Method for controlling the speed of an internal combustion engine

Abstract

A method for controlling the speed of an internal combustion engine and a speed control circuit for carrying out the method. For controlling, fuel energy is used as an output variable. The control units are calculated in accordance with a stationary proportion gain which is calculated proportionally to the fuel energy and inversely proportional to the engine speed.

Claims

1. A method for controlling speed of an internal combustion engine, comprising the steps of: injecting at least one fuel of one fuel type; providing a speed controller in a closed-loop speed control system, behavior of the speed controller being determined by controller parameters; generating a fuel energy of the at least one fuel as an output variable of the speed controller; and calculating components of the speed controller as a function of a stationary proportional coefficient, the stationary proportional coefficient being calculated proportionally to the fuel energy and inversely proportionally to the speed of the internal combustion engine, wherein an integrating component of the speed controller is calculated as a function of the stationary proportional coefficient.

2. The method according to claim 1, including calculating the components of the speed controller as a function of a dynamic proportional coefficient, wherein the dynamic proportional coefficient is also a function of a speed control deviation.

3. The method according to claim 1, wherein a proportionality factor of the stationary proportional coefficient is made up of two multipliers, wherein a first multiplier is a function of an application and has a value 2 for application as a ship, and a value 1 for application as a generator.

4. The method according to claim 3, wherein a second multiplier mirrors a loop gain, which is presettable by an operator, of an open-loop speed control system, and is independent of the application.

5. The method according to claim 2, wherein a proportional component of the speed controller is calculated as a function of the dynamic proportional coefficient.

6. The method according to claim 1, wherein a differential component of the speed controller is calculated as a function of the stationary proportional coefficient.

7. The method according to claim 6, including tracking a rate-action time linearly via the fuel energy for calculating the differential component.

8. The method according to claim 1, including adding a fuel energy load signal to an output signal of the speed controller for improving dynamic response of the speed controller.

9. The method according to claim 8, including calculating the fuel energy load signal from a system signal that is generated when load switching occurs.

10. The method according to claim 1, including injecting multiple fuels of different kinds into a cylinder in each case and combusting the fuels in a combustion operation.

11. A closed-loop speed control system for carrying out the method according to claim 1, comprising: a speed controller that generates a fuel energy as an output variable.

Description

(1) The present invention is schematically depicted in the drawing based on exemplary embodiments and is described in greater detail below with reference to the drawing.

BRIEF DESCRIPTION OF THE DRAWING

(2) FIG. 1 shows one embodiment of a closed-loop speed control system for carrying out the provided method.

(3) FIG. 2 shows a discrete algorithm of a PI(DT.sub.1) speed controller.

(4) FIG. 3 shows the calculation of a dynamic proportional coefficient.

(5) FIG. 4 shows the calculation of a static proportional coefficient.

(6) FIG. 5 shows the calculation of a rate-action time tv.

(7) FIG. 6 shows the calculation of a load signal.

DETAILED DESCRIPTION OF THE INVENTION

(8) FIG. 1 depicts a closed-loop speed control system in a block diagram, which is collectively designated by the reference numeral 10. This closed-loop speed control system 10 functions based on fuel energy. The depiction shows a controller 12, in this case a PI(DT.sub.1) controller, a block 14 for limiting fuel energy, a filter 16, a speed filter 18, an engine management 20, and an internal combustion engine 22. Instead of a PI(DT.sub.1) controller, a PI controller, PID controller, or (PID) T.sub.1 controller may in particular also be used.

(9) The input signal of the closed-loop speed control system 10 is the set speed 30. The difference between this set speed 30 and the measured engine speed 32 constitutes the speed control deviation 34. The speed control deviation 34 is the input variable of the PI(DT.sub.1) speed controller 12. The output variable of the PI(DT.sub.1) speed controller is the PI(DT.sub.1) fuel energy 36, which is related to an injection for a combustion operation of a cylinder of the internal combustion engine. The load signal-fuel energy 38 is added to the output variable 36 of the speed controller 12. This addition constitutes a disturbance-variable compensation. It is used to improve the dynamic response of the speed controller 12. The sum of the speed controller output 36 and the load signal-fuel energy 38 is subsequently limited upward to the maximum fuel energy 40 and downward to the negative fuel frictional energy 42 per cylinder via the block 14.

(10) In this case, the maximum fuel energy 40 is a function of the engine speed, the charge air pressure, and other variables. The limited fuel energy 44 constitutes the correcting variable of the closed-loop speed control system and is likewise related to an injection. The fuel frictional energy 46 is subsequently added to the limited fuel energy. Fuel frictional energy 46 may be understood to be the fuel energy which corresponds to the frictional losses of the internal combustion engine. In this case, frictional losses are inter alia frictional losses in the cylinders of the internal combustion engine. The sum of the required fuel energy is finally transferred to the engine management 20 and is converted by it into the injection quantity. In a diesel injection system, this is the injection quantity 48, and in an injection system with diesel and gasoline injection (dual-fuel injection), it is also the gasoline injection quantity 50. The engine speed 52 is detected and filtered with the aid of the speed filter 18. The output variable of the speed filter 18 is the measured speed 32.

(11) FIG. 2 shows the time-discrete algorithm of the PI(DT.sub.1) speed controller, which is collectively designated by the reference numeral 12. The output variable 36 of the speed controller algorithm is a sum of three components: the proportional component 74, the integrating component 76, and the DT.sub.1 component 78. In this case, the proportional component 74 constitutes the product of the speed control deviation 34 and the so-called dynamic proportional coefficient 80. The dynamic proportional coefficient 80 is a controller parameter of the speed controller algorithm; the calculation of this parameter is depicted in detail in FIG. 3.

(12) The integrating component 76 of the speed controller, the I component, constitutes the sum of an instantaneous limited integrating component delayed by one scanning step (delay element 82) and the product of the gain 84 and the sum of the instantaneous speed control deviation 34 delayed by one scanning step (delay element 86). In this case, the integrating component of the speed controller is limited upward to the maximum fuel energy 40 and downward to the negative fuel frictional energy 42.

(13) The calculation of the DT.sub.1 component 78 is depicted in the lower portion of FIG. 2. The DT.sub.1 component 78 is obtained as the sum of two products. The first product 92 results from the multiplication of the factor 94 by the DT.sub.1 component 78 delayed by one scanning step (delay element 96). The second product 98 is obtained from the multiplication of the factor 100 by the output of a switch 102. Depending on the position of the switch 102, the factor 100 is either multiplied by the difference between the instantaneous speed control deviation 34 and the speed control deviation delayed by a scanning step (delay element 86), (switch position 1), or by the difference between the measured engine speed delayed by one scanning step (delay element 108) and the instantaneously measured engine speed 32 (switch position 2).

(14) In this case, the switch position 2 is always favored if the engine set speed 30 does not change, or changes only slightly, as is the case, for example, in generator applications. The gains 84 and 100 of the I component or the DT.sub.1 component are functions of the so-called stationary proportional coefficient kpStat, while the proportional component is a function of the dynamic proportional coefficient 80. The calculation of the stationary proportional coefficient kpStat takes place according to:
kpStat=(f*v*E.sup.I.sub.target)/n.sub.actual

(15) Here, the measured engine speed n.sub.actual is indicated by the reference numeral 32, and the integrating component E.sup.I.sub.target is indicated by the reference numeral 76. The stationary proportional coefficient is thus proportional to the integrating component E.sup.I.sub.target and inversely proportional to the measured engine speed n.sub.actual. The proportionality factor is the product of two multipliers. The first multiplier is the factor f, and the second multiplier is the closed-loop gain v.

(16) The factor f is a function of the application. In the ship application, f assumes the value 2, and in the generator application, it assumes the value 1. The closed-loop gain v may be preset by the operator; in this case, it is the dimensionless closed-loop gain of the open-loop speed control system. If v assumes large values, the dynamic response of the closed-loop speed control system is large; on the other hand, if v assumes small values, the dynamic response of the closed-loop speed control system is small. The stationary proportional coefficient kpStat is limited downward to the presettable minimal proportional coefficient kpmin:
kpStatkpmin

(17) FIG. 3 depicts the calculation of the dynamic proportional coefficient 80. The dynamic proportional coefficient 80 is calculated additively from the stationary proportional coefficient kpStat 152 and a component 154 which is a function of the speed control deviation 34. It is then activated if the switch 156 assumes the position 1. On the other hand, if the switch 156 assumes the position 0, the dynamic proportional coefficient 80 is identical to the stationary proportional coefficient kpStat 152.

(18) The switch 156 assumes the position 1 if the switch 158 changes to the position 2. In this case, the switch 158 switches through a logical 1 to the switch 156, whereby it assumes the position 1. The switch 158 assumes the position 2 if the signal 160 has the logical value 1. This is then the case if the measured engine speed 32 becomes greater than or equal to the presettable activation speed 164 and the speed control deviation 34 simultaneously becomes smaller than or equal to the value 0. For the starting process of the engine, this means the following: after the engine is started, if the engine speed 32 reaches the activation speed 164, for example 1500 1/min, and if the engine speed 32 simultaneously reaches the set speed 30 (speed control deviation equal to 0), the switch 156 changes to the position 1, whereby the dynamic proportional coefficient 80 is calculated additively from the stationary proportional coefficient kpStat 152 and a component 154 which is a function of the speed control deviation 34. If an engine shutdown is detected, the logical signal 165 thus has the value 1 and the switch 158 assumes the position 1. Therefore, a logical 0 is switched through by the switch 158, so that the switch 156 assumes the position 0. In this case, the dynamic proportional value 80 is again identical to the stationary proportional coefficient kpStat 152.

(19) The component 154 which is a function of the speed control deviation 34 is calculated as follows: if the speed control deviation 34 becomes greater than the presettable value e.sup.min.sub.pos, the additive component 154 of the dynamic proportional coefficient 80 which is a function of the speed control deviation 34 is increased linearly until the speed control deviation 34 reaches the value e.sub.max. In the case of a further increase in the speed control deviation, the additive component 154 remains constant. On the other hand, if the speed control deviation 34 is negative and less than the presettable value e.sup.min.sub.neg, the additive component 154 is increased linearly until the speed control deviation 34 reaches the negative presettable value e.sub.max. If the speed control deviation is further reduced, the additive component 154 in turn remains constant.

(20) By calculating the dynamic proportional coefficient 80 as a function of the speed control deviation 34, the dynamic response of the closed-loop speed control system may be decisively improved in the case of non-stationary operations, in particular in the case of load connection and load disconnection operations, since, in the case of the occurrence of a speed control deviation, the proportional coefficient of the speed controller increases, and thus the proportional component also increases.

(21) FIG. 4 shows the calculation of the stationary proportional coefficient kpStat 152. If the engine speed n.sub.actual 32 is equal to 0, the switch 200 assumes the position 1, so that the value 80 is switched through. On the other hand, if the engine speed n.sub.actual 32 is not equal to 0, the engine speed n.sub.actual 32 is limited downward to the presettable value n.sub.min 202 and switched through by the switch 200, since it assumes the position 0 in this case. Subsequently, the reciprocal value 206 (block 204) is formed from the output value of the switch 200. This reciprocal value 206 is multiplied by the factor f 208, the closed-loop gain v 209, and the I component E.sup.I.sub.target 212 which is limited downward to the presettable value E.sub.min 210. The result 214 of this multiplication is still limited to the presettable value kpmin 216 and constitutes the stationary proportional coefficient kpStat 152. Altogether, kpStat 152 is thus calculated as follows:
kpStat=(f*v*E.sup.I.sub.target)/n.sub.actual(1)
E.sup.I.sub.targetE.sub.min
n.sub.actualn.sub.min
kpStatkpmin
where f=1 (generator) f=2 (ship)

(22) The I component E.sup.I.sub.target must be limited downward to the value E.sub.min, so that the stationary proportional coefficient kpStat does not become too small or equal to 0 and so that the speed controller does not have a dynamic response which is too low. In the case of a proportional coefficient of 0, the proportional component of the speed controller would no longer be active. The engine speed n.sub.actual must be limited downward at least to the detection limit of the engine speed; this is, for example, 80 1/min. For additional safety, kpStat is finally still limited on the whole to the lower limit value kpmin.

(23) Alternatively, instead of the I component E.sup.I.sub.target the filtered fuel energy E.sub.target.sup.filtered 53 may be used for calculating the stationary proportional coefficient kpStat:
kpStat=(f*v*E.sub.target.sup.filtered)/n.sub.actual

(24) where

(25) E.sub.target.sup.filteredE.sub.min

(26) n.sub.actualn.sub.min

(27) kpStat>kpmin

(28) where

(29) f=1 (generator)

(30) f=2 (ship)

(31) Equation (1) constitutes the control law of the fuel energy-based speed controller. This control law characterizes the calculation of the stationary proportional coefficient kpStat. The stationary proportional coefficient kpStat is proportional to the fuel energy E.sup.I.sub.target or E.sub.target.sup.filtered and inversely proportional to the engine speed n.sub.actual. In this case, the proportionality factor is a product of two multipliers: the factor f and the closed-loop gain v, wherein the factor f is a function of the application and the closed-loop gain v is preset by the operator.

(32) For deriving the control law, the engine and the system are modeled as a one-mass oscillator. If the principle of conservation of angular momentum is applied to this one-mass oscillator, the following equation is obtained for the case of the propeller drive (ship application):
*dw/dt=M.sub.mk.sub.B*n.sub.actual.sup.2

(33) where

(34) =.sub.engine+.sub.load

(35) total moment of inertia [kg m.sup.2]

(36) wangular velocity [1/s]

(37) M.sub.mengine momentum [Nm]

(38) k.sub.Bproportionality factor [Nm min.sup.2]

(39) n.sub.actualengine speed [1/min]

(40) The angular velocity w is calculated as follows:
w=2*pi*n.sub.actual

(41) The following nonlinear model of the one-mass oscillator is thus obtained:
*2pi*dn.sub.actual/dt+k.sub.B*n.sub.actual.sup.2=M.sub.m

(42) If this equation is linearized, the following linear model of the one-mass oscillator is obtained:
*2*pi*d(n)/dt+2*k.sub.B*n.sub.op*n=M.sub.m

(43) where

(44) n.sub.op: engine speed operating point at which linearization is performed

(45) n, M.sub.m: deviations of the engine speed and the engine momentum from the operating point

(46) For the transfer function of the one-mass oscillator, the following thus applies:
G(s)=n(s)/M.sub.m(s)=k.sub.m/(1+T.sub.m*s)

(47) where
k.sub.m=1/(2*k.sub.B*n.sub.op)(2)
T.sub.m=(pi*)/(k.sub.B*n.sub.op)

(48) The fuel energy E.sub.target per injection is related to the engine momentum M.sub.m as follows:
E.sub.target=(pi*M.sub.m)/(250*z*)

(49) where

(50) E.sub.targetfuel energy per injection [kJ]

(51) M.sub.mengine momentum [Nm]

(52) znumber of cylinders [ ]

(53) efficiency [ ]

(54) The following thus applies for the engine momentum M.sub.m:
M.sub.m=kv*E.sub.target(3)

(55) where

(56) kv=(250*z*)/pi

(57) At the operating point (M.sub.m.sup.op, E.sub.target.sup.op), the following thus applies:
M.sub.m.sup.op=k.sub.v*E.sub.target.sup.op(4)

(58) For the load torque, the following applies:
M.sub.L.sup.op=k.sub.B*n.sub.op.sup.2

(59) The following thus applies:
k.sub.B*n.sub.op=M.sub.L.sup.op/n.sub.op(5)

(60) For the gain of the engine, the following applies:
v.sub.m=k.sub.v*k.sub.m

(61) With (2), the following applies:
v.sub.m=k.sub.v*[1/(2*k.sub.B*n.sub.op)]

(62) With (5), the following is obtained:
v.sub.m=(k.sub.v*n.sub.op)/(2*M.sub.L.sup.op)

(63) In stationary operation, the engine momentum and load torque are identical:
M.sub.m.sup.op=M.sub.L.sup.op

(64) The following thus applies:
v.sub.m=(k.sub.v*n.sub.op)/(2*M.sub.m.sup.op)

(65) With (4), the following applies:
v.sub.m=(k.sub.v*n.sub.op)/(2*k.sub.v*E.sub.target.sup.op)

(66) The following thus applies for the stationary gain of the engine:
v.sub.m=n.sub.op/(2*E.sub.target.sup.op)(6)

(67) For the closed-loop gain v of the open-loop speed control system, the following applies:
v=kpStat*v.sub.m

(68) The following control law is thus obtained:
kpStat=(2*v*E.sub.target.sup.op)/n.sub.op

(69) where

(70) kpStatstationary proportional coefficient [kJ min]

(71) vclosed-loop gain [ ]

(72) n.sub.opengine speed [1/min]

(73) E.sub.target.sup.optarget fuel energy [kJ]

(74) If the I component of the speed controller is used for E.sub.target.sup.op and the measured speed n.sub.actual is used for n.sub.op, the following equation is obtained for the ship application:
kpStat=(2*v*E.sup.I.sub.target)/n.sub.actual(ship)

(75) In the case of the generator application, a linear relationship between the load torque M.sub.L and the engine speed n.sub.actual applies. This results in a modified multiplication factor in the control law:
kpStat=(v*E.sup.I.sub.target)/n.sub.actual(generator)

(76) Altogether, the aforementioned control law (1) is obtained:
kpStat=(f*v*E.sup.I.sub.target)/n.sub.actual

(77) where

(78) f=1 (generator)

(79) f=2 (ship)

(80) E.sup.I.sub.targetE.sub.min

(81) n.sub.actualn.sub.min

(82) kpStat>kpmin

(83) Via this control law, the closed-loop gain of the open-loop speed control system is held constant over the entire operating range. Equation (6) shows that the gain of the engine is low at low engine speed and is high at high engine speed. In the case of low fuel energy, the gain of the engine is high, and in the case of high fuel energy, i.e., high load, it is low. Since, corresponding to the aforementioned control law, a large kpStat is calculated at low engine speed and a small kpStat is calculated at high engine speed, the closed-loop gain of the open-loop speed control system is held constant overall. The same applies for the fuel energy: in the case of low fuel energy, a small kpStat is calculated, and at high fuel energy, a large kpStat is calculated, so that the closed-loop gain may be held constant overall in this case as well.

(84) The closed-loop gain v is a presettable parameter. By increasing this parameter, the dynamic response of the closed-loop speed control system may be increased. The control law in the described form is characterized by the following features: The stationary proportional coefficient kpStat is tracked linearly via the fuel energy. The stationary proportional coefficient is inversely proportional to the engine speed. The stationary proportional coefficient is proportional to the closed-loop gain v, which may be preset by the operator. The stationary proportional coefficient is twice as large in the ship application as in the generator application. The stationary proportional coefficient is limited downward to the presettable value kpmin.

(85) The rate-action time tv is used in FIG. 2 in order to calculate the gain 100 of the DT.sub.1 component. In this case, the rate-action time tv may be constant or, as depicted in FIG. 5, may alternatively be calculated as a function of the fuel energy. In this case, either the I component E.sup.I.sub.target of the speed controller, or alternatively, the filtered target fuel energy E.sub.target.sup.filtered, may be used as the fuel energy.

(86) FIG. 5 shows the profile of the rate-action time tv 250 as a function of the fuel energy 248. The depiction shows that the rate-action time tv 250 is identical to the value tv.sub.min 252 if the fuel energy is less than the presettable value E.sub.min 254. If the fuel energy is greater than the presettable value E.sub.max 256, tv is identical to the value tv.sub.max 258. If the fuel energy is greater than E.sub.min 254 and less than E.sub.max 256, tv 250 is linearly tracked via the fuel energy 248. The values tv.sub.min 252 and tv.sub.max 258 may be preset by the operator.

(87) FIG. 1 shows that the load signal-fuel energy 38 is added to the output 36 of the PI(DT.sub.1) speed controller. In this case, the load signal-fuel energy 38 constitutes a disturbance variable of the closed-loop speed control system. It has the task of improving the dynamic response of the speed controller in the case of non-stationary operations, for example, in the case of load connection and load disconnection operations.

(88) FIG. 6 shows the calculation of the load signal-fuel energy 38. The load signal-fuel energy is calculated from a system signal, in which it is, for example, a generator power signal. The system signal is provided as a 0 to 10 volt signal or a 4 to 20 mA signal by the operator of the system. If the switch 341 has the position 1, a voltage signal U (0 to 10 volts) is used; if the switch 341 has the position 2, a current signal I (4 to 20 mA) is used.

(89) The respective input signal 302 or 304 is initially converted to percent via a two-dimensional curve 306 or 308. The signal 310 defined in percent is obtained. The presettable maximum load signal-fuel energy 312, for example, identical to the value 20,000 J, is divided by the value 100 and multiplied by this value converted into percent. The result 316 of this multiplication is now amplified by a DT.sub.1 element 318. The presettable parameters of the DT.sub.1 algorithm are the rate-action time tv.sub.Load and the delay time t1.sub.Load. Both parameters are depicted as input variables of the block 318. The output 320 of the DT.sub.1 system 318 is processed by the hysteresis block 322 as follows: if the output of the DT.sub.1 system exceeds an upper limit value, for example 1000 J, or if it falls below a lower limit value, for example 1000 J, the output of the DT.sub.1 system is switched through, i.e., activated. In this case, the output 324 of the hysteresis block is identical to the output of the DT.sub.1 system. On the other hand, if the magnitude of the output of the DT.sub.1 system falls below a further limit value, for example 50 J, it is switched off; i.e., in this case, the output of the hysteresis block equals 0. The limit values are depicted as input variables of the block 322.

(90) The load signal-fuel energy 38 is identical to the output 324 of the hysteresis block 322 if the switch 330 assumes the position 1. This is the case if the engine speed 32 becomes greater than or equal to the presettable speed 334, and the load signal active parameter 340 is simultaneously equal to 1. This means that the load signal-fuel energy 38 is enabled if the engine speed 32 reaches the presettable speed 334 and the presettable load signal active parameter 340 is set to the value 1. In all other cases, the load signal-fuel energy 38 equals 0. The task of the load signal-fuel energy 38 is to support the speed controller in the case of load connection and disconnection operations. If a load is connected or disconnected in the case of a generator, the generator power thus increases or decreases. If this is detected and read in by the engine electronics as a 0 to 10 volt signal or a 4 to 20 mA signal, the signal is amplified with the aid of the DT.sub.1 element and applied to the speed controller as a disturbance variable, whereby the dynamic response, i.e., the responsiveness of the closed-loop speed control system, is improved.