EP1134399B1 - Procedure and device for pressure control - Google Patents

Description

State of the art

The invention relates to a method and a device for pressure regulation, in particular for a common rail system, according to the preamble of claim 1 or 9.

(See: YANG WC ET AL: "DYNAMIC MODELING AND ANALYSIS OF AUTOMOTIVE MULTI-PORT ELECTRONIC FUEL DELIVERY SYSTEM" TRANSACTIONS OF THE ASME. JOURNAL DYNAMIC SYSTEMS MEASUREMENT, XX, XX, vol. 113, 1 March 1991 (1991-03- 01), pages 143-151)

From WO 96/03577 a system for controlling and regulating the injection pressure of an internal combustion engine is known in which a pump delivers fuel under a high pressure in a rail. The rail has a number of outlets connected to respective injectors. The system consists of a pressure regulator, arranged between the output of the pump and the input of the rail and which is acted upon by a drive signal, which can be predetermined by a controller.

With this control device, a leakage amount can not or only difficult to determine. In addition, this control device has only a low accuracy without further disturbance variable processing.

The pressure control according to the invention for a common rail system has the advantage that with a model for the rail pressure control, the hydraulic disturbance dQs / dt can be detected. This is on the one hand a fast response monitoring on a leak in the high pressure region of a common rail system over the entire operating range possible. On the other hand, one can Improved dynamics of the rail pressure control can be achieved by switching the disturbance on the manipulated variable.

According to the invention this is achieved in that by means of a model, a pressure regulator and / or an actuator of a pressure control loop is modeled, and that the model provides at least one signal that characterizes the disturbances of the pressure control loop.

The procedure according to the invention is particularly advantageous in systems in which an injection quantity QK, a leakage quantity QL, a control amount QS and a pressure drop amount act as disturbance variables on the pressure control loop.

The model includes at least one model for the actuator, which is preferably designed as a volume-controlled high-pressure pump. In such a high-pressure pump, the compressed amount of fuel and thus in the pressure accumulator, which is also referred to as a rail, conveyed amount of fuel can be controlled.

The model includes at least one model for the controlled system, which is also referred to as a route model. As a controlled system, the rail is considered. In the simplest embodiment, an integrator is used to model the rail.

The determined by the model disturbance is used to detect a leak and / or to form a Aufschaltgröße for the control loop. The injected amount of fuel may be included in the disturbance. But it can also be provided that the disturbance does not include the injected fuel amount.

Further advantageous embodiments of the invention will become apparent from the dependent claims.

Embodiments of the invention are illustrated in the figures.

1 is a system overview of a common rail system, FIG. 2 is a block diagram of a pressure regulator, FIG. 3 is a block diagram of a model of the pressure regulator circuit, FIGS. 4 to 6 are block diagrams of different variants of the model, FIG. 7 is a block diagram of a FIG. 8 shows a block diagram of an improved pressure regulator with monitoring of the observed disturbance variable.

1 shows by reference numeral 1 a controllable high-pressure pump. 2 is a valve which is in operative connection with the suction line of the tank 4 via the filter 5. 3 is a fuel filter. 6 is a gear pump. 7 is the metering unit. 8 is the pressure reduction valve, 9 is the rail, 10 is a rail pressure sensor, 11 is a flow restrictor, 12 are injectors, 13 is the accelerator pedal, 14 the crankshaft revolution, 15 the camshaft revolution, 16 is the injection or / and Ignition. 17 are further actuators, for example, for exhaust gas recirculation and 18 are other sensors.

1 shows a common rail system with pressure control via a quantity-controlled high-pressure pump 1 and optionally an additional pressure reduction valve 8. The pressure control takes place via the solenoid valve-controlled proportional valve, which is also referred to as Zumeßeinheit 7, which is the feed rate to the high-pressure pump 1 is set according to the manipulated variable of the controller.

The fuel passes through the fuel filter 3 to Zumeßeinheit 7. Depending on the applied to the metering control variable reaches an adjustable Fuel quantity in the high pressure pump. From there, the fuel is conveyed under high pressure into the rail 9. About the injectors 12 of the fuel enters the internal combustion engine. The control unit 16 controls the metering unit 7, the injectors 12, further actuators 17 and optionally the pressure reduction valve 8 depending on the signals of the sensors 13 to 15.

Fig. 2 shows a block diagram of a pressure regulator. 20 denotes a regulator. This is fed via a node the difference between an actual value Pist and a setpoint Psoll for the rail pressure. The controller 20 supplies a manipulated variable PRail, soll, to a node, at the second input of the node is a precontrol value dQvs / dt. The output signal of the node is fed to a controller 21.

The controller is preferably an element of the control unit 16. The desired value Psoll and the precontrol value dQvs / dt are specified as a function of different operating parameters, preferably depending on the rotational speed of the internal combustion engine and a variable characterizing the fuel quantity to be injected.

The actuator 21 is preferably the controllable high pressure pump 1, wherein the Zumeßeinheit 7 is acted upon by the manipulated variable. The metering unit 7 is preferably designed as a solenoid valve which supplies a specific amount of fuel to the high pressure pump depending on the signal applied to it. The high pressure pump delivers the actual value of the flow rate dQ _HDP, / dt is the high pressure pump in the rail. This quantity is reduced by the disturbances dQ _L / dt, dQ _S / dt, dQ _DAV / dt and the injection quantity dQ _I / dt. This results in the amount dQ _rail actually conveyed into the _{rail, is} / dt. This is by a downstream of the actuator Link point clarifies. With this size dQrail, / dt is the rail, which is also referred to as distance 22, applied. At the end of the controlled system, the actual value Pist of the rail pressure is displayed. This is detected by the rail pressure sensor 10.

The block diagram of Fig. 2 is intended to illustrate that behind the actuator 21 attack several disturbances. These are the injection amount dQ _I / dt, the leakage amount dQ _L / dt, the control amount of the injectors dQ _S / dt, and the pressure decrease amount dQ _{DAV / dt} flowing through the depressurizing valve. Apart from the injection quantity, these variables are initially unknown in the engine control unit. With the knowledge of these disturbing variables, on the one hand, the dynamic behavior of the pressure regulator could be improved and, on the other hand, leakage monitoring of the high-pressure region could be carried out.

The disturbance dQ _K / dt corresponds to the injection quantity. There are two known methods for determining the amount of leakage and control:

The determination of the leakage quantity takes place in coasting mode of the vehicle (injection quantity = 0). The leakage quantity is estimated by calculating the rail pressure gradient (dp / dt) as long as the manipulated variable dQ _{HDP, setpoint} / dt = 0.

If the manipulated variable of the pressure controller dQ _{HDP, setpoint} > 0, then the sum of the disturbance variables can be estimated by this manipulated variable. This has the disadvantage that the estimated disturbance is coupled to the dynamics of the pressure control.

The procedure according to the invention makes it possible, in the previously described common rail system with pressure control via a quantity-controlled high-pressure pump and optionally an additional pressure reduction valve, to control the disturbances in the pressure control loop (Leakage amount, control amount, injection quantity, pressure drop amount) to be determined constantly. Thus, the disturbance variables are known at each operating point and it can thus on the one hand, the dynamics of the pressure control can be improved and on the other a permanent leakage monitoring of the high pressure area done.

Fig. 3 shows the structure of an embodiment of the model. This structure essentially consists of a series connection of a system model 21.0 of the controller 21 and a system model 22.0 of the controlled system. At the output of this model is the replica PB of the controlled variable P on. The modifier model 21.0 is the input signal dQ _{HDP, Soll} / dt of the actuator 21 is supplied as input. The modifier 21.0 acts on a node with its output signal. The output signal of the connection point 30 reaches the system model 22.0. At the output of the distance model 22.0 is the modeled controlled variable P _B. This _is compared in a node 31 with the real controlled variable P _ist . The difference between the modeled and the real controlled variable reaches different factor specifications 29.1, 29.2, 29.n, 29.10 and 29.11. The factor specifications 29.1, 29.2 and 29.n are applied to the modifier model and the factor specifications 29.10 and 29.11 apply signals to node 30. At the output of the factor specification 29.11 is the modeled disturbance QB.

The factor specifications 29.1, 29.2 and 29.n specify quantities g1, g2, gn which influence the transmission behavior of the modifier model, depending on the deviation between the modeled and the real controlled variable. This means that the modifier model is adapted depending on the deviation between the modeled and the real controlled variable. 29.10 denotes a proportional factor, 29.11 represents an integrator with Tσ as the integrator time constant.

In technical terminology of control engineering, 29.10 represents a P element and 29.11 represents an I element.

The track model 22.0 simulates the rail 9 and has essentially an integrating behavior.

This is followed by a weighted connection of the difference between the real actual value and the simulated control variable to the individual timers of the modifier model. In this case, GA (s) represents the transfer function of the actuator. The actuator model 21.0 essentially contains delay elements which characterize the behavior of the electrical output stage, the solenoid valve 7 and the time-delayed delivery flow of the high-pressure pump.

At the entrance of the track model 22.0 are all disturbances. At the entrance of this model, the proportional interpolation of the observer difference (Pist - pB) takes place with the factor go weighted. All sizes of the input of the leg GB (s) other than the manipulated variable dQ _{HDP is} / dt attack are combined to the overall disturbance.

4 shows a second variant proposal for a pressure control loop with observer structure. This model differs from the embodiment of Fig. 3 essentially only in that the total disturbance is not the output of the element 29.11 but the sum of the outputs of the elements 29.10 and 29.11 is used as Gesamtstörgröße QB.

Fig. 5 shows another alternative possibility of the observer structure. Essentially, Figure 5 differs from the previous embodiment in that the actuator 21 is divided into a current regulator 21a and the actual actuator 21b. The signal dQ _{HDP, target} / dt is the Regulator 21a fed as a setpoint, the current IMPROP actually flowing through the solenoid valve 7 is fed back as an actual value to the regulator 21a from the actuator 21b. Based on the comparison between the actual value and the setpoint value for the current through the solenoid valve, the controller 21a forms a manipulated variable UMPROP for acting on the actuator 21b.

Since the lower the atomic number of an observer system the higher the quality of an observer system, a possibility is given below of how the observer's order can be reduced by feeding the track simulation with the last known system quantities. This is the manipulated variable of the current controller 32. Thus, the route simulation is reduced by the Ordungszahl the current controller.

Fig. 6 shows a further alternative of the observer structure. In this embodiment, the modifier model with the actual value UMPROP of the current controller 32 is applied. By doing so, the order of the observer can be further reduced.

With the observed disturbance dQ _{B / dt} , the following possibilities arise in the existing system:

It is possible to monitor the leakage of the high-pressure region of the injection system. Leakage in pipes or in the rail or by hanging nozzle needles on injection valves can lead to undesired escape of fuel into the environment or into the engine cylinder. The injection quantity dQ _K / dt is a known size in the system, as is the state of the pressure reduction valve in the engine control unit. In normal operation of the pressure regulator, the DAV (pressure reduction valve) is not activated, ie dQ _DAV / dt = 0. The sum of leakage quantity dQ _L / dt and control amount dQ _S / dt can therefore be determined as follows: $${\mathrm{dQ}}_{\mathrm{L}}/\mathrm{dt}+{\mathrm{DQ}}_{\mathrm{S}}/\mathrm{dt}=-{\mathrm{dQ}}_{\mathrm{B}}/\mathrm{dt}-{\mathrm{dQ}}_{\mathrm{I}}/\mathrm{dt}$$

Due to other operating conditions, such as the rail pressure P, the rotational speed N and the injection quantity dQ _K / dt, a block 34 specifies a maximum value b for the sum a of control and leakage quantity. The sum a of control and leakage quantity calculated by a connection point is then compared with this maximum value b in a comparator 35 according to FIG. Depending on the result of the comparison, an error is detected in block 36 if the sum of the leakage amount and the control amount exceeds the maximum value. Alternatively it can also be provided that errors are detected if the modeled disturbance dQ _B / dt exceeds a threshold value.

By means of this procedure, leakages in the system can be reliably and easily detected.

Furthermore, with the modeled disturbance dQ _B / dt an improvement of the dynamics of the pressure regulator is possible. For this purpose, this variable is taken into account in the control loop at a suitable location. This offers the advantage that the disturbance does not have to be compensated by the controller. The disturbance variable is compensated dynamically by the observer size.

A corresponding embodiment is shown in FIG. 8. There, the pressure regulator is shown as shown in Fig. 2 as a block diagram. Corresponding elements are designated by corresponding reference numerals. The individual disturbances are flat rate denoted by dQo / dt.

In the illustrated embodiment, the precontrol value dQ _VS / dt is specified as a function of the injection quantity dQ _K / dt. In addition, a connection size dQ _Auf / dt in another Joining point added to the output signal of the pressure regulator 20.

Since the injection quantity is already compensated by the pre-control value dQ _VS / dt, not the entire disturbance is switched, but only the difference between disturbance and injection quantity. The additional switch-on quantity dQ _Auf / dt therefore results in too $${\mathrm{dQ}}_{\mathrm{On}}/\mathrm{dt}=-{\mathrm{dQ}}_{\mathrm{B}}/\mathrm{dt}-{\mathrm{dQ}}_{\mathrm{I}}/\mathrm{dt}$$

Alternatively it can also be provided that the disturbance is taken into account in the formation of the pilot control quantity dQ _VS / dt.

Claims (9)

1. Pressure control method, in particular in a common rail system, having a quantity-controlled high-pressure pump and/or a pressure-reduction valve, wherein a pressure controller and/or an actuating element of a pressure-control circuit are simulated by means of a model, characterized in that the model supplies at least one signal which characterizes the interference variables of the pressure-control circuit.
2. Method according to Claim 1, characterized in that an injection quantity, a leakage quantity, a control quantity and a pressure-reduction quantity act as interference variables on the pressure-control circuit.
3. Method according to Claim 1, characterized in that the sum of the individual interference variables such as the injection quantity, leakage quantity, control quantity and/or quantity of pressure-reduction valve can be determined by means of the model and the leakage quantity can be determined from the knowledge of these variables.
4. Method according to one of the preceding claims, characterized in that the model includes at least one actuator model and a system model.
5. The method as claimed in one of Claims 1 to 4, characterized in that the actuator model simulates at least one actuator, and the system model simulates at least one pressure accumulator.
6. Method according to Claim 5, characterized in that the system model has an at least integrating behaviour.
7. Method according to one of Claims 1 to 6, characterized in that the interference variable can be used to detect a leakage.
8. Method according to one of Claims 1 to 7, characterized in that the interference variable can be used to form a superimposed variable for the control circuit.
9. Pressure control device, in particular in a common-rail system, having a quantity-controlled high-pressure pump and/or a pressure-reduction valve, wherein a pressure controller and/or an actuating element of a pressure control circuit are simulated as a model, characterized in: that the model supplies at least one signal which characterizes the interference variables of the pressure control circuit.

Images