DE4019161C2 - Multi-fuel machine control with fuel viscosity correction depending on the fuel composition - Google Patents

Description

The invention relates to a multi-fuel machine control and in particular a fuel control of the type in which propellants Substance with essentially constant pressure during impulses by opening a machine fuel injector while before certain pulse duration is injected. Any such impulse duration or pulse length has a fuel flow part and an injector opening portion in which there is no flow takes place. The length of the fuel flow portion is the amount of fuel delivered with each pulse, and this amount is essentially proportional to the length this section.

Such a machine can with a mixture of Brennba fuels such as petrol and methanol. It is known the injection pulse lengths of such machines depending on a fuel composition signal to adjust the variable volumetric heat stop balancing these fuels. Such fuels but also have different viscosities. The visco The quantity of the fuel mixture changes with the fuel substance composition as well as with the fuel temperature, and the changing viscosity changes the actual one Amount of fuel that is the same machine fuel injector delivers at a predetermined pressure. With one Calibration process of the fuel control due to a pre the viscosity of the fuel is the amount of fuel delivered every Im pulses may be faulty when the fuel is assembled tion changes, even if the fuel pulse calculation changes the un different volumetric heat content of the fuel mix charges. A known arrangement this type is described in US-PS 47 06 630.

The invention provides for the correction of the fuel pulse length depending on a fuel composition, about the different viscosity of the changing propellant  to balance the mixture of substances.

For this purpose a propellant according to the invention is distinguished material control from through the characteristic features of the An saying 1.

In particular, the invention provides fuel control for a machine of a motor vehicle with a fuel line, means for supplying a liquid fuel mixes first and second combustible fuels with un Different volumetric heat content and viscosity th through the fuel line with a predetermined Pressure, with a fuel composition sensor in the Fuel line based on a physical parameter the fuel mixture together with the production of a fuel setting signal reacts, which for the relative parts of the first and the second fuel in the fuel mixture is significant, and with a fuel injection device tion with a machine fuel injector in the fuel line and control means for opening the machine fuel injector during predetermined pulse lengths to the machine to deliver single portions of fuel. The blowing The fuel injection device determines the pulse lengths in dependence of the fuel composition signal for one predetermined amount of fuel, calibrated with flow characteristics machine fuel injector and viscosity risks of the first fuel, and is still dependent on that Fuel composition signal to change the prep agreed pulse lengths to compensate for the changing visco quantity of the fuel mixture so as to achieve the predetermined propellant To create quantities with the individual fuel quantities, too if the viscosity of the fuel mixture changes with its Composition changes.

In another version, the fuel control changes  further the pulse length to compensate for the changing Viscosity of the fuel mixture depending on the Fuel temperature sensor. If the pulse length increases an additional injector opening length connects in the no fuel flows, the compensation is only on the Im Pulse length portion applied, in which a fuel flow takes place.

It will be submitted on the same day with this Application related patent applications referenced with the file number P 4 019 187.7, P 4 019 188.5, P 4 019 189.3, P 4 019 159.1, P 4 019 160.5.

The present invention is based, for example, on the Drawing explained; in this shows:

Fig. 1, a vehicle with an inventive fuel control,

Fig. 2 is a schematic diagram of a controller for uses in the vehicle according to Fig. 1,

Fig. 3 to 6 are flow charts for the operation of the controller of FIG. 2,

Fig. 7 is a graphical representation of the vaporization rate in the fuel tank of a motor vehicle with a propellant, which is a mixture of gasoline and methanol, and

Fig. 8 is a flowchart of a modification of the fuel control and operation according to the previous figures.

According to Fig. 1, a motor vehicle 10 is provided with a combustion engine 11 provided in an engine compartment 12, the machine 11 receives fuel from a fuel tank at the other end of the motor vehicle via a fuel line 15 and excess fuel returned through a return line 14 to the fuel tank 13. The fuel line 15 includes a sensor 16 for the fuel composition, which is arranged in the engine compartment 12 at a location near the engine 11 . The sensor 16 for the fuel composition generates a for the relative proportions of alcohol and gasoline in the fuel flowing through the signal. Various such sensors are known, but the preferred sensor is a capacitive dielectric sensor which measures the dielectric constant of the fuel. Such a sensor is to be regarded as universal in that it gives an output signal for a mixture with some alcohol, such as ethanol, methanol, etc., as well as for various motor fuel additives. One sensor that can be used here is a capacitive dielectric constant-fuel composition sensor as described in U.S. Patent 4,915,084, filed November 7, 1988 by Eugene V. Gonze, and assigned to the assignee of this application. A normal fuel vapor collection container 17 is connected via a steam line 18 to the fuel tank 13 to receive steam therefrom, and has a further steam line 19 which leads to the Einführungssy system of the machine 11 .

The operation of the machine 11 is controlled by an (electronic) controller 20 , which can be arranged in the manner shown at the rear in the engine compartment, but also at other suitable locations. The controller 20 may be a programmed digital computer currently used for machine control in automobiles. This type of control is well known and includes a microprocessor, RAM and ROM memory and the associated input / output circuitry, with a corresponding program stored in the ROM to receive input information from various sensors, the execution of calculations and the call of values as well as coordinate output commands to various actuators of the machine-related components.

The controller 20 is shown in Fig. 2 with its various input / output connections to different machine-related components. The controller 20 receives motor vehicle battery voltage from a battery 50 at the BAT input and is connected to ground at the GND input. The controller 20 of course also includes a conventional power supply circuit (which is not shown here) in order to derive its own regulated operating voltage from the output voltage of the battery 50 of typically 9 to 16 V, which actually includes the entire power supply of the vehicle with the machine-driven three-phase generator and voltage regulator.

The controller 20 receives an IGN input signal from the ignition switch of the motor vehicle, which has a value when the ignition switch is closed and another when the ignition switch is open. It can receive a coolant temperature input signal TCOOL from a coolant temperature sensor 21 , a KNOCK input signal from a knock sensor 22 , an air mass flow input signal MAF from an air mass flow sensor 23 , an engine speed signal RPM from an engine speed sensor 24 , a vehicle speed signal VSS from the vehicle speedometer sensor 25 , an accelerator pedal position signal TPS from the accelerator pedal sensor 26 , an intake absolute pressure signal MAP from a corresponding sensor 27 , an intake air temperature signal MAT from an air temperature sensor 28 , an oxygen sensor input signal OXY from an exhaust gas composition sensor 29 , a fuel composition input signal ALC from the fuel composition sensor 16 and a fuel temperature input signal FTS from a fuel temperature sensor 30 , the latter included in the use of the fuel composition sensor 16 .

Next, the controller 20 of FIG. 2 has an output INJ1 for simultaneous control of the engine fuel in injectors 31 to 33 and an output INJ2 for simultaneously controlling the engine fuel injectors 34 to 36. Furthermore, a container drain control signal output CCP is present, which controls the duty ratio of a CCP magnet 37 in the fuel vapor collecting container 17 . The controller 20 also has an EGR output (exhaust gas recirculation output) to an EGR magnet 38 of an EGR valve 40 and an input for a valve position return signal PINT, which is dependent on a valve position-dependent potentiometer 41 in the EGR valve 40 is coming. The potentiometer 41 reacting to the valve position receives a constant 5 V reference voltage against ground. The controller 20 also outputs a fuel pump relay control signal FPRD to a loading coil 43 of a fuel pump relay for a fuel pump 45 in the fuel tank 13 and contains a fuel pump input signal PPSW from a machine oil pressure switch 46 , through which the armature 47 of the fuel pump relay is connected to the battery 50 is when the fuel pump relay is not energized.

The fuel for the engine 11 is passed from the fuel tank 13 of the fuel pump 45 through the fuel line 15 to a normal standard fuel injection device for the engine 11 , which contains the engine fuel injectors 31 to 36 . The fuel pump 45 may include a pressure control device to supply the machine with a constant pressure fuel 11 , with excess fuel being returned to the fuel tank 13 . Alternatively, some implementations, particularly those described with reference to FIG. 8, may have a variable fuel pump pressure control in which an output signal FPS controls a variable voltage power supply to drive the fuel pump at a controlled speed and thus a controlled variable To create fuel pressure.

When the controller 20 outputs an injector opening signal at the INJ1 or INJ2 output, the respective engine fuel injectors are opened to inject fuel under the regulated pressure of the fuel pump 45 into the fuel injection ports of the engine 11 near the engine cylinder intake valves. The engine fuel injectors close when the injector opening signal stops, thereby ending the fuel delivery. The fuel is therefore injected into pulses with a controlled length of time, during which there is a constant nominal flow, and it can therefore be assumed that the supplied fuel depends on the pulse length. In the particular machine 11 shown , all the machine fuel injectors 31 to 36 are normally acted on simultaneously once per crankshaft revolution, with each machine fuel injector delivering half of its total calculated fuel quantity (during half of the total pulse length calculated) per machine cycle with each application . The illustrated machine 11 is a six-cylinder machine with one machine fuel injector per cylinder. Since two crankshaft revolutions are required for a complete cycle in the six-cylinder engine, each machine fuel injector 31 to 36 normally delivers the full calculated amount of fuel for each cycle in two crankshaft revolutions.

Air is admitted to the cylinders of engine 11 through a standard air filter in the same standard intake device, with air flow controlled by a throttle valve and sensed by MAF sensor 23 , while air temperature is sensed by MAT sensor 28 . The throttle position sensor 26 detects the position of the throttle valve just described, and the MAP sensor 27 detects the pressure in the inlet device in the flow direction behind the throttle valve. The output signal IAC can be used to control an idle air flow device 48 , either by changing the position of a throttle stop or by moving a valve in an idle air redirection, as is known to those skilled in the art. The machine 11 is further equipped with an ignition spark system of normal construction and operation as far as this description is concerned, and this will not be shown and described further.

In general, the supply of fuel to engine 11 is affected by the presence of alcohol in the fuel in two ways. The first is the different volume heat content and therefore different stoichiometric air / fuel ratio of the different fuels such as methanol and gasoline. The engine 11 is designed to operate normally with a stoichiometric air / fuel ratio of 14.6 for optimal combustion of gasoline in accordance with a three-way catalyst converter and a fuel control with an oxygen sensor in the exhaust; the corresponding stoichiometric air / fuel ratio for methanol is about 6.5. Therefore, the injector pulse length calculation should normally be modified for the changing stoichiometric air / fuel ratio of the fuel supplied to the engine 11 , as denoted by a fuel composition factor ALC% derived from the sensor output signal ALC, later in this Description given. As a result, the different volumetric heat content of the two fuels is balanced, as is generally known to the person skilled in the art.

The second way in which alcohol influences the fuel supply to machine 11 is the changed viscosity, for example with different additions of methanol to gasoline, and the change in the viscosity of the fuel mixture with temperature. In general, constant viscosity is assumed in the normal calculation of the fuel pulse length. However, since the viscosity affects the fuel flow rate through the engine fuel injectors 31 through 36 , the total amount of fuel supplied at a given pulse length changes. The fuel pulse length calculated in each case should also be tracked with a viscosity factor which is a function of the fuel composition factor ALC% and also a function of the fuel temperature FTS. This viscosity factor is multiplied by the total injector pulse length except for the portion that represents the correction quantity for the injector opening.

For example, the normal fuel pulse length increases Start NCRANKPW for pure gasoline in the following form:

NCRANKPW = BCPW + INJCORR,

where a calculated basic starting pulse length BCPW with a Injection correction length INJCORR is corrected. The reason Duration of the starting pulse can be according to any algorithm are calculated according to the prior art, but contains generally at least one factor from the coolant temperature TCOOL depends. It is calibrated based on the knew injector flow properties and viscosity of gasoline at a predetermined fuel temperature to the required at a predetermined fuel pressure To create the amount of fuel by the regulated propellant  material pump pressure is determined. INJCORR is added to to consider the injector opening time. That term is not related to the fuel since it is the equivalent represents time that the machine fuel injector opens needs before the fuel flow begins, and des half this term is a function of mechanical and electrical trical properties of the injector. He can depend of the electrical supply voltage for the Ma engine fuel injectors will change, however changing fuel composition not corrected.

The modification of the normal starting pulse length equation to a similar equation for a multi-component starting pulse length MCRANKPW contains two steps. The first step is the modification of the base pulse length to a more MBCPW basic drain pulse length, which depends on the fuel composition depends. A convenient way to change this the 2D value table of the Coolant temperature factor in a 3D table of values for the Coolant temperature TCOOL and the fuel composition ALC% expand. The second step is one Viscosity factor VISC for the modified basic pulse length to accomplish:

MCRANKPW = (MBCPW) (VISC) + INJCORR.

The viscosity multiplier VISC itself is made by one 3D table of values as a function of fuel composition factor ALC% and the fuel temperature FTS derived and corrects the viscosity changes when Ben changes Zin / alcohol shares in the fuel and the viscosity changes tion of an alcohol-containing mixture with the temperature, so that the calculated fuel pulse length is the correct fuel amount of substance results. As the fuel pulse length changes as a result the change in viscosity of the fuel mixture is not necessary  usually changes in the same way as the fuel pulse length due to changes in the stoichiometric A / F Ratio (A / F = air / fuel = air / fuel) with the Fuel composition, and since the fuel temperature The table of values only changes with the viscosity two corrections do not appear in a single table of values trimmed. There will be no injection correction length INJCORR Fuel composition correction applied as this is not affected by fuel properties. To when the engine starts, the fuel algorithm goes gradually from the starting fuel equation to normal fuel equation over which is a normal propellant Pulse injector pulse length BPINJ results in the following way:

BPINJ = [(BPW) ((BLM) (DE) + AE)] (VISC) + CORRCL + INJCORR.

The equation above is

BPW the calculated basic pulse length,
BLM a block learning multiplier,
DE a delay lean multiplier,
AE an acceleration enrichment term,
CORRCL a rule correction term,
INJCORR the injector corrector term and
VISC the fuel viscosity multiplier.

It can be seen that the normal fuel equation for loading consideration of different fuel viscosity in the we is substantially corrected in the same way as in the starting fuel equation was the case: i.e. the main one part of the pulse length controlling the amount of fuel is included multiplied the viscosity correction factor VISC by a table of values according to the fuel composition ALC% and the fuel temperature FTS is derived, while the injector correction term is not affected. If desired, the control loop correction term can also  compensated for viscosity in the following manner become:

BPINJ = [(BPW) ((BLM) (DE) + AE) + CORRCL] (VISC) + INJCORR.

In this case, the total amount of fuel ge controlling pulse length multiplied by the viscosity correction factor. However, it is generally not necessary dig to do this because the rule correction factor is an increase size is what is actually required to approximate that air / fuel ratio is added and none amount of fuel calculated with open control loop, and so much less affected by the alcohol content.

The basic pulse length term BPW can be derived from the Air mass flow rate, engine speed, the desired Fuel / air ratio and injector flow rate. The air mass flow rate in g air per s and the inverse Engine speed in computer clock pulses per cylinder combined and multiplied by a constant to one Load variables LV8. The basic pulse length is then determined by fol given equation:

BPW = K₁ (LV8) (INJ) (F / A),

where K _{1 is} the scale constant, LV8 is the load factor just defined, INJ is the injector flow rate and F / A is the desired fuel / air ratio, ie the scaled inverse of the air / fuel ratio. When calculating BPW, LV8 is calculated from engine operating parameters as in the state of the art, regardless of the fuel composition. The desired fuel / air ratio F / A is calculated as a straight-line interpolation between the desired ratio values for gasoline and methanol (or the fuel otherwise used) on the basis of the detected alcohol content ALC%. The injector flow rate is a constant term that is calibrated for the particular machine and depends on the fuel viscosity for gasoline and injector properties. Those skilled in machine control will be aware that there are other basic engine fuel control algorithms, such as engine speed and engine load factors such as absolute pressure or vacuum in the intake manifold. The respective method for determining the basic pulse length is not essential for the invention described here. The block learning multiplier BLM is an adaptive control term, which is stored in the memory as a function of the machine operating state and is used to transmit a large part of the control through an adaptive learning process of the control and thus to reduce the required control correction. The operation of such adaptive controls is well known in the art and is not modified by the fuel composition except for the one listed below.

The delay lean multiplier DE is used to reduce the fuel supply during delays due to the throttle position and / or other corresponding parameters. The multiplier can have a nonzero value to reduce fuel, or can be set to zero to stop fueling. The non-zero value is adjusted with a factor of a value table based on the fuel composition ALC% and the coolant temperature. For the machine 11 described , the value DE decreases with temperature in the case of pure gasoline and increases with the temperature in the case of pure methanol. Fuel compositions between these extreme values produce a mixture of these curves. If the term DE becomes zero, no fuel is delivered. However, asynchronous initiation pulses can be delivered at the end of the fuel shutdown time and are set up based on ALC%. The acceleration enrichment term AE is used to create additional fuel due to a positive change in the load parameter LV8 during acceleration processes. AE is modified based on the fuel composition and possibly the coolant temperature from a table of values. It should also be mentioned that the fuel is subjected to a choke function which reacts to the coolant temperature, as a result of which extra fuel is supplied when the machine is warmed up. Such a function can use a multiplier that decreases with increasing coolant temperature until it becomes and remains essentially one at a certain coolant temperature. Both the value of this multiplier and the drop rate can be functions of ALC% as well as the coolant temperature.

The rule correction term CORRCL is added when the rule is released. It includes integral and proportional terms. CORRCL is derived from the rich / lean status of the fuel mixture, as it is detected by the oxygen sensor in the exhaust system of the engine 11 . A standard zirconia oxygen sensor generates a voltage that changes depending on the excess free oxygen detected in the exhaust gas, and this is determined by the fuel's lean / lean status with respect to the stoichiometry. The output signal of the oxygen sensor is a voltage that changes rapidly in a narrow range around the stoichiometric value. This voltage, or a number derived therefrom, for use in a computer can be compared to a reference value that has a predetermined relationship to the stoichiometric value as part of a method of generating a signal indicative of the rich / lean status of the actual air / fuel. Ratio of the machine is significant.

It should be expected that the output of the oxygen sensor will not be significantly affected by changing fuel compositions since the oxygen sensor will respond to factors related to the stoichiometric value rather than the absolute air / fuel ratio. However, it has been shown that the operation of such oxygen sensors is affected by the presence of methanol in the fuel, thereby displacing the output voltage of the oxygen sensor to the rich side, and thus driving the engine fuel system to the lean side when controlled. The change is small, on the order of 0.1 to 0.2 A / F ratio values, but shifting this size toward the lean side of the stoichiometric ratio has a significant impact on NO _x conversion through a reducing catalyst . Therefore, the system is corrected by moving the reference voltage or voltages with which the output signal of the oxygen sensor is compared, in the same direction and in the same size as the displacement of the sensor output voltage, so that the oxygen sensor a ma gerere reading results, which corrects the change by the methanol.

An embodiment of a control loop correction system stems based on an oxygen sensor is in US-PS 46 25 698 is shown. With this control loop correction system the actual sensor output voltage is processed for Formation of a quick filter value (ff) and a slow fil ter values (sf), and these filtered values are or sf voltage windows compared. Every ff and sf fil terwert is determined as fat or lean if it is outside the respective window on the fat or lean Page, or if it is within the respective Window, depending on whether it is fat or lean  Changes direction. For adjustment in multi-fuel operation are the ff and sf windows in both US-PS 46 25 698 such a direction that the algorithm for is shifted towards lean reading, in a size that balances the effect of alcohol in the fuel, so that a consistent relationship between the reference people ten and the stoichiometric ratio is maintained. The shift depends on the fuel composition ALC% and the machine load nonlinearly and is thus of derived from a table of values. Any normal move Exercise of the windows with machine load is also increased, so that the value table is a 3D table depending on the Fuel composition ALC% and a machine load factor is. The result affects not only the determination of whether the fuel is recorded as rich or lean, but also the size of the error terms used to calculate the proportion channel correction factor can be used as in the mentioned US-PS is described.

The output signal ALC of the fuel composition sensor 16 is processed by a control correction system to increase the accuracy and stability of the fuel control. When an initial engine runtime IERT has expired after the engine is started, the output of the fuel composition sensor 16 is read on a regular basis, such as a 100 ms loop, along with the output of the fuel temperature sensor 30 and with analog / digital converted Compare values stored in RAM in controller 20 . A flow chart of the process is shown in FIG. 3. In an initial decision block 60 , the fuel pump runtime is compared to a reference initial engine runtime IERT, which can be done by comparing the contents of a memory byte used as a counter with another stored reference IERT. The fuel pump runtime can be counted in the controller from the first generation of the FPRD signal, which is output to act on the fuel pump relay of the fuel pump 45 . The fuel pump runtime is used as an indication of the actual initial engine runtime from the time engine start is started. The reasons for the IERT reference and its calibration are discussed below. If the reference time IERT has not yet expired, a further decision block 61 determines whether the fuel pump 45 is running. If not, for example if machine 11 has not yet started, the routine is exited. If so, the fuel pump runtime memory location is increased in step 72 before the routine is executed. In decision block 60, if the fuel pump runtime is determined to be greater than or equal to the reference IERT, the A / D converted fuel composition input signal ALC and the fuel temperature signal FTS are read in step 63 . The input signal ALC% is derived from ALC by converting ALC into a mathematical form, limiting its value within the upper and lower limits and filtering the result in step 65 in a low-pass filter routine of the first order. The supply voltage BAT can be checked in a decision block 66 with a reference voltage to see whether it is sufficient to give a good ALC signal, the derived ALC% value being stored only if the BAT is sufficient.

However, during the initial machine run time up to the reference time IERT, the ALC% value used by the controller 20 is a value that was held in a non-volatile memory location from the last machine operating time. The stored value is used for various reasons. First, the gasoline and methanol components of the fuel in the fuel tank and lines may have separated, even in the fuel composition sensor 16 itself, while the vehicle was at rest. Thus the last Able solution of the fuel composition sensor 16 before from the machine turn a more accurate reading of ALC% than the initial reading of the fuel composition sensor 16 before the fuel is mixed again. In general, several seconds should be allowed until the fuel composition sensor 16 is used again. In addition, the on-board electrical system voltage of the motor vehicle can fluctuate greatly during starting, and this can interfere with the precise operation of the fuel composition sensor 16 in some arrangements. This process can take longer, especially in cold weather, when the engine 11 cannot be started easily and the vehicle electrical system voltage drops further. An additional factor to be considered is the fact that the engine 11 is normally stopped when new fuel is added to the motor vehicle 10 , and refueling can significantly change the composition of the fuel in the fuel tank 13 . It is therefore definitely desirable to detect this change in time to re-enact the engine fuel control when the new fuel reaches engine 11 . However, there is still a time delay of 10 to 15 s before the new fuel is pumped from the fuel tank 13 to the fuel composition sensor 16 and can be detected there. It is not so necessary to try to detect a major change in fuel during this time, and during this time the fuel system receives fuel with the old composition. Therefore, using the stored ALC% during the time it takes for the new fuel to get from the fuel tank 13 to the fuel composition sensor 16 is a good fuel composition signal during the cranking process, and the reference time IERT can be equal to a magnitude essentially 10 to 15 s set and the value kept constant, with the assumption that a fuel pump 45 is used, which is regulated to a substantially constant pressure or constant flow. If the fuel pump pressure or flow changes, the value of IERT described in the previous sentence should be set according to the fastest possible flow or kept variable depending on any flow display parameter, such as the voltage applied to the fuel pump motor. Before this time has elapsed, the fuel separation in the fuel line and at the fuel composition sensor 16 is then ended, and in most cases the starting process is over. After the reference time IERT the fuel composition sensor 16 is read normally, so that the system notices a significant change in fuel due to refueling as soon as the new fuel reaches the fuel composition sensor. However, if a certain vehicle device is known to have difficulties with low vehicle electrical system voltages in the precise determination of the fuel composition during starting in cold weather, the supply voltage according to FIG. 3 can also be monitored to support the reference runtime IERT by this length of time without changing the ALC% In this case, extend the value until a good reading of the fuel composition sensor is ensured.

The desired air / fuel ratio A / F (or its reciprocal F / A) is not always adjusted depending on the value ALC% for use in calculating the fuel pulse length. While it is important to respond to any change in fuel when a tank is filled, and it is a good idea to continue capturing fuel composition during transition conditions, which include fuel transfers, there are other engine operating modes that are very stable in Regarding the fuel control. Such operating conditions generally include the closed-loop fuel control, in which a quantity of fuel is first calculated, but the calculation is adjusted due to the display of the exhaust gas composition sensor 29 , which provides an air / fuel feedback signal. As a result of practical cost restrictions, the A / D conversion and the computer device have only limited resolution, and a very small change in composition detected by the fuel composition sensor 16 , which is caused, for example, by an air bubble in the fuel line 16 or by electrical noise may be a change of, for example, 0.7 in the size of the air / fuel ratio. The value of A / F (or F / A) is added at the end of so that the integrators of the control do not have to repeatedly run back from such small insignificant changes, and thus to promote the stability of the fuel control when the machine is assumed to be stable previous operating state locks to always be used during the stable state. It remains locked until the stable operating state ends or until a change in the fuel composition, as shown by the fuel composition sensor 16 , exceeds a predetermined size, for example a 7% change in ALC%. The previous or first operation takes a long enough time so that a new fuel mixture in the fuel tank 13 due to refueling has a chance to reach the fuel composition sensor 16 . It has been shown that this process gives good results since the changes in the fuel composition during stable operation are very small. Such operation is described below.

The machine control of this version contains how many according to the prior art, an adaptive learning process for the fuel control, in which a block of storage space zen for correction factors used in machine re gel operation can be improved and included in the tax portion  the calculation of the fuel pulse length. It will in the block learning multiplier BLM of the normal injection pulse length equation taken into account. In practice, the BLM factor from a table of values based on the read Values of machine speed and a load factor like that Flow read out when calculating the injection pulse length becomes. However, a saved BLM factor can only be changed become when "block learning" is released and that occurs only on when certain machine operating states exist are there that be a balance machine operation to draw. Typical specific conditions, all during a calibrated length of time must be available Operation of the oxygen sensor, coolant temperature inside semi-calibrated limits, A / F ratio equal to one calibrated value, machine load factor at least equal to one value and no signs of overtemperature of the catalytic converter goal converter. If this or similar equivalence conditions conditions during the calibration period, it is true apparently that the machine operating conditions are stable Machine operation result, and the BLM multiplier for that existing machine speed and load can be due to the off signal from the oxygen sensor can be corrected. This can be done, for example that the state of the rule integrator, as in US-PS 46 25 698 with the current output signal of the Sauer substance sensor is compared, the BLM multiplier is increased accordingly in the rich or lean direction.

As described above, it is not desirable to introduce control operation according to a fuel control parameter derived from a fuel composition signal that changes for reasons that are insignificant from a fuel control standpoint. Therefore, the value of the desired A / F (or F / A) ratio is locked against changes in ALC% while learning control is enabled. This is a modification of the basic fuel injection pulse length equation described above, in which the term desired A / F ratio depended on the ALC% value. The process is shown in the flow diagram of FIG. 4, in which it is determined in decision block 70 when learning is enabled. If not, the desired A / F (F / A) ratio is released in step 72 and the routine is exited; when this happens, a bit is set to step 71 to lock the value of the fuel control parameter A / F (F / A). The learning control is locked and the value of A / F (F / A) is released if a malfunction of the oxygen sensor is detected, the machine 11 is switched off or a change in fuel is detected.

The value of ALC% can be temporarily locked in the event of a fuel change: ie if a significant change in the ALC% value is detected. This could be considered unnecessary if the fuel composition sensor 16 were incorporated directly into the fuel inlet device, but generally there will be a fuel run time between the sensor 16 and the fuel introduction device. If the fuel composition changes slowly and gradually, the runtime can be ignored, but if there is a sudden change, it is desirable to keep the old ALC% during the time that the new fuel composition is in transition from the sensor 16 to the insertion device, since during this time the machine 11 still receives the composition existing before the change. The routine that accomplishes this is included in the flow chart of Figure 5, which describes the full fuel transfer logic.

The routine first checks a fuel transition bit in decision block 75 . If it is not set, then there is no fuel transition and the routine next checks for the occurrence of a new transition in decision block 76 by comparing the last received value of ALC% (NEWALC%) with a stored previous value of ALC% (OLDALC% ). If the absolute value of the difference exceeds, for example, 7% of OLDALC%, then a fuel transition is considered to be detected and the fuel transition bit is set in step 77 , the desired A / F ratio is locked to its most recent value in step 78 and a fuel transfer Timer byte in RAM is cleared in step 79 . If no fuel transition has been detected in decision block 76 , the routine is exited.

From step 79 or decision block 75 if the fuel transfer bit was already set, the routine continues to decision block 80 where the fuel transfer timer is checked for the lapse of a predetermined fuel transfer time that corresponds to the time the fuel from the fuel composition Sensor 16 for the fuel introduction device of the machine 11 needs. In this embodiment, the timer is a memory byte in RAM, which is increased each time the routine is run during the fuel transition. If the fuel transition time has not expired, the timer byte is increased in step 81 and the routine is exited. If the time has expired, however, the A / F ratio is readjusted in step 82 to a new value which corresponds to the new value of ALC%, the fuel transition bit is set again in step 83 and OLDALC% is stored in memory by NEWALC% replaced in step 84 before exiting the routine. If the machine 11 is operating with the adaptive learning control switched on and the value A / F (F / A) is locked, the detection of a fuel transition results in an adjustment of A / F (F (A) based on the fuel composition signal or an end of the adaptive one Learning control process or locking A / F. This can occur as soon as the fuel transfer is detected.

An important additional part of the fuel control for the machine 11 is the discharge control for the fuel tank 17 , which is modified from the standard production control according to FIG. 6. As already described, evaporated fuel from the fuel tank 13 is stored in the fuel vapor storage container 17 , to be delivered to machine 11 in controlled quantities at controlled times. The fuel vapor collecting container 17 is a normal container of the type known from the prior art, with activated carbon or a similar substance which absorbs or adsorbs hydrocarbons. Normal amounts of vaporized fuel are captured by the activated carbon or other substances until the time when the control for the machine 11 outputs a signal CCP to the CCP magnetic coil 37 of the container control valve for the fuel vapor collection container 17 . The CCP signal is a pulse-length-modulated signal, which allows the container control valve to reach an average open position due to the switch-on ratio of the signal. The valve throttles the flow of fuel and / or air from the fuel vapor reservoir 17 through the steam line 19 to the fuel introduction device of the engine 11 . The container control valve thus controls a container discharge flow of an additional air / fuel mixture to the engine 11 . The air / fuel ratio of this mixture is uncontrolled, so that the rate at which it is added to the overall mixture must be kept sufficiently low so that the control of the exhaust gas composition sensor 29 is not disturbed. When a tank drain is released, the CCP value for pure gasoline is derived from a value table based on a measured machine air flow or a similar factor such as intake manifold absolute pressure or negative pressure. Various container drain algorithms are known in the prior art for gasoline powered machines.

Using a different fuel, such as methanol, may result in greater or lesser fuel evaporation that changes with both the fuel composition and temperature. Pure methanol is considerably less volatile than gasoline at the same temperature, but mixtures of methanol and gasoline in some compositions can be more volatile than either of the two per se. A typical behavior of the evaporation over the fuel composition at a certain fuel temperature is shown in FIG. 7. The volatility or evaporation rate increases from pure gasoline and pure alcohol to a maximum with a mixture somewhere in between. Similar curves at higher or lower fuel temperatures generally have similar shapes, but are shifted upwards as the fuel temperature rises. Since a multi-fuel machine has to be designed to work over a whole range of such mixtures, the container control has to be modified in order to be ready for different fuel evaporation rates. It can also be seen that some motor vehicles have to be equipped with a larger container in order to cope with the expected increased amount of fuel vapor.

The normal container drainage control derives a CCP value as a function of, for example, the air mass flow MAF in step 90 of FIG. 6. In order to adapt the container draining algorithms to the multi-fuel operation, the fuel temperature FTS is determined in a decision block 91 with a reference value TREF1, e.g. B. 17 ° C compared. If it exceeds this reference value, a CCP multiplier CCPMULT is derived in step 92 from a value table based on the fuel temperature FTS and the fuel composition ALC%. Also in step 92 , the switch-on ratio to be output is determined CCP% = (CCP) (CCPMULT). The CCP multiplier CCPMULT can be from 0 to 4, so that CCP values larger or smaller than the normal values for petrol are possible.

According to Fig. 7 shows curve 95 a typical change of the steam production rate at a predetermined rich Treibstofftemperaturbe on the fuel composition. Pure gasoline is shown as point 93 at the left end of the curve and pure methanol as point 94 at the far right end of the curve. It can be seen that the vapor formation in the lower range of the methanol concentration is higher than that of pure gasoline, but also than that of pure methanol and reaches a maximum at about 20% of methanol. The CCP multiplier will therefore rise in the same way for mixtures of gasoline and methanol. Thus, the container drainage rate is optimized between the other interfering targets to remove fuel vapor from the fuel vapor storage container 17 while it evaporates from the fuel tank 13 , and to change the air / fuel ratio of the engine as little as possible. In addition, the rate of vapor formation tends to increase as the vapor temperature increases, and this in itself can be disruptive, particularly for motor vehicles with fuel injection systems that return unused fuel, with heat absorbed in the engine compartment 12 being returned to the fuel tank 13 . This allows the CCP multiplier to increase at least during a predetermined fuel temperature range with increasing fuel temperature.

With some machines, a non-volatile permanent memory can be used to hold some values learned in adaptive learning control. If the vapor formation rate is high, the fuel control can be pushed towards a lean mixed value in order to compensate for the additional fuel vapor from the fuel vapor collecting container 17 . If the engine 11 is stopped and the motor vehicle 10 is parked overnight, for example, the fuel temperature will be lower the next time the engine is started, and the values learned from the previous period with higher fuel temperatures no longer apply. The fuel temperature FTS is thus compared with a reference value TRF2 in decision block 96 of FIG. 6. If this is higher, the long-term memory part is blocked in step 97 , so that the non-applicable values are no longer maintained after the new machine operation.

An alternative to readjusting the fuel pulse length for the different desired A / F ratio values is an ALC% dependent fuel pump pressure control. A fuel pump driven by an electric motor, which with a regulated voltage as described above Keeping the speed constant and thus keeping it constant of the discharge pressure can be supplied by a controlled Deviation of the applied voltage, he changes the speed hold. This allows the fuel pump pressure to be controlled motor control can be changed controllably. The above be wrote equations for the fuel pulse lengths are then essentially in the normal way for pure gasoline calculated up to the viscosity correction, if this correction does not also apply to the fuel pulse pressure tax is included, and the changing volumetric heat The content of the changing fuel mixture is released compared to a change in fuel pressure, whereby the amount of fuel supplied with the same pulse duration will be changed.

The process of the fuel pump pressure control is shown in the flow chart of FIG. 8, in which the fuel composition ALC% is derived in step 100 in the manner described. Then, in step 101, the desired fuel pump pressure DPRES is derived from the pump pressure for pure gasoline GPRES according to the equation.

DPRES = (GPRES) [1 + (STK-1) (ALC%) / 100] 2.

In this equation, STK is the stoichiometric ratio of gasoline to the other fuel. For methanol it is 14.6 / 6.5 or about 2.25. Finally, in step 102 the desired pressure is output to a fuel pressure control circuit which changes the supply voltage for the fuel pump accordingly to generate the desired pressure. Alternatively, the desired fuel pump pressure can be derived from a value table based on the value ALC%.

If the procedure with fuel pump pressure is used, all terms in the previous fuel pulse lengths equations for pure gasoline can be used without the need agility, the value tables for another dependency expand from the fuel composition. The driver substance viscosity factor can be used in the calculation Det be due to the fuel composition and the Fuel temperature. However, if the fuel pump pressure The viscosity can be derived from a table of values correction also included in this table of values which will then also be based on the recorded value of the fuel temperature can react. The shift of oxygen reference value, the fuel transfer and the initial Fuel pump delays as well as the one described Container drain modification are all used, though may be desirable, the fuel transfer and the An Delay in catching the fuel pump run with the propellant material pump pressure to change the variable flow rate  through the system to balance with fuel pressure. The A / F lock in adaptive learning control is not necessary dig because the used value of the desired A / F ratio does not change with ALC%.

Claims (8)

1. Fuel control for a machine ( 1 ) of a motor vehicle ( 10 ) with a fuel line ( 15 ); Means ( 45 ) for supplying a liquid fuel mixture of first and second combustible fuels, which have different volumetric heat contents and viscosities, through the fuel line at a predetermined pressure; a fuel composition sensor ( 16 ) in the fuel line which, in response to a physical parameter of the fuel mixture, generates a fuel composition signal indicative of the relative proportions of the first and second combustible fuels, and a fuel injector with a machine fuel injector ( 31-36 ) in the fuel line and with control means ( 20 ) for opening the machine fuel injector during predetermined pulse durations in order to deliver pulse quantities of fuel to the machine, the fuel injection device, the pulse durations depending on the fuel composition signal for a predetermined fuel quantity calculated, calibrated for flow characteristics of the engine fuel injector and the viscosity of the first fuel, characterized in that the fuel injection device further responds to the fuel composition signal by at a change in the composition of the fuel mixture is accompanied by a change in the viscosity of the mixture of fuels, a change in the predetermined pulse durations to compensate for the changing viscosity of the mixture of fuels is carried out in order to deliver the predetermined amounts of fuel in the fuel pulse lengths. 2. Fuel control according to claim 1, characterized, that the predetermined pulse duration is a subsequent one Share of a total injector opening duration is that too contains an injector opening time during which the injector properties essentially no blowing material flows, and only the predetermined pulse duration one derived from the fuel composition signal th viscosity factor is multiplied. 3. Fuel control according to claim 1, characterized in that the predetermined pulse duration is a subsequent section of the total injector opening duration, which also includes an injector opening time during which essentially no fuel flows as a result of the injector characteristics, the fuel control further comprising a fuel control means ( 20 , 29 ) which causes the generation of a fuel control signal as a function of an exhaust gas composition sensor ( 29 ) which emits a signal indicative of a stoichiometric air / fuel ratio, the fuel control responding to the fuel control signal at least temporarily so that in the Total injector opening time, an additional rule correction period is included, the fuel control acting so that only the predetermined te pulse duration and not the injector opening time, during which due to the injector characteristics substantially no fuel f flows, or the additional control correction period is multiplied by a fuel viscosity factor derived from the fuel composition signal. 4. Fuel control according to claim 1, wherein the pre-determined pulse duration is a subsequent portion of a total injector opening duration, which also includes an injector opening time during which essentially no fuel flows due to the injector properties, characterized in that the fuel control further comprises a fuel Control means ( 20 , 29 ) which is effective for generating a fuel control correction signal as a function of a designating the exhaust gas composition sensor for a stoichiometric air / fuel ratio, the fuel control depending on the control correction signal at least temporarily in the total injector opening duration includes an additional control correction period, the fuel control acting so that the predetermined pulse duration and the additional control correction duration, but not the injector opening time, during which, due to the injector properties, essentially lic hen no fuel flows, are multiplied by a fuel viscosity factor derived from the fuel composition signal. 5. Fuel control according to one of claims 1 to 4, in which the viscosity of the fuel mixture also changes with the fuel temperature, characterized in that the fuel control also comprises a fuel temperature sensor ( 30 ) in the fuel line ( 15 ) and the fuel injection device ( 20 , 31-36 ) also responds to the fuel temperature sensor. 6. Fuel control according to claim 5, characterized net that only the predetermined pulse duration with one of a stored value table derived viscosity factor due to the detected fuel composition and fuel temperature is multiplied. 7. Fuel control according to claim 1, characterized net that the predetermined pulse durations sections of Ge Velvet injector opening times are BPINJ, derived from of the following equation: BPINJ = [(BPW) ((BLM) (DE) + AE)] (VISC) + CORRCL + INJCORR, where BPW is a calculated basic pulse duration, BLM on Block learning multiplier, DE a delay lean multi plikator, AE an acceleration enrichment term, CORRCL a rule correction term, INJCORR an injector correction turterm and VISC one of the fuel composition Signal derived fuel viscosity multiplier is. 8. Fuel control according to claim 1, characterized in net that the predetermined pulse durations sections of Ge Velvet injector opening times are BPINJ, derived from of the following equation: BPINJ = [(BPW) ((BLM) (DE) + AE) + CORRCL] (VISC) + INJCORR, where BPW is a calculated basic pulse duration, BLM on Block learning multiplier, DE a delay lean multi plikator, AE an acceleration enrichment term, CORRCL a rule correction term, INJCORR an injector correction turterm and VISC one of the fuel composition Signal derived fuel viscosity multiplier is.