Method for determining driving dynamics state of bicycle
Patent Information
- Application Number
- JP2022066068
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2021-04-14
- Filing Date
- 2022-04-13
- Publication Date
- 2025-04-16
AI Technical Summary
Existing methods for determining the driving dynamics state of bicycles, particularly eBikes, are hindered by the need for costly and bulky measurement equipment and high-resolution sensors, which are impractical for installation on bicycles.
A method and device using an inertial measurement device and incremental encoder to provide signals for estimating driving dynamics state, incorporating a Kalman filter with pseudo-measurements to avoid signal drift during stationary conditions, allowing accurate estimation of driving dynamics at various speeds and eliminating the need for separate estimation methods during static conditions.
Enables accurate and reliable determination of driving dynamics states on bicycles, including at high and low speeds, avoiding signal drift and reducing costs by using affordable components and methods.
Smart Images
Figure 00000000_0000_ABST
Abstract
Description
[Technical Field]
[0001] The present invention is a method for determining the driving dynamics state of a bicycle, - A step of providing signals of acceleration in at least one spatial direction and rotational velocity in at least one spatial direction using an inertial measurement device, - A step of providing a speed signal for an incremental encoder, - A step of determining the driving dynamics state based on an estimation method, based on the provided signals. Regarding methods including
[0002] Furthermore, the present invention relates to an apparatus for determining the driving dynamics state of a bicycle, comprising: an inertial measuring device configured to provide signals of acceleration in at least one spatial direction and rotational velocity in at least one spatial direction; an incremental encoder configured to provide a velocity signal; and a state determination device configured to determine the driving dynamics state based on an estimation method, based on the provided signals.
[0003] Furthermore, this invention relates to bicycles. [Background technology]
[0004] While this invention is applicable to any estimation method, we will describe the estimation method using a Kalman filter. The present invention is applicable to any bicycle, particularly eBikes, Pedelecs, motorcycles, etc., but will focus on eBikes.
[0005] Key variables describing the dynamic behavior or driving dynamics of an eBike include, among others, the speed variable, and the roll angle variable, which reflects the lateral tilt of the bicycle. This allows for improvements to various functions of the eBike, while on the other hand, some functions are only possible with accurate knowledge of the driving dynamics. However, these driving dynamics cannot usually be measured directly, or measuring them requires measuring equipment that is too large to install on an eBike, and too expensive. [Overview of the project]
[0006] It is well known that to determine driving dynamics, for example, cameras, GPS, inertial sensors, or speed sensors are used, and the data from these sensors is evaluated. Depending on the presence of the corresponding sensors, high-resolution signals from the sensors are required, which is also costly.
[0007] In one embodiment, the present invention is a method for determining the driving dynamics state of a bicycle, The steps include providing signals of acceleration in at least one spatial direction and rotational velocity in at least one spatial direction using an inertial measurement device, The steps include providing a speed signal for an incremental encoder, A method comprising the step of determining the driving dynamics state based on an estimation method based on the provided signal, Once the current riding state is determined, and the determined current riding state is that the bicycle is stationary, an alternative speed signal is provided in place of the incremental encoder speed signal for the estimation method in order to estimate the driving dynamics state. Provide a method.
[0008] In a further embodiment, the present invention provides an apparatus for determining the driving dynamics state of a bicycle, comprising: an inertial measuring device configured to provide signals of acceleration in at least one spatial direction and rotational velocity in at least one spatial direction; an incremental encoder configured to provide a velocity signal; and a state determination device configured to determine the driving dynamics state based on an estimation method, wherein the current driving state is determined, and when the bicycle is stationary as the determined current driving state, an alternative velocity signal is used in place of the velocity signal of the incremental encoder with respect to the estimation method in order to estimate the driving dynamics state.
[0009] In a further embodiment, the present invention provides a bicycle equipped with the device described in claim 11. One potential advantage is that accurate estimation of the driving dynamics state on a bicycle becomes possible for each riding condition, both at high and low speeds, especially on inclines. A further advantage is that drift in the speed estimation is avoided, especially when stationary or at very low speeds. By providing an alternative speed signal, the driving dynamics state can still be reliably and accurately identified. This avoids the use of alternative estimation methods when stationary. Another advantage is that when transitioning from a stationary state to a riding state with positive velocity, the provision of an alternative speed signal eliminates the need to correct the drifted estimated speed signal, thus enabling a reliable determination of the driving dynamics state.
[0010] The term “bicycle” should be understood in its broadest sense, and in particular herein, preferably in the claims, relating to a bicycle having at least two wheels that can be operated manually and / or by a drive mechanism. The term “bicycle” also refers in particular to eBikes, Pedelecs, and motorcycles.
[0011] Further features, advantages, and embodiments of the present invention are described or disclosed below. According to an advantageous development of the present invention, the driving dynamics state is determined based on at least one of the variables mileage, yaw rate, roll angle, and / or pitch angle, as well as the first time derivative of each variable. Thus, each driving dynamics state can be reliably and sufficiently accurately identified.
[0012] In a more advantageous development of the present invention, the second time derivatives of each variable are further determined in order to estimate the driving dynamics state. One of the advantages that may arise from this is that the driving dynamics state can be identified more accurately.
[0013] According to a more advantageous development of the present invention, in order to estimate the driving dynamics state, possible changes in the variables, particularly the second time derivatives of the variables, are taken into account with respect to additional noise terms. This makes it easy to account for inaccuracies in modeling or in determining the driving dynamics state.
[0014] According to a more advantageous development of the present invention, the stationary state of the bicycle is - Fluctuations in the acceleration signal below a predetermined value, and / or -Fluctuations in the rotational speed signal below a predetermined value, and / or - Driver torque exceeding a predetermined value and driver cadence below a predetermined value It is determined by [the following].
[0015] This makes it possible to easily and reliably determine or detect a stationary state. According to a more advantageous development of the present invention, the estimation method includes the use of a Kalman filter, particularly a nonlinear Kalman filter. Its advantages are sufficient accuracy of estimation and, at the same time, acceptable computational resource requirements.
[0016] According to a further advantageous development of the invention, the incremental encoder is provided as a single-pulse incremental encoder and particularly includes lead contacts and / or rim magnets on the wheels of a bicycle. The advantage is that it is a simple and inexpensive incremental encoder.
[0017] According to a further advantageous development of the invention, in the estimation method, the driving dynamics state is estimated based on a model, and when the measured value is changed, the driving dynamics state estimated based on the signal is adapted. This enables a reliable and accurate estimation of the driving dynamics state.
[0018] According to a further advantageous development of the invention, the driving dynamics state is determined based on at least one sensor-specific parameter of the sensor, particularly the offset of the sensor and / or the position of the sensor on a bicycle. This enables a more accurate determination of the driving state.
[0019] According to a further advantageous development of the invention, when determining the current driving state, the lateral and / or vertical speed of the rear wheel of the bicycle is set to be zero. The advantage is that the driving dynamics state can be determined more simply and quickly.
[0020] The further important features and advantages of the invention will become apparent from the dependent claims, from the drawings, and from the description of the related drawings based on the drawings. It should be understood that, without departing from the scope of the invention, the features described above and below can be used not only in the presented combinations but also in other combinations or individually.
[0021] The preferred forms and embodiments of the invention are shown in the drawings and will be described in more detail in the following description, where the same reference numerals represent the same or similar or functionally identical components or elements.
Brief Description of the Drawings
[0022] [Figure 1A] This figure shows a comparison of various driving dynamics state variables between estimated and reference values, as well as their estimation errors, determined using the method according to an embodiment of the present invention. [Figure 1B] This figure shows a comparison of various driving dynamics state variables between estimated and reference values, as well as their estimation errors, determined using the method according to an embodiment of the present invention. [Figure 2] This figure shows the speed of a bicycle based on a reference value, an estimated value, and an estimated value with a speed substitute signal, determined using the method according to an embodiment of the present invention. [Figure 3] This figure shows the steps of a method for determining the driving dynamics state of a bicycle according to an embodiment of the present invention. [Modes for carrying out the invention]
[0023] Figures 1A and 1B show a comparison of various driving dynamics state variables between estimated and reference values, and their estimation errors, determined using the method according to embodiments of the present invention.
[0024] Figures 1A and 1B show, from top to bottom, the driving dynamics variables roll angle, pitch angle, and velocity plotted separately over specific time windows. On the left side of Figures 1A and 1B, the reference and estimated values for the corresponding variables are recorded within the graphs. On the right side of Figures 1A and 1B, the absolute estimation errors for each variable are plotted.
[0025] We can clearly see the excellent agreement between the estimated and reference values of each variable, as well as the small absolute errors of each estimated variable. The estimation method for estimating the driving dynamics state of a bicycle with front and rear wheels is based here on a nonlinear Kalman filter with state constraints for use with bicycles. The estimation method uses information about the current driving state, particularly "driving" or "stationary," to add a pseudo-measurement or speed substitute signal for the stationary state, thereby avoiding drift in the speed signal when stationary. When stationary, no signal is generated from the incremental encoder, which can lead to drift in the estimated speed. By inserting the pseudo-measurement or speed substitute signal, all states of the bicycle can still be identified, and there is no need to use a second Kalman filter in particular. This also avoids switching between Kalman filters.
[0026] In detail, in the Kalman filter, a state vector relating to the driving dynamics state is estimated in the prediction step based on the system model, and then the estimated state is modified using the measurement model and existing measurements (if new measurements exist). For example, for accurate estimation of the velocity, roll angle, pitch angle, and yaw rate states, further states appearing in the accurate measurement model must also be estimated. The estimated state vector is constructed as follows:
[0027]
number
[0028] Here, s is the distance the rear wheel contact point has advanced, ν x α is the velocity of the rear wheel contact point in the direction of the bicycle (corresponding to the speed of the bicycle), x This is the acceleration at the rear wheel contact point in the direction of the bicycle.
[0029]
number
[0030] is yaw rate,
[0031]
number
[0032] φ is the yaw acceleration, φ is the roll angle.
[0033]
number
[0034] is roll rate,
[0035]
number
[0036] θ is the roll acceleration, θ is the pitch angle.
[0037]
number
[0038] is pitch rate,
[0039]
number
[0040] This is the pitch acceleration. This state vector can be extended with further states, such as sensor offsets or system parameters, for example, the position of the inertial measurement unit for measuring acceleration and rotational velocity, if necessary.
[0041] In the following, the order of rotation is yaw-roll-pitch, the inertial frame is a North-East-Down coordinate system, the bicycle frame has its origin at the rear wheel hub, the x-axis of the bicycle frame indicates the direction of travel, the y-axis points to the right, and the z-axis points down.
[0042] In this case, the continuous system model of the Kalman filter is as follows:
[0043]
number
[0044] Here, w α ,
[0045]
number
[0046] ,
[0047]
number
[0048] , and
[0049]
number
[0050] This represents the noise term in various acceleration models. The noise term is used to account for inaccuracies in the system modeling. Acceleration is modeled as if it were constant, while the noise term allows for variation within certain limits.
[0051] To enable the use of the system model for the prediction step of the Kalman filter, the system model is discretized using a known method. Various sensor measurement models are used in the correction step of the Kalman filter.
[0052] Regarding the lead sensor, the following measurement model can be obtained. θ R =-s / r R -θ r R θ is the rear wheel radius.R is the rear wheel rotation angle. This rear wheel rotation angle is updated when a new lead pulse (or another pulse) exists.
[0053] θ R,new = θ R,lastPulse + 2π The measurement model of the rotational speed sensor is as follows.
[0054]
Number
[0055] The measurement model of the acceleration sensor incorporates the state constraints of the bicycle. Here, in particular, it is assumed that the rear wheel is not skidding, that is, the lateral velocity ν y = 0 at the rear wheel contact point. To derive the measurement model of the acceleration sensor, first, the positions of the inertial measurement unit in the bicycle coordinate system and the inertial / world coordinate system are specified. Here, hereinafter, the z - direction dynamics in the inertial / world coordinate system of the bicycle and the associated change in the z - coordinate are ignored.
[0056]
Number
[0057] The distances x IMU , y IMU , and z IMU between the rear wheel hub and the inertial measurement unit are fixed, and x and y are the coordinates of the rear wheel contact point in the inertial system. R is a rotation matrix representing the position of the bicycle in space.
[0058] By differentiating the position of the inertial measurement unit with respect to time, the velocity of the inertial measurement unit is obtained.
[0059]
Number
[0060] Here,
[0061]
number
[0062] and
[0063]
number
[0064] (Velocity of the rear wheel contact point)
[0065]
number
[0066] Replace with. As mentioned above, we assumed there was no slippage in the rear wheels, so the lateral velocity is zero (ν y It is set to (=0). Further time differentiation gives the sensor's acceleration in world coordinates.
[0067]
number
[0068] To obtain a measurement model for the acceleration sensor, gravitational acceleration g is considered, and the acceleration is transformed into the sensor coordinate system using a rotation matrix R that represents the position of the bicycle.
[0069]
number
[0070] The measurement equation for the acceleration sensor does not depend on the coordinates (x,y) of the rear wheel contact point and the yaw angle ψ, and can therefore be expressed by the conditions described above. Each measurement and measurement model is used only in the modification step when new information exists in the corresponding sensor.
[0071] If an incremental encoder stops generating pulses, the stationary state must first be recognized to avoid drift in the velocity signal during a stationary state. One or more of the following options can be used for this purpose.
[0072] - Small fluctuations in acceleration signal → No movement / vibration caused by cycling. - Small fluctuation in rotational speed signal → No movement of the bicycle. - Existing driver torque, driver cadence = 0 → The driver's feet are on the pedals with the brakes applied, and the bicycle is stationary.
[0073] When a stationary state is recognized, a pseudo-measured velocity or alternative velocity signal is fed to the Kalman filter. This causes the estimated velocity signal to approach zero when stationary. The corresponding measurement models are as follows:
[0074]
number
[0075] In a further embodiment, rotational acceleration in the measurement equation of an acceleration sensor.
[0076]
number
[0077] ,
[0078]
number
[0079] , and
[0080]
number
[0081] By ignoring this, we can reduce the state vector to three states. In this case, the system model is fitted using "constant rotational velocity" instead of "constant rotational acceleration" as appropriate. This improves the efficiency of the estimation method.
[0082] In further embodiments, the z-direction dynamics of the bicycle can also be taken into consideration. In this case, it is assumed that the rear wheel of the bicycle is always in contact with the road, i.e., not lifted off the ground. Accordingly, the vertical velocity of the rear wheel contact point is zero (ν z It is assumed that (=0). This assumption is used in the acceleration measurement equation a IMU This is incorporated into the derivation. Furthermore, when converting velocity from inertia / world coordinates to bicycle coordinates, the road gradient or the bicycle's pitch angle is also taken into further consideration.
[0083] Figures 1A and 1B show a comparison between the estimated state and the reference value, as described above. Figure 2 below shows the effect of the pseudo-measured value in the stationary state. Figure 2 shows the bicycle speed based on a reference value, an estimated value, and an estimated value with a speed substitute signal, as determined using the method according to an embodiment of the present invention.
[0084] In the upper part of Figure 2, reference value 1 and estimated values 2 and 3 for speed are plotted over a time window, with the latter one plotted using an alternative speed signal (curve 2) and the latter without an alternative speed signal (curve 3). In the lower part of Figure 2, the reed pulse 5 is plotted over the corresponding time window, and no reed pulse 5 occurs in the range of approximately 623 seconds to 667 seconds, thus providing an alternative speed pulse 4. In other words, in this range, the bicycle is stationary, and the reed sensor no longer transmits speed signals. It can be clearly seen that using the alternative speed signal results in a more accurate estimated driving dynamics state and avoids drift in the estimated driving dynamics state.
[0085] Figure 3 shows the steps of a method for determining the driving dynamics state of a bicycle according to an embodiment of the present invention. Figure 3 details the steps of the method for determining the driving dynamics state of a bicycle.
[0086] This delicious, - Step S1 provides signals of acceleration in at least one spatial direction and rotational velocity in at least one spatial direction using an inertial measurement device, - Step S2 provides the speed signal of the incremental encoder, -The step includes determining the driving dynamics state based on an estimation method based on the provided signal, Once the current driving state S4 is determined, and the determined current driving state is that the bicycle is stationary, an alternative speed signal is provided in place of the speed signal of the incremental encoder for the purpose of estimating the driving dynamics state.
[0087] In summary, at least one embodiment of the present invention has at least one of the following features and / or provides at least one of the following advantages. - Simple, reliable, and accurate determination of the bicycle's driving dynamics state.
[0088] - Avoiding drift in speed estimation. While the present invention has been described based on preferred exemplary embodiments, it is not limited thereto and can be modified in various ways. [Explanation of Symbols]
[0089] 1 Reference value 2. Estimates using alternative velocity signals 3. Estimates without using alternative velocity signals 4. Alternative velocity pulses 5 Lead pulses
Claims
1. 1. A method for determining a driving dynamics state of a bicycle, comprising: Providing signals of acceleration in at least one spatial direction and of rotational velocity in at least one spatial direction by means of an inertial measurement device (S1); providing an incremental encoder velocity signal (S2); and (S3) determining the driving dynamics state based on the provided signal based on an estimation method, A current riding state (S4) is determined, and when the bicycle is stationary as the determined current riding state, a substitute speed signal is provided for the estimation method instead of the speed signal of the incremental encoder in order to estimate the driving dynamics state. method.
2. The method of claim 1 , wherein the driving dynamics state is determined based on at least one of the following variables: distance traveled, yaw rate, roll angle, and / or pitch angle, and a first order time derivative of each said variable.
3. The method of claim 2 , further comprising determining a second time derivative of each of the variables to estimate the driving dynamics state.
4. 2. The method according to claim 1, wherein possible changes of the variables, in particular their second time derivatives, with respect to an additional noise term are taken into account to estimate the driving dynamics state.
5. The stationary state of the bicycle is Fluctuations in the acceleration signal below a predetermined value, and / or Fluctuations in the rotational speed signal below a predetermined value, and / or Rider torque above a certain value and cadence below a certain value The method of claim 1 , wherein the
6. The method according to claim 1 , wherein the estimation method comprises the use of a Kalman filter, in particular a non-linear Kalman filter.
7. 2. The method according to claim 1, wherein the incremental encoder is provided as a single pulse incremental encoder, in particular comprising a reed contact and / or a rim magnet at a bicycle wheel.
8. 2. The method of claim 1, wherein in the estimation method, the driving dynamics states are estimated based on a model and the estimated driving dynamics states are adapted based on the signals when measurements are changed.
9. The method according to claim 1 , wherein the driving dynamics state is determined based on at least one sensor-specific parameter of a sensor, in particular the offset of the sensor and / or the position of the sensor on the bicycle.
10. The method of claim 1 , wherein when determining the current riding state, the lateral and / or vertical speed of the rear wheel of the bicycle is set as non-existent.
11. 1. An apparatus for determining driving dynamics conditions of a bicycle, comprising: an inertial measurement device configured to provide signals of acceleration in at least one spatial direction and rotation rate in at least one spatial direction; an incremental encoder configured to provide a velocity signal; a state determination device configured to determine a driving dynamics state based on the provided signal and based on an estimation method; a current riding state is determined, and when the bicycle is stationary as the determined current riding state, a substitute speed signal is used instead of the speed signal of the incremental encoder for the estimation method to estimate the driving dynamics state. Device.
12. A bicycle comprising a device according to claim 11.