine the power loss using the inverted first mapping and thus to determine the threshold value for an absolute value of the setpoint speed, the last determined adaptable parameter can be used. This allows for gradual improvement in the accuracy of determining the threshold value for an absolute value of the setpoint speed and thus in the method for controlling the drive motor.

[0058] A second aspect of the present invention relates to a computer program. The computer program may consist of steps of embodiments of the method shown in the first aspect of the present invention.

[0059] A third aspect of the present invention relates to a machine-readable data carrier with such a computer program.

[0060] A fourth aspect of the present invention relates to a control device configured to carry out a method according to an embodiment of the first aspect of the present invention. The control device may have a read-in interface. Thus, for example, nominal setpoints determined with another control device can be read in to then be used in the further steps of this method described here for derating and controlling the drive motor. For example, if the nominal setpoints are merely read in using the method described here, the nominal setpoints can be determined with another control device, for example a control device of the drive motor or the drive train. Alternatively, this other control device for determining the nominal setpoints can also be a PC, a server, a central computer of the vehicle, a mobile phone, or a smartwatch.Furthermore, the control device can have an output interface. This allows the method to provide specific values, such as the target speed, for example, for other control devices and for other methods. Alternatively or additionally, these specific values ​​can be transmitted and stored. Alternatively or additionally, these specific values ​​can be sent, further processed, and alternatively or additionally visualized. For this purpose, a display can be provided, for example, on the vehicle's handlebars, on the smartwatch with an application, on the mobile phone with an application, on the server, on the PC, or on the vehicle's central computer to display these specific values.For example, when derating is active, the derating can be signaled to the user visually, acoustically, or haptically via vibration on the drive motor, the drivetrain, or, for example, on the display of the vehicle's driver as the vehicle's control unit, or on another control device such as a smartwatch or the user's mobile phone. Alternatively or additionally, information can be displayed, processed, and alternatively or additionally stored, indicating when the derating was active and how intensively it was intervened, for example, how much the specified setpoint deviates from the nominal setpoint.

[0061] A fifth aspect of the present invention relates to a drive train comprising a drive motor, a shaft connected thereto, and a control device according to an embodiment of the fourth aspect of the present invention. The drive train may, for example, comprise a sensor for measuring a rotational speed of the shaft and, alternatively or additionally, a sensor for measuring a torque of the shaft.

[0062] A sixth aspect of the present invention relates to a human-powered vehicle with a drive train according to an embodiment of the fifth aspect of the present invention. Such a vehicle can be, for example, a bicycle, an e-bike, a pedelec, or a cargo bike.

[0063] According to a further embodiment, the vehicle can be characterized in that the vehicle can further comprise a user interface for reading the user input. The user input can comprise a target speed of the vehicle, a target speed of the crank, and alternatively or additionally, a target speed of the driven wheel. The target speed can be used for pushing as push assistance and alternatively or additionally for assisting while driving. The user interface can, for example, be a button, a slider, a rotary knob, a rocker switch, a touch-sensitive surface, or a touch-sensitive display attached or attachable to the handlebars and alternatively or additionally integrated into the housing of the drive train. For example, the user interface can also comprise a data interface for communication with an application, a smartwatch, or a mobile phone. Fig. 1 schematically shows steps of a method for controlling a drive motor of a muscle-powered vehicle. Fig. Figure 2 shows schematically a muscle-powered vehicle with a drive train and a control device for carrying out the functions schematically shown in Fig. 1 shown steps of the procedure.

[0064] Fig. 1 schematically shows steps of a method for controlling S5 a drive motor 4 of a muscle-powered vehicle 2. Schematically in Fig. 2, the vehicle 2 is shown with a control device 10, which is arranged to schematically Fig. 1. Furthermore, a drive train 12 is shown as part of the vehicle 2, which drive train comprises the drive motor 4, a shaft 6 mechanically connected thereto, and the control device 10. The control device 10 is communicatively connected to the drive motor 4. Furthermore, the vehicle 2 has a user interface 14, which is designed here as a rocker switch on the handlebar of the vehicle 2. The user interface 14 is communicatively connected to the control device 10. Furthermore, in Fig. Figure 2 schematically shows an element 8 of the vehicle 2, here a part of the drive motor 4. In the embodiment shown, the vehicle 2 is a pedelec with a drive motor 4 designed as an electric motor for propulsion assistance. Element 8 is a stator of the drive motor 4.

[0065] In addition, the vehicle 2 has a crank 16. In an alternative embodiment, the vehicle 2 does not have a user interface 14, but rather a sensor (not shown) for measuring a rotational speed of the crank 16, also referred to as cadence. Furthermore, the vehicle 2 has a driven wheel 18. In the alternative embodiment in which the vehicle 2 does not have a user interface 14, the vehicle 2 has a sensor (not shown) for measuring a rotational speed of the wheel 18. Both the crank 16 and the driven wheel 18 are part of the drive train 12. The sensors of the crank 16 and the wheel 18 are communicatively connected to the control device 10.

[0066] In the embodiment shown, the driver specifies a target speed via the user interface 14 as user input. This target speed refers to a speed of the driven wheel 18. The driver thus specifies a target speed of the vehicle 2. Push assistance is thus specified and selected by the driver, with the driver defining the speed for the push assistance by specifying the target speed. Alternatively, the target speed refers to a speed of the crank 16. In the aforementioned alternative embodiment, the target speed is at least proportional to a measured speed.

[0067] The method comprises determining S1 a nominal setpoint value of a rotational speed of the shaft 6 connected to the drive motor 4. The determination S1 is carried out as a function of the target rotational speed.

[0068] Furthermore, the determination of S1 of the nominal setpoint is carried out depending on a gear ratio of vehicle 2. The gear ratio is a ratio between the target speed and the nominal setpoint. The following formulas describe the relationship when determining S1 of the nominal setpoint: ωref,nom=kpedl ωpedl ωref,nom→kpedl ωpedl ωref,nom=kwhl ωwhl ωref,nom→kwhl ωwhl

[0069] Where ω ref,nom the nominal speed setpoint, k pedl is the ratio between the target speed of crank 16 and the nominal setpoint speed. k whl is the ratio between the target speed of the driven wheel 18 and the nominal speed setpoint. ω pedl is the target speed at the crank 16 and ω whlis the target speed at the driven wheel 18. In the embodiment shown here, the nominal setpoint is determined as a value proportional to the target speed, i.e., when an equation is used to determine S1 of the nominal setpoint, as defined above by the first and third expressions. In an alternative embodiment, the nominal setpoint should converge to the product of the gear ratio and the target speed, as defined above by the second and fourth expressions. The first and second expressions relate to pedal assistance, and the third and fourth expressions relate to walking assistance.

[0070] In one embodiment, the determination S1 of the nominal setpoint is performed recursively, wherein the determination S1 is stateful. The determined nominal setpoint depends on previously determined nominal setpoints.

[0071] The method comprises determining S2 at least one adaptable parameter. The adaptable parameter is determined and adapted while performing the steps of the method.

[0072] The method further comprises determining S3 a threshold value for a value of the target speed as a function of a specific temperature of element 8 of vehicle 2 and as a function of the specific parameter. The purpose of determining S3 is to ensure that the following condition applies at all times: ϑ≤ϑmax where ϑ is the specific temperature and ϑ max the maximum temperature that may prevail at the element 8 without risking overheating of the drive motor 4.

[0073] Determining S3 of the threshold value comprises determining S3.1 a power loss of the drive motor 4 using a first mapping. The first mapping is configured as an operator Φ. The first mapping maps the power loss P of the drive motor 4, environmental parameters ψ, and the at least one adaptable parameter µ to the temperature ϑ. The first mapping is described by the following formula: ϑ=Φ[P,ψ,μ]

[0074] Furthermore, determining S3 includes determining S3.2 the threshold value using a second mapping. The second mapping is configured as a function γ. This second mapping maps the speed ω and drive motor parameter ξ to the power loss P of the drive motor 4. The following relationship defines the second mapping: P=γ(ω,ξ)

[0075] Determining S3.1 the power loss involves inverting S3.1.1 of the first figure. The power loss P limis represented by the following formula: Plim=Φ^−1[ϑlim,ψ,μ]

[0076] The inverted first mapping Φ̂ -1 is evaluated using a specific function. An explicit representation can be used for the power loss: Plim=α(δ−τ0ϑlim(0),δ−τ1ϑlim(1),…,δ−τnϑlim(n),δ−τψ,ψ,μ)

[0077] Here, α is a function, ϑlim(i) is the i-th time derivative of ϑ lim , δ is a temporal shift operator and τ i with i=0, ..., n and τ ψ are dead times. Furthermore, threshold values ​​for the magnitudes of setpoints of the speed |ω ref | lim determined by the following inequality: |ωref|lim≤γ−1(Plim,ξ)

[0078] In the embodiment shown here, the following formula is used to determine the threshold values ​​in order to be able to determine the highest possible speed as the target value: |ωref|lim=γ−1(Plim,ξ)

[0079] Thus, threshold values ​​at a given permissible temperature ϑ lim , and thus certain power loss P lim , maximal. The determination S3.2 of the threshold value comprises inverting S3.2.1 the second mapping γ in order to match the inverted second mapping γ -1 to determine threshold values.

[0080] The method further comprises determining S6 at least one value of the temperature of the element 8. This determination S6 is carried out for different points in time, so that several values ​​of the temperature ϑ i , i = 1, ..., L, L ∈ ℕ are determined as samples of temperature ϑ. The determination S6 is performed using a sensor.

[0081] Furthermore, the method comprises determining S7 at least one power loss of the drive motor 4. This determination S7 is carried out for different points in time, so that several power losses P i, i = 1,..., L, L ∈ ℕ. The determination S7 is performed using sensors, whereby the speed and torque are determined, and the power losses are determined as a function of these. In an alternative embodiment, a determination S3.1 of the power losses is performed using the first mapping.

[0082] The determination S2 of the at least one adaptable parameter is carried out as a function of the determined temperature value and the determined power loss. For this purpose, the determination S2 of the at least one adaptable parameter comprises determining S2.1 an estimated temperature value ϑ̂ iof element 8 based on the first image and depending on the determined power loss, on certain environmental parameters, and on an initial value of the adaptable parameter. The initial value of the adaptable parameter is a value of the at least one adaptable parameter determined before carrying out the method. The determination S2.1 is carried out for several estimated values ​​of the temperature ϑ̂ i (µ), each depending on the respective specific values ​​of the power loss (P k ) k=1,...,i , respective environmental parameters (ψ k ) k=1,..,i and an initial adaptable parameter µ according to the formula: ϑ^i(μ)=Φ[(Pk)k−1,…,i,(ψk)k=1,…,i,μ].

[0083] Determining S2 of the at least one adaptable parameter comprises minimizing S2.2 a metric between the determined value of the temperature ϑ i and the determined estimate of the temperature ϑ̂ i(µ). According to one embodiment, the metric d is a sum of squared errors: d((xi)i=1,…,L,(yi)i=1,…,L)=∑i=1L(xi−yi)2 with xi,yi∈ℝ, i=1,…,L.

[0084] The determination S2 of the at least one adaptable parameter is carried out by means of successive approximation using a gradient method. The determination S2 of the at least one adaptable parameter µ̂ k is based on at least one previously determined adaptable parameter µ̂ k-1 according to the following formula according to the gradient method: μ^k=μ^k−1−λk∂e∂μ(ϑk,Pk,ψk,μ^k−1) with λ k ∈ ℝ a learning rate, ∂(⋅)∂μ a gradient of the expression (.) with respect to the parameter µ, e a model error of the first mapping Φ, i.e. the difference between an initial value of the first mapping Φ and the temperature ϑ, e(ϑ,P,ψ,μ)=Φ[P,ψ,μ]−ϑ

[0085] The temperature ϑ in this equation is the determined value of the temperature of element 8. Using the above expression for the estimated value of the temperature, one obtains: μ^k=μ^k−1−λk∂∂μ(ϑ^i(μ^k−1)−ϑ)

[0086] Thus, few calculations are necessary to determine S2 of the adaptable parameter.

[0087] Furthermore, the method comprises determining S4 a setpoint value of the rotational speed ω ref of wave 6. This is done through the following relationships: ωref={ωref,nom,if |ωref,nom|≤|ωref|lim,|ωref|lim,if |ωref,nom|>|ωref|lim∧ωref,nom≥0,−|ωref|lim,if |ωref,nom|>|ωref|lim∧ωref,nom<0 The equations shown above mean that the setpoint of the speed ω ref the determined nominal setpoint ω ref,nom is determined if an amount of the nominal setpoint |ω ref,nom | is less than or equal to the threshold value for an amount of the setpoint |ω ref | lim. For this purpose, a comparison of the nominal setpoint and the threshold value is carried out during determination S4. In the embodiment shown here, nominal setpoints are also determined as setpoints of the speed if the magnitude of the nominal setpoint is equal to the threshold value for the magnitude of the setpoint. In another embodiment, only nominal setpoints are determined as setpoints if the magnitudes of the nominal setpoints are smaller than the threshold values ​​of the magnitudes of the setpoint. Otherwise, i.e. if the relationship |ωref,nom|≤|ωref|lim does not apply, the setpoint is determined such that an amount of the setpoint corresponds to the determined threshold value and a sign of the setpoint corresponds to a sign of the determined nominal setpoint.

[0088] The method further comprises controlling S5 the drive motor 4 as a function of the determined target speed. The control device 10 is configured to determine control parameters for the control S5 as a function of the determined target speed. The control device 10 is further configured to send these determined control parameters to the drive motor 4. During this control S5, the drive motor 4 is controlled such that the speed of the shaft 6 is adjusted to the determined target speed. In an alternative embodiment, the speed of the shaft 6 converges towards the determined target speed.

[0089] The method is used to control S5 the drive motor 4, determining the setpoint speed and thus implementing an adaptive derating function. The derating function is adaptive because the adaptable parameter µ is successively adapted and approximated more precisely. This allows the determination S3 of the threshold value for an amount of the setpoint speed to be carried out more precisely. The determination S3 is carried out using a situationally determined parameter. This allows a situational heat transfer from introduced power loss to the temperature of element 8 to be mapped to the adaptable parameter, so that derating can be carried out more precisely. The user of the vehicle 2 can thus be protected, for example, from excessively hot temperatures at element 8. A situational cooling capacity can thus be reflected in the adaptable parameter.

[0090] The target speed is determined in such a way that the speed of the shaft 6 is controlled as a function of the target speed of the crank 16 during pedal assistance, and alternatively, the speed of the shaft 6 is controlled as a function of the target speed of the driven wheel 18 during push assistance. This increases the driving comfort for the driver of the vehicle 2, since the pedals of the crank 16 are neither pushed into the driver's legs by the control S5 of the drive motor 4, nor does the vehicle 2 impart a shock via the handlebars to the driver when pushing. Reference symbol 2 vehicles 4 drive motor 6 Wave 8 elements 10 Control device 12 Drivetrain 14 User interface 16 crank 18 driven wheel S1 Determining a nominal setpoint of a speed S2 Determine at least one adaptable parameter S2.1 Determining an estimated temperature of an element of the vehicle S2.2 Minimize a metric between a given value of temperature and the given estimate of temperature S3 Determining a threshold value for an amount of the setpoint speed S3.1, S7 Determining the power loss of the drive motor S3.1.1 Inverting the first image S3.2 Determining the threshold using a second image S3.2.1 Inverting the second image S4 Determining a setpoint for the shaft speed S5 Control of the drive motor S6 Determining at least one value of the temperature of the element QUOTES CONTAINED IN THE DESCRIPTION

[0000] This list of documents submitted by the applicant was generated automatically and is included solely for the convenience of the reader. This list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions. Cited patent literature

[0000] DE 10 2022 200 541 A1

[0002]

Claims

[1] A method for controlling (S5) a drive motor (4) of a muscle-powered vehicle (2), comprising the steps of: determining (S1) a nominal setpoint value of a rotational speed of a shaft (6) connected to the drive motor (4); determining (S2) at least one adaptable parameter; determining (S3) a threshold value for an absolute value of the setpoint value of the rotational speed as a function of a specific temperature of an element (8) of the vehicle (2) and as a function of the at least one specific parameter; determining (S4) a setpoint value of the rotational speed of the shaft (6), wherein the specific nominal setpoint value is determined as the setpoint value if an absolute value of the nominal setpoint value is smaller than the threshold value for an absolute value of the setpoint value, and otherwise the setpoint value is determined such that an absolute value of the setpoint value corresponds to the specific threshold value and a sign of the setpoint value corresponds to a sign of the specific nominal setpoint value;and controlling (S5) the drive motor (4) as a function of the determined target value of the speed; [2] Method according to claim 1, characterized by that the determination (S1) of the nominal setpoint is carried out as a function of a target speed. [3] Method according to one of the preceding claims, characterized by that the determination (S1) of the nominal setpoint is carried out recursively. [4] Method according to one of the preceding claims, characterized by that determining (S3) the threshold value comprises determining (S3.1) a power loss of the drive motor (4) by means of a first mapping which maps the power loss of the drive motor (4), environmental parameters and the at least one adaptable parameter to the temperature, and further determining (S3.2) the threshold value by means of a second mapping which maps rotational speed and drive motor parameters to the power loss of the drive motor (4). [5] Method according to one of the preceding claims, characterized by that a determination (S6) of at least one value of the temperature of the element (8) is carried out, that a determination (S3.1; S7) of at least one power loss of the drive motor (4) is carried out and that the determination (S2) of the at least one adaptable parameter is carried out as a function of the determined value of the temperature and the determined power loss. [6] Method according to claim 5, characterized byin that a first mapping maps the power loss of the drive motor (4), environmental parameters and the at least one adaptable parameter to the temperature, in that the determination (S2) of the at least one adaptable parameter comprises determining (S2.1) an estimated value of the temperature of the element (8) based on the first mapping and as a function of the determined power loss, of certain environmental parameters and of an initial value of the adaptable parameter, and in that the determination (S2) of the at least one adaptable parameter comprises minimizing (S2.2) a metric between the determined value of the temperature and the determined estimated value of the temperature. [7] Method according to claim 6, characterized by that the determination (S2) of the at least one adaptable parameter is carried out based on at least one previously determined adaptable parameter. [8] Control device (10) which is arranged to carry out a method according to the preceding claims. [9] Drive train (12) with a drive motor (4), a shaft (6) connected thereto and a control device (10) according to claim 8. [10] Muscle-powered vehicle (2) with a drive train (12) according to claim 9.

Images