Method for estimating longitudinal acceleration of at least one rail vehicle

Abstract

A method for estimating a longitudinal acceleration of at least one rail vehicle by means of an acceleration sensor arrangement (100), comprising the following steps: - performing a calibration phase, the calibration phase comprising the following steps: solving the following system of equations to determine values for a first direction cosine k1, a second direction cosine k2 and a third direction cosine k3: - after the calibration phase, determining an estimated longitudinal acceleration value a lon (t i1 ) by means of the sum of: a product of the first direction cosine k1 and a first acceleration a x (t i1 ), a product of the second direction cosine k2 and a second acceleration a y (t i1 ), and a product of the third direction cosine k3 and a third acceleration a z (t i1 ).

Description

Technical Field

[0001] This invention generally belongs to the field of rail vehicles; in particular, this invention relates to a method for estimating the longitudinal acceleration of at least one rail vehicle. Background Technology

[0002] Slip (also known as skidding, drifting, or coasting) is understood as a difference between the rotational speed of the axle and the vehicle's travel speed. This difference is defined as the slip velocity.

[0003] The slip velocity can be calculated using the following formula:

[0004] V sliding =V RV -ω axle *R (1)

[0005] Among them, V RV It is the longitudinal speed of the rail vehicle, ω axle R is the angular velocity of the axle, and R is the radius of the wheel.

[0006] More modern rail vehicles are equipped with electronic systems that typically include subsystems for controlling wheel slippage, designed to intervene when the vehicle is in traction or braking phases. These subsystems are called anti-skid or skid-resistant systems, or WSP (Wheel Slip Protection) systems.

[0007] Figure 1 A system for controlling wheel adhesion in an anti-skid function, according to the prior art, is illustrated schematically. Figure 1 This involves a vehicle comprising n controlled axles A1, A2, ..., An. Axles A1, A2, ..., An include relative axles S1, S2, ..., Sn and relative wheel pairs W1, W2, ..., Wn that rotate integrally with them.

[0008] In the accompanying diagrams, only one wheel per axle is typically shown.

[0009] Figure 1 The WSP system includes an electronic control unit (ECU), typically based on a microprocessor architecture. This ECU receives rotational speed signals related to the angular velocity of each axle (A1, A2, ..., An) from sensors SS1, SS2, ..., SSn, each associated with its respective axle. The ECU is also connected to torque control units TC1, TC2, ..., TCn, each associated with its corresponding axle (A1, A2, ..., An).

[0010] The electronic control unit (ECU) is configured to adjust the torque applied to each axle according to a predetermined algorithm. If torque is applied during traction or braking when traction is reduced, one or more wheels on each axle may experience initial slippage. Adjusting the torque in this way prevents the axles from completely locking up, and in this way may allow each axle to enter a controlled slip condition to restore traction; however, traction is reduced throughout the duration of the condition.

[0011] It is obvious that knowing the vehicle's instantaneous speed V RV (t) is the basis for proper slip control.

[0012] A known method for accurately tracking the speed of a rail vehicle requires maintaining idle axles, i.e., axles not subjected to traction or braking torque. This necessitates ensuring that the measured speed is the actual speed V of the rail vehicle. real This provides optimal reproduction. This solution is particularly effective when the adhesion between the wheels and the rails is extremely low. In such cases, all wheels may slip under traction or braking, thus failing to provide accurate information about the vehicle's actual speed. The idle axle, not subjected to traction or braking torque, can continue to accurately track the vehicle's speed.

[0013] Modern rail vehicle architecture, especially in underground railways, often has a very limited configuration, such as consisting of two cars. In this case, using "idle" axles results in a significant decrease in the train's traction and braking capabilities.

[0014] Figure 2a An example combination containing two separate carriages is shown, while Figure 2b An example combination of two cars constrained by Jacobs bogies is shown. It is clear that in the first case, using an idle axle disadvantageously reduces traction and braking capacity by 12.5%, and in the second case by as much as 16.7%.

[0015] In the existing technology, there are also systems based on acceleration sensors for measuring the forward speed of vehicles.

[0016] Given the increasing availability and decreasing cost of MEMS (“microelectromechanical systems”), more and more electronic devices, regardless of their primary application, are integrating accelerometers, typically triaxial accelerometers, onto the board.

[0017] In principle, it is easy to apply the use of accelerometers to estimate the longitudinal acceleration of a vehicle.

[0018] Longitudinal acceleration is understood as the acceleration of a vehicle in its direction of travel. By integrating this longitudinal acceleration over time, the longitudinal velocity, i.e., the vehicle's speed, is obtained. This method is clearly unaffected by the axle slippage that may occur when traction decreases.

[0019] refer to Figure 3 The rail vehicle 1 travels on track 3 and is equipped with an acceleration sensor 2.

[0020] In the Earth / gravity reference frame, the z-axis can be defined as the direction of gravitational acceleration, and the x and y axes can be defined as the transverse directions on a plane perpendicular to z.

[0021] However, in a reference frame integrated with the vehicle, axis y' can be defined as the longitudinal direction of the vehicle, axis x' can be defined as the lateral direction of the vehicle, and axis z' can be defined as the direction perpendicular to the plane ("floor") of the vehicle.

[0022] In addition, x”, y”, and z” can be defined as the sensitive axes of a triaxial accelerometer.

[0023] Now consider the ideal scenario where the sensors are integrated with the vehicle, where these axes are perfectly aligned with the vehicle's axes. This would result in:

[0024] x'≡x”

[0025] y'≡y”

[0026] z'≡z”

[0027] Also considering the special case where vehicle 1 travels on a completely straight section of track 3 with no slope, this will result in:

[0028] x'≡x”≡x

[0029] y'≡y”≡y

[0030] z'≡z”≡z

[0031] Under these ideal conditions, the longitudinal acceleration of the vehicle can be directly derived from the accelerometer measurement, that is:

[0032] a train =a y (2)

[0033] Among them, a train It is the longitudinal acceleration of the vehicle, a y It is the acceleration measured by the accelerometer on the sensor's y" axis.

[0034] Then, the vehicle's speed can be calculated as the time integral of the acceleration value:

[0035]

[0036] Alternatively, in the case of a discrete acquisition system:

[0037]

[0038] Where ΔT is the sampling period of the electronic acquisition system, and n is the number of samples acquired at time t.

[0039] However, the above assumptions have greatly simplified this approach, but it is not actually applicable to real-world situations.

[0040] Although they can be precisely installed, the accelerometers integrated on the vehicle's electronic circuit board do not align their measuring axes x”, y”, and z” perfectly with the vehicle's axes x’, y’, and z’.

[0041] Furthermore, the assumption that vehicle 1 travels on a completely straight section of track 3 without any gradient is not applicable in practice, as rail vehicles may travel on curved sections and / or sections with non-zero gradients.

[0042] The invalidity of the above assumptions paves the way for geometric scenarios with three reference frames (gravity, vehicle, and accelerometer) at relative angles on three axes.

[0043] For axes x, y, and z, the rotation angles between the vehicle reference frame and the gravity reference frame can be defined as α, β, and , respectively.

[0044] For axes x, y, and z, the rotation angles between the accelerometer reference frame and the vehicle reference frame can be defined as α', β', and α', respectively.

[0045] Angles α, β The electronic units that acquire the accelerometer data are unknown, but they are constant over time because they depend only on the accelerometer's installation relative to the vehicle.

[0046] Angles α', β', The electronic units that acquire the accelerometer data are also unknown and not constant over time, as they depend on the curvature and gradient of local portions of the track.

[0047] Due to angles α, β, and angles α', β', Since the values ​​are unknown and unrelated to each other, the problem of determining the longitudinal acceleration of a vehicle based on measurements from a triaxial accelerometer installed on the vehicle may not be solvable analytically and / or geometrically.

[0048] For example, WO2017042138 proposes an analytical / geometric solution method for determining the orientation of an accelerometer relative to a vehicle on which the sensor is mounted. This method is based on the availability of the following:

[0049] - "Free state" refers to the state in which the rail vehicle is stationary or moving at a constant speed. In this state, the accelerometer is only affected by gravity.

[0050] - Availability of reliable and external rail vehicle speed sources, even discontinuous availability.

[0051] If we assume that the angular velocity of the axle (is determined by...) Figure 1 Since the accelerometer (SS measurement) is reliable in all non-degrading adhesion conditions (which is still the majority), it seems reasonable that reliable measurement of rail vehicle speed, even discontinuously, is also possible. The role of the accelerometer and the method for estimating rail vehicle speed based on it will be useful in situations where the axle slips completely and its angular velocity does not represent the speed of the rail vehicle or train, indicating a decrease in adhesion.

[0052] However, the method proposed in WO2017042138 is basically based on two assumptions:

[0053] 1) During the “free state”, the vehicle is in the completely horizontal part of the track (gradient is zero);

[0054] 2) During the phase where the external velocity of the accelerometer is used, the rail vehicle or train is in a completely straight section of the track. The lateral acceleration component (x-axis) is therefore ignored and appears in the curved sections.

[0055] The assumptions upon which WO2017042138 is based have significant limitations, and the effectiveness of the method in practical applications cannot be guaranteed.

[0056] Under braking conditions, this is used to estimate the actual vehicle speed V. V The most common algorithm for (t) typically uses the following function:

[0057] V v (T j )=max[S1(T j ),..,S n (T j ),(V v (T j-1 )+a max ·T)] (5)

[0058] In the case of traction, it uses the following function:

[0059] V v (Tj )=min[S1(T j ),..,S n (T j ),(V v (T j-1 )+a max ·T) (6)

[0060] Where a max This represents the maximum permissible acceleration of the vehicle during operation, which is positive under traction conditions and negative under braking conditions. The contribution (V) in equations (5) and (6) is... v (T j-1 )+a max ·T) is used for the speed V calculated using formulas (5) and (6) when excessive instantaneous and simultaneous changes in axle speed due to a significant decrease in adhesion during traction or braking may occur. V (t) When there is a significant loss, V V The change in (t) is suppressed within the physical limits allowed by the train.

[0061] More accurate variations of formulas (5) and (6) are known, but they are still based on instantaneous measurements of the individual speeds of the axles. It is clear here that the availability of idle axles would make formulas (5) and (6) very accurate if all axles bearing torque were in the slipping phase. Summary of the Invention

[0062] Therefore, the object of the present invention is to provide a method for estimating the longitudinal acceleration of at least one rail vehicle and a method for estimating the longitudinal velocity of at least one rail vehicle, which allow for the measurement of the longitudinal acceleration and longitudinal velocity of the vehicle, respectively, even when all axles of the vehicle are in a slipping phase due to reduced adhesion.

[0063] Therefore, another object of the present invention is to allow the use of idle axles to be fully restored for traction and braking purposes, even under particularly low adhesion conditions, thereby increasing the traction and braking capabilities of the train, while allowing the axles to accurately track the train's speed in order to accurately estimate the longitudinal forward speed.

[0064] The methods for estimating longitudinal velocity and longitudinal acceleration can be applied to slippage during the traction phase (negative V). slippage ) and slippage during braking (positive V) slippage Both.

[0065] This invention allows for the most accurate understanding of the longitudinal speed of a rail vehicle, thereby facilitating and improving, for example, the piloting of onboard control systems, anti-skid systems, and odometer references.

[0066] According to one aspect of the invention, the above and other objects and advantages are achieved by methods for estimating the longitudinal acceleration of at least one rail vehicle having the features defined in claims 1 and 2, respectively, and by methods for estimating the longitudinal acceleration of at least one rail vehicle having the features defined in claims 10 and 12, respectively. Preferred embodiments of the invention are defined in the dependent claims, the contents of which should be understood as an integral part of this specification. Attached Figure Description

[0067] The functional and structural features of some preferred embodiments of the method for estimating the longitudinal velocity of at least one rail vehicle according to the present invention will now be described. Referring to the accompanying drawings, wherein:

[0068] Figure 1 The WSP system produced according to existing technology is shown;

[0069] Figure 2a An example combination consisting of two separate carriages is shown, while Figure 2b An example combination of two cars constrained by Jacobs bogies is shown;

[0070] Figure 3 The image shows a rail vehicle traveling on a track and equipped with an acceleration sensor.

[0071] Figure 4 An example system that can be used to implement embodiments of the present invention is shown; and

[0072] Figure 5 The diagram shows the trend of rail vehicle speed over time, the availability of an independent longitudinal reference speed, and several illustrative graphs of the direction cosine. Detailed Implementation

[0073] Before describing several embodiments of the invention in detail, it should be understood that the application of the invention is not limited to the construction details and configurations of the components presented in the following description or shown in the drawings. The invention can be implemented or constructed in practice in other ways. It should also be understood that the wording and terminology are descriptive for purposes of description and should not be construed as limiting. The use of “comprising” and “including” and variations thereof should be understood to include the elements listed below and their equivalents, as well as additional elements and their equivalents.

[0074] This invention proposes a method for calculating the estimated longitudinal acceleration of at least one rail vehicle, and then proposes a method for calculating the estimated longitudinal velocity of the rail vehicle using an acceleration sensor device, such as a triaxial acceleration sensor device.

[0075] A method for estimating the longitudinal acceleration of at least one rail vehicle requires the availability, or even discontinuous availability, of train speed measurements independent of accelerometers, hereinafter referred to as the independent longitudinal reference speed V of the rail vehicle. ref The method for estimating the longitudinal acceleration of at least one rail vehicle is based on obtaining the direction cosine of a 3x3 orientation matrix. Multiplying this cosine by measurements from sensors with three orthogonal axes yields the estimated longitudinal acceleration of the vehicle. Integrating the estimated longitudinal acceleration of the vehicle, the estimated longitudinal velocity can be obtained.

[0076] The main equations used are as follows:

[0077] k1*a x (t)+k2*a y (t)+k3*a z (t)=a ref (t) (7)

[0078] in:

[0079] -k1, k2, and k3 are the numerical values ​​of the direction cosines, and k1, k2, and k3 are the variables of the equation;

[0080] -a x (t), a y (t), a z (t) is the acceleration detected by the accelerometer device on its three orthogonal axes at an instant t;

[0081] -a ref (t) is the independent reference acceleration at time t, derived from the independent reference longitudinal velocity (i.e., the travel speed) v. ref The derivative at time t is determined, that is:

[0082]

[0083] In order to calculate the derivative of the independent longitudinal reference velocity at time t, the independent longitudinal reference velocity must be measured within at least one time interval including time t. For example, this interval could be 20 ms, 100 ms, etc.

[0084] The unknowns in this equation are the values ​​of k1, k2, and k3.

[0085] Equation (7) applies only to the aforementioned independent longitudinal reference velocity v ref Effective where available.

[0086] Regardless of the technology used, at an independent longitudinal reference velocity v refThe available phase solution equation (7) means that the coefficients k1, k2 and k3 can be dynamically updated so that the estimated longitudinal acceleration of the rail vehicle or train can be obtained by a linear combination of them with the acceleration measured by the sensor.

[0087] The following describes a first embodiment of a method for estimating the longitudinal acceleration of at least one rail vehicle using an acceleration sensor device 100. The acceleration sensor device 100 is arranged to measure a first acceleration a along a first axis x. x The second acceleration a along the second axis y y and the third acceleration a along the third axis z z The first axis x, the second axis y, and the third axis z are orthogonal to each other.

[0088] A method for estimating the longitudinal acceleration of at least one rail vehicle includes a first calibration phase, which comprises the following steps:

[0089] -In the independent longitudinal reference speed v of the rail vehicle ref Available first calibration time t c1 The first value a of the first acceleration was measured. x (t c1 ), the first value of the second acceleration a y (t c1 ) and the first value of the third acceleration a z (t c1 ), where the independent longitudinal reference velocity v ref It is unrelated to the acceleration sensor device (100);

[0090] -The independent longitudinal reference speed v of the rail vehicle ref Available second calibration time t c2 The second value a of the first acceleration was measured. x (t c2 ), the second value of the second acceleration a y (t c2 ) and the second value of the third acceleration a z (t c2 The second calibration time t c2 Unlike the first calibration time t c1 ;

[0091] -The independent longitudinal reference speed v of the rail vehicle ref Available third calibration time t c3 The third value a of the first acceleration was measured. x (t c3 The third value of the second acceleration, a y (t c3 ) and the third value of the third acceleration az (t c3 The third calibration time t c3 Unlike the first calibration time t c1 and the second calibration time t c2 ;

[0092] -Based on the first calibration time t c1 The first value of the measured independent longitudinal reference velocity v ref (t c1 ) Calculate the first calibration time t c1 The value of the first independent longitudinal reference acceleration a ref (t c1 );

[0093] -Based on the second calibration time t c2 The second value of the measured independent longitudinal reference velocity v ref (t c2 ) Calculate at the second calibration time t c2 The value of the second independent longitudinal reference acceleration a ref (t c2 );

[0094] -Based on the third calibration time t c3 The third value of the measured independent longitudinal reference velocity v ref (t c3 ) Calculate at the third calibration time t c3 The value of the third independent longitudinal reference acceleration a ref (t c3 );

[0095] Solve the following system of equations to determine the values ​​of the first direction cosine k1, the second direction cosine k2, and the third direction cosine k3:

[0096]

[0097] Clearly, the calibration times can all be obtained within a single, continuous calibration interval where the independent longitudinal reference speed of the rail vehicle is available, or the calibration times can be obtained within several calibration intervals where the independent longitudinal reference speed of the rail vehicle is available. In the second case, the various calibration intervals can be separated by intervals where the independent longitudinal reference speed of the rail vehicle is not available.

[0098] In other words, if the independent longitudinal reference velocity v ref If available, then at three different calibration times t c1 t c2 t c3 Measure a x a y a z and vref The values ​​of k1, k2, and k3 are thus determined, resulting in a system of three equations. This system of three equations with three unknowns can be solved using both analytical and numerical methods. Furthermore, by obtaining measurements from more distant times (i.e., more than three times), the accuracy of the solutions k1, k2, and k3 can be recursively increased and updated over time.

[0099] In this case, the independent longitudinal reference velocity v ref The statement that it is unrelated to the acceleration sensor device 100 is understood to mean that the independent longitudinal reference velocity is not obtained through the acceleration sensor device.

[0100] In this case, the independent longitudinal reference velocity v ref The available statements can be understood to imply an independent longitudinal reference velocity v ref This can be used in a method for estimating the longitudinal acceleration of at least one rail vehicle (which is the subject of this invention) because it is available and reflects the true longitudinal velocity of the rail vehicle moving along the track.

[0101] The method for estimating the longitudinal acceleration of at least one rail vehicle also includes another measurement phase following the calibration phase.

[0102] This measurement phase includes the following steps:

[0103] -At least for the first measurement time t when the independent longitudinal reference velocity of the rail vehicle is unavailable. i1 Determine the estimated longitudinal acceleration value at for at least one rail vehicle. lon (t i1 ).

[0104] Estimate the longitudinal acceleration value a lon (t i1 ) and the measurement time t i1 The relevant factors are estimated by summing the following:

[0105] - The first direction cosine k1 determined during the calibration phase and the first measurement time t i1 The fourth value a of the first acceleration obtained x (t i1 The product of )

[0106] - The second direction cosine k2 determined during the calibration phase and the first measurement time t i1 The fourth value a of the second acceleration obtained y (t i1 The product of )

[0107] -The third-direction cosine k3 determined during the calibration phase and at the first measurement time t i1The fourth value a of the third acceleration obtained z (t i1 The product of ).

[0108] In an alternative embodiment, reference Figure 4 The method for estimating the longitudinal velocity of at least one rail vehicle again includes a calibration phase, which comprises the following steps:

[0109] -At least for the independent longitudinal reference speed v of the rail vehicle ref Available first calibration time t c1 The estimated longitudinal acceleration value 'a' of at least one rail vehicle is estimated using a linear filter. lon (t c1 In this case, the independent longitudinal reference velocity v ref It is also unrelated to the acceleration sensor device 100.

[0110] Estimate the longitudinal acceleration value a lon (t c1 ) and the calibration time t c1 The correlation is estimated by a linear filter through the sum of the following terms:

[0111] - A first direction cosine k1 with a predetermined value and at the first calibration time t c1 The first value a of the first acceleration was obtained. x (t c1 The product of )

[0112] - A second direction cosine k2 with a predetermined value and at the first calibration time t c1 The first value a of the second acceleration was obtained. y (t c1 The product of )

[0113] - A third-direction cosine k3 with a predetermined value and at the first calibration time t c1 The first value a of the third acceleration obtained z (t c1 The product of ).

[0114] In this embodiment, the calibration phase further includes the following steps:

[0115] -by relative to the first calibration time t c1 The estimated longitudinal acceleration value a lon (t c1 ) and relative to the first calibration time t c1 And according to the first calibration time t c1 The first value of the measured independent longitudinal reference velocity v ref (t c1The determined independent longitudinal reference acceleration value a ref (t c1 The difference between the two is used to determine the estimation error.

[0116] - The corresponding update values ​​of the first direction cosine k1, the second direction cosine k2, and the third direction cosine k3 to be applied to the linear filter are determined by an adaptive filter (104) to minimize the estimation error.

[0117] When the calibration phase begins for the first time, the values ​​of the first direction cosine k1, the second direction cosine k2, and the third direction cosine k3 can be predetermined. For example, they can be predetermined default values ​​that are gradually adjusted at various calibration times as calibration progresses, or they can be predetermined values ​​equal to the values ​​of the first direction cosine k1, the second direction cosine k2, and the third direction cosine k3 calibrated during previous runs of the rail vehicle.

[0118] In other words, the equations can be solved using an adaptive algorithm. This error can be corrected by comparing the output of the linear filter with the independent longitudinal reference velocity v through differentiation. ref The error is obtained by comparing the reference longitudinal acceleration value. The adaptive filter uses this error to dynamically recalculate the coefficients k1, k2, k3 of the linear filter in order to recursively minimize the error.

[0119] In this alternative embodiment, the method for estimating the longitudinal speed of at least one rail vehicle also includes a further measurement phase following the calibration phase.

[0120] This measurement phase includes the following steps:

[0121] -At least for the first measurement time t when the independent longitudinal reference velocity of the rail vehicle is unavailable. i1 The linear filter is used to determine the estimated longitudinal acceleration value a of at least one rail vehicle. lon (t i1 ).

[0122] Estimate the longitudinal acceleration value a lon (t i1 ) and the measurement time t i1 The relevant factors are estimated by summing the following:

[0123] - The updated value of the first direction cosine k1 determined during the calibration phase and the value at the first measurement time t i1 The value of the first acceleration a was obtained. x (t i1 The product of )

[0124] - The updated value of the second direction cosine k2 determined during the calibration phase is compared with the value at the first measurement time t.i1 The value of the second acceleration a was obtained. y (t i1 The product of )

[0125] - The updated value of the third-direction cosine k3 determined during the calibration phase and the value at the first measurement time t i1 The value of the third acceleration a was obtained. z (t i1 The product of ).

[0126] refer to Figure 5 Available time t av It can be defined as an independent longitudinal reference velocity v ref The last available moment. Available moment t av The subsequent moment will be an independent longitudinal reference velocity v ref The moment when it will no longer be available.

[0127] In t av Independent longitudinal reference velocity v detected at all times ref (t av ) is the independent longitudinal reference velocity v ref The last reliable value that is available.

[0128] Recovery of available time t ret_av (where t) ret_av >t av It can also be defined as an independent longitudinal reference velocity v. ref The moment it became available again.

[0129] Measurement time t i1 Recovery t with available time ret_av Consistent.

[0130] Equation k1*a x (t)+k2*a y (t)+k3*a z (t)=a ref () applies only to independent longitudinal reference velocity v ref Available time period, i.e., time period (t) <t av ) and (t>t ret_av ).

[0131] By solving this equation, the coefficients k1, k2, and k3 will be dynamically updated until the available time t. av In t av With t ret_av During the time interval between these intervals, the values ​​of k1, k2, and k3 will be frozen at their available time t. av The last updated value k1(t) av k2(t) av), k3(t av Then, when t>t av In the case of k1, k2, and k3, the values ​​can be dynamically updated by solving equation (7).

[0132] For example, the first measurement time t i1 It can be compared with the independent longitudinal reference velocity v ref (t ret_av The timing of becoming available again after being unavailable is consistent.

[0133] The adaptive filter can be arranged to determine the corresponding update values ​​of the first direction cosine k1, the second direction cosine k2, and the third direction cosine k3 by an adaptive algorithm based on the least mean square technique (LMS).

[0134] For all the above embodiments, the calibration phase can be repeated for multiple calibration times, such as the second calibration time t. i2 Third calibration time t i3 ..., nth calibration time t in .

[0135] Clearly, the calibration procedure can be performed on the first ignition of at least one rail vehicle.

[0136] The independent longitudinal reference speed v of the rail vehicle will be given below. ref Some examples.

[0137] For example, the independent longitudinal reference velocity v ref This could be the longitudinal velocity obtained from the angular velocity of the rail vehicle's axles. In this case, when the axles are not slipping, the independent longitudinal reference velocity v ref Available.

[0138] In another example, the independent longitudinal reference velocity v of the rail vehicle ref The longitudinal speed of the rail vehicle is provided by the positioning device. In this case, when the positioning device communicates with a satellite, the independent longitudinal reference speed v ref Available. The positioning device can be a GPS system / device that uses appropriate signals to communicate to obtain positioning information and thus the train's speed. For example, when a rail vehicle is in a tunnel and cannot communicate with the satellite, an independent longitudinal reference speed v is used. ref It may not be available.

[0139] In another aspect of the invention, the independent longitudinal reference speed v of the rail vehicle can be considered. ref Multiple available measurement times t i1 ,t i2 ,...,t in (e.g., the second measurement time t)i2 Third measurement time t i3 ..., nth measurement time t in Determine the estimated longitudinal acceleration a lon It can be based on the acquisition period ΔT i To select multiple measurement times t i1 ,t i2 ,...,t in .

[0140] Alternatively, estimate the longitudinal acceleration a. lon From available time t av and measurement time (e.g., the first measurement time t) i1 Or continuous measurement time t i2 ,...,t in The time t is continuously determined. av Compared with the independent longitudinal reference speed v directly located on the rail vehicle ref The time before the unavailable time is no longer available is consistent.

[0141] The present invention also relates to a method for estimating the longitudinal velocity of at least one rail vehicle.

[0142] When considering the independent longitudinal reference speed v of the rail vehicle ref Multiple available measurement times t i1 ,t i2 ,...,t in Determine the estimated longitudinal acceleration a lon At that time, those measurement times are based on the acquisition period ΔT i If selected, then the method for estimating the longitudinal velocity of at least one rail vehicle includes the available time t av The value of v of the independent longitudinal reference velocity of at least one axle of a rail vehicle is measured. ref (t av The steps are as follows.

[0143] A method for estimating the longitudinal velocity of at least one rail vehicle further includes determining the longitudinal velocity v of the rail vehicle according to the following steps. RV (t in The steps are as follows:

[0144] -Calculate at multiple measurement times t i1 ,t i2 ,…,t in The sum of estimated longitudinal accelerations;

[0145] - will be measured at multiple times t i1 ,t i2 ,…,t inThe sum of the determined longitudinal accelerations multiplied by the acquisition period ΔT i ;

[0146] - to place the at least one rail vehicle at available time t av The value of the independent longitudinal reference velocity V ref (t av ) and at multiple measurement times t i1 ,t i2 ,…,t in The sum of the estimated longitudinal accelerations and the acquisition period ΔT i Add the results of multiplication together.

[0147] For example, determining the longitudinal velocity v of a rail vehicle. RV The steps can be performed using the following formula:

[0148]

[0149] in:

[0150] -v ref (t av ) is the at least one rail vehicle at available time t av Independent longitudinal reference velocity V ref (t av );

[0151] - At multiple measurement times t i1 ,t i2 ,…,t in The sum of the estimated longitudinal accelerations, and n at the available time t. av With determining the longitudinal velocity v RV (t in The measurement time (t) in The number of acquisition cycles between ( ) and ( )

[0152] -ΔT i It refers to the data collection cycle.

[0153] Conversely, when estimating longitudinal acceleration a lon From the independent longitudinal reference velocity v directly located ref The available time t is consistent with the unavailable time before the unavailable time. av And, for example, the first measurement time t i1 When continuously determined, a method for estimating the longitudinal velocity of at least one rail vehicle includes the following steps: at available time t av The value v of the independent longitudinal reference velocity of the at least one rail vehicle is measured. ref (t av ).

[0154] A method for estimating the longitudinal velocity of at least one rail vehicle further includes determining the longitudinal velocity v of the rail vehicle according to the following steps. RV The steps of (t):

[0155] -Calculate from available time t av up to the first measurement time t i1 Continuously determined estimated longitudinal acceleration a lon The integral;

[0156] - to place the at least one rail vehicle at available time t av The value of the independent longitudinal reference velocity V ref (t av Add the integral result of the estimated longitudinal acceleration to the sum.

[0157] For example, determining the longitudinal velocity v of a rail vehicle. RV (t i1 The steps can be performed using the following formula:

[0158]

[0159] in:

[0160] -v ref (t av ) is the available time t av Independent longitudinal reference velocity V ref (t av );

[0161] - From available time t av To determine the longitudinal velocity v RV (t i1 The first measurement time t) i1 Continuously determined estimated longitudinal acceleration a lon The points.

[0162] The above also applies to subsequent measurement times. For example, considering the nth measurement time t... in Determine the longitudinal velocity v of the rail vehicle. RV (t in The steps can be performed using the following formula:

[0163]

[0164] in:

[0165] -v ref (t av ) is at the available time t av Independent longitudinal reference velocity V ref (tav );

[0166] - From available time t av To determine the longitudinal velocity v RV (t in The nth measurement time t) in Continuously determined estimated longitudinal acceleration a lon The points.

[0167] Refer again Figure 5 Therefore, we can conclude that in t av With t ret_av During the time interval between these intervals, at the independent longitudinal reference velocity v ref If unavailable, the acceleration of the rail vehicle will be calculated based on the accelerometer readings.

[0168] In t av With t ret_av During the time interval between these intervals, at the independent longitudinal reference velocity v ref In the event that the accelerometer is unavailable, the speed of the rail vehicle will be measured using the accelerometer readings and the last reliable independent longitudinal reference speed value v. ref Calculate based on the equation shown above.

[0169] Advantageously, due to the longitudinal speed v of the rail vehicle already described above in this specification... RV It will always be available, even at an independent longitudinal reference speed v. ref The same applies during unusable phases. Therefore, even under conditions of particularly low adhesion, the traction and braking uses of the idle axles can be fully restored, thereby improving the train's traction and braking capabilities.

[0170] Various aspects and embodiments of methods for estimating the longitudinal acceleration and longitudinal velocity of at least one rail vehicle according to the present invention have been described. It is understood that each embodiment can be combined with any other embodiment. Furthermore, the invention is not limited to the described embodiments, but can be varied within the scope defined by the appended claims.

Claims

1. A method for estimating the longitudinal acceleration of at least one rail vehicle by means of an acceleration sensor device (100), said acceleration sensor device (100) being arranged to measure a first acceleration a along a first axis x. x The second acceleration a along the second axis y y and the third acceleration a along the third axis z z ,in, The first axis x, the second axis y, and the third axis z are orthogonal to each other; The method for estimating the longitudinal acceleration of at least one rail vehicle includes the following steps: - Perform a calibration phase, which includes the following steps: • The independent longitudinal reference speed v of the rail vehicle ref Available first calibration time t c1 The first value a of the first acceleration is measured. x (t c1 ), the first value a of the second acceleration y (t c1 ) and the first value a of the third acceleration z (t c1 The independent longitudinal reference velocity v ref It is unrelated to the acceleration sensor device (100); • The independent longitudinal reference speed v of the rail vehicle ref Available second calibration time t c2 The second value a of the first acceleration is measured. x (t c2 ), the second value of the second acceleration a y (t c2 ) and the second value a of the third acceleration z (t c2 ), the second calibration time t c2 Unlike the first calibration time t c1 ; • The independent longitudinal reference speed v of the rail vehicle ref Available third calibration time t c3 The third value a of the first acceleration is measured. x (t c3 The third value of the second acceleration, a y (t c3 ) and the third value a of the third acceleration. z (t c3 The third calibration time t c3 Unlike the first calibration time t c1 and the second calibration time t c2 ; • Based on the first calibration time t c1 The first value v of the measured independent longitudinal reference velocity ref (t c1 ) Calculate at the first calibration time t c1 The value of the first independent longitudinal reference acceleration a ref (t c1 ); • Based on the second calibration time t c2 The second value of the effective independent longitudinal reference velocity v measured ref (t c2 ) Calculate at the second calibration time t c2 The value of the second independent longitudinal reference acceleration a ref (t c2 ); •According to the third calibration time t c3 The third value v of the measured independent longitudinal reference velocity ref (t c3 ) Calculate at the third calibration time t c3 The value of the third independent longitudinal reference acceleration a ref (t c3 ); • Solve the following system of equations to determine the values ​​of the first direction cosine k1, the second direction cosine k2, and the third direction cosine k3: - After the calibration phase, at least for the first measurement time t when the independent longitudinal reference velocity of the rail vehicle is unavailable. i1 Determine the estimated longitudinal acceleration value a of the at least one rail vehicle. lon (t i1 ), wherein the estimated longitudinal acceleration value a lon (t i1 ) and the first measurement time t i1 The relevant information is estimated by summing the following items: • The first direction cosine k1 determined during the calibration phase and the first measurement time t i1 The fourth value a of the first acceleration obtained x (t i1 The product of ) • The second direction cosine k2 determined during the calibration phase and at the first measurement time t i1 The fourth value a of the second acceleration obtained y (t i1 The product of ) • The third-direction cosine k3 determined during the calibration phase and at the first measurement time t i1 The fourth value a of the third acceleration obtained z (t i1 The product of ).

2. The method for estimating the longitudinal acceleration of at least one rail vehicle according to claim 1, wherein, The calibration phase is repeated for multiple calibration times.

3. The method for estimating the longitudinal acceleration of at least one rail vehicle according to claim 1 or 2, wherein, The calibration phase is performed during each first ignition of at least one rail vehicle.

4. The method for estimating the longitudinal acceleration of at least one rail vehicle according to claim 1 or 2, wherein, The independent longitudinal reference speed v of the rail vehicle ref It is the longitudinal velocity determined based on the angular velocity of the axle of the rail vehicle; When the axle does not slip, the independent longitudinal reference speed v ref Available.

5. The method for estimating the longitudinal acceleration of at least one rail vehicle according to claim 1 or 2, wherein, The independent longitudinal reference speed v of the rail vehicle ref The longitudinal speed of the rail vehicle is provided by the positioning device; When the positioning device communicates with the satellite, the independent longitudinal reference velocity v ref Available.

6. The method for estimating the longitudinal acceleration of at least one rail vehicle according to claim 1 or 2, wherein, The independent longitudinal reference velocity v of the rail vehicle ref Multiple available measurement times (t) i1 , t i2 , …, t in Determine the estimated longitudinal acceleration a lon The multiple measurement times (t) i1 , t i2 , …, t in According to the collection cycle T i Choose one.

7. The method for estimating the longitudinal acceleration of at least one rail vehicle according to claim 1 or 2, wherein, The estimated longitudinal acceleration a lon The independent longitudinal reference speed v directly located on the rail vehicle was obtained from the speed of the rail vehicle. ref The available time t is consistent with the time before the unavailable time that is no longer available. av and the first measurement time t i1 And continuously determined.

8. A method for estimating the longitudinal acceleration of at least one rail vehicle by means of an acceleration sensor device (100), said acceleration sensor device (100) being arranged to measure a first acceleration a along a first axis x. x The second acceleration a along the second axis y y and the third acceleration a along the third axis z z ,in, The first axis x, the second axis y, and the third axis z are orthogonal to each other; A method for estimating the longitudinal acceleration of at least one rail vehicle includes the following steps: - Perform a calibration phase, which includes the following steps: • At least for the independent longitudinal reference speed v of the rail vehicle ref Available first calibration time t c1 The estimated longitudinal acceleration value 'a' of the at least one rail vehicle is estimated using a linear filter. lon (t c1 ), wherein the independent longitudinal reference velocity v ref It is independent of the acceleration sensor device (100); wherein the estimated longitudinal acceleration value a lon (t c1 ) and the first calibration time t c1 The correlation is estimated by the linear filter through the sum of the following terms: o has a predetermined value of first direction cosine k1 and at the first calibration time t c1 The first value a of the first acceleration obtained x (t c1 The product of ) o has a predetermined value of second direction cosine k2 and at the first calibration time t c1 The first value a of the second acceleration was obtained. y (t c1 The product of ) o has a predetermined value for the third-direction cosine k3 and at the first calibration time t c1 The first value a of the third acceleration obtained z (t c1 The product of ) • By relative to the first calibration time t c1 The estimated longitudinal acceleration value a lon (t c1 ) and relative to the first calibration time t c1 And based on the first calibration time t c1 The first value of the measured independent reference longitudinal velocity v ref (t c1 The determined reference longitudinal acceleration value a ref (t c1 The difference between the two is used to determine the estimation error. • The corresponding update values ​​of the first direction cosine k1, the second direction cosine k2, and the third direction cosine k3 to be applied to the linear filter are determined by an adaptive filter (104) to minimize the estimation error; - After the calibration phase, at least for the first measurement time t when the independent longitudinal reference velocity of the rail vehicle is unavailable. i1 The estimated longitudinal acceleration value 'a' of the at least one rail vehicle is determined using the linear filter. lon (t i1 ), wherein the estimated longitudinal acceleration value a lon (t i1 ) and the first measurement time t i1 The relevant information is estimated by summing the following items: • The updated value of the first direction cosine k1 determined during the calibration phase and the value at the first measurement time t i1 The second value a of the first acceleration obtained x (t i1 The product of ) •- The updated value of the second direction cosine k2 determined during the calibration phase and at the first measurement time t i1 The second value a of the second acceleration obtained y (t i1 The product of ) •- The updated value of the third-direction cosine k3 determined during the calibration phase and the second value a of the third acceleration acquired at the first measurement time ti1. z (t i1 The product of ).

9. The method for estimating the longitudinal acceleration of at least one rail vehicle according to claim 8, wherein, The adaptive filter is configured to determine the corresponding update values ​​of the first direction cosine k1, the second direction cosine k2, and the third direction cosine k3 using an adaptive algorithm based on the least mean square (LMS) technique.

10. The method for estimating the longitudinal acceleration of at least one rail vehicle according to claim 8 or 9, wherein, The calibration phase is repeated for multiple calibration times.

11. The method for estimating the longitudinal acceleration of at least one rail vehicle according to claim 8 or 9, wherein, The calibration phase is performed during each first ignition of at least one rail vehicle.

12. The method for estimating the longitudinal acceleration of at least one rail vehicle according to claim 8 or 9, wherein, The independent longitudinal reference speed v of the rail vehicle ref It is the longitudinal velocity determined based on the angular velocity of the axle of the rail vehicle; When the axle does not slip, the independent longitudinal reference speed v ref Available.

13. The method for estimating the longitudinal acceleration of at least one rail vehicle according to claim 8 or 9, wherein, The independent longitudinal reference speed v of the rail vehicle ref The longitudinal speed of the rail vehicle is provided by the positioning device; When the positioning device communicates with the satellite, the independent longitudinal reference velocity v ref Available.

14. The method for estimating the longitudinal acceleration of at least one rail vehicle according to claim 8 or 9, wherein, The independent longitudinal reference velocity v of the rail vehicle ref Multiple available measurement times (t) i1 , t i2 , …,t in Determine the estimated longitudinal acceleration a lon The multiple measurement times (t) i1 , t i2 , …, t in According to the collection cycle T i Choose one.

15. The method for estimating the longitudinal acceleration of at least one rail vehicle according to claim 8 or 9, wherein, The estimated longitudinal acceleration a lon The independent longitudinal reference speed v directly located on the rail vehicle was obtained from the speed of the rail vehicle. ref The available time t is consistent with the time before the unavailable time that is no longer available. av and the first measurement time t i1 And continuously determined.

16. A method for estimating the longitudinal velocity of at least one rail vehicle, comprising the following steps: - Perform the method for estimating the longitudinal acceleration of at least one rail vehicle as described in claim 14; - At available time t av The value v of the independent longitudinal reference velocity of the at least one rail vehicle is measured. ref (t av The available time t av The independent longitudinal reference speed v directly located on the rail vehicle ref The time before the unavailable moment is no longer available; - Determine the longitudinal speed v of the rail vehicle according to the following steps. RV : • Calculate at multiple measurement times (t) i1 , t i2 , …, t in The sum of the estimated longitudinal accelerations; • At the multiple measurement times (t) i1 , t i2 , …, t in The sum of the longitudinal accelerations determined by the acquisition period is multiplied by the acquisition period. T i ; • The at least one rail vehicle will be at the available time t av The value of the independent longitudinal reference velocity V ref (t av ) and at the multiple measurement times (t) i1 , t i2 , …, t in The sum of the estimated longitudinal accelerations determined by the acquisition period T i Add the results of multiplication together.

17. The method for estimating the longitudinal velocity of at least one rail vehicle according to claim 16, wherein, The determination of the longitudinal speed v of the rail vehicle RV (t in The steps are performed using the following formula: in: - The at least one rail vehicle at the available time t av The independent longitudinal reference velocity V ref (t av ); - It is at the multiple measurement times t i1 , t i2 , …, t in The sum of the determined longitudinal accelerations, and n It is at the available time t av With determining the longitudinal velocity v RV (t in The measurement time (t) in The number of acquisition cycles between ( ) and ( ) - This refers to the acquisition period.

18. The method for estimating the longitudinal velocity of at least one rail vehicle according to claim 16 or 17, wherein, The first measurement time t i1 Recovery t with available time ret_av Consistent, the recovery t at the available time ret_av The independent longitudinal reference velocity V ref It becomes available again after being unavailable.

19. A method for estimating the longitudinal velocity of at least one rail vehicle, comprising the following steps: - Perform the method for estimating the longitudinal acceleration of at least one rail vehicle as described in claim 15; - at the available time t av The value v of the independent longitudinal reference velocity of the at least one rail vehicle is measured. ref (t av ); - Determine the longitudinal speed v of the rail vehicle according to the following steps. RV (t i1 ): • Calculate from the available time t av up to the first measurement time t i1 The continuously determined estimated longitudinal acceleration a lon The integral; • The at least one rail vehicle will be at the available time t av The value of the independent longitudinal reference velocity V ref (t av Add the integral result of the estimated longitudinal acceleration to the integral.

20. The method for estimating the longitudinal velocity of at least one rail vehicle according to claim 19, wherein, The determination of the longitudinal speed v of the rail vehicle RV (t i1 The steps are performed using the following formula: in: - It is at the available time t av The independent longitudinal reference velocity V ref (t av ); - From the available time t av To determine the longitudinal velocity v RV The first measurement time t of (t) i1 The continuously determined estimated longitudinal acceleration a lon The points.

21. The method for estimating the longitudinal velocity of at least one rail vehicle according to claim 19 or 20, wherein, The first measurement time t i1 Recovery t with available time ret_av Consistent, the recovery t at the available time ret_av The independent longitudinal reference velocity V ref It becomes available again after being unavailable.

Images