AIR-FUEL RATIO REGULATION DEVICE

Abstract

Air-fuel ratio control device comprising: a resistance-type oxygen sensor provided in contact with exhaust gas of an internal combustion engine exhibiting exhaust pulses, the sensor having a detection unit whose resistance value changes substantially stepwise at oxygen concentrations near the stoichiometry of the exhaust gas, and in which a detected value obtained from the resistance value of the detection unit produces a pulsed waveform with a peak value that responds to the temperature of the detection unit and to the exhaust gas pulses; a temperature reading unit that estimates or detects the temperature of the detection unit; and an excess-fuel ratio calculation unit that calculates an excess-air ratio with respect to a data map.which specifies the plurality of air-fuel ratio values ​​with correspondence relationships between a plurality of first scale values ​​for temperature and a plurality of second scale values ​​for the detected value, wherein the air-fuel ratio control device performs an air-fuel ratio feedback control based on a deviation between the air-fuel ratio and a target air-fuel ratio, wherein the air-fuel ratio control device further comprises: a storage unit which, with the correspondence relationships to the first scale values, stores the data map and a lookup table, wherein the lookup table specifies a rich-side threshold and a lean-side threshold to distinguish which air-fuel ratio range the detected value corresponds to under a rich range, a stoichiometric range and a lean range; a characteristic test unit,which sets the target air-fuel ratio close to stoichiometry in order to capture a peak value of the detected value and checks whether the peak value corresponds to one of the air-fuel ratio ranges; and a calibration unit which calibrates the data map and the lookup table by means of an affine transformation procedure, according to the test, so that the peak value corresponds to the threshold on the rich side and the threshold on the lean side.

Description

Technical FieldThe present invention relates to an air-fuel ratio control device that controls the air-fuel ratio in an internal combustion engine based on the oxygen concentration in the exhaust gas.Background ArtConventionally, it is known to perform air-fuel ratio feedback control in the combustion process of an internal combustion engine on the basis of an air-fuel ratio obtained from output information of an oxygen sensor that is in contact with the exhaust gas of the internal combustion engine (see, for example, U.S. Pat. No. 8,959,987 B2).In the technique of U.S. Pat. No. 8,959,987 B2, a resistance type oxygen sensor that detects oxygen based on an internal resistance of a detection unit is used as the oxygen sensor. In this oxygen sensor, a first value indicating an oxygen content in the exhaust gas is determined based on the resistance of the detection unit of the oxygen sensor. Moreover, a second value indicating the temperature of the exhaust gas sensor is determined based on the resistance of the heating part of the oxygen sensor. The air-fuel ratio is determined as a third value as a function of the first and second values.According to this technique, the third value determination function includes the second value that influences the first value in real time, so that the air-fuel ratio is accurately detected even in a case where the output characteristics of the oxygen sensor change with temperature.SUMMARY OF THE INVENTIONTechnical ProblemHowever, according to the technique described in U.S. Pat. No. 8,959,987 B2, the air-fuel ratio is calculated based on the resistance value of the detection unit of the oxygen sensor and the resistance value of the heating part. Therefore, if the resistance values vary due to manufacturing tolerance or temporal changes, the air-fuel ratio obtained based on these resistance values could vary significantly. Accordingly, the accuracy of the air-fuel ratio feedback control may also be significantly degraded.In view of the problem of the conventional technique, it is an object of the present invention to provide an air-fuel ratio feedback control apparatus capable of performing accurate air-fuel ratio feedback control regardless of variation in resistance values caused by manufacturing tolerance and temporal changes of the oxygen sensor.Solution of the ProblemAn air-fuel ratio control apparatus according to the present invention comprises:a resistance type oxygen sensor provided in contact with exhaust gas of an internal combustion engine including exhaust gas pulses, comprising a detection unit whose resistance value changes substantially stepwise at the oxygen concentration in the vicinity of stoichiometry of the exhaust gas, and wherein a detected value obtained from the resistance value of the detection unit produces a pulsed waveform having a peak value responsive to a temperature of the detection unit and to the exhaust gas pulses;a temperature reading unit that estimates or detects the temperature of the detection unit; andan excess ratio calculation unit that calculates an excess air ratio with reference to a data map indicating the plurality of excess air ratio values having correspondence relations between a plurality of first scale values for the temperature and a plurality of second scale values for the detected value,wherein the air-fuel ratio control device performs air-fuel ratio feedback control based on a deviation between the excess air ratio and a target excess air ratio,wherein the air-fuel ratio control device further comprises:a storage unit that stores, with the correspondence relations with the first scale values, the data map and a look-up table indicating a rich side threshold and a lean side threshold for discriminating which air-fuel ratio range the detected value corresponds to among a rich range, a stoichiometric range, and a lean range;a characteristic checking unit that sets the target excess air ratio to be near stoichiometry to detect a peak value of the detected value, and checks whether the peak value corresponds to one of the air-fuel ratio ranges; anda calibration unit that calibrates the data map and the lookup table by an affine transformation procedure according to the test so that the peak value corresponds to the rich side threshold value and the lean side threshold value.In this configuration, the value detected by the detection unit of the oxygen sensor produces a pulsed waveform having a peak value in response to exhaust pulses of the internal combustion engine. Moreover, the temperature of the detection unit of the oxygen sensor affects the above wave height. For example, at low temperatures (100° C. or lower), the wave height becomes approximately zero (oxygen sensor is inactive).Thus, according to the present invention, the air-fuel ratio (excess air ratio) is calculated by the data map indicating the plurality of excess air ratio values with the correspondence relations between the resistance values (sensor voltage values) of the detection unit of the oxygen sensor with respect to the above detected value and the temperature values of the detection unit obtained from the resistance values of the heating unit, and therefore, if the resistance values of the detection unit or the resistance values of the heating unit vary, the excess air ratio obtained based on these resistance values is also inaccurate, which might result in problems in the appropriate air-fuel ratio feedback control.Therefore, in the present invention, the wave height of the detected oxygen sensor value currently measured is compared with a specified standard (median) peak value by setting the target excess air ratio in the vicinity of the stoichiometric value for inspection with threshold values on the rich side and the lean side, so that the correspondence relationships between the detected value and the temperature in the aforementioned data map provide appropriate relationships for calculating an accurate excess air ratio, and the first and second scale values of the above data map, which provide the correspondence relationships in which the peak value of the detected value currently measured on the basis of the result of the inspection matches the threshold values on the rich side and the lean side (the wave height peaks overlap the threshold values), and wherein the threshold values at the rich and the lean sides of the look-up table are obtained by an affine transformation, thereby calibrating the data map and the look-up table.Therefore, according to the present invention, the excess air ratio is calculated using the data map and the look-up table calibrated in response to the variation in the resistance values to enable accurate calculation of the excess air ratio to perform appropriate air-fuel ratio feedback control regardless of variation in the resistance values of the oxygen sensor.In the present invention, the characteristic checking unit may include a temperature difference checking unit that detects an amount of deviation of the temperature of the detection unit from a reference value if, with the engine stopped, a predetermined period of time sufficient for the temperature of the detection unit to converge to an ambient temperature of the detection unit has elapsed, and a value obtained by correcting the temperature of the detection unit based on the amount of deviation may be used as the temperature of the detection unit for control used in the checking and the excess ratio calculation unit.Accordingly, after a lapse of a predetermined time period sufficient for the temperature of the detection unit once heated to a higher temperature to converge to the ambient temperature of the oxygen sensor, control is performed based on the control temperature of the detection unit corrected based on the amount of deviation from the reference value, thereby enabling more accurate air-fuel ratio feedback control.In this case, the characteristic checking unit may include a voltage difference determining unit that, in a case where the temperature of the detecting unit of the controller is equal to or lower than a predetermined value, detects a deviation between an average value within a predetermined period of time of the detected value obtained from the resistance value of the detecting unit and the threshold value on the lean side, and, as the detected value for control used in the checking and the excess ratio calculating unit, there may be used a value obtained by correcting the detected value from the resistance value based on the deviation.Accordingly, if the detection unit temperature is equal to or lower than a sufficiently low predetermined value, in other words, a condition in which the oxygen concentration sensor is inactive and the voltage indicates that the detected values do not move, the deviation from the threshold value is obtained from the lean side, and the control based on the detected value for control corrected based on the deviation is performed, thereby enabling more accurate air-fuel ratio feedback control.In this case, the calibration unit may store a first calibration magnification value to calibrate the data map and the look-up table by scaling up or down the plurality of first scale values, and in a case where the wave height peak near the lean side threshold in the wave height of the detected value for control that changes over time may be compared with the lean side threshold, and the calibration unit may increase the first calibration magnification value for the lean region peak and decrease the first calibration magnification value for the stoichiometric region peak.Accordingly, since the wave height peak near the lean side threshold of the detected value approaches the lean side threshold, the data map and the lookup table are calibrateable such that the lean side peak corresponds to the lean side threshold.In the present invention, the calibration unit may include a second calibration magnification value for calibrating the data map and the look-up table by multiplying the plurality of second scale values, the rich side threshold value, and the lean side threshold value, and the calibration unit may increase the second calibration magnification value in a case where the wave height peak closer to the rich side threshold value in the wave height of the detected value for control that changes over time is in the rich range, and the calibration unit may decrease the second calibration magnification value in a case where the peak is in the stoichiometric range.Accordingly, the wave height peak that is closer to the rich side threshold than the detected value approaches the rich side threshold, and therefore the data map and the lookup table are calibrateable such that the rich side peak corresponds to the rich side threshold.Brief Description of the DrawingsFIG. 1 is a schematic diagram schematically illustrating the configuration of the main part of an internal combustion engine including an air-fuel ratio control device according to an embodiment of the present invention. FIG. 2 is a block diagram illustrating the main structure of an ECU of the internal combustion engine illustrated in FIG. 1. FIG. 3 is a flowchart illustrating an excess ratio calculation process for calculating an excess air ratio by an excess ratio calculation unit in the ECU illustrated in FIG. 2. FIG. 4 is a graph illustrating how the air-fuel ratio in the stoichiometric range is calculated in the process of FIG. 3. FIG. 5 is a diagram illustrating a graph corresponding to a look-up table for finding a lean-side threshold LREF and a rich-side threshold RREF, and a data map for calculating the excess air ratio in the process of FIG. 3. FIG. 6 is a graph illustrating schematic changes in an air excess ratio λ calculated in the process in FIG. 3. FIG. 7 is a waveform diagram illustrating voltage waveforms representing changes in voltage values output from a voltage calculation unit over time obtained while a vehicle equipped with an internal combustion engine is traveling. FIG. 8A is a diagram illustrating an example of the inspection by a characteristic inspection unit and calibration by a calibration unit in the ECU illustrated in FIG. 2. FIG. 8B is a diagram illustrating another example of the inspection by the characteristic inspection unit and the calibration by the calibration unit. FIG. 8C is still another example of the inspection by the characteristic inspection unit and the calibration by the calibration unit. FIG. 8D is still another example of the inspection by the characteristic inspection unit and the calibration by the calibration unit.DESCRIPTION OF THE EMBODIMENTSHereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 illustrates the configuration of the main part of a four-stroke internal combustion engine equipped with an air-fuel ratio control device according to an embodiment of the present invention. This air-fuel ratio control apparatus has a function of performing air-fuel ratio feedback control based on a deviation between an excess air ratio obtained based on oxygen concentrations in the exhaust gas of the internal combustion engine and a target excess air ratio.As shown in FIG. 1, an engine body 1 of this internal combustion engine has an intake pipe 2 provided at an intake port and a throttle valve 3 provided in the intake pipe 2 to adjust the amount of intake air supplied from an air cleaner 4 to the intake port according to an opening degree.The throttle valve 3 is provided with a throttle sensor 5 that detects the opening degree of the throttle valve 3. A fuel injection valve 6 is provided in the vicinity of the inlet port of the intake pipe 2 to inject fuel. Fuel is pumped from a fuel tank (not shown) to the fuel injection valve 6 by a fuel pump.The intake pipe 2 is provided with an intake pressure sensor 7 that detects an intake pressure in the intake pipe 2, and an intake air temperature sensor 8 that detects the temperature of intake air in the intake pipe 2. In an exhaust pipe 10 connected to an exhaust port of the engine body 1, there are provided a catalyst 11 that reduces unburned components in the exhaust gas of the exhaust pipe 10, and an oxygen sensor 12 that detects the oxygen concentration in the exhaust gas.A spark plug 13 connected to an ignition device 14 is fixed to the engine body 1. An electronic control unit (ECU) 15 outputs an ignition timing command to the ignition device 14 to cause spark discharge in the cylinder combustion chamber of the engine 1.The ECU 15 receives analog voltage inputs indicating the detected values of the throttle sensor 5, the intake pressure sensor 7, the intake air temperature sensor 8, the oxygen sensor 12, a cooling water temperature sensor 17, and an atmospheric pressure sensor 20 that detects an atmospheric pressure. The ECU 15 is also connected to the above-described fuel injection valve 6.The ECU 15 further receives an input signal indicative of the rotational angle position of a crankshaft 18 from a crank angle sensor 19. In other words, the crank angle sensor 19 magnetically or optically detects, by means of a pickup 19 bdisposed near the circumference of a rotor 19 a, a plurality of convex portions provided at each predetermined angle (for example, 15°) on the circumference of the rotor 19 athat rotates along with the crankshaft 18, and generates a pulse (crank signal) from the pickup 19 bfor each predetermined angular rotation of the crankshaft 18.Specifically, each time the piston 9 reaches its top dead center or each time the crankshaft 18 rotates 360°, the crank angle 19 outputs to the ECU 15 a signal indicating the reference angle.FIG. 2 illustrates the main structure of the ECU 15. As illustrated in FIG. 2, the oxygen sensor 12 that supplies a detection signal of an oxygen concentration in the exhaust gas to the ECU 15 is in contact with the exhaust gas of an internal combustion engine including exhaust pulses, and the oxygen sensor 12 includes a sensor element 12 aas a detection unit that detects the oxygen concentration in the exhaust gas and a sensor heater 12 badjacent to the sensor element 12 ato heat the sensor element 12 a.The sensor element 12a has a resistance value that changes substantially stepwise when the exhaust gas from the internal combustion engine has an oxygen concentration in the stoichiometric vicinity, and a detected value obtained from the resistance value generates a pulsed waveform having a peak value that responds to a temperature of the sensor element 12 and to the exhaust pulses. The sensor element 12 aused in this embodiment is a titanium dioxide sensor element which is a resistance type oxygen sensor in which the resistance value changes with the oxygen concentration.The ECU 15 includes a heater controller 22 that controls the sensor heater 12 b, a temperature calculation unit 23 as a temperature reading unit that calculates a temperature value T indicating the temperature of the sensor element 12 a, and a voltage calculation unit 24 that converts an output signal of the sensor element 12 ainto a voltage value VHG as a detected value indicating the oxygen concentration in the exhaust gas.The temperature control of the sensor heater 12 bby the heater controller 22 by the ECU 15 is performed by pulse width modulation (PWM) control of the magnitude of excitation current supplied to the sensor heater 12 bfrom a power supply (storage battery) not illustrated. The calculation of the temperature value T by the temperature calculation unit 23 is performed by, for example, reading the values of the heater voltage and the excitation current supplied from the ECU 15 to the sensor heater 12 bto obtain a resistance value of the sensor heater 12 b, and then converting the resistance value using table data or a calculation formula representing the correspondence relationship between the heater resistance value and the temperature value T and prepared in advance in the ECU 15. The calculation results of the temperature calculation unit 23 and the voltage calculation unit 24 are supplied to a substitute value calculation unit 26 of an excess ratio calculation unit 25 described later.The ECU also includes: a rotation speed calculation unit 27 that calculates a rotation speed NE and an angular speed NETC of the internal combustion engine on the basis of the detection results of the crank angle sensor 19; and an excess ratio calculation unit 25 that calculates an excess air ratio λ on the basis of a temperature value T from the temperature calculation unit 23, a voltage value VHG from the voltage calculation unit 24, and an angular speed NETC from the rotation speed calculation unit 27.Further, the ECU 15 includes: a target value calculation unit 28 that calculates a target excess air ratio λcmd based on an estimated value or the like of an amount of oxygen stored in the catalyst 11; a basic injection amount calculation unit 29 that calculates a basic injection amount BJ based on a rotational speed NE from the rotational speed calculation unit 27 and a pressure PM in the intake pipe 2 from the intake pressure sensor 7; a feedback coefficient calculation unit 30 that finds a feedback coefficient k for use in correcting the basic injection amount BJ calculated by the basic injection amount calculation unit 29 so that the excess air ratio λ calculated by the excess ratio calculation unit 25 matches the target excess air ratio λcmd; and an injection amount calculation unit 31 that calculates an injection amount Ti based on the feedback coefficient k and the basic injection amount BJ and that activates the fuel injection valve 6.In the feedback coefficient calculation unit 30, the feedback coefficient k is calculated by performing PID control based on a deviation between the excess air ratio λ and the target excess air ratio λcmd. Based on the injection amount Ti calculated by the injection amount calculation unit 31 based on the feedback coefficient k and the basic injection amount BJ, the fuel injection valve 6 is opened for the corresponding time. Thus, the fuel of the amount corresponding to the feedback coefficient k of the above PID control is injected into the cylinder combustion chamber of the engine body 1 based on a comparison between the excess air ratio λ and the target excess air ratio λmcd.The excess ratio calculation unit 25 calculates the excess air ratio λ of the exhaust gas using data LD obtained by linearizing the voltage value VHG to the excess air ratio while compensating for temperature characteristics thereof on the basis of the voltage value VHG from the voltage calculation unit 24 and the temperature value T from the temperature calculation unit 23. However, this calculation is applied in a case where the voltage value VHG is equal to or less than the lean side threshold LREF as described later, and in the case where the voltage value VHG is greater than the lean side threshold LREF, the excess air ratio λ is obtained by another method.The excess ratio calculation unit 25 includes: a torque calculation unit 32 that calculates a torque value TQ of the engine based on the crank angle speed NETC of the engine; a limit threshold value setting unit 33 that sets the conversion limit threshold value for the above-described linearization conversion; a storage unit 34 that stores data required for calculation of a substitute value R for the excess air ratio λ or a data map and a lookup table as described later; and a substitute value calculation unit 26 that calculates the substitute value R.The limit threshold setting unit 33 sets a lean-side threshold LREF, which is a conversion limit threshold on the lean side, and a rich-side threshold RREF, which is a conversion limit threshold on the rich side, as conversion limit thresholds for the voltage value VHG from the voltage calculation unit 24. however, in the titanium dioxide sensor element 12 a, the dynamic range of output values (minimum and maximum values of each linear range of the sensor output voltage) changes with change in temperature, and therefore, the conversion limit threshold needs to be changed according to the temperature T from the temperature calculation unit 23.Referring also to FIG. 5, FIG. 5 illustrates the data map having a first scale value G 1 in the left-right direction in FIG. 5 corresponding to the temperature value T calculated by the temperature calculation unit 23, and a second scale value G 2 in the vertical direction in FIG. 5 corresponding to the voltage value VHG calculated by the voltage calculation unit 24, and having set numerical values of a plurality of the data LD associated with each other with the voltage value VHG and the temperature value T as coordinates. FIG. 5 also shows a diagram in which, on the data map, examples of lookup tables corresponding to graphs 35 and 36, respectively, are superimposed to obtain the lean-side and rich-side threshold values LREFand RREF.In other words, the data map shows a plurality of excess air ratio values having correspondence relations between the plurality of first scale values G 1 for the temperature value T and the plurality of second scale values G 2 for the voltage value VHG (detected value). The look-up table shows, with correspondence relations with the first scale value G 1, the rich side threshold value RREF, and the lean side threshold value LREF, which are used to distinguish which air-fuel ratio range the voltage value VHG corresponds to among the rich range, the stoichiometric range, and the lean range.The storing of the above data map and the lookup table in advance corresponds to the graphs 35 and 36 in the storage unit 34 in the ECU 15 enables easy acquisition and setting of the data LD linearized from the voltage value VHG using the data map and the lookup table and the lean side threshold LREF and the rich side threshold RREF.The graph 35 is prepared by, for example, setting an air-fuel ratio λ as the boundary between the lean region and the stoichiometric region to 1.02, obtaining a plurality of points on the above data map with the voltage value VHG and the temperature value T corresponding to this value as coordinates, and connecting the plurality of points to each other by line interpolation. Moreover, the graph 36 is prepared by, for example, setting an air-fuel ratio λ as the boundary between the stoichiometric range and the rich range to 0.98, obtaining a plurality of points on the above data map with the voltage value VHGand the temperature value corresponding to this value as coordinates, and connecting the plurality of points to each other by linear interpolation.For example, from the lookup table corresponding to the graph 35, in the case where the temperature value T from the temperature calculation unit 23 is t 0, the limit threshold value setting unit 33 is capable of setting the voltage value v 0 derived from the coordinate t 0 as the lean side threshold value LREF for the boundary between the lean range and the stoichiometric range. Similarly, from the lookup table corresponding to the graph 36, in the case where the temperature value T from the temperature calculation unit 23 is t 0, the limit threshold setting unit 33 is capable of setting the voltage value v 1 derived from the coordinate t 0 as the rich side threshold RREF for the boundary between the stoichiometric range and the rich range.Moreover, the storage unit 34 stores the execution time Ti 1 of the fuel injection by the fuel injection valve 6, the torque value TQ 1, and the excess air ratio λb with respect to the conversion limit threshold value LREF in the case where the voltage value VHG from the voltage calculation unit 24 is equal to or less than the conversion limit threshold value LREF as data required for calculation of the substitute value R.If the voltage value VHG überschreitet the conversion limit threshold value LREF, the equivalent value calculation unit 26 calculates the equivalent value R by the following equation (1) with the last fuel injection execution time as Ti2and the last torque value as TQ2:If the voltage value VHG überschreitet the conversion limit threshold value LREF, the excess ratio calculation unit 25 regards the substitute value R as the exhaust excess air ratio λ as the linearized data LD described above, instead of the excess air ratio λ.FIG. 3 illustrates the excess ratio (lambda) calculation process for calculating the excess air ratio λ in the excess ratio calculation unit 25. control by the ECU 15 including this excess ratio calculation process is performed on the basis of the pulse signal indicating the rotational angle position of the crankshaft 18 from the crank angle sensor 19 and in synchronization with the stroke of the internal combustion engine.At the start of the excess ratio calculation process, the torque calculation unit 32 calculates the engine torque TQ based on the crank angle speed NETC from the rotation speed calculation unit 27 in step S 1.In the calculation of the torque TQ, two angular speeds of the crankshaft of the internal combustion engine corresponding to the two successive strokes of the internal combustion engine including intake, compression, combustion expansion and exhaust strokes of the internal combustion engine are calculated, and based thereon, the torque generated by the internal combustion engine is accurately calculated (see Patent No. JP 6 254 633 B2).Then, in step S 2, the limit threshold setting unit 33 sets, based on the temperature value T from the temperature calculation unit 23, the lean-side threshold LREF, and the rich-side threshold RREFusing the lookup tables in FIG. 5 corresponding to the graphs 35 and 36.Subsequently, in step S 3, the voltage value VHGis acquired by the voltage calculation unit 24, and the voltage value VHGis corrected by the deviation VD erhalten by a voltage difference determination unit 43 described later to set the voltage value (detected value) VHGconfor control.Then, in step S 4, the aforementioned data map (FIG. 5 ) is sampled on the basis of the temperature value T acquired in step S 2 and the voltage value VHGcon acquired in step S 3, thereby obtaining the data LD obtained by linearizing the voltage value VHGcon to the excess air ratio λ while compensating for temperature characteristics thereof.Subsequently, in step S 5, it is determined whether the voltage value VHGcon detected in step S 3 is less than the rich side threshold value RREF set in step S 2. If the voltage value VHGcon is determined to be smaller, a flag F_DETECT is set to zero in subsequent step S 6, and the process proceeds to step S 16 to set the value of the above data LD as the excess air ratio value λ, and then to end the excess calculation process in FIG. 3.If it is determined in step S5 that the voltage value VHGcon is not less than the rich side threshold value RREF, then it is determined in step S7 whether the voltage value VHGcon detected in step S3 is greater than the lean side threshold value LREF set in step S2.If it is determined in step S7 that the above voltage value VHGcon is not greater, then in step S8, a calculation is made on the excess air ratio λ as the data LD obtained by linearizing the voltage value VHG from the voltage calculation unit 24 to the excess air ratio while compensating the temperature characteristics of the oxygen sensor 12, on the basis of: the voltage value Iref of the lean side threshold Iref and the voltage value rref of the rich side threshold rref detected in step S2; the excess air ratio λ value (in this embodiment, λ=1.02) as the boundary between the predetermined stoichiometric range and the lean range corresponding to the voltage value Iref; the excess air ratio λ value (in this embodiment, λ=0.98) as the boundary between the predetermined rich range and the stoichiometric range corresponding to the voltage value rref; and the voltage value VHG acquired in step S 3, and then the process proceeds to step S 9.Also referring to FIG. 4, if the excess air ratio λ is a variable #LMD (for example, 1.02) that allows the excess air ratio λ to be numerically set in advance as the boundary between the predetermined stoichiometric range and the lean range, and if the excess air ratio λ is a variable #RLMD (for example, 0.98) that allows the excess air ratio λ to be numerically set in advance as the boundary between the predetermined rich range and the stoichiometric range, the excess air ratio λ is a graph as shown in FIG. 4. In the graph, the horizontal axis in the left-right direction in FIG. 4 represents the voltage value VHG, and the vertical axis in the vertical direction in FIG. 4 represents the excess air ratio λ. Thus, for example, if the voltage value VHGis vhg1, the corresponding value λ1of the excess air ratio λ can be calculated by the following equation (2):In step S 9, the execution time Ti of the last fuel injection by the fuel injection valve 6 and the torque TQ calculated in step S 1 are stored as Ti 1 and TQ 1, respectively, and the excess air ratio λ with respect to the lean side threshold LREF from the storage unit 34 is stored as λb. At about the same time, a count-down timer value TIMER indicating an effective time of the aforementioned storage is reset with its predetermined initial value, #TMINT. Then, the flag F_DETECT is set to 1, and the process proceeds to step S16 to set the value of the data LD acquired in the above step S8 as the excess air ratio λ value LAMBDA, and then to end the excess ratio calculation process in FIG. 3.Here, the value of the data LD acquired in step S 8 is stored as λb. Here, it is preferable to store the moving average of the values of the data LD as λb. For example, an exponential moving average λa of the excess air ratio λ (data LD) obtained by the following equation (3) is stored as λb:Above, k1 is a moving average coefficient, and λab is a moving average value in the previous control cycle stored in the storage unit 34. For example, as the moving average coefficient k1, 0.34 is used.Moreover, above, the storage unit 34 preferably stores moving average values as the fuel injection execution time Ti 1 and the torque value TQ 1, respectively. For example, the exponential moving average TiFLT of the fuel injection execution time Ti is calculated by the following equation (4) and stored as Ti 1, and the exponential moving average TQFLT of the torque TQ is calculated by the following equation (5) and stored as TQ 1:Above, k2 and k3 are moving average coefficients, and TiFLTb and TQFLTb are moving average values in the previous control cycle stored in the storage unit 34. In this embodiment, as the moving average coefficients k1, k2, and k3, different values can be used, respectively.Then, if it is determined in step S 7 that the voltage value VHGcon detected in step S 3 is greater than the lean side threshold value LREF, it is determined in step S 10 whether the above-described down-count timer value TIMER has reached zero. If the TIMER has reached zero, the flag F_DETECT is reset to 0 (step S11).Then, the process proceeds to step S 12 to determine whether the flag F_DETECT=1. If F_DETECT=1, it indicates that the storage unit 34 stores the excess air ratio λb with respect to the lean side threshold LREF, the fuel injection execution time Ti 1, and the torque value TQ 1, and therefore, the process proceeds to step S 13 to calculate the substitute value R in the substitute value calculation unit 26 with the above equation (1), and to set the value of the data LD to the substitute value R.Then, in step S14, it is determined whether the value of the data LD set in step S13 is larger than the predetermined upper limit value #LLMT. If the value of the data LD set in step S13 is larger than the upper limit value #LLMT, the value of the data LD is set to the upper limit value #LLMT (step S15). In this case, for example, 1.25 may be used as the upper limit value #LLMT.If F_DETECT=0 in the above step S 12, the substitute value R cannot be calculated because the equation indicates that the storage unit 34 does not store any valid values for the excess air ratio λb with respect to the lean side threshold LREF, the fuel injection execution time Ti 1, and the torque value TQ 1. In this case, the value of the data LD is still set to the above upper limit value #LLMT (step S15).The value of the data LD set in step S13 or step S15 is then set as the excess air ratio λ value LAMBDA (step S16), thereby completing the excess ratio calculation process in FIG. 3.After the excess ratio calculation process in FIG. 3 is completed, the ECU 15 uses PID control of the feedback coefficient calculation unit 30 to, in the control of the injection amount by the fuel injection valve 6, match the excess ratio λ value LAMBDA calculated in the excess ratio calculation process in FIG. 3 to the target air-fuel ratio λcmd from the target value calculation unit 28, as described above.FIG. 6 is a graph schematically illustrating changes in excess air ratio λ value LAMBDA calculated by the excess ratio calculation process in FIG. 3. The horizontal axis of the graph represents numerical values indicating the passage of time, and the vertical axis represents the excess air ratio λ.A graph 37 in FIG. 6 illustrates changes in numerical values of the excess air ratio λ value in the case of calculating the excess air ratio λ value using the data obtained by directly linearizing the voltage value VHG read by the voltage calculation unit 24 of the ECU 15 to the excess air ratio λ while compensating the temperature characteristics in the case where the current excess exhaust air ratio λ is gradually increased at a constant change rate in the range from the left end to the center of the horizontal axis in the left-right direction in FIG. 6, and then the current excess exhaust air ratio λ is gradually decreased at a constant change rate in the subsequent range from the center to the right end of the horizontal axis.Similarly, in a graph 38, in the case where the actual exhaust excess air ratio λ is gradually increased or decreased at a constant rate of change from the left end to the right end of the above horizontal axis, the excess air ratio λ value is calculated using data obtained by directly linearizing the voltage value VHG with the above data map (FIG. 5 ) or equation (2), in the case where the voltage value VHG from the voltage calculation unit 24 is equal to or less than the voltage value Iref of the lean-side threshold value Iref, while in the case where the voltage value VHG from the voltage calculation unit 24 exceeds the voltage value Iref (corresponding to an excess air ratio λ value of 1.020) of the lean-side threshold value LREF, The graph 38 illustrates changes in numerical values of the excess air ratio λ value when, instead of the data obtained by linearizing the above voltage value VHG, the substitute value R acquired in the above formula (1) is used as the excess air ratio λ value.Thus, if the current exhaust air excess ratio λ is 1.020 or less, the voltage value VHG from the voltage calculation unit 24 changes in response thereto in proportion (linearly) with respect to the above current exhaust air excess ratio λ. Therefore, if the current exhaust excess air ratio λ is 1.020 or less, both graphs 37 and 38 are linear transitions following the above constant changes in the current exhaust excess air ratio λ. However, if the exhaust air excess ratio λ exceeds 1.020, the voltage value VHGwhich may exhibit nonlinearity rapidly changes in the increasing direction, and therefore, the graph 37 which indicates the air excess ratio λ value based on the data obtained by directly linearizing the voltage value VHGalso changes steeply and nonlinearly in the increasing direction. On the other hand, in the graph 38, the excess air ratio λ value LAMBDA changes linearly with respect to the excess exhaust air ratio λ (air-fuel ratio) even in the case where the excess exhaust air ratio λ exceeds 1.020 (the above #LMD) where the change is accompanied with the current excess exhaust air ratio λ.Therefore, it is understood that the excess ratio calculation process is used to calculate the excess air ratio λ using the linearized data described above, if the voltage value VHG is equal to or less than the lean side threshold value LREF, and if the voltage value VHG exceeds the lean side threshold value LREF, the excess air ratio LAMBDA is calculated using the above formula (1) (graph 38), whereby the excess ratio calculation unit 25 is capable of supplying the feedback coefficient calculation unit 30 with an excess air ratio λ value that changes proportionately along with the current excess exhaust air ratio for the entire range of the graph in FIG. 6. This prevents the PID control from being interrupted by the feedback coefficient calculation unit 30.Then, a variation in the resistance values of the sensor element 12 (the detection unit) or the sensor heater 12 b(the heating unit) caused by manufacturing tolerances or the like results in an inaccurate excess air ratio obtained based on these resistance values, which may interfere with the air-fuel ratio feedback control.Therefore, the excess ratio calculation unit 25 includes: a characteristic checking unit 40 that checks the characteristics of the voltage value VHG (the detected value) that changes according to variation of the aforementioned resistance values; and a calibration unit 41 that calibrates the aforementioned data map and the lookup tables on the basis of the check results.The characteristic checking unit 40 includes: a temperature difference determining unit 42 that determines the amount of deviation TD of the temperature T from the reference value and sets a temperature Tcon for control; and a voltage difference determining unit 43 that determines a deviation between an average value VHGSTD within a predetermined period of time of the voltage value VHG of the sensor element 12 aand the threshold value on the lean side LREF and sets a voltage value (detected value) VHGcon for control, and then the characteristic checking unit 40 is configured to check the output characteristics of the oxygen sensor 12 connected to the ECU 15 according to the set temperature Tcon for control and the voltage value (detected value) VHG for control.The temperature difference determination unit 42 detects the deviation amount TD of the temperature T from the reference value if a predetermined time period sufficient for the temperature T of the sensor element 12 ato converge to the ambient temperature of the sensor element 12 ais elapsed while the internal combustion engine is stopped. The temperature difference determination unit 42 then sets the value Tcon (in this embodiment, Tcon=T+TD) obtained by correcting the temperature T of the sensor element 12 abased on the deviation amount TD as the temperature for checking the characteristics of the voltage value VHGand for control used in the excess ratio calculation unit 25.The reference value Tref for setting the above temperature Tcon can be obtained, for example, by adding the air intake temperature obtained at the intake air temperature sensor 8 to the engine temperature obtained from the cooling water temperature sensor 17, and then dividing the result by 2 {Tref=(engine temperature+inair air temperature)÷2}. Thereby, the above temperature Tcon for control can be accurately set by reducing the influence of a quantization error at each reading of the cooling water temperature sensor 17 and the air inlet temperature sensor 8.In this case, the temperature Tcon for controlling the engine temperature and the air intake temperature is set to be equal to each other under normal temperature and before starting. Thus, the temperature difference determination unit 42 moves the data map and the lookup tables corresponding to the above graphs 35 and 36 substantially parallel to the left-right direction (in the direction of the first scale value G 1) in FIG. 5.Subsequently, the voltage difference determination unit 43 detects a deviation VD between the average value VHGSTD of the voltage value VHG obtained from the resistance value of the sensor element 12 awithin a predetermined period and the lean side threshold value LREFin a case where the above temperature Tcon for controlling the sensor element 12 ais equal to or less than a predetermined value (for example, equal to the above reference value Tref). The voltage difference determination unit 43 then sets the voltage value VHGcon(=VHG+VD) obtained by correcting the voltage value VHG erhaltenen from the aforementioned resistance value on the basis of the deviation VD, as the voltage value (detected value) for checking the above characteristics and controlling the excess ratio calculation unit 25.In this case, in a situation where the temperature Tcon of the sensor element 12 ais sufficiently low and the oxygen sensor 12 is inactive, the wave height of the voltage value VHGis approximately zero (the voltage value VHGdoes not move), and the voltage value VHGrefis readable from the sensor element 12 a, the standard resistance value of which has the same values as those in the graphs 35 and 36 (the voltage value VHGrefis overlapped with the graphs 35 and 36). In other words, if the above temperature Tcon for control is equal to or less than a predetermined value, the voltage value VHGref is equal to the graph 35 (threshold value on the lean side LREF), and therefore, at the above sufficiently low temperature Tcon, the reference value for setting the voltage value (detected value) VHGcon for control is obtainable by sampling the look-up table corresponding to the graph 35, and can set the voltage value VHGcon for control with high accuracy.Thus, the voltage value (detected value) VHGconfor control is corrected to be the same as the graph 35 (lean side threshold value LREF) obtained in the case where the temperature Tconfor control is equal to or less than a predetermined value. This causes the voltage difference of the determination unit 43 to move the above data map substantially parallel to the vertical direction (toward the second scale value G 2) in FIG. 5.When the oxygen sensor 12 currently connected to the ECU 15 is checked, the characteristic checking unit 40 sets the target excess air ratio λcmd by the target value calculating unit 28 in the vicinity of stoichiometry to detect a peak value of the voltage value VHGcon, and also checks whether this peak value corresponds to one of the air-fuel ratio ranges.In the calibration by the calibration unit 41, the lookup tables corresponding to the above data map and the graphs 35 and 36 are calibrated in response to the results of the above test.FIG. 7 shows a voltage waveform 44 representing temporal changes in the voltage value (detected value) VGHref obtained by the characteristic checking unit 40 while a vehicle equipped with the oxygen sensor 12 having a standard resistance value in the internal combustion engine with the target excess air ratio λcmd set near stoichiometry is operating. FIG. 7 also shows an air-fuel ratio waveform 45 representing the waveform of an output signal of a wide-band air-fuel ratio sensor which is temporarily mounted for verification purposes.In a four-stroke internal combustion engine, the exhaust gas in the exhaust pipe pulsates and the oxygen concentration in the exhaust gas. Therefore, as illustrated in FIG. 7, the voltage waveform 44 of the voltage value (detected value) VHGref read out from the sensor element 12 aof the oxygen sensor 12 having standard resistance values and the air-fuel ratio waveform 45 of the attached air-fuel ratio sensor are observed as waveforms whose wave height has oscillating transitions over time. In particular, however, during the illustrated period P over which the air-fuel ratio waveform 45 spans the level where the excess air ratio λ is 1.0, the resistance value of the sensor element 12 afirstly has a characteristic that abruptly changes in the oxygen concentration in a stepwise (staircase) manner in the vicinity of stoichiometry. Therefore, the wave height M (the peak value of the detected value) of the voltage waveform 44 is observed as a waveform that oscillates very significantly as compared with the wave height of the air-fuel ratio waveform 45 during the same period P. Moreover, lookup tables corresponding to the graph 36 (rich side threshold RREF) and the graph 35 (lean side threshold LREF) are respectively set along the transition of the wave height of the voltage waveform 44 that oscillates greatly in this manner.Specifically, in the illustrated period P, among the peaks of the respective wave heights M in the voltage waveform 44 obtained from the oxygen sensor 12 having standard resistance values, a large number of rich side peaks 36 in the vicinity of the rich side threshold RREFin the lower part in FIG. 7 are approximately on the rich side threshold RREF, while a large number of the lean side peaks 47 in the vicinity of the lean side threshold LREFin the upper part in FIG. 7 are approximately on the lean side threshold LREF.In other words, the ECU 15 measures the wave height M of the voltage value (detected value) VHGcon of the oxygen sensor 12 currently connected to the ECU 15 while performing the air-fuel ratio feedback control so as to reproduce the illustrated state of the period P by setting the target excess air ratio λcmd in the vicinity of stoichiometry, and checks whether the peak of the wave height P matches the lean-side threshold LREF and the rich-side threshold RREF to be able to grasp and grasp the degree of deviation of the characteristics of the voltage value VHGcon from the characteristics of the voltage value VHGref in the above standard resistance value.Thus, the characteristic checking unit 40 is configured to detect the peak value of the wave height M of the voltage value (detected value) VHGcon of the oxygen sensor 12 connected to the ECU 15 while performing the air-fuel ratio feedback control with the target excess air ratio λcmd set in the vicinity of stoichiometry, and check whether the peak value of the wave height M corresponds to any one of the rich range, the stoichiometric range, and the lean range. Moreover, based on this check result, the calibration unit 41 is capable of calibrating the data map and the look-up table by applying affine transformation so that the peak value of the wave height M coincides with the rich side threshold RREF and the lean side threshold LREF.Specifically, the calibration unit 41 has a first calibration magnification value C 1 and a calibration magnification value C 2 for calibrating the data map and the lookup table by enlarging or reducing the plurality of first scale values G 1 and the plurality of second scale values G 2 by a so-called affine transformation procedure.For example, as will be described in detail with reference to FIGS. 8A to 8D, if the lean-side peak 47 vis in the lean range, when comparing the lean-side peak 47 vof the voltage value VHGconfor control that changes over time with the lookup table (lean-side threshold LREF) corresponding to the graph 35, the calibration unit 41 increases the first calibration increase value C 1 to increase the data map and the lookup table toward the right side as illustrated in FIG. 8A, whereas if the lean-side peak 47 vis in the stoichiometric range, on the other hand, the calibration unit 41 reduces the first calibration magnification value C 1 and reduces the data map and the look-up table to the left side as illustrated in FIG. 8B.In other words, the calibration unit 41 calibrates the data map and the lookup table by substantially enlarging or reducing the data map and the lookup table in the so-called affine transformation procedure by multiplying the first scale value G 1 of the data map and the lookup table by the first calibration enlargement value C 1 enlarged or reduced as above. Moreover, zero Kelvin (minus 273.15° C.; absolute zero point) may be applied as the origin of expansion or contraction of the first scale value G 1. When the value before calibration of the plurality of first scale values G 1 stored in the storage unit 34 is Tk (temperature in degrees Celsius), the first scale value Tkcal for post-calibration control is obtained by the following formula (6 a):For example, if the rich side peak 46 vis in the rich range, when comparing the rich side peak 46 vof the voltage value VHGconfor control that changes over time with the lookup table corresponding to the graph 36 (the rich side threshold RREF), the calibration unit 41 increases the second calibration multiplication value C 2 to increase the data map and the lookup table downward as illustrated in FIG. 8D, whereas if the rich side peak 46 vis in the stoichiometric range, the calibration unit 41 decreases the second calibration increase value C 2 to decrease the data map and the lookup table upward as illustrated in FIG. 8C.In other words, the calibration unit 41 calibrates the data map and the lookup table by substantially enlarging or reducing the data map and the lookup table in the so-called affine transformation procedure by multiplying the second scale value G 2 of the data map and the lookup table by the second calibration enlargement value C 2 enlarged or reduced as above. Moreover, the value of the graph 35 (lean side threshold LREF) used when the above temperature Tcon for control is equal to or less than the predetermined value is applicable as the origin of expansion or contraction of the second scale value G 2.In this state in which LREFanchor is the value of the graph 35 in the case where the above temperature Tcon for control is equal to or less than the predetermined value and in the Vk is the value stored in the storage unit 34 of the plurality of second scale values G 2 before calibration, the second scale value Vkcal for control after calibration is obtained by the following equation (7):In this embodiment, the calibration unit 41 is configurable to include a transition process to cause a gradual transition to the completed calibration state by gradually increasing or decreasing the first calibration magnification value C 1 or the second calibration magnification value C 2 by intervening a predetermined gradual increase or decrease rate as the first calibration magnification value C 1 and the second calibration magnification value C 2 become larger or smaller.As described above, according to this embodiment, the data map and the look-up table are calibrated such that the peak values of the voltage value VHGcon correspond to the rich side threshold value RREF and the lean side threshold value LREF in the case where the target excess air ratio λcmd is set near stoichiometry, and therefore, an accurate excess air ratio λ can be calculated to perform the appropriate air-fuel ratio feedback control regardless of the tolerance (variation) of respective resistance values of the sensor element 12 aand the sensor heater 12 b.Moreover, the temperature difference determination unit 42 detects the deviation amount TD of the temperature T of the sensor element 12 afrom the reference value to correct the temperature T on the basis of the deviation amount TD, thereby being able to perform more accurate air-fuel ratio feedback control.When the temperature of the sensor element 12 ais equal to or lower than a predetermined value which is sufficiently low, for example, the voltage difference determination section 43 obtains a correction value (deviation VD) for the voltage value VHG, for example, the wave height of the voltage value VHG is approximately zero and the voltage VHG does not fluctuate to correct the voltage value VHG by the correction value, so as to be able to perform more accurate air-fuel ratio feedback control.Further, the characteristic checking unit 40 and the calibration unit 41 increase (gradually increase) the first calibration increase value C 1 in the case where the lean-side peak value 47 vis in the lean range, and decrease (gradually decrease) the first calibration increase value C 1 in the case where the lean-side peak value 47 vis in the stoichiometric range to calibrate the data map and the lookup table such that the observed lean-side peak value 47 vis closer to the lean-side threshold value LREF, thereby being able to perform more accurate air-fuel ratio feedback control.Moreover, the characteristic checking unit 40 and the calibration unit 41 increase (gradually increase) the second calibration increase value C 2 in the case where the rich side peak 46 vlies in the rich range, and decrease (gradually decrease) the second calibration increase value C 2 in the case where the rich side peak 46 vlies in the stoichiometric range, to calibrate the data map and the lookup table such that the observed rich side peak 46 vlies closer to the rich side threshold RREF, thereby being able to perform more accurate air-fuel ratio feedback control.Moreover, in the control of the fuel injection amount Ti by PID control based on the excess air ratio λ, if the voltage value VHG indicating the oxygen concentration in the exhaust gas exceeds the lean side threshold value LREF, the substitute value R calculated by the above formula (1) is regarded as the excess air ratio λ to prevent interruption of the PID control, to increase the control accuracy, and to improve the efficiency of the exhaust purification.Further, the storage unit 34 stores the moving average values of the fuel injection execution time Ti 1, the torque value TQ 1, and the excess air ratio λb as these values, respectively, thereby enabling a decrease in quantification noise (error) when converting these measured values into digital values.Although the embodiments of the present invention have been described above, the present invention is not limited to the above embodiments, and various structural changes can be made without departing from the scope of the present invention described in the claims.For example, in the above embodiments, the plurality of first scale values G 1 of the data map and the lookup table are respectively multiplied by the first calibration magnification values C 1 to calibrate the data map and the lookup table by substantially enlarging or reducing the data map and the lookup table by the so-called affine transformation procedure (formula (6 a)). However, the present invention is not limited thereto. For example, it is also possible to substantially increase or decrease the data map and the lookup table even in the configuration in which the temperature for control of sampling the data map and the lookup table is multiplied by the first calibration magnification value C 1. Here, if Tk (temperature in degrees Celsius) is the value before calibration of the plurality of first scale values G 1 stored in the storage unit 34 and Tkcal is the first scale value for control after calibration, the temperature value Tcon (temperature in degrees Celsius) for scan control to scan the data map and the look-up table is obtained by the following formula (6 b):Regardless of whether the above formula (6a) or (6b) is used, calibration is performed such that the peak values of the actually measured voltage values (detected values) VHGcon conform to the threshold values on the rich side and the lean side (the wave height peaks overlap the threshold values) to calculate an accurate excess air ratio, thereby being able to perform appropriate air-fuel ratio feedback control.In the above-described embodiment, the reference value Tref for setting the temperature Tcon is set by adding the intake air temperature obtained from the intake air temperature sensor 8 to the engine temperature obtained from the cooling water temperature sensor 17, and dividing the result by 2 {Tref=(engine temperature+introair air temperature)÷ 2}, but the present invention is not limited thereto. For example, the reference value Tref for setting the temperature Tcon may also be calculated from either the engine temperature obtained from the cooling water temperature sensor 17 or the air inlet temperature obtained from the air inlet temperature sensor 8.Description of Reference Numerals1 Engine body 2 Air intake pipe 3 Throttle valve 4 Air cleaner 5 Throttle sensor 6 Fuel injection valve 7 Intake pressure sensor 8 Intake air temperature sensor 9 Piston 10 Exhaust pipe 11 Catalyst 12 Oxygen sensor 12 aSensor element (detection unit) 12 b Sensor heater 13 Spark plug 14 Ignition device 15 ECU (electronic control unit) 17 Cooling water temperature sensor 18 Crankshaft 19 Crank angle sensor 19 a Rotor 19 b Aufnehmer 20 Atmospheric pressure sensor 22 Heater controller 23 Temperature calculation unit 24 Voltage calculation unit 25 Excess ratio calculation unit 26 Substitute value calculation unit 27 Rotational speed calculation unit 28 Target value calculation unit 29 Basic injection amount calculation unit 30 Feedback coefficient calculation unit 31 Injection amount calculation unit 32 Torque calculation unit 33 Limit threshold value setting unit 34 Storage unit 35 to 38, and 35 b, 36 b, 35 c, 36 c, 35 d, 36 d, 35 e, 36 e Graph 40 Characteristic checking unit 41 Calibration unit 42 Temperature difference determination unit 43 Voltage difference determination unit 44 Voltage waveform 45 Air-fuel ratio waveform 46 Rich-side peak 47 Lean-side peak 46 v, 47 v

Claims

An air-fuel ratio control apparatus comprising: a resistance type oxygen sensor provided in contact with exhaust gas of an internal combustion engine including exhaust pulses, including a detection unit whose resistance value changes substantially stepwise at the oxygen concentration in the vicinity of stoichiometry of the exhaust gas, and in which a detected value obtained from the resistance value of the detection unit produces a pulsed waveform having a peak value that responds to a temperature of the detection unit and to the exhaust pulses; a temperature reading unit that estimates or detects the temperature of the detection unit; and an excess ratio calculation unit that calculates an excess air ratio with respect to a data map, which indicates the plurality of excess air ratio values having correspondence relations between a plurality of first scale values for temperature and a plurality of second scale values for the detected value, wherein the air-fuel ratio control device performs an air-fuel ratio feedback control based on a deviation between the excess air ratio and a target excess air ratio, wherein the air-fuel ratio control device further comprises: a storage unit which, having the correspondence relations with the first scale values, stores the data map and a look-up table, wherein the look-up table indicates a rich side threshold and a lean side threshold for discriminating which air-fuel ratio range the detected value is below a rich range, a characteristic checking unit that sets the target excess air ratio to be near stoichiometry to detect a peak value of the detected value and checks whether the peak value corresponds to one of the air-fuel ratio ranges; and a calibration unit that calibrates the data map and the look-up table by an affine transformation procedure according to the check so that the peak value corresponds to the rich side threshold value and the lean side threshold value.The air-fuel ratio control device according to claim 1, wherein: the characteristic checking unit includes a temperature difference checking unit that detects an amount of deviation of the temperature of the detection unit from a reference value if, with the engine stopped, a predetermined period of time sufficient for the temperature of the detection unit to converge to an ambient temperature of the detection unit has elapsed; and a value obtained by correcting the temperature of the detection unit based on the amount of deviation is used as the temperature of the detection unit for control used in the check and the excess ratio calculation unit.The air-fuel ratio control device according to claim 2, wherein: the characteristic checking unit includes a voltage difference determination unit that detects, in a case where the temperature of the detection unit of the controller is equal to or lower than a predetermined value, a deviation between an average value within a predetermined period of time of the detected value obtained from the resistance value of the detection unit and the threshold value on the lean side; and as the detected value for control used in the check and the excess ratio calculation unit, there is used a value obtained by correcting the detected value from the resistance value based on the deviation.The air-fuel ratio control device according to claim 3, wherein: the calibration unit stores a first calibration magnification value to calibrate the data map and the look-up table by scaling up or down the plurality of first scale values; and in a case where the peak wave height near the lean side threshold value is compared with the lean side threshold value for control that changes over time, the calibration unit increases the first calibration magnification value for the lean region peak and decreases the first calibration magnification value for the stoichiometric region peak.The air-fuel ratio control device according to claim 3, wherein: the calibration unit has a second calibration magnification value for calibrating the data map and the look-up table by multiplying the plurality of second scale values, the rich side threshold value, and the lean side threshold value; and the calibration unit enlarges the second calibration magnification value in a case where the wave height peak closer to the rich side threshold value in the wave height of the detected value for control that changes over time is in the rich range, and the calibration unit reduces the second calibration magnification value in a case where the peak is in the stoichiometric range.

Images