METHOD FOR DETERMINING MOTION QUANTITIES OF A TWO-WHEELER

Description

State of the art

[0001] The present invention relates to a method for determining movement quantities of a two-wheeler, as well as to a two-wheeler.

[0002] Sensor systems for two-wheelers are well known, by means of which movement variables such as speed, distance traveled, acceleration and rotation rates of two-wheelers can be recorded. Speed ​​is an important movement variable which is also used for other systems, for example. For example, in an electric bicycle, the drive unit of the electric bicycle is often controlled depending on the speed. In bicycles in particular, the speed is often recorded using so-called reed sensors. A magnet is usually attached to one of the bicycle's wheels. A magnetic sensor attached to the bicycle frame records a pulse for each rotation of the wheel in order to determine the speed of the bicycle based on the pulse frequency and the wheel circumference.For reasons of cost, simplicity, and weight, a single-pulse sensor with a single magnet on the wheel is often used. However, this often results in high inaccuracies, especially at low speeds. It is also known to increase accuracy with multi-pulse sensors, but this increases complexity, cost, and weight.

[0003] US 2012 / 259526 A1, US 2017 / 285065 A1 and US 2012 / 323485 disclose known systems and methods for determining movement quantities of vehicles. Disclosure of the invention

[0004] The method according to the invention with the features of claim 1 is distinguished by a particularly simple and cost-effective method with which the movement variables of a two-wheeler can be determined very precisely. In particular, a high level of accuracy can be achieved even at very low speeds. This is achieved according to the invention by a method for determining the movement variables of a two-wheeler, wherein the two-wheeler comprises a sensor system that has a yaw rate sensor, an acceleration sensor, and a wheel speed sensor. The wheel speed sensor is, in particular, a rotation sensor and is designed to detect at least one measurement pulse per rotation of a wheel of the two-wheeler.Preferably, the wheel speed sensor is a single-pulse reed sensor, which has precisely one magnet attached to the wheel and rotating with the wheel, and in particular a receiver that detects precisely one measurement pulse when the magnet passes. The method comprises the following steps: . Detecting, in particular three-dimensional, rotation rates of the two-wheeler of the rotation rate sensor, detecting acceleration values ​​of the two-wheeler by means of the acceleration sensor, estimating a state of motion of the two-wheeler based on the detected rotation rates, wherein the state of motion comprises estimated values ​​for estimated acceleration values ​​and for an estimated speed and for an estimated distance traveled, first correcting the estimated state of motion based on the detected acceleration values, and determining a distance traveled by the two-wheeler, or a current speed of the two-wheeler and a distance traveled by the two-wheeler based on the corrected estimated state of motion.

[0005] Estimates are, in particular, determined or calculated values ​​of the respective parameters, i.e., the estimated acceleration values, the estimated speed, and the estimated distance traveled. In other words, an estimate is, in particular, a numerical value, preferably including the corresponding unit of measurement. In particular, the motion state comprises a separate estimate for each such parameter. In particular, the estimated values ​​are iteratively optimized by the method in order to be able to determine the desired parameters based on them.

[0006] Preferably, three-dimensional yaw rates are detected by means of the yaw rate sensor, each of which comprises a yaw rate about a longitudinal axis, which is aligned in particular in the direction of travel, about a vertical axis, and about a pitch axis, which is perpendicular to the longitudinal axis and the vertical axis.

[0007] In other words, in the method, the yaw rate sensor detects the yaw rates, in particular three-dimensional yaw rates, and based on these, a general motion state of the two-wheeler, which also includes other motion variables such as the estimated speed and the estimated distance traveled, is estimated. This estimated motion state is then corrected based on the additional acceleration values ​​available from the acceleration sensor, in particular based on a comparison of the estimated acceleration values ​​with the actual acceleration values. In particular, the estimated acceleration values ​​can be corrected directly based on the available measured acceleration values. At the same time, preferably based on this correction step, the other motion variables of the motion state are corrected, in particular the estimated speed and the estimated distance traveled.From this, the current speed and / or the distance currently covered by the two-wheeler can then be determined.

[0008] The method is therefore characterized by the fact that particularly comprehensive and precise sensor data about the bicycle's movement can be obtained using comparatively simple and inexpensive sensors. In particular, simple and inexpensive sensors can be used to record precise and high-resolution speed values ​​even at low speeds, which is particularly advantageous when used on a bicycle.

[0009] The subclaims contain preferred developments of the invention.

[0010] The method preferably further comprises the following step: a second correction of the motion state based on the measurement pulses detected by the wheel speed sensor. The second correction preferably occurs each time a measurement pulse detected by the wheel speed sensor is present. This allows for a particularly high degree of accuracy in determining the motion variables, since the measurement pulses detected by the wheel speed sensor, whenever such measurement pulses are present, provide particularly precise data that can be used to optimize the estimated motion state.

[0011] Particularly preferably, the second correction is carried out based on the following equation: y2 = [ x 5 , old + 2 π r] . Here, y2 is a corrected value for the distance traveled by the bicycle, x5,old is an old, i.e., past, value for the distance traveled by the bicycle, and r is the radius of a wheel of the bicycle. In particular, y2 is the estimated distance traveled in the motion state. In other words, at each time a measurement pulse is detected by the wheel speed sensor, the estimated distance traveled in the motion state is replaced by the exact measured value of the wheel speed sensor.

[0012] Preferably, one or more of the following motion variables of the two-wheeler are determined based on the corrected motion state: roll angle, pitch angle, and longitudinal acceleration. This allows particularly precise information to be obtained about the current motion of the two-wheeler.

[0013] More preferably, the first correction is performed using a nonlinear Kalman filter. This allows for a particularly efficient and precise correction of the motion state in a simple manner.

[0014] Preferably, the estimation of the motion state of the two-wheeler is carried out by means of a state vector x = x 1 x 2 x 3 x 4 x 5 , using an input vector u = u 1 u 2 u 3 , and based on the following system equation: x ˙ = u 1 + tan x 2 sin x 1 u 2 + tan x 2 cos x 1 u 3 cos x 1 u 2 − sin x 1 u 3 0 x 3 x 4 .

[0015] In particular, the input vector u can be used as input of the system equation ẋ Where x1 is a roll angle, x2 is a pitch angle, x3 is a longitudinal acceleration, x4 is a longitudinal velocity, and x5 is a distance traveled. Furthermore, u1, u2, and u3 are the three-dimensional angular rates. In particular, the system equation ẋ a temporal change in the state of motion.

[0016] Preferably, the estimation of the state of motion of the two-wheeler is based on a calculation of an integral of the system equation ẋ In particular, after integrating the system equation, the corresponding components of the state of motion can be used directly as estimated values ​​for the acceleration values, the speed, and the distance traveled.

[0017] Further preferably, the estimation of the state of motion of the two-wheeler is further carried out based on the following equations: Rx = 1 0 0 0 cos x 1 sin x 1 , Ry = cos x 2 0 − sin x 2 0 1 0 sin x 2 0 cos x 2 , ψ ˙ = u 2 sin x 1 + u 3 cos x 1 cos x 2 , y 1 = Rx Ry x 3 − x 4 ψ ˙ g .

[0018] This is ψ̇ a yaw rate of the two-wheeler, and y1 are the estimated acceleration values ​​y1 of the two-wheeler. The yaw rate is considered to be, in particular, the rotational speed of the two-wheeler around its longitudinal axis. In particular, the first correction is performed by correcting the estimated acceleration values ​​y1.

[0019] The method preferably further comprises the step of determining a standstill of the two-wheeler based on the estimated state of motion. Preferably, a distinction can additionally be made between a total of three driving modes: standstill, driving, and transition. The driving modes can preferably be determined based on predefined threshold values, for example, the estimated speed. This allows particularly simple and unambiguous information about the current driving state of the two-wheeler to be obtained. Furthermore, the determined driving states, such as standstill, can be used to further optimize the method to increase the accuracy of determining the state of motion.

[0020] More preferably, the method further comprises the steps: Reducing the state vector x and the system equation x to the following states: x = x 1 x 2 and x ˙ = u 1 + tan x 2 sin x 1 u 2 + tan x 2 cos x 1 u 3 cos x 1 u 2 − sin x 1 u 3 , if no measurement pulses are detected by the wheel speed sensor for at least a predefined period of time or if a standstill of the vehicle has been determined, and extending the state vector x and the system equation ẋ to the original conditions before reduction when measurement pulses are again detected by the wheel speed sensor.

[0021] This means that when the vehicle is at a standstill, or if the wheel speed sensor does not detect any measuring pulses for another reason, the state vector x and the system equation ẋ reduced to the respective first two states. This prevents drift in the estimated values ​​of the motion state, which can occur if correction is not possible due to a missing velocity signal.

[0022] Particularly preferably, the method further comprises the step of determining a steering angle of the two-wheeler based on the corrected motion state. The steering angle is defined as an angle between a longitudinal direction of the two-wheeler and a front wheel of the two-wheeler projected onto a plane perpendicular to the vertical axis, for example, onto a ground plane. The steering angle can be determined particularly precisely through the high-resolution and precise detection of the motion state.

[0023] The steering angle is preferably determined δ based on the following equation: δ = arctan ψ ˙ L x 4 , with the yaw rate ψ̇, a wheelbase L of the two-wheeler, and that of a longitudinal speed x4. In particular, the wheelbase corresponds to a distance between the two wheel hubs or axles of the two-wheeler. Preferably, it is additionally provided that the calculation of this equation is carried out in such a way that the expression in the denominator does not assume values ​​around zero in order to avoid numerical problems that may arise. Preferably, a determined minimum speed is used for this purpose as the longitudinal speed x4. Preferably, the calculation of the steering angle is only carried out during a detected movement of the two-wheeler, and in particular prevented when a detected standstill of the two-wheeler is detected.

[0024] Furthermore, the invention leads to a two-wheeler comprising a sensor system that has a yaw rate sensor, an acceleration sensor, and a wheel speed sensor. The two-wheeler also comprises a control device configured to carry out the described method for determining movement variables of the two-wheeler. The wheel speed sensor is preferably a single-pulse reed sensor that has precisely one magnet that is attached to the wheel and rotates with the wheel. The two-wheeler is characterized in that the movement variables can be determined with high temporal resolution and high accuracy, while the sensor system is particularly simple and cost-effective.

[0025] Preferably, the two-wheeler is designed as an electrically driven, in particular and / or drivable, bicycle, which can also be referred to in particular as an electric bicycle. Short description of the drawings

[0026] The invention is described below using exemplary embodiments in conjunction with the figures. In the figures, functionally identical components are identified by the same reference numerals. Here: Figure 1 shows a simplified schematic view of a two-wheeler with a sensor system and a control device for carrying out a method according to a preferred embodiment of the invention, Figure 2 shows an alternative view of the two-wheeler of the Figure 1 to illustrate a steering angle, Figure 3 an alternative view of the two-wheeler of the Figure 1 to illustrate an inclined position, and Figure 4 shows a simplified schematic view of an implementation of the method according to the preferred embodiment of the invention. Preferred embodiments of the invention

[0027] Figure 1shows a simplified schematic view of a two-wheeler 1 with a sensor system 2 and a control device 20 for carrying out a method for determining movement variables of the two-wheeler 1 according to a preferred embodiment of the invention.

[0028] The two-wheeler 1 is an electric bicycle, which has a drive unit 12 in the area of ​​a bottom bracket, by means of which the manually generated pedaling force of a rider of the two-wheeler 1 can be motor-assisted. The drive unit 12 is supplied with electrical energy by an electrical energy storage device 14.

[0029] The control device 20 is arranged on a handlebar of the two-wheeler 1 and can, for example, be part of an on-board computer.

[0030] The sensor system 2 comprises several sensors. Specifically, the sensor system 2 comprises a yaw rate sensor 21 and an acceleration sensor 22, both of which are integrated into the control device 20.

[0031] The yaw rate sensor 21 detects three-dimensional yaw rates of the two-wheeler 1 during a ride. In this case, a yaw rate is measured by the Figure 1 indicated axes x, y, z (see also Figures 2 and 3 ) recorded.

[0032] The x-axis is parallel to a longitudinal axis L of the two-wheeler 1 (see Figure 2 ), which is parallel to a direction of travel A when the two-wheeler 1 is traveling straight ahead. The z-axis corresponds to a vertical axis H (see Figure 3), which is particularly parallel to a gravitational direction of the Earth's gravity field (not shown). The y-axis is perpendicular to the x-axis and perpendicular to the z-axis. The y-axis can also be referred to as the pitch axis. Furthermore, the z-axis can also be referred to as the yaw axis.

[0033] By means of the acceleration sensor 22, acceleration values ​​of the two-wheeler 1 are recorded, preferably a total of three acceleration values ​​along each of the axes x, y, z.

[0034] The sensor system 2 further comprises a single-pulse wheel speed sensor 23, which is designed as a rotation sensor to detect exactly one measurement pulse per rotation of a wheel 11 of the two-wheeler 1. For this purpose, the wheel speed sensor 23 is configured to detect the measurement pulse exactly once per rotation of the wheel 11 each time it passes a magnet 23a, which is attached, for example, to a spoke of the wheel 11. Based on the measurement pulses detected by the wheel speed sensor 23, a rotational speed of the wheel 11 can thus be determined.

[0035] The movement variables of the two-wheeler 1 are determined by means of the method 50 as a current speed of the two-wheeler 1, a distance travelled and a current steering angle δ determined.

[0036] The steering angle δ is in the Figure 2 clarified. Figure 2 shows a view of the bicycle 1 along the z-axis. As in the Figure 2to recognize, the steering angle δ an angle between the longitudinal axis L and the front wheel 11. When driving straight ahead, the steering angle δ equal to zero, and correspondingly higher, the smaller the curve radius of the curve through which the two-wheeler 1 is driven.

[0037] When cornering with the two-wheeler 1, the two-wheeler 1 is as shown in the Figure 3 shown, placed in an inclined position. Figure 3 shows schematically an angle of inclination β of the two-wheeler 1. The angle of inclination β is the angle by which the two-wheeler 1 is inclined from the vertical axis H.

[0038] The implementation of the method 50 for determining the movement quantities of the two-wheeler 1 is described below with reference to the Figure 4 described.

[0039] In the method 50, the three-dimensional rotation rates of the two-wheeler 1 are first detected 51 by means of the rotation rate sensor 21. At the same time, the acceleration values ​​of the two-wheeler 1 are detected 52 by means of the acceleration sensor 22. Based on the detected three-dimensional rotation rates, an estimation 53 of a motion state of the two-wheeler 1 is then carried out.

[0040] The motion state of the two-wheeler 1 includes estimated values ​​for estimated acceleration values ​​and an estimated speed, as well as an estimated distance traveled. In detail, the motion state is estimated using a state vector that has the following parameters: roll angle, pitch angle, longitudinal acceleration, longitudinal speed, and distance traveled. In particular, the roll angle corresponds to the inclination angle β, i.e., a deflection or rotation of the two-wheeler 1 about the longitudinal axis H. Preferably, the pitch angle corresponds to a deflection or rotation of the two-wheeler 1 about the Y-axis, i.e., transverse to the longitudinal axis H.

[0041] Based on the state vector and an input vector, where the input vector has the three-dimensional rotation rates, a system equation is subsequently created which in particular represents a temporal change of the state vector.

[0042] The motion state of two-wheeler 1 is then estimated by calculating an integral of this system equation. This yields the estimated motion quantities of two-wheeler 1.

[0043] Subsequently, correction steps 54, 55 of the motion state follow. First, a first correction 54 of the motion state is performed based on the acceleration values ​​actually detected by the acceleration sensor 22.

[0044] In addition, a second correction 55 of the motion state occurs each time a measurement pulse from the wheel speed sensor 23 is detected. Specifically, the motion state is corrected based on the actual distance traveled determined by the wheel speed sensor 23. Since the actual distance traveled can be determined very precisely due to the geometric relationship of the measurement pulses across the circumference of the wheel 11, a particularly precise correction step of the motion state can be performed by means of the second correction 55.

[0045] Subsequently, based on the corrected movement state, the determination 57 of the steering angle δ of the two-wheeler 1.

[0046] Method 50 can also be implemented in a modification (not shown) that additionally takes into account a standstill of the two-wheeler 1. In this case, a standstill of the two-wheeler 1 is additionally determined based on the estimated state of motion.

[0047] If it has been determined that the two-wheeler 1 is stationary, the state vector and the system equation can be reduced to the first two states. This prevents the estimated movement variables from drifting over time due to the absence of a measurement pulse, which can be used for the second correction 55. As soon as it has been determined that the two-wheeler 1 is moving again, or as soon as a measurement pulse is again detected by the wheel speed sensor 23, the state vector and the system equation are expanded back to the original states before the reduction, thus subsequently enabling an accurate determination of all movement variables.

[0048] As an alternative to a standstill of the two-wheeler 1, the absence of a measuring pulse from the wheel speed sensor 23 can also be used to reduce the state vector and the state equation to the first two states.

[0049] The singly or doubly corrected motion state thus provides particularly accurate estimates for the motion variables of the two-wheeler 1. In particular, a desired motion variable, such as speed, can be read off at any time based on the corrected motion state and used, for example, for other systems or methods of the two-wheeler 1. Furthermore, the method 50 can be used to precisely determine the motion variables of the two-wheeler 1 even at very low speeds, since the method 50 is based in particular on the measured values ​​of the yaw rate sensor 21 and the acceleration sensor 22, which can deliver precise and reliable measured values ​​even at low speeds.

Claims

1. Method for ascertaining movement variables of a two-wheeled vehicle (1), - wherein the two-wheeled vehicle (1) comprises a sensor system (2) having a rotation rate sensor (21), an acceleration sensor (22), and a wheel speed sensor (23), - wherein the wheel speed sensor (23) is designed to detect at least one measurement pulse per revolution of a wheel (11) of the two-wheeled vehicle (1), and - wherein the method comprises the following steps: - detecting (51) rotation rates, in particular three-dimensional rotation rates, of the two-wheeled vehicle (1) by means of the rotation rate sensor (21), - detecting (52) acceleration values of the two-wheeled vehicle (1) by means of the acceleration sensor (22), characterized by: - estimating (53) a motion state of the two-wheeled vehicle (1) on the basis of the detected rotation rates, - wherein the motion state comprises estimated values for estimated acceleration values and for an estimated speed and for an estimated distance covered, - first correcting (54) of the estimated motion state on the basis of the detected acceleration values, and - ascertaining (56) a distance covered by the two-wheeled vehicle (1) or an instantaneous speed of the two-wheeled vehicle (1) and a distance covered by the two-wheeled vehicle (1) on the basis of the corrected estimated motion state.

2. Method according to Claim 1, further comprising the following step: - second correcting (55) of the estimated motion state on the basis of the measurement pulses detected by means of the wheel speed sensor (23).

3. Method according to Claim 2, wherein the second correcting (55) is carried out on the basis of the following equation: y 2 = x 5 , old + 2 π r with a corrected value for a distance y2 covered by the two-wheeled vehicle (1), an old value for a distance x5,old covered by the two-wheeled vehicle (1), and a radius r of a wheel (11) of the two-wheeled vehicle (1).

4. Method according to any of the preceding claims, wherein one or more of the following motion variables of the two-wheeled vehicle (1) are ascertained on the basis of the corrected estimated motion state: roll angle, pitch angle and longitudinal acceleration.

5. Method according to any of the preceding claims, wherein the first correcting (54) is carried out by means of a non-linear Kalman filter.

6. Method according to any of the preceding claims, wherein estimating (53) the motion state of the two-wheeled vehicle (1) is effected by means of a state vector x = x 1 x 2 x 3 x 4 x 5 , with a roll angle x1, a pitch angle x2, a longitudinal acceleration x3, a longitudinal speed x4, and a distance covered x5, and using an input vector u = u 1 u 2 u 3 , with the three-dimensional rotational rates u1, u2, and u3, and based on the following system equation: x ˙ = u 1 + tan x 2 sin x 1 u 2 + tan x 2 cos x 1 u 3 cos x 1 u 2 − sin x 1 u 3 0 x 3 x 4 .

7. Method according to Claim 6, wherein estimating (53) the motion state of the two-wheeled vehicle (1) is effected on the basis of a calculation of an integral of the system equation ẋ.

8. Method according to Claim 7, wherein estimating (53) the motion state of the two-wheeled vehicle (1) is furthermore carried out on the basis of the following equations: Rx = 1 0 0 0 cos x 1 sin x 1 , Ry = cos x 2 0 − sin x 2 0 1 0 − sin x 2 0 cos x 2 , ψ ˙ = u 2 sin x 1 + u 3 cos x 1 cos x 2 , y 1 = Rx Ry x 3 − x 4 ψ ˙ g , with a yaw rate ψ̇ of the two-wheeled vehicle (1), and the estimated acceleration values y1 of the two-wheeled vehicle (1).

9. Method according to any of Claims 6 to 8, furthermore comprising the following steps: - reducing the state vector x and the system equation ẋ to the following states: x = x 1 x 2 and x ˙ = u 1 + tan x 2 sin x 1 u 2 + tan x 2 cos x 1 u 3 cos x 1 u 2 − sin x 1 u 3 , if the wheel speed sensor (23) detects no measurement pulses over a predetermined period of time, or if a standstill of the vehicle has been ascertained, and - expanding the state vector x and the system equation x to the original states before the reduction if measurement pulses are again detected by the wheel speed sensor (23).

10. Method according to any of the preceding claims, furthermore comprising the following step: - ascertaining a standstill of the two-wheeled vehicle (1) on the basis of the estimated motion state.

11. Method according to any of the preceding claims, furthermore comprising the following step: - ascertaining (57) a steering angle of the two-wheeled vehicle (1) on the basis of the corrected estimated motion state.

12. Method according to Claims 6 to 8 and 11, wherein the ascertaining (57) of the steering angle δ is carried out on the basis of the following equation: δ = arctan ψ ˙ L x 4 , with a wheelbase L of the two-wheeled vehicle (1).

13. Two-wheeled vehicle, comprising: - a sensor system (2) having a rotation rate sensor (21), an acceleration sensor (22), and a wheel speed sensor (23), and - a control device (20) configured to carry out a method according to any of the preceding claims.

14. Two-wheeled vehicle according to Claim 13, characterized in that the two-wheeled vehicle is in the form of an electrically driven bicycle.