Sensor unit and method for controlling sensor unit

Abstract

To improve the accuracy of a calculated air-fuel ratio.SOLUTION: A sensor unit 20 includes a sensor element 21 that outputs a detection signal Ip corresponding to the oxygen concentration of detected gas Ge, and a calculation unit 22A that converts the detection signal Ip into an air-fuel ratio of the detected gas Ge using a conversion formula F. The calculation unit 22A offsets the conversion formula F based on distribution information D that indicates the statistical distribution of the detection signal Ip, which has been acquired in advance.SELECTED DRAWING: Figure 4

Description

[Technical field]

[0001] The present invention relates to a sensor unit and a method for controlling a sensor unit. [Background technology]

[0002] In order to measure the air-fuel ratio of exhaust gas discharged from an internal combustion engine, an oxygen sensor (sensor unit) is known that determines the air-fuel ratio from the output of a gas sensor element (sensor element) provided in an exhaust path. In such an oxygen sensor, a detection signal output by the gas sensor element is converted into an air-fuel ratio using a predetermined conversion formula. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2022-15533 Summary of the Invention [Problem to be solved by the invention]

[0004] The conversion equation for calculating the air-fuel ratio from the detection signal is created based on the design (theoretical) output characteristics, but the actual output characteristics vary due to individual differences. The causes of the variation include, for example, leakage of the gas to be measured and the environmental temperature during measurement. Therefore, the air-fuel ratio calculated using a conversion equation that does not reflect the variation in the output characteristics may be a value different from the actual air-fuel ratio.

[0005] The present invention aims to improve the accuracy of the calculated air-fuel ratio by offsetting the conversion equation based on previously acquired information on the variation in output characteristics (distribution information). [Means for solving the problem]

[0006] (1) A gas sensor unit comprising a sensor element that outputs a detection signal corresponding to the oxygen concentration of a gas to be detected, and a calculation unit that converts the detection signal into an air-fuel ratio of the gas to be detected using a conversion equation, and the conversion unit offsets the conversion equation based on distribution information that indicates a statistical distribution of the detection signal, which has been acquired in advance.

[0007] The detection signal of a sensor element is affected by individual differences in the sensor element and the surrounding environment at the time of detection. Therefore, even if the oxygen concentration of the detected gas is the same, the magnitude of the detection signal varies from sensor element to sensor element. The distribution information is information that indicates the statistical distribution of the detection signal, which is obtained in advance experimentally or by simulation. In addition, in this specification, "offset" refers to changing the value of the intercept of the conversion equation.

[0008] By offsetting the conversion formula based on previously acquired statistical distribution information, the air-fuel ratio can be calculated taking into account the influence of variations, thereby improving the accuracy of the air-fuel ratio calculated by the sensor unit.

[0009] (2) The distribution information may be information indicating a distribution of the detection signals when the detection target gas having a theoretical air-fuel ratio is measured by a plurality of the sensor elements.

[0010] By offsetting the conversion equation based on the distribution information obtained when the stoichiometric air-fuel ratio is measured, the accuracy of the calculated air-fuel ratio can be improved, particularly when the gas to be detected has an air-fuel ratio that is the stoichiometric air-fuel ratio or an air-fuel ratio close to the stoichiometric air-fuel ratio.

[0011] (3) The calculation unit may offset the conversion equation so that, when the detection signal output by the sensor element matches any one of an average value, a median value, or a mode value of the detection signal included in the distribution information, an air-fuel ratio calculated using the detection signal matches a theoretical air-fuel ratio.

[0012] In this way, for example, when the sensor outputs a value identical to the average value included in the distribution information, the difference between the calculated air-fuel ratio and the theoretical air-fuel ratio can be set to zero. This makes it possible to improve the accuracy of the calculated air-fuel ratio. Note that, similarly, the accuracy of the calculated air-fuel ratio can be improved when the detection value output by the sensor matches the median or mode included in the distribution information.

[0013] (4) The calculation unit may offset the conversion equation so that, when the detection signal output by the sensor element matches a central value of a variation range of the detection signal included in the distribution information, an air-fuel ratio calculated using the detection signal matches a theoretical air-fuel ratio.

[0014] By using the offset conversion formula, the detection signal when the sensor measures the gas to be detected having the stoichiometric air-fuel ratio can be set to the center value of the air-fuel ratio variation range. As a result, in the worst sensor where the detection signal when measuring the gas having the stoichiometric air-fuel ratio indicates an end value of the variation range, the difference between the calculated air-fuel ratio and the stoichiometric air-fuel ratio becomes small, and the detection accuracy can be improved. Effect of the Invention

[0015] According to the present invention, the accuracy of the calculated air-fuel ratio can be improved. [Brief description of the drawings]

[0016] [Figure 1] An explanatory diagram showing the relationship between the sensor unit and peripheral devices. [Diagram 2] Conversion formula graph [Diagram 3] Graph showing the distribution curve of pump current [Figure 4] FIG. 13 is an explanatory diagram showing the relationship between the pump current distribution, the conversion formula, and the calculated air-fuel ratio deviation Δ. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] <Embodiment 1> 1. Overall structure A first embodiment of the present invention will be described with reference to Fig. 1 to Fig. 4. Fig. 1 is an explanatory diagram that shows a schematic configuration of an internal combustion engine 10 provided with a sensor unit 20 according to the first embodiment, and an ECU (Electronic Control Unit) 11 that controls the operation of the internal combustion engine 10. The sensor unit 20 measures the air-fuel ratio of exhaust gas (an example of a "gas to be detected") Ge discharged from the internal combustion engine 10 mounted on a vehicle. The air-fuel ratio is the mixture ratio (mass ratio) of air and gasoline.

[0018] The internal combustion engine 10 is a four-stroke spark ignition engine. An intake pipe 12 through which air drawn in from the outside flows is connected to the upstream side of the internal combustion engine 10. An exhaust pipe 13 through which exhaust gas Ge generated by the combustion of fuel flows is connected to the downstream side of the internal combustion engine 10.

[0019] A fuel supply unit 14 is installed in the intake path of the intake pipe 12. The fuel supply unit 14 is composed of a throttle valve that adjusts the amount of intake air, an injector that injects fuel, etc. The fuel supply unit 14 supplies a mixture of air and fuel to the internal combustion engine 10 in response to a control signal from the ECU 11.

[0020] The catalytic device 15 is provided in the exhaust pipe 13. The catalytic device 15 is, for example, a three-way catalyst or an oxidation catalyst. The catalytic device 15 purifies emissions (hydrocarbons (HC), carbon monoxide (CO), nitrogen oxides (NOx), etc.) contained in the exhaust gas Ge moving through the exhaust pipe 13.

[0021] The ECU 11 is a device for electronically controlling the operation of the internal combustion engine 10, and is configured with a microcomputer chip equipped with a CPU, a ROM, a RAM, etc. The ROM stores various control programs and maps to be referred to when executing the various control programs. Information on the air-fuel ratio of the exhaust gas Ge obtained by the sensor unit 20 is input to the ECU 11. The ECU 11 feeds back the input information on the air-fuel ratio to the operation control of the internal combustion engine 10.

[0022] The sensor unit 20 includes a sensor element 21 and a sensor control unit 22. The sensor element 21 is, for example, a full-range air-fuel ratio sensor (linear lambda sensor) that outputs a signal whose magnitude changes linearly according to the oxygen concentration. The sensor element 21 is installed on the path of the exhaust pipe 13, downstream of the catalytic device 15.

[0023] The sensor element 21 is electrically connected via a signal line to a sensor control unit 22 disposed at a position away from the sensor element 21. The sensor element 21 is provided with a heater 21A. When the sensor element 21 is heated and activated by the heater 21A, it becomes possible to detect the oxygen concentration. The activated sensor element 21 outputs a pump current (an example of a "detection signal") Ip having a magnitude corresponding to the concentration of oxygen contained in the exhaust gas Ge (an example of a "detection process").

[0024] The sensor control unit 22 has a calculation unit 22A and a storage unit 22B. The calculation unit 22A is a so-called CPU, and executes a control program stored in the storage unit 22B to control the energization of the sensor element 21 and the heater 21A. The storage unit 22B is a storage medium such as a ROM or a RAM, and stores a conversion formula F and distribution information D in addition to the control program.

[0025] 2. About conversion formula F The conversion formula F is a formula showing the relationship between the pump current Ip output by the sensor element 21 and the air-fuel ratio deviation Δ of the exhaust gas Ge. Here, the air-fuel ratio deviation Δ is the difference between the air-fuel ratio of the exhaust gas Ge and the stoichiometric air-fuel ratio. Δ=0 means that the air-fuel ratio of the exhaust gas Ge matches the stoichiometric air-fuel ratio (stoichiometric). Δ<0 means that the air-fuel ratio of the exhaust gas Ge is rich (more fuel, less oxygen). 0<Δ means that the air-fuel ratio of the exhaust gas Ge is lean (less fuel, more oxygen).

[0026] The theoretical air-fuel ratio is a constant, for example 14.6. Therefore, if the air-fuel ratio is known, the air-fuel ratio deviation Δ is uniquely determined, and conversely, if the air-fuel ratio deviation Δ is known, the air-fuel ratio is uniquely determined. In this specification, the sensor unit 20 calculates the air-fuel ratio deviation Δ, but "calculation of the air-fuel ratio deviation Δ" and "calculation of the air-fuel ratio" are essentially synonymous. In the claims, the "air-fuel ratio" is calculated as the "air-fuel ratio deviation Δ" in this specification.

[0027] In this embodiment, the conversion formula F is offset as described below, but the initial conversion formula before the offset is defined as conversion formula F1. The conversion formula F1 is stored in the storage unit 22B.

[0028] Conversion formula F1 is a theoretical relational expression between the air-fuel ratio deviation Δ and the pump current Ip corresponding to the air-fuel ratio deviation Δ, assuming that the sensor element 21 has ideal characteristics. It is known that conversion formula F1 is expressed by combining linear expressions with a common intercept b and different slopes, such as the following formulas (1R) and (1L).

[0029] Ip=a R Δ+b (Δ<0: rich side)···equation (1R) Ip=a L Δ+b (Δ>0:Lean side)...Formula (1L) Figure 2 shows a graph plotting conversion formula F1 with the air-fuel ratio deviation Δ on the horizontal axis and the pump current Ip on the vertical axis. The region on the left, Δ<0, is the rich side with less oxygen and more fuel, and the region on the right, Δ>0, is the lean side with more oxygen and less fuel. Δ=0 means that the air-fuel ratio of the exhaust gas Ge is equal to the theoretical air-fuel ratio.

[0030] The conversion formula F1 consists of a rich side formula (1R) and a lean side formula (1L). R is the slope of the rich side, a L is the lean side slope. a R and a L The magnitude relationship of a R >a L b is the intercept, but in the transformation formula F1, b=0, and the graph passes through the origin.

[0031] The graph of conversion formula F1 shown in Figure 2 is a graph in which the pump current Ip changes linearly on both the rich and lean sides versus the air-fuel ratio deviation Δ (or air-fuel ratio). The intercepts are the same value on both the rich and lean sides (both zero in Figure 2), and the graph of formula (1R) and the graph of formula (1L) are continuous across Δ=0.

[0032] Calculation of the air-fuel ratio deviation Δ using the conversion formula F1 will be described with reference to Fig. 2. For example, when the value of the pump current Ip output by the sensor element 21 is Ip0, the air-fuel ratio deviation Δ calculated by the calculation unit 22A is Δ0. This shows that the air-fuel ratio of the exhaust gas Ge is larger (leaner) than the stoichiometric air-fuel ratio by Δ0. Also, when the value of the pump current Ip output by the sensor element 21 is zero, the air-fuel ratio deviation Δ of the exhaust gas Ge is zero. In this case, it is shown that the air-fuel ratio of the exhaust gas Ge is equal to the stoichiometric air-fuel ratio.

[0033] The calculation process (an example of a "conversion process") for calculating the air-fuel ratio deviation Δ from the pump current Ip using the conversion formula F1 and a conversion formula F2 described later is performed by the calculation unit 22A.

[0034] 3. Distribution Information D Conversion formula F1 is a formula that expresses the relationship between the pump current Ip and the air-fuel ratio deviation Δ when the characteristics of the sensor element 21 are as theoretical values ​​assumed at the time of designing the sensor. However, various variation factors affect the characteristics of each sensor element 21 that is actually manufactured, so there are individual differences in the characteristics of each sensor element 21.

[0035] The variation factors include, for example, the dimensions of the sensor, the measuring device and measuring method, the environmental temperature during the measurement and the temperature of the sensor element 21, gas leakage into and out of the sensor, and differences in fuel composition.

[0036] 3 shows a graph of distribution information D that statistically represents the distribution of the characteristics of the sensor element 21. The distribution information D is information obtained by measuring a reference gas known to have a theoretical air-fuel ratio (in other words, an air-fuel ratio deviation Δ=0) using a plurality of sensor elements 21 and summarizing the values ​​of the pump current Ip output by each sensor element 21 as a statistical distribution. The distribution information D is experimentally obtained in advance and stored in the memory unit 22B.

[0037] The distribution curve 31 shown in Fig. 3 indicates the variation distribution when the pump current Ip exhibits a normal distribution centered on zero. In this case, the range (3σ) of the pump current Ip is from -X (rich side) to +X (lean side), and the average value, median value, and mode value are all zero. The variation width is 2X.

[0038] If the influence of all variation factors on the sensor characteristics is symmetrical between the rich side and the lean side, the distribution information D will be a distribution in which the center of the variation range of the pump current Ip indicates zero, as shown in distribution curve 31. However, among the variation factors, there are variation factors that affect only the rich side of the sensor characteristics and variation factors that affect only the lean side. Therefore, as a result of combining the influences of multiple variation factors, the center of the variation range may not be zero.

[0039] The distribution curve 32 is a curve showing a distribution in which the shape of the curve showing the variation in the pump current Ip is the same as that of the distribution curve 31, but the center of the variation range is not zero but shifted to the lean side. In this specification, a case will be described in which the variation range of the pump current Ip when the exhaust gas Ge having a theoretical air-fuel ratio is measured has sensor characteristics shifted to the lean side as shown by the distribution curve 32.

[0040] The distribution information D is statistical information including the distribution curve 32 and the variation range, maximum value, minimum value, average value, median value, mode value, and the like of the distribution curve 32.

[0041] Distribution curve 32 is a curve obtained by shifting distribution curve 31 in parallel by p to the lean side. The variation range (3σ) of pump current Ip is from -X+p (rich side) to +X+p (lean side). The variation of pump current Ip is, for example, a normal distribution, with the mean, median, and mode each being p. The variation width is 2X.

[0042] 4. About the offset of conversion formula F1 Next, the offset (offset processing) of the conversion formula F1 will be described. The offset of the conversion formula F1 means changing the magnitude of the intercept b of the above-mentioned formula (1R) and formula (1L) based on the distribution information D (distribution curve 32, etc.). When the magnitude of the intercept b is changed, the graph of the conversion formula F1 shown in FIG. 2 moves in parallel in the vertical direction on the paper.

[0043] Fig. 4 is a diagram for explaining the offset of conversion formula F1 by combining the graph of conversion formula F1 in Fig. 2 with the graph of distribution curve 32 in Fig. 3. The graph of distribution curve 32 located in the upper left of the page of Fig. 4 is obtained by rotating and inverting the graph of distribution curve 32 described above so that the axis of pump current Ip is up and down (Ip is larger at the top) and the occurrence frequency is left and right (occurrence frequency is higher at the right).

[0044] In addition, the vertical axis of the graph of conversion formula F located at the upper right of the page and the vertical axis of the graph of distribution curve 32 located at the upper left both indicate the magnitude of pump current Ip, and the zero points and scales of the two vertical axes are assumed to be the same. At the lower right of the page, arrows (51, 55, 56) indicating the calculation results of the air-fuel ratio deviation Δ and distribution curves (61, 62, 63) of the calculated air-fuel ratio deviation Δ are shown.

[0045] 4.1 When using conversion formula F1 as is First, the calculation performed when the air-fuel ratio deviation Δ is calculated using the initial conversion formula F1 as it is when the sensor characteristics vary as shown in the distribution curve 32 will be visually explained with reference to Fig. 4. This calculation is performed by the calculation unit 22A.

[0046] The point on the distribution curve 32 in FIG. 4 that corresponds to the average value p is designated as point 40. A dashed line 41 extending in the left-right direction from point 40 intersects with the graph of conversion formula F1 at point 42. A dashed line 43 extending in the up-down direction from the intersecting point 42 intersects with the horizontal axis (Δ axis) at point 44. The value of Δ at point 44 is designated as the air-fuel ratio deviation Δ1C corresponding to the pump current Ip=p. This means that when the pump current Ip output by the sensor element 21 is Ip=p, if the pump current Ip is converted to the air-fuel ratio deviation Δ using the conversion formula F1, the calculated air-fuel ratio deviation Δ becomes Δ1C. Since the intercept of the graph of conversion formula F1 is zero (Δ=0 when Ip=0), Δ1C does not become zero if the average value p is not zero.

[0047] When the pump current Ip is at the minimum value (-X+p) in the distribution range, a dashed line 46 extending left and right through point 45 indicating the minimum value (-X+p) intersects with the graph of conversion formula F1 at point 47. The air-fuel ratio deviation Δ corresponding to point 47 is defined as Δ1R. Similarly, when the pump current Ip is at the maximum value (X+p) in the distribution range, a dashed line 49 extending left and right from point 48 intersects with the graph of conversion formula F1 at point 50. The air-fuel ratio deviation Δ corresponding to point 50 is defined as Δ1L.

[0048] As described above, when the air-fuel ratio deviation Δ is calculated using the initial conversion formula F1 based on the distribution information D (the minimum, maximum, and average values ​​that can be read from the distribution curve 32), the variation range of the air-fuel ratio deviation Δ is in the range of Δ1R-0-Δ1L around Δ=0, as shown by the arrow 51. Moreover, the air-fuel ratio deviation Δ corresponding to the average value p of the pump current Ip is Δ1C. The distribution of the air-fuel ratio deviation Δ calculated from the distribution curve 32 of the pump current Ip using the initial conversion formula F1 is shown in the distribution curve 61.

[0049] A distribution curve 61 is a curve showing the distribution of the air-fuel ratio deviation Δ calculated using the conversion formula F1 when the exhaust gas Ge having the theoretical air-fuel ratio is measured by the sensor element 21 having the characteristics shown in the distribution curve 32. The peak of the distribution curve 61 occurs when the air-fuel ratio deviation Δ=Δ1C corresponding to Ip=p, and the peak air-fuel ratio deviation Δ is greater than zero. In other words, the sensor unit 20 using the conversion formula F1 does not calculate the value Δ=0 even with a sensor element 21 having average characteristics (measuring the exhaust gas Ge having the theoretical air-fuel ratio (Δ=0) and outputting the pump current Ip=p).

[0050] The slope of the conversion formula F1 is such that the slope on the rich side is larger than the slope on the lean side at Δ=0 (a R >a L ), the distribution curve 61 has an asymmetric shape with a wider base on the lean side than on the rich side.

[0051] 4.2 Offsetting the conversion formula F1 based on the average value p As described above, when the air-fuel ratio deviation Δ is calculated from the pump current Ip using the initial conversion formula F1, when the pump current Ip is the average value p, the air-fuel ratio deviation Δ does not become 0, but becomes Δ=Δ1C. In other words, even if the exhaust gas Ge with an air-fuel ratio deviation Δ=0 is measured using the conversion formula F1, the calculated air-fuel ratio deviation Δ does not become zero, and the calculated value includes an error (Δ1C). When the conversion formula F1 is used, the calculated air-fuel ratio deviation Δ and the actual air-fuel ratio deviation Δ do not match, so there is room for improvement in the calculation accuracy of the air-fuel ratio deviation Δ.

[0052] In this embodiment, the calculation unit 22A generates a conversion equation F2 by offsetting the initial conversion equation F1 based on the average value p of the pump current Ip included in the distribution information D. The calculation unit 22A also calculates the air-fuel ratio deviation Δ using the conversion equation F2. The process of offsetting the conversion equation F1 to generate the conversion equation F2 and then calculating the air-fuel ratio deviation Δ from the pump current Ip using the conversion equation F2 will be described with reference to the graph in FIG. 4. This process is performed by the calculation unit 22A.

[0053] In the distribution curve 32 in FIG. 4, the point where the dashed line 41 extending in the left-right direction from the point 40 corresponding to the average value p of the distribution curve 32 intersects with the vertical axis (Ip axis) of the conversion formula graph is defined as point 52. Point 52 is the point where the air-fuel ratio deviation Δ=0 and the pump current Ip=p. The graph of conversion formula F2 (dash-dotted line) is obtained by translating the graph of conversion formula F1 upward (towards +Ip) so as to pass through point 52. In other words, the conversion formula F2 is an equation in which the intercept b of conversion formula F1 is changed from zero to p. The offset amount when generating conversion formula F2 by offsetting conversion formula F1 is the difference p between the intercepts of both formulas F1 and F2.

[0054] The formula (2R) and formula (2L) that express the conversion formula F2 are as follows. The conversion formula F2 differs from the conversion formula F1 only in the intercept, and the slope (a R , a L ) are identical. Ip=a R Δ+p (Δ<0: rich side)...Formula (2R) Ip=a L Δ+p (Δ>0:Lean side)...Formula (2L)

[0055] Next, the air-fuel ratio deviation Δ is calculated using the generated conversion equation F2. A dashed line 46 extending in the left-right direction from a point 45 indicating the minimum value (-X+p) of the pump current Ip intersects with the graph of the conversion equation F2 at a point 53. The air-fuel ratio deviation Δ corresponding to the point 53 is set as Δ2R. The magnitude relationship between Δ1R and Δ2R is Δ2R<Δ1R.

[0056] A dashed line 49 extending left and right from point 48 indicating the maximum value (X+p) of the pump current intersects with the graph of conversion formula F2 at point 54. The air-fuel ratio deviation Δ corresponding to point 54 is set to Δ=Δ2L. The magnitude relationship between Δ1L and Δ2L is Δ2L<Δ1L.

[0057] A dashed line 41 extending left and right from point 40, which indicates the average value p of the pump current, intersects with the graph of conversion formula F2 at point 52. The air-fuel ratio deviation Δ corresponding to point 52 is defined as Δ2C. Point 52 is a point where the air-fuel ratio deviation Δ=0, and Δ2C=0. The magnitude relationship between Δ1C and Δ2C is Δ2C<Δ1C.

[0058] The variation range of the air-fuel ratio deviation Δ calculated using the conversion formula F2 and the distribution of the air-fuel ratio deviation Δ are indicated by an arrow 55 and a distribution curve 62 (a dashed dotted line), respectively. The variation range is Δ2R-0-Δ2L, with Δ=0 in between.

[0059] 5.Effects In the conversion formula F1 and conversion formula F2, the slope of the rich side a R Leaner side inclination a L is smaller, the difference between Δ1L and Δ2L on the lean side is larger than the difference between Δ1R and Δ2R on the rich side (see arrows 51, 55). Therefore, the variation range (Δ2R to Δ2L) obtained by conversion formula F2 is smaller than the variation range (Δ1R to Δ1L) obtained by conversion formula F1.

[0060] Also, the point (Δ2L) farthest from Δ=0 in the range of arrow line 55 is closer to Δ=0 (theoretical air-fuel ratio) than the point (Δ1L) farthest from Δ=0 in the range of arrow line 51. The range of arrow line 55 includes Δ=0.

[0061] When conversion formula F2 is used instead of conversion formula F1, the variation in the calculated air-fuel ratio deviation Δ changes from arrow 51 to arrow 55, and the variation range becomes smaller. In addition, since the variation range indicated by arrow 55 is a range including Δ=0, when conversion formula F2 is used, the variation range becomes smaller in a range including Δ=0. This makes it possible to improve the calculation accuracy of the air-fuel ratio deviation Δ, particularly in the vicinity of the stoichiometric air-fuel ratio (Δ=0).

[0062] When the calculation accuracy of the air-fuel ratio deviation Δ in the vicinity of the theoretical air-fuel ratio is improved, the sensor unit 20 transmits a more accurate air-fuel ratio of the exhaust gas Ge to the ECU 11. The ECU 11 controls the fuel supply unit 14 and the internal combustion engine 10 based on the air-fuel ratio, and as a result, the emitted emissions are reduced.

[0063] Conversion formula F2 is generated by offsetting conversion formula F1 so that the air-fuel ratio deviation Δ corresponding to pump current Ip=p (average value of distribution curve 32) becomes zero. Using conversion formula F2, Δ=0 (stoichiometric air-fuel ratio) is calculated as the air-fuel ratio deviation Δ corresponding to the average value p. In other words, when exhaust gas Ge having a stoichiometric air-fuel ratio is measured using sensor element 21 with average characteristics, conversion formula F2 can be used to calculate the air-fuel ratio deviation Δ=0 without error. This can improve the accuracy of the air-fuel ratio deviation Δ calculated when exhaust gas Ge has a stoichiometric air-fuel ratio or its vicinity.

[0064] In the present embodiment, the conversion formula F2 is generated by offsetting the conversion formula F1 so that the air-fuel ratio deviation Δ corresponding to the average value p becomes zero, but the conversion formula F2 may be offset so that the air-fuel ratio deviation Δ becomes zero at a value other than the average value p. For example, if the distribution curve 32 is not a normal distribution, the average value, median value, and mode value of the distribution of the pump current Ip may be different values. In this case, instead of the above-mentioned average value, it is possible to offset the conversion formula F1 so that the air-fuel ratio deviation Δ corresponding to the median value or mode value of the pump current Ip becomes zero. This makes it possible to zero the air-fuel ratio deviation Δ corresponding to the median value or mode value of the pump current Ip, and improves the accuracy of the air-fuel ratio deviation Δ when the output pump current Ip is close to the median value or mode value.

[0065] <Embodiment 2> Next, the sensor unit 120 according to the second embodiment will be described. In the second embodiment, the conversion formula F1 is offset so that the variation range of the air-fuel ratio deviation Δ corresponding to the variation range of the pump current Ip is a symmetric range centered around Δ=0 (theoretical air-fuel ratio), and a new conversion formula F3 is generated. The second embodiment differs from the first embodiment only in the method of generating the conversion formula F3, and the device configurations of the sensor unit 20 and the sensor unit 120 and other configurations are the same as those of the first embodiment shown in FIG. 1. The contents of the distribution information D acquired in advance are also the same as those of the first embodiment, and the variation distribution of the pump current Ip is represented by a distribution curve 32. The same components as those of the first embodiment are denoted by the same reference numerals as those of the first embodiment, and description thereof will be omitted.

[0066] If the offset amount when generating the conversion formula F3 from the conversion formula F1 is B, the conversion formula F3 is expressed by the following formulas (3R) and (3L). Ip=a R Δ+B (Δ<0: rich side)···equation (3R) Ip=a L Δ+B (Δ>0: lean side) ··· formula (3L)

[0067] Solving for Δ respectively gives the following equations (4R) and (4L). Δ=(Ip-B) / a R (Δ<0: rich side) Equation (4R) Δ=(Ip-B) / a L (Δ>0: lean side) ··· formula (4L)

[0068] As shown in Fig. 4, the ends of the variation range of the pump current Ip are the minimum value Ip = -X+p on the rich side indicated at point 45 and the maximum value Ip = X+p on the lean side indicated at point 48. Substituting the minimum value Ip = -X+p into equation (4R) and the maximum value Ip = X+p into equation (4L), the following equations (5R) and (5L) are obtained. Δ=(-X+pB) / a R (Δ<0: rich side) Formula (5R) Δ=(X+pB) / a L (Δ>0: lean side) ··· formula (5L)

[0069] If the offset amount B is set so that the magnitude of the air-fuel ratio deviation Δ corresponding to each of the two ends of the variation range of the pump current Ip is equal on the negative side (lean side, Δ<0) and the positive side (rich side, Δ>0), the variation range of the air-fuel ratio deviation Δ can be made symmetrical with Δ=0 as the center. When the right side of equation (5R) multiplied by -1 is equal to the right side of equation (5L), the following equation (6) is established. -(-X+pB) / a R =(X+pB) / a L ...Equation (6)

[0070] By solving equation (6) for B, equation (7) is obtained, and the offset amount B can be calculated. The calculation unit 22A offsets the conversion equation F1 by the offset amount B to generate the conversion equation F3. A graph of the conversion equation F3 (two-dot chain line) is shown in FIG. B = (a R (X+p)+a L (-X+p)) / (a R +a L )...Equation (7)

[0071] The calculation of the air-fuel ratio deviation Δ using the conversion formula F3 will be explained with reference to Fig. 4. A dashed line 46 extending in the left-right direction from a point 45 indicating the minimum value of the pump current (Ip = -X + p) intersects with the graph of the conversion formula F3 at a point 57. The air-fuel ratio deviation Δ corresponding to the point 57 is defined as Δ3R. The magnitude relationship between Δ1R, Δ2R, and Δ3R is Δ3R<Δ2R<Δ1R.

[0072] A dashed line 49 extending left and right from point 48 indicating the maximum value of the pump current (Ip=X+p) intersects with the graph of conversion formula F3 at point 58. The air-fuel ratio deviation Δ corresponding to point 58 is defined as Δ3L. The magnitude relationship between Δ1L, Δ2L, and Δ3L is Δ3L<Δ2L<Δ1L.

[0073] The point on the conversion formula F3 that corresponds to the average value of the pump current (Ip=p) is point 59. The air-fuel ratio deviation Δ that corresponds to point 59 is defined as Δ3C. The magnitude relationship between Δ1C, Δ2C, and Δ3C is Δ3C<Δ2C<Δ1C.

[0074] The arrow 56 indicates that the variation range of the air-fuel ratio deviation Δ calculated using the conversion formula F3 is within the range of Δ3R to Δ3L. The distribution curve 63 (two-dot chain line) is a curve that indicates the distribution of the air-fuel ratio deviation Δ calculated using the conversion formula F3.

[0075] <Effects of the second embodiment> When conversion formula F1 is used, the variation range of the air-fuel ratio deviation Δ is larger on the lean side (0 to Δ1L) than on the rich side (Δ1R to 0), as shown by arrow 51 and distribution curve 61. Similarly, when conversion formula F2 is used, the variation range of the air-fuel ratio deviation Δ is larger on the lean side (0 to Δ2L) than on the rich side (Δ2R to 0), as shown by arrow 55 and distribution curve 62.

[0076] When the conversion formula F1 and the conversion formula F2 are used, the variation range of the air-fuel ratio deviation Δ is significantly wider on the lean side, and the center of the variation range is on the lean side. This is because, in each of the conversion formulas F1 and F2, the gradient a R Leaner side inclination a L When the variation in the air-fuel ratio deviation Δ is large on the lean side, the error included in the calculated value of the air-fuel ratio deviation Δ becomes large on the lean side.

[0077] On the other hand, when conversion formula F3 is used, the variation range of the air-fuel ratio deviation Δ becomes symmetrical with respect to Δ=0, as shown by arrow 56. The size of the variation range on the rich side (Δ3R to 0) is equal to the size of the variation range on the lean side (0 to Δ3L). By using conversion formula F3, the variation range of the air-fuel ratio deviation Δ becomes the same on the rich side and the lean side, and is not biased to either side. As a result, the variation range on the lean side becomes relatively smaller compared to the rich side, and the calculation accuracy of the air-fuel ratio deviation Δ can be improved, especially on the lean side.

[0078] Also, consider the sensor element 21 that outputs pump current Ip=X+p when exhaust gas Ge with a theoretical air-fuel ratio is measured. Such a sensor element 21 has sensor characteristics corresponding to point 48 at the end of distribution curve 32, and is sometimes called the worst sensor. When exhaust gas Ge with a theoretical air-fuel ratio is measured with this worst sensor, and conversion formula F3 is used, the air-fuel ratio deviation Δ3L is calculated from point 58, which is the intersection point between dashed line 49 and the graph of conversion formula F3.

[0079] Even with the same worst sensor, when conversion formula F1 is used, the air-fuel ratio deviation Δ1L is calculated from point 50, which is the intersection point between the dashed line 49 and the graph of conversion formula F1, and when conversion formula F2 is used, the air-fuel ratio deviation Δ2L is calculated from point 54, which is the intersection point between the dashed line 49 and the graph of conversion formula F2. Compared with Δ1L and Δ2L, Δ3L is closer to Δ=0 and has a smaller absolute value.

[0080] Even for a sensor element 21 whose sensor characteristics are far from the center of the variation range, such as the worst sensor, the calculated air-fuel ratio deviation Δ becomes small by using conversion formula F3. The fact that the air-fuel ratio deviation Δ calculated when measuring exhaust gas Ge with a theoretical air-fuel ratio becomes small means that the calculation accuracy of the air-fuel ratio deviation Δ becomes high. The sensor unit 20 transmits a more accurate air-fuel ratio of the exhaust gas Ge to the ECU 11, and as a result, the emissions discharged from the internal combustion engine 10 are reduced.

[0081] <About the upper limit of offset amount B> As in the above-described first and second embodiments, by using the conversion equations F2 and F3 generated by offsetting, it is possible to reduce the variation range of the air-fuel ratio deviation Δ and bring the center of the variation range closer to Δ = 0. For example, when the conversion equation F1 is offset by the offset amount B calculated using the above-described equation (7), the variation range is reduced from the range of the arrow 51 to the range of the arrow 56, and further, the center of the variation range becomes Δ = 0.

[0082] However, depending on the offset amount B, the use of the offset conversion equation may result in a larger variation range than the use of the conversion equation before the offset. In terms of reducing the maximum absolute value of the air-fuel ratio deviation Δ, there is a certain upper limit to the offset amount B where a reduction effect is expected.

[0083] Specifically, within the variation range of the air-fuel ratio deviation Δ calculated by the conversion formula F3 after offset (arrow 56, Δ3R to Δ3L), if the maximum value of the absolute value on the rich side, Δ = Δ3R, does not exceed the maximum value of the absolute value on the lean side, Δ = Δ1L, within the variation range of the air-fuel ratio deviation Δ calculated by the initial conversion formula F1 (arrow 51, Δ1R to Δ1L), the maximum value of the absolute value of the air-fuel ratio deviation Δ can be reduced.

[0084] The upper limit value of the offset amount where Δ3R does not exceed Δ1L is defined as the offset amount B1. The offset amount B1 can be calculated as follows.

[0085] In the conversion formula F1, the maximum value of the air-fuel ratio deviation Δ on the lean side is Δ1L. Substituting Δ = Δ1L into Equation (1L) gives the following Equation (8L). Δ1L=(X + p) / aL ··· Equation (8L)

[0086] The maximum value (absolute value) of the air-fuel ratio deviation Δ expected on the rich side in the conversion formula F3 is obtained by substituting Δ = -Δ3R into Equation (3R), resulting in the following Equation (8R). -Δ3R=(-X + p - B) / aR ··· Equation (8R)

[0087] Solving -Δ3R < Δ1L, Equation (8R), and Equation (8L) for the offset amount B gives the following Equation (9). B < aR((X + p) / aL + (-X - p) / aR) ··· Equation (9)

[0088] From Equation (9), the upper limit value of the offset amount B can be obtained. When the offset amount B satisfies Equation (9), the maximum value of the absolute value of the air-fuel ratio deviation Δ calculated using the conversion formula after offset is smaller than the maximum value of the absolute value of the air-fuel ratio deviation Δ calculated using the initial conversion formula F1. In this case, the effect of improving the accuracy of the air-fuel ratio deviation Δ due to offset can be obtained. Note that the upper limit value of the offset amount B shown in Equation (9) is applicable to both Embodiment 1 and Embodiment 2.

[0089] <Other Embodiments> The present invention is not limited to the embodiments described above and illustrated in the drawings, and the following embodiments, for example, are also included within the technical scope of the present invention.

[0090] (1) In the above embodiment, a "full range air-fuel ratio sensor (linear lambda sensor)" that detects oxygen concentration is exemplified as the sensor element 21, but the present invention is not limited to this and can be applied to a linear type sensor element whose output changes linearly with respect to gas concentration. The type of gas to be detected is not limited to oxygen. In addition, two types of sensor elements, one-cell type and two-cell type, are known as sensor elements that detect oxygen concentration, and the present invention can be applied to either type of sensor.

[0091] (2) In the above embodiment, the sensor element 21 is installed downstream of the catalytic converter 15, but the sensor element 21 may be installed upstream of the catalytic converter 15 (between the internal combustion engine 10 and the catalytic converter 15).

[0092] (3) In the above embodiment, the distribution information D is information obtained experimentally in advance. However, the distribution information D may be information obtained by simulation.

[0093] (4) In the above embodiment, the distribution curve 32 is a normal distribution with p as the mean value. However, the distribution shown in the distribution curve is not limited to a normal distribution.

[0094] (5) In the above embodiment, the distribution information D is information indicating the distribution of detection signals when exhaust gas Ge having a theoretical air-fuel ratio is measured by a plurality of sensor elements. The gas measured when acquiring the distribution information does not have to have a theoretical air-fuel ratio, and may have a known air-fuel ratio.

[0095] (6) In the above embodiment, the conversion formula is offset based on any one of the average, median, and mode of the pump current Ip included in the distribution information D, or the center value of the variation range, but the conversion formula may be offset based on a value other than these. For example, the conversion formula may be offset based on any value other than the maximum or minimum value of the variation range. [Explanation of symbols]

[0096] 10: internal combustion engine, 12: intake pipe, 13: exhaust pipe, 14: fuel supply unit, 15: catalytic device, 20, 120: sensor unit, 21: sensor element, 21A: heater, 22: sensor control unit, 22A: calculation unit, 22B: memory unit, D: distribution information, F1 to F3: conversion formula, Ge: exhaust gas (an example of a "detected gas"), Ip: pump current (an example of a "detection signal")

Claims

1. a sensor element that outputs a detection signal corresponding to the oxygen concentration of a gas to be detected; a calculation unit that converts the detection signal into an air-fuel ratio of the detection target gas using a conversion formula, The calculation unit offsets the conversion equation based on distribution information indicating a statistical distribution of the detection signals, which is acquired in advance.

2. The sensor unit according to claim 1 , The distribution information is information indicating a distribution of the detection signals when the detection target gas having a theoretical air-fuel ratio is measured by a plurality of the sensor elements.

3. The sensor unit according to claim 2, the calculation unit offsets the conversion equation so that an air-fuel ratio calculated using the detection signal coincides with a theoretical air-fuel ratio when the detection signal output by the sensor element coincides with any one of an average value, a median value, or a mode value of the detection signal included in the distribution information.

4. The sensor unit according to claim 2, the calculation unit offsets the conversion equation so that an air-fuel ratio calculated using the detection signal coincides with a theoretical air-fuel ratio when the detection signal output by the sensor element coincides with a central value of a variation range of the detection signal included in the distribution information.

5. A detection process for outputting a detection signal corresponding to the oxygen concentration of the gas to be detected; an offset process for offsetting a conversion equation based on distribution information indicating a statistical distribution of the detection signals, the distribution information having been acquired in advance; A conversion process for converting the detection signal into an air-fuel ratio of the gas to be detected using an offset conversion equation.

Images