DE102009024996A1 - Method for determining correction value for controlling analog valves in motor vehicle brake system, involves forming correction value based on comparison of actual and target parameters for correcting signals to correct valve opening flow - Google Patents

Abstract

The method involves determining wheel rotation characteristics of a motor vehicle wheel, and determining wheel acceleration force from the characteristics. Valve control signals of pressure reduction- and/or pressure build-up valves are evaluated, where the valves regulate brake pressure of the wheel. Target and actual parameters are formed from the signals and characteristics, respectively. A comparative examination of the target and actual parameters is executed. A correction value is formed based on the comparison for correcting the signals to correct valve opening flow of an inlet valve.

Description

The Invention relates to a method according to the preamble of claim 1.

Automotive brake control units with partially analog or analogized driven solenoid hydraulic valves (hereinafter called analog valves) are known per se. The invention refers to such motor vehicle brake control devices Analog valves that adjust the wheel brake pressure individually without pressure sensors to regulate in the wheel brake cylinder. Corresponding motor vehicle brake systems are usually designed with just one form sensor with the aim of using pressure models over the pressure in the wheel brake cylinders other data accessible to the brake control electronics, such as valve opening times, wheel speeds, etc. too determine.

In such brake pressure control systems without Raddrucksensoren be even the intake valves of an ABS / ESP brake or Raddruckregelungssystems operated analog or analog. Due to the lack of wheel pressure sensors is no direct Raddruckmesssignal available. For the individual brake pressure control is known per se, the wheel pressure over Pressure models for each wheel in real time in the controller of the brake control unit to calculate. The print model thus places within certain limits of accuracy a theoretical print size available.

In the braking systems described above or more specifically, in the used in printing models or the printing model, there is generally the goal of the most accurate pressure control of analogized Inlet valves in a wheel brake pressure control system without Allow wheel pressure sensors and in particular the accuracy the regulation and thus the reliability of known wheel pressure models for analogized pressure valves to further improve.

Of the Need for improved pressure control of analog operated intake valves in a motor vehicle brake system is especially when the actual driver pressure above the usual existing pre-pressure sensor detectable range is because in this Fall the pressure difference across the inlet valve, anyway can only be estimated to contain an additional unknown.

In connection with the control of analogized brake pressure control valves, it is known per se to carry out an adaptive valve control, which incorporates the wheel rotational behavior for dimensioning the valve control. Such a procedure goes out of the EP 1 625 058 B1 the same applicant.

The known pressure model is based on the simplistic assumption of a constant blocking pressure level. With this assumption becomes the pressure built up within the print model with the previously (via digital pulses) degraded model pressure compared to the sizing If necessary, correct the valve current in the event of a deviation. To this In this way, a correction value can be determined with which future values can be determined Valve controls based on the adaptive model more accurate can.

Of the Pressure build-up and pressure reduction occurs in brake control procedures pulsed very often. As the known pressure model Blocking pressure changes not taken into account, can lower pressure deviations between pressure build-up and previous pressure reduction are not taken into account, which the model does not model printing, especially in the adaptation phase fast enough to the actual pressure. In case of deviations between model pressure and actual pressure but then works a wheel control such as B. ABS not optimal.

The The object of the present invention is now a method to improve a brake pressure control and in particular the time to bring model pressure to actual pressure (Adaptation phase) in an anti-lock control system.

These The object is achieved by the method according to claim 1 solved.

The invention takes into account the idea that the wheel acceleration of a wheel is determined by its mass inertia and the sum of all braking torques acting on the wheel: J · φ .. = ΣM i ≈ -M br + F Along · r dyn , With J = moment of inertia, φ .. = wheel acceleration, M _i = wheel torques, -M _Br = braking torque, F _longitudinal = tire _longitudinal force, r _dyn = dynamic rolling radius

Accordingly, if sufficient decoupling is assumed, the wheel torques M _{i are} essentially limited to the decelerating braking torque -M _Br resulting from the brake pressure and the acceleration-resultant supporting torque F LF resulting from the tire _{longitudinal force} . A strong wheel deceleration occurs when the brake torque exceeds the support torque.

According to the invention, the wheel deceleration is set in relation to an applied brake torque change. According to the method of the invention, therefore, a comparative consideration is carried out leads, in which the occurring after a pressure control wheel dynamics is compared with the applied according to the pressure model braking torque change. The comparison can be quantitative or only qualitative. In the simplest case, the comparison merely checks whether the wheel behavior is plausible.

• a) Betrachtung der auf eine inkrementale Bremsmomentänderung folgenden Radverzögerungsänderung (Radbeschleunigungsänderung, siehe Ansprüche 2 und 3).
• b) Betrachtung der seit Beginn einer Druckaufbauphase aufgebrachten Bremsmomentänderung und der unmittelbar nach dieser Phase, insbesondere während der darauffolgenden Druckabbauphase, beobachteten Radverzögerungsänderung (siehe Ansprüche 4 und 5).
• c) Betrachtung des absoluten Modelldrucks und Vergleich mit der Radverzögerungsänderung (siehe Anspruch 6).

The invention specifies three preferred method variants for determining correction values which differ from one another with regard to the possible accuracy as a function of the driving condition and the road conditions. Depending on the current driving and driving condition conditions, one of the methods may have advantages over the error correction accuracy over the other method.

• a) Consider the following on an incremental braking torque change wheel deceleration change (Radbeschleunigungsänderung, see claims 2 and 3).
• b) Considering the brake torque change applied since the beginning of a pressure build-up phase and the wheel deceleration change observed immediately after this phase, in particular during the subsequent pressure reduction phase (see claims 4 and 5).
• c) Consideration of the absolute model pressure and comparison with the Radverzögerungsänderung (see claim 6).

In method a), an incremental braking torque change or a pressure build-up pulse is used as the desired characteristic. The comparison with the actual characteristic takes place on the basis of the wheel delay change directly following this moment change or the pulse. Thereafter, at least a check is made as to whether the result is plausible. Method a) allows for the most accurate comparison between expected momentum change and actual moment change under constant environmental conditions (no longitudinal force change) compared to methods b) and c), but is sensitive to external disturbances including both coefficient of friction changes and longitudinal force decreasing with increasing slip , In these cases, the wheel acceleration is no longer causally related to the applied moment change. Therefore, method 1) is particularly suitable in situations where an immediate evaluation of the wheel acceleration is possible (high sampling time), or the tire has a pronounced range of constant longitudinal force. In these cases, the momentum value calculated from the acceleration approximates the torque change applied by the build-up pulse and can be related to the expected torque change ΔM _{Br, real} : J · Δφ · ≈ -ΔM Br, real (1)

In method b), the actual brake torque change is calculated from the maximum deceleration during the following unstable phase. This is compared with the brake torque change from the pressure model from the beginning of the pressure buildup phase to the entry into the unstable phase. Accordingly, according to method b), the total braking torque (not derived from the absolute model pressure or the stored blocking pressure level as in method c) during a pressure build-up phase (as desired parameter) is compared with the maximum occurring during the following pressure reduction phase Wheel deceleration (actual parameter) is set. The actual brake torque change results from the extreme value of the wheel deceleration during the unstable phase. It can be assumed that the braking torque level M _Br at the beginning of the pressure build-up phase is below the braking torque level which is necessary for the entire wheel recuperation slip region which has passed through during the previous unstable phase: M br ≤ F (λ) Along · r dyn (2)

In formula 2) F (λ) _longitudinal the longitudinal force in function of the slip λ according to the instantaneous adhesion coefficient μ which is valid for the assumed simplified case usually a constant wheel loading. Method b), like method a), assumes that there are no jumps in the coefficient of friction. Unlike method a), method b) assumes that the excess torque present during an unstable phase is never greater than the brake torque change applied during the pressure build-up phase. This is especially true if a wide range of slip has already been passed before the start of the pressure build-up phase.

at Method c) is the wheel deceleration as an actual parameter in a comparative consideration to the absolute braking torque or absolute setpoint pressure (setpoint characteristic), possibly reduced by a hysteresis moment, set. During process b) the in process a) still present risk of misjudgment by a wheel deceleration increase caused by the tire itself excludes always exists in the methods a) and b) the risk of a misjudgment due to friction value changes during the scheme. The absolute target pressure is in particular taken directly from the pressure model calculated in the controller. Thereby, that the absolute braking torque is considered, the danger is prevented that after a Reibwertsprung existing, from the longitudinal force resulting support torque below the beginning the pressure buildup existing torque is located. Therefore this procedure is safe against misrecognition than the methods described above a) and b).

Out J · φ .. = ΣM i ≈ -M br + F Along · r dyn (3) then follows: J · φ .. ≥ -M br (4)

Method c) is particularly suitable for detecting valves that are too wide open in the previous pressure build-up cycle on lanes with low friction (too high valve coil current). At low friction, the longitudinal force term is small compared to the braking torque. In this case: J · φ .. ≈ -M br (5)

As already in process b) is the actual Braking torque from the maximum deceleration during the unstable phase is calculated; the reference braking torque from the model however corresponds to the model moment at the beginning of the unstable phase, preferably minus the hysteresis torque (brake pressure change, due to friction to no significant brake force change has led).

method c) is characterized by a comparatively high level of security, which is a very fast correction of the valve opening current allows. However, the scope is as already mentioned, mainly on rather rare situations limited to lanes with low friction coefficient. Driving situation, in which the methods a) and b) are feasible, result However, considerably more often. But that is due to the risk of misrecognition a slower correction of the Valve opening current necessary.

Prefers is in the inventive method a Correction taking into account at least two of the above Method, in particular all the above method performed. The formation of the correction factor for the drive current for the valve expediently but only when a previously executed Plausibility check requires this.

ermittelt.Preferably, based on at least one of the above methods, a correction value R according to the formula

determined.

In all three methods a) to c), the value for ΔM _{Br, real} in particular from the term -J · φ .. certainly.

In the case of method a), the value for ΔM _{Br, model} is determined, in particular, from the momentary jump which has taken place by the last pressure build-up pulse.

In method b), the value for ΔM _{Br, model} results from the momentary jump attributable to the total pressure buildup, and
in method c), this value is determined in particular taking into account the stored blocking pressure level (absolute model pressure) and a proportionality factor.

The correction value R is a measure of the error in the valve opening behavior with which the pressure model which actuates the valve (s) can be corrected. Appropriately, the application of the correction factor described here is carried out together with other correction factors, their determination, for example, in documents EP 1 625 058 B1 is described.

Out the valve map can be the volume flow error determine appropriate valve current offset. This offset will be in particular the valve current used is corrected with the correction value R, so that an improved control is obtained.

According to one Another preferred embodiment of the method is a weighting of the above method (s), which of the Robustness of the error detection method and / or the road friction coefficient and / or the current driving situation.

Especially For this purpose, the determined error offset, particularly preferred only partially provided with a factor in the interval between [0 ... 1], so that a correction of the valve current only up to a predetermined maximum value is possible.

Further preferred embodiments will be apparent from the dependent claims and the following description of an embodiment Hand of figures.

It demonstrate

1 a μ-slip diagram for explaining method a),

2 a μ-slip diagram for explaining procedures b), and

3 a μ-slip diagram for explaining method c).

The μ-slip curve in 1 shows the course of a typical brake pressure control for explaining method a). On the Y-axis is the friction value μ and the slip S is plotted on the X-axis. The course of the brake pressure during an anti-lock braking of a passenger vehicle on an asphalt road surface (high friction coefficient) shows curve 4 , First, the vehicle is braked by applying the brake. In the case of exceeding the adhesive limit, at least one wheel slips, which causes pressure to be reduced by the ABS regulator on this wheel (pressure reduction phase). In the subsequent pressurization phase, the brake pressure regulator successively generates the pressure buildup pulses P1, P2 and P3 at the pressure inlet valve. In the diagram pulse P3 leads to a renewed tearing of the wheel; that is, the roadway, the requested brake pressure or the braking torque, which is approximately proportional to μ in the example shown, no longer transmitted to the road. The brake pressure curve (curve loop) running above the μ-slip curve moves in the direction of the arrow 5 , The bike slides. The wheel speed signals of the wheels indicate excessive slippage, causing the pressure regulator to again engage with one or more pressure reduction pulses in the braking process. The pressure falls in the direction of arrow 1 and go through point 2 , Below the μ-slip curve the wheel engages again. In the course of this, the current brake pressure in the direction of the arrow moves in the brake pressure curve 6 to point 3 in the rising branch of the μ-slip curve. Thereafter, the second control cycle begins with renewed pressure build-up.

Δp _i denotes a pressure build-up pulse, which for method a) designates the desired characteristic with which the wheel behavior is compared. From the wheel spin behavior the value ΔM _ist can be determined. A comparison of the torque change on the basis of the pressure build-up pulse and the moment change after the pulse allows a plausibility check. If the nominal torque change approximately matches the actual torque change on the basis of the wheel information within predefined limits (arrow I ), the regulation works in the optimum range. Due to the decreasing longitudinal force with increasing slip, the wheel absorbs an excess torque. Therefore, the wheel speed initially increases until the braking torque is reduced via a pressure reduction (arrow II ).

2 illustrates a typical control cycle associated with method b). At the beginning, the pressure is increased by the regulator, for example by means of the two pressure buildup pulses P1 and P2. The pressure buildup can also take place with fewer or more pressure buildup pulses. According to method b), the desired characteristic value ΔP is a sum over the two pressure build-up pulses P1 and P2 (or the sum of the pulses which were required during the pressure build-up). The amount of arrow III shows the variation of the actual torque _is .DELTA.M, which can be determined from the wheel rotational behavior as described above. The actual torque change is preferably plausible if it is less than or equal to the pressure build-up torque change that results from the pressure difference ΔP. If, during pressure regulation, a coefficient of friction jump occurs at a lower coefficient of friction (dashed μ-slip curve) 7 ), the change of the actual moment would be the larger amount of arrow IV correspond. In this case, method 2) is not feasible.

The in 3 illustrated control cycle refers to method c). Again, the cycle begins with, for example, pressure buildup P1 and P2. In this case, the target characteristic corresponds to the absolute pressure P, which is reached after the last pressure buildup pulse P2. Even in the case of friction coefficient jump during control, the torque change ΔM _is in the direction of arrow IV (Arrow ends on solid friction curve) or arrow V (Arrow ends on dashed friction curve) plausibility always less than or equal to the resulting from the stored blocking pressure level reference braking torque. This method leads to advantages especially at low friction, as already described above.

QUOTES INCLUDE IN THE DESCRIPTION

This list The documents listed by the applicant have been automated generated and is solely for better information recorded by the reader. The list is not part of the German Patent or utility model application. The DPMA takes over no liability for any errors or omissions.

Cited patent literature

• - EP 1625058 B1 [0006, 0028]

Claims (9)

Method in a motor vehicle brake control system with a pressure control by means of at least one analogized Raddruckventil taking into account a Raddruckmodells, which does not use pressure data from a wheel pressure sensor and in which a correction value for correcting the drive current for the analogized valve during an ongoing anti-lock control is determined with the steps - Determining the Raddrehverhaltens at least one vehicle wheel and preferably determining the wheel acceleration from the wheel rotation behavior and - Evaluating the valve control signal the pressure reduction and / or pressure build-up valve, which / which the Regulates brake pressure of the at least one vehicle wheel, in through the steps - Forming a target characteristic from the valve control signal (s), - Make one Actual parameter from the wheel rotational behavior, - Carry out a comparative analysis of nominal characteristic and actual parameter and - Make a Correction value R for correcting the valve drive signal s / e, in particular, the valve opening current of the intake valve according to the comparative consideration, or merely Plausibility of the comparison result. A method according to claim 1, characterized in that the actual characteristic variable formed from the Raddrehverhalten the change of the wheel acceleration .DELTA..PH .. and the target characteristic is an acting on the wheel torque change .DELTA.M, in particular braking torque change .DELTA.M _Br , is. Method according to claim 2, characterized in that that the moment change in incremental steps (for Example pressure change pulses) takes place and in the comparative Consider an increment of moment change with a following wheel acceleration change is compared. Method according to claim 2, characterized in that that the moment change in incremental steps, in particular in the form of pressure change pulses, and in the comparative Consider a sum of incremental moment changes with a subsequent wheel acceleration change during or one of the subsequent depressurization phase (s). Method according to claim 2 or 4, characterized that the actual moment difference from the maximum delay during the unstable one following the pressurization phase Phase of an anti-lock control is calculated. Method according to Claim 2 or 5, characterized that the target characteristic from the absolute model pressure the pressure buildup phase is determined. Method according to at least one of the claims 1 to 6, characterized in that a form pressure sensor with a limited pressure measuring range is present and the upstream pressure above of the form recordable by the form. Method according to at least one of the claims 1 to 7, characterized in that the desired characteristic is obtained from pressure values of the wheel pressure model and as an actual parameter the wheel acceleration φ .. is used. Method according to at least one of the claims 1 to 8, characterized in that the correction value R from the Quotient of the actual moment difference and itself calculated from the pressure model resulting torque difference.