CH648247A5 - Braking force control device, with anti-blocking protection, on a rail vehicle - Google Patents

Description

The invention relates to an anti-lock braking force control device on a rail vehicle with a pressure-operated brake device which has a pressure control valve and at least one brake cylinder interacting with it, a controller being provided which, depending on signals which are proportional to the wheel circumferential speed of a monitored wheel, derives from this the circumferential wheel deceleration - or -acceleration signals, from threshold value signals and from signals simulating the vehicle speed lowers the brake fluid pressure in the brake cylinder, keeps it constant or increases it. Such a braking force control circuit has become known, for example, from US Pat. No. 2,914,359 (Yarber). Most of today's pneumatic and hydraulic

55 see brakes, a braking force control (anti-lock system) in principle always raises, lowers and maintains the brake pressure (cf. FIG. 1). The control criteria are obtained as follows:

Braking force control requires an encoder that delivers a frequency fj proportional to the rotational speed <p £ 6o. This frequency then (digital or analog) values for both the rotational speed <p; as well as for the spin cp; won each wheel. Signals for triggering the anti-lock braking force control circuit are then obtained from this, for example by comparing the wheel speed with a fictitious vehicle speed that is as close as possible to the actual vehicle speed. Also a block

3rd

648 247

Kierschutz triggering via differential speeds of individual wheels or reaching certain threshold values of wheel deceleration or wheel acceleration as well as the combination of these criteria is possible and leads to the following conditions: lowering brake pressure, keeping brake pressure constant and increasing brake pressure, the aim of the control being to increase the wheel speed in this way regulate,

that the optimal slip between bike and road is maintained.

If the wheel starts to lock, its peripheral speed TjCPi drops steeply and its peripheral deceleration —rjtpj exceeds a certain value - (rcp) E, which usually causes the brake pressure Pi to be increased after a response time xE according to pj = pEie ~ kE (t-tEL> decreases. By lowering the brake pressure the wheel can accelerate again.

One starts from the idea that when this acceleration threshold is passed, the wheel has left the stable state and has entered the unstable area of the brake force coefficient-brake slip curve (grip value slip curve). This is particularly often not the case at the beginning of a regulation. For this reason, at the start of a braking force control, if r ^> v2 applies, falling below the - (np) E value should only keep the brake pressure (Pi = Peì), while lowering solely by the speed criterion r ^ <v , a replacement logic can be achieved.

The tendency that the wheel no longer collapses is now indicated by the fact that the circumferential deceleration becomes smaller and smaller and finally turns into acceleration. This is another signal that the pressure drop must be stopped. With the main logic, this is achieved in such a way that after a value is exceeded

(np) G, which is small and always positive and after a response time xG the brake pressure is kept constant (Pi = Paì) -

The bike can continue to accelerate during this stopping time. If the circumferential acceleration r ^ exceeds a high value + (rtp) A, this is a sign that the frictional relationships between the wheel and the roadway allow an immediate sharp increase in pressure. If in this case r; <Pj> vt is still fulfilled, then after a response time io xA the brake pressure becomes pi = Ps

, -Q-

~ ks * (t-tAi

+ TAi>]

with T., = 4- •! "for.

'Ai psi ~ pAi

20th

raised until the acceleration falls below the value + (r <p) A again and the response time xA has elapsed. The brake pressure is then kept constant (p; = pAi) until the circumferential acceleration has decreased so much that it reaches the very small positive value + (rcp) G— (rcp) v.

If the high value -I- (rcp) A is not exceeded when the wheel is re-accelerated due to poor friction between the wheel and the road or if 30 TjCpj <vb applies, the brake pressure is kept constant until + (np) G— (rtp) v is below. This criterion always means that after a response time xA the brake pressure according to the formula

Pi = PSi

1 - e

"N / A -twI + Twi)

wi k.

P.

with T..i = -r-— • jn 51

psi "PAi is made flatter than possible (kA <ks) again (see FIG. 2).

The most important criteria of the main logic are secured by decisions that are based on the speed curve rj (j> j. For this purpose, it is necessary to reproduce the vehicle speed v well, because for cost reasons it should be avoided to measure exactly.

The simulation of the vehicle speed is based on the measured circumferential speeds of all wheels of a car (or a bogie), in which deep breaks in the logical evaluation are bridged by replacement straight lines. These straight lines then begin to replace the actual speed profiles when the circumferential decelerations exceed a limit value - (r <p) H. The slope of these replacement curves is of the same negative amount - (np) H if the brake pressure increases on one of the linked wheels.

If not, i. H. If the brake pressure drops everywhere or is kept constant, the gradient takes on the smaller negative value - (rcp) w. The vehicle replacement speed v is now always formed from the highest value of the peripheral speeds or their replacement lines.

On the basis of the fictitious or simulated vehicle speed v mentioned above, for the present

45 two braking speeds v, and v2 are created by subtracting proportional and constant portions from v:

Vi = (l-? Ici) v-vc,

50

v2 = (1-Ic2) v-vc2.

If the peripheral speed r ^ falls below the comparison speed v ,, the brake pressure is reduced after 55 a response time xç as if the deceleration criterion - (rq>) E had been met. The same can be said for the two pressure maintenance commands. If the speed r ^ j exceeds the comparison speed v2, then after the response time xG the brake pressure is also kept constant 60, although the acceleration threshold + (rcp) G has not been reached. The first of the two lowering and pressure maintenance criteria is always effective.

These two speed signals therefore prevent the wheel from locking or the brake pressure being completely reduced if the main logic fails. In order to prevent the brake pressure from remaining in a holding phase for the rest of the braking, the time tHA is also provided. It begins when the wheel locks after a rise in pressure

648 247

ren starts and the brake pressure is kept constant or immediately lowered and causes at the end that the brake pressure is increased significantly after the corresponding response time. The formulas already given apply.

Previous attempts have shown that the brake force control considered works best for rail vehicles if the exponent for the brake pressure drop kE ~ 2 - and the exponent for the pressure increase s kA ~ 1 Ì. However, the volumes of the brake cylinders in the individual rail vehicles can fluctuate between 2.51 and 7.51 and more. As a result, the pressure control valve for the respective brake cylinder must be set. So far, this has been done by exchangeable nozzles in the pressure control valve or by a valve with a unitary effect.

Recently, a uniform and inexpensive pressure control valve has been required for all types of rail vehicles, particularly for cost reasons. Adapting the braking force control to the different brake cylinders by means of exchangeable nozzles in the pressure regulating valve or by means of a valve with a uniform effect is therefore no longer acceptable.

The object of the present invention is therefore to provide an anti-lock brake force control circuit for rail vehicles of the type specified above, in which an optimum brake pressure curve can be achieved using a pressure control valve and at least one brake cylinder which is not specially adapted to the pressure control valve.

This object is achieved by the present invention in that the pressure control valve is designed such that a predetermined brake pressure curve can still be achieved in the case of a large-volume brake cylinder, and in that the controller is designed such that the actual brake pressure curve in a brake cylinder with a smaller volume as a function of the mentioned signals about alternating pressure rise and pressure hold phases or alternate pressure drop and pressure hold phases with time adjustable by the controller is approximated to the predetermined brake pressure curve. Advantageous refinements and developments of the invention can be found in the dependent claims. Essentially, in the present invention, therefore, a pressure control valve with unchangeable nozzles is preferably used, which are dimensioned such that the predetermined values for the brake pressure drop or the brake pressure increase are achieved for a large-volume brake cylinder, the actual brake pressure profile being transmitted to the pulse process predetermined brake pressure curve is adjusted. In the case of smaller brake cylinder volumes, the excessively steep changes in brake pressure are consequently flattened by pulses. According to an advantageous development of the invention, the times for the pressure rise and fall phases are fixed, while the times of the pressure hold phases can be changed, and according to a further embodiment of the invention, the pressure hold phases are linked to the pressure rise and fall phases by factors. The optimal brake pressure curve is optimally set by selecting these factors (aA and aE). As will be described in more detail below, these factors are determined iteratively and, if necessary, changed so that the braking force control and thus the optimal braking pressure curve adapt to the actual conditions of the braking.

Of course, it is also possible to make the times of the pressure rise and fall phases adaptively variable, it being noted that these time intervals are in principle not related to the measurement time of digitally operating brake force control circuits. However, the effort of digital logic is significantly reduced if these time intervals are a multiple of the measurement time.

The invention is explained in more detail below on the basis of an exemplary embodiment in connection with the figures. It shows:

Figure 1 shows the brake pressure curve of a general brake force control circuit;

FIG. 2 shows the course over time of the wheel circumference acceleration or deceleration of the wheel circumference speed and of the controlled brake pressure;

FIG. 3 shows the time course of the wheel peripheral speed for several control cycles and the time course of the simulated vehicle speed;

FIG. 4 shows a block diagram of the brake force control circuit according to the invention;

Figure 5 shows the brake pressure curve over time in the brake force control circuit according to the invention;

6 shows the brake pressure curve and the wheel peripheral speed for a control cycle in the brake force control circuit according to the invention and

FIG. 7 shows a representation of the wheel circumference acceleration, the wheel circumference speed and the average brake pressure over time in the brake force control circuit according to the invention to explain its principle of operation.

FIG. 1 shows the basic time course of the brake pressure in a conventional brake lcraft control circuit with the phases: “raising”, “lowering” and “holding”.

FIG. 2 shows the time course of the wheel circumference acceleration, the wheel circumference speed and the brake pressure as a function of the control criteria described above.

Figure 3 shows the circumferential wheel speed and the simulated vehicle speed for several control cycles. Details of Figures 2 and 3 were explained in the introduction and do not need to be repeated here.

Figure 4 shows a block diagram of the brake force control circuit according to the invention. A monitored wheel 1 of the vehicle is coupled to a generator 2, which outputs a signal proportional to the peripheral speed of the wheel to an anti-skid device 3. The wheel 1 is braked by a brake 4, the brake 4 being actuated by an anti-skid valve or brake cylinder 5. The anti-skid valve or brake cylinder 5 is connected pneumatically or hydraulically to a pressure control valve 6, which in turn is connected to a pressure vessel 7. A manual brake actuation initiated by the driver of the vehicle acts in a known manner on the pressure control valve, as a result of which the anti-skid valve 5 is actuated. Via the anti-skid device 3, the braking process is monitored in accordance with the above-mentioned criteria, wherein devices are provided between the anti-skid device 3 and the anti-skid valve 5, which ensure a pressure increase 8, a pressure decrease 9 and a pressure maintenance 10 depending on the signals mentioned. Furthermore, devices 11 and 12 are provided which are assigned to devices 8 and 9 and contain devices for the learning process for increasing the pressure or the learning process for reducing the pressure. The factors aA and aE are thus determined via the devices 11 and 12. The anti-skid device 3 and the devices 8 to 12 represent the controller of the brake force control circuit according to the invention.

4th

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

5

648247

Figure 5 shows the time course of the brake pressure in the pressure drop or pressure increase phase. The relationship between the exponent k * of the unregulated pressure curve and the exponent îc of the middle pulsed pressure curve exists for both the drop and the increase in the brake pressure

TT =

*

1 + a -

2n.

n denotes the number of holding phases and a the factors aA and aE mentioned above.

For n = 1, kl ~ follows

1 +

where V! is always less than v2 and v2 is always less than v.

A rise in pressure was therefore favorable over time if the circumferential speed of the wheel r ^ between v2 and Vj is below the brake pressure drop threshold - (r; q>;) E.

In this case the factor aA has a favorable value that does not need to be changed. If, on the other hand, k <p [2: v2 applies to this criterion, this means that the ventilation io was too shallow, so that aA must be reduced, which in logic is indicated by the command aA = aAn-l - ^ aA

i5 is realized. ÀaA symbolizes a constant value with which aA is optimized. If, on the other hand, the circumferential speed r ^ falls below the comparison speed vb, the equivalent condition r ^: £ v, is fulfilled, then the increase in the brake pressure was too steep and aA is too large:

while for n - * oo we get:

*

1 + a

From this it can be seen that at the beginning of a pulsed pressure drop or rise (n = 1) the fc value is greater than afterwards (n-> oo), where an approximately constant E can be expected. This is favorable for the control, because the hysteresis of the brake is overcome even faster.

In order not to make the service life of solenoid valves too short, the number of holding stages is limited, for example to four. This means that afterwards the brake pressure change is not pulsed.

If one chooses the times of the pressure rise and fall phases 9a and 9e constant and the pressure maintenance phases according to aA. 3A (for the pressure increase phase) or aE 9e (for the pressure reduction phase), the K values (tcE or kA for brake pressure reduction or brake pressure build-up) can be selected based on the selection of the two factors aE (brake pressure reduction) and aA (for the Brake pressure build-up). The two variables aE and aA must be linked to the control logic, whereby in the present invention the factor aE is determined solely from criteria at the end of the pressure reduction phases, while the factor aA from the data at the end of an increase in brake pressure (with an exception to be explained later) be determined.

It can be demonstrated that, for example, the ventilation or ventilation times are not suitable quantities for determining the factors aE or aA. This data does not only depend on the È values, but also strongly on the braking hysteresis, vehicle data, the friction conditions and the controller data. According to an advantageous development of the invention, it is therefore proposed to find the optimal aE and aA values iteratively. This route is explained below in connection with FIGS. 6 and 7. It is assumed that the speed v (FIG. 7), which reproduces the vehicle speed v as faithfully as possible, two reference speeds v, and v2 being derived therefrom, which obey the following laws:

v, = (1- X.cl) v-vcl v2 = (1-ac2) v-vc2.

aAn - aA „-l + AaA-

It must be ensured that aA never becomes negative. At the beginning of controlled braking, aA = 0 must be set so that the pulsation can adapt to each brake cylinder.

Deviating from this, the factor aE is linked to the difference in the circumferential wheel speed Ave, which is formed during the ventilation phase, because during this time the circumferential acceleration r ^ is negative and positive, i.e. the circumferential speed first drops and then increases again after passing through its minimum. As a rule, the speed increase is low until the acceleration threshold + (rq>) E is exceeded. Ave is then a suitable quantity to influence the factor aE.

30th

35

In the case of very poor friction conditions, however, the wheel may spin up for a long time before this criterion is reached. This difficulty is avoided by specifying desired speed differences Ave and reducing the logical decisions to times calculated from them.

45 Fig. 7 shows primarily the courses of brake pressure, speed and acceleration during a bleeding phase. In the time span ATP, the acceleration is always less than - (r (p) E, while between Tp and Tp + ATG it changes its value from - (r (p) E to - + - (rcp) G. One takes this approximately while ATp assumes a constant acceleration value - (rcp) E and while ATG a linear acceleration change between - (np) E and 0, Ave calculates to iv £ • - (rf),

AT,

60

If one now starts from the idea that the speed difference Ave should always assume values between v2-v and v-v in a favorable braking force control, the following logical decisions result:

a) If the acceleration threshold + (np) G is exceeded and AvE> v2-vb d applies. H.

648247

6

ATß> 2-

v2 "V1

37717

-AT

■ AT,

R,

then the time course of the pressure drop was favorable and the factor aE need not be changed.

b) On the other hand, if one registers with this criterion ÀvE ^ v2-V], d. H.

ATfi4 2-

V2 "V1

(r ^) E "ATf

»ATc then this means that the ventilation was too steep, ie aE has to be increased, which is achieved by the instruction ae = ae aae. AaE again represents a constant value that needs to be optimized. At the beginning of each braking force control, of course, aE must be set to zero, c) If you find that AvE ^ v — vh d. H.

ATfi è 2 '

V - Vi

- (r j5) E "^ TP

"If ATr is fulfilled, then the wheel may lock up (eg as a result of a sudden change in the coefficient of friction). To prevent this from happening under all circumstances, aE = 0 must be set immediately. In this case, one must not wait until whether the + rcp) G threshold is exceeded, but this criterion must be monitored during the entire venting phase.

s For all these decisions, the time ATp must be determined using logic. If the acceleration value - (rcp) E is not undershot when the pressure is reduced, which can happen if the grip value is very poor when the wheel no longer comes up, then ATp = 0 must be set and io count the time span ATE from time i.

The above situation can also be described as follows (see FIG. 6): A pulsed pressure increase was favorable in terms of control technology if shortly before its end, i.e. H. at the moment of the decision the control speed vR between two-15 see two limit speeds v, and v2 lies. In this case, the size aA has the optimal value that is maintained. Conversely, if vR => v2, this means that the ventilation was too shallow, so aA should be reduced the next time the pressure rises. If vR 20 fell below the limit speed V] at this point, the increase in brake pressure was too steep and aA must be increased for further control.

When selecting aE, it is assumed that with a favorable braking force control during a pulsed pressure reduction, the greatest difference in the wheel speed AvM (the speed drop) should always take values between Avmin and Avmax. If this is determined at the end of the lowering phase, then the AE need not be changed. Registering AvM <Avmin at this moment means that

that the venting is too steep, that is to say increased for the next pressure drop. If, on the other hand, you recognize at any point in the venting phase that Avm> Avmax, the Ra-35 des may be blocked. To prevent this from happening under any circumstances, aE must be set to zero immediately so that the brake cylinder is vented with the greatest possible slope.

40

45

50

55

60

65

s

6 sheets of drawings

Claims (7)

648 247 2nd PATENT CLAIMS 1. An anti-lock brake force control device on a rail vehicle with a pressure-medium-operated brake device, which has a pressure control valve and at least one brake cylinder interacting with it, a controller being provided which, depending on signals that are proportional to the wheel circumferential speed of a monitored wheel, derives from this circumferential wheel deceleration - or acceleration signals, from threshold signals and from signals simulating the vehicle speed, the pressure medium pressure in the brake cylinder is lowered, kept constant or increased, characterized in that - The pressure control valve (6) is designed such that a predetermined brake pressure curve can be achieved with a large-volume brake cylinder (5) - And that the controller (3.8-12) is designed so that in a brake cylinder (5) with a small volume, the actual brake pressure curve depending on the signals mentioned via alternating pressure rise (SA) and pressure maintenance phases (aA $ A) or alternating pressure drop (3E) and pressure holding phases (aE9E) with time adjustable by the controller is approximated to the predetermined brake pressure curve. 2. Brake force control device according to claim 1, characterized in - That the pressure rise (9A) and pressure drop phases (& E) a predetermined fixed period - And the pressure holding phases (aASA; aE9E) have variable time periods that can be set by the controller (3, 8-12). 3. Braking force control device according to claim 2, characterized in that the pressure holding phases have a correspondingly adjustable factors (aA or aE) changed time period compared to the time period of the pressure rise or pressure drop phases (SA or 3E). 4. Braking force control device according to claim 3, characterized in that the factors (aA or aE) for a control cycle depending on the signals mentioned shortly before the end of the total pressure increase or. Pressure drop phase of a previous control cycle are defined. 5. Braking force control device according to claim 4, characterized in that the factors (aA or aE) are equal to zero in the first control cycle and are increased or decreased or kept constant by a predetermined value (AaA or AaE) in further control cycles . 6. Brake force control device according to one of claims 3 to 5, characterized in that the factor (aA) for the constant phases during a brake pressure increase cycle indicates the following variables if the following conditions occur shortly before the end of this brake pressure increase cycle: aAn = aAn_! if v, <vR <v2 io aAn = aAn. [- AaA if vR> v2 aAn = aAn-, + ÄaA if vR <Vj where aAn is the factor for a brake pressure increase cycle n 15, aAn-i is the factor for the previous brake pressure increase cycle n-1, ÄaA is a constant, predetermined value, vR is the peripheral speed of a monitored Ra-20 des v, and v2 are quantities generated proportional to the modeled vehicle speed v, where v> v2> V [. 7. Braking force control device according to one of claims 3 to 5, characterized in that the factor (aE) for 25 the constant hold phases during a brake pressure reduction cycle assume the following sizes: aEn = aEn_i if the following applies at the end of the brake pressure reduction cycle: Avmin <Ave <Avmax aEn = aEn.! + AAe if the following applies at the end of the brake pressure reduction cycle: AvE <Avmin aEn = 0 if at any time of the Brake pressure reduction cycle applies: AvE ^ vmax, in which 35 aEn is the factor for a brake pressure reduction cycle n, aEn-i is the factor for the previous brake pressure reduction cycle (n-1), AAE is a constant, predetermined value, AvE is the difference between the wheel circumference speed at the beginning of the brake pressure reduction cycle and the minimum of the wheel circumference speed during the brake pressure reduction cycle, Avmax a first predetermined fixed value, Avmin is a second predetermined fixed value, where: 45 Avmax> Avmin and where the end of the brake pressure reduction cycle is determined by the fact that the wheel circumference acceleration has exceeded a predetermined, positive quick value.