JPH0986476A - Bicycle with electric motor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To generate motor output torque corresponding to human power torque without using a torque detection mechanism. SOLUTION: An electric motor 2 for assisting a human power drive force is connected to the wheel of a bicycle mechanism 1 operable with human power, and a rotational angle sensor 3 for detecting a wheel rotational angle is fitted to the wheel. In addition, a power assist controller 4 is mounted for controlling the output torque of the electric motor 2, on the basis of the wheel rotational angle detected with the rotational angle sensor 3. Furthermore, an inverse function for a bicycle mechanism transfer function is established in the power assist controller 4.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a bicycle equipped with an electric motor for assisting the driving force by human power by connecting an electric motor to a bicycle mechanism that can be driven by human power.

[0002]

13 and 14 show the structure of a conventional control system for a bicycle with an electric motor. The manual torque generated by stepping on the pedal is detected by the torque detection mechanism (5), and the detection signal is the motor torque adjustment circuit (6).
Sent to According to this, motor torque adjustment circuit (6)
Then, a motor torque command corresponding to the input torque detection signal is created and supplied to the electric motor (2). As a result, the bicycle mechanism (1) is supplied with the motor output torque of a magnitude corresponding to the human power torque value in addition to the human power torque,
The driving force by human power is assisted. As the torque detection mechanism (5), for example, a method of connecting a planetary gear mechanism to an output shaft of an electric motor to detect torque between gears (for example, JP-A-5-58379), torque between a pedal and a sprocket, and A method of detecting by torsion of a torsion bar (for example, Japanese Patent Laid-Open No. 4-321482) is known.

[0003]

However, since the structure of the torque detecting mechanism is generally complicated, there is a problem in that the structure of the entire electric motor-equipped bicycle becomes complicated and the weight becomes large. An object of the present invention is to provide a bicycle with an electric motor having a simple structure capable of generating a motor output torque according to a human torque without using a torque detection mechanism.

[0004]

In the bicycle with an electric motor according to the present invention, an electric motor is connected to a bicycle mechanism that can be driven manually, and a signal corresponding to a wheel rotation angle or a rotation angular velocity of the bicycle mechanism is detected. And a control circuit that controls the output torque of the electric motor based on the sensor detection signal.

The control circuit described above is based on the inverse function of the bicycle mechanism transfer function, in which the torque applied to the bicycle mechanism by the human power and the electric motor is an input signal, and the signal corresponding to the wheel rotation angle or the rotational angular velocity is an output signal. First arithmetic processing means for deriving an input torque value for the bicycle mechanism from the sensor detection signal, and second for deriving a manual torque value by subtracting the output torque value of the electric motor from the derived input torque value for the bicycle mechanism. It is provided with an arithmetic processing means and a motor torque adjusting means for creating a motor torque command according to the derived manual torque value and supplying it to the electric motor.

In the bicycle with the electric motor of the present invention, the human power and the torque generated by the electric motor are input to the bicycle mechanism, and the sensor detects a signal corresponding to the wheel rotation angle or the rotation angular velocity of the bicycle mechanism. The sensor detection signal is input to the first arithmetic processing means. Here, since the first arithmetic processing means has an inverse function of the bicycle mechanism transfer function as its transfer function, the input / output relationship is opposite to that of the bicycle mechanism transfer function, and the wheel rotation angle or rotation angular velocity is changed. By inputting the corresponding sensor detection signal (output value of the bicycle mechanism transfer function), the input torque value to the bicycle mechanism, that is, the total value of the torque component due to human power and the torque component due to the electric motor (input value of the bicycle mechanism transfer function) ) Will be derived. The input torque value for the bicycle mechanism thus derived is input to the second arithmetic processing means,
The torque component due to human power is obtained by subtracting the torque component due to the electric motor (output torque value of the electric motor).
(Human torque value) is derived. Therefore, the manual torque value can be obtained without using the torque detection mechanism.
The manpower torque value thus derived is input to the motor torque adjusting means, a motor torque command is created in accordance with the manpower torque value, and the motor torque command is supplied to the electric motor. As a result, the electric motor generates a motor output torque according to the manual torque.

Specifically, the bicycle mechanism transfer function uses a human power and a torque generated by an electric motor as input signals, and a subtraction processing unit for subtracting the loss torque value from the input signals, and the inertial force of the bicycle mechanism and the subtraction result. And a transfer function unit for deriving a rotational angular velocity by performing a calculation process using a transfer function having a viscous resistance as a parameter.

In this specific configuration, the torque generated by the human power and the electric motor is input to the subtraction processing unit, and the loss torque value is subtracted, whereby the driving torque value of the bicycle mechanism that actually contributes to bicycle running is calculated. It The calculated driving torque value of the bicycle mechanism is input to the transfer function unit. Here, the relationship between the drive torque of the bicycle mechanism and the wheel rotation angular velocity can be defined by using the inertial force and the viscous resistance of the bicycle mechanism as parameters. Therefore, the wheel rotation angular velocity in actual bicycle traveling is derived from the calculated driving torque value by the transfer function unit having this relationship as the transfer function.

Further, specifically, a second transfer function unit for deriving a first loss torque component due to air resistance received when the bicycle is traveling, based on the derived wheel rotation angular velocity, is provided at the subsequent stage of the transfer function unit. It is connected and the first loss torque component is input to the subtraction processing unit as the loss torque value.

Here, since it is considered that the air resistance during bicycle traveling is proportional to the square of the bicycle traveling speed, the second transfer function unit calculates the square of the wheel rotation angle and multiplies it by an appropriate coefficient. , The first loss torque component due to the air resistance can be derived. by this,
The air resistance received while riding the bicycle is incorporated into the bicycle mechanism transfer function.

More specifically, an addition processing unit for adding a second loss torque component due to disturbance random noise to the first loss torque component due to the air resistance is provided between the second transfer function unit and the subtraction processing unit. The addition result is input to the subtraction processing unit as the loss torque value.

In the specific configuration, the addition processing section adds the second loss torque component due to the disturbance random noise to the first loss torque component received during traveling of the bicycle.
By this, in addition to the air resistance received when riding a bicycle,
Disturbance random noise will be incorporated into the bicycle mechanism transfer function.

Further, specifically, a noise removing means for removing a high frequency component having a frequency higher than the rotation frequency of the pedal is provided at a stage subsequent to the first arithmetic processing means.

When the disturbance random noise is not incorporated in the bicycle mechanism transfer function, the output signal containing the second loss torque component due to the disturbance random noise is derived from the first arithmetic processing means. Here, generally, the frequency of the second loss torque component due to the disturbance random noise is sufficiently higher than the rotation frequency of the pedal. Therefore, by removing the high frequency component from the output signal of the first arithmetic processing means by the noise removing means, the second loss torque component due to the disturbance random noise can be subtracted from the output signal, resulting in human power. The total value of the torque component and the torque component due to the electric motor (input torque value for the bicycle mechanism transfer function) will be derived.

More specifically, the third arithmetic processing means having a transfer function for stabilizing the operation of the control circuit is interposed between the second arithmetic processing means and the motor torque adjusting means. Therefore, even when the control circuit has a configuration that becomes unstable due to disturbance, the operation of the control circuit is stable and an appropriate human power torque value is always calculated.

In another configuration of the electric motor-operated bicycle according to the present invention, the control circuit uses the torque value obtained by subtracting the load torque from the driving torque by the human power and the electric motor as an input signal to the bicycle mechanism, and the wheel rotation angle or the rotation angular velocity. And a first arithmetic processing means for deriving an input torque value for the bicycle mechanism from the sensor detection signal based on an inverse function of the bicycle mechanism transfer function whose output signal is a signal corresponding to Second calculation processing means for deriving the observed torque value by subtracting the output torque value of the electric motor from the value, and a minimum value that appears in 1/2 cycle of the rotation cycle of the pedal in the variation of the derived observed torque value. Then, the load torque value for the bicycle mechanism is estimated, and the load torque value is added to the observed torque value and given to the bicycle mechanism. A third arithmetic processing means for deriving the human power torque value, to create a motor torque command corresponding to the derived human power torque value, which comprises a motor torque adjusting means for supplying to the electric motor.

In the bicycle with the electric motor of the present invention, the torque value (τh + τm−τa) obtained by subtracting the load torque τa from the driving torque (τh + τm) by the human power and the electric motor is input to the bicycle mechanism to rotate the wheels of the bicycle mechanism. A signal corresponding to the angle or the rotational angular velocity is detected by the sensor.
The sensor detection signal is input to the first arithmetic processing means. Here, since the first arithmetic processing means has an inverse function of the bicycle mechanism transfer function as its transfer function, the input / output relationship is opposite to that of the bicycle mechanism transfer function, and the wheel rotation angle or rotation angular velocity is changed. By inputting the corresponding sensor detection signal (output value of the bicycle mechanism transfer function), the input torque value for the bicycle mechanism, that is, the torque value obtained by subtracting the load torque τa from the driving torque (τh + τm) by the human power and the electric motor (the bicycle value The input value of the mechanism transfer function: τh + τm−τa) will be derived. The input torque value for the bicycle mechanism thus derived is input to the second arithmetic processing means, and the torque component due to the electric motor (output torque value of the electric motor) is subtracted, whereby the torque component due to human power (human power A torque value (observed torque value: τh−τa) is derived by subtracting the load torque component τa from the torque value τh).
The observed torque value is input to the third arithmetic processing means.

By the way, in the process of pedaling a bicycle and applying a manual torque to the bicycle mechanism, the magnitude of the manual torque depends on the relationship between the pedaling direction to the pedal and the inclination angle of the crank arm. Becomes zero when reaches the top dead center or the bottom dead center, and becomes maximum when the rotation angle of the pedal is 90 degrees. Therefore, in the fluctuation of the manual torque value, a minimum value that becomes zero in 1/2 cycle of the rotation cycle of the pedal appears. On the other hand, the load torque component applied to the bicycle mechanism is dominated by the wind pressure applied to the bicycle and the human body and the load due to the inclination of the road surface, and therefore the magnitude thereof fluctuates at a cycle sufficiently longer than the rotation cycle of the pedal. Therefore, from the human torque component τh to the load torque component τa
Even in the variation of the observed torque value (τh−τa), which is the value obtained by subtracting, the minimum value appears in 1/2 cycle of the rotation cycle of the pedal. Since the human torque component τh is zero when the minimum value is generated, the minimum value of the observed torque value matches the value (−τa) obtained by inverting the sign of the load torque component τa.

Then, based on the minimum value of the observed torque value obtained periodically, the load torque value τa which is continuous in time is estimated and added to the observed torque value (τh-τa), the human power is reduced. The torque value τh will be derived. Thus, the manual torque value can be obtained without using the torque detection mechanism. The human torque value thus derived is input to the motor torque adjusting means,
A motor torque command corresponding to the human power torque value is created,
It is supplied to the electric motor. As a result, the electric motor generates a motor output torque according to the manual torque.

Specifically, the bicycle mechanism transfer function is a transfer function having the inertial force and the viscous resistance of the bicycle mechanism as parameters, and the torque value obtained by subtracting the load torque from the driving torque by the human power and the electric motor is the bicycle mechanism. And the rotational angular velocity is derived from the input signal.

Here, the input signal to the bicycle mechanism is
It is the drive torque value of the bicycle mechanism that actually contributes to bicycle traveling, and the relationship between the drive torque of the bicycle mechanism and the wheel rotation angular velocity can be defined by using the inertial force and viscous resistance of the bicycle mechanism as parameters. Therefore, the wheel rotation angular velocity in actual bicycle running is derived from the input signal by the bicycle mechanism transfer function having this relationship as the transfer function.

More specifically, the bicycle mechanism transfer function is
The torque value obtained by subtracting the load torque from the driving torque by the human power and the electric motor is used as an input signal to the bicycle mechanism,
A subtraction processing unit that subtracts a load torque component due to air resistance received during bicycle running from the input signal, and a rotational angular velocity is derived by performing a calculation process using a transfer function with the inertia force and viscous resistance of the bicycle mechanism as parameters on the subtraction result. First to do
Based on the transfer function part and the derived rotational angular velocity,
And a second transfer function unit for deriving a load torque component due to air resistance received during bicycle traveling, and the derived load torque component is input to the subtraction processing unit.

In the concrete structure, the torque value obtained by subtracting the load torque from the driving torque by the human power and the electric motor is input to the subtraction processing unit, the load torque component is subtracted, and the bicycle mechanism actually contributes to the bicycle running. The drive torque value of is calculated. The calculated driving torque value of the bicycle mechanism is input to the first transfer function unit. Here, the relationship between the drive torque of the bicycle mechanism and the wheel rotation angular velocity can be defined by using the inertial force and the viscous resistance of the bicycle mechanism as parameters. Therefore, the first transfer function unit having this relationship as a transfer function derives the wheel rotation angular velocity in the actual bicycle running from the calculated drive torque value. or,
The air resistance during bicycle traveling is considered to be proportional to the square of the bicycle traveling speed. Therefore, the second transfer function unit calculates the square of the wheel rotation angle and multiplies it by an appropriate coefficient to determine the load torque due to the air resistance. The components can be derived. As a result, the air resistance received while riding the bicycle is incorporated in the bicycle mechanism transfer function.

More specifically, a fourth arithmetic processing means having a transfer function for stabilizing the operation of the control circuit is interposed between the third arithmetic processing means and the motor torque adjusting means. . Therefore, even when the control circuit has a configuration that becomes unstable due to disturbance, the operation of the control circuit is stable and an appropriate human power torque value is always calculated.

[0025]

According to the bicycle with the electric motor of the present invention, since the torque detecting mechanism can be omitted, the structure of the entire bicycle with the electric motor can be simplified and the weight can be reduced. is there.

[0026]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, an embodiment of the present invention will be specifically described with reference to the drawings based on four examples. First Embodiment A bicycle with an electric motor according to the present embodiment shown in FIG. 2 has a bicycle mechanism transfer function in which air resistance and disturbance random noise received during traveling of the bicycle are incorporated.

FIG. 1 shows the overall construction of a bicycle with an electric motor according to the present embodiment. The wheels of a bicycle mechanism (1) which can be driven by human power are equipped with electric motors (2) for assisting the driving force by human power. And a rotation angle sensor (3) for detecting the rotation angle of the wheel is attached. Further, a power assist controller (4) for controlling the output torque of the electric motor based on the detection signal of the wheel rotation angle obtained from the rotation angle sensor (3) is provided. As the rotation angle sensor (3), a sensor having a simple structure such as a well-known rotary encoder can be adopted.

The human power torque generated by stepping on the pedal and the motor output torque generated by the electric motor (2) are applied to the bicycle mechanism (1), whereby the wheels of the bicycle mechanism (1) are rotationally driven. . The wheel rotation angle is detected by the rotation angle sensor (3), and the rotation angle detection signal is input to the power assist controller (4). According to this, power assist controller (4)
Creates a motor torque command and supplies it to the electric motor (2). This drives the wheels.

FIG. 2 is a block diagram showing the control system of the electric motor-operated bicycle of this embodiment. First, the power assist controller (4) will be specifically described, and then the modeling of the bicycle mechanism (1) necessary for designing the power assist controller (4) will be described.

In FIG. 2, the human power torque value τh generated by stepping on the pedal and the motor output torque value τm generated by the electric motor are simultaneously given to the bicycle mechanism (1), so that the wheels of the bicycle mechanism (1) move. Driven. Then, the rotation angle θ of the wheel is detected by the rotation angle sensor (3), and the detection signal of the wheel rotation angle θ is input to the power assist controller (4).

In the power assist controller (4), the wheel rotation angle θ obtained from the rotation angle sensor (3)
Is input to the differential calculator (41), and the wheel rotation angular velocity ω is calculated. The calculated wheel rotation angular velocity ω is input to the inverse function calculator (42). Where the inverse function calculator (42)
Is a bicycle mechanism transfer function Pn (s) from the input of the human power torque value τh and the motor output torque value τm to the derivation of the wheel rotation angular velocity ω in the model of the bicycle mechanism (1) described later.
Of the inverse function Pn (s) ⁻¹ of
^-1 has an input / output relationship opposite to that of the bicycle mechanism transfer function Pn (s). Therefore, by inputting the wheel rotation angular velocity ω to the inverse function Pn (s) ^-1 , the human torque value τh
Then, the total value (τh + τm) of the motor output torque value τm is derived. The total value (τh + τm) of the derived human power torque value τh and the motor output torque value τm is input to the subtractor (43), and the motor output torque value τm is subtracted to derive the human power torque value τh. . The derived human torque value τh is input to the stabilization processing circuit (44).
Here, the stabilization processing circuit (44) has a transfer function Q (s) for stabilizing the operation of the control circuit, and an appropriate human power torque value τh ′ is always obtained by calculation processing by the transfer function Q (s). Is calculated. A well-known method for determining the transfer function Q (s)
(For example, Journal of the Robotics Society of Japan, Vol. 11, No. 4, May 1993.
Months, see pages 6 to 13). The human power torque value τh ′ thus derived is input to the motor torque adjusting circuit (45), a motor torque command corresponding to the human power torque value τh ′ is created, and supplied to the electric motor (2). . The adjustment gain of the motor torque adjusting circuit (45) can be defined by a constant value or an arbitrary function with the traveling speed of the bicycle as a parameter. As a result, the electric motor (2) generates a motor output torque corresponding to the human torque generated by stepping on the pedal.

Next, modeling of the bicycle mechanism (1) will be described. FIG. 3 shows a basic configuration of a model of the bicycle mechanism (1), which is a total value (τh + τm) of a human power torque value τh generated by stepping on a pedal and a motor output torque value τm generated by an electric motor. Is input to the subtraction unit (11), and the loss torque value τl is subtracted from the subtraction unit (11) to calculate the drive torque value τ of the bicycle mechanism (1) that actually contributes to bicycle traveling. The calculated drive torque value τ is input to the torque-angular velocity conversion unit (12). here,
The relationship between the drive torque value τ of the bicycle mechanism (1) and the wheel rotation angular velocity ω can be defined by using the inertial force J and the viscous resistance D of the bicycle mechanism (1) during traveling of the bicycle as parameters. (12) is the transfer function G of the following equation 1
It is represented by (s).

[0033]

[Equation 1] G (s) = 1 / (Js + D)

Accordingly, the torque-angular velocity converting section (12) derives the wheel rotational angular velocity ω during actual bicycle traveling. The derived wheel rotation angular velocity ω is calculated by the integral calculation unit (1
The wheel rotation angle θ is derived by inputting to 3) and performing integration.

FIG. 4 shows a concrete structure of a model of the bicycle mechanism (1) which further incorporates the loss torque value τl as an element. The wheel rotation angular velocity ω derived by the torque-angular velocity conversion unit (12) is input to the integral calculation unit (13) and the square calculation unit (14). here,
Air resistance when riding a bicycle is considered to be proportional to the square of the bicycle running speed. Therefore, after the wheel rotation angular velocity ω is squared in the square calculation unit (14), the square value is multiplied by the coefficient C1 according to the viscosity of air in the wind pressure torque calculation unit (15) to obtain the One loss torque component τe will be derived. Derived first loss torque component τe
Is input to the addition unit (16), and the second loss torque component τd due to the disturbance random noise is added to derive an accurate loss torque value τl that is received during bicycle traveling. The derived loss torque value τl is input to the subtraction unit (11). As described above, after modeling the bicycle mechanism (1) and deriving the bicycle mechanism transfer function Pn (s), the wheel rotation angular velocity ω is input by obtaining the inverse function Pn (s) ⁻¹ by a known method. Input torque given as a signal to the bicycle mechanism (1) by human power and an electric motor
The transfer function of the inverse system with (τh + τm) as the output signal is obtained.

FIG. 5 shows a control procedure of the power assist controller (4) of the present invention. When the bicycle starts to be driven by human power and motor output, first in step S1, the motor output torque value τm is initialized to 0,
In step S2, the detection signal of the wheel rotation angle θ is fetched from the rotation angle sensor (3).

In step S3, the wheel rotational angular velocity ω is calculated by inputting the detected wheel rotational angle θ to the differential calculator (41). Then, in step S4, is the wheel rotational angular velocity ω positive? If it is determined NO, the process returns to step S1. YE in step S4
If it is determined to be S, the process proceeds to step S5, and the wheel rotation angular velocity ω is input to the inverse function calculator (42),
The sum of the human power torque value τh and the motor output torque value τm (τh
+ Τm) is derived. Then, in step S6, the total value (τh + τm) of the human power torque value τh and the motor output torque value τm
Is input to the subtractor (43) to subtract the motor output torque value τm, thereby calculating the human torque value τh. Then, in step S7, by inputting the human power torque value τh into the stabilization processing circuit (44), an appropriate human power torque value τh ′ for stabilizing the control circuit operation is calculated. Next, in step S8, the human torque value τh ′ is input to the motor torque adjusting circuit (45) to output the human torque value τh.
Create a motor torque command according to h ′ to generate an electric motor
Supply to (2) and return to step S2.

According to the above-mentioned procedure, the output torque of the electric motor (2) can be changed according to the manpower torque generated by the depression of the pedal, whereby an appropriate auxiliary driving force can be obtained. For example, on an uphill road, pedaling with a large force increases the human-powered torque input to the bicycle mechanism, and as a result, the motor output torque supplied to the bicycle mechanism by the electric motor is increased, and Driving force is reduced. On the other hand, on a downhill, a large force is not required to step on the pedal, so the human torque input to the bicycle mechanism is small, and as a result, the motor output torque supplied to the bicycle mechanism by the electric motor is reduced. Becomes smaller.

Second Embodiment In the first embodiment described above, the air resistance and the disturbance random noise received during bicycle running are incorporated in the bicycle mechanism transfer function, while the electric motor of the present embodiment shown in FIG. 6 is included. In a bicycle, only the air resistance received during traveling of the bicycle is incorporated into the bicycle mechanism transfer function, and the power assist controller (40) is provided with a low-pass filter circuit (47) to remove disturbance random noise. The overall configuration of the electric motor-equipped bicycle of this embodiment is the same as that of the first embodiment shown in FIG.

FIG. 6 is a block diagram showing the control system of the electric motor-operated bicycle according to the present embodiment. First, the power assist controller (40) will be specifically described, as in the first embodiment. The modeling of the bicycle mechanism (1) will be described.

In FIG. 6, the human power torque value τh generated by stepping on the pedal and the motor output torque value τm generated by the electric motor are simultaneously given to the bicycle mechanism (1) so that the wheels of the bicycle mechanism (1) are Driven. Then, the rotation angle θ of the wheel is detected by the rotation angle sensor (3), and the detection signal of the wheel rotation angle θ is input to the power assist controller (40).

In the power assist controller (40), the wheel rotation angle θ obtained from the rotation angle sensor (3)
Is input to the differential calculator (41), and the wheel rotation angular velocity ω is calculated. The calculated wheel rotation angular velocity ω is input to the inverse function calculator (46). Where the inverse function calculator (46)
Is a model of the bicycle mechanism (1) to be described later, with the torque value (τh + τm−τd) obtained by subtracting the loss torque component τd due to disturbance random noise from the total value of the human power torque value τh and the motor output torque value τm as an input signal, It has an inverse function Pn ′ (s) ⁻¹ of the bicycle mechanism transfer function Pn ′ (s) up to the derivation of the wheel rotation angular velocity ω. The inverse function Pn '(s) ^-1
Has an input / output relationship opposite to that of the bicycle mechanism transfer function Pn ′ (s), and the wheel rotation angular velocity ω is input to the inverse function Pn ′ (s) ⁻¹ , whereby the input torque value ( τh +
τm−τd) will be derived.

The derived torque value (.tau.h + .tau.m-.tau.d) is input to the low pass filter circuit (47), and the frequency component higher than the rotation frequency of the pedal is removed. Here, generally, the frequency of the loss torque component τd due to the disturbance random noise is sufficiently higher than the rotation frequency of the pedal. Therefore, from the low-pass filter circuit (47), a signal in which the loss torque component τd due to the disturbance random noise is removed from the derived torque value (τh + τm−τd), that is, the human power torque value τh and the motor output torque value τm Total value (τh ＋ τm)
Is output.

The total value (τh + τm) of the derived human power torque value τh and the motor output torque value τm is input to the subtractor (43), and the motor output torque value τm is subtracted to obtain the human power torque value τh. Derived. After that, the same processing as in the first embodiment is performed on the human power torque value τh, and the electric motor (2) outputs a motor output torque corresponding to the human power torque generated by stepping on the pedal. become.

Next, modeling of the bicycle mechanism (1) will be described. As shown in FIG. 6, the total value (τh + τm) of the human power torque value τh generated by stepping on the pedal and the motor output torque value τm generated by the electric motor is the first value.
The loss torque component τd due to the disturbance random noise is input to the subtraction unit (17) and is subtracted. The subtraction result is input to the second subtraction unit (18), and the loss torque component τe due to the air resistance received during traveling of the bicycle is subtracted to actually drive the bicycle mechanism (1). τ is calculated. The calculated drive torque value τ is torque −
It is input to the angular velocity conversion unit (12). Here, the torque-angular velocity conversion unit (12) is represented by the transfer function G (s) of the above-mentioned equation 1, and the wheel rotational angular velocity ω in actual bicycle traveling is derived as in the first embodiment. . The derived wheel rotation angular velocity ω is input to the integration calculation unit (13) and integrated to derive the wheel rotation angle θ. The derived wheel rotation angular velocity ω is also input to the square calculation unit (14) at the same time, and after the wheel rotation angular velocity ω is squared, the square pressure value is calculated by the wind pressure torque calculation unit (15) according to the viscosity of air. Coefficient C
By multiplying by 1, the loss torque component τe due to the air resistance is derived. The derived loss torque component τe is input to the subtraction unit (18).

By modeling the bicycle mechanism (1) as described above, the torque value (τh + τm−τd) obtained by subtracting the loss torque component τd due to disturbance random noise from the total value of the human power torque value τh and the motor output torque value τm. ) As an input signal, it is possible to obtain the bicycle mechanism transfer function Pn '(s) up to the derivation of the wheel rotation angular velocity ω. Then, the wheel transfer angular velocity ω is used as an input signal by obtaining the inverse transfer function Pn ′ (s) ⁻¹ from the transfer function Pn ′ (s),
The transfer function of the inverse system with the input torque value (τh + τm−τd) as the output signal is obtained.

According to the bicycle with the electric motor of the present embodiment, even when the element for eliminating the disturbance random noise is not incorporated in the inverse function calculator (46), the disturbance random noise is generated by the low-pass filter circuit (47). The loss torque component τd due to can be removed. by this,
The manual torque generated by actually stepping on the pedal can be obtained, and an appropriate auxiliary drive force can be obtained by applying a motor output torque corresponding to the manual torque to the electric motor (2).

Third Embodiment In the bicycle with an electric motor of the present embodiment shown in FIG. 7, the power assist controller (8) has a human torque value calculating circuit (82).
Is provided, the load torque value for the bicycle mechanism (1) is estimated, and the torque value given to the bicycle mechanism (1) by human power is derived based on the estimated value. The overall configuration of this embodiment is the same as that of the first embodiment shown in FIG.

FIG. 7 is a block diagram showing the control system of the electric motor-operated bicycle of this embodiment. First, the power assist controller (8) will be specifically described, and then modeling of the bicycle mechanism (1) will be described.

In FIG. 7, the torque value (τh + τm) obtained by subtracting the load torque τa from the total value of the human power torque value τh generated by stepping on the pedal and the motor output torque value τm generated by the electric motor, that is, the drive torque value. −τa)
Is input to the bicycle mechanism (1), and the wheels of the bicycle mechanism (1) are driven. Here, the load torque τa is given as a total value (τe + τs + τd) of the load torque component τe due to the air resistance, the load torque component τs due to the inclination of the road surface, and the load torque component τd due to the disturbance random noise. Then, the rotation angle θ of the wheel is detected by the rotation angle sensor (3), and the detection signal of the wheel rotation angle θ is input to the power assist controller (8).

In the power assist controller (8), the wheel rotation angle θ obtained from the rotation angle sensor (3)
Is input to the differential calculator (41), and the wheel rotation angular velocity ω is calculated. The calculated wheel rotation angular velocity ω is input to the inverse function calculator (81). Where the inverse function calculator (81)
In the model of the bicycle mechanism (1) described later, using the input torque value (τh + τm−τa) as an input signal, the bicycle mechanism transfer function Rn up to the derivation of the wheel rotation angular velocity ω.
It has the inverse function Rn (s) ⁻¹ of (s). Inverse function Rn (s) ^-1
Has an input / output relationship opposite to that of the bicycle mechanism transfer function Rn (s). By inputting the wheel rotation angular velocity ω to the inverse function Rn (s) ^-1 , the input torque value (τh + τm
−τa) will be derived.

The derived torque value (τh + τm-τa) is input to the subtractor (43), and the motor output torque value τm is subtracted to derive the observed torque value (τh-τa). The derived observed torque value (τh−τa) is input to the human power torque value calculation circuit (82).

Here, the operation principle of the manual torque value calculation circuit (82) will be specifically described. When a pedaling force is applied vertically downward to one pedal (9a) as indicated by arrow A in FIG. 9, the crank arm (90) rotates clockwise as indicated by arrow B. Here, the inclination angle (crank angle) of the crank arm (90) is assumed to be 0 degree and clockwise when the pedal (9a) on one side reaches the top dead center. At this time,
In the process of pedaling a bicycle, the magnitude of the human torque τh applied to the bicycle mechanism (1) is as shown in FIG. 10, when one pedal (9a) reaches the top dead center or the bottom dead center, that is, It becomes zero when the crank angle is 0 degrees or 180 degrees, and becomes maximum when the crank angle is 90 degrees or 270 degrees. Therefore, in the fluctuation of the human torque τh, a minimum value that becomes zero in 1/2 cycle of the pedal rotation cycle appears.

On the other hand, the load torque τa is the total value (τe + τs + τd) of the load torque component τe due to the air resistance, the load torque component τs due to the inclination of the road surface and the load torque component τd due to the disturbance random noise, as described above.
The load torque component τe due to the air resistance and the load torque component τs due to the inclination of the road surface are dominant. Therefore, the magnitude of the load torque τa fluctuates in a cycle sufficiently longer than the rotation cycle of the pedal.

For example, when the slope of the road surface changes as shown in FIG. 8A, the load torque is substantially zero on a flat road surface.
On the ascending slope, a positive load torque corresponding to the inclination is generated, and on the descending slope, a negative load torque according to the inclination is generated. That is, the load torque τa fluctuates in a cycle corresponding to the change in the inclination of the road surface, and this cycle is generally sufficiently longer than the pedal rotation cycle (see FIG. 8 (c)).

As a result, even when the observed torque value (τh-τa), which is the value obtained by subtracting the load torque value τa from the human power torque value τh, changes as shown in FIG. The minimum value m appears in two cycles. Since the human power torque value τh is zero at the time when the minimum value m is generated, the minimum value m of the observed torque value matches the value (−τa) obtained by inverting the sign of the load torque value τa.

Therefore, based on the minimum value m of the observed torque value periodically obtained by the human power torque value calculating circuit (82) as shown in FIG. 8B, the load torque value τ shown in FIG.
a is estimated, and this estimated load torque value τa is observed torque value.
By adding it to (τh−τa), the human power torque value τh shown in FIG.

The load torque value τa can be estimated by a method of predicting the load torque until the next minimum value is obtained based on the fluctuation of the minimum value of the observed torque in the past, or a minimum value of the current observed torque. Various well-known methods such as a method of updating the load torque with a delay of ½ cycle of the pedal rotation cycle by holding the value for a half cycle of the pedal rotation cycle can be adopted. As a result, the estimated load torque value approximate to the load torque τa shown in FIG. 8C can be obtained.

After that, the first torque is calculated for the human torque τh.
Processing similar to that of the embodiment is performed, and from the electric motor (2),
The motor output torque corresponding to the manual torque generated by stepping on the pedal is output.

Next, modeling of the bicycle mechanism (1) will be described. As shown in FIG. 7, a torque value (τh + τm−τa) obtained by subtracting the load torque τa from the total value (τh + τm) of the human power torque value τh generated by stepping on the pedal and the motor output torque value τm generated by the electric motor, That is, the driving torque value τ of the bicycle mechanism (1) that actually contributes to bicycle running is input to the torque-angular velocity conversion unit (12). Here, the torque-angular velocity conversion unit (12) is represented by the transfer function G (s) of the above-mentioned equation 1, and the wheel rotational angular velocity ω in actual bicycle traveling is derived as in the first embodiment. . The derived wheel rotation angular velocity ω is input to the integration calculation unit (13) and integrated to derive the wheel rotation angle θ.

By modeling the bicycle mechanism (1) as described above, the torque value obtained by subtracting the load torque value τa from the total value of the human power torque value τh and the motor output torque value τm.
(τh + τm−τa) as input signal, wheel rotation angular velocity ω
It is possible to obtain the bicycle mechanism transfer function Rn (s) up to the derivation of Then, by obtaining the inverse transfer function Rn (s) ⁻¹ from the transfer function Rn (s), the wheel rotation angular velocity ω is used as an input signal, and the input torque value (τh + τm−
The transfer function of the inverse system with τa) as the output signal is obtained.

FIG. 11 shows the control procedure of the power assist controller (8) of the present invention. When the bicycle starts to be driven by human power and motor output, first in step S10, the motor output torque value τm and the load torque value τa are initialized to 0, and in step S11, the wheel rotation angle θ is read from the rotation angle sensor (3). Capture the detection signal of.

In step S12, the detected wheel rotation angle θ is input to the differential calculator (41) to calculate the wheel rotation angular velocity ω, and then in step S13, is the wheel rotation angular velocity ω positive? Determine whether or not. Step S
If NO is determined in 13, the process returns to step S10. If YES is determined in step 13, the process proceeds to step S14, and the wheel rotation angular velocity ω is calculated by the inverse function calculator.
By inputting to (81), the torque value (τh + τm−τa) is derived by subtracting the load torque value τa from the total value of the human power torque value τh and the motor output torque value τm. Then, in step S15, the observed torque value (τh−τa) is calculated by inputting the torque value (τh + τm−τa) to the subtractor (43) and subtracting the motor output torque value τm. next,
In step S16, the pedal rotation angle φ is 0 ° or 1
If it is determined to be YES, the process proceeds to step S17 to estimate the load torque value τa as described above. Then, in step S18, the estimated load torque value τa is added to the observed torque value (τh−τa) to derive the human power torque value τa. On the other hand, if NO in step S16, the process proceeds to step S18, and the current observed torque value (τh−τa)
The human power torque value τh is derived by adding the estimated load torque value τa at that time to. Then, in step S19, the human torque value τh is input to the stabilization processing circuit (44) to calculate an appropriate human torque value τh ′ for stabilizing the operation of the control circuit. Next, in step S20, the human torque value τh ′ is input to the motor torque adjusting circuit (45) to generate a motor torque command corresponding to the human torque value τh ′ and supply it to the electric motor (2). Return to S11.

Fourth Embodiment In the above-mentioned third embodiment, the load torque is not incorporated in the bicycle mechanism transfer function, whereas in the bicycle with the electric motor of this embodiment shown in FIG. It incorporates a load torque component due to the air resistance received when riding a bicycle. The overall configuration of the electric motor-equipped bicycle of this embodiment is the same as that of the first embodiment shown in FIG.

FIG. 12 is a block diagram showing the control system of the electric motor-operated bicycle of this embodiment. First, the power assist controller (80) will be specifically described, and then modeling of the bicycle mechanism (1) will be described.

In FIG. 12, the sum of the human power torque value τh generated by stepping on the pedal and the motor output torque value τm generated by the electric motor, that is, the drive torque value (τ
The torque value (τh) obtained by subtracting the load torque τa 'from (h + τm)
+ Τm−τa ′) is input to the bicycle mechanism (1) and the wheels of the bicycle mechanism (1) are driven. Where load torque τ
a'is the sum of the load torque component τs due to the road surface inclination and the load torque component τd due to the disturbance random noise (τs + τ
It is d), and does not include the load torque component τe due to the air resistance received while riding the bicycle. Then, the rotation angle θ of the wheel is detected by the rotation angle sensor (3), and the detection signal of the wheel rotation angle θ is detected by the power assist controller (80).
Is input to

In the power assist controller (80), the wheel rotation angle θ obtained from the rotation angle sensor (3)
Is input to the differential calculator (41), and the wheel rotation angular velocity ω is calculated. The calculated wheel rotation angular velocity ω is input to the inverse function calculator (83). Where the inverse function calculator (83)
In the model of the bicycle mechanism (1) described later, using the input torque value (τh + τm−τa ′) as an input signal, the bicycle mechanism transfer function R until the wheel rotation angular velocity ω is derived.
It has an inverse function Rn '(s) ^-1 of n' (s). Inverse function R
The input / output relation of n '(s) ^-1 is opposite to that of the bicycle mechanism transfer function Rn' (s), and the wheel rotation angular velocity ω is input to the inverse function Rn '(s) ^-1. Thus, the input torque value (τh + τm−τa ′) is derived.

The derived torque value (.tau.h + .tau.m-.tau.a ') is input to the subtractor (43), and the motor output torque value .tau.m is subtracted to derive the observed torque value (.tau.h-.tau.a'). The derived observed torque value (τh−τa ′) is input to the human power torque value calculation circuit (84), the load torque value τa ′ is estimated by the same principle as the third embodiment, and the estimated load torque value is calculated. The human torque value τh is derived by adding τa ′ to the observed torque value (τh−τa ′). Here, the load torque value τa ′ is the total value (τs + τd) of the load torque component τs due to the inclination of the road surface and the load torque component τd due to the disturbance random noise, and the load torque component τs due to the inclination of the road surface is dominant. Therefore, the fluctuation cycle of the load torque value τa ′ is sufficiently longer than the rotation cycle of the pedal. Therefore, the observed torque value is the same as in the third embodiment.
In the fluctuation of (τh−τa ′), 1 / (1) of the rotation cycle of the pedal
A minimum value appears in two cycles, and the load torque value τa 'can be estimated based on the minimum value.

After that, the first value is calculated with respect to the human power torque τh.
Processing similar to that of the embodiment is performed, and from the electric motor (2),
The motor output torque corresponding to the manual torque generated by stepping on the pedal is output.

Next, modeling of the bicycle mechanism (1) will be described. As shown in FIG. 12, the torque value (τh + τm−τa ′) obtained by subtracting the load torque τa ′ (τs + τd) from the total value of the human power torque value τh generated by stepping on the pedal and the motor output torque value τm generated by the electric motor. )
Is input to the subtraction unit (71) and the load torque component τe due to the air resistance received during traveling of the bicycle is subtracted, whereby the driving torque value τ of the bicycle mechanism (1) that actually contributes to traveling of the bicycle is calculated. The calculated drive torque value τ is input to the torque-angular velocity conversion unit (12). Here, the torque-angular velocity conversion unit (12) is represented by the transfer function G (s) of the above-mentioned equation 1, and the wheel rotational angular velocity ω in actual bicycle traveling is derived as in the first embodiment. . The derived wheel rotation angular velocity ω is input to the integration calculation unit (13) and integrated to derive the wheel rotation angle θ.
Further, the derived wheel rotation angular velocity ω is the square calculation unit (14)
Is input at the same time, and after squared the wheel rotation angular velocity ω,
The load torque component τe due to the air resistance is derived by multiplying the squared value by the coefficient C1 according to the viscosity of the air in the wind pressure torque calculation unit (15). The derived load torque component τe is input to the subtraction unit (71).

As described above, in the bicycle mechanism (1), the load torque component τe due to the air resistance received during bicycle traveling is separated from the load torque for the bicycle mechanism (1), and the load torque component τe is used as the bicycle mechanism (1). By modeling, the bicycle mechanism transfer function Rn ′ incorporating air resistance
(s) can be obtained. Then, the transfer function Rn ′
By calculating the inverse transfer function Rn ′ (s) ⁻¹ from (s), the wheel rotation angular velocity ω is used as an input signal, and the load torque component τe is calculated from the total value of the human torque value τh and the motor output torque value τm. A transfer function of the inverse system having the torque value (τh + τm−τa ′) obtained by subtracting the load torque value τa ′ not included as an output signal is obtained.

According to the electric motor-operated bicycles of the first to fourth embodiments described above, the motor output torque corresponding to the human torque can be generated by installing the rotation angle sensor having a simple structure, which is complicated. The torque detection mechanism having such a structure is not necessary, and thus, the bicycle mechanism can be made simpler and lighter than conventional ones.

The above description of the embodiments is for explaining the present invention, and should not be construed as limiting the invention described in the claims or reducing the scope.
In addition, the configuration of each part of the present invention is not limited to the above-described embodiment, and it goes without saying that various modifications can be made within the technical scope described in the claims.

For example, in any of the above embodiments, the input signal of the power assist controller is not limited to the detection signal of the wheel rotation angle, but the rotation angular velocity detected by the rotation angular velocity sensor is directly input to the inverse function calculator. It is also possible to use the method. In this case, the differential calculator can be omitted.

[Brief description of drawings]

FIG. 1 is a diagram showing an overall configuration of a bicycle with an electric motor according to the present invention.

FIG. 2 is a block diagram showing a control system of the bicycle with the electric motor according to the first embodiment.

FIG. 3 is a block diagram showing a basic configuration of a model of a bicycle mechanism.

FIG. 4 is a block diagram showing a specific configuration of a model of a bicycle mechanism that incorporates loss torque as an element.

FIG. 5 is a flowchart showing a control procedure by the power assist controller of the first embodiment.

FIG. 6 is a block diagram showing a control system of a bicycle with an electric motor according to a second embodiment.

FIG. 7 is a block diagram showing a control system of a bicycle with an electric motor according to a third embodiment.

FIG. 8 is a graph showing variations in observed torque value, load torque value, and human power torque value during bicycle traveling.

FIG. 9 is a diagram illustrating a relationship between a pedaling force direction with respect to a pedal and a crank angle.

FIG. 10 is a graph showing a relationship between a human power torque value applied to a bicycle mechanism and a crank angle.

FIG. 11 is a flowchart showing a control procedure by the power assist controller of the third embodiment.

FIG. 12 is a block diagram showing a control system of a bicycle with an electric motor according to a fourth embodiment.

FIG. 13 is a diagram showing an overall configuration of a conventional bicycle with an electric motor.

FIG. 14 is a block diagram showing a control system of the bicycle with an electric motor.

[Explanation of symbols]

 (1) Bicycle mechanism (2) Electric motor (3) Rotation angle sensor (4) Power assist controller

Claims (10)

[Claims] 1. A bicycle with an electric motor for connecting an electric motor to a bicycle mechanism that can be driven by human power to assist the driving force by human power, and the bicycle mechanism has a wheel rotation angle or a rotational angular velocity of the bicycle mechanism. A sensor that detects a signal and a control circuit that controls the output torque of the electric motor based on the sensor detection signal are provided, and the control circuit uses the torque applied to the bicycle mechanism by the human power and the electric motor as an input signal, First arithmetic processing means for deriving an input torque value for the bicycle mechanism from the sensor detection signal based on an inverse function of a bicycle mechanism transfer function whose output signal is a signal corresponding to a wheel rotation angle or a rotation angular velocity; Second arithmetic processing means for deriving a human torque value by subtracting the output torque value of the electric motor from the input torque value for the bicycle mechanism , To create a motor torque command corresponding to the derived human power torque value, bicycles electric motor, characterized in that it comprises a motor torque adjusting means for supplying to the electric motor. 2. A bicycle mechanism transfer function is obtained by performing a subtraction processing unit for subtracting a loss torque value from the input signal, and performing a calculation process by a transfer function having the inertia force and viscous resistance of the bicycle mechanism as parameters on the subtraction result. The bicycle with an electric motor according to claim 1, comprising a transfer function unit that derives a rotational angular velocity. 3. A second transfer function section for deriving a loss torque component due to air resistance received during bicycle traveling based on the derived wheel rotation angular velocity is connected to the latter stage of the transfer function section, and the loss torque component is connected. The bicycle with an electric motor according to claim 2, wherein is input to the subtraction processing unit as the loss torque value. 4. An addition processing unit for adding a second loss torque component due to disturbance random noise to the first loss torque component due to the air resistance is interposed between the second transfer function unit and the subtraction processing unit, The bicycle with an electric motor according to claim 3, wherein the addition result is input to the subtraction processing unit as the loss torque value. 5. The bicycle with an electric motor according to claim 3, further comprising noise removing means for removing a high-frequency component having a frequency higher than a rotation frequency of the pedal, provided at a stage subsequent to the first arithmetic processing means. 6. The third arithmetic processing means having a transfer function for stabilizing the operation of the control circuit is interposed between the second arithmetic processing means and the motor torque adjusting means. A bicycle with an electric motor according to claim 5. 7. A bicycle with an electric motor, which is connected to an electric motor to a bicycle mechanism that can be driven by human power and assists the driving force by human power, wherein the bicycle mechanism corresponds to a wheel rotation angle or a rotational angular velocity of the bicycle mechanism. A sensor that detects the signal and a control circuit that controls the output torque of the electric motor based on the sensor detection signal are provided, and the control circuit calculates the torque value obtained by subtracting the load torque from the driving torque by the human power and the electric motor. As an input signal to the bicycle mechanism,
First arithmetic processing means for deriving an input torque value for the bicycle mechanism from the sensor detection signal based on an inverse function of a bicycle mechanism transfer function whose output signal is a signal corresponding to a wheel rotation angle or a rotation angular velocity; Second arithmetic processing means for deriving the observed torque value by subtracting the output torque value of the electric motor from the input torque value for the bicycle mechanism; and 1/2 cycle of the rotation cycle of the pedal in the variation of the derived observed torque value. A third arithmetic processing means for estimating a load torque value for the bicycle mechanism based on the minimum value appearing in the above, and adding the load torque value to the observed torque value to derive a human power torque value given to the bicycle mechanism; A motor torque command according to the derived human torque value is created, and a motor torque adjusting means for supplying to the electric motor is provided. Bicycle with an electric motor. 8. The electric motor-equipped motor according to claim 7, wherein the bicycle mechanism transfer function is a transfer function having the inertial force and the viscous resistance of the bicycle mechanism as parameters, and the rotational angular velocity is derived from the input signal. bicycle. 9. The bicycle mechanism transfer function includes a subtraction processing unit that subtracts a load torque component due to air resistance received during bicycle traveling from the input signal, and a transfer that uses the inertia force and the viscous resistance of the bicycle mechanism as parameters to the subtraction result. A first transfer function unit that performs a calculation process by a function to derive a rotational angular velocity, and a second transfer function unit that derives a load torque component due to air resistance received during bicycle traveling based on the derived rotational angular velocity. The bicycle with an electric motor according to claim 7, wherein the derived load torque component is input to the subtraction processing unit. 10. The fourth arithmetic processing means having a transfer function for stabilizing the operation of the control circuit is interposed between the third arithmetic processing means and the motor torque adjusting means. A bicycle with an electric motor according to claim 9.