A method for determining a correction of a fuel amount to be injected into each cylinder of an internal combustion engine to compensate for the effects of a fuel amount lost from each cylinder by backflow into the combustion air supply system

Abstract

There is described a method for determining a correction of a fuel amount to be inj ected into each cylinder of an internal combustion engine in order to compensate for the ef fects of a fuel amount rejected by backflow from each cylinder into the combustion air supply system during transitions of the engine operating cycle from an Atkinson cycle to a Miller Cycle and vice versa. The method enables avoiding air-to- fuel ratio variations during the transition, thereby limiting the emission of pollutants.

Description

[0001] "A method for determining a correction of a fuel amount to be injected into each cylinder of an internal combustion engine to compensate for the effects of a fuel amount lost from each cylinder by backflow into the combustion air supply system"

[0002] ★ ★ ★ ★

[0003] TEXT OF THE DESCRIPTION

[0004] Field of the Invention

[0005] The present invention refers to internal combustion engines . The invention was developed with speci fic reference to internal combustion engine with variable transmission parameters , so as to favour the transitions of the engine operating cycle , speci fically the transitions from an Atkinson cycle to a Miller Cycle and vice versa .

[0006] Prior Art

[0007] As a general theoretical introduction, in an Otto thermodynamic cycle the closing of the intake valves ideally takes place at the bottom dead centre , as soon as the intake stroke has ended . In a Miller thermodynamic cycle , the intake valves are closed before reaching the bottom dead centre , while in an Atkinson thermodynamic cycle the intake valves are closed after reaching the bottom dead centre . Among the three operating cycles , the Atkinson cycle exhibits the most serious problems as regards the air-to- fuel ratio oscillations during the occurrence of a transition .

[0008] Considering the three di f ferent closing timings of the intake valves , it is immediately apparent that , in the Otto or Miller cycle operation, the closing timing of the intake valves does not trigger any signi ficant event of fuel backflow towards the intake ports and towards the intake mani fold ( comprising a plenum and runners individually associated with each cylinder ) , while it is equally apparent that the closing timing of the intake valves may lead to rather significant backflow events during the transition to the Atkinson cycle, because the intake valves remain open during part of the compression stroke.

[0009] The operation with the Atkinson cycle does not imply problems during stationary operating conditions; the fuel flowing back into the runner-port assembly (i.e. into the connection between the plenum and the cylinder) or into the plenum of the intake manifold is drawn back in on the following cycle, and even an open-loop injection pattern is fundamentally stable. Generally speaking, the backflow in stationary condition is mapped in the control unit of the engine, so that the injected mass of fuel may be adjusted accordingly. The problem appears during the transition from Miller to Atkinson and vice versa.

[0010] As an example, let us consider the case of a transition of operation from a Miller cycle - therefore, a cycle without any fuel backflow - to an Atkinson cycle; the injected fuel, deriving either from a direct or from an indirect injection upstream the intake valves, is subjected to a partial backflow along the compression stroke, during which the intake valves remain open. If it is the first Atkinson cycle for the cylinder, the combustion taking place in that cylinder is necessarily a lean burn, due to the fuel backflow. On the following cycle, is situation is reversed; the fuel which has flown back to the supply conduit and to the plenum of the intake manifold is drawn back in together with the new charge, therefore increasing the global amount of fuel in the combustion chamber and leading to a rich burn. After a few cycles, the engine spontaneously tends to a condition of balance, therefore eliminating the problem. However, during the cycles needed to achieve a new balance, the polluting emissions of the engine essentially depend on the oscillations between rich burn conditions and lean burn conditions . As the operation during a transition ( the transition Atkinson-Miller and vice versa being a type thereof ) is a signi ficant part of the global operation of an internal combustion engine , the spontaneous evolution of the engine towards new balance conditions cannot be accepted i f it is necessary to control consumptions and emissions .

[0011] Obj ect of the Invention

[0012] The invention aims at overcoming the technical problems outlined in the foregoing . Speci fically, the invention aims at contrasting the air-to- fuel ratio oscillations during the transitions from a Miller cycle to an Atkinson cycle and vice versa .

[0013] Summary of the Invention

[0014] The obj ect of the invention is achieved by means of a method having the features forming the subj ect of the claims that follow, which form an integral part of the technical disclosure provided herein in relation to the invention .

[0015] Brief Description of the Figures

[0016] The invention will now be described with reference to the annexed Figures , which are provided by way of non-limiting example only, wherein :

[0017] - Figure 1 shows a flow diagram which exempli fies the method according to the invention,

[0018] - Figure 2 shows a flow diagram referring to a sequence of combustion cycles to which the method according to the invention refers to , in a transition from a Miller cycle operation to an Atkinson cycle operation, and

[0019] - Figure 3 shows a flow diagram referring to a sequence of combustion cycles to which the method according to the invention refers to , in an transition from an Atkinson cycle operation to a Miller cycle operation .

[0020] Detailed Description

[0021] As a general introduction, according to the embodiments of the invention there is defined a method for determining a correction of an amount , speci fically of a mass , of fuel to be inj ected in each cylinder of an internal combustion engine to compensate for the ef fects of an amount of fuel lost from each cylinder by backflow into the combustion air supply system, particularly during timing and / or li ft transitions of the intake valves , as is the case in the transition of the engine operating cycle from an Atkinson cycle to a Miller cycle and vice versa .

[0022] The method is applicable to any ( especially gasoline- fuelled) spark-ignition internal combustion engine , whether it is turbocharged or naturally charged, comprising one or more cylinders which are supplied with combustion air through corresponding intake ports ( at least one for each cylinder ) , which are generally arranged in the head which closes the cylinders , and an intake mani fold comprising a ( common) plenum and corresponding runners , individually associated with each cylinder and provided upstream of the respective intake ports . In a typical configuration, with an aligned arrangement of cylinders (which applies for inline engines , V engines , boxer engines as well as flat engines ) , the intake mani fold comprises a plenum, which is in fluid communication with the one or more cylinder ( s ) of the engine by means of supply conduits - one for each cylinder - corresponding to the assembly of runner and port of each cylinder, and therefore comprising the at least one intake port of the cylinder and the runner which starts from the plenum of the intake mani fold and is connected in fluid communication with the at least one intake port of the cylinder . Referring to Figure 1, the method according to the invention comprises: determining an amount of charge rejected by backflow RCH by each cylinder of said internal combustion engine, determining (block 2) , for each cylinder, a distribution of the rejected charge, said determining a distribution of the rejected charge comprising determining a first fraction of the rejected charge which is stored in the at least one inlet supply conduit (runner + port, block 4) , and determining a second fraction of said rejected charge which is stored in the plenum of the intake manifold (block 6) ,

[0023] - determining an amount of fuel present in the first fraction of the rejected charge (blocks 8, 10) of each cylinder, and determining an amount of fuel stored in the plenum of the intake manifold (block 12) ,

[0024] - determining (blocks 14, 16) an amount of fuel drawn in from the plenum of the intake manifold by each cylinder on each cycle following an engine operating cycle transition from an Atkinson cycle to a Miller cycle, or vice versa,

[0025] - determining, for each cylinder, a correction of the amount of fuel (FC) to be injected on each cycle following an engine operating cycle transition from an Atkinson cycle to a Miller cycle or vice versa, based on said amount of fuel drawn in from the plenum of the intake manifold into the cylinder and based on the amount of fuel present in the first fraction of the charge rejected from the cylinder (block 18) .

[0026] The correction FC is to be understood as a variation of the amount of fuel (particularly of the mass of fuel) to be injected into the nthcylinder in order to compensate for the effect of the fuel lost in the supply conduit (runner + port) which connects the plenum to the cylinder, and / or in the plenum of the intake mani fold, net of the fuel drawn back in by the cylinder . The division into the two above-mentioned control volumes is essentially linked to the physics of the charge intake event . The first control volume of each cylinder, which comprises the runner and the intake port , corresponds to a control volume which pertains substantially exclusively to the cylinder, because it is fundamentally impossible for the fluid present therein - whether stationary or transiting - to have interactions with other cylinders ( for the purposes of the method, it is virtually impossible for the fluid to be attracted and drawn in by other cylinders ) . The second control volume , i . e . the plenum of the intake mani fold, corresponds to a control volume the fluid whereof may be attracted and drawn in generally by any cylinder . It shall be borne in mind, however, that the separation of the two control volumes might - depending on the speed of rotation of the engine and on the engine load - be slightly ( albeit variably) di f ferent from the physical separation of the components ( the head and parts of the intake mani fold) which physically define the control volumes .

[0027] This will become clearer from the exemplary calculations as per Figures 2 and 3 . The correction FC generally has a positive sign in the transition from a Miller cycle to an Atkinson cycle - an event which leads , as observed in the foregoing, to a loss of fuel in the charge supply system, and therefore the positive sign indicates the need to refill the loss - and generally has a negative s ign in the transition from an Atkinson cycle to a Miller cycle - an event which leads to the depletion of the fuel storage in the supply conduits and in the plenum of the intake mani fold .

[0028] Determining the amount of rej ected charge RCH and the amount of fuel in the rej ected charge , meaning both the fuel stored in the plenum of the intake manifold and the fuel stored in each supply conduit, is performed based on a difference between a maximum expected volumetric efficiency value and an estimated current volumetric efficiency value, based on the closing angle of the intake valve (s) (which is mapped according to a calculation model) .

[0029] Determining the amount of fuel present in the rejected charge RCH is based on an estimate (block 8) of the mass of fuel as a function of an engine speed of rotation, of an engine load (and therefore, inter alia, of a volumetric efficiency) and, if necessary, of the injection distribution, i.e. the distribution between a sum of a mass of fuel injected into an intake port and into the corresponding cylinder with one or more intake valves are open (deriving, mainly or totally, from an injection system into the port, and minimally or not at all from a direct injection system) , and the mass injected into the cylinder with one or more intake valves closed (the latter in the case of a direct injection into the cylinder) . If the engine operates only on the basis of either system, of course the respective variable will take on values or coefficients which reflect the one-sidedness of the injection. Generally speaking, the injection system (which may be either into the port or direct) influences the calculation of the amount of fuel (gasoline) which is rejected. In the system with injection into the port, the amount of charge which is rejected generally has a stoichiometric air-to-fuel ratio (or anyway the air-to-fuel ratio mapped for that operating point) , while for a system with direct injection the air-to-fuel ratio of the rejected charge is a function of the fraction percentage (on the total of injected mass) of the fuel injected with the valves open. A simplification which is implemented in preferred embodiments of the invention, albeit not being strictly necessary or the only possible solution, consists in assuming that all the fuel inj ected on the first posttransition cycle of each cylinder may contribute to the rej ection of the charge by backflow, while the inj ections on the cycles following the transition take place with the intake valve ( s ) closed or nearly closed, so that the contribution thereof in terms of mass of fuel in the rej ected charge is zero or nearly zero .

[0030] Moreover, determining the amount of fuel present in the rej ected charge is performed as a function of a relative air-to- fuel ratio mapped for the speci fic operating point of the engine . Generally speaking, in the engines which may operate indi f ferently both according to the Atkinson cycle and according to the Miller cycle , and which are adapted to accommodate such transitions , the inj ection mapping envisages an operation with stoichiometric air-to- fuel ratio ; therefore , generally speaking, the mapped value of relative air-to- fuel ratio amounts to 1 , and determining the fuel present in the rej ected charge is based on such assumption . Of course , determining the fuel takes place whatever the mapped value of relative air-to- fuel ratio , and therefore it is possible to achieve a determination also based on the operating points involved in a Miller- Atkinson or Atkinson-Miller transition with a non- unitary value of relative air-to- fuel ratio .

[0031] The calculation o f the mass of fuel present in the plenum of the intake mani fold and present in the supply conduit of each cylinder is performed on the basis of two separate calculation models , which correspond to the di f ferent phenomenology the two masses are subj ected to during engine operation .

[0032] The mass of fuel of the rej ected charge which is present in the supply conduits is generally drawn back in by the same cylinder on the following cycle , because it cannot be attracted and drawn in by other cylinders . As a consequence , such a mass may be assumed to remain in the supply conduit without a signi ficant trans fer towards the plenum of the intake mani fold .

[0033] The mass of fuel of the rej ected charge which is present in the plenum of the intake mani fold, on the contrary, is subj ected to a layered evolution . Generally speaking, it does not exhibit stationary features either as regards the distribution or as regards the amount , and it is progressively drawn back in by the other cylinders on the cycles following the cycle which initially triggered the rej ection of the charge by backflow .

[0034] The di f ferent behaviour is reflected in the blocks 14 , 16 , 18 : block 18 corresponds to an operator which implements the timing of the individual cylinders of the engine . As the mass of fuel present in the supply conduit of the cylinder is substantially drawn back in by the same cylinder, the value of the mass present in the supply conduit to be taken into account for the correction is the value which refers to the previous cycle of the same cylinder ; therefore , block 18 practically introduces a sort of delay to the following cycle in considering the value of the mass of fuel which remains in the supply conduit of the cylinder .

[0035] Block 16 - and, together with it , block 14 - take into account the fact that the mass of fuel dispersed towards the plenum of the intake mani fold is progressively drawn back in ( and possibly further supplied) by the intake events of the following cylinders . The calculation of the correction FC for each individual cylinder, therefore , starts from a value of mass of fuel present in the plenum of the intake mani fold which equals the latest stored value (block 16 ) , from which there is subtracted the mass of fuel of the (nth) cylinder involved in the correction on the cycle involved in the correction (block 14 and following operator . It shall be observed, moreover, that the result of the subtraction is recycled as the latest stored value of the mass of fuel in the plenum of the mani fold .

[0036] The sum of the contributions to the correction, with the respective delays or updates (block 18 , 16 respectively) participates in the definition of the correction FC . Irrespective of the operators and "+" shown in Figure 1 , each contribution to the correction has a sign of its own, which depends on the physical phenomenon which generated it .

[0037] By way of example , let us consider the calculation progressions shown in Figures 2 and 3 . The text provided in the Figure has an immediately evident meaning, and therefore further numerical references will be omitted . Very simply, each box corresponds to the implementation of the method according to the invention, as per Figure 1 , on the individual cylinders of a spark-ignition internal combustion engine having four cylinders in line , with the usual ignition order 1-3-4-2 .

[0038] Referring to Figure 2 , it shows an operating transition from a Miller cycle to an Atkinson cycle , which starts at cylinder n . 3 , meaning that cylinder n . 1 is still operating with a Miller cycle ( reference C1M) , while the following cylinder in the order of ignition is already operating according to the Atkinson cycle ( reference C3A / 1 ) . The coding of the post-transition cycles may be CNA / K or CNM / K, wherein N is the number of the cylinder, A and M respectively represent the Atkinson cycle and the Miller cycle , K represents the iteration of the cycle according to which the cylinder operates ( 1 = first cycle , 2 = second cycle , etc . ) .

[0039] The last Miller cycle for the cylinder n . 1 - block C1M) clearly shows zero values of the rejected mass of fuel (rejected charge = 0) in the supply conduit (abbreviated as "s.c." in the Figures) , and zero values of the mass of fuel rejected into the plenum of the intake manifold, as well as zero value of the flow which is drawn back in from the plenum of the intake manifold. As a consequence, the value of correction FC is zero.

[0040] The first Atkinson cycle of cylinder n. 3 - block C3A / 1 - shows the first effects of the rejection of charge at intake. The amount of the rejected charge RCH is 100 mg of charge per cubic centimetre of the cylinder, which results in 10 mg of fuel lost in the supply conduit and 4.4 mg of fuel lost in the plenum of the intake manifold of cylinder n. 3. Obviously, the amount of fuel drawn back in from the plenum of the intake manifold amounts to zero, because the previous cycle C1M did not generate any backflow. Because it is the first Atkinson cycle for cylinder 3, no re-intake of the fuel lost in the supply conduit takes place. The total correction FC, therefore, amounts to the sum of the contributions 10 mg and 4.4 mg and has a positive sign, because it will be necessary to restore the mass lost outside the cylinder.

[0041] Meanwhile, the first Atkinson cycle for cylinder n. 3 has caused an increase of the mass of fuel in the plenum of the intake manifold. As regards cylinder n. 4, block C4A / 1, the first Atkinson cycle of the cylinder features a rejection of charge always amounting to 100 mg / cc, corresponding to a rejection of fuel again amounting to 14.4 mg, of which 10 mg are lost in the supply conduit and 4.4 mg are lost in the plenum of the intake manifold (which are summed to the 4.4 mg which are already present from the cycle C3A / 1) , but at the same time a re-intake takes place, from the plenum of the intake manifold, of 2 mg of fuel, i.e. of a part of the mass of fuel which was lost during the first cycle of cylinder n. 3. As it is the first Atkinson cycle for cylinder 4, no re-intake takes place of the fuel lost in the supply conduit. This reduces the amount of the correction FC of the mass drawn back in from the plenum of the intake manifold, because the latter mass is burnt inside cylinder n. 4. The correction FC for cylinder 4 therefore amounts to 10 mg + 4.4 mg - 2 mg = 12.4 mg, with positive sign, for the same reasons as for cylinder 3.

[0042] As regards cylinder n. 2, block C2A / 1, on the first Atkinson cycle of the cylinder a rejection of charge takes place which again amounts to 100 mg / cc, corresponding to a rejection of fuel which always amounts to 14.4 mg, of which 10 mg are lost in the supply conduit and 4.4 mg are lost in the plenum of the intake manifold (which are summed to the 6.8 mg which are already present as a residue of the cycles C3A / 1, C4A / 1) , but at the same time a re-intake takes place from the plenum of the intake manifold of 3 mg of fuel, i.e. of part of the mass of fuel lost during the first cycles of cylinders n. 3 and n.4. This further reduces the amount of the correction FC of the mass drawn back in from the plenum of the intake manifold, because the latter is burnt within cylinder n. 2. As it is the first Atkinson cycle for cylinder 2, no re-intake takes place of the fuel lost in the supply conduit. Therefore, the correction FC for cylinder 2 amounts to 10 mg + 4.4 mg - 3 mg = 11.4 mg, positive sign, for the same reasons as for cylinders 3, 4.

[0043] As regards cylinder n. 1, block C1A / 1, the first Atkinson cycle of the cylinder features a rejection of charge which again amounts to 100 mg / cc, corresponding to a rejection of fuel which always amount to 14.4 mg, of which 10 mg are lost in the supply conduit and 4.4 mg are lost in the plenum of the intake manifold (which are summed to the 8.2 mg which are already present as a residue of the cycles C3A / 1, C4A / 1, C2A / 1) , but at the same time a re-intake takes place from the plenum of the intake manifold of 4 mg of fuel, i.e. of part of the mass of fuel which was lost during the first cycles of the cylinders n. 3, n. 4 and n. 2. This further reduces the amount of the correction FC of the mass drawn back in from the plenum of the intake manifold, as the latter mass is burnt inside cylinder n. 1. As it is the first Atkinson cycle for cylinder 1, no re-intake takes place of the fuel lost in the supply conduit. Therefore, the correction FC for cylinder 2 amounts to 10 mg + 4.4 mg - 4 mg = 10.4 mg, positive sign, for the same reasons as for cylinders 3, 4, 2.

[0044] As regards the second Atkinson cycles for the engine cylinders, the mass balance of the fuel starts tending towards neutrality on the second Atkinson cycle of cylinder n. 3, block C3A / 2. The second Atkinson cycle of the cylinder features a rejection of charge always amounting to 10 mg / cc, corresponding to a rejection of fuel always amounting to 14.4 mg, of which 10 mg are lost in the supply conduit and 4.4 mg are lost in the plenum of the intake manifold (which are summed to the 8.6 mg which are already present as a residue of the cycles C3A / 1, C4A / 1, C2A / 1, C1A / 1) , but at the same time a re-intake takes place from the plenum of the intake manifold of 4 mg of fuel (i.e., of part of the mass of fuel which was lost during the first cycles of the cylinders n. 3, n. 4, n. 2 and n. 1) and, being it a second cycle, a complete re-intake takes place of the 10 mg of fuel which were lost in the supply conduit on cycle C3 / 1. This greatly reduces the amount of correction FC with respect to the first cycle, because now it amounts only to 10 mg + 4.4 mg - 4 mg - 10 mg = 0.4 mg, again with a positive sign for the same reasons as for cylinders 3, 4, 2, 1.

[0045] A stable condition is achieved on the following cycle, corresponding to a second Atkinson cycle for cylinder n. 4. The second Atkinson cycle of the cylinder features a rejection of charge which always amounts to 100 mg / cc, corresponding to a rejection of fuel always amounting to 14.4 mg, of which 10 mg are lost in the supply conduit and 4.4 mg are lost in the plenum of the intake manifold (which are summed to the 9 mg which are already present as a residue of the cycles C3A / 1, C4A / 1, C2A / 1, C1A / 1, C3A / 2) , but at the same time a re-intake takes place from the plenum of the intake manifold of 4.4 mg of fuel (i.e., of part of the mass of fuel which was lost during the first cycles of the cylinders n. 3, n. 4, n. 2 and n. 1 and during the second cycle of cylinder n. 3) and, being it a second cycle, a complete re-intake takes place of the 10 mg of fuel which were lost in the supply conduit on cycle C4 / 1. This greatly reduces the amount of correction FC with respect to the first cycle, because now it amounts only to 10 mg + 4.4 mg - 4.4 mg - 10 mg = 0.0 mg, which leads to the absence of a correction for this cycle and for the following cycles, showing the achievement of a new stability.

[0046] Referring to Figure 3, it shows an operating transition reversed with respect to the one of Figure 2, i.e. from an Atkinson cycle to a Miller cycle, which always begins with cylinder n. 3, indicating therefore that cylinder n. 1 is still operating according to an Atkinson cycle (reference CIA) , while the following cylinder in the ignition order is already operating according to a Miller cycle (reference C3M / 1) . The coding of the post-transition cycles is the same as described in the foregoing.

[0047] The last Atkinson cycle for cylinder n. 1 - block CIA - features a rejection of charge which always amounts to 100 mg / cc, corresponding to a rejection of fuel always amounting to 14.4 mg, of which 10 mg of fuel are lost in the supply conduit and 4.4 mg of fuel are lost in the plenum of the intake manifold of cylinder n. 3. This is summed to the 9 mg of fuel which are present in the plenum of the intake manifold due to a previous sequence of cycles, corresponding to Figure 2, after the achievement of the conditions of stability, but at the same time a re-intake takes place from the plenum of the intake manifold of 4.4 mg of fuel and, being it a second cycle, a complete re-intake takes place of the 10 mg of fuel which were lost in the supply conduit on the previous Atkinson cycle of cylinder 4. Again, thus confirming the achievement of stability, this leads to a correction FC amounting to zero: 10 mg + 4.4 mg - 4.4 mg - 10 mg = 0.0 mg .

[0048] The first Miller cycle of cylinder n. 3 - block C3M / 1 - features an amount of rejected charge RCH amounting to 0 mg of charge, with the obvious consequence of a zero amount of fuel lost in the plenum of the intake manifold and in the supply conduit. However, a re-intake takes place of part of the 9 mg of fuel present in the plenum of the intake manifold (specifically, 4.4 mg) , as well as a complete re-intake of the 10 mg of fuel which have remained in the supply conduit since the previous Atkinson cycle of cylinder 3. The correction FC will therefore have a negative sign, being it a sum of terms which all have a negative sign (fuel which has been drawn back in) : FC = -4.4 mg - 10 mg = - 14.4 mg. In other words, in order to compensate for the re-intake of fuel, which would be in excess with respect to the target air- to-fuel ratio, the amount of fuel supplied to cylinder 3 is corrected by subtracting 14.4 mg of fuel.

[0049] The first Miller cycle of cylinder n. 4 - block C4M / 1 - features again an amount of rejected charge RCH amounting to 0 mg of charge, with the obvious consequence of a zero amount of fuel lost in the plenum of the intake manifold and in the supply conduit. Again, a re-intake takes place of part of the 4.6 mg of fuel which remain in the plenum of the intake manifold (specifically, 2.2 mg) , as well as the complete re-intake of the 10 mg of fuel which have remained in the supply conduit since the previous Atkinson cycle of cylinder 4. The correction FC will have a negative sign again, because it is a sum of terms which all have a negative sign (fuel which is drawn back in) : FC = -2.2 mg - 10 mg = - 12.2 mg. In other words, in order to compensate for the re-intake of fuel, which would be in excess with respect to the target air- to-fuel ratio, the amount of fuel supplied to cylinder 3 is corrected by subtracting 12.2 mg of fuel.

[0050] The first Miller cycle of cylinder n. 2 - block C2M / 1 - features again an amount of rejected charge RCH amounting to 0 mg of charge, with the obvious consequence of a zero amount of fuel lost in the plenum of the intake manifold and in the supply conduit. A further re-intake takes place of part of the 2.4 mg of fuel which remain in the plenum of the intake manifold (specifically, 1.1 mg) , as well as the complete re-intake of the 10 mg of fuel which have remained in the supply conduit since the previous Atkinson cycle of cylinder 2. The correction FC will have a negative sign again, because it is a sum of terms which all have a negative sign (fuel which is drawn back in) : FC = -1.1 mg - 10 mg = - 11.1 mg. In other words, in order to compensate for the re-intake of fuel, which would be in excess with respect to the target air- to-fuel ratio, the amount of fuel supplied to cylinder 2 is corrected by subtracting 11.1 mg of fuel.

[0051] Moreover, the first Miller cycle of cylinder n. 1 - block CIM / 1 - features again an amount of rejected charge RCH amounting to 0 mg of charge, with the obvious consequence of a zero amount of fuel lost in the plenum of the intake manifold and in the supply conduit. A further re-intake takes place of the residues, 1.3 mg of fuel, from the plenum of the intake manifold (specifically, 0, 6 mg) , as well as the complete re-intake of the 10 mg of fuel which have remained in the supply conduit since the previous Atkinson cycle of cylinder 1. The correction FC will have a negative sign again, because it is a sum of terms which all have a negative sign (fuel which is drawn back in) : FC = -0.6 mg - 10 mg = - 10.6 mg. In other words, in order to compensate for the re-intake of fuel, which would be in excess with respect to the target air-to-fuel ratio, the amount of fuel supplied to cylinder 2 is corrected by subtracting 10.6 mg of fuel .

[0052] Similarly to the transition as per Figure 2, with the second cycles the correction FC definitely tends towards new stability conditions.

[0053] The second Miller cycle of cylinder n. 3 - block C3M / 2 - features again an amount of rejected charge RCH amounting to 0 mg of charge, with the obvious consequence of a zero amount of fuel lost in the intake manifold and in the plenum of the supply conduit. A further re-intake takes place of the residues, 0,7 mg of fuel, in the plenum of the intake manifold (specifically, 0,3 mg) , but unlike the first cycles there is no re-intake of fuel from the supply conduit, since the fuel has already been drawn back in on the previous cycle and has no longer been replaced by a new rejected charge.

[0054] The correction FC keeps the negative sign, but it is much smaller than in the previous cycles, because it is no longer influenced by the contribution of fuel from the supply conduit: FC amounts to -0.3 mg, i.e. substantially to the residue drawn in from the plenum of the intake manifold. In other words, in order to compensate for the re-intake of fuel, which would be in excess with respect to the target air-to-fuel ratio, the amount of fuel supplied to cylinder 3 is corrected by subtracting 0.3 mg of fuel.

[0055] Stability is achieved on the second Miller cycle of cylinder 4, block C4M / 2. The second Miller cycle of cylinder n. 4 features again an amount of rejected charge RCH amounting to 0 mg of charge, with the obvious consequence of a zero amount of fuel lost in the plenum of the intake manifold and in the supply conduit. A further re-intake takes place of the residues, 0.3 mg of fuel, in the plenum of the intake manifold (specifically, 0.1 mg) , but unlike the first cycles there is no reintake of fuel from the supply conduit, since the fuel has already been drawn back in on the previous cycle and has no longer been replaced by a new rejected charge.

[0056] The correction FC keeps the negative sign, but it is approximately zero: FC amounts to -0.1 mg, which is substantially the residue drawn in from the plenum of the intake manifold. It is an almost negligible correction, which indicates the achievement of a new stability .

[0057] Thanks to the method according to the invention, it becomes possible to manage the operating transitions from an Atkinson cycle to a Miller cycle and vice versa, without significant deviations from the target air-to- fuel ratio (such deviations would occur if the method were not implemented) , with a consequent elimination or nearly total limitation of undesirable peaks of polluting emissions (NOx in case of lean burn, HC in case of rich burn) .

[0058] Of course, the implementation details and the embodiments may amply vary from what has been described and illustrated, without departing from the scope of the present invention as defined in the annexed claims.

Claims

CLAIMS1. A method for determining a correction of a fuel amount to be injected in each cylinder of an internal combustion engine to compensate the effects of a fuel amount rejected by backflow from each cylinder into the combustion air supply system during transitions of the engine operating cycle from an Atkinson cycle to a Miller Cycle and vice versa, the engine comprising one or more cylinders, an intake manifold including a plenum, and a supply conduit for each cylinder that establishes a fluid communication between the plenum of the intake manifold and the cylinder, the method comprising: determining an amount of charge rejected by backflow (RCH) by each cylinder of said internal combustion engine, determining (2) , for each cylinder, a distribution of said rejected charge, said determining a distribution of said rejected charge comprising: determining a first fraction of said rejected charge which is stored in the inlet supply conduit (4) , and determining a second fraction of said rejected charge which is stored in the plenum of the intake manifold (6) ,- determining an amount of fuel present in the first fraction of the rejected charge (8, 10) of each cylinder, and determining a quantity of fuel stored in the plenum of the intake manifold (12) ,- determining an amount of fuel drawn in from the plenum of intake manifold (14, 16) by each cylinder on each cycle following an engine operating cycle transition from an Atkinson cycle to a Miller cycle, or vice versa,- determining, for each cylinder, a correction to the amount of fuel to be injected in each cycle followingan engine operating cycle transition from an Atkinson cycle to a Miller cycle or vice versa based on the amount of fuel drawn in from the plenum of the intake manifold into the cylinder and the amount of fuel present in the first fraction of the charge rej ected from the cylinder ( 18 ) .2 . The method according to claim 1 , wherein said determining an amount of rej ected charge by backflow by a cylinder of an internal combustion engine comprises determining an amount of rej ected charge by di f ference between a maximum expected volumetric ef ficiency value and a current volumetric ef ficiency value .

3. The method according to claim 1 , wherein said determining an amount of fuel drawn in from the plenum of the intake mani fold is operated as a function of an engine speed of rotation, an engine load, and an amount of fuel present in the plenum of the intake mani fold .4 . The method according to any one of the preceding claims , wherein said amount of fuel is a mass of fuel .5 . The method according to claim 4 , wherein said mass of fuel is determined as a function of an engine speed of rotation, an engine load, and a fuel inj ection pattern, said fuel inj ection pattern comprising a distribution between a sum of a mass of fuel inj ected into an intake port and into the corresponding cylinder with one or more intake valves open, and a mass of fuel inj ected into the cylinder with one or more intake valves closed .

6. The method according to any one of the preceding claims , wherein said correction ( FC ) comprises , for each cylinder, summing the mass of fuel rej ected into the plenum of the intake mani fold to the mass rej ected into the supply conduit of the cylinder, and subtracting the mass of fuel drawn in from the plenum of the intake mani fold and the mass of fuel drawn in from the supplyconduit of the cylinder .7 . The method according to any one of the preceding claims , wherein the supply conduit of each cylinder comprises at least one cylinder intake port and a runner starting from the plenum of the intake mani fold plenum and connected in fluid communication with the at least one cylinder intake port .

Images