DE10209378A1 - Detecting and minimizing centrifugal and sliding processes of short-range electric railway vehicle involves locking control loop if maximum revolution rate difference exceeded and nulling - Google Patents

Abstract

The method involves locking a control loop for a time determined by a control signal if a maximum value for the difference between the engine revolution rate and a reference revolution rate characterizing the vehicle speed is exceeded, setting to a null initial value followed by releasing the control loop. The revolution rate difference is then controled back to the initial desired value.

Description

The invention relates to a method for the detection and minimization of Spin / slide operations on rail-bound electrically driven Commercial vehicles, in particular with asynchronous motor drive, with the Features from the preamble of claim 1.

For rail-bound, electrically powered mass transit vehicles A skid / slide protection primarily serves to maximize the over Wheel / rail contact transferable forces when requested by the driver Total traction available between the wheels and the rail Adhesion forces exceeds. It is based on the following basics went out.

Due to the friction, a wheel can roll in due to the wheel contact force generate a usable tangential force at the point of contact with the rail. For this purpose, in addition to the rolling movement, there must also be a relative movement Run the rail in the contact area. This relative or Hatching speed arises from the difference between the Circumferential speed of the wheel and the driving speed. At constant Driving speed increases with increasing slip speed Power transmission ability of the rolling wheel. At higher speeds determines the ratio of in the case of acceleration Slip speed to vehicle speed the adhesion. In the low Speed range up to approx. 10 km / h results in a maximum adhesion at Slip speed of approx. 0.1 km / h. At higher speeds at a slip of approx. 1.2%. If the rail conditions are bad, it will Maximum adhesion with partially significantly higher slip is reached.

If you allow slip values that are greater than the slip at maximum Are non-positive, so the reduced friction results in a smaller one Maximum deceleration or acceleration and a smaller one Maximum traction.

Operation at high slip speeds can vary depending on Drive constellation and rail condition often so-called chatter or Cause rolling vibrations. In addition to unpleasant noise emissions these vibrations cause heavy wear in the mechanical Draw the drive train. A very large delay Slip values (up to locking the wheels) also cause a strong one Wear of the wheels. Acceleration with skidding wheels can especially in heavy rail traffic, lead to deformation of the rail.

There are a number of methods for recognizing and mastering Skidding / sliding operations.

On the one hand, the acceleration of the can in a so-called dv / dt limitation Motor and thus indirectly the acceleration of the mechanically coupled Wheels are monitored. The instantaneous acceleration is from the Engine speed (circumferential wheel speed) calculated by differentiating. If it exceeds a certain value, there is a sliding or The wheels spin and the dv / dt limit causes a reduction the target torque responsible for the process. The differentiation the noisy speed signal is problematic and continues to be the speed determination and thus also the differentiation at low Speeds inaccurate.

The so-called dv limitation is the torque to be set Regulated taking into account the current rail condition. Doing so the slip between a reference speed and the speed of the decelerating or driving wheel is limited. If the requested adhesion the maximum in the present rail condition transferable, the DV limitation prevents spinning or blocking the wheels by limiting the slip to the reference speed or by regulating the target torque.

Determining an exact reference speed is expensive or inaccurate. When using a reference wheel, the slip limitation must be between Reference and drive wheel speed out of consideration for cornering, different wheel diameters, measurement noise, wheel flange cages etc. very much be carried out roughly. The consequence is that the drive wheel is frequent is operated clearly outside the adhesion maximum. this leads to lower degrees of utilization of the wheel / rail contact as well as to the described mechanical vibrations.

The invention is based, based on an object Slip limitation a method for the detection and control of skid / To create sliding processes in which the speed difference monitoring is very can be focused and therefore responds very quickly, so that the available frictional connection in the best possible way can be exploited. The occurrence of vibrations should be as possible be avoided.

This object is achieved according to the invention with the features the characterizing part of claim 1. Advantageous Further developments of the inventions are described in the dependent claims.

The basic idea of the invention is, in the case of a too large one Slip between the driven or decelerated wheel of the vehicle and the rail, which is caused by an excessively large difference between the actual value the engine speed and the vehicle speed Reference wheel speed is marked, the current machine torque to reduce by setting directly in the control loop Torque is intervened. As below using an exemplary embodiment the current forming for the torque is explained in more detail responsible PI controller reset directly. This direct intervention in the torque Regardless of its settings, the controller leads to maximum speed Reduction of the actual torque.

It has proven to be advantageous if the method has the characteristics of Claim 2 has. As below using a Exemplified embodiment is thus a particularly rapid torque optimization possible.

In a particularly advantageous, described in claim 3 The reference wheel speed is the embodiment of the method according to the invention carried out as a "virtual wheel speed" by means of a "speed monitor", the expected theoretical speed curve from the measured Engine speed, the setpoint and brake means dependent Vehicle target acceleration and various virtual quantities are calculated. In contrast to conventional reference speed recordings, for example using an impeller, In principle, this type of reference speed calculation does not apply different wheel diameters, cornering etc. resulting Problem, the reference speed only after offset additions with the measured value to be able to compare. With slow speed changes z. B. by The determined reference value follows the cornering due to the hard coupling Measured value offset-free.

This determination of the reference wheel speed makes use of Calculation methods as used in a known method for damping mechanical torsional vibrations between a torque-controlled, variable-speed drive motor and one with this drive motor mechanically coupled low-frequency vibratory drive train of a Vehicle are used, which is described in DE 199 43 067 A1.

The process according to the invention and an example of its application is explained in more detail with reference to the accompanying drawings.

The drawings show:

Fig. 1 in a schematic circuit diagram of the electro-mechanical drive of a rail vehicle in model-like representation;

Figure 2 shows the control depiction of a device for calculation of the virtual wheel.

Fig. 3, the control depiction of a resettable numerical PI controller for torque control;

Fig. 4 is a schematic representation of the learning algorithm for optimizing torque;

Fig. 5 a diagram showing the timing in a typical braking operation using the method to identify and control of spin / sliding operations.

The basic structure of the electromechanical drive of a rail vehicle is explained below with reference to FIG. 1.

A tensile force setpoint F _{Z is} transmitted from the vehicle bus to the drive control module ASM. The torque setpoints M ₁ or M _{2 to be set are} formed from this value and are transmitted to the converter control modules USM. The drive motors Mo1 and Mo2 are controlled via the control unit DPU. Their speed is recorded by speed meters Dr1 and Dr2 and passed on as measured values V1 and V2 to the converter control modules USM. The devices of the method described below for limiting the speed difference between the actual value of the motor speed and the reference wheel speed are localized to the drive in the converter control module USM, specifically for each machine. There is no functional difference between the anti-skid device.

To determine the reference wheel speed as a virtual wheel speed, a device is used, which is referred to below as the “speed monitor” and is installed in each converter control module USM. The speed monitor continuously calculates the expected theoretical speed curve n _F = ≙ _{F, virt.} from the setpoint and brake medium-dependent vehicle target acceleration a _REF and the measured engine speed n = ≙ _M according to the following formula:

(α _M - α _{F, virt.} ) .C _virt. + (≙ _M - ≙ _{F, virt.} ) .D _virt. - J _{F, virt.} , ≙ _{F, virt.} = M _Roll + M _Wind + a _REF .km _F .r

The virtual vehicle _inertia corresponds to J _{F, virt.} that of the fully loaded vehicle. The virtual torsional stiffness C _virt. and the virtual damping D _virt. are determined empirically. In contrast to terrain-dependent torques, the wind torque M _wind and the rolling torque M _roll are taken into account as theoretical values. m _F corresponds to the mass of the fully loaded vehicle, r is the wheel diameter and k is the gear ratio. In the differential equation shown above, α _{M is} the rotation angle of the motor and α _{F is} the virtual wheel rotation angle.

The calculation of the virtual wheel speed n _F = ≙ _{F, virt.} can be read from the control representation of the speed monitor in Fig. 2. The only variable input variable required is the measured engine speed n = ≙ _M.

Fig. 3 shows in control engineering representation of the resettable numerical PI controller for torque control. The setpoint of the torque-generating current I _{q, soll} and the actual value of the torque-generating current I _{q are} fed to this controller, with Q0 and Q1 being the usual parameters for determining the gain and the time constant of the numerical PI controller. Furthermore, a signal SGS_AKT is supplied to the controller which has the value 0 as long as the difference between n and n _F which characterizes the slip does not exceed a permissible maximum value. It takes the value 1 as soon as the maximum permissible value for the slip is exceeded. In this case, as symbolically indicated in FIG. 3 at the output of the controller, the output of the controller is set to the value 0. This direct engagement in the slider leads, regardless of its settings Q0 and Q1 to a maximum of rapid reduction of the actual torque _M. If, after the very rapid torque reduction, the slip between the observer speed n _F and the measured speed n again corresponds to the predetermined limit value, the torque controller is released again (SGS_AKT = 0) and the target torque is set to a somewhat lower value (for example by 5% reduced) than the original value requested by the driver. If skidding / sliding occurs again, the value SGS_AKT = 1 occurs again for the control signal and the process of dynamic torque reduction is repeated, the value M _REF requested by the driver again being slightly reduced (for example by 5%). If the limiting function is not triggered any further, the setpoint for the torque is ramped down to 100%.

The processes when SGS_AKT = 1 occurs only for a very short time referred to as micro-spin / gliding.

The processes taking place in connection with this intervention in the controller are shown schematically in FIG. 4 in a time diagram.

The possible course of the signal SGS_AKT is shown in an exemplary manner on the upper time axis. The timing diagram below shows the curve for the nominal value M _{REF of} the torque, the curve drawn with thick lines representing the curve of the nominal value, while the thinly drawn lines show the strong reduction of the actual value M _{IST of} the torque when the control circuit is blocked. The whole procedure corresponds to a learning algorithm. After several micro-spinning / sliding operations, the reduced setpoint M _REF-LA for the torque M _REF ideally converges against the maximum transferable value.

If the torque _setpoint M _{REF falls below} a predeterminable value (for example 85% M _REF ) during this process, a signal for discharging sand is emitted. This is shown in the SAND time diagram (SAND = 1). _Sanding is stopped again after another limit has been exceeded (e.g. 92% M _REF ). If the threshold falls below an even lower level (e.g. 70% M _REF ), a signal is issued to trigger the rail brake. This is shown in the diagram RAIL (RAIL = 1). The two functions have a hysteresis.

Fig. 5 shows measurement results for a typical braking operation, in which the method described above was used for the slip limitation. The following variables are plotted in the time diagram:
the setpoint M _REF for the torque
the reduced setpoint M _REF-LA for the torque
the actual value M _IST for the torque
the engine speed N
the SAND signal for sand delivery.

The strong fluctuations in the actual value M _IST for the torque during the micro-spinning / sliding operations can be clearly seen, as well as the successive changes in the reduced setpoint M _REF-LA for the torque, which in each case return to the after the micro-spinning / sliding operations have ended original setpoint M _{REF is} returned.

Claims (6)

1.Procedure for the detection and control of skidding / sliding processes in rail-bound, electrically powered local transport vehicles, in particular with an asynchronous motor drive, by means of a control circuit, to which a setpoint value for the torque and a variable characterizing the actual value of the torque is fed and which to an actuator outputs a correction signal acting to change the torque, the control circuit being supplied with a control signal which characterizes the slip between the driven or decelerated wheel of the vehicle and the rail and a function of the difference between the actual value of the engine speed and a reference wheel speed characterizing the vehicle speed is characterized in that when a predetermined maximum value of this speed difference is exceeded, the control circuit blocks for a period of time predetermined by the control signal and is set to the initial value ZERO and then again is released, after which the setpoint value for the torque is additionally set to a value that is reduced compared to the output setpoint value and, if the maximum value of the speed difference continues to be undershot, is returned to the output setpoint value in a time-controlled manner. 2. The method according to claim 1, characterized in that the control signal is a digital signal which, when the maximum value of the Speed difference assumes the value 1 and when the frequency increases of the control signal, the setpoint for the torque is reduced several times becomes. 3. The method according to claim 1 or 2, characterized in that the reference wheel speed as "virtual wheel speed" in a model calculation from the actual value of the engine speed (n = ≙ _M ), a virtual moment of inertia (J _{F, virt.} ) _Of the fully loaded Vehicle, a virtual torsional stiffness (C _virt. ) Of the drive train, a virtual damping constant (D _virt. ), Virtual values for the rolling torque (M _Roll ), the wind torque (M _Wind ) and a target vehicle acceleration (a _REF ), the mass (m _F ) of the fully loaded vehicle, the wheel diameter (r) and the gear ratio (k) according to a differential equation:
(α _M - α _{F, virt.} ) .C _virt. + (≙ _M - ≙ _{F, virt.} ) .D _virt. - J _{F, virt.} , ≙ _{F, virt.} = M _Roll + M _Wind + a _REF .km _F .r
[α _M = angle of rotation of the motor, α _F = virtual angle of rotation of the wheel] is determined by means of a computer, the virtual variables being determined and, after forming the difference with the actual value of the engine speed, the control signal to be supplied to the control circuit is generated by comparison with the predetermined maximum value. 4. The method according to any one of claims 1 to 3, characterized in that if the setpoint for the torque is reduced several times If the signal falls below a first minimum value, the signal for Sand is released. 5. The method according to claim 4, characterized in that at If a second minimum value is undershot, a signal for triggering the Rail brake is released. 6. The method according to any one of claims 1 to 5, characterized in that the return of the reduced value of the target value for the Torque to the output setpoint ramped with a given slope he follows.