RU2804825C2 - Method for controlling a railway vehicle, controller and railway vehicle - Google Patents

Abstract

FIELD: railway.

SUBSTANCE: group of inventions relates to the control of the traction system of electric vehicles. The method for controlling a railway vehicle is to determine the actual speed of the railway vehicle, determine the actual rotation speed of the wheel of the railway vehicle, obtain the estimated torque from the motor controller, estimate the maximum adhesion coefficient, determine the reference torque, and provide the reference torque to the engine controller. In this case, the maximum coefficient of adhesion is estimated based on the wheel speed, the actual speed of the railway vehicle and the calculated torque. The reference torque is determined by summing the torque calculated from the estimated adhesion coefficient and the output torque provided by the feedback control. Also claimed are a torque controller, a rail vehicle and a processor.

EFFECT: preventing the excitation of possible transmission resonance.

12 cl, 6 dwg

Description

The present invention relates to a method for controlling a railway vehicle, comprising a motor controller for an electrical machine driving at least one wheel of the railway vehicle.

Furthermore, the present invention relates to a torque controller for a railway vehicle, comprising a motor controller for an electrical machine driving at least one wheel of the railway vehicle.

Finally, the present invention relates to a railway vehicle comprising such a controller and a computer program product.

Anti-skid devices in the prior art perform the function of controlling the anti-lock braking system (ABC), typically maintaining the force at the traction limit through a threshold of speed, acceleration and jerk. Two strategies are usually used. The first strategy aims to eliminate ("reduce") slippage in order to operate within the micro-slip range of the clutch curve as much as possible. The first strategy requires rapid torque modulation, creating a periodic triangular initial torque profile. The second aims to "hold" a constant amount of slip to take advantage of the cleaning action due to the increase in wheel surface temperature (clutch regeneration). The second strategy, controlling a constant amount of slip, would apply a less variable force, but by assuming predefined thresholds for wheel slip/slip, acceleration, and jerk, in real-world applications wheel force and wheel speed can sometimes cause unwanted fluctuations.

In the publication "Adaptive nonlinear control of braking in railway vehicles" Caporale et al., "52nd IEEE Conference on Decision and Control", December 10-13, 2013, Florence, Italy, the methodology for the functioning of the traction control system in railway vehicles. The article discloses a simplified model for estimating the adhesion curve.

The purpose of the invention is to provide a reference torque as constant as possible to prevent the excitation of possible transmission resonance and to follow changing rail conditions.

In accordance with an aspect, there is provided a method for controlling a railroad vehicle comprising a motor controller for an electrical machine operating at least one wheel of the railroad vehicle. The method includes:

- receiving a torque request from the central controller;

- determination of the actual speed of a railway vehicle;

- determination of the actual speed of rotation of a wheel of a railway vehicle; - obtaining the calculated torque from the engine controller;

- estimation of the maximum coefficient of adhesion based on the wheel speed and the estimated torque received from the engine controller;

- determination of the reference torque based on the estimated adhesion coefficient; And

- providing a reference torque to the motor controller.

In accordance with embodiments, the present invention may include one or more of the following features:

- the main controller receives data from the driver of the railway vehicle and/or the train safety system;

- to estimate the maximum adhesion coefficient, a linear filter is used;

- the linear filter is a Kalman filter;

- an electric machine is used to accelerate or decelerate a railway vehicle;

- the calculated torque is based on the voltage measured at the inverter input and/or on the phase currents of the electrical machine;

- to estimate the maximum adhesion, a three-parameter model is used, in particular using the following equation:

,

Where is the vector of independent variables in regression, and

- vector of parameters, μ – clutch, and σ – slip.

- the reference torque is calculated for each motorized axis separately, the motorized axis being the axis that is acted upon by the electric machine;

- the electric machine is a permanent magnet electric machine or an asynchronous motor;

- feedback control is used to determine the reference torque;

- in feedback control, PI or PID controllers are used; and/or

- use the difference between the optimal wheel speed and the actual wheel speed, with the optimal wheel speed being determined based on the maximum coefficient of adhesion.

According to another aspect, there is provided a torque controller for a railroad vehicle, comprising a motor controller for an electrical machine acting on at least one wheel of the railroad vehicle, wherein the torque controller is adapted to

- receive a torque request from the central controller; - obtain or estimate the actual speed of a railway vehicle;

- obtain the actual speed of rotation of the wheel of a railway vehicle; - receive calculated torque from the motor controller;

- estimate the maximum coefficient of adhesion based on the wheel speed and the calculated torque received from the engine controller;

- determine the reference torque based on the estimated adhesion coefficient; And

- send reference torque to the motor controller.

In accordance with yet another aspect, a railroad vehicle is provided including a motor controller for an electrical machine operating at least one wheel of the railroad vehicle, a plurality of wheels, one or more wheel sensors 14 for detecting wheel speed, wherein the railroad vehicle comprises a torque controller according to the embodiment described herein.

Finally, the present invention relates to a computer program product containing instructions for implementing a method in accordance with the embodiment described herein when loaded and executed on a processor.

So that the manner in which the above features of the present invention can be understood in detail, the more specific description of the invention summarized above can be read with reference to the embodiments. The accompanying drawings relate to embodiments of the invention and are described as follows:

In fig. 1 schematically shows a railway vehicle;

in fig. 2 is a schematic block diagram of a controller in accordance with the invention;

in fig. 3 shows a schematic block diagram for determining the reference torque;

in fig. 4 schematically shows the adhesion curves depending on the longitudinal slip;

in fig. 5 schematically shows the torque curve; and the curve of train speed and wheel speed, and

in fig. 6 is a schematic block diagram for determining a reference torque in accordance with the invention.

In fig. 1 shows a railway vehicle 1 running on rails 2. The railway vehicle 1 includes wheels 3 which are driven by at least one electric machine 5. Only one electric machine 5 is mentioned in the following, however, embodiments also apply to railway vehicles with multiple electric machines 5. The electric machine 5 is designed to accelerate and decelerate the wheels 3. In addition, the railway vehicle includes a motor controller 7. Inverter 8 supplies power to the electrical machine. The motor controller 7 controls the state of the electric machine 5 and estimates the motor torque T _est by measuring the phase currents of the electric machine 5 and by measuring or estimating the phase voltages of the electric machine 5, which can be derived from measuring the DC bus voltage, the inverter input current and the state inverter commands.

According to an embodiment, the electric machine 5 is a permanent magnet machine. In another embodiment, the electrical machine is an induction motor.

The railway vehicle also includes a central controller 9 that is adapted to receive commands from a driver or other system that determine a reference speed and/or a reference torque T _ref for the wheels 3. In other words, T _ref corresponds to a torque request. For example, another system may be a train safety system, such as the European Train Control System.

In an embodiment, the railway vehicle 1 includes at least one speed sensor 12 for determining the speed v of the railway vehicle 1. Alternatively, the speed v of the railway vehicle 1 can be estimated by updating the measured wheel speed performed by the ABC device. In addition, the railroad vehicle 1 has one or more wheel sensors 14 for determining the rotation speed ω of the wheels 3. In an embodiment that may be combined with other embodiments described herein, the railroad vehicle 1 includes a torque controller 16 moment. According to an embodiment, the torque controller is an anti-lock braking system (ABC) control device. The torque controller 16 is configured to calculate a reference torque, which is then supplied to the engine controller 7. In an embodiment, the torque controller 16 is coupled to the engine controller 7 to receive the calculated engine torque. In addition, the torque controller 16 is connected to the main controller 9 to receive the torque request T _ref . In addition, the torque controller 16 is directly or indirectly coupled to the speed sensor 12 and the wheel sensor 14 to receive data for determining the actual speed of the railway vehicle 1 and the rotation speed of the wheels 3. In an embodiment, the torque controller 16 is configured to use an output signal one or more wheel sensors 14 for estimating the speed v of the railway vehicle 1.

In an embodiment, the central controller 9, the engine controller 7 and/or the torque controller 16 are configured as one controller or two or more controllers.

The torque controller 16 is provided to prevent slipping or slipping of the wheels 3. For example, the torque controller 16 is configured to calculate the maximum adhesion between the wheels 3 and the rail 2.

In fig. 2 shows the units provided in the torque controller 16 in more detail. The first block 20 is designed to evaluate the maximum adhesion (µ _est ) of the railway vehicle 1, in particular the wheels 3, with the rail. In addition, the first block 20 calculates the optimal wheel speed ω ₀ , slip σ and the absolute value of slip σ _Absolute . The first block 20 also provides the (actual) wheel speed ω as an output signal.

As already described above, the adhesion µ of a wheel to a rail depends on the condition of the wheel/rail surface (for example, dry or wet) and on the speed v of the railway vehicle. The sliding or slipping depends on the clutch µ, on the rotational speed of the wheels 3, on the speed v of the railway vehicle 1 and on the actual driving or braking force. In other words, sliding or slipping is the difference between the tangential speed of the wheel (at the point of contact on the rail) and the speed v of railway vehicle 1.

In addition, the second block 22 is configured to calculate the torque T _contr that must be transmitted by the electric machine 5 to the wheel 3. In accordance with the embodiments, the second block 22 implements the control itself.

The torque of the wheel 3 depends on the torque of the electric machine 5. The output torque T _contr of the second unit 22 is adapted to be supplied to the motor controller 7. The second block 22 is configured to control torque based on the estimated maximum grip μ _est from the first block 20, slip σ and torque request T _ref provided by the central controller 9.

In an embodiment, the torque controller 16, in particular the second unit 22, includes a first selection device for selecting whether a reference torque T is provided_ref directly as output torque T_contr or not. In a further example, the torque controller 16, in particular the second unit 22, includes a second selection device for distinguishing between traction and braking. This information is passed from the second block to the first block 20. For example, the second block 22 provides the values TSWITSCH and TBSWITCH to the first block 20.

TSWITCH is a logical (or boolean) value that is 0 (false) or 1 (true). If TSWITCH is 0, then the reference torque T _contr transmitted to the engine is the value T _ref determined by the driver. If TSWITCH is equal to 1, then the torque T _ref is no longer set as an output, but the output T _contr is calculated to use the maximum available adhesion force. This signal is always less than or equal (in absolute value) to the torque T _ref set by the driver. TBSWITCH is another boolean value that allows you to differentiate between traction and braking.

The first block 20 and therefore the torque controller 16 is configured to determine the optimal wheel speed ω ₀ and/or the maximum grip μ _est based on the railway vehicle speed v, the wheel rotation speed ω and the calculated torque T _est received from the controller 7 engine.

In fig. 3 shows a block diagram of various components included in the torque controller 16.

Block 24 implements a linear filter for estimating the maximum adhesion µ _est and for calculating the torque T _comp .

For example, a linear filter uses T _est and wheel speed ω to calculate an estimate of maximum grip µ _est . In an embodiment, a Kalman filter is used for this purpose.

The filter is adapted to calculate the adhesion curve. For example, the adhesion curve is used for this, as described in the article "Adaptive nonlinear control of braking in railway vehicles" by Caporale et al., "52nd Conference on Decision and Control", December 10-13, 2013, Florence , Italy.

The adhesion curve is calculated based on the following equation:

Equation (1),

Where is the vector of independent variables in regression, and
- vector of parameters. In other words, it is clear that θ ₁ =a, θ ₂ =b and θ ₃ =c. Equation (1) provides a simplified three-parameter model and is used to estimate the maximum adhesion in block 24. This is a function of longitudinal slip σ and three parameters a, b and c. In fig. Figure 4 shows the curves in accordance with equation (1) in comparison with the model presented by Polach in the article "Creep forces in simulations of traction vehicles on adhesion limit", WEAR 258, pp. 992-1000, 2005, publishing house "Elservier".

The curves calculated according to Polach's work are solid, while the curves obtained according to the above equation are dotted. The curves are shown for several railway speeds, in this case for several tram speeds. It should be noted that the left side of the coupling curve, where the peak is located, is quite accurate, while on the right side at least the sign of the derivative of the fitted curve matches the model.

Compared to the more complex Polach estimate, the parameters in accordance with equation (1) are calculated more simply, in particular in real time. In other words, the three parameters a, b and c are capable of describing all types of behavior that can be encountered in real trains.

According to an embodiment, filter 24 is implemented in block 20 of FIG. 2.

According to an embodiment, the adaptive control with linear filtering is built on the adhesion curves in accordance with equation (1) and their parameterization. Thus, it is possible that block 20 is adapted to calculate and/or track the point of maximum adhesion μ_est. In other words, the torque calculated by the motor controller 7 is used to calculate the coefficient μ_est clutch, in particular with a corresponding filter. From this information, the output torque T is calculated_contr, which must be applied by the motor controller 7.

In addition, in FIG. 3 shows a second filter 26 configured to determine the optimal slip σ ₀ based on the maximum adhesion coefficient µ _est . For example, you can use a least squares filter to fit a curve. Using the vehicle speed and the optimal slip σ ₀ , the optimal wheel rotation speed ω ₀ can be calculated. According to an embodiment, the second filter 26 is implemented in block 20 of FIG. 2.

In fig. 3 shows a control 28, such as a PI controller, for monitoring reference values. The controller 28 uses the wheel rotation speed error signal ω _e as an input signal. The error signal ω _e is calculated based on the difference between the optimal rotation speed ω ₀ and the actual rotation speed ω of the wheel. In other embodiments, other controls may also be used, particularly closed-loop controls such as a PID (proportional-integral-derivative) controller.

In an embodiment that may be combined with other embodiments described herein, controller 28 is implemented in block 22 of FIG. 2.

The controller 28 provides an adjustable torque T _reg as an output signal.

Finally, in FIG. 3 shows an additional feedforward control element 30 for feedforward operation. Feedforward control 30 is used to provide dynamic adjustment of the control torque T _contr based on process knowledge in the form of a mathematical model. The example uses the reference optimal rotation speed ω ₀ and its derivative as input data (not shown in Fig. 3). The output of the feedforward control is the feedforward torque _Tff .

According to an embodiment, the torque T _comp provided by the first (linear) filter 24, the output torque T _reg provided by the controller 28, and, optionally, the feedforward torque T _ff are added to calculate the control torque T _contr .

According to embodiments, the control torque T _contr exhibits reduced fluctuations, which provides benefits in terms of mechanical wear of components.

Thus, in accordance with the invention, the optimal value for the wheel rotation speed ω ₀ is estimated. Using the optimal speed ω ₀ of wheel rotation, the torque control loop becomes closed.

According to an embodiment, the optimal wheel speed ω ₀ and/or the maximum adhesion coefficient μ _est are updated in real time. Therefore, an accurate reference wheel speed is always used, and the torque controller 16 prevents torque fluctuations. Typically, the relative slip or slip is limited to within a few percent, so that mechanical stress on the transmission components, such as torsional vibrations of the wheelset and the resulting maintenance costs, is reduced while the service life of the wheel is increased.

In fig. 5, Graph A) schematically shows the torque curve, and Graph B) schematically shows the curve of train speed and wheel speed. As can be seen, the control torque T _contr , the calculated torque T _est and the real torque T _motor of the engine are very close to each other or even almost identical. In addition, they do not show strong changes. Moreover, as can be seen from graph B), which shows the speed vel _train of the train and the speed vel _wheel of the wheel rotation, the speed vel _wheel of the wheel rotation does not show much change.

Thus, the invention makes it possible to estimate the instantaneous value of the wheel/rail friction coefficient and thus estimate the actual wheel rotation force that the railway vehicle can use in order to accordingly update the reference torque set by the controllers 7, 9, 16, in particular, independently from the axis.

According to the invention, it is possible to ensure maximum traction and/or braking force under all operating conditions, for example under different rail adhesion characteristics. The torque estimated by the engine controller 7 is used to estimate the optimal usable wheel force and to update the torque T _contr accordingly.

According to the invention, the proposed torque control device or ABC device is capable of estimating current rail conditions based on measurements and provides a reference torque profile that is as close as possible to its optimal value. Moreover, according to the invention, when the optimal wheel speed is assessed, the torque control loop becomes closed and the resulting torque becomes less subject to fluctuation. Moreover, when traction, the distance traveled increases, and when braking, the braking distance decreases.

In fig. 6 is a flowchart of reference torque determination according to an embodiment, and includes more information compared with the embodiment shown in FIG. 3. The same reference numerals are used for the same functions. Thus, in general, the description of FIG. 3 also applies to FIG. 6.

In fig. 6 in addition to FIG. 3 shows the first selection device 32, which has already been described with reference to FIG. 3, depending on the value, either the torque T _AKO or the torque request T _ref required by the driver is transmitted to the engine controller 7 as T _contr . The torque T _comp provided by the first (linear) filter 24, the output torque T _reg provided by the regulator 28, and optionally the feedforward torque T _ff are added to calculate the control torque T _AKO .

In block 26, the vehicle speed is denoted by v. For example, as explained above, the speed v of a railway vehicle is determined using a speed sensor or an ABC device.

In fig. 3 and also in FIG. 6, T _cont or T _AKO , particularly calculated from the combination of the torque T _comp provided by the first (linear) filter 24, the output torque T _reg provided by the controller 28, and, optionally, the feedforward torque T _ff , is provided as a reference torque to the engine controller 7, overriding the torque request T _ref if slipping or slipping is detected, for example, if the difference between the tangential wheel speed and the railway vehicle speed exceeds a predetermined threshold value. For example, in the case of detection of slippage or slippage, the value of TSWITCH may be 1. Otherwise, the value of TSWITCH is 0. In case the value of TSWITCH is 0, the torque request T _ref is transmitted as T _contr to the engine controller 7. Otherwise, if the value of TSWITCH is 1, the value T _contr (in FIG. 3) or T _AKO (in FIG. 6) is transmitted to the motor controller 7 as T _contr .

According to an embodiment, T _comp may be calculated as follows, particularly at block 24:

Equation (2),

where R is the wheel radius, M is the axle mass, g is the gravitational acceleration, τ is the engine transmission ratio, and μ _est is the estimated current clutch.

According to an embodiment, a Kalman filter is used to estimate the current available traction, particularly in real time, for example at block 24, based on an estimate of the motor torque T _est applied to the wheel and the wheel rotation speed ω.

In accordance with embodiments, the vector θ is estimated using an adaptive extension of the Kalman filter (eg, an adaptive filter H at infinity (H∝)). The current available coupling µ _est is used by an adaptive extension of the Kalman filter as the basis for estimating the parameters θ ₁ = a, θ ₂ = b and θ ₃ = c of equation (1), a simplified coupling model, in particular in block 26.

In an embodiment, the adaptive filter H at infinity, which may be a recursive filter, provides robust attenuation of the error-to-noise ratio. In one example, filtering was extended to include natural positivity constraints on the parameters (e.g., parameters a, b, c must be greater than 0) along with other convex constraints to ensure that the optimal slip σ ₀ is practicable, which must fall within a given interval included in (0, k), for k < 1, an interval that can be adjusted according to operating conditions, such as vehicle speed. The optimal slip σ ₀ is the reference for the control loop that provides the maximum usable clutch µ, see FIG. 4.

For example, the parameters (eg parameters a, b, c) of the curve to be estimated are constrained to force σ ₀ to belong to a configurable sub-interval (0,1). The optimal slip σ ₀ corresponds to the maximum adhesion on the curve in accordance with equation (1). In other words, the maximum adhesion point is calculated based on the actual vehicle speed and, in particular, indirectly, based on the wheel speed and the calculated torque T _est in block 24.

The estimated optimal slip σ ₀ and the actual speed are used to calculate a reference optimal rotation speed ω ₀ which is then used for the rotation speed control loop which finally instructs the engine on the torque that is designed to use the maximum available clutch, see FIG. 6. For example, the optimal reference speed ω ₀ can be calculated as follows:

If thrust occurs, for example if TBSWITCH is greater than or equal to zero,

Equation (3)

where ω ₀ is the optimal reference speed, R is the wheel radius, v is the actual speed of the railway vehicle and σ0 is the optimal slip.

When braking, for example, if the TBSWITCH value is below zero,

Equation (4)

where ω ₀ is the optimal reference speed, R is the wheel radius, v is the actual speed of the railway vehicle and σ0 is the optimal slip.

According to embodiments, an object of the present invention is to provide real-time estimation of wheel-rail adhesion to improve anti-slip/slip performance by adapting a reference engine torque value (T _contr ) to actual adhesion conditions. An additional effect is to provide a torque reference that varies slowly, preventing the transmission from causing mechanical resonance.

The resulting engine torque is smooth and allows for maximum traction while reducing fatigue stress on the wheel axle.

Claims (26)

1. A method for controlling a railway vehicle, comprising a motor controller for an electric machine acting on at least one wheel of the railway vehicle, characterized in that determine the actual speed of the railway vehicle; determine the actual speed of rotation of the wheel of the railway vehicle; receive the calculated torque from the motor controller; estimating the maximum adhesion coefficient based on the wheel rotation speed, the actual speed of the railway vehicle and the estimated torque obtained from the engine controller, using a linear filter to estimate the maximum adhesion coefficient (μ _est ); determining a reference torque by summing the torque calculated from the estimated adhesion coefficient and the output torque provided by the feedback control, wherein the feedback control uses the difference between the optimal wheel speed and the actual wheel speed, wherein the optimal speed wheel rotation is determined based on the maximum coefficient of adhesion; And provide a reference torque to the motor controller. 2. The method according to claim 1, in which the main controller receives data from the driver of the railway vehicle and/or the train safety system. 3. The method according to claim 1 or 2, wherein the linear filter is a Kalman filter. 4. The method as claimed in any one of the preceding claims, wherein the electric machine is used to accelerate or decelerate the railway vehicle. 5. A method as claimed in any one of the preceding claims, wherein the calculated torque is based on a voltage measured at the inverter input and/or on the phase currents of the electrical machine. 6. A method according to any of the preceding claims, wherein a three-parameter model is used to estimate maximum adhesion, in particular using the following equation: , Where is the vector of independent variables in the regression, and is the vector of parameters, μ is the clutch, and σ is the slip. 7. The method as claimed in any one of the preceding claims, wherein the reference torque is calculated separately for each motorized axis. 8. The method as claimed in any one of the preceding claims, wherein the electric machine is a permanent magnet electric machine or an asynchronous motor. 9. The method as claimed in any one of the preceding claims, wherein the closed-loop control uses a PI or PID controller. 10. A torque controller for a railway vehicle, comprising a motor controller for an electrical machine acting on at least one wheel of the railway vehicle, wherein the torque controller is configured to obtain or estimate the actual speed of a railway vehicle; obtain the actual rotation speed of the wheel of a railway vehicle; receive calculated torque from the motor controller; estimate the maximum adhesion coefficient based on the wheel rotation speed, the actual speed of the railway vehicle and the estimated torque obtained from the motor controller, whereby a linear filter is used to estimate the maximum adhesion coefficient (μ _est ); determine a reference torque by summing the torque calculated from the estimated adhesion coefficient and the output torque provided by the feedback control, wherein the feedback control uses the difference between the optimal wheel speed and the actual wheel speed, wherein the optimal speed wheel rotation is determined based on the maximum coefficient of adhesion; And send reference torque to the motor controller. 11. A railway vehicle comprising a motor controller for an electrical machine acting on at least one wheel of the railway vehicle, a plurality of wheels, one or more wheel sensors (14) for detecting wheel speed, wherein the railway vehicle comprises a torque controller according to clause 10. 12. A processor executing a computer program containing commands that, when executed by the processor, implement the steps of the method according to one of claims. 1-9.

Images