Method for updating allowable minimum injection duration in self-learning mode

Through the method of self-learning and updating the minimum injection time allowed, the problem of poor injection quantity accuracy during multiple injections of direct injection engines is solved, the fuel combustion economy is optimized, and the fuel accuracy is improved.

CN118934310AActive Publication Date: 2024-11-12DONGFENG MOTOR GRP
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202411211758.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-30
Publication Date
2024-11-12
Estimated Expiration
2044-08-30

AI Technical Summary

Technical Problem

During multiple injections, due to the inaccurate control of the minimum injection time duration, the injection volume accuracy is extremely poor, and the fuel oil is lean or thick, which has not been effectively solved by the prior art.

Method used

The method of self-learning update allows minimum injection time, and by judging the conditions for activating self-learning update, calculate the relationship between the current injection time and the allowable minimum injection time, and update the allowable minimum injection time according to the preset conditions and the number of times. Optimize injection duration to improve fuel accuracy.

Benefits of technology

Through self-learning update, the minimum injection time is allowed, which avoids the problem of lean or thick injection fuel, optimizes the combustion economy of fuel, and improves the accuracy of injection fuel injection volume.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118934310A_ABST
    Figure CN118934310A_ABST
Patent Text Reader

Abstract

The invention discloses a method for updating the allowable minimum injection duration through self-learning, which is used for controlling the minimum injection duration of a direct injection engine and comprises the following steps: calculating the relationship between the injection duration of a current injection cylinder and the allowable minimum injection duration determined by the last self-learning updating, and updating the minimum injection duration according to the relationship, whether fuel oil is too thick or too thin and other preset conditions. The working conditions are divided into four kinds of spraying duration increasing, four kinds of spraying duration decreasing and other working conditions, and self-learning updating is further carried out according to the classification result of the working conditions. According to the invention, the accurate control of the allowable minimum injection duration can be realized when the direct injection engine injects for multiple times.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of engine control, and more particularly to a method for self-learning and updating the minimum allowed injection duration. Background Art

[0002] In order to reduce the vibration and noise of direct injection engines, improve engine cold start and emission performance, and reduce the risk of oil dilution, it is required to set different injection times according to different engine working conditions. In multiple injections, if the injection time of a certain injection is too short, due to the nonlinear characteristics of the injector at low injection volume, it may lead to extremely poor injection amount accuracy, affecting fuel economy and emissions.

[0003] In the prior art, some technical solutions provide multiple injection control strategies for engines, but do not consider the specific calculation of injection pulse width; and for the technical solutions that propose the calculation method of injection pulse width, they do not consider the optimization of injection pulse width when the number of injections is reduced. Therefore, it is urgent to propose a method for self-learning and updating the minimum allowed injection time to solve the problem of fuel being too lean or too rich when a direct injection engine injects fuel multiple times. Summary of the invention

[0004] The technical problem to be solved by the present invention is to propose a method for self-learning and updating the minimum allowed injection time to solve the problem that when a direct injection engine is injected multiple times, the injection amount accuracy is extremely poor due to inaccurate control of the minimum injection time.

[0005] The technical solution adopted by the present invention to solve the above technical problems is: a method for self-learning and updating the minimum injection duration allowed, judging whether the conditions for activating the self-learning and updating are met, and if so, allowing the self-learning and updating value of the minimum injection duration to enter the self-learning and updating stage; calculating the relationship between the injection duration of the current injection cylinder and the minimum injection duration allowed according to the self-learning and updating value of the minimum injection duration allowed after the last learning and updating;

[0006] When the fuel concentration satisfies the preset condition 1, under the same conditions, according to the relationship between the injection duration of the current injection cylinder and the minimum allowable injection duration, and the number of times the preset conditions 3 and 4 are respectively satisfied, the injection duration increase condition 1, the injection duration decrease condition 1, the injection duration increase condition 2, the injection duration decrease condition 2 and other conditions are obtained;

[0007] When the fuel concentration satisfies the preset condition 2, under the same conditions, according to the relationship between the injection time of the current injection cylinder and the minimum allowable injection time, and the number of times the preset conditions 3 and 4 are respectively satisfied, the injection time increase working condition 3, the injection time decrease working condition 3, the injection time increase working condition 4, the injection time decrease working condition 4 and other working conditions are obtained;

[0008] For the above four injection duration increase conditions, the corresponding preset value is added to the allowable minimum injection duration self-learning update value after the last learning update to obtain an updated allowable minimum injection duration self-learning update value; for the above four injection duration reduction conditions, the corresponding preset value is reduced from the allowable minimum injection duration self-learning update value after the last learning update to obtain an updated allowable minimum injection duration self-learning update value; for the above other conditions, the allowable minimum injection duration self-learning update value after the last learning update is not updated;

[0009] The allowed minimum injection duration is updated according to the updated allowed minimum injection duration self-learning update value.

[0010] Furthermore, the relationship between the current injection cylinder injection time and the minimum allowable injection time is specifically t PxRaw Value and f(p FuelDiff_Px )×(1+r Adapt (z)) values, where t PxRaw is the injection duration of the current injection cylinder, f(p FuelDiff_Px )×(1+r Adapt (z)) is the minimum injection duration allowed, p FuelDiff_Px is the average value of the difference between the fuel rail pressure and the cylinder pressure during the current injection cylinder injection cycle, f(p FuelDiff_Px ) is based on p FuelDiff_Px The initial value of the minimum injection duration allowed by the current injection cylinder is determined, r Adapt (z) is the self-learning update value of the minimum injection duration allowed after the last learning update, and the default initial value is 0.

[0011] Furthermore, the preset condition 1 is r AFR_DiffAvg >k0×f(x)×AFR Req ×(1+r Lrn ), and the preset condition 2 has not been met in the previous learning process, and the preset condition 2 is r AFR_DiffAvg <(-1)×k1×f(x)×AFR Req ×(1+r Lrn ), and the preset condition 1 has not been met in the previous learning process, where r AFR_DiffAvg is the average value of the difference in air-fuel ratio, k0 and k1 are the air-fuel ratio threshold coefficients, f(x) is the correction coefficient obtained based on different cylinder tests, AFR Req is the target air-fuel ratio of the engine, r Lrn is the threshold correction factor.

[0012] Furthermore, the injection time increase condition 1, injection time reduction condition 1, injection time increase condition 2, injection time reduction condition 2 and other conditions are specifically: under the same conditions, when the fuel concentration meets the preset condition 1, if t PxRaw >f(p FuelDiff_Px )×(1+r Adapt (z)), record the number of times CNT1 that the preset condition three is satisfied. If CNT1 is greater than the preset value, this is the injection time increase condition one. Record the number of times CNT2 that the preset condition four is satisfied. If CNT2 is greater than the preset value, this is the injection time decrease condition one. If t PxRaw <f(p FuelDiff_Px )×(1+r Adapt (z)), record the number of times CNT3 that preset condition three is met, if CNT3 is greater than the preset value, this situation is the second condition where the injection duration is increased, record the number of times CNT4 that preset condition four is met, if CNT4 is greater than the preset value, this situation is the second condition where the injection duration is reduced; the conditions that do not meet the above four conditions are other conditions.

[0013] Furthermore, the three conditions of increasing the injection time, decreasing the injection time, increasing the injection time, decreasing the injection time and other conditions are as follows: under the same conditions, when the fuel concentration meets the preset condition two, if t PxRaw >f(p FuelDiff_Px )×(1+r Adapt (z)), record the number of times CNT5 that the preset condition three is met. If CNT5 is greater than the preset value, this is the injection time increase condition three. Record the number of times CNT6 that the preset condition four is met. If CNT6 is greater than the preset value, this is the injection time reduction condition three. If t PxRaw <f(p FuelDiff_Px )×(1+r Adapt (z)), record the number of times CNT7 that preset condition three is met, if CNT7 is greater than the preset value, this situation is the injection duration increase condition four, record the number of times CNT8 that preset condition four is met, if CNT8 is greater than the preset value, this situation is the injection duration decrease condition four; the conditions that do not meet the above four conditions are other conditions.

[0014] Furthermore, the third precondition is |r DsrdAFRAvg -r AFRFilterAvg |≥C1×r AFRFilterAvg , the fourth precondition is |r DsrdAFRAvg -r AFRFilterAvg |≤C2×r AFRFilterAvg , where r DsrdAFRAvg is the average value of the target air-fuel ratio of the engine, r AFRFilterAvg is the actual air-fuel ratio filtered average value, and C1 and C2 are preset values.

[0015] Furthermore, when the fuel concentration meets the preset condition 1, if

[0016]

[0017] Or when the fuel concentration meets the preset condition 2, if

[0018]

[0019] Then the threshold correction coefficient is updated by reducing the corresponding preset value from the threshold correction coefficient after the last self-learning update, r Lrn (z) is the threshold correction coefficient after the last self-learning update, and its default initial value is 0.

[0020] Furthermore, if the number of consecutive occurrences of the other operating conditions exceeds a preset number, the threshold correction coefficient is not updated; otherwise, the threshold correction coefficient after the last self-learning update is updated by adding the corresponding preset value, and the number of consecutive occurrences of the other operating conditions is cleared after the update.

[0021] Furthermore, the same conditions are defined as: the current injection cylinder number, the number of injections, the injection duration of the current injection cylinder, the average fuel rail pressure, the average fuel rail temperature, the actual average density of fresh air intake into the cylinder, the engine target air-fuel ratio, the long-term fuel correction average, the short-term fuel correction average and the engine cooling water temperature average are all the same.

[0022] Furthermore, the four operating conditions where the fuel concentration meets the preset condition one or the four operating conditions where the fuel concentration meets the preset condition two have different priorities. When the conditions of the high priority operating condition are met, the low priority operating condition is no longer judged.

[0023] Furthermore, once the adjustment of the self-learning update value of the minimum injection duration allowed under one of the working conditions is executed, the accumulation of the number of times of all working conditions under the same injection cylinder is reset to zero.

[0024] Furthermore, the method for determining whether the self-learning update value of the minimum injection duration is allowed to meet the conditions for self-learning update includes: determining whether preset condition five is satisfied; if satisfied, the self-learning update value of the minimum injection duration is allowed to enter the self-learning update stage; if not satisfied, entering the self-learning unactivated stage; in the self-learning unactivated stage, when preset condition five is satisfied, entering the self-learning stabilization stage; in the self-learning stabilization stage, when preset condition five and preset condition six are satisfied at the same time, entering the self-learning activation stage; in the self-learning activation stage, after completing the continuous reading of various parameters within the preset time, entering the self-learning update stage.

[0025] Furthermore, preset condition five includes that the engine speed fluctuation is within a preset range, the activation of the oxygen sensor upstream of the catalyst exceeds the preset time, no fuel cut-off occurs in any cylinder, the engine target air-fuel ratio remains unchanged, the fresh air density entering the cylinder fluctuates within a preset range, the number of injections does not change, the injection duration of the current injection cylinder is not less than the preset time and the fluctuation does not exceed the preset time, the intake temperature fluctuation is within the preset range, the fuel rail pressure fluctuation is within the preset range, the fuel rail temperature fluctuation is within the preset range, the engine cooling water temperature fluctuation is within the preset range, the long-term fuel correction coefficient remains unchanged, the short-term fuel correction coefficient fluctuation is within the preset range, the ignition angle efficiency is not lower than the preset value, the engine speed closed loop is not entered, the transmission system does not enter the gear shifting process state and the delay time after the transmission system exits the gear shifting process state exceeds the preset time, and no fault is detected; preset condition six includes entering the self-learning stabilization stage for more than the preset time and the self-learning update value of the minimum injection duration allowed is not updated for more than the preset time.

[0026] Furthermore, preset condition six includes that the self-learning stabilization phase is entered for a period exceeding a preset time and the self-learning update value of the allowed minimum injection duration is not updated for a period exceeding a preset time.

[0027] Furthermore, after entering the self-learning stabilization stage: if preset condition five is met and preset condition six is ​​not met, the self-learning stabilization stage is maintained; if both preset condition five and preset condition six are not met, the self-learning is returned to the self-learning inactive stage.

[0028] Further, the injection duration t of each injection can be obtained according to the updated minimum allowed injection duration Px , specifically: if t PxRaw <f(p FuelDiff_Px )×(1+r Adapt ), then t P x=0; if t PxRaw ≥f(p FuelDiff_Px )×(1+r Adapt ), then t Px =t PxRaw .

[0029] The beneficial effects of the present invention are as follows: by calculating the relationship between the injection duration of the current injection cylinder and the minimum allowable injection duration, when the fuel is too rich or too lean, the self-learning update value of the minimum allowable injection duration is updated according to different working conditions, and the minimum allowable injection duration is further updated, thereby avoiding the problem of the injection of fuel being too lean or too rich due to inaccurate control of the minimum allowable injection duration during multiple injections of the direct injection engine, thereby optimizing the combustion economy of the fuel. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1A flow chart of a method for self-learning and updating the minimum allowed injection duration of the present invention;

[0031] Figure 2 The present invention is a flow chart of a method for determining whether conditions for self-learning update are met. DETAILED DESCRIPTION

[0032] The present invention will be further described below in conjunction with specific embodiments and accompanying drawings.

[0033] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present application should have the common meanings understood by persons having ordinary skills in the field to which the present disclosure belongs.

[0034] In order to facilitate the understanding of this application document, the following keywords need to be explained as follows:

[0035] r AFR_DiffAvg : average value of air-fuel ratio difference; k0, k1: air-fuel ratio threshold coefficients; f(x): correction coefficient based on different cylinder tests; AFR Req : Engine target air-fuel ratio;

[0036] r Lrn : Threshold correction coefficient; r Lrn (z): Threshold correction coefficient after the last learning update;

[0037] t Px_Diff :t PxRaw and f(p FuelDiff_Px )×(1+r Adapt (z)) difference; t PxRaw : The current injection cylinder injection duration; p FuelDiff_Px : The average value of the difference between the fuel rail pressure and the cylinder pressure during the current injection cylinder injection cycle; f(p FuelDiff_Px ): the initial value of the minimum injection duration allowed by the current injection cylinder;

[0038] r DsrdAFRAvg : Average value of target air-fuel ratio of the engine; r AFRFilterAvg : Average value of actual air-fuel ratio filter value;

[0039] r Adapt : Allow the minimum injection time to be updated by self-learning; r Adapt (z): the self-learning update value of the minimum injection duration allowed after the last learning update;

[0040] r AFRFilter : Actual air-fuel ratio filtering value; r AFRFilter (N): Actual air-fuel ratio filtered value of the Nth sampling period; r AFRRaw(N): the actual air-fuel ratio original value of the Nth sampling period, N = 1.2.3...; K AFR : coefficient, and k AFR : air-fuel ratio filter coefficient; m: number of engine cylinders; n: engine speed.

[0041] This embodiment provides a method for self-learning and updating the minimum injection duration value, such as Figure 1 As shown, the following steps are included:

[0042] Determine whether the conditions for activating self-learning update are met. If so, allow the minimum injection duration self-learning update value to enter the self-learning update stage. Specifically, the judgment process is as follows: Figure 2 shown.

[0043] When the following activation conditions are met at the same time, the minimum injection duration self-learning update value is allowed to be updated by self-learning:

[0044] The engine speed fluctuation range is within ±20rpm; the oxygen sensor upstream of the catalyst has been activated for a period of time, which is 0.5s in this embodiment; there is no fuel cut-off in any cylinder; the engine target air-fuel ratio is fixed; the fresh air intake density entering the cylinder fluctuates within ±15mg / l; the number of injections has not changed (the definition of the number of injections is given in patent CN202410313240.0 "A direct injection engine multi-injection control method and system"); the injection duration of the current injection cylinder fluctuates by no more than ±0.2ms; the injection duration of the current injection cylinder is not less than 0.5ms; the intake temperature fluctuates within ±2℃; the oil rail pressure fluctuates by no more than ±2kPa; the oil rail temperature fluctuates by no more than ±1℃; the engine cooling water temperature fluctuates by no more than ±1 ℃; the long-term fuel correction coefficient has not changed (the definition of the long-term fuel correction coefficient is given in CN202010246641.0 "A self-learning method for long-term fuel correction of a gasoline engine"); the fluctuation range of the short-term fuel correction coefficient does not exceed ±0.01 (the definition of the short-term fuel correction coefficient is given in CN202111202932.0 "A method and control system for controlling short-term fuel correction of an engine"); the ignition angle efficiency is not lower than the preset value, and the present embodiment takes 0.9 to avoid mislearning due to poor combustion caused by too low ignition angle efficiency; the engine speed closed loop is not entered; the transmission system has not entered the gear shifting process state and the delay time after the transmission system exits the gear shifting process state exceeds the preset time, and the present embodiment takes 0.05s; no faults are detected.

[0045] During the self-learning process, if any of the activation conditions is not met at any stage, the self-learning is terminated and the self-learning inactive stage is entered. In the self-learning inactive stage, if the above activation conditions are met, the self-learning stabilization stage of allowing the self-learning update value of the minimum injection duration is entered. The purpose of the stabilization stage is to ensure that the self-learning activation conditions are stable and reliable.

[0046] During the self-learning stabilization phase, when the reactivation conditions are met, the self-learning activation phase of allowing the self-learning update value of the minimum injection duration is entered. The reactivation conditions include: entering the self-learning stabilization phase for more than a preset time t0, which is 5s in this embodiment; allowing the self-learning update value of the minimum injection duration to be not updated for more than a preset time t1, which is 2880h in this embodiment.

[0047] After entering the self-learning stabilization stage: if the reactivation condition is not met and the activation condition is met, it remains in the self-learning stabilization stage; if both the reactivation condition and the activation condition are not met, it returns to the self-learning inactive stage; when both the reactivation condition and the activation condition are met, it enters the self-learning active stage.

[0048] When entering the self-learning activation stage that allows the self-learning update value of the minimum injection duration, the average value of the fuel rail pressure, the average value of the fuel rail temperature, the average value of the engine cooling water temperature, the engine target air-fuel ratio, the actual average value of the fresh air intake density entering the cylinder, the actual air-fuel ratio filter value, the actual air-fuel ratio filter value average value, the number of injections, the long-term fuel correction factor, the short-term fuel correction factor average value, and the air-fuel ratio difference average value in the continuous time t2 after entering the self-learning activation stage are cumulatively read. In this embodiment, t2 is 3s.

[0049] Among them, the actual air-fuel ratio filtering value r AFRFilter The calculation formula is:

[0050] r AFRFilter (N) = K AFR ×[r AFRRaw (N)-r AFRFilter (N-1)]+r AFRFilter (N-1).

[0051] In this embodiment, Δt is 10ms, the number of engine cylinders m is 4, and k AFR The calibration speed is 1000rpm, k AFR Take 0.023.

[0052] set up The purpose is to calibrate the engine with 4 cylinders and a speed of 1000 rpm without any special calibration. AFR , thereby reducing the calibration test work and achieving normalization processing.

[0053] After time t2 is met, the self-learning update phase is entered to allow the self-learning update value of the minimum injection duration.

[0054] According to the self-learning update value of the minimum allowable injection time after the last learning update, the relationship between the current injection cylinder injection time and the minimum allowable injection time is calculated.

[0055] Specifically, calculate t PxRaw and f(p FuelDiff_Px )×(1+r Adapt (z)) Px_Diff .

[0056] When the fuel concentration meets preset condition one, under the same conditions, according to the relationship between the injection time of the current injection cylinder and the minimum allowable injection time, and the number of times preset conditions three and four are met respectively, injection time increase condition one, injection time decrease condition one, injection time increase condition two, injection time decrease condition two and other conditions are obtained.

[0057] For the condition of increasing injection duration, the corresponding preset value is added to the self-learning update value of the minimum allowable injection duration after the last learning update to obtain an updated self-learning update value of the minimum allowable injection duration; for the condition of decreasing injection duration, the corresponding preset value is subtracted from the self-learning update value of the minimum allowable injection duration after the last learning update to obtain an updated self-learning update value of the minimum allowable injection duration; for the other conditions mentioned above, the self-learning update value of the minimum allowable injection duration after the last learning update is not updated.

[0058] Specifically, if r AFR_DiffAvg >k0×f(x)×AFR Req ×(1+r Lrn ), if the fuel was lean in the previous self-learning process, the self-learning update value of the minimum injection duration allowed will not be updated this time; if the fuel was not lean in the previous process, it means that the actual air-fuel ratio control is too small, and it is judged that the fuel is rich, then the self-learning update value of the minimum injection duration allowed will be updated.

[0059] Record the same conditions to meet t Px_Diff is greater than 0 and satisfies |r DsrdAFRAvg -r AFRFilterAvg |≥C1×r AFRFilterAvg The number of times CNT1 is set, C1 in this embodiment is 0.02. Once CNT1 is greater than the preset value, it is set to 5 in this embodiment, indicating that the current injection time is short, which is easy to cause air-fuel ratio fluctuations, and the minimum injection time needs to be increased. This situation is called injection time increase condition 1, and r is updated Adapt =r Adapt (z)+0.02.

[0060] Record the same conditions to meet t Px_Diff is greater than 0 and satisfies |r DsrdAFRAvg -r AFRFilterAvg |≤C2×r AFRFilterAvg The number of times CNT2 is calculated, C2 in this embodiment is 0.01. Once CNT2 is greater than the preset value, in this embodiment it is 5, indicating that the current injection time is long and it is not easy to cause air-fuel ratio fluctuations, and the minimum injection time needs to be reduced. This situation is called the injection time reduction condition 1, and r is updated. Adapt =r Adapt (z)-0.01.

[0061] Record the same conditions to meet t Px_Diff Less than 0, and satisfies |r DsrdAFRAvg -r AFRFilterAvg |≥C1×r AFRFilterAvg Once CNT3 is greater than the preset value, this embodiment takes 5, indicating that the current injection duration is very short, which is easy to cause air-fuel ratio fluctuations, and the minimum injection duration needs to be increased to a large extent. This situation is called the injection duration increase condition 2, and r is updated Adapt =r Adapt (z)+0.05.

[0062] Record the same conditions to meet t Px_Diff Less than 0, and satisfies |r DsrdAFRAvg -r AFRFilterAvg |≤C2×r AFRFilterAvg Once CNT4 is greater than the preset value, the value is 5 in this embodiment, indicating that the current injection duration is very short and is not likely to cause air-fuel ratio fluctuations. The minimum injection duration needs to be reduced to a large extent. This situation is called the injection duration reduction condition 2, and r is updated. Adapt =r Adapt (z)-0.03.

[0063] Under other working conditions Adapt Not updated.

[0064] If the number of times other working conditions occur continuously exceeds the preset number, which is 20 in this embodiment, then r Lrn Nor update; otherwise r Lrn =r Lrn (z)+0.006, at r L After rn is updated, the consecutive times of other working conditions are reset to zero, and then continue to accumulate from 0.

[0065] In any of the above 4 situations, if Then update r Lrn =r Lrn (z)-0.015, that is, due to r AdaptAfter the update is performed, the minimum injection duration also needs to be updated. At this time, the operating condition range that is allowed to enter this situation is increased, and it is further determined whether the minimum injection duration needs to be updated under other operating conditions.

[0066] The priorities of the injection time increasing condition 2, the injection time increasing condition 1, the injection time decreasing condition 1, and the injection time decreasing condition 2 are getting lower and lower. If the high priority condition is met, the low priority condition will no longer be judged, and the low priority learning coefficient update will not be executed.

[0067] When the fuel concentration meets preset condition two, under the same conditions, according to the relationship between the injection time of the current injection cylinder and the minimum allowable injection time, and the number of times preset condition three and preset condition four are met respectively, injection time increase condition three, injection time decrease condition three, injection time increase condition four, injection time decrease condition four and other conditions are obtained.

[0068] For the condition of increasing injection duration, the corresponding preset value is added to the self-learning update value of the minimum allowable injection duration after the last learning update to obtain an updated self-learning update value of the minimum allowable injection duration; for the condition of decreasing injection duration, the corresponding preset value is subtracted from the self-learning update value of the minimum allowable injection duration after the last learning update to obtain an updated self-learning update value of the minimum allowable injection duration; for the other conditions mentioned above, the self-learning update value of the minimum allowable injection duration after the last learning update is not updated.

[0069] Specifically, if r AFR_DiffAvg <(-1)×k1×f(x)×AFR Req ×(1+r Lrn ), if the fuel was too rich in the previous learning process, the self-learning update value of the minimum injection duration allowed will not be updated this time; if the fuel was not too rich in the previous learning process, it means that the actual air-fuel ratio control is too large, and it is judged that the fuel is too lean, then the self-learning update value of the minimum injection duration allowed will be updated.

[0070] Record the same conditions to meet t Px_Diff is greater than 0 and satisfies |r DsrdAFRAvg -r AFRFilterAvg |≥C1×r AFRFilterAvg The number of times CNT5 is calculated. Once CNT5 is greater than the preset value, the value is 5 in this embodiment, indicating that the current injection duration is short, which is likely to cause air-fuel ratio fluctuations, and the minimum injection duration needs to be increased. This situation is called the injection duration increase condition 3. The injection duration that needs to be increased in the injection duration increase condition 3 is not less than the injection duration that needs to be increased in the injection duration increase condition 1. The main reason is that the fuel injection is lean at this time, and the minimum injection duration needs to be further increased to avoid further leaning, and r is updated. Adapt =r Adapt(z)+0.04.

[0071] Record the same conditions to meet t Px_Diff is greater than 0 and satisfies |r DsrdAFRAvg -r AFRFilterAvg |≤C2×r AFRFilterAvg The number of times CNT6 is calculated. Once CNT6 is greater than the preset value, the value of 5 is selected in this embodiment, indicating that the current injection duration is relatively long and is not likely to cause air-fuel ratio fluctuations. The minimum injection duration needs to be reduced. This situation is called the injection duration reduction condition 3. The injection duration that needs to be reduced in the injection duration reduction condition 3 should not be greater than the injection duration that needs to be reduced in the injection duration reduction condition 1. The main reason is that the fuel injection is lean at this time. The injection duration cannot be reduced too much to avoid further leaning, and r is updated. Adapt =r Adapt (z)-0.006.

[0072] Record the same conditions to meet t Px_Diff Less than 0, and satisfies |r DsrdAFRAvg -r AFRFilterAvg |≥C1×r AFRFilterAvg The number of times CNT7 is reached. Once CNT7 is greater than the preset value, the value of 5 is set in this embodiment, indicating that the current injection duration is extremely short, which is likely to cause air-fuel ratio fluctuations, and the minimum injection duration needs to be increased to a large extent. This situation is called injection duration increase condition 4. The injection duration that needs to be increased in injection duration increase condition 4 must not be less than the injection duration that needs to be increased in injection duration increase condition 2. The main reason is that the fuel injection is lean at this time, and the minimum injection duration needs to be further increased to avoid further leaning, and r is updated. Adapt =r Adapt (z)+0.08.

[0073] Record the same conditions to meet t Px_Diff Less than 0, and satisfies |r DsrdAFRAvg -r AFRFilterAvg |≤C2×r AFRFilterAvg The number of times CNT8 is reached. Once CNT8 is greater than the preset value, this embodiment takes 5, indicating that the current injection duration is extremely short, but it is not easy to cause air-fuel ratio fluctuations, and the minimum injection duration needs to be reduced to a large extent. This situation is called injection duration reduction condition 4. The injection duration that needs to be reduced in injection duration reduction condition 4 should not be greater than the injection duration that needs to be reduced in injection duration reduction condition 2. The main reason is that the fuel injection is lean at this time, and the injection duration cannot be reduced too much to avoid further leaning, and update r Adapt =r Adapt (z)-0.01.

[0074] Under other working conditions, the self-learning update value of the minimum injection duration is not updated.

[0075] If the number of times other working conditions occur continuously exceeds the preset number, which is 20 in this embodiment, then r Lrn Nor update; otherwise r Lrn =r Lrn (z)+0.006, at r Lrn After the update, the consecutive times of other working conditions are reset to zero, and then continue to accumulate from 0.

[0076] In any of the above 4 situations,

[0077] like Then update r Lrn =r Lrn (z)-0.015, that is, due to r Adapt After the update is performed, the minimum injection duration also needs to be updated. At this time, the operating condition range that is allowed to enter this situation is increased, and it is further determined whether the minimum injection duration needs to be updated under other operating conditions.

[0078] The priorities of the injection time increasing condition four, the injection time increasing condition three, the injection time decreasing condition three, and the injection time decreasing condition four are getting lower and lower. If the high priority working condition is met, the low priority working condition will no longer be judged, and the low priority learning coefficient update will not be executed.

[0079] Once the self-learning update value adjustment of the minimum allowable injection duration under one of the above-mentioned working conditions is performed, the cumulative number of times of all working conditions under the same injection cylinder will be cleared.

[0080] In this embodiment, parameters such as the current injection cylinder number, the number of injections, the injection duration of the current injection cylinder, the average value of the fuel rail pressure, the average value of the fuel rail temperature, the average value of the actual fresh air intake density entering the cylinder, the engine target air-fuel ratio, the long-term fuel correction average value, the short-term fuel correction average value and the engine cooling water temperature average value are all the same and are defined as the same conditions.

[0081] The self-learning update values ​​of the minimum injection duration allowed under different working conditions are stored in the non-volatile memory EEPROM. There is an initial default value of 0 in the EEPROM, and the stored value in the EEPROM is updated after the self-learning update value of the minimum injection duration allowed is completed.

[0082] The allowed minimum injection duration is updated according to the updated allowed minimum injection duration self-learning update value.

[0083] The present invention obtains the relationship between the injection duration of the current injection cylinder and the minimum allowable injection duration determined by the self-learning update value of the minimum allowable injection duration after the last self-learning update, and divides the working conditions into four injection duration increase working conditions, four injection duration reduction working conditions and other working conditions according to the relationship and the judgment of whether the fuel is too rich or too lean, and further updates the self-learning update value of the minimum allowable injection duration, and combines the obtained updated self-learning update value of the minimum allowable injection duration with the minimum allowable injection duration determined based on the difference between the oil rail pressure and the cylinder pressure to calculate the updated minimum allowable injection duration. The present invention optimizes the control of the minimum allowable injection duration, solves the problem of poor injection injection amount accuracy due to nonlinear characteristics of the injector at low injection amount when the direct injection engine is injected multiple times, and optimizes the combustion economy of the fuel.

[0084] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all these improvements and changes should fall within the scope of protection of the appended claims of the present invention.

Claims

1. A method for self-learning and updating the minimum allowed injection duration, characterized in that: include: Determine whether the conditions for activating self-learning update are met. If so, allow the minimum injection duration self-learning update value to enter the self-learning update stage; Calculate the relationship between the injection time of the current injection cylinder and the minimum injection time allowed according to the self-learning update value of the minimum injection time allowed after the last learning update; When the fuel concentration satisfies the preset condition 1, under the same conditions, according to the relationship between the injection duration of the current injection cylinder and the minimum allowable injection duration, and the number of times the preset conditions 3 and 4 are respectively satisfied, the injection duration increase condition 1, the injection duration decrease condition 1, the injection duration increase condition 2, the injection duration decrease condition 2 and other conditions are obtained; When the fuel concentration satisfies the preset condition 2, under the same conditions, according to the relationship between the injection time of the current injection cylinder and the minimum allowable injection time, and the number of times the preset conditions 3 and 4 are respectively satisfied, the injection time increase working condition 3, the injection time decrease working condition 3, the injection time increase working condition 4, the injection time decrease working condition 4 and other working conditions are obtained; For the above four injection duration increase conditions, the corresponding preset value is added to the allowable minimum injection duration self-learning update value after the last learning update to obtain an updated allowable minimum injection duration self-learning update value; for the above four injection duration reduction conditions, the corresponding preset value is reduced from the allowable minimum injection duration self-learning update value after the last learning update to obtain an updated allowable minimum injection duration self-learning update value; for the above other conditions, the allowable minimum injection duration self-learning update value after the last learning update is not updated; The allowed minimum injection duration is updated according to the updated allowed minimum injection duration self-learning update value.

2. A method for self-learning and updating the minimum allowed injection duration according to claim 1, characterized in that: The relationship between the current injection cylinder injection time and the minimum allowable injection time is specifically t PxRaw Value and f(p FuelDiff_Px )×(1+r Adapt (z)) values, where t PxRaw is the injection duration of the current injection cylinder, f(p FuelDiff_Px )×(1+r Adapt (z)) is the minimum injection duration allowed, p FuelDiff_Px is the average value of the difference between the fuel rail pressure and the cylinder pressure during the current injection cylinder injection cycle, f(p FuelDiff_Px ) is based on p FuelDiff_Px The initial value of the minimum injection duration allowed by the current injection cylinder is determined, r Adapt (z) is the self-learning update value of the minimum injection duration allowed after the last learning update, and the default initial value is 0.

3. A method for self-learning and updating the minimum allowed injection duration according to claim 1, characterized in that: The preset condition 1 is r AFR_DiffAvg >k0×f(x)×AFR Req ×(1+r Lrn ), and the preset condition 2 has not been met in the previous learning process, and the preset condition 2 is r AFR_DiffAvg <(-1)×k1×f(x)×AFR Req ×(1+r Lrn ), and the preset condition 1 has not been met in the previous learning process, where r AFR_DiffAvg is the average value of the difference in air-fuel ratio, k0 and k1 are the air-fuel ratio threshold coefficients, f(x) is the correction coefficient obtained based on different cylinder tests, AFR Req is the target air-fuel ratio of the engine, r Lrn is the threshold correction factor.

4. A method for self-learning and updating the minimum allowable injection duration according to claim 2, characterized in that: The aforementioned injection time increase condition 1, injection time decrease condition 1, injection time increase condition 2, injection time decrease condition 2 and other conditions are specifically when the fuel concentration meets the preset condition 1, under the same conditions: If PxRaw >f(p FuelDiff_Px )×(1+r Adapt (z)), record the number of times CNT1 that the preset condition three is satisfied. If CNT1 is greater than the preset value, this is the first condition that the injection duration is increased. Record the number of times CNT2 that the preset condition four is satisfied. If CNT2 is greater than the preset value, this is the first condition that the injection duration is reduced. If PxRaw <f(p FuelDiff_Px )×(1+r Adapt (z)), record the number of times CNT3 that the preset condition three is satisfied. If CNT3 is greater than the preset value, this is the second condition where the injection duration is increased. Record the number of times CNT4 that the preset condition four is satisfied. If CNT4 is greater than the preset value, this is the second condition where the injection duration is reduced. Other working conditions are those that do not meet the above four working conditions.

5. A method for self-learning and updating the minimum allowed injection duration according to claim 2, characterized in that: The aforementioned injection time increase condition 3, injection time decrease condition 3, injection time increase condition 4, injection time decrease condition 4 and other conditions are specifically when the fuel concentration meets the preset condition 2, under the same conditions: If PxRaw >f(p FuelDiff_Px )×(1+r Adapt (z)), record the number of times CNT5 that the preset condition three is satisfied. If CNT5 is greater than the preset value, this is the injection duration increase condition three. Record the number of times CNT6 that the preset condition four is satisfied. If CNT6 is greater than the preset value, this is the injection duration decrease condition three. If PxRaw <f(p FuelDiff_Px )×(1+r Adapt (z)), record the number of times CNT7 that the preset condition three is satisfied. If CNT7 is greater than the preset value, this is the fourth condition where the injection duration is increased. Record the number of times CNT8 that the preset condition four is satisfied. If CNT8 is greater than the preset value, this is the fourth condition where the injection duration is reduced. Other working conditions are those that do not meet the above four working conditions.

6. A method for self-learning and updating the minimum allowed injection duration according to claim 1, characterized in that: Precondition three is |r DsrdAFRAvg -r AFRFilterAvg |≥C1×r AFRFilterAvg , the fourth precondition is |r DsrdAFRAvg -r AFRFilterAvg |≤C2×r AFRFilterAvg , where r DsrdAFRAvg is the average value of the target air-fuel ratio of the engine, r AFRFilterAvg is the actual air-fuel ratio filtered average value, and C1 and C2 are preset values.

7. A method for self-learning and updating the minimum allowed injection duration according to claim 1, characterized in that: When the fuel concentration meets the preset condition 1, if Or when the fuel concentration meets the preset condition 2, if Then the threshold correction coefficient is updated by reducing the corresponding preset value from the threshold correction coefficient after the last self-learning update, r Lrn (z) is the threshold correction coefficient after the last self-learning update, and its default initial value is 0.

8. A method for self-learning and updating the minimum allowed injection duration according to claim 1, characterized in that: If the number of consecutive occurrences of the other operating conditions exceeds a preset number, the threshold correction coefficient is not updated. Otherwise, the threshold correction coefficient after the last self-learning update is updated by adding the corresponding preset value, and the number of consecutive occurrences of the other operating conditions is cleared after the update.

9. A method for self-learning and updating the minimum allowed injection duration according to claim 1, characterized in that: The same conditions are defined as: the current injection cylinder number, injection times, injection duration of the current injection cylinder, average fuel rail pressure, average fuel rail temperature, average density of fresh air actually entering the cylinder, target engine air-fuel ratio, long-term fuel correction average, short-term fuel correction average and average engine cooling water temperature are all the same.

10. The method for self-learning and updating the minimum allowable injection duration according to claim 1, characterized in that: The four operating conditions where the fuel concentration meets the preset condition one or the four operating conditions where the fuel concentration meets the preset condition two have different priorities. When the conditions of the high priority operating condition are met, the low priority operating condition is no longer judged.

11. A method for self-learning and updating the minimum allowed injection duration according to claim 1, characterized in that: Once the adjustment of the self-learning update value of the minimum injection duration allowed under one of the working conditions is performed, the accumulation of the number of times of all working conditions under the same injection cylinder is cleared.

12. A method for self-learning and updating the minimum allowed injection duration according to claim 1, characterized in that: The method for judging whether the self-learning update value of the minimum injection duration is allowed to meet the conditions for self-learning update includes: judging whether preset condition five is satisfied; if satisfied, the self-learning update value of the minimum injection duration is allowed to enter the self-learning update stage; if not satisfied, the self-learning unactivated stage is entered; when in the self-learning unactivated stage, when preset condition five is satisfied, the self-learning stabilization stage is entered; when in the self-learning stabilization stage, when preset condition five and preset condition six are satisfied at the same time, the self-learning activation stage is entered; when in the self-learning activation stage, after completing the continuous reading of various parameters within the preset time, the self-learning update stage is entered.

13. A method for self-learning and updating the minimum allowed injection duration according to claim 12, characterized in that: Preset condition five includes that the engine speed fluctuation is within the preset range, the activation of the oxygen sensor upstream of the catalyst exceeds the preset time, there is no fuel cut-off in any cylinder, the engine target air-fuel ratio remains unchanged, the density of fresh air entering the cylinder fluctuates within the preset range, the number of injections does not change, the injection time of the current injection cylinder is not less than the preset time and the fluctuation does not exceed the preset time, the fluctuation of the intake temperature is within the preset range, the fluctuation of the fuel rail pressure is within the preset range, the fluctuation of the fuel rail temperature is within the preset range, the fluctuation of the engine cooling water temperature is within the preset range, the long-term fuel correction factor remains unchanged, the fluctuation of the short-term fuel correction factor is within the preset range, the ignition angle efficiency is not lower than the preset value, the engine speed closed loop is not entered, the transmission system has not entered the gear shifting process state and the delay time after the transmission system exits the gear shifting process state exceeds the preset time, and no fault is detected.

14. A method for self-learning and updating the minimum allowed injection duration according to claim 12, characterized in that: Preset condition six includes entering the self-learning stabilization phase for more than a preset time and the self-learning update value of the allowed minimum injection duration is not updated for more than a preset time.

15. The method for self-learning and updating the minimum allowed injection duration according to claim 12, characterized in that: After entering the self-learning stabilization stage: if preset condition five is met and preset condition six is ​​not met, it remains in the self-learning stabilization stage; if both preset condition five and preset condition six are not met, it returns to the self-learning inactive stage.

Citation Information

Patent Citations

  • Natural gas engine spraying control method and device

    CN110273773A

  • Self-learning method for gasoline engine long-term fuel correction

    CN111412074A

  • Internal combustion engine controlling method for use in vehicle, involves determining minimum controlling duration based on response of rail pressure signal during stopping of internal combustion engine and when fuel injection is detected

    DE102008054409A1

  • Fuel injection control device for internal combustion engine

    JP2011179389A

  • Air-fuel ratio feedback control apparatus for engines

    US20060150962A1