METHOD FOR CONTROLLING AN ELECTRIC MOTOR OF AN ELECTRIC BICYCLE

Description

Technical area

[0001] The technical field relates to methods for controlling an electric motor of an electric bicycle. State of the art

[0002] Methods for controlling an electric motor of an electric bicycle are known from the prior art, wherein the electric motor is controlled as a function of a movement of the crank.

[0003] EP 2 604 499 A1 relates to an electrically assisted bicycle capable of efficiently applying input torque from a motor to pedaling force. Such a bicycle comprises a pedaling force sensor for detecting the pedaling force applied to a pedal, a motor for assisting the driving force based on the output of the pedaling force sensor, a crank angle sensor for detecting the crank angle of a crankshaft connected to the pedal, a means for calculating the rotational frequency of the crankshaft from the crank angle, and a control unit for adjusting the assist force generated by the motor depending on the phase of the crank angle and the calculated rotational frequency. Description of the invention

[0004] The invention relates to a method for controlling an electric motor of an electric bicycle. The electric bicycle may have a battery as an energy storage device. The electric motor may have a motor control unit, and the motor control unit may be configured to execute the method for controlling the electric motor. The motor control unit may be configured to control a power supply to the electric motor with electrical energy stored in the battery by means of the method.

[0005] The bicycle has an electric motor and a crank. A chain can drive a rear wheel of the bicycle via a chainring connected to the crank. The electric motor and crank can be connected in such a way that the electric motor can influence the rotational movement of the crank. The crank can have a pedal for a cyclist's foot at each end of the crank. The cyclist can apply input torque to the crank using muscle power. The electric motor can apply additional torque to the crank. The sum of the input torque and the additional torque of the electric motor can drive the rear wheel via the chain.

[0006] The method comprises a step of detecting the input torque at at least two different points in time. The input torque can be detected periodically and, alternatively or additionally, the input torque can be detected at more than two different points in time. The bicycle can have a sensor with which the input torque can be detected. The method further comprises a step of detecting a cadence of the crank at at least two different points in time. The cadence of the crank can be detected periodically and, alternatively or additionally, the cadence can be detected at more than two different points in time. The bicycle can have a sensor with which the cadence of the crank can be detected.

[0007] The method further comprises a step of determining a temporal change in the input torque based on the detected input torque at at least two different points in time. The method further comprises a step of determining a temporal change in the cadence based on the detected cadence at at least two different points in time. The step of determining the temporal change in the input torque and, alternatively or additionally, the temporal change in the cadence can be carried out based on all detected input torques and all detected cadences. In the determining step, a function of the input torque and, alternatively or additionally, a function of the cadence can be determined as a function of time.In the determining step, a function of the temporal change in the input torque and, alternatively or additionally, a function of the temporal change in the cadence can be determined as a function of time. Alternatively or additionally, in the step of determining the temporal changes in the input torque and cadence, a relative change in the detected value of the input torque or cadence compared to the previously detected value can be determined.

[0008] The method further comprises a step of determining a base torque as a function of the temporal change in the input torque and as a function of the temporal change in the cadence. The step of determining the base torque can be performed iteratively, wherein the base torque from the n-1st step of determining the base torque can be used for an nth step of determining the base torque. In other words, the step of determining the base torque can also represent a filtering of the base torque from previous steps of determining the base torque. The step of determining the base torque can be performed as a function of the input torque, in particular as a function of the absolute value of the input torque. Thus, in a first step of determining the base torque, the base torque can be assumed to be equal to the input torque, and the iteration can be started.A dynamic function can be used to reduce the iteration steps. This allows a base torque to be determined more quickly or filtered in response to changes in the input torque and cadence over time.

[0009] In the step of determining the base torque, the dependence of the base torque on the temporal change in the input torque decreases with increasing cadence. The dependence of the base torque on the temporal change in the input torque as a function of cadence can be represented as a monotonically decreasing or strictly monotonically decreasing function. The function can be represented by a linear, polynomial, or exponential function. In other words, the base torque may depend more strongly on the temporal change in the input torque at smaller cadences than at larger cadences. The weighting of the change in the input torque when determining the base torque can decrease with increasing cadence.

[0010] In the step of determining the base torque, the dependence of the base torque on the temporal change in cadence increases with increasing cadence. The dependence of the base torque on the temporal change in cadence as a function of cadence can be represented as a monotonically increasing or strictly monotonically increasing function. The function can be represented by a linear, polynomial, or exponential function. In other words, the base torque may depend less strongly on the temporal change in cadence at lower cadences than at higher cadences. The weighting of the change in cadence when determining the base torque can increase with increasing cadence.

[0011] The sum of the weights for the changes in input torque and cadence may be constant. In other words, the weights may be inversely related to each other. If the weight for the change in torque over time is greater at lower cadences, the weight for cadence may be lower. Conversely, the weight for torque may be lower at higher cadences, and the weight for cadence may be greater.

[0012] The base torque can be determined using one or more tables. For example, a table can describe the dependence of the base torque on the change in the input torque over time. Columns in the table can describe relative changes in the base torque with respect to a column-specific change in the input torque over time. These relative changes in the base torque for different cadences can be described in the rows. The table can contain entries for a relative change in the base torque for discrete values of the change in the input torque and cadence. Entries can also be zero if no relative change in the base torque is to occur for the associated change in the input torque and cadence. Entries can be positive or negative, which can increase or decrease the base torque.The current base torque can be determined using the base torque from the previous determination step and the relative change in the base torque determined from the table. A similarly structured table can describe the dependence of the base torque on the temporal change in the cadence. The entries can behave inversely to one another, according to the inverse weightings described above. The relative changes in the base torque taken from the two tables can be offset against one another to obtain an overall relative change in the base torque. The dynamic function can contain tables with modified values for the relative change in the base torque. This allows the base torque to be determined to be reached or filtered more quickly.

[0013] The method further comprises a step of controlling the electric motor with a target torque as a function of the base torque. The target torque can be determined after the base torque has been determined and as a function of the base torque. The target torque can be equal to the base torque. The target torque can represent the torque to be provided by the electric motor, which, together with the rider's input torque, can represent the total torque. The total torque, together with the cadence, can represent the total power that can be used to propel the bicycle. After the base torque has been determined and before controlling with the target torque, the base torque can still be changed; in particular, the base torque can be changed before the target torque is determined.Thus, the target torque determined based on the changed base torque can be used to control the electric motor. Alternatively or additionally, the determined target torque can be changed before controlling the electric motor.

[0014] Advantageously, a method for controlling an electric motor of an electric bicycle can thus be shown, in which the control takes place at different cadences with differently weighted dependencies on the temporal changes in the input torque and the cadence. This makes it easier to recognize a rider's command regarding the control of the electric motor, whereby the rider's command can only be detected via the movement of the crank. The dependency of the base torque can be such that at low cadences there is a high dependency on the temporal change in the input torque. At low cadences, the input torque can be relatively large due to a general power relationship between cadence and input torque. The general power relationship can describe the power applied by the rider, given by cadence and input torque at the crank, which can be constant over wide cadence ranges.For example, high input torque can prevail at low cadences and lower input torque at higher cadences, which can lead to almost constant power. A change in input torque can be detected better at high input torque values. This means that at low cadences, a driver's desired change can be better taken into account by the method due to an input torque that changes over time. At higher cadences, it can be advantageous for the dependence of the base torque on the change in cadence over time to increase with increasing cadence, since the cadence also increases in absolute value and a change in cadence can be better detected and determined. This means that at high cadences, a driver's desired change can be more easily recognized via the changing cadence.

[0015] According to a further embodiment, the base torque can be determined depending on a riding condition. The riding condition can be at least one of standstill, starting, riding, accelerating, kickdown, rolling without motor assistance, and boost. The riding condition can be determined depending on a relative speed of the bicycle relative to the ground. Thus, in different riding conditions, the dependencies of the base torque on the temporal change in the input torque and on the temporal change in cadence as a function of cadence can differ. For each riding condition, riding condition-specific tables can be provided regarding the dependency between base torque and temporal change in the input torque and cadence. Functions that describe the weightings of the changes in the input torque and cadence against the cadence can be different for different riding conditions.For example, in the starting state, the base torque may depend more strongly on the temporal change of the input torque than with the same cadence in the accelerating state while riding, i.e. with an already existing relative speed of the bicycle compared to the ground.

[0016] Advantageously, the method can thus shift the weighting of the temporal changes in input torque and cadence depending on the driving conditions so that the driver's input can be optimally detected in each driving condition based on the changes in input torque and cadence. For example, when starting off, it may be more appropriate to give greater weight to the input torque at a low cadence, and when accelerating while moving, it may be more appropriate to give greater weight to the change in cadence when the absolute cadence is otherwise the same to determine the base torque. This allows the driver's input to be optimally detected in each driving condition, and thus the electric motor can be optimally controlled according to the driving condition and driver's input.

[0017] According to a further embodiment, the method can further comprise a step of determining a crank position as a function of the temporal change in the input torque. The step of determining the crank position can take place as a function of a temporal progression of the input torque. The temporal progression can be determined using values of the input torque recorded over time. For example, a sinusoidal function can describe the input torque as a function of time. Here, in particular, a sine-squared function can describe the input torque as a function of time. The local maxima of the function can describe the two horizontal positions of the crank over the course of a complete revolution of the crank, since, assuming a constant force from the driver on the pedal, the input torque is at a maximum in these crank positions due to the maximum lever.

[0018] Advantageously, even in cases where the engine control unit has no information regarding the absolute crank position, the crank position can be determined by means of the determined temporal change of the input torque.

[0019] According to a further embodiment, the step of determining the base torque can include a step of adjusting the input torque for the crank position. The crank position can be stored in a memory, which can be accessed by the engine control unit executing the method. Alternatively or additionally, the crank position can be determined in the step of determining the crank position. Due to the lever between a bottom bracket of the crank and the pedal, which changes sinusoidally with one crank revolution, the input torque can also change. Alternatively or additionally, the input torque can change due to the changing force applied by the rider to the pedal. Thus, at an angular position between 0°, where 0° describes the crank at its highest point, and 45°, the rider can exert no or only minimal force on the pedal. The same applies to angular positions between 135 and 180°.In other words, the base torque can only depend on input torques at crank positions between 45° and 135°. The adjustment step can be performed for input torques above a certain limit cadence, since below the limit cadence, for example, no crank position can be determined.

[0020] Thus, the method can advantageously be designed such that only a subset of detected input torques describing the driving behavior and the desired change can be used to determine the base torque. Thus, the base torque can be independent of input torque values that can be detected at a crank position that cannot reflect the driver's actual input torque, such as 0° - 45° and, alternatively or additionally, 135° - 180°. This allows the driver's input to be detected more precisely, and the control of the electric motor with the target torque can correspond even better to the actual driver's input.

[0021] According to a further embodiment, the method may further comprise a step of adjusting the base torque depending on a driving mode. The adjusting may include increasing, maintaining constant, or decreasing the determined base torque. The driving mode may adopt a specific mode depending on physical driving variables, such as an incline, lateral acceleration, an incline, and alternatively or additionally the input torque. The driving mode may be one of an ECO mode, a touring mode, a trail mode, a user-specific mode, and an auto mode. For example, in ECO mode, the base torque may be reduced, and in touring mode, the base torque may be increased by a factor of 2.5. The factor by which the base torque can be adjusted depending on the driving mode may be referred to as a gain factor.The driving mode can be specified by the driver through an active driver request via a user interface, for example via a switch on the handlebar.

[0022] Advantageously, the base torque and thus the control of the electric motor with the target torque can be adjusted from a user request actively specified by the driver and via a user interface.

[0023] Adjusting the base torque as a function of the driving mode, like further steps of adjusting the base torque as a function of other variables, can occur after the step of determining the base torque and before the step of determining the target torque as a function of the base torque. Determining the target torque can occur as a function of the adjusted base torque. One or more steps of adjusting the base torque as a function of different variables can occur independently of one another and in any order.

[0024] According to a further embodiment, the method can further comprise a step of adjusting the base torque as a function of the cadence, wherein the step of adjusting as a function of the cadence can be carried out independently of the detected input torque. In other words, after determining the base torque as a function of the temporal changes in the input torque and the cadence, the determined base torque can be adjusted as a function of the cadence. The adjustment can comprise increasing, maintaining constant, or decreasing the determined base torque. The adjustment of the base torque as a function of the cadence can be carried out differently in different modes. For example, in a first mode, the base torque can be adjusted constantly across all cadences by a factor greater than 1 for a cadence not equal to 0, i.e., increased.In a second mode, the base torque can be increased with increasing cadences, depending on the cadence, with factors greater than 1.

[0025] Advantageously, the absolute value of the cadence can thus be weighted when adjusting the base torque.

[0026] According to a further embodiment, the method may further comprise a step of adjusting the base torque depending on a speed of the bicycle. The adjustment may comprise increasing, maintaining constant, or decreasing the determined base torque. The adjustment of the base torque depending on the speed may occur differently in different modes. For example, in a first mode, the base torque may be constantly increased by a factor greater than 1 for all speeds other than 0. In a second mode, the base torque may be increased by a factor that increases with speed. Alternatively or additionally, the method may further comprise a step of adjusting the base torque depending on a driving resistance during travel. The driving resistance may comprise air resistance and, alternatively or additionally, rolling friction resistance.

[0027] Advantageously, the method can thus adapt the base torque to resistance due to rolling and air friction increasing with increasing speed and thus counteract the resistance.

[0028] According to a further embodiment, the method can further comprise a step of adapting the base torque depending on the gradient of a driving surface. The adapting can comprise increasing, maintaining a constant, or reducing the determined base torque. As the gradient of the driving surface increases positively, i.e., when driving uphill, the base torque can have an increasingly increasing factor greater than 1 for adaptation. As gradients decrease and become negative, i.e., when driving downhill and on an increasingly steep hill, the base torque can be adapted with a factor less than 1 and increasingly smaller. In other words, when driving uphill and on an increasingly steep uphill, the base torque can be increasingly increased, and when driving downhill and on an increasingly steep downhill, the base torque can be increasingly reduced.The step of adjusting the base torque depending on the gradient can be performed differently in different modes. In a first mode, the gradient-dependent factors for adjusting the base torque can be multiples of factors in a second mode. Thus, the gradient-related assistance can be performed differently in different modes. The mode can be selected by the driver via active user input via a user interface.

[0029] Thus, the method can advantageously take into account the gradient of the driving surface and the target torque can be determined, adapted to the gradient, and used to control the electric motor.

[0030] According to a further embodiment, the method can further comprise a step of adjusting the base torque depending on an inclination of the bicycle relative to the direction of the weight force. The adjustment can comprise increasing, maintaining a constant, or decreasing the determined base torque. The inclination can, for example, be defined in positive values to the right and in negative values to the left. The bicycle can have an inclination in one direction or the other when cornering. For example, if the inclination increases in absolute values, i.e., regardless of the direction of inclination, the base torque can be adjusted and thus reduced by a factor less than 1 that decreases with increasing inclination. Different factors can be used in different modes for the same inclination angles.

[0031] Advantageously, the method can thus adapt the base torque when cornering or in other riding conditions in which the bicycle may be tilted, thus determining the target torque for controlling the electric motor. For example, particularly strong acceleration can be avoided when cornering, allowing the method to respond to the lateral accelerations occurring when cornering with reduced longitudinal accelerations compared to straight-line travel due to the activation of the electric motor. This prevents the bicycle's wheels from skidding or slipping due to excessive acceleration. Different factors can be used in different modes for the same tilt angles.For example, sportier riders who can also handle higher assistance on steep inclines can choose a mode in which the incline-dependent factors for adjusting the base torque can lead to larger values of the base torque.

[0032] According to a further embodiment, before the step of controlling the electric motor with the target torque, the target torque can be limited to a maximum value in a limiting step. The maximum value can be stored in a memory and read out by the motor control unit executing the method. The maximum value can be absolute. The maximum value can comprise several individual maximum values, each of which must not be exceeded individually. Alternatively or additionally, the maximum value can be relative, so that a maximum power, calculated from the target torque and cadence, cannot be exceeded. Alternatively or additionally, the maximum value can depend on external variables, such as the speed of the bicycle over ground, the temperature, or external signals.External signals can be, for example, signals from traffic control systems such as traffic lights and information regarding traffic flow, such as a green wave. This allows the electric motor to be controlled depending on the traffic situation. Alternatively or additionally, the electric motor can be controlled depending on the outside temperature, so that a temperature-dependent maximum power of the electric motor cannot be exceeded and permanent damage to the electric motor can be avoided. The dependence of the maximum value on speed can be due to a legal requirement. The maximum value can be dependent on position data, such as GPS data. This allows range control to be implemented by the process. The maximum value can be dependent on data for accident prevention.For example, the maximum value can be dependent on information from another vehicle in the immediate vicinity, for example to avoid a collision. The maximum value can alternatively or additionally depend on ABS data so that the bicycle can remain maneuverable. The maximum value can be dependent on the cadence. For example, if the cadence is 0, the maximum value of the target torque can be set to 0. Through user input, the rider can activate a push assist on the bicycle, whereby the maximum value of the target torque can be set so that the electric motor can be controlled even if the cadence is 0, allowing the bicycle to accelerate to a maximum speed of around 6 km / h.

[0033] Alternatively or in addition to the step of limiting the target torque, the method may include a step of increasing the target torque to a minimum value. For example, the increase may be dependent on the driver's health data. For example, the driver's input power may be detected, and the target torque may be increased in such a way that the driver's maximum power is not exceeded.

[0034] A further aspect of the invention relates to an engine control unit. The engine control unit can be configured to execute the method according to an embodiment of the preceding aspect of the invention for controlling an electric motor.

[0035] Another aspect of the invention relates to a bicycle comprising an electric motor and a motor control unit according to the previous aspect of the invention. The bicycle may further comprise a crank, pedals, a battery, a user interface, and / or cabling for connecting the battery, user interface, electric motor, and motor control unit. Short description of the characters

[0036] Figure 1 schematically shows steps of a method for controlling an electric motor according to one embodiment. Figure 2 schematically shows the weighting of the dependencies for determining the base torque in the method for controlling an electric motor according to one embodiment. Detailed description of embodiments

[0037] Figure 1schematically shows steps of a method for controlling an electric motor according to one embodiment. The electric motor is part of an electric bicycle, and the bicycle has a crank. The electric motor is connected to the crank. The bicycle has a motor control unit for controlling the electric motor, wherein the motor control unit is configured to carry out the method. The bicycle has a torque sensor, attached to the crank, for detecting S1.1 an input torque. The input torque is a torque applied to the crank by a rider of the bicycle. The sensor detects the input torque periodically and thus at multiple points in time. Another sensor of the bicycle detects the cadence of the crank. The detection S1.2 of the cadence of the crank occurs periodically and thus for multiple points in time. The detection S1.1 of the input torque and the detection S1.2 of the cadence occur independently of one another.

[0038] The method comprises a step of determining S2.1 a temporal change in the input torque. The step of determining S2.1 is carried out based on the detected input torque at different points in time. The method comprises a step of determining S2.2 a temporal change in the cadence. The step of determining S2.2 is carried out based on the detected cadence at different points in time. The steps of determining S2.1, S2.2 the temporal changes in the input torque and the cadence are carried out independently of one another. The temporal changes in the input torque and the cadence are relative changes in the input torque and cadence to previously detected values of input torque and cadence. Depending on the step of determining S2.1 the temporal change in the input torque, a step of determining S2.3 a crank position is carried out. The determination of the crank position S2.3 is dependent on a temporal progression of the input torque. The temporal progression of the input torque depends on the specific temporal change of the input torque.

[0039] The method comprises a step S3 of determining a base torque. The step S3 of determining the base torque is performed as a function of the temporal change in the input torque and as a function of the temporal change in the cadence. Figure 2shows a weighting of the dependencies. The weightings of the dependencies are plotted along the vertical axis B, the temporal change in the input torque, curve C, and the temporal change in the cadence, curve D, are plotted as a function of the cadence, shown on the horizontal axis A. The weightings determine the dependencies when determining S3 the base torque. For example, at low cadences, the base torque depends strongly on the temporal change in the input torque. With increasing cadences, this weighting decreases. At the same time, the weighting and thus the dependence of the base torque on the temporal change in cadence increases with increasing cadence.Since the product of cadence and torque results in power and the power applied by the driver can be assumed to be almost constant over large parts of the ride, it follows that the summed weightings of the temporal changes in input torque and cadence should be almost constant over large parts of the cadence.

[0040] The method includes a step S3.1 of adjusting the input torque for the crank position. The adjusting step S3.1 occurs during the step S3 of determining the base torque, so that the base torque determined in the step S3 of determining is adjusted for the crank position.

[0041] The method further comprises a step S4.1 of adjusting the base torque depending on the riding mode. The method further comprises a step S4.2 of adjusting the base torque depending on the cadence, wherein the adjusting step S4.2 occurs independently of the detected input torque. The method further comprises a step S4.3 of adjusting the base torque depending on the speed of the bicycle. The method further comprises a step S4.4 of adjusting the base torque depending on the gradient of the riding surface. The method further comprises a step S4.5 of adjusting the base torque depending on the inclination of the bicycle relative to the direction of the weight force. The adjusting steps S4.1 - S4.5 occur simultaneously, wherein each adjusting step represents a single factor of a product for adjusting the base torque. The adjusting steps S4.1 - S4.5 are performed independently of each other. Thus, in one case, only a subset of the adjustment steps S4.1 - S4.5 can be performed. After the last adjustment step, S4.1 - S4.5, a target torque is determined based on the determined and adjusted base torque.

[0042] The method further includes a step of limiting S5 the target torque to a maximum value. The maximum value is an absolute value in newton meters. The maximum value represents the maximum value of the electric motor, which is limited by its design. By limiting S5 the target torque to the maximum value, continuous use of the electric motor up to this maximum value can be ensured without the risk of damage occurring.

[0043] The method includes a step S6 of controlling the electric motor with the limited target torque. Here, the supply of electrical energy from a battery to the electric motor is controlled according to the target torque so that the electric motor can provide the target torque. Reference symbol

[0044] S1.1(step) Detecting an input torque S1.2(step) Detecting a cadence of the crank S2.1(step) Determining a change in the input torque over time S2.2(step) Determining a change in the cadence over time S2.3(step) Determining a crank position S3(step) Determining a base torque S3.1(step) Adjusting the input torque for the crank position S4.1(step) Adjusting the base torque depending on a riding mode S4.2(step) Adjusting the base torque depending on the cadence S4.3(step) Adjusting the base torque depending on a speed of the bicycle S4.4(step) Adjusting the base torque depending on an incline of a riding surface S4.5(step) Adjusting the base torque depending on the inclination of the bicycle S5(step) Limiting a target torque S6(step) Controlling the electric motor with the target torque depending on the base torque ACadence BWeighting of the dependencies CTime-dependent change in the input torque DTime-dependent change in the cadence.

Claims

1. Method for controlling an electric motor of an electric bicycle, wherein the bicycle comprises the electric motor and a crank, wherein the method comprises the steps of: - detecting (S1.1) an input torque at at least two different points in time, wherein the input torque is a torque applied to the crank by a rider of the bicycle, - detecting (S1.2) a cadence of the crank at at least two different points in time, characterized by the steps of - determining (S2.1) a change in the input torque with respect to time based on the detected input torque at at least two different points in time, - determining (S2.2) a change in the cadence with respect to time based on the detected cadence at at least two different points in time, - determining (S3) a base torque depending on the change in the input torque with respect to time and depending on the change in the cadence with respect to time, wherein in the determining step (S3) the dependence of the base torque on the change in the input torque with respect to time decreases as the cadence increases, and in the determining step (S3) the dependence of the base torque on the change in the cadence with respect to time increases as the cadence increases, and - driving (S6) the electric motor with a target torque depending on the base torque.

2. Method according to Claim 1, characterized by determining (S3) the base torque depending on a riding state.

3. Method according to either of Claims 1 and 2, characterized in that the method further comprises a step of determining (S2.3) a crank position depending on the change in the input torque with respect to time.

4. Method according to any of the preceding claims, characterized in that the step of determining (S3) the base torque comprises a step of adjusting (S3.1) the input torque about the crank position.

5. Method according to any of the preceding claims, characterized in that the method further comprises a step of adapting (S4.1) the base torque depending on a riding mode.

6. Method according to any of the preceding claims, characterized in that the method further comprises a step of adapting (S4.2) the base torque depending on the cadence, wherein the step of adapting (S4.2) depending on the cadence is performed independently of the detected input torque.

7. Method according to any of the preceding claims, characterized in that the method further comprises a step of adapting (S4.3) the base torque depending on a speed of the bicycle.

8. Method according to any of the preceding claims, characterized in that the method further comprises a step of adapting (S4.4) the base torque depending on a gradient of a surface being ridden on.

9. Method according to any of the preceding claims, characterized in that the method further comprises a step of adapting (S4.5) the base torque depending on an inclination of the bicycle relative to the direction of the weight.

10. Method according to any of the preceding claims, characterized in that, before the step of driving (S6) the electric motor with the target torque, the target torque is limited to a maximum value in a limiting step (S5).

11. Motor control unit which comprises means for executing the method according to any of the preceding claims for controlling an electric motor.

12. Bicycle, which comprises an electric motor and the motor control unit according to Claim 11.