Control device for internal combustion engines
The control device accurately estimates mixed gas temperature in internal combustion engines by accounting for heat transfer with the intake manifold, using a neural network to improve fuel injection control and reduce emissions and fuel consumption.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- KK TOYOTA CHUO KENKYUSHO
- Filing Date
- 2024-12-16
- Publication Date
- 2026-06-26
AI Technical Summary
Existing techniques for estimating the temperature of a mixed gas in internal combustion engines do not account for heat exchange between the mixed gas and the intake manifold wall, leading to inaccuracies in fuel injection control.
A control device that estimates the temperature of the mixed gas by considering heat transfer between the mixed gas and the intake manifold wall, using a neural network to account for nonlinear relationships and incorporating variables such as rotational speed, fuel injection amount, and intake air temperature.
Accurate estimation of mixed gas temperature reduces emissions of carbon monoxide, hydrocarbons, and nitrogen oxides, and minimizes fuel consumption by optimizing fuel injection timing.
Smart Images

Figure 2026105187000001_ABST
Abstract
Description
[Technical Field]
[0001] This invention relates to a control device for an internal combustion engine. [Background technology]
[0002] Conventionally, techniques for estimating the temperature of a target gas are known in order to suitably control the fuel injection amount and fuel injection timing in an internal combustion engine. Patent Document 1 discloses a device for calculating the temperature of a filled gas, which is the gas filling the cylinder, which is the combustion chamber, using the intake gas temperature, the amount of intake gas in the filled gas, the residual gas temperature, and the amount of residual gas in the filled gas. Patent Document 2 discloses a control device for estimating the temperature of a mixed gas of fresh air and recirculated exhaust supplied to the combustion chamber, based on the target intake air flow rate, intake air temperature, recirculated exhaust temperature, and recirculated exhaust flow rate. Patent Document 3 discloses an intake gas temperature estimation device for estimating the cylinder gas temperature, which is the combustion chamber, from the difference between a calculated gas temperature value calculated using the amount, temperature, and specific heat of cylinder intake air and the amount, temperature, and specific heat of the mixed gas, and the mixture temperature drop rate. [Prior art documents] [Patent Documents]
[0003] [Patent Document 1] Japanese Patent Application Publication No. 11-148419 [Patent Document 2] Japanese Patent Publication No. 2009-287479 [Patent Document 3] Patent No. 4158679 [Overview of the Initiative] [Problems that the invention aims to solve]
[0004] In an internal combustion engine, heat is exchanged between the mixed gas (a mixture of intake air and recirculated gas / exhaust gas) and the wall of the intake manifold. Although this heat exchange affects the temperature of the mixed gas, Patent Documents 1-3 did not disclose any consideration of this point. Therefore, there was a need for a technique to improve the accuracy of estimating the temperature of the mixed gas.
[0005] The present invention has been made to solve at least some of the above-mentioned problems, and aims to provide a technique that can accurately estimate the temperature of a mixed gas obtained by mixing intake air and reflux gas. [Means for solving the problem]
[0006] The present invention has been made to solve at least some of the above-mentioned problems and can be realized in the following forms.
[0007] (1) According to one embodiment of the present invention, a control device for an internal combustion engine is provided. The control device for an internal combustion engine comprises a combustion chamber for burning fuel and air, a fuel injection unit for injecting the fuel to be burned in the combustion chamber, an intake pipe for supplying air to the combustion chamber, an exhaust pipe for discharging post-combustion gas from the combustion chamber, and a recirculation pipe for recirculating at least a portion of the post-combustion gas to the intake pipe as recirculation gas, and further comprises a heat transfer amount estimation unit for estimating the amount of heat transfer between a mixed gas, which is a mixture of intake air sent from the upstream side of the intake pipe above the connection portion with the recirculation pipe, and the wall surface of the intake pipe, and a mixed gas temperature estimation unit for estimating the temperature of the mixed gas using the amount of heat transfer estimated by the heat transfer amount estimation unit.
[0008] With this configuration, the amount of heat exchange between the mixed gas and the intake manifold wall is taken into account when estimating the temperature of the mixed gas, allowing for accurate estimation of the mixed gas temperature. This suppresses excess or insufficient fuel injection due to estimation errors in the mixed gas temperature and enables optimization of fuel injection timing, thereby reducing emissions of carbon monoxide, hydrocarbons, and nitrogen oxides from the internal combustion engine, as well as reducing fuel consumption.
[0009] (2) In the control device of the above form, the mixed gas temperature estimation unit may estimate the temperature of the mixed gas based on the law of conservation of energy and the law of conservation of mass. This configuration allows for a more accurate estimation of the mixed gas temperature by estimating it based on the laws of conservation of energy and mass.
[0010] (3) In the control device of the above form, the heat transfer amount estimation unit may estimate the heat transfer amount using a neural network. This configuration allows for accurate estimation of heat transfer by using a neural network capable of remembering the nonlinear relationship between input and output variables in the training data, to estimate the amount of heat transfer between the mixed gas flowing inside the intake pipe, which has a highly nonlinear and extremely complex flow field, and the wall of the intake pipe.
[0011] (4) In the control device of the above form, at least one of the following variables may be used as input variables to the heat transfer amount estimation unit: the rotational speed of the internal combustion engine, the amount of fuel injected from the fuel injection unit, the temperature of the cooling water that cools the internal combustion engine, the temperature of the intake air, the flow rate of the intake air, the temperature of the recirculating gas, and the flow rate of the recirculating gas. With this configuration, variables that directly or indirectly affect the amount of heat transfer between the gas mixture and the intake pipe wall are used as input variables to the neural network, allowing for accurate estimation of the heat transfer amount.
[0012] (5) The control device of the above form further includes a recirculating gas temperature estimation unit that estimates the temperature of the recirculating gas using a neural network, and at least one of the following variables may be used as input variables to the recirculating gas temperature estimation unit: the rotational speed, the fuel injection amount, the temperature of the coolant, the temperature of the intake air, the flow rate of the intake air, the internal pressure of the intake pipe, and the opening degree of the recirculating control valve that adjusts the flow rate of the recirculating gas. With this configuration, variables that directly or indirectly affect the temperature of the refluxing gas are used as input variables to the neural network, allowing for accurate estimation of the refluxing gas temperature.
[0013] (6) The control device of the above form further includes a recirculation gas flow rate estimation unit that estimates the flow rate of the recirculating gas using a neural network, and at least one of the following variables may be used as input variables to the recirculation gas flow rate estimation unit: the rotational speed, the fuel injection amount, the temperature of the cooling water, the temperature of the intake air, the flow rate of the intake air, the internal pressure of the intake pipe, and the opening degree of the recirculation control valve that adjusts the flow rate of the recirculating gas. With this configuration, variables that directly or indirectly affect the reflux gas flow rate are used as input variables to the neural network, allowing for accurate estimation of the reflux gas flow rate.
[0014] Furthermore, the present invention can be realized in various forms, for example, in the form of a vehicle equipped with the control device, or a control method for such a vehicle. [Brief explanation of the drawing]
[0015] [Figure 1] This is an explanatory diagram illustrating the configuration of the control device and internal combustion engine according to the embodiment. [Figure 2] This flowchart shows the procedure for the temperature estimation process performed by the control device. [Figure 3] This is a schematic diagram of a neural network. [Modes for carrying out the invention]
[0016] <Embodiment> Figure 1 is an explanatory diagram illustrating the configuration of the control device 40 (details of which will be described later) and the internal combustion engine 1 controlled by the control device 40 in this embodiment. The internal combustion engine 1 is a four-cylinder engine having four combustion chambers CR1 to CR4. An intake pipe 10 and an exhaust pipe 20 are connected to each of the combustion chambers CR1 to CR4. Hereafter, the combustion chambers CR1 to CR4 may be collectively referred to as combustion chamber CR.
[0017] The intake pipe 10 is a tubular member that supplies air to the combustion chamber CR. The intake pipe 10 is equipped with, in order from upstream, an air flow meter 12, a compressor 14, an intercooler 16, and a throttle valve 18.
[0018] The air flow meter 12 detects the flow rate of air drawn into the intake pipe 10. The compressor 14 is connected to the turbine 24 via shaft SH. The compressor 14 rotates in conjunction with the turbine 24, which is rotated by the post-combustion gas exhausted from the combustion chamber CR, via shaft SH, compressing the air sent from the upstream side and sending it to the downstream side. The compressor 14, shaft SH, turbine 24, and variable nozzle NZ together are called the supercharger CH.
[0019] The intercooler 16 cools the air compressed by the compressor 14. The throttle valve 18 adjusts the flow rate of air sent to the combustion chamber CR. The intake manifold 19 is a branched portion of the intake pipe 10 that connects to each of the combustion chambers CR1 to CR4. Air drawn into the intake pipe 10 from outside the internal combustion engine 1 is supplied to each of the combustion chambers CR1 to CR4 via the intake manifold 19.
[0020] Each of the combustion chambers CR1 to CR4 is equipped with an injector, which is a fuel injection unit (not shown). These fuel injection units inject fuel to be burned in the combustion chamber CR. The amount of fuel injected and the timing of fuel injection are controlled by the control device 40. In this embodiment, the internal combustion engine 1 injects fuel directly into the combustion chamber CR, but instead of providing fuel injection units in each of the combustion chambers CR1 to CR4, fuel may be injected into the intake pipe 10 from a fuel injection unit provided in the intake pipe 10. The combustion chamber CR outputs power by burning fuel and air.
[0021] The exhaust pipe 20 is a tubular member that discharges post-combustion gas from the combustion chamber CR. Post-combustion gas is the gas remaining after fuel and air have been burned in the combustion chamber CR. The exhaust manifold 22 is a branched portion of the exhaust pipe 20 connected to each of the combustion chambers CR1 to CR4. A turbine 24 is provided in the portion of the exhaust pipe 20 downstream of the exhaust manifold 22. As described above, the turbine 24 rotates using the post-combustion gas exhausted from the combustion chamber CR, thereby rotating the compressor 14. The turbine 24 is equipped with a variable nozzle NZ that adjusts the flow rate of post-combustion gas flowing into the turbine 24. A catalyst (not shown) is provided in the portion of the exhaust pipe 20 downstream of the turbine 24. Examples of this catalyst include oxidation catalysts and NOx storage reduction type three-way catalysts. After passing through the catalyst, the post-combustion gas is exhausted to the outside of the internal combustion engine 1.
[0022] The recirculation pipe 30 is a tubular member that connects the intake pipe 10 and the exhaust pipe 20. The recirculation pipe 30 is a tubular member that recirculates at least some of the post-combustion gas to the intake pipe 10 as recirculated gas. The recirculated gas is so-called EGR (Exhaust Gas Recirculation) gas. The recirculation pipe 30 is equipped with an EGR cooler 36 and a recirculation control valve 38.
[0023] The EGR cooler 36 cools the recirculated gas that flows into the recirculation pipe 30. The recirculation control valve 38 adjusts the flow rate of the recirculated gas that is recirculated to the intake manifold 10. More specifically, the control device 40, which will be described later, adjusts the flow rate of the recirculated gas that is recirculated to the intake manifold 10 by adjusting the opening degree of the recirculation control valve 38 and the opening degree of the throttle valve 18.
[0024] The control device 40 controls each component of the internal combustion engine 1 in accordance with information received from various sensors (not shown) installed in the internal combustion engine 1. For example, the control device 40 receives information from the various sensors indicating the rotational speed of the internal combustion engine and the amount the accelerator pedal is pressed, and controls the fuel injection unit, throttle valve 18, and recirculation control valve 38, etc., in accordance with this information.
[0025] The control device 40 includes a heat transfer amount estimation unit 41, a mixed gas temperature estimation unit 43, a recirculating gas temperature estimation unit 45, and a recirculating gas flow rate estimation unit 47. The heat transfer amount estimation unit 41 estimates the amount of heat transferred between the mixed gas and the wall surface of the intake pipe 10. The mixed gas is a mixture of intake air sent from the upstream side of the intake pipe 10 above the connection point with the recirculating pipe 30 and the recirculating gas. Air drawn into the intake pipe 10 from outside the internal combustion engine 1 mixes with the recirculating gas, which is the post-combustion gas recirculated by the recirculating pipe 30, before flowing into the combustion chamber CR to become a mixed gas. This mixed gas is used for combustion in the combustion chamber CR together with the fuel injected from the fuel injection unit. The mixed gas exchanges heat with the wall surface of the intake pipe 10 as it passes through before reaching the combustion chamber CR.
[0026] The mixed gas temperature estimation unit 43 estimates the temperature of the mixed gas using the heat transfer amount estimated by the heat transfer amount estimation unit 41. The reflux gas temperature estimation unit 45 estimates the temperature of the reflux gas using a neural network. The reflux gas flow rate estimation unit 47 estimates the flow rate of the reflux gas using a neural network. Details of the estimations performed by each of the heat transfer amount estimation unit 41, the mixed gas temperature estimation unit 43, the reflux gas temperature estimation unit 45, and the reflux gas flow rate estimation unit 47 will be described later.
[0027] Figure 2 is a flowchart showing the procedure for the temperature estimation process performed by the control device 40. The temperature estimation process involves determining the temperature T of the mixed gas. im This is a process for estimating the temperature. The temperature estimation process is performed periodically while the internal combustion engine 1 is in operation. When the temperature estimation process starts, the control device 40 first determines whether the start condition is met (step S1). The start condition is considered to be met only if all the various sensors that send signals to the control device 40 in the processes of steps S2 to S8 described later are functioning normally. If the start condition is not met (step S1: NO), the control device 40 terminates the temperature estimation process. At this time, the control device 40 may notify about the sensor that has malfunctioned.
[0028] If the starting condition is met (Step S1: YES), the control device 40 acquires the rotational speed Ne[k] (rpm) of the internal combustion engine 1 (Step S2). From this point onward, variables represented by the letter [k] indicate that they are variables at time t = kΔt (k: discrete time, Δt: sampling time interval). Information indicating the rotational speed Ne[k] is transmitted to the control device 40 from a rotational speed sensor (not shown) that measures the rotational speed of the internal combustion engine 1 (engine).
[0029] After obtaining the rotational speed Ne[k] (step S2), the control device 40 determines the fuel injection amount Qf[k](mm) from the fuel injection unit. 3 Step S3 obtains the / st). In this embodiment, since each of the combustion chambers CR1 to CR4 is provided with a fuel injection unit, the fuel injection amount Qf[k] refers to the total amount of fuel injected into each of the combustion chambers CR1 to CR4 at time t=kΔt. Information indicating the fuel injection amount Qf[k] is transmitted to the control device 40 from an injection amount sensor (not shown) that measures the amount of fuel injected from the fuel injection unit. Since the amount of fuel injected is controlled by the control device 40, the control device 40 can also obtain information indicating the fuel injection amount Qf[k] without receiving it from the injection amount sensor (not shown).
[0030] After obtaining the fuel injection quantity Qf[k] (step S3), the control device 40 measures the temperature T water [k] (K) of the cooling water that cools the internal combustion engine 1 (step S4). Information indicating the temperature T water [k] of the cooling water is transmitted from a temperature sensor (not shown) that measures the temperature of the cooling water to the control device 40.
[0031] After obtaining the temperature T water [k] of the cooling water (step S4), the control device 40 measures the temperature T air [k] (K) of the intake air that is sent from the upstream side of the connection portion of the intake pipe 10 with the reflux pipe 30 (step S5). Information indicating this temperature T air [k] of the intake air is transmitted from a temperature sensor (not shown) provided inside the portion of the intake pipe 10 upstream of the connection portion with the reflux pipe 30 to the control device 40.
[0032] After obtaining the temperature T air [k] of the intake air (step S5), the control device 40 measures the flow rate m air [k] (kg / s) of the intake air that is sent from the upstream side of the connection portion of the intake pipe 10 with the reflux pipe 30 (step S6). Information indicating this flow rate m air [k] of the intake air is transmitted from a flow rate sensor (not shown) provided inside the portion of the intake pipe 10 upstream of the connection portion with the reflux pipe 30 to the control device 40.
[0033] After obtaining the flow rate m air [k] of the intake air (step S6), the control device 40 measures the internal pressure P im [k] (Pa) of the intake pipe 10 (step S7). Information indicating the internal pressure P im [[ID=3[k] is obtained (step S8). Opening degree u of the recirculation control valve 38. EGR Information indicating [k] is transmitted to the control device 40 from an opening degree sensor (not shown) located near the recirculation control valve 38.
[0035] Opening degree u of the recirculation control valve 38 EGR After obtaining [k] (step S8), the reflux gas temperature estimation unit 45, which is part of the control device 40, uses a neural network to estimate the reflux gas temperature T EGR Estimate [k](K) (Step S9).
[0036] Figure 3 is a schematic diagram illustrating a neural network NN. The neural network NN has an input layer IL into which input variables are input, an output layer OL that outputs output variables, and an intermediate layer HL that performs various calculations between the input layer IL and the output layer OL. The reflux gas temperature estimation unit 45 estimates the reflux gas temperature T EGR Using a neural network NN that has been pre-trained to output [k], the temperature T of the refluxing gas is used. EGR Estimate [k] (step S9).
[0037] Temperature T of refluxing gas EGR When estimating [k], the reflux gas temperature estimation unit 45 determines the reflux gas temperature T EGR Variables that directly or indirectly affect [k] are input as input variables. The input variables input to the reflux gas temperature estimation unit 45 are rotational speed Ne[k], fuel injection amount Qf[k], and coolant temperature T water [k] and the intake air temperature T air [k] and the flow rate of the intake air m air [k] and the internal pressure P of the intake manifold 10. im [k] and the opening degree u of the recirculation control valve 38 EGR [k] and are used.
[0038] Temperature T of refluxing gas EGR After estimating [k] (step S9), the reflux gas flow rate estimation unit 47, which is part of the control device 40, uses a neural network NN to estimate the reflux gas flow rate m EGR[k](kg / s) is estimated (step S10). In detail, the reflux gas flow rate estimation unit 47, similar to the reflux gas temperature estimation unit 45, estimates the reflux gas flow rate m EGR Using a neural network NN that has been pre-trained to output [k], the flow rate m of the reflux gas is measured. EGR Estimate [k] (step S10).
[0039] Flow rate of recirculating gas m EGR When estimating [k], the reflux gas flow rate estimation unit 47, similar to the reflux gas temperature estimation unit 45, estimates the reflux gas temperature T EGR Variables that directly or indirectly affect [k] are input as input variables. The input variables input to the recirculating gas flow rate estimation unit 47 are the same as those input to the recirculating gas temperature estimation unit 45: rotational speed Ne[k], fuel injection amount Qf[k], and cooling water temperature T water [k] and the intake air temperature T air [k] and the flow rate of the intake air m air [k] and the internal pressure P of the intake manifold 10. im [k] and the opening degree u of the recirculation control valve 38 EGR [k] and are used.
[0040] Flow rate of recirculating gas m EGR After estimating [k] (step S10), the control device 40 determines the constant-pressure specific heat C of air. p , air [k](J / (kg·K)) Specific heat of refluxing gas at constant pressure C p , EGR [k](J / (kg·K)) and the constant-pressure specific heat C of the mixed gas p , im [k](J / (kg·K)) is obtained (step S11). The constant-pressure specific heat of each is pre-stored in the control device 40. Note that the constant-pressure specific heat C of the mixed gas p , im [k] is the temperature T of the refluxing gas. EGR [k] and the flow rate of reflux gas m EGR [k] and the intake air temperature T air [k] and the flow rate of the intake air m air [k] may be used to calculate the value each time the process in step S11 is performed.
[0041] After obtaining the specific heat at constant pressure for each component (step S11), the heat transfer amount estimation unit 41, which is part of the control device 40, uses a neural network NN to estimate the amount of heat transfer Q between the mixed gas and the wall of the intake pipe 10. loss,im [k] is estimated (step S12). In detail, the heat transfer amount estimation unit 41 estimates the heat transfer amount Q. loss,im Using a neural network NN that has been pre-trained to output [k], we use the heat transfer quantity Q. loss,im [k] is estimated (step S12). Note that the amount of heat transfer Q loss,im When [k] is a positive value, it means that heat is transferred from the gas mixture to the wall of the intake pipe 10, and the amount of heat transfer Q loss,im When [k] is a negative value, it means that heat is transferred from the wall of the intake pipe 10 to the gas mixture.
[0042] The input variables input to the heat transfer amount estimation unit 41 are rotational speed Ne[k], fuel injection amount Qf[k], and coolant temperature T. water [k] and the intake air temperature T air [k] and the flow rate of the intake air m air [k] and the temperature T of the refluxing gas EGR [k] and the flow rate m of the reflux gas EGR [k] is used. In addition to these variables, the input variables input to the heat transfer amount estimation unit 41 are rotational speed Ne[k-1], fuel injection amount Qf[k-1], and coolant temperature T water [k-1] and the intake air temperature T air [k-1] and the flow rate of the intake air m air [k-1] and the temperature T of the refluxing gas EGR [k-1] and the flow rate of the refluxing gas m EGR [k-1] is used.
[0043] The temperature of the intake pipe 10 wall is a variable used to estimate the amount of heat exchange between the mixed gas and the wall surface of the intake pipe 10, but it is difficult to directly measure this wall temperature on the vehicle. On the other hand, it is possible to indirectly calculate the change in the wall temperature of the intake pipe 10 at time t=kΔt. For such an indirect calculation, the amount of heat exchange with the gas (intake air and recirculating gas), the amount of heat exchange with the coolant, the amount of heat received from the cylinder head, and the heat capacity of the wall surface of the intake pipe 10 are used at time t=(k-1)Δt.
[0044] Variables related to the amount of heat exchange with the gas (intake air and return gas) are the intake air temperature T. air [k-1] and the flow rate of the intake air m air [k-1] and the temperature T of the refluxing gas EGR [k-1] and the flow rate of the refluxing gas m EGR It is thought to be [k-1]. The variables related to the amount of heat exchange with the cooling water are rotational speed Ne[k] and the temperature of the cooling water T water It is thought that [k] is the relevant variable for the amount of heat absorbed from the cylinder head. The variables related to the amount of heat absorbed from the cylinder head are rotational speed Ne[k], fuel injection amount Qf[k], and coolant temperature T. water [k] is considered to be the case. Therefore, as a substitute for the wall temperature of the intake manifold 10, the rotational speed Ne[k-1] at time t=(k-1)Δt, the fuel injection amount Qf[k-1], and the coolant temperature T are used. water [k-1] and the intake air temperature T air [k-1] and the flow rate of the intake air m air [k-1] and the temperature T of the refluxing gas EGR [k-1] and the flow rate of the refluxing gas m EGR [k-1] and are used as input variables to the heat transfer amount estimation unit 41.
[0045] Heat exchange amount Q loss,im After estimating [k] (step S12), the mixed gas temperature estimation unit 43, which is part of the control device 40, determines the amount of heat transfer Q loss,im Using [k], the temperature T of the mixed gas im [k](K) is estimated. At this time, the mixed gas temperature estimation unit 43 estimates the temperature T of the mixed gas based on the energy conservation law and the mass conservation law. im Estimate [k](K).
[0046] Temperature T of the mixed gas based on the energy conservation law and the mass conservation law im The estimation method of [k] will be described. Temperature T of the mixed gas im [k] is estimated using equations (1) to (3). Equation (1) is expressed as follows.
Equation
[0047] Equation (2) is expressed as follows.
Equation
[0048] Equation (3) is expressed as follows based on equations (1) and (2).
Equation
[0049] According to the control device 40 of the embodiment described above, the temperature of the mixed gas T im In estimating [k], the amount of heat transfer Q between the mixed gas and the wall of the intake pipe 10 is considered. loss,im Since [k] is taken into consideration (see the right-hand side of equation (3)), the temperature of the mixed gas T im [k] can be estimated with high accuracy. Therefore, the temperature T of the mixed gas can be estimated with high accuracy. im By suppressing the excess or deficiency of fuel injection amount caused by estimation errors of [k] and optimizing the fuel injection timing, it becomes possible to reduce the emissions of carbon monoxide, hydrocarbons, and nitrogen oxides emitted from the internal combustion engine 1, as well as reduce fuel consumption. In the simulation using the control device 40 of the embodiment, the temperature T of the mixed gas is im The coefficient of determination, which indicates the estimation accuracy of [k], was 0.984. On the other hand, the heat transfer amount Q loss,im The coefficient of determination without considering [k] was 0.965.
[0050] Furthermore, in the control device 40 of the embodiment, the mixed gas temperature estimation unit 43 estimates the temperature of the mixed gas T based on the energy conservation law and the mass conservation law. im By estimating [k], the temperature T of the mixed gas can be determined. im [k] can be estimated with greater accuracy.
[0051] Furthermore, in the control device 40 of the embodiment, a neural network NN capable of storing the nonlinear relationship between the input variable and the output variable in the training data is used to determine the amount of heat exchange Q between the mixed gas flowing inside the intake pipe 10, which has a highly nonlinear and extremely complex flow field, and the wall surface of the intake pipe 10. loss,im By estimating [k], the amount of heat transfer Q loss,im [k] can be estimated with high accuracy.
[0052] A neural network (NN) can estimate phenomena that are difficult to express explicitly with physical equations, provided there is a causal relationship between the input and output variables. On the other hand, since a neural network (NN) does not take physical laws into consideration, it is preferable to satisfy physical laws in order to improve the estimation accuracy when estimating other variables using the output variables estimated using a neural network (NN). In this regard, the control device 40 of the embodiment uses a neural network (NN) to estimate variables (reflux gas temperature T EGR [k] Flow rate of reflux gas m EGR [k] and the amount of heat transfer Q between the mixed gas and the wall of the intake pipe 10. loss,im [k]) and variables measured by various sensors (intake air temperature T air [k] and the flow rate of the intake air m air By substituting [k]) into equation (3) based on the laws of conservation of energy and conservation of mass, the temperature T of the mixed gas can be obtained. im [k] is estimated. To do this, by using equation (3), the temperature T of the mixed gas is estimated while satisfying the physical laws for various variables, including the variables estimated using the neural network NN. im Since [k] is being estimated, the temperature T of the mixed gas im [k] can be estimated with high accuracy.
[0053] Furthermore, in the control device 40 of the embodiment, the amount of heat exchange Q between the mixed gas and the wall surface of the intake pipe 10 is controlled. loss,im Since variables that directly or indirectly affect [k] are used as input variables to the neural network NN, the amount of heat exchange Q between the mixed gas and the wall of the intake pipe 10 is determined. loss,im [k] can be estimated with high accuracy. In the simulation using the control device 40 of the embodiment, the heat transfer amount Q loss,im The coefficient of determination, which indicates the estimation accuracy of [k], was 0.980.
[0054] Furthermore, in the control device 40 of the embodiment, the temperature T of the reflux gas EGR Since variables that directly or indirectly affect [k] are used as input variables to the neural network NN, the temperature T of the refluxing gas is also important. EGR[k] can be estimated with high accuracy. In the simulation using the control device 40 of the embodiment, the temperature T of the reflux gas is EGR The coefficient of determination, which indicates the estimation accuracy of [k], was 0.987.
[0055] Furthermore, in the control device 40 of the embodiment, the flow rate of the reflux gas m EGR Since variables that directly or indirectly affect [k] are used as input variables to the neural network NN, the flow rate of the reflux gas m EGR [k] can be estimated with high accuracy. In the simulation using the control device 40 of the embodiment, the flow rate m of the reflux gas EGR The coefficient of determination, which indicates the estimation accuracy of [k], was 0.996.
[0056] A characteristic of neural networks (NN) is that they have a relatively light computational load. In this respect, in the control device 40 of the embodiment, the amount of heat exchange Q between the mixed gas and the wall surface of the intake pipe 10 is loss,im [k], temperature of reflux gas T EGR [k] and the flow rate of reflux gas m EGR Each of [k] is estimated using a neural network NN. Therefore, since these three variables can be estimated with a relatively light computational load, the control device 40 can determine the temperature T of the mixed gas even on a vehicle equipped with an internal combustion engine 1. im [k] can be estimated accurately and with low computational load. Furthermore, if both the data acquired when the internal combustion engine 1 is in a transient state and the data acquired when the internal combustion engine 1 is in a steady state are used to train the neural network NN, then the temperature T of the mixed gas can be estimated in both the transient and steady states. im [k] can be estimated with high accuracy and low computational load.
[0057] Furthermore, in the control device 40 of the embodiment, the temperature T of the mixed gas im By accurately estimating [k], it becomes possible to accurately estimate the amount of gas flowing into the combustion chamber. (Gas flowing into the combustion chamber m) cylIt is known that this can be expressed as shown in equation (4) below. Note that the amount of gas m cyl This is the flow rate m of the mixed gas included in equation (1). im It is synonymous with [the above].
number
[0058] <Modified form of this embodiment> The present invention is not limited to the embodiments described above, and can be implemented in various forms without departing from its spirit, for example, the following modifications are also possible.
[0059] In the embodiment described above, the fuel injection amount Qf[k] referred to the total amount of fuel injected into each combustion chamber CR1 to CR4 at time t=kΔt, but it is not limited to this. For example, if fuel is injected into the intake pipe 10 from a fuel injection unit provided in the intake pipe 10 instead of providing a fuel injection unit in each of the combustion chambers CR1 to CR4, then the fuel injection amount Qf[k] refers to the amount of fuel injected into the intake pipe 10 at time t=kΔt.
[0060] In the embodiment described above, the temperature of the reflux gas T EGR [k] and the flow rate of reflux gas m EGR[k] was estimated using a neural network NN, but is not limited to this. The temperature T of the refluxing gas. EGR [k] and the flow rate of reflux gas m EGR At least one of [k] may be measured by a temperature sensor or a flow sensor installed inside the portion of the reflux pipe 30.
[0061] In the embodiment described above, the input variables input to the reflux gas temperature estimation unit 45 are the rotational speed Ne[k], the fuel injection amount Qf[k], and the cooling water temperature T. water [k] and the intake air temperature T air [k] and the flow rate of the intake air m air [k] and the internal pressure P of the intake manifold 10. im [k] and the opening degree u of the recirculation control valve 38 EGR [k] was used, but it is not limited to this. The input variables input to the reflux gas temperature estimation unit 45 are rotational speed Ne[k], fuel injection amount Qf[k], and coolant temperature T water [k] and the intake air temperature T air [k] and the flow rate of the intake air m air [k] and the internal pressure P of the intake manifold 10. im [k] and the opening degree u of the recirculation control valve 38 EGR It is sufficient that at least one of the variables [k] and is used.
[0062] In the embodiment described above, the input variables input to the reflux gas flow rate estimation unit 47 are the same as those input to the reflux gas temperature estimation unit 45: rotational speed Ne[k], fuel injection amount Qf[k], and cooling water temperature T. water [k] and the intake air temperature T air [k] and the flow rate of the intake air m air [k] and the internal pressure P of the intake manifold 10. im [k] and the opening degree u of the recirculation control valve 38 EGR [k] was used, but it is not limited to this. The input variables input to the recirculating gas flow rate estimation unit 47 are the same as those input to the recirculating gas temperature estimation unit 45: rotational speed Ne[k], fuel injection amount Qf[k], and cooling water temperature T water [k] and the intake air temperature T air [k] and the flow rate of the intake air mair [k] and the internal pressure P of the intake manifold 10. im [k] and the opening degree u of the recirculation control valve 38 EGR It is sufficient that at least one of the variables [k] and is used.
[0063] In the embodiment described above, the input variables input to the heat transfer amount estimation unit 41 are the rotational speed Ne[k], the fuel injection amount Qf[k], and the coolant temperature T. water [k] and the intake air temperature T air [k] and the flow rate of the intake air m air [k] and the temperature T of the refluxing gas EGR [k] and the flow rate m of the reflux gas EGR [k] was used, but it is not limited to this. The input variables input to the heat transfer amount estimation unit 41 are rotational speed Ne[k], fuel injection amount Qf[k], and coolant temperature T water [k] and the intake air temperature T air [k] and the flow rate of the intake air m air [k] and the temperature T of the refluxing gas EGR [k] and the flow rate m of the reflux gas EGR It is sufficient that at least one of the variables [k] and is used. The same applies to the input variables at time t=(k-1)Δt that are input to the heat transfer amount estimation unit 41.
[0064] The embodiments of this specification have been described above based on the embodiments and modifications described above. The embodiments described above are for the purpose of facilitating understanding of this specification and do not limit it. This specification may be modified and improved without departing from its spirit and the scope of the claims, and equivalents thereof are included in this specification. Furthermore, any technical features that are not described as essential in this specification may be deleted as appropriate.
[0065] The present invention can also be realized in the following forms. [Application Example 1] A control device for an internal combustion engine comprising: a combustion chamber for burning fuel and air; a fuel injection unit for injecting the fuel burned in the combustion chamber; an intake pipe for supplying air to the combustion chamber; an exhaust pipe for discharging post-combustion gas from the combustion chamber; and a recirculation pipe for recirculating at least a portion of the post-combustion gas as recirculating gas to the intake pipe, A heat transfer amount estimation unit estimates the amount of heat transfer between the intake air supplied from upstream of the connection point with the return pipe in the intake pipe, the mixed gas obtained by mixing the intake air with the return gas, and the wall surface of the intake pipe. A control device for an internal combustion engine, comprising: a mixed gas temperature estimation unit that estimates the temperature of the mixed gas using the amount of heat transfer estimated by the heat transfer amount estimation unit; and [Application Example 2] A control device for an internal combustion engine as described in Application Example 1, The aforementioned mixed gas temperature estimation unit is a control device for an internal combustion engine that estimates the temperature of the mixed gas based on the law of conservation of energy and the law of conservation of mass. [Application Example 3] A control device for an internal combustion engine as described in Application Example 1 or Application Example 2, The heat transfer amount estimation unit is a control device for an internal combustion engine that estimates the heat transfer amount using a neural network. [Application Example 4] A control device for an internal combustion engine described in any one of Application Examples 1 to 3, A control device for an internal combustion engine, wherein the input variables input to the heat transfer amount estimation unit include at least one of the following variables: the rotational speed of the internal combustion engine, the amount of fuel injected from the fuel injection unit, the temperature of the cooling water that cools the internal combustion engine, the temperature of the intake air, the flow rate of the intake air, the temperature of the recirculating gas, and the flow rate of the recirculating gas. [Application Example 5] A control device for an internal combustion engine described in any one of Application Examples 1 to 4, further comprising: The system includes a reflux gas temperature estimation unit that estimates the temperature of the reflux gas using a neural network, A control device for an internal combustion engine, wherein at least one of the following variables is used as an input variable to the recirculating gas temperature estimation unit: rotational speed, fuel injection amount, coolant temperature, intake air temperature, intake air flow rate, internal pressure of the intake manifold, and opening degree of a recirculation control valve that adjusts the flow rate of the recirculating gas. [Application Example 6] A control device for an internal combustion engine described in any one of Application Examples 1 to 5, further comprising: The system includes a reflux gas flow rate estimation unit that estimates the flow rate of the reflux gas using a neural network, A control device for an internal combustion engine, wherein at least one of the following variables is used as an input variable to the recirculating gas flow rate estimation unit: the rotational speed, the fuel injection amount, the temperature of the coolant, the temperature of the intake air, the flow rate of the intake air, the internal pressure of the intake manifold, and the opening degree of a recirculation control valve that adjusts the flow rate of the recirculating gas. [Explanation of Symbols]
[0066] 1…Internal combustion engine 10…Intake pipe 12... Air flow meter 14…Compressor 16…Intercooler 18... Throttle valve 19…Intake Manifold 20... Exhaust pipe 22…Exhaust manifold 24... Turbine 30…reflux pipe 36…EGR cooler 38…Refrigeration control valve 40…Control device 41…Heat exchange amount estimation part 43... Mixed gas temperature estimation unit 45...Reflux gas temperature estimation unit 47...Reflux gas flow rate estimation unit CH... Supercharger CR1~CR4... Combustion chamber HL…middle layer IL…Input layer NN...Neural Network NZ... Variable nozzle OL…Output layer SH... Shaft
Claims
1. A control device for an internal combustion engine comprising: a combustion chamber for burning fuel and air; a fuel injection unit for injecting the fuel burned in the combustion chamber; an intake pipe for supplying air to the combustion chamber; an exhaust pipe for discharging post-combustion gas from the combustion chamber; and a recirculation pipe for recirculating at least a portion of the post-combustion gas as recirculating gas to the intake pipe, A heat transfer amount estimation unit estimates the amount of heat transfer between the intake air supplied from upstream of the connection point with the return pipe in the intake pipe, the mixed gas obtained by mixing the intake air with the return gas, and the wall surface of the intake pipe. A control device for an internal combustion engine, comprising: a mixed gas temperature estimation unit that estimates the temperature of the mixed gas using the amount of heat transfer estimated by the heat transfer amount estimation unit; and
2. A control device for an internal combustion engine according to claim 1, The aforementioned mixed gas temperature estimation unit is a control device for an internal combustion engine that estimates the temperature of the mixed gas based on the law of conservation of energy and the law of conservation of mass.
3. A control device for an internal combustion engine according to claim 1 or claim 2, The heat transfer amount estimation unit is a control device for an internal combustion engine that estimates the heat transfer amount using a neural network.
4. A control device for an internal combustion engine according to claim 3, A control device for an internal combustion engine, wherein the input variables input to the heat transfer amount estimation unit include at least one of the following variables: the rotational speed of the internal combustion engine, the amount of fuel injected from the fuel injection unit, the temperature of the cooling water that cools the internal combustion engine, the temperature of the intake air, the flow rate of the intake air, the temperature of the recirculating gas, and the flow rate of the recirculating gas.
5. A control device for an internal combustion engine according to claim 4, further, The system includes a reflux gas temperature estimation unit that estimates the temperature of the reflux gas using a neural network, A control device for an internal combustion engine, wherein at least one of the following variables is used as an input variable to the recirculating gas temperature estimation unit: rotational speed, fuel injection amount, coolant temperature, intake air temperature, intake air flow rate, internal pressure of the intake manifold, and opening degree of a recirculation control valve that adjusts the flow rate of the recirculating gas.
6. A control device for an internal combustion engine according to claim 4, further, The system includes a reflux gas flow rate estimation unit that estimates the flow rate of the reflux gas using a neural network, A control device for an internal combustion engine, wherein at least one of the following variables is used as an input variable to the recirculating gas flow rate estimation unit: the rotational speed, the fuel injection amount, the temperature of the coolant, the temperature of the intake air, the flow rate of the intake air, the internal pressure of the intake manifold, and the opening degree of a recirculation control valve that adjusts the flow rate of the recirculating gas.