An engine mixture information processing method, device, medium and controller

Abstract

The present application belongs to the technical field of engine control, and particularly relates to an engine mixture information processing method, device, medium and controller; a first data temporary storage step / unit thereof records working condition information corresponding to a single self-learning value and uses the same as a dynamic window to screen working conditions; a second learning freezing step / unit thereof filters information according to the deviation of the dynamic window and the working conditions, ensures that the self-learning process is implemented under acceptable working conditions, and freezes data beyond the range of the dynamic window until the working condition returns to the range defined by the dynamic window; a third data updating step / unit thereof uses the self-learning value obtained after the dynamic window screening to solve a correction value; and a fourth fault evaluation step / unit thereof diagnoses the fault of the fuel supply system through the relationship between the correction value and a monitoring threshold value; the present scheme can be used in structures such as an electronic control unit (ECU) and an engine management system (EMS) to improve the efficiency of emission control and fault detection.

Description

Technical Field

[0001] This invention belongs to the field of engine control technology, and particularly relates to a method, device, medium and controller for processing engine mixture information. Background Technology

[0002] Mixture control is one of the core technologies of engine combustion control; for spark-ignition engines, it is a decisive factor in determining emission levels. Factors such as the dispersion of related components (intake flow sensor, injector, fuel rail, rail pressure sensor, etc.), vehicle dispersion, and vehicle durability can all lead to mixture pre-control deviations, requiring electronic control units (ECUs) or engine management systems (EMSs) to implement similar measures. Figure 1 The closed-loop control shown constrains the mixture parameters within a predetermined range.

[0003] However, since the aforementioned closed-loop control of the air-fuel mixture is not yet activated after a cold start of the engine or vehicle, the impact on emission parameters is significantly increased during this stage. Related data shows that a 3% deviation in the air-fuel mixture can lead to a risk of emission grade downgrading. Ideally, the self-learning process should only occur after the air-fuel mixture closed-loop control is activated and under certain conditions. The resulting self-learning value can then correct for air-fuel mixture pre-control deviations when the air-fuel mixture closed-loop control is not activated after the next start. Specifically, the self-learning process occurs only after the air-fuel mixture closed-loop control is activated and preset conditions are met. The resulting self-learning value can then correct for air-fuel mixture pre-control deviations when the air-fuel mixture closed-loop control is not activated after the next start or under other conditions.

[0004] The applicant's research found that: in adopting such Figure 2 The single-point gas mixture self-learning shown or as Figure 3 When the zoned gas mixture self-learning is shown, on the one hand, there is the phenomenon of mislearning. The introduction of inaccurate or undesirable data will lead to abnormal pre-control deviation correction and emission deterioration. On the other hand, the zoned gas mixture self-learning cannot represent the overall systemic deviation, and is only applicable to the correction of the pre-control deviation of the gas mixture in a specific area. For operating conditions outside the specific area, its correction process fails.

[0005] Furthermore, sufficient self-learning requires a significant amount of machine time, which may not meet the real-time and processing efficiency requirements of practical application scenarios. Consequently, it imposes excessive constraints on related processing procedures, such as carbon canister desorption. Moreover, the multi-region self-learning process in related technologies may also lead to a decrease in the completeness and real-time performance of fault diagnosis. Therefore, it is urgent to improve the aforementioned intervention process for pre-control deviations to ensure that the engine's original emissions are controlled and operate efficiently throughout the entire process. Summary of the Invention

[0006] This invention discloses a method for processing engine air-fuel mixture information, used for a self-learning process of the air-fuel mixture information. The core process includes a first data storage step and a second learning freeze step. The first data storage step saves the operating condition information corresponding to a single self-learning value and uses this operating condition information as a dynamic window for filtering operating conditions. The operating conditions include the current operating condition and historical operating conditions. The second learning freeze step filters the self-learning process based on the deviation between the dynamic window and the operating conditions. If the current operating condition exceeds the range defined by the dynamic window, the current self-learning process is frozen until the current operating condition returns to within the range defined by the dynamic window. The self-learning process is performed when the engine's air-fuel mixture is under closed-loop control or when the engine's emission control reaches a preset standard.

[0007] Furthermore, the self-learning process can be used to correct the first deviation information in the mixture information; the mixture information is processed by self-learning through single-point self-learning and / or regional self-learning processes, thereby intervening in and / or correcting the mixture information; its current operating condition is the operating condition when the self-learning process is being executed; the historical operating condition is at least one operating condition or operating condition sequence when the self-learning process was performed before entering the current operating condition.

[0008] The operating condition information includes a first speed and / or a second load; the second load can be characterized by the required torque and / or preset parameters or parameter sets of the engine; the first deviation information includes a first systematic pre-control deviation and / or a second non-systematic pre-control deviation; and the self-learning process is limited to the multiplicative self-learning process, i.e., FRA, to correct proportional offset systematic pre-control deviations.

[0009] Specifically, its dynamic window can use operating condition information as a basis and set the range of the dynamic window according to a preset first offset set; the first offset set is used to limit the value range of preset parameters in the operating condition information.

[0010] The operating condition information can be obtained by using the first offset of the first speed and the second load as parameters to obtain a rectangular window area, and the rectangular window area can be used as the basis for filtering the current operating condition and / or historical operating condition based on the value of the dynamic window at different times.

[0011] Furthermore, the self-learning process can use a low-pass filter to approach the current mixture pre-control deviation, i.e., xfrai, by a neutral value of 1.0; at the same time, a first positive process variable, i.e. flpfrai, and a second negative process variable, i.e. flnfrai, are obtained; and the variable with the larger absolute value between the first positive process variable and the second negative process variable is used as a third process variable to characterize the progress depth or learning degree of the self-learning process.

[0012] The first positive process variable and the second negative process variable approach the target value from the direction with the larger absolute value through a low-pass filtering process. The target value is the difference between the current mixture pre-control deviation and the current self-learning value. As the self-learning process progresses, if the current self-learning value reaches the current mixture pre-control deviation level, the target value approaches 0. If the third process variable is less than the preset fourth process threshold and remains at the preset fifth time, the self-learning process ends.

[0013] Furthermore, the mixed gas information processing method may also include a third data update step, which compares the historical dynamic window or set of historical dynamic windows corresponding to historical operating conditions with the current dynamic window corresponding to the current self-learning value. If the elements of the historical dynamic window or set of historical dynamic windows overlap with the current dynamic window or their intersection is greater than a preset ratio, then the current self-learning value replaces the elements in the historical self-learning value or the historical self-learning value sequence; otherwise, a self-learning value sequence is constructed with at least one historical self-learning value and the current self-learning value until the number of elements in the self-learning value sequence reaches or exceeds a positive integer N; if the number of elements in the self-learning value sequence when the current self-learning value is added is equal to a positive integer N, and N is greater than or equal to 2, then the current self-learning value replaces the earliest historical self-learning value added to the historical self-learning value sequence; for ease of description, the historical self-learning value is taken as the first element of a self-learning value sequence, so the minimum number of elements N in the self-learning sequence is 3; when the number of self-learning values ​​reaches the upper limit N, if the current dynamic window does not overlap with other self-learning values, the earliest historical self-learning value is replaced first.

[0014] If the number of elements in the historical self-learning sequence is greater than or equal to 3, the maximum and minimum values ​​in the historical self-learning sequence are removed, and the weighted average of the remaining elements in the historical self-learning sequence is used as the correction value for the self-learning process; if the number of elements in the historical self-learning sequence is less than or equal to 2, the correction value remains unchanged.

[0015] Furthermore, the mixture information processing method may also include a fourth fault assessment step, which obtains emission test data of a preset set of engines or vehicles. If the correction value is greater than a preset monitoring threshold, the fuel supply system is considered to be faulty.

[0016] The dynamic window can be a fixed window range or correspond to a preset range of operating conditions; its operating conditions are constrained or given by preset engine parameters or air-fuel mixture parameters.

[0017] Accordingly, this invention also discloses a fault diagnosis device, including a first data temporary storage unit and a second learning freeze unit; wherein: the first data temporary storage unit stores the operating condition information corresponding to a single self-learning value and uses the operating condition information as a dynamic window for operating condition filtering, the operating conditions including the current operating condition and historical operating conditions; the second learning freeze unit filters the self-learning process according to the deviation between the dynamic window and the operating condition; if the current operating condition exceeds the range limited by the dynamic window, the current self-learning process is frozen until the current operating condition returns to the range limited by the dynamic window; the self-learning process is performed when the engine's air-fuel mixture is under closed-loop control or when the engine's emission control reaches a preset standard; the self-learning process can be used to correct the first deviation information in the air-fuel mixture information; the self-learning processing of the air-fuel mixture information is performed using a single-point self-learning and / or regional self-learning process, thereby intervening in and / or correcting the air-fuel mixture information; the current operating condition is the operating condition when the self-learning process is being executed; the historical operating conditions are at least one operating condition or operating condition sequence when the self-learning process was performed before entering the current operating condition.

[0018] Specifically, its operating condition information can be characterized by a first speed and / or a second load; the second load can be characterized by the required torque and / or preset parameters or parameter sets of the engine; its first deviation information includes a first systematic pre-control deviation and / or a second non-systematic pre-control deviation; its self-learning process is limited to a multiplicative self-learning process, i.e., FRA, used to correct proportional offset systematic pre-control deviations; its dynamic window is based on the operating condition information, and the range of the dynamic window is set according to a preset first offset set; the first offset set is used to limit the value range of preset parameters in the operating condition information.

[0019] Furthermore, the operating condition information can be obtained by using the first offset set of the first speed and the second load as parameters to obtain a rectangular window area set, and the rectangular window area set can be used as the value of the dynamic window at different times to filter the current operating condition and / or historical operating condition.

[0020] Correspondingly, its self-learning process can also use low-pass filtering to approach the current mixture pre-control deviation, i.e., xfrai, from the neutral value of 1.0; at the same time, the first positive process variable, i.e. flpfrai, and the second negative process variable, i.e. flnfrai, are obtained; and the one with the larger absolute value among the first positive process variable and the second negative process variable is used as the third process variable to characterize the progress depth or learning degree of the self-learning process.

[0021] The first positive process variable and the second negative process variable approach the target value from the direction with the larger absolute value through a low-pass filtering process. The target value is the difference between the current mixture pre-control deviation and the current self-learning value. As the self-learning process progresses, if the current self-learning value reaches the current mixture pre-control deviation level, the target value approaches 0. If the third process variable is less than the preset fourth process threshold and remains at the preset fifth time, the self-learning process ends.

[0022] Furthermore, the fault diagnosis device may also be equipped with a third data update unit, which compares the historical dynamic window or set of historical dynamic windows corresponding to historical operating conditions with the current dynamic window corresponding to the current self-learning value. If the elements of the historical dynamic window or set of historical dynamic windows overlap with the current dynamic window or the intersection of them is greater than a preset ratio, then the elements in the historical self-learning value or the historical self-learning value sequence are replaced with the current self-learning value; otherwise, the earliest stored historical self-learning value is used as the first element of the self-learning value sequence, and the minimum number of elements N in the self-learning sequence is 3. When the number of historical self-learning values ​​reaches the upper limit N, if the current dynamic window does not overlap with the historical self-learning value of interest or the difference is greater than a preset range, then the historical self-learning values ​​that exceed the range are replaced with the data corresponding to the current dynamic window.

[0023] If the number of elements in the historical self-learning sequence is greater than or equal to 3, the maximum and minimum values ​​in the historical self-learning sequence are removed, and the weighted average of the remaining elements in the historical self-learning sequence is used as the correction value for the self-learning process; if the number of elements in the historical self-learning sequence is less than or equal to 2, the correction value remains unchanged.

[0024] Furthermore, the fault diagnosis device may also be equipped with a fourth fault assessment unit; it can acquire emission test data of a preset set of engines or vehicles, and if the correction value is greater than the preset monitoring threshold, it is considered that there is a fault in the fuel supply system.

[0025] The dynamic window can also use a fixed window range or correspond to a preset range of operating conditions; its operating conditions are constrained or given by preset engine parameters or air-fuel mixture parameters.

[0026] Accordingly, embodiments of the present invention also disclose a computer storage medium and a controller; the computer storage medium includes a storage medium body for storing a computer program; when the computer program is executed by a microprocessor, it is used to implement any of the above-mentioned mixed gas information processing methods; the controller includes any of the above-mentioned fault diagnosis devices and / or the computer storage medium can be used to solve the same technical problem.

[0027] In summary, the first data temporary storage step / unit of this invention records the operating condition information corresponding to a single self-learning value and uses it as a dynamic window to filter operating conditions; the second learning freeze step / unit filters information based on the deviation between the dynamic window and the operating conditions to ensure that the self-learning process is implemented under acceptable operating conditions, and freezes data exceeding the range of the dynamic window until the operating conditions return to the range defined by the dynamic window; the third data update step / unit uses the self-learning value obtained after dynamic window filtering to solve for the correction value; and the fourth fault assessment step / unit performs fault diagnosis of the fuel supply system through the relationship between the correction value and the monitoring threshold. The product can be used in structures such as Electronic Control Units (ECUs) and Engine Management Systems (EMSs) to improve the efficiency of emission control and fault detection.

[0028] It should be noted that the terms "first," "second," and similar terms used in this article are merely for describing the constituent elements of the technical solution and do not constitute a limitation on the technical solution, nor should they be interpreted as an indication or implication of the importance of the corresponding elements; elements with terms such as "first," "second," or similar terms indicate that at least one of the elements is included in the corresponding technical solution. Attached Figure Description

[0029] To more clearly illustrate the technical solution of the present invention and facilitate a further understanding of its technical effects, features, and objectives, the present invention will be described in detail below with reference to the accompanying drawings. The drawings constitute an essential part of the specification and are used together with Embodiment 1 of the present invention to illustrate the technical solution of the present invention, but do not constitute a limitation on the present invention.

[0030] The same reference numerals in the attached diagram represent the same parts, as detailed below.

[0031] Figure 1 This is a schematic diagram of the engine's air-fuel mixture control system.

[0032] Figure 2 This is a schematic diagram of the self-learning region of a single-point mixed gas and its addition and multiplication self-learning regions.

[0033] Figure 3 This is a schematic diagram of the self-learning region for the mixed gas.

[0034] Figure 4 This is a schematic diagram of a dynamic window sequence according to an embodiment of the present invention.

[0035] Figure 5 This is a set of curves representing a single self-learning process according to an embodiment of the present invention.

[0036] Figure 6 This is a schematic diagram illustrating data correction in an embodiment of the present invention.

[0037] Figure 7 This is a schematic diagram illustrating data storage and updating in an embodiment of the present invention.

[0038] Figure 8 This is a schematic diagram of the process of an embodiment of the method of the present invention.

[0039] Figure 9 The data simulation verification results are for embodiments of the present invention.

[0040] Figure 10 This is a schematic diagram of the structural composition of an embodiment of the device of the present invention.

[0041] Figure 11 This is a schematic diagram of the composition structure of an embodiment of the product of the present invention. Figure 1 .

[0042] Figure 12 This is a schematic diagram of the composition structure of an embodiment of the product of the present invention. Figure 2 .

[0043] Figure 13 This is a schematic diagram of the composition structure of an embodiment of the product of the present invention. Figure 3 .

[0044] Figure 14 This is a timing diagram illustrating the self-learning process in an embodiment of the present invention.

[0045] Figure 15 for Figure 14 The dynamic window corresponding to the process is a schematic diagram of operating condition information (speed and load).

[0046] in: 001 - Engine; 002 - Electronic Control Unit, or ECU; 003 - Injector; 004 - First oxygen sensor, i.e., upstream oxygen sensor; 005 - Catalyst; 006 - Second oxygen sensor; i.e., downstream oxygen sensor; 010 - Single-point self-learning addition self-learning area, i.e., ORA; 011 - Single-point self-learning multiplication self-learning zone, i.e., FRA; 099 - Partition self-learning region division; 100 - First data temporary storage step; 200 - Second learning freeze step; 201 - Collection of rectangular window regions; 211 - First dynamic window; 212 - Second dynamic window; 213 - Third dynamic window; 300 - Third data update step; 333 - Exceeding the current operating point of the dynamic window; 400 - Fourth Fault Assessment Step; 404 - Single self-learning value; 444 - Self-learning correction value; 555 - Intermittent gas mixture interference; 600 - Fault Diagnosis Device; 610 - First temporary data storage unit; 620 - Second Learning Freeze Unit; 630 - Third Data Update Unit; 640 - Fourth Fault Assessment Unit; 900 - Vehicles; 901 - Controller; 903 - Computer storage media. Detailed Implementation

[0047] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Of course, the specific embodiments described below are merely illustrative of the technical solutions of the present invention, and not intended to limit the invention. Furthermore, the parts described in the embodiments or drawings are merely illustrative examples of relevant parts of the present invention, and not the entirety of the invention.

[0048] like Figure 8 The engine mixture information processing method shown includes a first data temporary storage step 100 and a second learning freeze step 200. The first data temporary storage step 100 saves the operating condition information corresponding to a single self-learning value and uses the operating condition information as a dynamic window for filtering operating conditions. The operating conditions include the current operating conditions and historical operating conditions. The second learning freeze step 200 filters the self-learning process based on the deviation between the dynamic window and the operating conditions. If the current operating condition exceeds the range limited by the dynamic window, the current self-learning process is frozen until the current operating condition returns to the range limited by the dynamic window. The self-learning process is carried out when the engine mixture composition is under closed-loop control or when the engine emission control reaches the preset standard.

[0049] The self-learning process is used to correct the first deviation information in the mixture information; the mixture information is processed by self-learning through single-point self-learning and / or regional self-learning processes, thereby intervening in and / or correcting the mixture information; the current operating condition is the operating condition when the self-learning process is being executed; the historical operating condition is at least one operating condition or operating condition sequence when the self-learning process was performed before entering the current operating condition.

[0050] Specifically, its operating condition information includes a first speed and / or a second load; the second load is characterized by the required torque and / or preset parameters or parameter sets of the engine; the first deviation information includes a first systematic pre-control deviation and / or a second non-systematic pre-control deviation; the self-learning process is limited to the multiplicative self-learning process, i.e., FRA, to specifically correct proportional offset systematic pre-control deviations.

[0051] The dynamic window is based on the working condition information and the range of the dynamic window is set according to the preset first offset set; the first offset set is used to limit the value range of the preset parameters in the working condition information.

[0052] Specifically, its operating condition information is obtained by using the first offset set of the first speed and the second load as parameters, as follows: Figure 4 The rectangular window region set 201 is shown, and the rectangular window region set 201 is used as a dynamic window to filter the current working condition and / or historical working condition at different times.

[0053] Furthermore, the self-learning process uses a low-pass filter to approach the current mixture pre-control deviation, i.e., xfrai, with a neutral value of 1.0; at the same time, it obtains the first positive process variable, flpfrai, and the second negative process variable, flnfrai; and then uses the larger absolute value of the first positive process variable and the second negative process variable as the third process variable to characterize the progress depth or learning degree of the self-learning process.

[0054] Among them, the first positive process variable and the second negative process variable approach the target value from the direction with larger absolute value through a low-pass filtering process; the target value is the difference between the current mixture pre-control deviation and the current self-learning value; as the self-learning process deepens, if the current self-learning value reaches the current mixture pre-control deviation level, the target value approaches 0; if the third process variable is less than the preset fourth process threshold and remains at the preset fifth duration, the self-learning process ends.

[0055] Furthermore, such as Figure 8As shown, the mixed gas information processing method also includes a third data update step 300, which compares the historical dynamic window or set of historical dynamic windows corresponding to historical operating conditions with the current dynamic window corresponding to the current self-learning value. If the elements of the historical dynamic window or set of historical dynamic windows overlap with the current dynamic window or the intersection of the elements is greater than a preset ratio, then the current self-learning value replaces the elements in the historical self-learning value or the historical self-learning value sequence; otherwise, a self-learning value sequence is constructed with at least one historical self-learning value and the current self-learning value until the number of elements in the self-learning value sequence reaches or exceeds a positive integer N; if the number of elements in the self-learning value sequence when the current self-learning value is added is equal to a positive integer N, and N is greater than or equal to 3, then the current self-learning value replaces the earliest historical self-learning value added to the historical self-learning value sequence; wherein, the upper limit of the number of elements in the self-learning sequence is N, and the minimum of N is 3, which is used to calculate the correction value for coordination; the previous self-learning correction value is used as the first value of the historical self-learning sequence, and the historical dynamic window and the windows of subsequent self-learning values ​​have no intersection.

[0056] Among them, such as Figure 7 As shown, if the number of elements in the historical self-learning sequence is greater than or equal to 3, the maximum and minimum values ​​in the historical self-learning sequence are removed, and the weighted average of the remaining elements in the historical self-learning sequence is used as the correction value for the self-learning process; if the number of elements in the historical self-learning sequence is less than or equal to 2, the correction value remains unchanged.

[0057] Furthermore, such as Figure 8 As shown, the mixture information processing method also includes a fourth fault assessment step 400; obtaining emission test data of a preset engine set or vehicle set, and if the correction value is greater than the preset monitoring threshold, it is considered that there is a fault in the fuel supply system.

[0058] In addition, its dynamic window can adopt a fixed window range or correspond to a preset operating condition range; its operating conditions can be constrained or given by preset engine parameters or air-fuel mixture parameters.

[0059] Accordingly, such as Figure 10 The fault diagnosis device 600 shown includes a first data temporary storage unit 610 and a second learning freeze unit 620. The first data temporary storage unit 610 stores the working condition information corresponding to a single self-learning value and uses the working condition information as a dynamic window as the basis for working condition filtering. The working conditions include the current working condition and historical working conditions.

[0060] Furthermore, its second learning freeze unit 620 filters the self-learning process based on the deviation between the dynamic window and the operating condition; if the current operating condition exceeds the range limited by the dynamic window, the current self-learning process is frozen until the current operating condition returns to the range limited by the dynamic window; the self-learning process is carried out when the engine's air-fuel mixture is under closed-loop control or when the engine's emission control reaches the preset standard; its self-learning process is used to correct the first deviation information in the air-fuel mixture information.

[0061] Among them, the mixture information is processed by self-learning through single-point self-learning and / or regional self-learning processes, and then the mixture information is intervened and / or corrected; the current operating condition is the operating condition when the self-learning process is being executed; the historical operating condition is at least one operating condition or operating condition sequence when the self-learning process was performed before entering the current operating condition.

[0062] Specifically, its operating condition information includes a first speed and / or a second load; the second load is characterized by the required torque and / or preset parameters or parameter sets of the engine; the first deviation information includes a first systematic pre-control deviation and / or a second non-systematic pre-control deviation; its self-learning process is limited to a multiplicative self-learning process, i.e., FRA, used to correct proportional offset systematic pre-control deviations; its dynamic window is based on the operating condition information, and the range of the dynamic window is set according to the preset first offset set; the first offset set is used to limit the value range of the preset parameters in the operating condition information.

[0063] Specifically, such as Figure 4 As shown, the operating condition information uses the first offset set of the first speed and the second load as parameters to obtain a rectangular window region set 201, and uses the rectangular window region set 201 as the dynamic window to filter the current operating condition and / or historical operating condition at different times. Among them, the dynamic window corresponding to the current operating condition determined by the first speed and the second load is a rectangular window region formed by adding or subtracting the first offset, with the operating condition point determined by the filtered values ​​of the first speed and the second load as the center. That is, the dynamic window changes with the change of the current operating condition point, but it is not centered on the current operating condition point. Instead, it freezes the current self-learning process when the current operating condition point exceeds the range of the dynamic window. In addition, when a single self-learning is completed, the current dynamic window is stored. It is not necessary to save all the changes of the dynamic window during the single self-learning process. The dynamic window storage only needs to reflect the acceptable range of this self-learning or be sufficient to express the dynamic window at the time of completion.

[0064] The self-learning process employs a low-pass filter to approach the current pre-control deviation of the mixture, xfrai, by a neutral value of 1.0. Simultaneously, a first positive process variable, flpfrai, and a second negative process variable, flnfrai, are obtained. The variable with the larger absolute value between the first positive and second negative process variables is then used as a third process variable to characterize the progress depth or learning degree of the self-learning process.

[0065] Specifically, the first positive process variable and the second negative process variable approach the target value from the direction with the larger absolute value through a low-pass filtering process; the target value is the difference between the current mixture pre-control deviation and the current self-learning value; as the self-learning process deepens, if the current self-learning value reaches the current mixture pre-control deviation level, the target value approaches 0; if the third process variable is less than the preset fourth process threshold and remains at the preset fifth duration, the self-learning process ends.

[0066] Furthermore, such as Figure 10 The fault diagnosis device 600 shown also includes a third data update unit 630, which compares the historical dynamic window or set of historical dynamic windows corresponding to historical operating conditions with the current dynamic window corresponding to the current self-learning value. If the elements of the historical dynamic window or set of historical dynamic windows overlap with the current dynamic window or their intersection is greater than a preset ratio, then the current self-learning value replaces the elements in the historical self-learning value or the historical self-learning value sequence; otherwise, a self-learning value sequence is constructed with at least one historical self-learning value and the current self-learning value until the number of elements in the self-learning value sequence reaches or exceeds a positive integer N; if the number of elements in the self-learning value sequence when the current self-learning value is added is equal to a positive integer N, and N is greater than or equal to 3, then the current self-learning value replaces the earliest historical self-learning value added to the historical self-learning value sequence.

[0067] If the number of elements in the historical self-learning sequence is greater than or equal to 3, the maximum and minimum values ​​in the historical self-learning sequence are removed, and the weighted average of the remaining elements in the historical self-learning sequence is used as the correction value for the self-learning process; if the number of elements in the historical self-learning sequence is less than or equal to 2, the correction value remains unchanged.

[0068] Furthermore, such as Figure 10 The fault diagnosis device 600 shown is also equipped with a fourth fault assessment unit 640; it acquires emission test data of a preset engine set or vehicle set, and if the correction value is greater than the preset monitoring threshold, it is considered that there is a fault in the fuel supply system.

[0069] Similarly, its dynamic window can adopt a fixed window range or correspond to a preset operating condition variation range; its operating conditions are constrained or given by preset engine parameters or mixture parameters.

[0070] Similarly, such as Figures 11 to 13 The computer storage medium 903 and controller 901 shown adopt the same inventive concept and can be used to solve the same technical problem; specifically, the computer storage medium 903 includes a storage medium body for storing computer programs; when the computer program is executed by the microprocessor, it can implement any of the above-mentioned engine mixture information processing methods; the controller 901 includes any of the above-mentioned fault diagnosis device 600 and / or computer storage medium 903, the implementation process of which will not be described in detail.

[0071] In practical applications, the multi-point gas mixture self-learning process based on dynamic window sequences in the embodiments of the present invention can retain the advantages of gas mixture self-learning schemes in related technologies; at the same time, it can also make up for their defects and deficiencies and avoid the risk of misdiagnosis.

[0072] On the one hand, this invention avoids the mislearning process of single-point gas mixture self-learning when dealing with local non-systematic pre-control deviations; it can perform self-learning values ​​for multiple representative operating points and obtain compensation information for systematic pre-control deviations, avoiding the influence of a single local non-systematic pre-control deviation; wherein, the representativeness of the operating points is achieved through a dynamic window sequence; if the dynamic windows of two self-learning operating points are different, then when the two are sufficiently far apart, the above-mentioned risks can be avoided, and the situation of self-learning in the same local area can be eliminated.

[0073] On the other hand, considering that the main problem with zone self-learning is that the self-learning value cannot represent the overall systemic deviation, which will result in the untimely correction of the mixture deviation; in addition, the related technology requires more self-learning processes, which will affect the completeness of fuel system diagnosis and the timeliness of fault detection; the embodiments of the present invention correct the systemic deviation through the self-learning value correction process; at the same time, since its learning process does not depend on a single self-learning, it avoids the influence of local deviations.

[0074] Specifically, in such Figure 4 The self-learning operating condition area shown can still be used for single-point self-learning; however, considering that the operating condition range of additive self-learning is relatively simple and there is no local deviation influence, the region and self-learning strategy of additive self-learning ORA can remain unchanged; at the same time, its dynamic window sequence can be used for multi-point mixed gas self-learning process, only multiplicative self-learning, and obtain the correction of the proportional offset type systematic pre-control deviation in the multiplicative self-learning region.

[0075] Furthermore, such as Figure 4 As shown, its dynamic window sequence filters the engine's operating speed and load, and takes the filtered operating point as the center, with a certain upper and lower range of speed and load, such as speed ±200rpm and load ±10%, as the range defined as the dynamic window sequence of the current operating condition.

[0076] The dynamic window changes with the filtered operating point, resulting in a series of dynamic windows 211, 212, 213, etc. Its function is twofold: firstly, it can identify the operating condition represented by the current self-learning value; secondly, it can identify dynamic operating conditions and freeze the self-learning process in a timely manner. Under normal steady-state operating conditions, the current operating point coincides with the filtered operating point. However, when the operating condition changes, the current operating point will deviate from the filtered operating point. If the current operating point exceeds the range of the dynamic window sequence, it indicates that the degree of dynamic operating conditions is too great, and the self-learning process must be frozen in a timely manner.

[0077] Specifically, such as Figure 5 The single self-learning process shown is initiated when the self-learning conditions are met, i.e. when self-learning is enabled and the operating condition is within the range of multiplicative self-learning FRA, such as when the current operating condition is within the dynamic window sequence.

[0078] Among them, the self-learning value fri approaches the current mixture pre-control deviation xfrai by low-pass filtering (filter coefficient f0); at the same time, the positive and negative process variables (flpfrai and flnfrai) and the maximum process variable are calculated; the maximum process variable is the larger of the absolute values ​​of the positive and negative process variables, which is used to reflect the depth of self-learning.

[0079] Furthermore, the inputs to its positive and negative process variables are the difference between the current mixture pre-control deviation and the current self-learning value (xfrai-frai). The two process variables approach (xfrai-frai) from the larger positive initial value and negative initial value respectively through low-pass filtering (filtering coefficient f1).

[0080] Furthermore, as self-learning progresses, the current mixture pre-control deviation xfrai tends to be consistent with the current self-learning value, and the two process variables will also tend to 0. When the maximum process variable is lower than the set threshold and remains so for a certain period of time, it can be considered that self-learning has been fully completed. At this point, the self-learning process can be terminated.

[0081] Since the filtration coefficient f1 is much larger than the filtration coefficient f0, therefore, in the event of... Figure 5 When there is an occasional gas mixture disturbance at 17s, its maximum process variable will deviate from 0 at a faster rate, thus avoiding the impact of the occasional gas mixture disturbance on the final self-learned value; in addition, such as Figure 14 , 15 As shown, there is a load change in the 10-15 second interval, and the engine speed also changes accordingly. During the time interval T0-T1-T2, there is a process of the operating condition changing from stable to dynamic and then returning to steady state. When the operating condition is dynamic, it is not suitable to perform self-learning, and the self-learning process needs to be frozen.

[0082] Specifically, this can be achieved through dynamic windows; among which, Figure 15 The system provides three time points: the current operating point (lighter, larger point) and the filtered operating point (darker, smaller point), as well as the dynamic window (dashed box) for each time point. At time T0, the current operating point coincides with the filtered operating point. At time T1, the operating condition is dynamic, and the current operating point jumps out of the dynamic window, allowing the self-learning process to be frozen. At time T2, the operating condition tends to be steady again, and the current operating point returns to the dynamic window, allowing the self-learning process to be unfrozen.

[0083] The self-learning value is defined as: frai = (1-f0)(frai_old-1) + f0(xfrai-1); The forward process variable is defined as: flpfrai = (1-f1)flpfrai_old + f1(xfrai-frai); The negative process variable is defined as: flnfrai = (1-f1)flnfrai_old + f1(xfrai-frai); The maximum process variable is defined as: flmxfrai = max(flpfrai, flnfrai).

[0084] Furthermore, such as Figure 6 The self-learning values ​​can be saved, and relevant information can be corrected and updated. When a single self-learning is completed, the result of this self-learning can be saved. For example, a maximum of 5 self-learning values ​​can be saved (i.e., N=5). Before saving, it is first determined that the current self-learning condition (the filtered condition) is within the corresponding dynamic window sequence. If it is, it means that the current self-learning is close to the condition of the self-learning.

[0085] At this point, the current self-learning value can be used to replace the original self-learning value; at the same time, the dynamic window data (filtered operating conditions) and the number of times the current driving cycle has completed self-learning can be updated; if the dynamic windows corresponding to the previously saved self-learning values ​​do not contain the operating conditions of the current self-learning, then the earliest stored self-learning value and its dynamic window are updated; if the current driving cycle has not yet stored 5 self-learning values, then they are saved sequentially.

[0086] The self-learning correction value is obtained by averaging the remaining self-learning values ​​after removing the maximum and minimum values ​​from all saved self-learning values. When the number of self-learning values ​​is less than 5, the current correction value will be included in the calculation. If the number of self-learning values ​​stored in this driving cycle is less than 2, the operation of removing the maximum and minimum values ​​cannot be performed. In this case, the self-learning correction value will not be updated.

[0087] Specifically, the system stores the information corresponding to a single self-learning value and a dynamic window (such as rotational speed and load). Before storing, it needs to check whether there is a situation where the dynamic window corresponding to a previously stored self-learning value (i.e., an element in the historical self-learning value sequence) overlaps with the current self-learning dynamic window (i.e., the reference rotational speed and load point are located within the current self-learning dynamic window). If such a situation exists, it is replaced with the current self-learning value (self-learning value and dynamic window information). If not, a new set of self-learning information (self-learning value and dynamic window data) is added. If the number of stored self-learning values ​​exceeds the upper limit N (e.g., N=5), the earliest stored self-learning value is replaced with the current self-learning value.

[0088] Furthermore, such as Figure 7 As shown, the effect of the dynamic window multi-point mixture self-learning disclosed in this embodiment can also be verified. By simulating the emission test files of multiple test vehicles, the comparison between the single self-learning value and the self-learning correction value shows that although there are local mixture deviations that cause large deviations in the single self-learning value, the self-learning correction value can still remain relatively stable, which can better reflect the overall systematic pre-control deviation of the vehicle. Therefore, it can also better ensure the accuracy of mixture control and ensure the consistency of vehicle emissions. Among them, the initial correction value is the previous self-learning correction value, and at the same time, it is the first value of the self-learning storage sequence. Its historical dynamic window is centered on speed 0 and load 0, and has no intersection with the historical dynamic window of subsequent self-learning.

[0089] In summary, compared with the single-point self-learning process in related technologies, this embodiment can avoid the influence of local non-systematic mixture pre-control deviations, avoid erroneous self-learning value correction processes, ensure stable vehicle emission performance, and to a certain extent avoid misdiagnosis of fuel system faults. On the other hand, compared with the single-point self-learning process, even without local non-systematic mixture pre-control deviations, the self-learning correction value of the mixture can be made more stable by averaging the maximum and minimum values, avoiding interference from differences in mixtures under different operating conditions on the self-learning value and avoiding its impact on emissions.

[0090] Furthermore, compared to the related partition self-learning process, this embodiment can obtain a more representative systematic pre-control deviation after only two rounds of single self-learning, thus simplifying the required self-learning process and time, and making the correction of mixture deviation more timely. At the same time, this embodiment also effectively promotes the completion of fuel system diagnosis and the timeliness of engine fault diagnosis. In particular, for situations where a fixed window or fixed partition is used, but the purpose of the window or partition is consistent with that of the dynamic window, that is, when the fixed window or fixed partition can also reflect the representativeness of the operating conditions, the same technical effect can be achieved, which also falls under the solution covered by this embodiment.

[0091] It should be noted that the above embodiments are only for more clearly illustrating the technical solution of the present invention. Those skilled in the art will understand that the implementation of the present invention is not limited to the above content. Any obvious changes, substitutions or replacements made based on the above content do not exceed the scope of the technical solution of the present invention. Other implementations will also fall within the scope of the present invention without departing from the concept of the present invention.

Claims

1. A method for processing engine air-fuel mixture information, used in a self-learning process of the air-fuel mixture information, characterized in that... The process includes a first data temporary storage step (100) and a second learning freeze step (200); wherein: the first data temporary storage step (100) saves the operating condition information corresponding to a single self-learning value and uses the operating condition information as a dynamic window for operating condition filtering, the operating condition including the current operating condition and historical operating conditions; the second learning freeze step (200) filters the operating conditions of the self-learning process according to the deviation between the dynamic window and the operating condition; if the current operating condition exceeds the range limited by the dynamic window, the current self-learning process is frozen until the current operating condition returns to the range limited by the dynamic window; the self-learning process is carried out when the engine's air-fuel mixture is under closed-loop control or when the engine's emission control reaches a preset standard; The self-learning process is used to correct the first deviation information in the mixture information; the mixture information is processed by self-learning through single-point self-learning and / or regional self-learning processes, thereby intervening in and / or correcting the mixture information; the current operating condition is the operating condition when the self-learning process is being executed; the historical operating condition is at least one operating condition or operating condition sequence when the self-learning process was performed before entering the current operating condition. The method further includes a third data update step (300), which compares the historical dynamic window or set of historical dynamic windows corresponding to the historical operating conditions with the current dynamic window corresponding to the current self-learning value. If the elements of the historical dynamic window or set of historical dynamic windows overlap with the current dynamic window or the intersection of the elements is greater than a preset ratio, then the current self-learning value replaces the elements in the historical self-learning value or the historical self-learning value sequence. Otherwise, the earliest stored historical self-learning value is used as the first element of the self-learning value sequence, and the minimum number of elements N in the self-learning sequence is 3. When the number of historical self-learning values ​​reaches the upper limit N, if the current dynamic window does not overlap with the historical self-learning value being monitored, then the historical self-learning value being monitored will be replaced first.

2. The engine mixture information processing method as described in claim 1, wherein: The operating condition information includes a first speed and / or a second load; the second load is characterized by the required torque and / or the preset parameters or parameter set of the engine; the first deviation information includes a first systematic pre-control deviation and / or a second non-systematic pre-control deviation; the self-learning process is limited to the multiplicative self-learning process, i.e., FRA, used to correct proportional offset systematic pre-control deviations.

3. The engine mixture information processing method according to any one of claims 1 to 2, wherein: The dynamic window is based on the operating condition information and the range of the dynamic window is set according to a preset first offset set; the first offset set is used to limit the value range of preset parameters in the operating condition information.

4. The engine mixture information processing method as described in claim 3, wherein: The operating condition information uses the first offset set of the first rotational speed and the second load as parameters to obtain a rectangular window area set (201), and uses the rectangular window area set (201) as the value of the dynamic window at different times to filter the current operating condition and / or the historical operating condition.

5. The engine mixture information processing method as described in any one of claims 1, 2, or 4, wherein: The self-learning process uses a low-pass filter to approach the current mixture pre-control deviation, i.e., xfrai, by a neutral value of 1.

0. At the same time, a first positive process variable, flpfrai, and a second negative process variable, flnfrai, are obtained. The variable with the larger absolute value between the first positive process variable and the second negative process variable is used as a third process variable to characterize the progress depth or learning degree of the self-learning process.

6. The engine mixture information processing method as described in claim 5, wherein: The first positive process variable and the second negative process variable approach the target value from preset positive initial value and preset negative initial value, respectively, through the low-pass filtering process; The target value is the difference between the current pre-control deviation of the mixture and the current self-learning value. As the self-learning process progresses, if the current self-learning value reaches the current pre-control deviation level of the gas mixture, the target value approaches 0; if the third process variable is less than the preset fourth process threshold and remains at the preset fifth duration, the self-learning process ends.

7. The engine mixture information processing method as described in claim 6, wherein: If the number of elements in the historical self-learning sequence is greater than or equal to 3, then the maximum and minimum values ​​in the historical self-learning sequence are removed, and the weighted average of the remaining elements in the historical self-learning sequence is used as the correction value for the self-learning process; if the number of elements in the historical self-learning sequence is less than or equal to 2, then the correction value remains unchanged.

8. The engine mixture information processing method as described in claim 7 further includes a fourth fault assessment step (400): obtaining emission test data of a preset engine set or vehicle set, and if the correction value is greater than a preset monitoring threshold, then the fuel supply system is considered to have a fault.

9. The engine mixture information processing method according to any one of claims 1, 2, 4, 6, 7 or 8, wherein: The dynamic window adopts a fixed window range or corresponds to a preset operating condition variation range; the operating condition is constrained or given by preset engine parameters or air-fuel mixture parameters.

10. A fault diagnosis device (600), characterized in that... The system includes a first data storage unit (610) and a second learning freeze unit (620); wherein: the first data storage unit (610) stores the operating condition information corresponding to a single self-learning value and uses the operating condition information as a dynamic window for operating condition filtering, the operating condition including the current operating condition and historical operating conditions; the second learning freeze unit (620) filters the self-learning process according to the deviation between the dynamic window and the operating condition; if the current operating condition exceeds the range defined by the dynamic window, the current self-learning process is frozen until the current operating condition returns to the range defined by the dynamic window; the self-learning process is performed when the engine's air-fuel mixture is under closed-loop control or when the engine's emission control reaches a preset standard; the self-learning process is used to correct the first deviation information in the air-fuel mixture information; the self-learning processing of the air-fuel mixture information is performed using single-point self-learning and / or regional self-learning processes, thereby intervening in and / or correcting the air-fuel mixture information; the current operating condition is the operating condition when the self-learning process is being executed; the historical operating condition is at least one operating condition or operating condition sequence when the self-learning process was performed before entering the current operating condition. The device further includes a third data update unit (630), which compares the historical dynamic window or set of historical dynamic windows corresponding to the historical operating conditions with the current dynamic window corresponding to the current self-learning value. If the elements of the historical dynamic window or set of historical dynamic windows overlap with the current dynamic window or the intersection of the elements is greater than a preset ratio, then the current self-learning value replaces the elements in the historical self-learning value or the historical self-learning value sequence. Otherwise, the earliest stored historical self-learning value is used as the first element of the self-learning value sequence, and the minimum number of elements N in the self-learning sequence is 3. When the number of historical self-learning values ​​reaches the upper limit N, if the current dynamic window does not overlap with the historical self-learning value being monitored, then the historical self-learning value being monitored will be replaced first.

11. The fault diagnosis device (600) as described in claim 10, wherein: The operating condition information includes a first speed and / or a second load; the second load is characterized by the required torque and / or preset parameters or parameter sets of the engine; the first deviation information includes a first systematic pre-control deviation and / or a second non-systematic pre-control deviation; the self-learning process is limited to a multiplicative self-learning process, i.e., FRA, used to correct proportional offset systematic pre-control deviations; the dynamic window is based on the operating condition information, and the range defined by the dynamic window is set according to a preset first offset set; the first offset set is used to limit the value range of preset parameters in the operating condition information.

12. The fault diagnosis device (600) as described in claim 10 or 11, wherein: The operating condition information uses the first offset set of the first rotational speed and the second load as parameters to obtain a rectangular window area set (201), and uses the rectangular window area set (201) as the value of the dynamic window at different times to filter the current operating condition and / or the historical operating condition.

13. The fault diagnosis device (600) as described in claim 12, wherein: The self-learning process uses a low-pass filter to approach the current mixture pre-control deviation, i.e., xfrai, by a neutral value of 1.

0. At the same time, a first positive process variable, flpfrai, and a second negative process variable, flnfrai, are obtained. The variable with the larger absolute value between the first positive process variable and the second negative process variable is used as a third process variable to characterize the progress depth or learning degree of the self-learning process.

14. The fault diagnosis device (600) as described in claim 13, wherein: The first positive process variable and the second negative process variable approach the target value from preset positive initial value and preset negative initial value, respectively, through the low-pass filtering process; The target value is the difference between the current pre-control deviation of the mixture and the current self-learning value. As the self-learning process progresses, if the current self-learning value reaches the current pre-control deviation level of the gas mixture, the target value approaches 0; if the third process variable is less than the preset fourth process threshold and remains at the preset fifth duration, the self-learning process ends.

15. The fault diagnosis device (600) as described in claim 14, wherein: If the number of elements in the historical self-learning sequence is greater than or equal to 3, then the maximum and minimum values ​​in the historical self-learning sequence are removed, and the weighted average of the remaining elements in the historical self-learning sequence is used as the correction value for the self-learning process; if the number of elements in the historical self-learning sequence is less than or equal to 2, then the correction value remains unchanged.

16. The fault diagnosis device (600) as described in any one of claims 10, 11, 13, 14 or 15 further includes a fourth fault assessment unit (640); acquiring emission test data of a preset engine set or vehicle set, and if the correction value is greater than a preset monitoring threshold, then considering that there is a fault in the fuel supply system.

17. The fault diagnosis device (600) as described in claim 16, wherein: The dynamic window adopts a fixed window range or corresponds to a preset operating condition variation range; the operating condition is constrained or given by preset engine parameters or air-fuel mixture parameters.

18. A computer storage medium (903) comprising a storage medium body for storing a computer program; wherein the computer program, when executed by a microprocessor, implements the engine mixture information processing method as described in any one of claims 1 to 8.

19. A controller (901) comprising the fault diagnosis device (600) as claimed in any one of claims 10 to 16 and / or the computer storage medium (903) as claimed in any one of claims 18.

Images