Estimation apparatus and method for determining motion estimate

Abstract

The invention further relates to an estimation device (1) for determining a new motion estimation value (vvirt) which indicates a speed of a rail vehicle (200) having at least two axles, comprising: a Kalman filter (30) which applies at least torque values (mpi, mpi) on an input side; the torque values (mpi, mi), the vertical force values (Qi), the wheel size values (ri) and the wheel rotational speed values (omega i) are determined on the basis thereof and on the basis of Kalman filtering, a slip module (40) which applies the torque values (mpi, mi, mi) on the input side, and the slip module (40) determines a Kalman estimated vehicle speed value (vk) on the basis of the torque values (mpi, mi, mi). Mei), a wheel size value (ri), a wheel rotational speed value ([omega] i) and a respective previous motion estimation value (vvirt) of the device (1) are estimated and a wheel circumferential speed value (vumauf) and slip information indicative of a possible slip of the respective axle are determined, and an evaluation module (50), and determining a new motion estimation value (vvirt) of the estimation device (1) on the basis of the Kalman estimated vehicle speed value (vk), the wheel circumferential speed value (vumauf) and the slip information.

Description

[0001] In the field of railway technology, it is very important to determine the speed of rail vehicles relatively accurately in order to be able to meet the requirements that operators usually have today.

[0002] The object of the present invention is to provide an estimation device by means of which a movement estimate describing the movement of a rail vehicle can be determined with relatively high accuracy.

[0003] This object is achieved according to the invention by an estimation device having the features according to claim 1. Advantageous embodiments of the estimation device according to the invention are specified in the dependent claims.

[0004] According to this, an estimation device is provided for determining a new movement estimate which indicates the speed of a rail vehicle having at least two axles, wherein the estimation device comprises:

[0005] a Kalman filter which receives at least the torque value, the vertical force value, the wheel dimension value and the wheel speed value on the input side and determines a Kalman estimated vehicle speed value based on these and based on the Kalman filtering,

[0006] a slip module which applies the torque values, the wheel dimension values, the wheel speed values ​​and the corresponding previous movement estimates of the estimation device on the input side and determines wheel peripheral speed values ​​and slip information indicating a possible slip of the corresponding axle, and

[0007] An evaluation module which determines a new motion estimate of the estimation device based on the Kalman estimated vehicle speed value, said wheel peripheral speed value and the slip information.

[0008] An essential advantage of the estimation device according to the invention is that, by combining the Kalman filter provided according to the invention with the slip module according to the invention, very precise motion estimates that can be used for the operation of the rail vehicle can be determined with little hardware and time expenditure.

[0009] It is considered advantageous with respect to the slip module that the slip module outputs the wheel peripheral speed value of the axle with the maximum wheel peripheral speed value to the evaluation module when the rail vehicle is in braking operation and all axles are not slipping, outputs the wheel peripheral speed value of the axle with the minimum wheel peripheral speed value to the evaluation module when the rail vehicle is in driving operation and all axles are not slipping, and outputs full slip information to the evaluation module when all axles are slipping.

[0010] It is considered advantageous with respect to the evaluation module that the evaluation module outputs the wheel peripheral speed value of the axle with the maximum wheel peripheral speed value as a new motion estimate value when the rail vehicle is in braking operation and all axles are not slipping, outputs the wheel peripheral speed value of the axle with the minimum wheel peripheral speed value as a new motion estimate value when the rail vehicle is in driving operation and all axles are not slipping, and determines the new motion estimate value based on the vehicle speed value of the Kalman estimate of the Kalman filter when all axles are slipping.

[0011] Advantageously, the estimation device has a vertical force calculation module, which generates a vertical force value for each axle based at least on the basis of a torque value, a wheel size value and a corresponding previous motion estimation value of the estimation device, which is also used by the Kalman filter, wherein each vertical force value of the vertical force calculation module forms one of a plurality of vertical force values ​​used by the Kalman filter.

[0012] It is also advantageous if the estimation device has a friction braking torque calculation module which generates a friction braking torque value for each axle based at least on the brake pressure value for each axle, the wheel size value and a corresponding previous motion estimation value of the estimation device, wherein each friction braking torque value forms one of a plurality of torque values ​​used by the Kalman filter.

[0013] The Kalman filter preferably also uses axle-dependent engine torque values ​​as torque values, or at least also takes these into account.

[0014] The Kalman filter is preferably an extended Kalman filter.

[0015] The estimation device preferably outputs a vehicle acceleration value indicating the acceleration of the rail vehicle as a further movement estimation value.

[0016] Furthermore, it is considered advantageous if the Kalman filter also determines a friction fit value per axle which is related to the force-rail contact, and the estimation device also outputs the friction fit value.

[0017] The estimation device preferably includes a computing device and a memory, in which are stored: a Kalman filter software module, which forms a Kalman filter when executed by the computing device; a sliding software module, which forms a slip module when executed by the computing device; and an evaluation software module, which forms an evaluation module when executed by the computing device.

[0018] Also preferably stored in the memory is a vertical force calculation software module which, when executed by the computing device, forms a vertical force calculation module.

[0019] Preferably, a friction braking torque calculation software module is also stored in the memory, which, when executed by the computing device, forms a friction braking torque calculation module.

[0020] The invention also relates to a rail vehicle equipped with an estimation device as described above.

[0021] With regard to the advantages of the rail vehicle according to the invention and its advantageous designs, reference is made to the above remarks in connection with the estimation device according to the invention and its advantageous designs.

[0022] Preferably, the slip controller is connected to the estimation device, and controls the slip of the corresponding axle by using the estimated friction fit value of each axle, the wheel peripheral speed value and the corresponding movement estimation value of the estimation device.

[0023] It is also advantageous if a plausibility check device is connected to the estimation device and which performs a plausibility check on the motion information from other sources using the motion estimate of the estimation device.

[0024] The present invention also relates to a method for determining a new motion estimate value, which indicates the speed of a rail vehicle having at least two axles, in which method at least a torque value, a vertical force value, a wheel size value and a wheel speed value are input to a Kalman filter on the input side, and a Kalman-estimated vehicle speed value is determined based on this and on the basis of the Kalman filter, a wheel circumferential speed value and slip information are determined using the wheel size value, the wheel speed value and the previously determined motion estimate value for each axle, the slip information indicating a possible slip of the corresponding axle, and a new motion estimate value is determined based on the Kalman-estimated vehicle speed value, the wheel circumferential speed value and the slip information.

[0025] With regard to the advantages of the method according to the invention and its advantageous configurations, reference is made to the above remarks in connection with the estimation device according to the invention and its advantageous configurations.

[0026] The present invention is further described below based on the embodiments, wherein:

[0027] Figure 1 An embodiment of an estimation device according to the present invention is shown in the form of a block diagram;

[0028] Figure 2 A computing facility is shown in which implementation according to Figure 1 The estimation device and accordingly form the basis Figure 1 The estimation device,

[0029] Figure 3 Show equipped with Figure 1 and Figure 2 The estimation device for the rail vehicle.

[0030] For the sake of clarity, the same reference numbers are used in the drawings for the same or similar components.

[0031] Figure 1An embodiment of an estimation device 1 according to the present invention is shown in the form of a block diagram. The estimation device 1 comprises a friction braking torque calculation module 10 , a vertical force calculation module 20 , an extended Kalman filter 30 , a slip module 40 and an evaluation module 50 .

[0032] The estimation device 1 has an input terminal for

[0033] - brake pressure value in the form of brake cylinder pressure p_i,

[0034] -Electric engine torque value m_e_i (hereinafter also referred to as engine torque value),

[0035] For example, wheel size values ​​in the form of dynamic wheel radius r_i, and

[0036] - Wheel speed value ω_i, also referred to as axle speed below.

[0037] The index i indicates the corresponding axle. Figure 1 In the embodiment of FIG. 5 , it is assumed, for example, that the rail vehicle in which the estimation device is used has four axles, that is, i can take a value between 1 and 4.

[0038] The estimation device 1 has an output terminal for

[0039] - Friction fit value in the form of adhesion coefficient μ_i, which describes the wheel / rail contact,

[0040] - a new motion estimate in the form of a velocity estimate v_virt, and

[0041] - A new motion estimate in the form of the acceleration estimate a_virt.

[0042] In the friction braking torque calculation module 10 , the friction braking torque value m_p_i per axle is calculated from the measured brake cylinder pressure p_i, the dynamic wheel radius r_i, geometric constants (eg friction radius) and the speed-dependent friction coefficient between the wheel and the brake disk.

[0043] In the vertical force calculation module 20, the dynamic wheel support force Q is calculated from the friction braking torque value m_p_i, the motor torque m_e_i, the dynamic wheel radius r_i, the axle speed ω_i, the known driving resistance, the known car body mass and the known geometric coefficients. i .

[0044] Driving resistance F w It can be determined approximately by empirically derived constants and from the corresponding vehicle speed.

[0045] The Kalman filter 30 is implemented based on an observer in control technology. The equations of motion for the wheels and the carriage are preferably used as mathematical model according to the following equations (1)-(5):

[0046]

[0047] In the translation direction, according to equation (1), Newton's law can be deduced:

[0048] - Driving resistance (F w ) and,

[0049] -Wheel support force Q i (calculated in the vertical force calculation module 20) and the adhesion coefficient μ i The product of

[0050] - Coupling force F between carriages Kupl , where, if the coupling forces are not measured, they can alternatively be neglected,

[0051] -Mass M and acceleration The product of .

[0052] The mass M consists of a static part known from engineering and a running part which can be measured via the spring force of the carriage.

[0053] In the direction of rotation, by considering equations (2) to (5) it can be derived that:

[0054] the resulting axle-dependent engine torque value m_e_i, which results from the electric motor drive torque and the braking torque, wherein the engine torque value m_e_i can be calculated, for example, from electrical variables of the engine,

[0055] - the pneumatic braking torque m_p_i generated by the friction braking torque calculation module 10,

[0056] - Based on wheel support force Q i (calculated in the vertical force calculation module 20), the adhesion coefficient μi and the wheel radius r i The torque,

[0057] -Moment of inertia I i and the product of the wheelset acceleration, where the moment of inertia of the axle I i It is known from engineering that the wheelset acceleration It can be derived from the axle speed.

[0058] The equations of motion are converted into a state space representation and fed to the Kalman filter as mathematical boundary conditions:

[0059]

[0060] State Vector With n=9 states, and the dimension of the process noise is Q (9x9) , where the diagonal entries of the matrix are occupied here.

[0061] Wheel speedω i is measured and returned as observations to the Kalman filter 30, where m=4 observations are available. The result is an observation matrix H (4x9) and the diagonal of the measurement noise occupies the covariance matrix R (4x4) .

[0062] The diagonal entries of the covariance matrix and the process noise matrix are preferably determined empirically.

[0063] According to the input The Kalman filter 30 estimates the current adhesion coefficient as a friction fit value and a speed v_k, hereinafter referred to as the Kalman speed.

[0064] The slip module 40 is based on the wheel speed ω i and wheel radius r_i to calculate the axle speed value v_umlauf according to:

[0065] During braking operation (identified by the sum of the friction braking torque value and the engine torque value), the fastest axle defines the axle speed value.

[0066] - In drive operation, the slowest axle defines the axle speed value.

[0067] Furthermore, the state "ist_gleiten" is determined with the aid of the returned speed or movement estimate v_virt. This state indicates whether the vehicle is in full axle slip. The following criteria are used for this purpose:

[0068] - variance of axle speed,

[0069] - first derivative of the axle speed in time,

[0070] - the second derivative of the axle speed in time, and

[0071] The slip of the axle speed relative to a reference speed, which may correspond, for example, to the last determined speed estimate v_virt.

[0072] In the evaluation module 50 , a new velocity estimate v_virt is formed from the Kalman velocity v_k and the axle velocity values, and its time derivative as a new acceleration estimate a_virt, preferably according to:

[0073] - In the absence of full axle slip (ist gleiten == 0), the axle speed value (v_umlauf) defines the speed estimate v_virt. -

[0074] If full axle slip is present (ist_gleiten == 1), the Kalman velocity v_k is used for the new speed estimate v_virt. To this end, the Kalman velocity v_k is debiased by "pulling it up" or "pulling it down" to the last axle velocity value (without full axle slip) and the new speed estimate v_virt is determined during the full axle slip.

[0075] By differentiating and smoothing the new velocity estimate v_virt, a new acceleration estimate a_virt is generated, which can be used for plausibility checking of the measured vehicle acceleration.

[0076] In summary, based on brake cylinder pressure, engine torque, dynamic wheel diameter and wheel speed, estimation device 1 can determine the unmeasurable adhesion coefficient and a new speed estimate v_virt and a new acceleration estimate a_virt, which can be used, for example, for plausibility verification of measured vehicle acceleration.

[0077] The post-controller can set the slip by the estimated adhesion coefficient and speed of each wheel group (i.e. the motion estimation value v_virt) to obtain the maximum possible adhesion coefficient, thereby minimizing the stopping distance or maximizing the traction.

[0078] As mentioned above, the acceleration estimate a_virt can be used for plausibility verification of the measured acceleration. Safety can be improved by this diverse redundancy, or the installation of additional hardware can be avoided, or the measured values ​​of the installed sensors can be given greater weight.

[0079] Furthermore, the new speed estimate v_virt can be used to adjust the time between the detection of full axle slip and the triggering of drift protection. Drift protection is preferably not executed after a fixed time, but after the difference between the "classic" reference speed (i.e. the axle speed limited by the steilheitsbegrenzt steepness) and the speed estimate v_virt exceeds a predetermined value. Impermissibly large slip values ​​and unnecessary brake releases of the axles can thus be prevented.

[0080] The estimation device 1 enables, for example, a greater weighting of the acceleration sensor, GPS or radar data for determining the reference speed, since this can be achieved through diversity ( The redundant) method is validated by existing data. In addition, a more accurate estimation of the adhesion coefficient is achieved, which leads to an optimization of the adhesion coefficient utilization in slip control in traction and braking situations.

[0081] Furthermore, drift protection interventions can be carried out in a manner adapted to the ambient conditions in order to improve the accuracy of the reference speed and shorten the stopping distance.

[0082] Figure 2 For example, according to Figure 1 The estimation device 1 is implemented in a computing facility 100 that forms or at least also forms the estimation device 1. The computing facility 100 and therefore the estimation device 1 comprises a computing device 110 and a memory 120, in which a Kalman filter software module M30, a sliding software module M40 and an evaluation software module M50 are stored, the Kalman filter software module M30 forming a sliding software module M40 according to the evaluation software module M50 when executed by the computing device. Figure 1 The Kalman filter 30, the sliding software module M40 forms a Figure 1 The slip module 40, the evaluation software module M50, when executed by the computing device, forms a Figure 1 An evaluation module 50 is provided.

[0083] In addition, the memory 120 stores a vertical force calculation software module M20 and a friction braking torque calculation software module M10. The vertical force calculation software module M20 generates a calculation result according to the calculation result when executed by the computing device. Figure 1 The vertical force calculation module 20 and the friction braking torque calculation software module M10 are formed according to Figure 1 The friction braking torque calculation module 10.

[0084] Figure 3 An embodiment for a rail vehicle 200 is shown, which is equipped with a Figure 2 The computing facility 100 is therefore equipped with Figure 1 The estimation device 1.

[0085] The rail vehicle 200 further comprises a slip controller 210 which controls the slip of the corresponding axle using the friction fit value μ_i per axle estimated by the estimation device 1 , the wheel peripheral speed value and the corresponding motion estimation value v_virt.

[0086] Furthermore, a plausibility check device 220 is connected to the estimation device 1 , which performs a plausibility check on the motion information from other sources using the motion estimate of the estimation device 1 .

[0087] Finally, it should be mentioned that the features of all the above-described embodiments can be combined with one another in any manner to form further other embodiments of the present invention.

[0088] Furthermore, all features of the dependent claims may be combined individually with each parallel claim and may be combined arbitrarily either individually or with one or other dependent claims to obtain further other embodiments.

Claims

1. An estimation device (1) for determining a new motion estimate (v_virt), the new motion estimate indicating the speed of a rail vehicle (200) having at least two axles, the estimation device comprising: a Kalman filter (30) which applies at least torque values ​​(m_p_i, m_e_i), vertical force values ​​(Q_i), wheel dimension values ​​(r_i) and wheel speed values ​​(ω_i) on the input side and determines a Kalman-estimated vehicle speed value (v_k) based on these and based on the Kalman filtering, a slip module (40) which applies the torque values ​​(m_p_i, m_e_i), the wheel dimension values ​​(r_i), the wheel speed values ​​(ω_i) and the corresponding previous motion estimate values ​​(v_virt) of the estimation device (1) on the input side and determines the wheel peripheral speed values ​​(v_umlauf) and slip information indicating a possible slip of the corresponding axle, and An evaluation module (50) which determines a new motion estimation value (v_virt) of the estimation device (1) based on the Kalman estimated vehicle speed value (v_k), the wheel peripheral speed value (v_umlauf) and the slip information.

2. The estimation device (1) according to claim 1, characterized in that Slip Module(40) - outputting the wheel peripheral speed value (v_umlauf) of the axle having the maximum wheel peripheral speed value to the evaluation module (50) in the case of a braking operation of the rail vehicle (200), - in the case of drive operation, outputting the wheel peripheral speed value (v_umlauf) of the axle having the smallest wheel peripheral speed value to the evaluation module (50), and - in the event of slippage on all axles, outputting total slip information (ist_gleiten) to the evaluation module (50).

3. The estimation device (1) according to claim 2, characterized in that Evaluation Modules (50) - outputting wheel peripheral speed values ​​(v_umlauf) as new motion estimates when the rail vehicle (200) is in braking or driving operation and all axles are free of slip, and - When all axles slip, a new motion estimate (v_virt) is determined based on the vehicle velocity value (v_k) estimated by the Kalman filter (30).

4. The estimation device (1) according to one of the preceding claims, characterized in that the estimation device (1) has a vertical force calculation module (20) which generates a vertical force value (Q_i) per axle based at least on the moment values ​​(m_p_i, m_e_i), the wheel size values ​​(r_i) and the corresponding previous motion estimation values ​​(v_virt) of the estimation device (1), which are also used by the Kalman filter (30), - wherein each vertical force value (Q_i) of the vertical force calculation module (20) forms one of a plurality of vertical force values ​​(Q_i) used by the Kalman filter (30).

5. The estimation device (1) according to one of the preceding claims, characterized in that the estimation device (1) has a friction braking torque calculation module (10) which generates a friction braking torque value (m_p_i) for each axle based at least on a brake pressure value (p_i) for each axle, a wheel size value (r_i) and a corresponding previous motion estimation value (v_virt) of the estimation device (1), -in, Each friction braking torque value (m_p_i) forms one of a plurality of torque values ​​(m_p_i, m_e_i) used by a Kalman filter (30).

6. The estimation device (1) according to one of the preceding claims, characterized in that The Kalman filter (30) also uses the engine torque values ​​(m_e_i) associated with the axles as torque values ​​(m_p_i, m_e_i), or at least also takes these values ​​into account.

7. The estimation device (1) according to one of the preceding claims, characterized in that The Kalman filter (30) is an extended Kalman filter (30).

8. The estimation device (1) according to one of the preceding claims, characterized in that The estimation device (1) outputs a vehicle acceleration value indicating the acceleration of the rail vehicle (200) as a further motion estimation value (a_virt).

9. The estimation device (1) according to one of the preceding claims, characterized in that - The Kalman filter (30) also determines the friction fit value (μ_i) per axle involved in the force-track contact, - And the estimation device (1) also outputs the friction fit value (μ_i).

10. The estimation device (1) according to one of the preceding claims, characterized in that The estimation device comprises a computing device (110) and a memory (120), wherein the memory stores - a Kalman filter software module (M30) which, when executed by a computing device, forms said Kalman filter (30), - a slip software module (M40) which, when executed by a computing device, forms said slip module (40), - and an evaluation software module (M50) which, when executed by a computing device, forms said evaluation module (50).

11. The estimation device (1) according to claim 10, characterized in that The memory (120) stores - a vertical force calculation software module (M20) which, when executed by a computing device, forms said vertical force calculation module (20), and / or - a friction braking torque calculation software module (M10) which, when executed by a computing device, forms the friction braking torque calculation module (10).

12. A rail vehicle (200), characterized in that: The rail vehicle is equipped with an estimation device (1) according to one of the preceding claims.

13. The rail vehicle according to claim 12, characterized in that The slip controller (210) is connected to the estimation device (1), and controls the slip of the corresponding axle by using the estimated friction fit value of each axle, the wheel peripheral speed value and the corresponding motion estimation value (v_virt) of the estimation device (1).

14. The rail vehicle according to claim 12 or 13, characterized in that The rationality verification device (220) is connected to the estimation device (1) and performs a rationality check on motion information of other sources through the motion estimation value of the estimation device (1).

15. A method for determining a new motion estimate (v_virt) indicating a velocity of a rail vehicle (200) having at least two axles, wherein in the method - at least the moment values ​​(m_p_i, m_e_i), the vertical force values ​​(Q_i), the wheel size values ​​(r_i) and the wheel speed values ​​(ω_i) are fed to a Kalman filter (30) on the input side and a Kalman-estimated vehicle speed value (v_k) is determined based on this and based on the Kalman filtering, - determining a wheel peripheral speed value (v_umlauf) and a slip information (ist_gleiten) which indicates a possible slip of the respective axle by means of a wheel dimension value (r_i), a wheel speed value (ω_i) and a previously determined motion estimate value (v_virt), respectively, - and determining a new motion estimate value (v_virt) based on the Kalman estimated vehicle speed value (v_k), the wheel peripheral speed value (v_umlauf) and the slip information.

Images