JP2005330922A - Data processing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accurately average the output of an AFM and the output of an AFM model with different sampling frequencies with a same response. <P>SOLUTION: A first averaging means 45 averages the output of the AFM for each specified period by sampling it at specified sampling periods. In this case, an averaged value is obtained by dividing the integrated value of the sampling data on the output of the AFM in the specified period by the number of times of sampling. On the other hand, a second averaging means 44 averages the output of the AFM model for each specified period by sampling it at specified sampling periods longer than the sampling periods of the output of the AFM. In this case, the number of averaged data is set by correcting the number of times of sampling in the specified period by using a correction factor, virtual data is set by correcting the sampling data on the output of the AFM model by using a correction factor, and an averaged value is obtained by dividing the sampling data on the output of the AFM model and the virtual data by the number of the averaged data. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a data processing apparatus that performs a process of averaging two types of signals having different sampling periods.

  In an intake air amount detection device for an internal combustion engine, for example, as described in Patent Document 1 (Japanese Patent Laid-Open No. 2002-147279), an AFM (air flow meter) that detects a flow rate of air flowing through an intake pipe of the internal combustion engine. And the output of an AFM model constructed so as to simulate the response delay of this AFM, the intake air amount is calculated.

In general, the output of the AFM arranged in the intake pipe of the internal combustion engine is influenced by the intake pulsation and becomes a pulsation waveform. For example, this is described in Patent Document 2 (Japanese Patent Publication No. 62-40645). As described above, the AFM output is sampled at a sampling period shorter than the pulsation period and averaged every predetermined period (for example, the intake stroke period), so that the average air flow rate in the predetermined period can be calculated with high accuracy. is there.
JP 2002-147279 A (page 4 etc.) Japanese Examined Patent Publication No. 62-40645 (Page 2 etc.)

  By the way, in the system for calculating the intake air amount using the output of the AFM and the output of the AFM model as in Patent Document 1, the output of the AFM is sampled at a sampling period shorter than the pulsation period as in Patent Document 2. When averaging is performed every predetermined period, it is necessary to average the output of the AFM model every same period.

  However, since the AFM model performs very complicated calculations, it is necessary to perform calculations with a calculation cycle longer than the sampling cycle of the AFM output (that is, a cycle shorter than the pulsation cycle) in order to reduce the load on the CPU. is there. For this reason, the output of the AFM model is sampled at a sampling period longer than the sampling period of the AFM output. As a result, the number of times of sampling of the output of the AFM model is reduced as compared with the number of times of sampling of the output of the AFM within a predetermined period in which the averaging process is performed, so that the accuracy of the averaging process of the output of the AFM model is deteriorated. There was a problem that the detection accuracy of the average air flow rate deteriorated.

  The present invention has been made in view of such circumstances, and therefore, an object of the present invention is to provide a data processing apparatus capable of averaging two types of signals having different sampling periods with high accuracy with the same response. There is to do.

  In order to achieve the above object, the data processing apparatus according to claim 1 of the present invention averages data obtained by sampling the first signal at the first sampling period at predetermined intervals by the first averaging processing means. At the same time, the data obtained by sampling the second signal in the second sampling period longer than the first sampling period is averaged at predetermined intervals by the second averaging processing means. Thereby, the timing which averages the sampling data of a 1st signal and the timing which averages the sampling data of a 2nd signal can be made the same timing each time.

  By the way, the first signal has a shorter sampling period and a larger number of samplings (the number of sampling data) within the predetermined period than the second signal, so that the sampling data is scattered in a balanced manner at short intervals throughout the predetermined period. Cheap. Therefore, if the data of the first signal sampled within a predetermined period is averaged, the average value of the first signal can be obtained with relatively high accuracy even during a transient change.

  On the other hand, the second signal has a longer sampling period and a smaller number of samplings (sampling data number) within the predetermined period than the first signal, and therefore a small number of sampling data within a predetermined period has a wide interval. Tend to be biased. For this reason, if only a small number of data sampled within a predetermined period is simply averaged, the calculation accuracy of the average value of the second signal within the predetermined period is deteriorated during a transient change.

  Therefore, in the invention according to claim 1, the second averaging processing means is configured to calculate the number of data to be averaged for each predetermined period based on at least the predetermined period and the second sampling period (hereinafter referred to as “averaged”). (The number of data ") is set to be larger than the actual number of data in the predetermined period, and the number of virtual data corresponding to the difference between the averaged number of data and the actual number of data is changed in the second signal. Based on the behavior, the averaging process is performed using the sampling data of the second signal within the predetermined period, the virtual data, and the averaged data number. In this way, even if a small number of second signal sampling data is biased within the predetermined period, the virtual data is used to substantially reduce the number of data of the second signal within the predetermined period. The second signal data (sampling data and virtual data) can be arranged in a well-balanced manner within a predetermined period, and the average value of the second signal within the predetermined period is relatively accurate even during transient changes. It can be calculated well. Accordingly, the average value of the first signal and the average value of the second signal within the predetermined period can be calculated accurately at the same timing each time, and the first signal and the second signal having different sampling periods can be calculated. Signals can be averaged with the same response and high accuracy.

  In this case, as in claim 2, when it is assumed that the first signal and the second signal exhibit the same behavior, the result of averaging the first signal by the first averaging processing means and the second A correction coefficient set so that the result of the averaging process of the second signal by the averaging processing means coincides is stored in advance, and the number of times of sampling of the second signal within a predetermined period is corrected with the correction coefficient. It is preferable to set the number of averaged data and obtain the virtual data by correcting the sampling data of the second signal close to the virtual data with the correction coefficient. In this way, the first signal and the second signal can be averaged with higher accuracy and the same response by a relatively simple calculation process.

  By the way, when the length of the predetermined period changes and the sampling number of the first signal and the sampling number of the second signal within the predetermined period change, the appropriate correction coefficient also changes accordingly. Therefore, as described in claim 3, it is preferable to set the correction coefficient according to at least one of the number of sampling times of the first signal and the number of sampling times of the second signal within a predetermined period. In this way, it is possible to set an appropriate correction coefficient in accordance with the number of times of sampling the first signal and the number of times of sampling the second signal within a predetermined period.

  In this case, the correction coefficient may be fixed to a predetermined value or the correction may be prohibited in a region where the number of times the second signal is sampled within a predetermined period. As the number of times of sampling (number of sampling data) of the second signal within the predetermined period increases, the bias of the sampling data within the predetermined period decreases, and the sampling data tends to be scattered throughout the predetermined period. Tends to approach a constant value, and the correction coefficient tends to be small and can be ignored. Therefore, in a region where the number of samplings of the second signal within a predetermined period is large, even if the correction coefficient is fixed to a predetermined value or the correction is prohibited, the first signal and the second signal are averaged with the same response and high accuracy. Can be In this case, if the correction coefficient is fixed to a predetermined value or prohibited from being corrected in a region where the number of samplings of the second signal is large, the arithmetic processing when performing the averaging process of the second signal can be simplified. .

  The present invention is not limited to a configuration in which the correction coefficient is changed according to the number of times of sampling the first signal and the number of the second signal within a predetermined period, and the correction coefficient is always set to a predetermined value as in claim 5. You may make it fix to. In this way, it is possible to further simplify the arithmetic processing when performing the averaging process of the second signal while ensuring the accuracy of averaging the first signal and the second signal with the same response. .

  Further, when setting the virtual data based on the sampling data of the second signal, the data sampled within the predetermined period and / or the sampling data of the second signal among the sampling data of the second signal as in claim 6 The virtual data may be calculated based on the data sampled immediately before. By doing this, it is possible to obtain virtual data that accurately reflects the change behavior of the second signal within the predetermined period of this time, and it is possible to perform averaging processing with higher accuracy.

  Further, the present invention is constructed so as to simulate the output of an air flow meter (hereinafter referred to as “AFM”) for detecting the flow rate of air flowing through the intake passage of the internal combustion engine and the output of this AFM. Applied to the system for calculating the intake air amount of the internal combustion engine based on the output of the AFM model, the first averaging processing means averages the output of the AFM as the first signal, and the second averaging processing The means may average the output of the AFM model as the second signal. In this way, even if the AFM model output is sampled at a sampling period longer than the sampling period of the AFM output, the AFM output and the AFM model output having different sampling periods are accurately averaged with the same response. It is possible to improve the detection accuracy of the average air flow rate.

  Hereinafter, two embodiments 1 and 2 in which the present invention is applied to an intake air amount detection device for an internal combustion engine will be described.

A first embodiment of the present invention will be described with reference to FIGS.
First, a schematic configuration of the entire engine control system will be described with reference to FIG. An air cleaner 13 is provided at the most upstream portion of the intake pipe 12 (intake passage) of the engine 11 that is an internal combustion engine, and an air flow meter (hereinafter referred to as “air flow meter”) that detects the air flow rate in the intake pipe 12 downstream of the air cleaner 13. 14) (referred to as “AFM”). On the downstream side of the AFM 14, a throttle valve 15 whose opening is adjusted by a DC motor or the like, and a throttle opening sensor 16 for detecting the throttle opening are provided.

  Further, a surge tank 17 is provided on the downstream side of the throttle valve 15, and an intake pipe pressure sensor 18 for detecting the intake pipe pressure is provided in the surge tank 17. The surge tank 17 is provided with an intake manifold 19 for introducing air into each cylinder of the engine 11, and a fuel injection valve 20 for injecting fuel is attached in the vicinity of the intake port of the intake manifold 19 of each cylinder. Yes. A spark plug 21 is attached to each cylinder of the engine 11 for each cylinder, and the air-fuel mixture in the cylinder is ignited by spark discharge of each spark plug 21.

  On the other hand, the exhaust pipe 22 of the engine 11 is provided with a catalyst 23 such as a three-way catalyst for purifying CO, HC, NOx and the like in the exhaust gas. / An exhaust gas sensor 24 (air-fuel ratio sensor, oxygen sensor, etc.) for detecting lean or the like is provided.

  A cooling water temperature sensor 25 that detects the cooling water temperature and a crank angle sensor 26 that outputs a pulse signal each time the crankshaft of the engine 11 rotates a predetermined crank angle are attached to the cylinder block of the engine 11. Based on the output signal of the crank angle sensor 26, the crank angle and the engine speed are detected.

  Outputs of these various sensors are input to an engine control circuit (hereinafter referred to as “ECU”) 27. The ECU 27 is mainly composed of a microcomputer, and executes various engine control programs stored in a built-in ROM (storage medium) to thereby determine the fuel injection amount of the fuel injection valve 20 according to the engine operating state. The ignition timing of the spark plug 21 is controlled.

  Next, the outline of the calculation processing of the intake air amount of the engine 11 by the ECU 27 will be described with reference to FIG. FIG. 2 is a block diagram illustrating the calculation processing function of the intake air amount of the ECU 27.

  As shown in FIG. 2, the electronic throttle model 36 outputs the predicted throttle opening TA0, the engine speed NE, the valve timing VT, and the like to the TA model 37 (throttle air model). Here, the predicted throttle opening degree TA0 is a throttle opening degree after a predetermined time has elapsed from the present time, and is estimated based on the current throttle opening degree TA or the like. The TA model 37 receives a predicted intake pipe pressure P0 output from an intake pipe model 38 to be described later.

  The TA model 37 calculates an air flow rate QA0 passing through the throttle valve 15 based on the predicted throttle opening degree TA0, the engine speed NE, the valve timing VT, the predicted intake pipe pressure P0, etc., and this air flow rate QA0 is calculated as the intake pipe model. 38.

  The intake pipe model 38 includes an intake valve model 39, and calculates a predicted intake pipe pressure P0 after a predetermined time has elapsed from the present based on the air flow rate QA0. The predicted intake pipe pressure P0 is a pressure predicted from the intake pipe pressure when the intake valve 35 is closed.

  On the other hand, the TA model 40 is input with the throttle opening degree TA, the engine speed NE, the valve timing VT and the like, and the current intake pipe pressure P1 output from the intake pipe model 41 described later.

  The TA model 40 calculates the current air flow rate QA1 passing through the throttle valve 15 based on the throttle opening degree TA, the engine speed NE, the valve timing VT, the current intake pipe pressure P1, and the like, and this air flow rate QA1 is taken into the intake air. Output to the tube model 41.

  The intake pipe model 41 includes an intake valve model 42, and calculates the current intake pipe pressure P1 based on the current air flow rate QA1.

  The current air flow rate QA 1 output from the TA model 40 is input to the AFM model 43. The AFM model 43 calculates an air flow rate QA2 having a time delay corresponding to the response delay of the AFM 14 with respect to the current air flow rate QA1. The air flow rate QA2 that is the output of the AFM model 43 is averaged by the second averaging processing means 44 described later and output to the intake pipe model 46.

  The intake pipe model 46 includes an intake valve model 47, and calculates an intake pipe pressure P2 having a time delay based on the air flow rate QA2.

  On the other hand, the air flow rate QA, which is the output of the AFM 14, is averaged by a first averaging processing unit 45 described later and input to the intake pipe model 48. The intake pipe model 48 includes an intake valve model 49, and calculates an intake pipe pressure P3 having a time delay based on the air flow rate QA. The intake pipe pressure P3 output from the intake pipe model 48 includes a time delay similarly to the intake pipe pressure P2 output from the intake pipe model 46, and has the same response as the intake pipe pressure P2.

  The intake pipe pressure P0 output from the intake pipe model 38 is added to the intake pipe pressure P3 output from the intake pipe model 48, and the intake pipe pressure P2 output from the intake pipe model 46 is subtracted. The predicted pressure P is obtained, and the intake air amount Q when the intake valve 35 is closed is calculated based on the predicted pressure P by a map or a mathematical formula. Thus, the intake air amount that compensates for the response delay of the AFM 14 is calculated.

  Next, the averaging process of the AFM 14 output by the first averaging processing unit 45 and the averaging process of the AFM model 43 output by the second averaging processing unit 44 will be described.

  The first averaging processing means 45 samples the AFM 14 output (first signal) at a predetermined sampling period (for example, 4 ms period), and the sampling data for a predetermined period (for example, in the case of a 4-cylinder engine, an intake stroke period). Are averaged every 180 ° C.).

  At this time, the first averaging processing means 45 first obtains the number of samplings N1 of the AFM 14 output within the predetermined period by dividing the time required for the predetermined period (for example, 180 ° C. A) by the sampling period of the AFM 14 output. Thereafter, the sampling data of the AFM 14 output sampled within a predetermined period is integrated, and the integrated value is divided by the number of times of sampling N1 to obtain the average value of the AFM 14 output within the predetermined period. Thus, the average value of the AFM 14 output within a predetermined period is calculated.

  On the other hand, the second averaging processing means 44 samples the output of the AFM model 43 (second signal) at a predetermined sampling period (for example, 8 ms period) longer than the sampling period of the AFM 14 output, and the sampling data is the first. Are averaged at the same predetermined period (for example, 180 ° C.) as the averaging processing means 45.

  At that time, the second averaging processing means 44 first divides the time required for a predetermined period (for example, 180 ° C. A) by the sampling period of the AFM model 43 output, and the number of samplings N2 of the AFM model 43 output within the predetermined period. Ask for. Thereafter, a correction coefficient α corresponding to the sampling number N1 of the AFM 14 output is obtained, and this correction coefficient α is added to the sampling number N2 of the output of the AFM model 43 to obtain the number of averaged data (N2 + α). Of the output sampling data, the NN2 (where NN2 is an integer satisfying N2 <NN2 ≦ N2 + 1) sampling data XQA2 is multiplied by a correction coefficient α from the end timing of the predetermined period to the past. Data (α × XQA2) is obtained. Thereby, virtual data is set based on the data sampled within the predetermined period this time or the data sampled immediately before the sampling data output from the AFM model 43.

  Then, (N 2 + α) data is accumulated from the end timing of the predetermined period to the past, that is, N 2 sampling data and virtual data (α × XQA 2) are accumulated toward the past, and the accumulated value is obtained. Is divided by the number of averaged data (N2 + α) to obtain the average value of the output of the AFM model 43 within a predetermined period. As a result, even if a small number of sampling data of the AFM model 43 output is biased with respect to the predetermined period, the number of data of the AFM model 43 output within the predetermined period is substantially increased by using virtual data. Thus, the AFM model 43 output data (sampling data and virtual data) can be arranged in a well-balanced manner within a predetermined period, and the average value of the AFM model 43 output within the predetermined period can be calculated relatively accurately even during a transient change. can do.

  Here, a method for setting the correction coefficient α will be described. 3 and 4, the correction coefficient α is calculated based on the result of the averaging process performed by the first averaging process unit 45 and the second value when it is assumed that the output of the AFM 14 and the output of the AFM model 43 show the same behavior. Are set so that the results of the averaging processing by the averaging processing means 44 coincide with each other.

  Hereinafter, in order to facilitate understanding, a case will be described in which the time required for rotation in a predetermined period (for example, 180 ° C. A) is 16 ms, the sampling period of the AFM 14 output is 4 ms, and the sampling period of the AFM model 43 output is 8 ms.

First, as shown in FIG. 3, the output of the AFM 14 is averaged by the averaging process by the first averaging processing unit 45. In this case, since the number of samplings of the AFM 14 output within the predetermined period is N1 = 16/4 = 4, four sampling data (a + 12b), (a + 11b), (a + 10b) from the end timing of the predetermined period to the past. ), (A + 9b), and the integrated value is divided by 4 to obtain an average value.
Average value = {(a + 12b) + (a + 11b) + (a + 10b) + (a + 9b)} / 4
= A + 10.5b (1)

  Next, as shown in FIG. 4, the output of the AFM model 43 is averaged by averaging processing by the second averaging processing means 44. In this case, since the number of samplings A2 of the AFM model 43 output within a predetermined period is N2 = 16/8 = 2, the correction coefficient α is added to the number of samplings N2 to set the number of averaged data (2 + α), and Virtual data {α × (a + 8b)} is set by multiplying the third sampling data (a + 8b) from the end timing of the predetermined period to the past by the correction coefficient α.

  Then, (2 + α) data is accumulated toward the past from the end timing of the predetermined period, that is, two sampling data (a + 12b) and (a + 10b) and virtual data {α × (a + 8b) toward the past. } And the average value is obtained by dividing the integrated value by the number of averaged data (2 + α).

Average value = {(a + 12b) + (a + 10b) + α × (a + 8b)} / (2 + α)
(2) Thereafter, when the equation with the right side of the above equation (1) = the right side of the above equation (2) is solved for the correction coefficient α, α = 0.4.

  When the correction coefficient α is calculated for a predetermined period of 20 ms, 24 ms, 28 ms,... In the same manner, as shown in FIG. 5, when the predetermined period is 20 ms, α = 0.5 and the predetermined period is 24 ms. In this case, α = 0.43, and when the predetermined period is 28 ms, α = 0.5,. In addition, since the sampling number N1 of the AFM 14 output and the sampling number N2 of the AFM model 43 output change according to the length of the predetermined period, in the first embodiment, the relationship between the sampling number N1 of the AFM 14 output and the correction coefficient α is shown. A prescribed map is created in advance and stored in the ROM of the ECU 27.

  Hereinafter, the ECU 27 realizes the processing contents of the first averaging processing program of FIG. 6 executed to realize the function of the first averaging processing means 45 and the function of the second averaging processing means 44. The processing contents of the second averaging processing program of FIG. 7 executed for this purpose will be described.

  The first averaging processing program shown in FIG. 6 is executed at the sampling period (for example, 4 ms) of the AFM 14 output. When this program is started, first, in step 101, the output QA of the AFM 14 is read as sampling data.

Thereafter, the process proceeds to step 102, where the number of samplings N1 of the AFM 14 output within the latest predetermined period (for example, 180 ° C. A) is calculated by the following equation.
N1 = (time required for a predetermined period) / (sampling period of AFM14 output)

  Thereafter, the process proceeds to step 103, and the count value of the counter C is incremented by “1”. The counter C counts up at the sampling period of the AFM 14 output (execution period of this program), thereby counting the total number of samplings of the AFM 14 output within a predetermined period.

Thereafter, the process proceeds to step 104, where the sampling data QASUM of the current AFM 14 output is added to the integrated value QASUM of the sampling data of the AFM 14 output up to the previous time to update the integration QASUM. At this time, the forward flow component is integrated as a positive value and the reverse flow component is integrated as a negative value in the AFM 14 output having a pulsation waveform.
QASUM = QASUM + QA (i)

  Thereafter, the routine proceeds to step 105, where it is determined whether or not the count value of the counter C has reached the number of sampling times N1. If the count value of the counter C has not reached the number of times of sampling N1, the process of counting up the count value of the counter C and integrating the sampling data of the AFM 14 output is repeated.

  Thereafter, when it is determined in step 105 that the count value of the counter C has reached the sampling number N1, the process proceeds to step 106, and the integrated value QASUM of the sampling data output from the AFM 14 is divided by the sampling number N1 to obtain a predetermined period. The average value QAav of the AFM output is obtained.

  Thereafter, the process proceeds to step 107, the count value of the counter C and the integrated value QASUM are cleared, and this program is terminated.

  The second averaging processing program shown in FIG. 7 is executed at the sampling period (for example, 8 ms) of the AFM model 43 output. When this program is started, first, in step 201, the output QA2 of the AFM model 43 is read as sampling data.

Thereafter, the process proceeds to step 202, where the number of samplings N2 of the AFM model 43 output within the latest predetermined period (for example, 180 ° C. A) is calculated by the following equation.
N2 = (time required for a predetermined period) / (sampling period of AFM model 43 output)

  Thereafter, the process proceeds to step 203, and the correction coefficient α corresponding to the sampling number N1 of the AFM 14 output is calculated using the map of the correction coefficient α shown in FIG.

  Thereafter, the process proceeds to step 204, where it is determined whether or not the count value of the counter C has reached the sampling number N1, and when it is determined that the count value of the counter C has reached the sampling number N, the process proceeds to step 205. Then, the correction coefficient α is added to the number of samplings N2 of the output of the AFM model 43 to obtain the averaged data number (N2 + α), and the sampling data of the AFM model 43 output from the end timing of the present predetermined period toward the past. Virtual data (α × XQA 2) is obtained by multiplying NN 2 (where NN 2 is an integer satisfying N 2 <NN 2 ≦ N 2 +1) by the correction coefficient α.

  Thereafter, the process proceeds to step 206, where (N 2 + α) data is accumulated toward the past from the end timing of the predetermined period, that is, N 2 sampling data and virtual data (α × XQA 2) are accumulated toward the past. The integrated value is divided by the number of averaged data (N 2 + α) to obtain the average value QA 2av of the AFM model 43 output within a predetermined period.

  For example, when the predetermined period is 16 ms, the sampling period of the AFM 14 output is 4 ms, and the sampling period of the AFM model 43 output is 8 ms, the number of samplings A2 of the AFM model 43 output within the predetermined period N2 = 2, and the correction coefficient α = 0. .4, the correction coefficient α is added to the number of samplings N2 to obtain the averaged data number (2 + 0.4), and the third sampling data QA2 ( Virtual data {0.4 × QA2 (i-2)} is obtained by multiplying i-2) by the correction coefficient α. In this case, the virtual data is calculated using the data sampled immediately before the predetermined period in the sampling data output from the AFM model 43.

Then, (2 + 0.4) pieces of data are accumulated toward the past from the end timing of this predetermined period, that is, two pieces of sampling data QA2 (i), QA2 (i-1) and virtual data are directed toward the past. {0.4 × QA2 (i-2)} is integrated, and the integrated value is divided by the number of averaged data (2 + 0.4) to obtain the average value QA2av of the AFM model 43 output within a predetermined period.
QA2av = [QA2 (i) + QA2 (i-1) + {0.4 × QA2 (i-2)}] / (2 + 0.4)

  When the predetermined period is 20 ms, the sampling period of the AFM 14 output is 4 ms, and the sampling period of the AFM model 43 output is 8 ms, the number of samplings A2 of the AFM model 43 output within the predetermined period is N2 = 2.5, and the correction coefficient α = 0.5, the correction coefficient α is added to the number of samplings N 2 to obtain the averaged data number (2.5 + 0.5), and the third data from the end timing of the predetermined period to the past Virtual data {0.5 × QA2 (i-2)} is obtained by multiplying the sampling data QA2 (i-2) by the correction coefficient α. In this case, virtual data is calculated using data sampled within the predetermined period of time among sampling data output from the AFM model 43.

Then, (2.5 + 0.5) pieces of data are accumulated toward the past from the end timing of this predetermined period, that is, 2.5 pieces of sampling data QA2 (i), QA2 (i-1) towards the past. ), {0.5 × QA2 (i-2)} and virtual data {0.5 × QA2 (i-2)} are integrated, and the integrated value is divided by the average number of data (2.5 + 0.5) Then, the average value QA2av of the output of the AFM model 43 within a predetermined period is obtained.
QA2av = [QA2 (i) + QA2 (i-1) + {0.5 × QA2 (i-2)}
+ {0.5 × QA2 (i-2)}] / (2.5 + 0.5)

  In the first embodiment described above, the sampling data of the AFM 14 output and the sampling data of the AFM model 43 output having different sampling periods are averaged for the same predetermined period, so the sampling data of the AFM 14 output is averaged. The timing and the timing at which the sampling data output from the AFM model 43 is averaged can be made the same every time.

  By the way, the output of the AFM 14 has a shorter sampling period and a larger number of samplings N1 (the number of sampling data) in the predetermined period than the output of the AFM model 43, so that the sampling data is likely to be distributed in a balanced manner throughout the predetermined period (see FIG. 3). ). Therefore, if the AFM 14 output data sampled during a predetermined period is averaged, the average value of the AFM 14 output can be calculated with relatively high accuracy even during a transient change.

  On the other hand, the output of the AFM model 43 has a longer sampling cycle and a smaller number of samplings N2 (the number of sampling data) within a predetermined period than the output of the AFM14. There is a tendency to be biased (see FIG. 4). For this reason, if only a small number of data sampled during a predetermined period is simply averaged, the accuracy of calculating the average value of the AFM model 43 output within the predetermined period is deteriorated during a transient change.

  In this respect, in the first embodiment, the number of samplings N2 of the AFM model 43 output is corrected with the correction coefficient α to set the number of averaged data (N2 + α), and the sampling data of the AFM model 43 output with the correction coefficient α. Since the corrected virtual data is set, the number of data output from the AFM model 43 within the predetermined period is substantially increased, and the data (sampling data and virtual data) output from the AFM model 43 within the predetermined period. They can be arranged in a well-balanced manner, and the average value of the output of the AFM model 43 within a predetermined period can be calculated with relatively high accuracy even during a transient change. As a result, the average value of the AFM 14 output and the average value of the AFM model 43 output within a predetermined period can be accurately calculated every time at the same timing, and the AFM 14 output and the AFM model 43 output having different sampling periods are the same response. Can be averaged with high accuracy.

  In addition, in the first embodiment, when it is assumed that the output of the AFM 14 and the output of the AFM model 43 exhibit the same behavior, the result of the averaging process by the first averaging processing unit 45 and the second averaging processing unit 44 Since the correction coefficient α is set so as to match the average processing result, the AFM 14 output and the AFM model 43 output can be averaged with higher accuracy and the same response.

  By the way, when the time required for a predetermined period (for example, 180 ° C.) changes due to the change of the engine speed and the sampling number N1 of the AFM 14 output and the sampling number N2 of the AFM model 43 output within the predetermined period change, accordingly. The appropriate correction coefficient α also changes. In consideration of such circumstances, in the first embodiment, since the correction coefficient α is set according to the sampling number N1 of the AFM 14 output within a predetermined period, the sampling number N1 of the AFM 14 output and the output of the AFM model 43 are output. An appropriate correction coefficient α can be set according to the number of samplings N2.

  In the first embodiment, the correction coefficient α is set according to the sampling number N1 of the AFM 14 output. However, the correction coefficient α is set according to the predetermined period and the sampling number N2 of the AFM model 43 output. May be.

  Further, in the first embodiment, the virtual data is set based on the data sampled within the predetermined period of time or the data sampled immediately before the predetermined period among the sampling data output from the AFM model 43. Virtual data that accurately reflects the behavior of the AFM model 43 output within a predetermined period can be set, and more accurate averaging processing can be performed.

  Note that the virtual data may be set based on both the data sampled within the predetermined period and the data sampled immediately before the predetermined period among the sampling data output from the AFM model 43.

  In the first embodiment, the correction coefficient α is changed in accordance with the sampling number N1 of the AFM 14 output in the entire region. However, the region in which the sampling number N2 of the AFM model 43 output is large within a predetermined period (that is, engine rotation). The correction coefficient α may be fixed to a predetermined value (for example, 0.5) or the correction may be prohibited in a region where the speed is low and the time required for the predetermined period is long. If the number of samplings N2 (the number of sampling data) of the output of the AFM model 43 within the predetermined period increases, the sampling data is less biased with respect to the predetermined period, and the sampling data tends to be scattered throughout the predetermined period. α tends to approach a constant value (for example, 0.5), or the correction coefficient tends to be small and can be ignored. Therefore, even if the correction coefficient α is fixed to a predetermined value (for example, 0.5) or correction is prohibited in a region where the number of samplings N2 of the AFM model 43 output within a predetermined period is large, the AFM 14 output and the AFM model 43 output are output. It is possible to average accurately with the same response. In addition, if the correction coefficient α is fixed to a predetermined value or prohibited in a region where the number of samplings N2 of the output of the AFM model 43 is large, the arithmetic processing when performing the averaging process of the AFM model 43 output can be simplified. it can.

  Alternatively, the correction coefficient α may always be fixed to a predetermined value (for example, 0.5). In this way, it is possible to further simplify the arithmetic processing when performing the averaging process of the AFM model 43 output while ensuring the accuracy of averaging the AFM 14 output and the AFM model 43 output with the same response.

  In the first embodiment, the correction coefficient α is obtained from the sampling number N1 of the AFM 14 output, and the averaged data number (N2 + α) is set by correcting the sampling number N2 of the AFM model 43 output using this correction coefficient α. However, in the second embodiment of the present invention, the process for obtaining the correction coefficient α is omitted, and the number of averaged data is directly calculated from the number of samplings N1 of the AFM 14 output using the map of the number of averaged data shown in FIG. N is calculated. Then, N pieces of data are integrated from the end timing of the predetermined period to the past, and the integrated value is divided by N to obtain the average value QA2av of the AFM model 43 output in the predetermined period.

For example, if the predetermined period is 16 ms, the sampling period of the AFM 14 output is 4 ms, and the sampling period of the AFM model 43 output is 8 ms, the number of samplings of the AFM 14 output within the predetermined period N1 = 4, and the number of average data N = 2 Therefore, 2.4 data is accumulated from the end timing of the predetermined period to the past, and the integrated value is divided by 2.4 to obtain an average value of the output of the AFM model 43 within the predetermined period. Find QA2av.
QA2av = [QA2 (i) + QA2 (i-1) + {0.4 * QA2 (i-2)}] / 2.4

When the predetermined period is 20 ms, the sampling period of the AFM 14 output is 4 ms, and the sampling period of the AFM model 43 output is 8 ms, the number of samplings A1 of the AFM 14 output within the predetermined period is N1 = 5, and the number of average data N = 3 Therefore, three data are accumulated from the end timing of the present predetermined period toward the past, and the integrated value is divided by 3 to obtain the average value QA2av of the AFM model 43 output in the predetermined period.
QA2av = {QA2 (i) + QA2 (i-1) + QA2 (i-2)} / 3
Also in the second embodiment described above, the same effect as the first embodiment can be obtained.

  In the second embodiment, the averaged data number N is set according to the sampling number N1 of the AFM 14 output. However, the averaged data number N is set according to the sampling period N2 of the AFM model 43 output for a predetermined period. You may make it set.

  In each of the first and second embodiments, the present invention is applied to a system that averages the output of the AFM 14 and the output of the AFM model 43 having different sampling periods. However, the present invention is not limited to this, and the present invention has different sampling periods. The present invention can be widely applied to a system that averages two types of signals. In particular, an in-vehicle CPU used for engine control and vehicle control needs to perform various kinds of control at high speed within the range of limited calculation capability, and thus it is difficult to newly add processing with a large calculation load. Since the two-signal averaging process of the present invention has a relatively small calculation load and can be sufficiently processed by an in-vehicle CPU, the effect of applying the present invention to engine control and vehicle control is great.

It is a schematic block diagram of the whole engine control system in Example 1 of this invention. It is a block diagram which shows the calculation processing function of intake air amount. FIG. 6 is a diagram (No. 1) for describing a method of setting a correction coefficient α. FIG. 6 is a diagram (No. 2) for describing a method of setting the correction coefficient α. It is a figure which shows notionally the map of correction coefficient (alpha). It is a flowchart which shows the flow of a process of a 1st averaging process program. It is a flowchart which shows the flow of a process of a 2nd averaging process program. It is a figure which shows notionally the map of the average data number N.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 11 ... Engine (internal combustion engine), 12 ... Intake pipe (intake passage), 14 ... AFM (air flow meter), 15 ... Throttle valve, 20 ... Fuel injection valve, 21 ... Spark plug, 22 ... Exhaust pipe, 27 ... ECU, 43 ... AFM model, 44 ... second averaging processing means, 45 ... first averaging processing means

Claims (7)

First averaging processing means for averaging data obtained by sampling the first signal at the first sampling period every predetermined period;
Second averaging processing means for averaging data obtained by sampling a second signal in a second sampling period longer than the first sampling period for each predetermined period;
The second averaging processing means calculates the number of data to be averaged for each predetermined period (hereinafter referred to as “averaged data number”) based on at least the predetermined period and the second sampling period. The number of virtual data corresponding to the difference between the averaged data number and the actual data number is estimated based on the change behavior of the second signal, A data processing apparatus that performs an averaging process using sampling data of the second signal within a period, the virtual data, and the averaged data number.   The second averaging processing means is configured to average the first signal by the first averaging processing means when it is assumed that the first signal and the second signal exhibit the same behavior. The correction coefficient set so that the result and the averaging process result of the second signal by the second averaging processing unit coincide with each other is stored in advance, and the sampling of the second signal within the predetermined period is stored. The average number of data is set by correcting the number of times with the correction coefficient, and the virtual data is obtained by correcting the sampling data of the second signal close to the virtual data with the correction coefficient. The data processing apparatus according to claim 1.   The second averaging processing means sets the correction coefficient in accordance with at least one of the number of sampling times of the first signal and the number of sampling times of the second signal within the predetermined period. The data processing apparatus according to claim 2.   4. The second averaging processing unit fixes the correction coefficient to a predetermined value or prohibits correction in an area where the number of times of sampling of the second signal within the predetermined period is large. The data processing apparatus described in 1.   3. The data processing apparatus according to claim 2, wherein the second averaging processing means always fixes the correction coefficient to a predetermined value.   The second averaging processing means calculates the virtual data based on data sampled within the predetermined period of time and / or data sampled immediately before the sampling data of the second signal. A data processing apparatus according to any one of claims 1 to 5. The intake of the internal combustion engine based on the output of an air flow meter (hereinafter referred to as “AFM”) that detects the flow rate of air flowing through the intake passage of the internal combustion engine and the output of an AFM model constructed to simulate the output of this AFM Applied to a system that calculates air volume,
The first averaging processing means averages the output of the AFM as the first signal, and the second averaging processing means averages the output of the AFM model as the second signal. The data processing apparatus according to claim 1, wherein:

Images