Electronic control device and flow measuring system

Abstract

The electronic control device of the present application is provided with: a flow rate calculation section that calculates the flow rate of intake air from an output signal of a flow rate measuring device attached to an intake pipe; a flow rate correction value calculation section that calculates an average value, a maximum value, a minimum value, and an amplitude of a signal of one or more frequencies included in the output signal of the flow rate measuring device and above a fundamental frequency of the output signal of the flow rate measuring device, for a prescribed period, of the flow rate of intake air calculated by the flow rate calculation section, and calculates a correction value for the flow rate of intake air from the calculation results; and a flow rate correction section that corrects the flow rate of intake air according to the correction value.

Description

Technical Field

[0001] This invention relates to an electronic control device and a flow measurement system. Background Technology

[0002] Previously, a control technology for an internal combustion engine was disclosed, which measures the air flow rate drawn into the engine using a flow sensor mounted in the engine's intake manifold, calculates the amount of air to be filled into the cylinder based on the measured intake air volume, and controls the fuel injection quantity and ignition timing based on the calculated air volume. For example, a technique was disclosed in which the detection error of the flow sensor increases as the pulsation of the air drawn into the engine increases, and therefore the detection error of the flow sensor is corrected based on the pulsation amplitude (for example, see Patent Document 1).

[0003] Existing technical documents

[0004] Patent documents

[0005] Patent Document 1: Japanese Patent Application Publication No. 2014-020212 Summary of the Invention

[0006] The problem the invention aims to solve

[0007] However, depending on the operating conditions of the internal combustion engine, sometimes corrections based solely on pulsation amplitude cannot completely correct the detection error of the flow sensor. Due to this flow sensor error, there are problems such as deterioration in the control accuracy of fuel injection quantity and ignition timing, as well as deterioration in the exhaust characteristics and fuel consumption rate of the internal combustion engine.

[0008] The present invention was made in view of the above circumstances, and its object is to be able to determine the intake air flow rate with high accuracy even under the operating conditions of an internal combustion engine where the intake air flow at the flow sensor location produces various pulsations.

[0009] Technical means to solve the problem

[0010] To address the aforementioned issues, an electronic control device according to one aspect of the present invention comprises: a flow calculation unit that calculates the flow rate of intake air based on the output signal of a flow measuring device mounted on an intake manifold; a flow correction value calculation unit that calculates the average value, maximum value, minimum value, and amplitude of a signal at or above the fundamental frequency of the output signal of the flow measuring device and including one or more frequencies in the output signal of the flow measuring device over a predetermined period of the flow calculation unit, and calculates a correction value for the flow rate of intake air based on the calculation results; and a flow correction unit that corrects the flow rate of intake air based on the correction value.

[0011] The effects of the invention

[0012] According to at least one aspect of the present invention, the detection error of the flow measuring device is appropriately corrected based on the average, maximum, and minimum values ​​of the output signal of the flow measuring device, and the amplitude information of one or more frequency components contained in the output signal of the flow measuring device. Therefore, even under operating conditions of an internal combustion engine where the airflow at the location of the flow measuring device produces various pulsations, the airflow rate can be determined with high accuracy. Thus, the deterioration of the exhaust characteristics or fuel consumption rate of the internal combustion engine that is a concern during periods of large pulsation can be prevented.

[0013] Other issues, structures, and effects not mentioned above will be clarified through the following description of the implementation methods. Attached Figure Description

[0014] Figure 1 This is a schematic diagram showing an example of the overall configuration of an internal combustion engine system that is the object of control of the electronic control device according to the first embodiment of the present invention.

[0015] Figure 2 This is a block diagram illustrating an example of the hardware configuration of an ECU.

[0016] Figure 3 It is a chart showing the operating range in which EGR is introduced within the operating range defined by the rotational speed and filling efficiency of the internal combustion engine.

[0017] Figure 4 It is a diagram showing the operating range of the Miller cycle within the operating range defined by the rotational speed and filling efficiency of the internal combustion engine.

[0018] Figure 5 It is a diagram showing the lift patterns of the intake and exhaust valves that achieve delayed Miller cycle closing and early Miller cycle closing.

[0019] Figure 6 This is a graph illustrating an example of the pulsation (intake pulsation behavior) of the flow sensor output signal when Miller cycle and EGR are implemented.

[0020] Figure 7 This is a conceptual diagram representing the airflow inside a flow sensor.

[0021] Figure 8 It is a graph showing the relationship between the average flow velocity of the main flow path and the average flow velocity of the bypass flow path measured by the flow sensor under different pulsating flow conditions.

[0022] Figure 9 It is a graph showing the differences in detection error (correction amount) of flow sensors under different pulsating flow conditions.

[0023] Figure 10 This is an example of a typical pulsation correction map used to correct the detection error of a flow sensor.

[0024] Figure 11 This is a chart showing examples of the maximum, minimum, and average values ​​of the output signals of flow sensors in different pulsating flows.

[0025] Figure 12 This is an example of a pulsation correction multidimensional mapping diagram in the first embodiment of the present invention, which uses the position of the average value between the maximum and minimum values ​​of the pulsating waveform as a parameter to correct the detection error of the flow sensor.

[0026] Figure 13 This is an explanatory diagram illustrating a method for extracting feature quantities of a pulsating waveform according to the second embodiment of the present invention.

[0027] Figure 14 It is a schematic diagram representing the weights and biases of each neuron that make up the neural network model.

[0028] Figure 15 This is an explanatory diagram illustrating a method for calculating the pulsation correction amount of a feature quantity based on a pulsating waveform using a neural network model according to the second embodiment of the present invention.

[0029] Figure 16 This is a block diagram illustrating an example of a pulsation correction logic installed in an ECU to correct the detection error of a flow sensor, according to the second embodiment of the present invention.

[0030] Figure 17 This is a flowchart illustrating an example of the sequence in which the ECU of the second embodiment of the present invention performs pulsation correction on the flow sensor.

[0031] Figure 18 It is a time-series diagram showing the changes in throttle opening, pressure sensor readings, and flow sensor readings as the vehicle accelerates from a throttle state to a boost state by opening the throttle valve.

[0032] Figure 19 This is a graph showing the relationship between the pulsation amplitude ratio and the pulsation correction value as effects of the first and second embodiments of the present invention.

[0033] Figure 20 This is a graph showing the relationship between the pulsation amplitude ratio and the pulsation correction value during high rotation, which are effects of the first and second embodiments of the present invention.

[0034] Figure 21 This is a block diagram illustrating an example of a pulsation correction logic installed in an ECU to correct the detection error of a flow sensor, according to the third embodiment of the present invention. Detailed Implementation

[0035] Hereinafter, examples of embodiments for carrying out the present invention will be described with reference to the accompanying drawings. In this specification and drawings, constituent elements having substantially the same function or structure are given the same reference numerals and repeated descriptions are omitted.

[0036] <First Implementation>

[0037] First, the overall configuration of the internal combustion engine system, which is the object of control of the electronic control device according to the first embodiment of the present invention, will be described.

[0038] [Overall Structure of an Internal Combustion Engine System]

[0039] Figure 1 This is a schematic diagram showing an example of the overall structure of an internal combustion engine system.

[0040] The internal combustion engine system includes: internal combustion engine 1, flow sensor 2, turbocharger 3, air bypass valve 4, intercooler 5, boost temperature sensor 6, throttle valve 7, intake manifold 8, boost pressure sensor 9, flow enhancement valve 10, intake valve 11, exhaust valve 13, fuel injection valve 15, spark plug 16, knock sensor 17, and crank angle sensor 18. The internal combustion engine system also includes exhaust bypass valve 19, air-fuel ratio sensor 20, exhaust gas catalytic converter 21, EGR (Exhausted Gas Recirculation) pipe 22, EGR cooler 23, EGR valve 24, temperature sensor 25, differential pressure sensor 26, and ECU (Electronic Control Unit) 27.

[0041] The intake airflow path (intake pipe 28) and the exhaust airflow path (exhaust pipe 29) are connected via the internal combustion engine 1. A flow sensor 2 and an intake air temperature sensor (not shown) are mounted in the intake airflow path. The turbocharger 3 consists of a compressor 3a and a turbine 3b. The compressor 3a is connected to the intake airflow path, and the turbine 3b is connected to the exhaust airflow path. The turbine 3b of the turbocharger 3 converts the energy of the exhaust gas from the internal combustion engine 1 into the rotational energy of the turbine blades. The compressor 3a of the turbocharger 3 compresses the intake air flowing in from the intake airflow path through the rotation of the compressor blades connected to the turbine blades.

[0042] Intercooler 5 is located downstream of compressor 3a of turbocharger 3 to cool the intake air whose temperature rises due to thermal compression by compressor 3a. Boost temperature sensor 6 is mounted downstream of intercooler 5 to measure the temperature (boost temperature) of the intake air cooled by intercooler 5.

[0043] Throttle valve 7 is located downstream of boost pressure sensor 6 and throttles the intake airflow path, controlling the amount of intake air flowing into the cylinders of internal combustion engine 1. Throttle valve 7 is an electronically controlled butterfly valve whose opening degree can be controlled independently of the driver's accelerator pedal input. Downstream of throttle valve 7 is intake manifold 8, on which boost pressure sensor 9 is assembled.

[0044] Alternatively, the intake manifold 8, located downstream of the throttle valve 7, and the intercooler 5 can be integrated into one unit. In this case, since the volume from downstream of the compressor 3a to the cylinder can be reduced, the responsiveness and controllability of acceleration and deceleration can be improved.

[0045] The flow enhancement valve 10 is located downstream of the intake manifold 8 and enhances the flow turbulence inside the cylinder by deflecting the air drawn into the cylinder. During exhaust recirculation combustion, as described later, the flow enhancement valve 10 is closed to promote and stabilize turbulent combustion.

[0046] The internal combustion engine 1 is provided with an intake valve 11 and an exhaust valve 13. The intake valve 11 and exhaust valve 13 each have a variable valve mechanism for continuously changing the phase of valve opening and closing. An intake valve position sensor 12 and an exhaust valve position sensor 14 for detecting the valve opening and closing phase are respectively mounted on the variable valve mechanisms of the intake valve 11 and exhaust valve 13. The cylinder of the internal combustion engine 1 has a direct fuel injection valve 15 that injects fuel directly into the cylinder. Alternatively, the fuel injection valve 15 can also be an intake port injection type, injecting fuel into the intake port.

[0047] A spark plug 16 is assembled on the cylinder of the internal combustion engine 1, with its electrode exposed inside the cylinder to ignite the combustible mixture via a spark. A knock sensor 17 is disposed in the cylinder block and detects knocking by detecting cylinder block vibrations caused by combustion pressure vibrations generated in the combustion chamber. A crankshaft angle sensor 18 is mounted on the crankshaft and outputs a signal corresponding to the crankshaft's rotational angle as a signal representing the crankshaft's rotational speed to the ECU 27.

[0048] The air-fuel ratio sensor 20 is located downstream of the turbine 3b of the turbocharger 3, outputting a signal to the ECU 27 indicating the detected exhaust composition, i.e., the air-fuel ratio. The exhaust purification catalyst 21, such as a three-way catalyst, is located downstream of the air-fuel ratio sensor 20, purifying harmful exhaust components such as carbon monoxide, nitrogen compounds, and unburned hydrocarbons through a catalytic reaction. Typically, the catalyst material uses platinum and rhodium, or a substance with palladium added.

[0049] The turbocharger 3 includes an air bypass valve 4 and an exhaust bypass valve 19. To prevent excessive pressure rise from downstream of the compressor 3a to upstream of the throttle valve 7, the air bypass valve 4 is positioned in the bypass path connecting the upstream and downstream of the compressor 3a. When the throttle valve 7 is abruptly closed during boost, the air bypass valve 4 is opened under the control of the ECU 27, allowing compressed intake air downstream of the compressor 3a to flow backward through the bypass path to the upstream of the compressor 3a. As a result, by immediately reducing the boost pressure, a phenomenon known as surge can be prevented, thus appropriately preventing damage to the compressor 3a.

[0050] An exhaust bypass valve 19 is configured in the bypass path connecting the upstream and downstream of the turbocharger 3b. The exhaust bypass valve 19 is an electrically operated valve controlled by the ECU 27, allowing free control of its opening relative to the boost pressure. When the opening of the exhaust bypass valve 19 is adjusted by the ECU 27 based on the boost pressure detected by the boost pressure sensor 9, a portion of the exhaust gas flows through the bypass path, thereby reducing the work done by the exhaust gas on the turbocharger 3b. As a result, the boost pressure can be maintained at the target pressure.

[0051] EGR pipe 22 connects the exhaust flow path downstream of exhaust purification catalyst 21 and the intake flow path upstream of compressor 3a, diverting exhaust gas downstream of exhaust purification catalyst 21 and returning it upstream of compressor 3a. An EGR cooler 23 installed in EGR pipe 22 cools the exhaust gas. EGR valve 24 is located downstream of EGR cooler 23 and controls the exhaust gas flow rate. A temperature sensor 25 detects the temperature of the exhaust gas flowing upstream of EGR valve 24, and a differential pressure sensor 26 detects the differential pressure between upstream and downstream of EGR valve 24.

[0052] ECU 27 is an example of an electronic control unit, which is a computational circuit that controls the various components of the internal combustion engine system or performs various data processing. Various sensors and actuators are communicatively connected to ECU 27. ECU 27 controls the operation of actuators such as throttle valve 7, fuel injection valve 15, intake valve 11, exhaust valve 13, and EGR valve 24. Furthermore, ECU 27 detects the operating state of the internal combustion engine 1 based on signals input from various sensors, and ignites spark plug 16 at a time determined according to the operating state. In this embodiment, flow sensor 2 and ECU 27 constitute a flow measurement system.

[0053] [ECU Hardware Configuration]

[0054] Next, the hardware configuration of ECU27 will be explained. Here, an example of the hardware configuration of the computer included in ECU 27 will be described.

[0055] Figure 2This is a block diagram illustrating an example of the hardware configuration of ECU 27.

[0056] The ECU 27 includes a control unit 31, a storage unit 32, and an input / output interface 33 interconnected via a system bus 36. The control unit 31 consists of a CPU (central processing unit) 31a, a ROM (Read Only Memory) 31b, and a RAM (Random Access Memory) 31c. The CPU 31a loads the control program stored in the ROM 31b into the RAM 31c and executes it to implement the various functions of the ECU 27. In other words, the control unit 31 is used as an example of a computer that controls the operation of an internal combustion engine system.

[0057] The storage unit 32 is an auxiliary storage device composed of a semiconductor memory or the like. For example, the storage unit 32 stores parameters used in the control program, transformation tables, and ripple correction mapping diagrams (see reference). Figure 10 , Figure 12 ), neural network model (refer to) Figure 15 This includes data such as data obtained from executing the control program. Alternatively, the control program can also be stored in the storage unit 32.

[0058] Input / output interface 33 is the interface for communication of signals and data with various sensors and actuators. ECU 27 includes an A / D (Analog / digital) converter (not shown) and drive circuitry, etc., for processing the input and output signals from various sensors. Input / output interface 33 can also function as an A / D converter. Furthermore, while a CPU is used as the processor, other processors such as an MPU (microprocessor unit) can also be used.

[0059] The following describes the control method of the internal combustion engine 1 for achieving low fuel consumption operation through the EGR system, Miller cycle system and exhaust bypass system of the internal combustion engine 1.

[0060] [Operating areas implementing EGR]

[0061] Figure 3 This is a graph showing the operating range of EGR (exhaust gas recirculation) within the operating range defined by the rotational speed and filling efficiency of the internal combustion engine 1. Filling efficiency is the ratio of the mass of air drawn into the cylinder in one cycle to the standard mass of air corresponding to the cylinder volume. Figure 3 The chart below is an example of importing cold EGR; the horizontal axis represents rotation speed, and the vertical axis represents fill efficiency. The area enclosed by the thick dashed line is the cold EGR import area.

[0062] The operating range of internal combustion engine 1 can be roughly divided into a non-turbocharged region and a turbocharged region. In the non-turbocharged region, the filling efficiency is controlled by the throttle valve. In the turbocharged region, the throttle valve is opened, and the boost pressure is controlled by the exhaust bypass valve, thereby controlling the filling efficiency. In this way, by adjusting the engine torque between the non-turbocharged and turbocharged regions, pump losses in internal combustion engine 1 can be reduced, achieving low fuel consumption operation. The thin dashed lines represent isobars (line of equal flow rate of fresh air).

[0063] Furthermore, the internal combustion engine 1 shown in this embodiment is equipped with an EGR system. From the higher load conditions (high filling efficiency) in the non-turbocharged region of the internal combustion engine 1 to the turbocharged region, the EGR gas cooled by the EGR cooler is returned to the cylinder, thereby diluting the air drawn into the cylinder with EGR gas as an inert gas. As a result, improper combustion, which is prone to occur under high load conditions, called knock, can be suppressed. Moreover, since knock can be suppressed, the ignition timing (advance angle, retarder angle) can be appropriately controlled, and low fuel consumption operation can be achieved.

[0064] [The operating area implementing the Miller cycle]

[0065] Figure 4 This is a diagram showing the operating region where the Miller cycle is implemented, within the operating range defined by the rotational speed and filling efficiency of the internal combustion engine 1. The area enclosed by the thick dashed line is the Miller cycle introductory region.

[0066] In the lower flow rate operating range of the internal combustion engine 1, the throttle valve 7 is further controlled towards the closing side to reduce the amount of air drawn into the cylinder. This tends to increase pump losses. By shifting the intake valve closing time from bottom dead center to the advance or retard side, the piston's compression work can be reduced, achieving the Miller cycle. Furthermore, if the intake air volume is controlled by controlling the intake valve phase instead of the throttle valve 7, the throttle valve 7 can be set further towards the open side, reducing pump losses. Through the effects of the Miller cycle and the reduction in pump losses, low fuel consumption operation can be achieved.

[0067] [Intake and exhaust valve lift modes]

[0068] Figure 5 This is a diagram showing the lift patterns of the intake valve 11 and exhaust valve 13 in achieving delayed Miller cycle closing and early Miller cycle closing.

[0069] If a configuration that can change the phase of the intake valve 11 is adopted, and the intake valve closing time is set to an advance or delay side based on bottom dead center, the amount of air drawn into the cylinder increases or decreases. In the delayed-closing Miller cycle shown in the upper part of the figure, the gas temporarily drawn into the cylinder is blown back into the intake manifold 28 after bottom dead center, thereby suppressing the amount of air drawn into the cylinder. On the other hand, in the advanced-closing Miller cycle shown in the lower part of the figure, the amount of air drawn into the cylinder is suppressed by closing the intake valve midway through the gas being drawn into the cylinder.

[0070] In the internal combustion engine system of this embodiment, a variable intake valve phase mechanism is used to realize the Miller cycle. However, a variable valve timing and lift mechanism or a continuously variable phase and lift mechanism that can switch the opening and closing period and lift of the intake valve can also be used to realize the Miller cycle.

[0071] [Intake pulsation behavior when implementing Miller cycle and EGR]

[0072] Figure 6 This is a diagram illustrating an example of the pulsation (intake pulsation action) of the output signal of flow sensor 2 when Miller cycle and EGR are implemented. Figure 6 The horizontal axis represents time [s], and the vertical axis represents flow velocity [m / s]. Furthermore, in this specification, the output signal of the flow sensor 2 is also referred to as the "flow signal". Additionally, the shape of the output signal of the flow sensor 2 is sometimes described as a "pulsating waveform".

[0073] As shown in the upper part of the figure, the internal combustion engine 1 only intermittently intakes air during the intake stroke of each cylinder, thus generating pulsations within the intake manifold 28. Particularly in the low-rotation, high-load region, there is a tendency to generate pulsations with large amplitude ratios at low frequencies, which is a major cause of the deterioration in the detection accuracy of the flow sensor 2. The pulsation amplitude ratio is the ratio of the amplitude (between maximum and minimum values) of the fundamental frequency component of the flow signal to the average value of that signal. Under low-rotation, high-load conditions where the pulsation amplitude increases relative to the average flow velocity, there are moments when the flow direction indicates countercurrent.

[0074] The middle section of the diagram represents the intake pulsation behavior (pulsation A) of flow sensor 2 when the Miller cycle (delayed shutdown) is implemented. For example... Figure 5 As shown, in the delayed-shutdown Miller cycle, the gas temporarily drawn into the cylinder is blown back into the intake manifold 28. Furthermore, in the delayed-shutdown Miller cycle, the throttle valve 7 is set to the open side compared to the normal cycle. Due to these effects, pulsations generated within the cylinder during the delayed-shutdown Miller cycle are more likely to reach the flow sensor 2.

[0075] The lower section of the figure shows the intake pulsation behavior (pulsation B) of flow sensor 2 during EGR introduction. The exhaust pulsation is a larger pulsation (higher flow rate and amplitude) compared to the intake side, and the pulsation propagates to the intake side through EGR pipe 22. If the EGR valve is set to the open side to allow more EGR gas to flow back, there is a tendency for the pulsation of flow sensor 2 to increase.

[0076] As described above, by implementing Miller cycle or EGR, turbulence is generated in the flow within the intake manifold 28. Therefore, compared to the pulsation caused by the intake stroke of a normal cycle, the pulsation when Miller cycle or EGR is implemented becomes a pulsation with a higher frequency component. In such pulsation phenomena, not only the pulsation amplitude rate, but also the pulsation frequency and the pulsation waveform determined by their combination are important.

[0077] [Airflow inside the flow sensor]

[0078] Figure 7 This is a conceptual diagram representing the airflow inside flow sensor 2.

[0079] The flow sensor 2 has a bypass flow path, within which a sensor element for detecting flow velocity is installed. By optimizing the shape of the bypass flow path, dust or water can be prevented from adhering to the sensor element. The flow sensor 2 outputs a voltage signal corresponding to the flow rate of the main flow in the portion of the intake pipe 28 where the flow sensor 2 is mounted, by detecting the heat dissipation caused by the local flow of the sensor element, which is mainly composed of a heating resistor. As shown in the figure, the shape (length L, l, inner diameter D, d), shape loss coefficient (Cp, cp), and friction loss coefficient (Cf, cf) of the flow field are different in the main flow and the bypass flow, thus becoming flow fields based on different kinematic equations. When the flow rate of the main flow path is U and the flow rate of the bypass flow path is u, the flow velocity dU / dt of the main flow path and the flow velocity du / dt of the bypass flow can be expressed by the following equations (1) and (2), respectively.

[0080]

[0081]

[0082] The relationship between the average flow velocity of the main flow path and the average flow velocity of the bypass flow path

[0083] Figure 8 This is a graph showing the relationship between the average flow velocity of the main flow path and the average flow velocity of the bypass flow path measured by flow sensor 2 under different pulsating flow conditions. Conditions 1 to 3 are different operating conditions. However, in any of the pulsating flows under conditions 1 to 3, the average flow velocity and pulsation amplitude ratio of the main flow path are the same for each condition.

[0084] The bypass flow path has a smaller inner diameter (d) than the main flow path and includes a tortuous shape, resulting in greater pressure loss. Furthermore, due to flow deviations in the main or bypass flow path, resonance, and the response delay of flow sensor 2, the characteristics vary with frequency, absolute value of flow velocity, and flow direction. Therefore, as shown in the figure, the flow velocity in the main flow path differs from the flow velocity in the bypass flow path measured by flow sensor 2, causing an error. As shown in conditions 1-3, even if the average flow velocity of the main flow path is the same (20 m / s) over a specified period (e.g., three cycles), if the pulsation waveform is different, the average flow velocity of the main flow path over the specified period will be different from the average flow velocity of the bypass flow path measured by flow sensor 2. Therefore, a pulsation correction amount corresponding to the pulsation conditions (influencing factors) is required.

[0085] [The average flow rate of mainstream flow rates and the average flow rate detected by flow sensors]

[0086] Figure 9 This is a graph showing the differences in detection error (correction amount) of the flow sensor under different pulsating flow conditions. The horizontal axis represents the average flow velocity detected by flow sensor 2, and the vertical axis represents the average flow velocity of the main flow.

[0087] Even if the average flow velocity of the mainstream is the same, the average flow velocity detected by flow sensor 2 is different for each pulsation condition 1 to 3. Therefore, under each pulsation condition 1 to 3, the correction amounts cv1 to cv3 corresponding to the detection error of flow sensor 2 relative to the average flow velocity of the mainstream are different.

[0088] [Typical pulsation correction map]

[0089] Figure 10 This is an example of a typical pulsation correction map that corrects for detection errors in flow sensors caused by pulsation. In the pulsation correction map, the pulsation frequency [Hz] is set horizontally, and the pulsation amplitude ratio [%] is set vertically.

[0090] In the pulsation correction mapping matching, the fundamental frequency (=pulsation frequency) is determined based on the rotational speed of the internal combustion engine 1, and frequencies above the fundamental frequency are considered pulsation components. The pulsation amplitude ratio is calculated based on the amplitude (between maximum and minimum values) of the fundamental frequency signal obtained by a low-pass filter that allows frequency bands below the fundamental frequency to pass through, and the average value of that signal. The pulsation correction amount is recorded in the pulsation correction mapping 1000 with the pulsation frequency and pulsation amplitude ratio as axes. When mounted in a vehicle, the pulsation frequency and pulsation amplitude ratio are calculated based on the detection value of the flow sensor 2 and its rotational speed, and the detection value of the flow sensor 2 is corrected according to the corresponding pulsation correction amount.

[0091] like Figure 8 and Figure 9As mentioned above, the matching value for pulsation correction varies depending on the pulsation waveform; therefore, a multi-dimensional pulsation correction mapping map is needed for each influencing factor affecting the pulsation waveform. Figure 10 The lower section shows the pulsation correction multidimensional mapping diagram 1100. The pulsation correction multidimensional mapping diagram 1100 consists of a pulsation correction mapping diagram 1000 for a baseline condition, a pulsation correction mapping diagram 1010 for a certain Miller cycle condition, a pulsation correction mapping diagram 1020 for a certain EGR condition, and a pulsation correction mapping diagram 1030 for differences in external air conditions. For example, the baseline condition is a condition where Miller cycle control and EGR control are not performed under certain external air conditions.

[0092] For example, external air conditions affect the speed of sound waves, and are therefore a factor influencing the pulsating waveform. Therefore, the pulsation correction multidimensional mapping 1100 needs to include pulsation correction mappings 1030 for different external air conditions, such as temperature or atmospheric pressure.

[0093] Furthermore, since intermediate states exist in Miller cycle control and EGR control, appropriate interpolation is required for these intermediate states. For example, when the EGR condition in the pulsation correction map 1020 is an EGR rate of 30%, the pulsation correction amount for an EGR rate of 10% as an intermediate state is obtained by interpolation using the correction amounts of the pulsation correction map 1000 with the reference condition (EGR rate of 0%) and the pulsation correction map 1020 with the EGR condition of 30%. Regarding external air conditions, the pulsation correction amount suitable for the current external air conditions can also be obtained through interpolation.

[0094] <First Implementation>

[0095] Hereinafter, as a first embodiment, a method for correcting the detection error of the flow sensor 2 caused by pulsation phenomenon based on the position of the average value between the maximum and minimum values ​​of the pulsating waveform will be described.

[0096] [Maximum, minimum, and average values ​​of the flow sensor output signal]

[0097] Figure 11 This is a graph showing the maximum, minimum, and average values ​​of the output signal of flow sensor 2 under different pulsating flow conditions.

[0098] like Figure 11 As shown, based on the waveform (pulsating waveform) of the output signal of flow sensor 2, the relative position of the average value with respect to the maximum and minimum values, i.e., the position of the average value between the maximum and minimum values, will change. Figure 11 In the upper segment of the pulsating waveform, the average value (20 m / s) is located close to the minimum value. Additionally, in Figure 11In the lower segment of the pulsating waveform, the minimum value is smaller than that in the upper segment, but the waveform near the minimum value becomes sharper, so the position of the average value (20m / s) is close to the maximum value.

[0099] Therefore, in this embodiment, a pulsation correction map is generated by utilizing the fact that the position of the average value between the maximum and minimum values ​​of the waveform (pulsating waveform) of the output signal of the flow sensor 2 changes. Specifically, the position of the average value between the maximum and minimum values ​​of the pulsating waveform of the flow detected by the flow sensor 2 is used as a parameter to generate a pulsation correction map instead of... Figure 10 The next section contains a pulsation correction multidimensional mapping diagram, Figure 1100, for each influencing factor affecting the pulsation waveform.

[0100] [A pulsation correction map considering the position of the average value relative to the maximum and minimum values]

[0101] Figure 12 This is an example of a pulsation correction multidimensional mapping diagram that uses the position of the average value between the maximum and minimum values ​​of the pulsating waveform as a parameter to correct the detection error of flow sensor 2 caused by pulsation. Figure 12 In the pulsation correction multidimensional mapping diagram 1200 shown, the pulsation frequency [Hz] is set horizontally, and the pulsation amplitude ratio [%] is set vertically. Here, in the pulsation waveform, the position of the minimum value is set to 0%, and the position of the maximum value is set to 100%. The pulsation correction multidimensional mapping diagram 1200 includes a pulsation correction mapping diagram 1210 where the average position between the maximum and minimum values ​​is 20%, a pulsation correction mapping diagram 1220 where the average position between the maximum and minimum values ​​is 50%, and a pulsation correction mapping diagram 1230 where the average position between the maximum and minimum values ​​is 80%.

[0102] According to the first embodiment described above, for each position of the average value between the maximum and minimum values ​​of the output signal of the flow sensor 2, a pulsation correction map 1210-1230 with the pulsation frequency and pulsation amplitude ratio as axes is prepared. Correction values ​​are obtained from these pulsation correction maps 1210-1230 to correct the detection error of the flow sensor 2. Therefore, even under operating conditions of the internal combustion engine 1 where Miller cycle control, EGR control, etc., are implemented, and various pulsations occur in the airflow at the flow sensor position, the intake air flow can be determined with high accuracy. Thus, the deterioration of the exhaust characteristics and fuel consumption rate of the internal combustion engine 1, which are concerns about large pulsations, can be prevented.

[0103] Furthermore, according to this embodiment, a pulsation correction multidimensional mapping diagram 1100 is prepared for each influencing factor affecting the pulsation waveform (see reference). Figure 10 Compared to the case of ), it can reduce the number of pulsation correction maps.

[0104] <Second Implementation Method>

[0105] Next, as a second embodiment, refer to Figures 13 to 15 This paper explains the method for calculating the pulsation correction amount of feature quantities based on pulsating waveforms using a neural network model.

[0106] [Methods for Extracting Pulsating Feature Quantities]

[0107] Figure 13 This is an explanatory diagram illustrating the method for extracting feature quantities from pulsating waveforms.

[0108] for Figure 13 The output signal (flow velocity data) of flow sensor 2 shown in the upper section is used for... Figure 13 The lower section shows the extraction of characteristic quantities from the pulsating waveform. As an example, Figure 13 The waveform (pulsating waveform) of the output signal of flow sensor 2 shown in the upper section is similar to... Figure 8 The same waveform appears in the lower segment. The average flow velocity of the main flow path (U with bar on the upper side) and the average flow rate detected by flow sensor 2 (u with bar on the upper side) show a difference.

[0109] In the extraction of pulsation features, the average value μ, maximum value max, and minimum value min of the output signal of the flow sensor 2 during the period specified by the rotational speed of the internal combustion engine 1 are calculated, and then filtered by bandpass filters (1) to (4) of different frequencies. By applying the above-mentioned bandpass filters (1) to (4) to the detection results of the flow sensor 2, the amplitudes σ1 to σ4 of each waveform in different frequency bands are obtained. These pulsation features are then correlated with the pulsation correction amount δ (U of the upper band bar / u of the upper band bar).

[0110] For example, the bandpass of bandpass filter (1) is the bandpass containing the fundamental frequency contained in the output signal of flow sensor 2. Additionally, the passbands of bandpass filters (2), (3), and (4) are each the bandpasses containing the frequencies of the second, third, and fourth harmonics. However, the passband of bandpass filter (1) may not be the bandpass containing the fundamental frequency. Similarly, the passbands of bandpass filters (2) to (4) may not be the bandpasses containing harmonic frequencies.

[0111] In addition, four bandpass filters (1) to (4) are set, but only one or more bandpass filters are required. Alternatively, the bandpass filter to be used can be selected from multiple bandpass filters according to specified rules, or a fixed number of bandpass filters can be used. For example, in Figure 13Multiple bandpass filters are defined, but only one bandpass filter can actually perform the filtering process. However, by using multiple bandpass filters, the accuracy of the ripple correction is improved.

[0112] [Weights, biases, and activation functions of each neuron]

[0113] Figure 14 It is a schematic diagram representing the weights and biases of each neuron that make up the neural network model.

[0114] A neural network model is a mathematical model that mimics the structure of neural circuits in the human brain. Neural networks are commonly used in machine learning as a means of deep learning. For example, backpropagation can be applied in machine learning algorithms. This embodiment utilizes a neural network, but it is not limited to this example as long as machine learning can be used to correct the detection error of the flow sensor 2.

[0115] like Figure 14 As shown, weights w and biases b are assigned to each neuron (unit) constituting the neural network model. Inputs a1~a n The inputs are fed into n neurons and multiplied by the weights w1 to w2 assigned to each neuron. n Then, in the neurons of the next layer, the neurons are multiplied by weights w1 to w... n Input a1~a n An addition operation (combination) is performed, and an output z is obtained by assigning a bias b to the result of the addition operation. The output of the next layer of neurons is a, represented by the function f(z).

[0116] In addition, an activation function is defined for each neuron. This activation function may contain appropriately chosen logistic functions (sigmoid function) or ramp functions (ReLU function). Figure 14 The diagram illustrates an example where neurons are more activated and their output y (= f(x)) is closer to 1 the larger the input x is than 0, while neurons are less activated and their output is closer to 0 the larger the input x is than 0. For example, when the input x is "5", the activation function outputs y as "1", and "1" is output from that neuron to the next layer of neurons.

[0117] [Neural Network]

[0118] Figure 15 This is an explanatory diagram illustrating the method of calculating the pulsation correction amount of feature quantities based on pulsating waveforms using a neural network model.

[0119] A neural network model consists of multiple neurons forming a layer, with intermediate layers between the input and output layers. By increasing the number of neurons or intermediate layers, more complex input-output relationships can be approximated. A trade-off exists between approximation accuracy and model size; a balance must be struck between these two factors.

[0120] Set the input layer separately Figure 13 The pulsation feature shown, matching the output layer with a set pulsation correction amount, can approximate the input-output relationship through machine learning on the weights w and offset b (with a teacher). The pulsation feature is the specified period of the output signal of the flow sensor 2. Figure 16 The average value μ, maximum value max, minimum value min, and amplitude σ1 to σ4 of each frequency component of the pulsating waveform during the feature detection period are used. The matching result of the pulsation correction is the pulsation correction δ. The learned model, having undergone such learning, performs operations based on the learned content when given pulsating feature values, and outputs the pulsation correction as the result of the operations.

[0121] In this embodiment, the position of the average value between the maximum and minimum values ​​of the pulsating waveform, the amplitude of one or more frequency components contained in the pulsating waveform, and the pulsation correction amount are correlated with the use of... Figure 12 Similarly, in the case of the pulsation correction multidimensional mapping diagram 1200 shown, it is possible to determine the pulsation correction amount corresponding to the difference in position from the average value.

[0122] [Pulse Correction Logic]

[0123] Figure 16 This is a block diagram illustrating an example of pulsation correction logic installed in ECU 27 to correct the detection error of flow sensor 2. ECU 27 and flow sensor 2 constitute a flow measurement system.

[0124] Here, we will first explain the principle of the hot-wire flow sensor used in flow sensor 2. The hot-wire flow sensor's main component is a heated resistor placed within the airflow being measured. In the hot-wire flow sensor, a bridge circuit is constructed such that the current flowing through the heated resistor increases when the amount of air drawn in is large, and conversely, decreases when the amount of air drawn in is small. The hot-wire flow sensor extracts the airflow as a voltage signal based on the current flowing through the heated resistor.

[0125] ECU 27 includes an A / D converter 1601, a voltage-flow conversion unit 1602, a rotational speed calculation unit 1603, a flow correction value calculation unit 1615, a voltage-pressure conversion unit 1616, and a flow correction unit 1620. The flow correction value calculation unit 1615 includes bandpass filters 1604-1607, a characteristic quantity detection period setting unit 1608, a maximum / minimum / average value calculation unit 1609, an amplitude detection unit 1610-1613, and a correction value derivation unit 1614.

[0126] The A / D conversion unit 1601 converts the analog voltage signals output from the flow sensor 2 and the boost pressure sensor 9 into digital voltage signals using an A / D converter. The A / D conversion unit 1601 outputs the digital voltage signals to the voltage-flow conversion unit 1602 and the voltage-pressure conversion unit 1616.

[0127] The voltage-flow conversion unit 1602 converts the digital voltage signal into a flow signal (flow value) through a voltage / flow conversion table, and outputs the calculation result to the bandpass filters 1604-1607, the maximum / minimum / average value calculation unit 1609, and the flow correction unit 1620 of the flow correction value calculation unit 1615.

[0128] In this embodiment, a voltage signal corresponding to the air volume is output as a voltage value. However, there is also a method where the voltage signal is converted into a frequency signal by a voltage / frequency conversion circuit (not shown) and then output. When the frequency signal after voltage / frequency conversion is input, the period of the signal is measured by inputting it through a port of the CPU 31a, thereby inputting the period or a value converted from the period to the frequency. The air volume conversion table referenced by the voltage-flow conversion unit 1602 retrieves values ​​from a pre-stored value based on the period or frequency. If no suitable value is found, interpolation is performed to convert the value into the detected air volume.

[0129] Figure 16 The characteristic curve of the voltage-flow conversion unit 1602 shown represents the relationship between the intake air volume and the output signal of a typical hot-wire flow sensor. This characteristic curve is a nonlinear curve showing a low voltage output signal when the intake air volume is low and a high voltage output signal when the intake air volume is high. The reason for using a nonlinear characteristic is that when converting the detection signal from the heating resistor into the air volume, the following formula (3), known as the Wang formula, is mainly used for the air volume Q.

[0130]

[0131] Here, Ih is the current value of the heating resistor, Rh is the resistance value of the heating resistor, Th is the surface temperature of the heating resistor, Ta is the air temperature, Q is the air volume, and α and β are constants determined by the specifications of the heating resistor. Generally, since the current value Ih of the heating resistor is controlled to keep (Th-Ta) constant, the air volume is converted into a voltage value V based on the voltage drop across the resistor for detection. As a result, the voltage value V becomes a fourth-order function. Therefore, when the detection signal from the heating resistor is converted into the air volume, the curvature of the fourth-order curve, i.e., the relationship between the output and the air volume, becomes non-linear.

[0132] The voltage-pressure conversion unit 1616 converts a digital voltage signal into a pressure signal using a voltage / pressure converter, and outputs the conversion result to the correction amount derivation unit 1614 of the flow correction value calculation unit 1615. Furthermore, similar to the voltage-flow conversion unit 1602, the voltage-pressure conversion unit 1616 outputs a voltage signal corresponding to the pressure as a voltage value. However, the voltage-pressure conversion unit 1616 may also be configured to obtain the pressure signal from a frequency signal converted from a digital voltage signal by a voltage / frequency conversion circuit (not shown).

[0133] The rotational speed calculation unit 1603 calculates the rotational speed of the internal combustion engine 1 based on the output signal of the crank angle sensor 18, and outputs the calculation result to the characteristic quantity detection period setting unit 1608 of the flow correction value calculation unit 1615.

[0134] Bandpass filters 1604-1607 perform bandpass filtering on the flow signal input from the voltage-flow conversion unit 1602, applying different passbands. Bandpass filters 1604-1607 then output the filtered flow signal to the corresponding amplitude detection units 1610-1613. Bandpass filters 1604-1607 are equivalent to... Figure 13 The bandpass filters (1) to (4) are used. By having multiple bandpass filters 1604 to 1607, various characteristic quantities of the output signal (pulsating waveform) of the flow sensor 2 can be used, and the inference accuracy of the pulsation correction quantity is improved.

[0135] in addition, Figure 16 Examples of four bandpass filters 1604 to 1607 are shown, but the number of bandpass filters can be more than four, i.e., 1 to 3 or more than 5. In the following description, bandpass filters 1604 to 1607 are sometimes referred to as bandpass filters (1) to (4).

[0136] The characteristic quantity detection period setting unit 1608 determines the characteristic quantity detection period based on the rotational speed input from the rotational speed calculation unit 1603, and outputs the determined characteristic quantity detection period to the maximum / minimum / average value calculation unit 1609 and the amplitude detection units 1610-1613. The characteristic quantity detection period is a period synchronized with the intake interval of the internal combustion engine 1, or an integer multiple of that period. For example, in a four-stroke engine, it is preferable to set the characteristic quantity detection period of each cylinder to the period of one or more strokes of the internal combustion engine 1. Figure 13 An example of extracting pulsating characteristic quantities during three strokes is shown. The period for characteristic quantity detection is determined by the rotational speed and number of cylinders of the internal combustion engine 1, but it can also be at intervals of a specified crank angle.

[0137] The maximum / minimum / average value calculation unit 1609 calculates the average, maximum, and minimum values ​​of the characteristic quantity of the output signal (pulsating waveform) of the flow sensor 2 during the detection period, and outputs the calculation result to the correction quantity derivation unit 1614. In addition to time averaging, other methods for calculating the average value include arithmetic averaging, filtering, and the DC component of Fourier transform. For example, filtering may include weighted averaging or harmonic averaging.

[0138] Amplitude detection units 1610 to 1613 detect the amplitude of the flow sensor signal that has passed through each bandpass filter 1604 to 1607 during the characteristic quantity detection period, and output the detection result to correction quantity derivation unit 1614. Hereinafter, amplitude detection units 1610 to 1613 are sometimes referred to as amplitude detection units (1) to (4).

[0139] Thus, the flow correction value calculation unit 1615 includes: one or more bandpass filters 1604-1607 corresponding to any one of the frequencies contained in the output signal of the flow sensor 2, and one or more amplitude detection units 1610-1613 corresponding to the one or more bandpass filters, which calculate the amplitude of the output signal after passing through the corresponding bandpass filter. Therefore, multiple frequency components can be extracted from the output signal of the flow sensor 2, and the amplitude of each frequency component can be calculated.

[0140] The correction quantity derivation unit 1614, based on the calculation results of the maximum / minimum / average value calculation unit 1609 and the detection results of the amplitude detection units (1) to (4), uses... Figure 15 The neural network model (learned model) described herein calculates the pulsation correction amount and outputs the calculation result to the flow correction unit 1620. Additionally, the correction amount derivation unit 1614 calculates the maximum value, minimum value, and amplitude of the pressure signal input from the voltage-pressure conversion unit 1616, and determines whether execution is permissible based on the calculation result. Figure 17 The pulsation correction is shown.

[0141] The flow correction unit 1620 corrects the flow rate before correction based on the flow signal (flow rate before correction) input from the voltage-flow conversion unit 1602 and the pulsation correction amount input from the correction amount derivation unit 1614. The ECU 27 uses the intake air flow rate (corrected flow rate) corrected by the flow correction unit 1620 to control the fuel injection amount, ignition timing, etc.

[0142] The flow correction value calculation unit 1615 and the flow correction unit 1620 perform calculations at a cycle synchronized with the intake interval (stroke) of the internal combustion engine 1. Thus, for a cycle synchronized with the intake interval of the internal combustion engine, such as each stroke, a pulsating correction value is calculated, and the intake air flow rate is corrected based on this correction value. Alternatively, the flow correction unit 1620 and the flow correction unit 1620 can also be configured as a single functional block.

[0143] By having a function to correct for detection errors caused by the pulsation of the flow sensor 2, the detection error of the flow sensor can be appropriately corrected even under the operating conditions of the internal combustion engine 1, where the flow sensor 2 is prone to detection errors due to pulsation. Therefore, the detection accuracy of the flow sensor 2 is ensured, and the accuracy of air-fuel ratio control is improved. Furthermore, by performing high-precision air-fuel ratio control, the deterioration of the exhaust gases from the internal combustion engine 1 can be prevented.

[0144] Furthermore, as mentioned above, the pulsation correction amount can also be based on the pulsation amplitude ratio of each pulsation frequency as the pulsation correction multidimensional mapping diagram 1200 (refer to). Figure 12 The correction amount is obtained from the rotation speed calculation unit 1603 or the feature quantity detection period setting unit 1608. Additionally, the maximum / minimum / average value calculation unit 1609 calculates the position of the average value between the maximum and minimum values ​​of the flow signal and the pulsation amplitude ratio, and outputs the calculation result to the correction amount extraction unit 1614. The correction amount extraction unit 1614 retrieves the pulsation correction multidimensional mapping map 1200 based on the pulsation frequency, pulsation amplitude ratio, and the position of the average value, and determines the appropriate pulsation correction amount.

[0145] [Pulse Correction Processing]

[0146] Figure 17 This is a flowchart illustrating the sequence of pulsation corrections performed by ECU27 on flow sensor 2. Figure 17 The processing steps shown are executed at regular intervals, for example, every 1ms.

[0147] First, the A / D converter 1601 converts the analog voltage signal output from the flow sensor 2 and the analog voltage signal output from the boost pressure sensor 9 located downstream of the throttle valve 7 into a digital voltage signal (S1).

[0148] Next, the voltage-flow conversion unit 1602 converts the digital voltage signal generated by A / D conversion of the analog voltage signal of the flow sensor 2 into a flow signal (S2).

[0149] In addition, the voltage-pressure conversion unit 1616 converts the digital voltage signal generated by A / D conversion of the analog voltage signal of the boost pressure sensor 9 into a pressure signal (S3).

[0150] Next, the maximum / minimum / average calculation unit 1609 accumulates the flow rate based on the flow rate signal calculated in step S2 during the feature quantity detection period set by the feature quantity detection period setting unit 1608 for calculating the average value of the flow rate signal (S4).

[0151] Next, multiple bandpass filters (1) to (4) perform bandpass filtering on the flow signal input from the voltage-flow conversion unit 1602 (S5).

[0152] Next, the maximum / minimum / average calculation unit 1609 updates the maximum and minimum values ​​of the flow signal during the aforementioned feature quantity detection period (equivalent to...). Figure 13 The maximum and minimum values ​​of the pressure signal are also measured. In addition, the amplitude detection units (1) to (4) update the maximum and minimum values ​​of the flow signals that have passed through each bandpass filter (1) to (4) (S6).

[0153] Next, the maximum / minimum / average calculation unit 1609 determines whether the crankshaft has passed a preset crankshaft angle position (S7) based on the output signal of the crankshaft angle sensor 18. The preset crankshaft angle is, for example, the top dead center of the compression stroke of each cylinder of the internal combustion engine 1. When the maximum / minimum / average calculation unit 1609 determines that the crankshaft has passed the preset crankshaft angle (S7 "Yes"), the ECU 27 performs the processing after step S8. On the other hand, if the crankshaft has not passed the preset crankshaft angle (S7 "No"), the processing of this flowchart ends.

[0154] Next, the maximum / minimum / average calculation unit 1609 calculates the average flow rate during the feature quantity detection period based on the cumulative flow rate value obtained in step S4 (equivalent to...). Figure 13 (S8). Next, the correction amount derivation unit 1614 calculates the pressure amplitude based on the maximum and minimum values ​​of the pressure signal obtained in step S6 (S9). Alternatively, the flow correction value calculation unit 1615 may also have an amplitude detection unit (not shown) that calculates the pressure amplitude based on the maximum and minimum values ​​of the pressure signal, and input the pressure amplitude calculated by the amplitude detection unit to the correction amount derivation unit 1614.

[0155] Next, the amplitude detection units (1) to (4) use the maximum and minimum values ​​of the flow signals that have passed through each bandpass filter (1) to (4) obtained in step S6 to calculate the amplitude (equivalent to) of the flow signals that have passed through each bandpass filter (1) to (4). Figure 13 σ1~σ4)(S10).

[0156] Next, the correction quantity derivation unit 1614, according to... Figures 13 to 15 The pulsation correction method shown derives the pulsation correction amount (equivalent to...). Figure 13 δ)(S11). That is, the correction amount derivation unit 1614 inputs the maximum and minimum values ​​of the flow signal obtained in step S6, the average value of the flow signal obtained in step S8, and the amplitude of the flow signal that has passed through each bandpass filter (1) to (4) obtained in step S10 into the learning completion model to obtain the pulsation correction amount.

[0157] Then, ECU 27 initializes the maximum, minimum, and cumulative values ​​of the flow signal and the maximum and minimum values ​​of the pressure signal (S12) for use in the next calculation.

[0158] Next, the correction amount derivation unit 1614 determines whether the pressure amplitude obtained in step S9 is below a predetermined value (S13). If the pressure amplitude exceeds the predetermined value ("No" in S13), the processing of this flowchart ends.

[0159] Then, if the pressure amplitude is below a predetermined value ("Yes" in S13), the flow correction unit 1620 will calculate the average value of the flow signal obtained in step S8 (equivalent to...). Figure 13 The upper bar of u) multiplied by the pulsation correction amount obtained in step S11 (equivalent to Figure 13 The δ) is used to correct the flow (S14). Through this series of processes, the mainstream flow is determined ( Figure 13 The upper side of the U-shaped bar). ECU 27 periodically repeats this process for each cylinder of internal combustion engine 1. Figure 17 The flowchart shown illustrates a series of processes.

[0160] Furthermore, while this flowchart illustrates performing all processing at regular intervals, steps S8 and beyond can also be executed separately as processes synchronized with the crank angle. That is, steps S8 and beyond can also be treated as interruptions synchronized with the crank angle.

[0161] [Changes in throttle opening, boost pressure sensor readings, and flow sensor readings]

[0162] Figure 18This is a timing diagram showing the changes in the opening degree of throttle valve 7, the detection value of boost pressure sensor 9, and the detection value of flow sensor 2 as the vehicle accelerates from a throttle state to a boost state by opening throttle valve 7. Figure 18 In the diagram, the horizontal axis represents time [s], and the vertical axis represents the throttle opening [deg], downstream pressure of throttle 7 [MPa], and flow rate [kg / s] in the order of upper, middle, and lower sections. The boost pressure sensor 9 detects the downstream pressure of throttle 7.

[0163] As shown in the figure, in the middle section, when the pressure downstream of throttle 7 accelerates from the non-pressurized region (below atmospheric pressure) to the pressurized region (above atmospheric pressure), during the period immediately after opening throttle 7 from the throttle state and filling air downstream of throttle 7 until atmospheric pressure conditions are met (equivalent to the "No" judgment in S13), there is no pulsation component in the flow detection results shown in the lower section. Therefore, during this period, since pulsation correction is not required, it is stopped (pulsation correction OFF). As shown in the lower section, pulsation occurs in the flow detection results after the pressure downstream of throttle 7 reaches atmospheric pressure conditions. During the time period before and after pulsation correction OFF, pulsation correction is set to ON. In this way, by appropriately stopping pulsation correction, flow errors caused by over-correction are prevented.

[0164] Thus, the electronic control device (ECU 27) of this embodiment is configured to stop the calculation of the correction value of the flow correction value calculation unit 1615 (correction amount derivation unit 1614) when the change in the downstream pressure of the throttle valve 7 assembled in the intake manifold exceeds a predetermined value during a specified period (characteristic quantity detection period) (equivalent to "No" in S13).

[0165] In addition, the electronic control device (ECU 27) of this embodiment may stop the calculation of the correction value of the flow correction value calculation unit (correction amount derivation unit 1614) if any one or more of the average value, maximum value, minimum value of the intake air flow rate calculated by the flow calculation unit (voltage flow conversion unit 1602) during a specified period (characteristic quantity detection period) and the amplitude of a signal of one or more frequencies included in the output signal of the flow measuring device (flow sensor 2) are outside the specified range.

[0166] If any one or more of the average, maximum, and minimum values ​​of the intake air flow rate, or the amplitude of a signal at more than one frequency, fall outside the specified range, an abnormality in the internal combustion engine 1 is suspected. In such cases, by appropriately stopping the pulsation correction, erroneous flow errors caused by abnormal values ​​or flow errors caused by overcorrection can be prevented. When an abnormality in the internal combustion engine 1 is detected, the ECU 27 can perform fault protection procedures.

[0167] [Relationship between pulsation amplitude ratio and pulsation correction value]

[0168] Figure 19 This is a graph showing the relationship between the pulsation amplitude ratio and the pulsation correction value as effects of the first and second embodiments described above. Here, the output signal of the flow sensor 2 was measured multiple times at different throttle opening points (operating conditions) 1 to 4 with a fixed rotation speed (e.g., 1500 rpm), and the pulsation amplitude ratio and pulsation correction value were calculated.

[0169] The upper section is a graph plotted by averaging multiple necessary correction values ​​calculated based on the pulsation amplitude ratios measured at action points 1 to 4 and the detection error of flow sensor 2. The plotted points indicated by "●" are the average values ​​of the measurements at each action point 1 to 4. For example, the necessary correction values ​​for multiple pulsation amplitude ratios obtained over a specified period such as 10 seconds are averaged.

[0170] The middle section is a graph showing the correction values ​​when the internal combustion engine 1 is operated using the technology described in Patent Document 1. The plotted points, indicated by "○", "◇", "△", and "□", are the average measurements taken when the control described in Patent Document 1 is applied at each of the operation points 1 to 4. The correction values ​​"△" for operation point 3 and "□" for operation point 4 are plotted vertically without overlapping. The solid line represents the correction value set in the pulsation correction mapping. The pulsation correction mapping can also be set to precisely set inflection points to track the measured values. For example, the solid line can be set to a shape connecting the operation point 2 side to the operation point 3 side, and then from the operation point 3 side to the operation point 4 side. However, for changes in the pulsation amplitude ratio, a sharp change in the correction amount is not ideal, so it is set as a solid line as shown in the middle section.

[0171] As shown in the middle section, even at each actuation point (under the same operating conditions) with a fixed rotational speed and throttle opening, the pulsation amplitude ratio will vary, especially where the slope of the correction is large (the sloping part of the solid line), where the change in correction becomes larger. In addition, at actuation points 3 and 4, the pulsation amplitude ratio is the same (approximately 300%), so although the necessary correction for actuation points 3 and 4 differs by 8%, they are the same correction value.

[0172] The following section is a graph showing the correction values ​​when the present invention is applied. The plotted points, indicated by "○", "◇", "△", and "□", are the average measurements taken when the control of the present invention is applied at each of the 1 to 4 action points. As shown in the graph, in the present invention (which also includes, for example,...) Figures 13-15 , Figure 12 In this process, since the necessary correction values ​​are calculated based on the pulsation characteristics of each action point, the correction values ​​for action points 3 and 4 can be separated. Therefore, the necessary correction values ​​can be calculated for each action point.

[0173] [Relationship between pulsation amplitude ratio and pulsation correction value at high rotation speed]

[0174] Figure 20 This is a graph showing the relationship between the pulsation amplitude ratio and the pulsation correction value at high rotation speed, which is an effect of the first and second embodiments described above. Here, the output signal of the flow sensor 2 was measured multiple times at different throttle opening points (operating conditions) 1 to 4 at a relatively high fixed rotation speed (e.g., 4000 rpm), and the pulsation amplitude ratio and pulsation correction value were calculated.

[0175] The upper section is a graph plotted by averaging several necessary correction values ​​calculated based on the pulsation amplitude ratio measured at operating points 1 to 4 at high rotational speeds (e.g., 4000 rpm) and the detection error of flow sensor 2. The plotted points indicated by "●" are the average values ​​of the measurements at each operating point 1 to 4.

[0176] The middle section plots the correction values ​​for the characteristic quantities of the pulsating waveform when the pulsating amplitude of the flow signal passing through bandpass filters 1604-1607 of different frequencies is used as the input to the correction value derivation unit 1614. The plotted points “○”, “◇”, “△”, and “□” are the measured average values ​​when the pulsating amplitude of different frequency components of the flow signal is applied at each of the operation points 1-4. That is, the correction values ​​shown at each plotted point do not include the average value between the maximum and minimum values ​​of the flow signal. As shown in the figure, during high rotation, the frequency contained in the pulsating waveform increases, but the higher the frequency, the greater the attenuation. Therefore, the difference in characteristic quantities at each operation point becomes smaller, and the detection error of the flow sensor 2 becomes larger. Therefore, the deviation of the correction value is large between operation points 1-4.

[0177] The following section is a graph showing the correction values ​​for the pulsating waveform characteristic quantity of the flow signal after passing through bandpass filters 1604-1607 of different frequencies, with the average, maximum, and minimum values ​​of the flow signal added to the pulsating amplitude. The plotted points “○”, “◇”, “△”, and “□” are the measured average values ​​when the control of the present invention is applied at each of the 1-4 action points. As shown in the figure, in the present invention, as in… Figure 11 as well as Figure 12 As explained, even with the attenuation of high-frequency components, the necessary correction values ​​can be determined at each action point because the positional relationship of the average, maximum, and minimum values ​​of the flow signal is included in one of the pulsating characteristic quantities.

[0178] As described above, the electronic control device (ECU27) of the second embodiment includes the following configuration: a flow calculation unit (voltage-flow conversion unit 1602) that calculates the flow rate of the intake air based on the output signal of the flow measuring device (flow sensor 2) mounted on the intake manifold; a flow correction value calculation unit (flow correction value calculation unit 1615) that calculates the average value (μ), maximum value (max), minimum value (mix) and the amplitude (σ1 to σ4) of a signal at or above the fundamental frequency of the output signal of the flow measuring device and included in the output signal of the flow measuring device during a predetermined period (characteristic quantity detection period) of the flow rate of the intake air calculated by the flow calculation unit, and calculates a correction value (δ) for the flow rate of the intake air based on the calculation result; and a flow correction unit (flow correction unit 1620) that corrects the flow rate of the intake air (u with bar on the upper side) based on the correction value.

[0179] According to the second embodiment configured as described above, the detection error of the flow measuring device (flow sensor 2) is corrected based on the average, maximum, and minimum values ​​of the output signal of the flow measuring device, and the amplitude information of one or more frequency components contained in the output signal of the flow measuring device. Therefore, even under operating conditions of an internal combustion engine where the airflow at the location of the flow measuring device produces various pulsations, the airflow rate can be determined with high accuracy. Thus, the deterioration of the exhaust characteristics or fuel consumption rate of the internal combustion engine that is a concern during periods of large pulsation can be prevented.

[0180] Furthermore, in the electronic control device (ECU 27) of this embodiment described above, the flow correction value calculation unit (flow correction value calculation unit 1615) includes a neural network model (correction amount derivation unit 1614), which takes as input the average value (μ), maximum value (max), minimum value (mix) of the flow rate of the intake air calculated by the flow calculation unit (voltage flow conversion unit 1602) during a specified period (characteristic quantity detection period), and the amplitude of a signal at or above the fundamental frequency of the output signal of the flow measuring device (flow sensor 2) and containing at least one frequency in the output signal of the flow measuring device, and outputs a correction value (e.g., pulsation correction amount δ) for the flow rate of the intake air.

[0181] According to this embodiment configured as described above, a neural network model is used, and... Figure 12 Similarly, in the case of the pulsation correction multidimensional mapping diagram 1200 shown, it is possible to determine the correction value corresponding to the difference in position between the maximum and minimum values ​​of the output signal of the flow measuring device (flow sensor 2).

[0182] Furthermore, in the electronic control device (ECU 27) of this embodiment described above, the correction value is a correction coefficient and / or a correction amount. The flow correction unit (flow correction unit 1620) multiplies the flow rate of the intake air (u with bar on the upper side) calculated by the flow calculation unit (voltage flow conversion unit 1602) by the correction coefficient (pulsation correction amount δ) and adds the correction amount, or multiplies it by the correction coefficient and adds the correction amount.

[0183] <Third Implementation>

[0184] The third embodiment involves appropriately selecting the configuration of the bandpass filter to be used from a plurality of bandpass filters 1604 to 1607 based on the rotational speed of the internal combustion engine 1.

[0185] [Pulse Correction Logic]

[0186] Figure 21 This is a block diagram illustrating an example of a pulsation correction logic installed in ECU 27A to correct the detection error of flow sensor 2, according to the third embodiment of the present invention. ECU 27A and ECU 27 (see reference ECU 27A). Figure 16 Compared to ECU 27, the difference lies in the presence of a frequency selection unit 1617. Hereinafter, the differences between ECU 27A and ECU 27 will be explained.

[0187] The frequency selection unit 1617 selects the bandpass filter to be used from a plurality of bandpass filters 1604 to 1607 based on the rotational speed of the internal combustion engine 1 calculated by the rotational speed calculation unit 1603, and outputs the selection signal to the corresponding bandpass filter.

[0188] The bandpass filter that receives the selection signal from the frequency selection unit 1617 filters the flow signal (pulsating waveform) input from the voltage-flow conversion unit 1602 and outputs the processing result to the corresponding amplitude detection unit.

[0189] Thus, the flow correction value calculation unit (flow correction value calculation unit 1615) of this embodiment includes a frequency selection unit (frequency selection unit 1617), which selects the frequency of the calculation object according to the rotational speed of the internal combustion engine when calculating the amplitude of a signal of more than one frequency included in the output signal of the flow measuring device (flow sensor 2).

[0190] When calculating the amplitude of a signal containing one or more frequencies in the output signal of the flow measuring device (flow sensor 2), the frequency selection unit (frequency selection unit 1617) selects a combination of at least one frequency from the fundamental frequency and the frequencies of higher harmonics, which are determined by the rotational speed of the internal combustion engine. For example, consider a configuration in which bandpass filters (1) to (4) are set to the fundamental frequency and bandpass filters (2) to (4) are set to the frequencies of the 2nd to 4th higher harmonics. In order to improve the inference accuracy of the correction value, it is useful to select multiple frequencies (bandpass filters) that include the fundamental frequency. However, it is also possible to select only the fundamental frequency or multiple frequencies that do not include the fundamental frequency.

[0191] Alternatively, the frequency selection unit (frequency selection unit 1617) can be configured to select at least one frequency from a combination of a fixed frequency, a fundamental frequency determined by the rotational speed of the internal combustion engine, and a higher harmonic frequency when the amplitude of a signal containing one or more frequencies in the output signal of the flow measuring device (flow sensor 2) is measured. For example, consider configuring the bandpass filters (1) to (4) such that the bandpass filter (1) is set to a fixed frequency, the bandpass filter (2) is set to the fundamental frequency, and the bandpass filters (3) and (4) are set to the higher harmonic frequencies.

[0192] In the third embodiment configured as described above, the same effects as in the first and second embodiments can be obtained. Furthermore, in the third embodiment, the following effects, which differ from those in the first and second embodiments, can also be obtained. According to the third embodiment, the frequency selection unit (frequency selection unit 1617) can select the appropriate frequency of the computation object based on the rotational speed of the internal combustion engine. Therefore, the correction value for the intake air flow rate can be calculated by taking into account the amplitude of the signal of any frequency contained in the output signal of the flow measuring device (flow sensor 2). Therefore, even under the operating conditions of an internal combustion engine where the intake air flow at the location of the flow measuring device produces various pulsations, the intake air flow rate can be determined with high accuracy.

[0193] <Variation Example>

[0194] Furthermore, in the first to third embodiments described above, the maximum, minimum, and average values ​​of the output signal (flow signal) of the flow sensor 2 are used, but this is not a limitation. For example, even if a combination of the upper amplitude (absolute value), lower amplitude (absolute value), and average value of the flow signal, or a combination of the position of the amplitude value, the average value, and the average value within the amplitude is used, the same effect as described above can be achieved.

[0195] Furthermore, in the first to third embodiments described above, one or more Fourier transform units (not shown) may be provided instead of one or more bandpass filters (bandpass filters (1) to (4)) configured to correspond to any one of the frequencies contained in the output signal of the flow sensor 2. For example, Fourier transform units (1) to (4) may be provided, and the frequency of the processing target may be set in the Fourier transform units (1) to (4) instead of the bandpass filters (1) to (4) may be provided. Figure 16 The bandpass filters shown are (1) to (4).

[0196] Fourier transform unit (1) performs a Fourier transform on the flow signal and extracts the fundamental frequency from the transform result, outputting it to amplitude detection unit (1). Additionally, Fourier transform unit (2) performs a Fourier transform on the flow signal and extracts the second harmonic from the transform result, outputting it to amplitude detection unit (2). Similarly, Fourier transform unit (3) extracts the third harmonic from the Fourier transform result of the flow signal and outputs it to amplitude detection unit (3), and Fourier transform unit (4) extracts the fourth harmonic from the Fourier transform result of the flow signal and outputs it to amplitude detection unit (4).

[0197] Furthermore, the amplitude detection units (1) to (4) corresponding to the Fourier transform units (1) to (4) respectively calculate the amplitude of the fundamental wave or higher harmonic output from the corresponding Fourier transform units (1) to (4). In this way, when a Fourier transform unit is provided instead of a bandpass filter, the same effect as when a bandpass filter is provided can be obtained.

[0198] Alternatively, in embodiments 1 to 3, the passband configuration of bandpass filters (1) to (4) can be changed. For example, the bandpass filter can change the configuration of the frequency band it can pass through by changing or switching its logic configuration, parameters, etc. If the frequency selection unit 1617 selects a frequency based on the rotational speed, it outputs a frequency selection signal to any one of the bandpass filters (1) to (4). The bandpass filter that has received the frequency selection signal changes its passband based on the frequency information contained in the frequency selection signal. Then, the bandpass filter outputs a signal that has passed through the changed passband to the corresponding amplitude detection unit.

[0199] Furthermore, the present invention is not limited to the above-described embodiments. Various other applications and modifications may be adopted without departing from the spirit of the invention as described in the claims.

[0200] For example, the above embodiments have described the configuration of the electronic control device (ECU 27, 27A) in detail for the purpose of easily understanding and explaining the present invention, and are not necessarily limited to having all the constituent elements described. Furthermore, a portion of the configuration of one embodiment may be replaced with constituent elements of other embodiments. Additionally, constituent elements of other embodiments may be added to the configuration of one embodiment. Furthermore, for a portion of the configuration of each embodiment, other constituent elements may be added, replaced, or deleted.

[0201] Furthermore, the aforementioned components, functions, and processing units can also be implemented in hardware, for example, by designing some or all of them using integrated circuits. As hardware, generalized processor devices such as FPGAs (Field Programmable Gate Arrays) or ASICs (Application Specific Integrated Circuits) can also be used.

[0202] In addition, Figure 17 As shown in the flowchart, multiple processes can be executed in parallel or the processing order can be changed without affecting the processing result.

[0203] Symbol Explanation

[0204] 1…Internal combustion engine, 2…Flow sensor, 7…Throttle valve, 9…Boost pressure sensor, 15…Fuel injection valve, 16…Spark plug, 18…Crank angle sensor, 22…EGR pipe, 24…EGR valve, 27, 27A…ECU, 1602…Voltage-flow conversion unit, 1603…Rotation speed calculation unit, 1604~1607…Bandpass filter, 1608…Characteristic detection period setting unit, 1609…Maximum / minimum / average value calculation unit, 1610~1613…Amplitude detection unit, 1614…Correction value derivation unit, 1615…Flow correction value calculation unit, 1616…Voltage-pressure conversion unit, 1620…Flow correction unit, 1615…Flow correction value calculation unit, 1620…Flow correction unit.

Claims

1. An electronic control device, characterized in that, have: The flow calculation unit calculates the flow rate of the intake air based on the output signal of the flow measuring device assembled on the intake pipe; The flow correction value calculation unit calculates the average, maximum, and minimum values ​​of the intake air flow rate calculated by the flow calculation unit within a specified period, as well as the amplitude of a signal at or above the fundamental frequency of the output signal of the flow measuring device and at a frequency included in the output signal of the flow measuring device. Using the relative position of the average value within the interval formed by the maximum and minimum values, and the amplitude of the signal at the one or more frequencies, as parameters, it calculates a correction value for the intake air flow rate. as well as The flow correction unit corrects the flow rate of the intake air based on the correction value.

2. The electronic control device according to claim 1, characterized in that, The flow correction value calculation unit includes: one or more bandpass filters corresponding to any one of the more than one frequencies contained in the output signal of the flow measuring device; and One or more amplitude detection units are set up corresponding to the one or more bandpass filters and calculate the amplitude of the output signal that has passed through the corresponding bandpass filter.

3. The electronic control device according to claim 1, characterized in that, The flow correction value calculation unit includes: one or more Fourier transform units corresponding to any one of the more than one frequencies included in the output signal of the flow measuring device; and One or more amplitude detection units are set up corresponding to the one or more Fourier transform units and calculate the amplitude of the fundamental wave or higher harmonic output from the corresponding Fourier transform unit.

4. The electronic control device according to claim 1 or 2, characterized in that, The flow correction value calculation unit has a neural network model, which takes as input the average value, maximum value, minimum value of the flow rate of the intake air calculated by the flow calculation unit over a specified period, and the amplitude of a signal at or above the fundamental frequency of the output signal of the flow measuring device and containing at least one frequency in the output signal of the flow measuring device, and outputs a correction value for the flow rate of the intake air.

5. The electronic control device according to claim 4, characterized in that, The flow correction value calculation unit includes a frequency selection unit, which selects the frequency of the calculation object based on the rotational speed of the internal combustion engine when calculating the amplitude of the signal of one or more frequencies.

6. The electronic control device according to claim 5, characterized in that, When calculating the amplitude of the signal at one or more frequencies, the frequency selection unit selects a combination of at least one frequency from the fundamental frequency and the frequencies of higher harmonics determined by the rotational speed of the internal combustion engine.

7. The electronic control device according to claim 5, characterized in that, When calculating the amplitude of the signal at one or more frequencies, the frequency selection unit selects a combination of at least one frequency from a fixed frequency and the fundamental frequency and higher harmonic frequencies determined by the rotational speed of the internal combustion engine.

8. The electronic control device according to claim 1, characterized in that, The calculations and corrections of the flow correction value calculation unit are performed at a cycle synchronized with the intake interval of the internal combustion engine.

9. The electronic control device according to claim 8, characterized in that, The specified period for the flow correction value calculation unit to calculate the average, maximum, and minimum flow rates of the intake air calculated by the flow calculation unit is a period synchronized with the intake interval of the internal combustion engine or an integer multiple of the specified period.

10. The electronic control device according to claim 1, characterized in that, The correction value is a correction coefficient and / or a correction amount. The flow correction unit multiplies the flow rate of the intake air calculated by the flow calculation unit by the correction coefficient, adds the correction amount, or multiplies by the correction coefficient and adds the correction amount.

11. The electronic control device according to claim 1, characterized in that, If the change in the downstream pressure of the throttle valve assembled in the intake manifold exceeds a predetermined value during the specified period, the calculation of the correction value by the flow correction value calculation unit is stopped.

12. The electronic control device according to claim 1, characterized in that, If any one or more of the average, maximum, and minimum values ​​of the intake air flow rate calculated by the flow calculation unit during the specified period, and the amplitude of one or more frequencies contained in the output signal of the flow measuring device, are outside the specified range, the calculation of the correction value by the flow correction value calculation unit shall be stopped.

13. A flow measurement system, comprising: Flow measurement device assembled on the intake pipe; and Electronic control device, The flow measurement system is characterized in that... The electronic control device includes: The flow calculation unit calculates the flow rate of the intake air based on the output signal of the flow measuring device; The flow correction value calculation unit calculates the average, maximum, and minimum values ​​of the intake air flow rate calculated by the flow calculation unit over a specified period, as well as the amplitude of a signal at or above the fundamental frequency of the output signal of the flow measuring device and containing at least one frequency. Using the relative position of the average value within the interval formed by the maximum and minimum values, and the amplitude of the at least one frequency signal, as parameters, it calculates a correction value for the intake air flow rate. The flow correction unit corrects the flow rate of the intake air based on the correction value.

Images