A misfire detection data processing method, test device, medium and calibration equipment

Abstract

The embodiment of the application discloses a misfire detection data processing method, a test device, a medium and a calibration equipment; the method adopts a preset data structure, under the engine operating condition, through activating the adaptive detection function, the engine condition and operation data are synchronously acquired by online scanning and according to the ignition phase, and then the corresponding detection threshold or calibration data is calculated online through the statistical method; the method or product of the application does not need to deliberately control the engine speed and load in the implementation or operation, each working parameter of the working condition point experienced in the engine operation process is automatically detected and analyzed, and then the misfire detection threshold data is automatically obtained; the method and product of the application can be realized by upgrading the engine control unit ECU software, so that the data acquisition process is simplified, the working hours are significantly shortened, the misfire detection threshold is obtained in real time, and the development efficiency is significantly improved.

Description

Technical Field

[0001] This invention belongs to the field of internal combustion engine technology, and particularly relates to a misfire detection data processing method, test device, medium and calibration equipment. Background Technology

[0002] Since the widespread application of the On-Board Diagnostics II (OBDII) system, the development of on-board diagnostic system functions has become essential to meet regulatory requirements. In vehicle research and development, production and debugging, wheel chock testing and related data analysis and processing typically consume a significant proportion of time. In particular, the acquisition, detection, analysis, and processing of misfire data involves numerous calibration steps, including condition conversions, numerical simulations, comparisons, and optimizations. This makes the testing process complex, with slow sampling speeds, long cycles, and high manpower and resource consumption, necessitating improved efficiency. Furthermore, existing systems and methods often rely on human experience, limiting sample observation points. Data processing depends on manual offline analysis, resulting in slow analysis speeds, impacting product data consistency, and increasing the workload of subsequent testing and quality verification, failing to meet the needs of efficient development. Summary of the Invention

[0003] This invention discloses a fire detection data processing method, including a first initialization step and a second data acquisition step; wherein: the first initialization step sets the coordinates of the center point of a self-learning region; its row coordinates correspond to different rotational speeds nmot, and its column coordinates correspond to different load states midmd; then the coordinates of the center point of the self-learning region can be numbered to store or locate the corresponding detection data; the coordinates of the center point of the self-learning region correspond to a preset operating point, and the preset operating point corresponds to a specific rotational speed nmot and / or a specific load state midmd.

[0004] Furthermore, by sending a status word command to the engine control unit (ECU) online to activate the matching quantity self-learning process, random misfires or continuous misfires are implanted by actively cutting off ignition or fuel injection. At this time, for each ignition phase, the engine misfires during the power stroke of the set ignition or fuel injection cut-off phase. The second data acquisition step is based on synchronously acquiring and / or storing engine data based on the engine ignition phase to form each set of data samples. Starting from a preset first speed, a misfire process or misfire state is created in the central region Zone(i,j) of each feature point defined by the coordinates of the center point of the self-learning region. In this misfire process and misfire state, misfires occur in the engine cylinders. The engine load state midmd can then be adjusted so that the load state midmd covers all regions from idle to full load, thereby scanning or simulating the corresponding misfire process or state.

[0005] Specifically, each set of data samples may include rotational speed nmot, load status midmd, random misfire characteristic signal luts, continuous misfire characteristic signal fluts, and misfire mode flag signal evz_austot.

[0006] Furthermore, its second data acquisition step can receive adaptive instruction codes and / or activate the matching quantity self-learning process. The matching quantity self-learning process selects valid data and discards invalid data in the data sample according to preset standards, thereby obtaining all operating points that would cause the expected fire.

[0007] Specifically, the aforementioned standard includes the values ​​or combinations of the above data samples during and / or under the fire failure process and state; the combination of values ​​includes at least two of the above-mentioned speed nmot, load state midmd, random fire failure characteristic signal luts, continuous fire failure characteristic signal fluts, and fire failure mode flag signal evz_austot; by repeatedly executing the processing procedure of the second data acquisition step and storing or transmitting its valid data, the operating condition information of the corresponding fire failure process or state, that is, the operating condition parameters of the fire failure point, can be obtained, including speed nmot and / or load state midmd.

[0008] To obtain valid data, the method may further include a third data filtering step; this third data filtering step detects the misfire process and / or misfire state; if the misfire process and / or misfire state actively implanted by the ECU exists, then the data sample at this time is recorded or stored as valid data; this valid data is used for subsequent calculation of the misfire characteristic parameter threshold for sub-regions; wherein, the feature point center region Zone(i,j) is determined according to the preset speed and load, i is the number of the speed zone in the feature point center region Zone(i,j), m is the number of speed zones m, i varies from 1 to m; j is the number of the load zone in the feature point center region Zone(i,j), n is the number of load zones, j varies from 1 to n.

[0009] Furthermore, the third data filtering step can classify each data point of the valid data, i.e., flutsk(nmot, midmd), into the corresponding self-learning region Zone(i, j) according to the speed nmot and load state midmd, and store them with numbers according to the speed nmot and load state midmd; based on the misfire mode flag signal evz_austot, if the misfire process and / or misfire state actively implanted by the ECU does not exist, the data samples obtained in the second data acquisition step are discarded.

[0010] Furthermore, to improve data fluctuation and reduce randomness, k sets of the above-mentioned types of valid data can be scanned and continuously accumulated by region in the third data filtering step, where k is a positive integer. For each newly added set of valid data, the random fire feature signal luts(nmot,midmd) and the continuous fire feature signal flutsk(nmot,midmd) are summed and counted in the feature point center region Zone(i,j) to which the valid data belongs, respectively, to obtain the region and SZone. k，evz_austot (i, j) and the corresponding sample count value CountZone k，evz_austot (i, j), where the subscript k represents the cumulative number of fires in the region; MI represents the fire fault characteristic signal, including the random fire fault characteristic signal luts(nmot, midmd) and the continuous fire fault characteristic signal flutsk(nmot, midmd)), and the subscript evz_austot distinguishes the fire fault mode, such that:

[0011] SZone k，evz_austot (i, j) = SZone k-1，evz_austot (i,j)+MI k，evz_austot (i, j), SZone k-1，evz_austot (i, j) represents the value of the previous summation;

[0012] CountZone k，evz_austot (i, j) = CountZone k-1，evz_austot (i, j) + 1, CountZone k-1，evz_austot (i, j) represents the previous count value;

[0013] When the number of valid data points reaches N (where N is a positive integer), the learning process corresponding to the feature point center region Zone(i,j)(999) ends, and the average value SZoneAve of the feature parameters corresponding to the feature point center region Zone(i,j)(999) is output. N，evz_austot (i, j), such that:

[0014] SZoneAve N，evz_austot (i, j) = SZone N，evz_austot (i, j) / N; and such that:

[0015] The learning attribute A(i,j) corresponding to the central region Zone(i,j) of the feature point is set from 0 to 1.

[0016] Furthermore, the method of the present invention may further include a fourth pulse spectrum generation step to obtain a series of threshold parameters for engine misfire detection; the misfire threshold parameters constitute a pulse spectrum MAP according to the coordinates of the engine speed and load condition points, specifically referring to the misfire characteristic parameter thresholds stored in the ECU according to the coordinates of the operating points. The fourth pulse spectrum generation step obtains the average value SZoneAve. N，evz_austot The fire threshold TAve corresponding to (i, j) evz_austot (i, j), such that: TAve evz_austot (i, j) = factor * SZoneAve N，evz_austot (i, j); where factor is the scaling factor;

[0017] Use this scaling factor to correct SZoneAve N，evz_austot (i, j) and fill in the pulse spectrum; the pulse spectrum is used to describe the threshold parameters for fire detection at the corresponding operating point when a fire occurs; the parameters of the operating point include the speed nmot parameter and / or the load state midmd parameter; by traversing the speed nmot and the load region midmd, all the operating points and / or related feature data corresponding to the self-learning region Zone (i, j) can be obtained, and the process of filling in its pulse spectrum can be completed.

[0018] Specifically, examining the learning attribute A(i,j) corresponding to the central region Zone(i,j) of the feature point, based on the rotational speed nmot and / or load state midmd coordinates, for regions with learned regions in the surrounding coordinates, they belong to regions not reached by the engine; for unlearned regions at the coordinate edges, they belong to regions not reached by the engine or operating regions that are impossible to reach. If unlearned regions exist, the above pulse spectrum is filled using interpolation. If there are regions that do not need to be diagnosed, the pulse spectrum of that point can be filled with a preset maximum value.

[0019] Furthermore, the aforementioned data samples can be extracted in real time from the vehicle and / or engine control unit and / or microprocessor, which includes the vehicle electronic control unit (ECU); its misfire or misfire data processing corresponds to random misfire and / or continuous misfire; the random misfire and continuous misfire correspond to different misfire mode flag signals evz_austot; its second data acquisition step can be completed under free operating conditions, which traverse the engine speed nmot range and load state midmd range.

[0020] Furthermore, this embodiment of the invention also discloses a test apparatus, including a first initialization unit and a second data acquisition unit; wherein: the first initialization unit has a preset self-learning region center point coordinate data structure; the row coordinates of this data structure correspond to different rotational speeds nmot, and the column coordinates correspond to different load states midmd; then, by numbering the center point coordinates of the self-learning region, the center point coordinates of the self-learning region correspond to preset operating points, which correspond to specific rotational speeds nmot and / or specific load states midmd.

[0021] Furthermore, its second data acquisition unit acquires and / or stores engine data according to the engine ignition phase to form each set of data samples; it traverses the engine's reachable operating condition range according to the engine speed nmot and load midmd; during this period, random misfire or continuous misfire states are implanted during engine operation; each set of data samples includes engine speed nmot, load state midmd, random misfire characteristic signal luts, continuous misfire characteristic signal fluts, and misfire mode flag signal evz_austot;

[0022] Specifically, its second data acquisition unit can receive adaptive instruction codes and / or activate the matching quantity self-learning process. The matching quantity self-learning process selects valid data and discards invalid data in the data sample according to preset standards. The standards include the values ​​or combinations of values ​​of the data sample during the fire process and / or under the fire state. The combinations of values ​​include rotational speed nmot, load state midmd, random fire characteristic signal luts, continuous fire characteristic signal flutst, and fire mode flag signal evz_austot. The processing of its second data acquisition unit is repeatedly executed and valid data is stored or transmitted, thereby obtaining all fire condition points that need to be tested or experimented on.

[0023] Furthermore, the hub testing device may also include a third data filtering unit; the third data filtering unit detects the fire process and / or fire state; if the fire process and / or fire state exists, the data sample at this time is recorded or stored as valid data; the valid data can be used for subsequent processing.

[0024] Among them, the central region of the feature point, Zone(i,j), can be determined according to the preset speed and load. i is the number of the speed zone in the central region of the feature point, m is the number of speed zones, and i varies from 1 to m; j is the number of the load zone in the central region of the feature point, n is the number of load zones, and j varies from 1 to n.

[0025] Specifically, its third data filtering unit can classify each set of data points flutsk(nmot, midmd) of valid data into the corresponding self-learning region Zone(i, j) according to the rotation speed nmot and load state midmd, and store them with numbers according to the rotation speed nmot and load state midmd; if the fire process and / or fire state do not exist, the data samples obtained in the second data acquisition unit are discarded.

[0026] Furthermore, the third data filtering unit can scan and continuously accumulate N valid data points by region, where N is a positive integer. For each newly added valid data point, the random fire feature signal luts(nmot,midmd) and the continuous fire feature signal flutsk(nmot,midmd) can be summed and counted in the central region Zone(i,j) of the feature point to which the valid data point belongs, respectively, to obtain the region and SZone. k，evz_austot (i, j) and the corresponding sample count value CountZone k，evz_austot (i, j), where the subscript k represents the cumulative number of fires in the region; MI represents the fire fault characteristic signal, including the random fire fault characteristic signal luts(nmot, midmd) and the continuous fire fault characteristic signal flutsk(nmot, midmd)), and the subscript evz_austot distinguishes the fire fault mode, such that:

[0027] SZone k，evz_austot (i, j) = SZone k-1，evz_austot (i,j)+MI k，evz_austot (i, j), SZone k-1，evz_austot (i, j) represents the value of the previous summation;

[0028] CountZone k，evz_austot (i, j) = CountZone k-1，evz_austot (i, j) + 1, CountZone k-1，evz_austot (i, j) represents the previous count value;

[0029] When the number of valid data points reaches N (where N is a positive integer), the learning process corresponding to the feature point center region Zone(i,j)(999) ends, and the average value SZoneAve of the feature parameters corresponding to the feature point center region Zone(i,j)(999) is output. N，evz_austot (i, j), such that:

[0030] SZoneAve N，evz_austot (i, j) = SZone N，evz_austot (i, j) / N; and such that:

[0031] The learning attribute A(i,j) corresponding to the central region Zone(i,j) of the feature point is set from 0 to 1.

[0032] Furthermore, the experimental setup may also include a fourth pulse spectrum generation unit; this fourth pulse spectrum generation unit obtains the average value SZoneAve N，evz_austot The fire threshold TAve corresponding to (i, j) evz_austot (i, j), such that:

[0033] TAve evz_austot (i, j) = factor * SZoneAve N，evz_austot (i, j), where factor is the scaling factor;

[0034] TAve evz_austot (i, j) are filled in the pulse spectrum diagram; this pulse spectrum diagram is used to describe the relevant characteristic data threshold of the corresponding operating point when a fire occurs. When the fire characteristic value exceeds the threshold, it is judged that a corresponding fire has occurred; the parameters of its operating point include the speed nmot parameter and / or the load state midmd parameter.

[0035] Specifically, iterate through all the working points and / or related feature data of the self-learning region Zone(i,j) and complete the process of filling in its pulse spectrum.

[0036] If there are operating areas that the engine has not reached or cannot reach, or areas that do not require diagnosis, then the pulse spectrum is filled with interpolated values; if there are areas that do not require diagnosis, then the pulse spectrum of that point is filled with the maximum value.

[0037] Furthermore, its data samples can be extracted in real time from the vehicle and / or engine control unit and / or microprocessor, the control unit including the vehicle electronic control unit (ECU); its misfire or misfire data processing corresponds to random misfire and / or continuous misfire; its random misfire and continuous misfire correspond to different misfire mode flag signals evz_austot; the second data acquisition unit can be completed under free operating conditions, under which the engine speed nmot and load midmd traverse the reachable area.

[0038] Furthermore, in order to improve the efficiency of the testing process and obtain intermediate results and testing progress in a timely manner, the testing device may also include a fifth human-computer interaction unit. This fifth human-computer interaction unit can receive and / or display pulse spectrum diagrams to provide the current or planned test subjects, intermediate results or related content of the testing device.

[0039] Furthermore, embodiments of the present invention also disclose a computer storage medium, including a storage medium body for storing a computer program; when the computer program is executed by a microprocessor, it can implement any of the above-described fire detection data processing methods.

[0040] Furthermore, embodiments of the present invention also disclose an engine calibration device, including any of the above-mentioned test devices; and / or the above-mentioned computer storage medium; an engine test module can also be constructed from the above-mentioned devices or media, the engine test module being an integrated unit or assembly system of standardized test devices and / or computer storage media; the engine test module can exchange information with the vehicle or test equipment through the vehicle bus or communication line.

[0041] The methods, apparatus, media, and equipment disclosed in the embodiments of the present invention do not require deliberate control of speed and load during implementation or operation, and can achieve adaptive detection and processing for every working condition swept by the engine.

[0042] By introducing the technical solution of this invention, the efficiency of data acquisition and processing is effectively improved, and the testing process is simplified; the working time of related tests is significantly shortened, and relevant thresholds or characteristic quantities can be obtained in real time, thus significantly improving the efficiency of the hub test.

[0043] In view of the above-mentioned beneficial effects, the method and product of the present invention can be implemented in the upgrade of existing systems mainly through software, which can significantly improve the efficiency of fire detection calibration test and data analysis, and greatly reduce the product development cycle.

[0044] 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

[0045] 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.

[0046] The same reference numerals in the attached diagrams represent the same parts, specifically:

[0047] Figure 1 This is a schematic diagram of the structure and process of an embodiment of the method of the present invention.

[0048] Figure 2 This is a schematic diagram of the virtual test data storage structure in an embodiment of the method of the present invention.

[0049] Figure 3 This is a partially rotated enlarged view of the virtual test curve illustration in an embodiment of the method of the present invention.

[0050] Figure 4 This is a virtual test curve for an embodiment of the method of the present invention.

[0051] Figure 5 This is a test data example for an embodiment of the method of the present invention.

[0052] Figure 6 This is a schematic diagram of the structural composition of the device, module, and equipment embodiments of the present invention.

[0053] Figure 7 This is a schematic diagram of the structural components of an embodiment of the hub testing device of the present invention.

[0054] in:

[0055] 001 - Missfire mode flag signal evz_austot,

[0056] 002 - Fire fault characteristic signals luts or fluts,

[0057] 003-Spindle speed nmot,

[0058] 004-Load Status midmd,

[0059] 011 - Data structure for the first speed range (2500 rpm)

[0060] ii - Data structure for the i-th rotational speed region (6500 rpm),

[0061] mm-mth rotational speed region (8500) data structure,

[0062] 100 - First initialization step,

[0063] 111 - Test data structure with 10% load

[0064] 200 - Second data acquisition step,

[0065] 300 - Third data filtering step,

[0066] Steps for generating the fourth pulse spectrum (400).

[0067] jjj - Test data structure under 70% load.

[0068] nnn-Test data structure with 90% load,

[0069] 600 - Test Apparatus

[0070] 610 - First Initialization Unit

[0071] 620 - Second Data Acquisition Unit

[0072] 630 - Third Data Filtering Unit

[0073] 640 - Fourth pulse spectrum generation unit,

[0074] 666 - Fifth Human-Computer Interaction Unit;

[0075] 700-Engine Test Module

[0076] 903 - Computer Storage Media

[0077] 800-Engine Calibration Equipment

[0078] 999 - Elements of the test point sequence

[0079] 1234 - Process parameters

[0080] 1465 - Data instances collected from the load range of 14% to 65%.

[0081] Data examples collected in the range of 2480-2400 (rpm) to 8000 (rpm) by region. Detailed Implementation

[0082] 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.

[0083] like Figure 1 The fire detection data processing method shown includes a first initialization step 100 and a second data acquisition step 200; wherein: the first initialization step 100 sets the coordinates of the center point of the self-learning area.

[0084] Among them, such as Figure 2 As shown, row coordinates 011, ii, mm correspond to different speeds nmot, and column coordinates 111, jjj, nnn correspond to different load states midmd; the coordinates of the center point of the self-learning region are numbered; the coordinates of the center point of the self-learning region correspond to the preset operating point, and the preset operating point corresponds to a specific speed nmot and / or a specific load state midmd.

[0085] Furthermore, the second data acquisition step 200 acquires and / or stores engine data at the engine ignition time to form each data sample; starting from a preset first speed, a misfire process or misfire state is created in the central region Zone(i,j) of each group of feature points defined by the coordinates of the center point of the self-learning region; in the misfire process and misfire state, there is a misfire phenomenon in the cylinder of the engine.

[0086] Furthermore, such as Figure 3 , Figure 4 As shown, the load state midmd of the engine can be adjusted so that the load state midmd covers all regions from idle state to full load state; each set of data samples may include at least one of the following: engine speed nmot, load state midmd, random misfire characteristic signal luts, continuous misfire characteristic signal fluts, and misfire mode flag signal evz_austot.

[0087] The second data acquisition step 200 can receive adaptive instruction codes and / or activate a matching quantity self-learning process. This matching quantity self-learning process selects valid data and discards invalid data from the data samples according to preset standards. The standards include the values ​​or combinations of values ​​of the data samples during the fire process and / or under the fire state. The combination of values ​​includes at least two of the following: rotational speed nmot, load state midmd, random fire characteristic signal luts, continuous fire characteristic signal fluts, and fire mode flag signal evz_austot. The processing of the second data acquisition step 200 is repeatedly executed and its valid data is stored or transmitted.

[0088] Furthermore, the method embodiment may also include a third data filtering step 300, used to detect the fire process and / or fire state; if the fire process and / or fire state exist, then the data sample at this time is recorded or stored as valid data.

[0089] Specifically, this valid data will be used for subsequent processing; wherein, the central region of the feature point, Zone(i,j), can be determined according to the preset speed and load, i is the number of the speed zone in the central region of the feature point, m is the number of speed zones, and i varies from 1 to m; j is the number of the load zone in the central region of the feature point, n is the number of load zones, and j varies from 1 to n.

[0090] Furthermore, the third data filtering step 300 can classify each group of data points flutsk(nmot, midmd) of valid data into the corresponding self-learning region Zone(i, j) according to the rotational speed nmot and the load state midmd, and store them with numbers according to the rotational speed nmot and the load state midmd.

[0091] Specifically, if the fire process and / or fire state do not exist, the data samples obtained in the second data acquisition step 200 are discarded.

[0092] Further, the third data filtering step 300 can scan and continuously acquire k groups of the aforementioned valid data, where k is a positive integer; for each newly added group of the aforementioned valid data, the random fire feature signal luts(nmot,midmd) and the continuous fire feature signal flutsk(nmot,midmd) are summed and counted in the feature point center region Zone(i,j) (999) to which the valid data belongs, respectively, to obtain the region and SZone. k，evz_austot (i, j) and the corresponding sample count value CountZone k，evz_austot (i, j), where the subscript k represents the cumulative number of fires in the region; MI represents the fire fault characteristic signal, including the random fire fault characteristic signal luts(nmot, midmd) and the continuous fire fault characteristic signal flutsk(nmot, midmd)), and the subscript evz_austot distinguishes the fire fault mode, such that:

[0093] SZone k，evz_austot (i, j) = SZone k-1，evz_austot (i,j)+MI k，evz_austot (i, j), SZone k-1，evz_austot (i, j) represents the value of the previous summation;

[0094] CountZone k，evz_austot (i, j) = CountZone k-1，evz_austot (i, j) + 1, CountZone k-1，evz_austot (i, j) represents the previous count value;

[0095] When the number of valid data points reaches N (where N is a positive integer), the learning process corresponding to the feature point center region Zone(i,j)(999) ends, and the average value SZoneAve of the feature parameters corresponding to the feature point center region Zone(i,j)(999) is output. N，evz_austot (i, j), such that:

[0096] SZoneAve N，evz_austot (i, j) = SZone N，evz_austot (i, j) / N; and such that:

[0097] The learning attribute A(i,j) corresponding to the central region Zone(i,j) (999) of the feature point is set from 0 to 1.

[0098] Furthermore, such as Figure 1 and Figure 5 As shown, the data processing method may further include a fourth pulse spectrum generation step 400; the fourth pulse spectrum generation step 400 obtains the average value SZoneAve. N，evz_austot The fire threshold TAve corresponding to (i, j)evz_austot (i, j), such that:

[0099] TAve evz_austot (i, j) = factor * SZoneAve N，evz_austot (i, j), where factor is the weighting factor.

[0100] Specifically, the SZoneAve can be corrected using a scaling factor. N，evz_austot (i, j) and fill in the pulse spectrum; the pulse spectrum is used to describe the corresponding operating point and related feature data when the misfire occurs; the parameters of the operating point include the speed nmot parameter and / or the load state midmd parameter; obtain the operating points and / or related feature data of all self-learning regions Zone (i, j) and complete the process of filling in its pulse spectrum; if there is an operating area that the engine has not reached or cannot reach, or an area that does not need to be diagnosed, then fill its pulse spectrum with interpolated values; if there is an area that does not need to be diagnosed, then fill the pulse spectrum of that point with the maximum value.

[0101] Furthermore, its data samples can be extracted in real time from the vehicle and / or engine control unit and / or microprocessor, which may be the vehicle electronic control unit (ECU).

[0102] Among them, misfire or misfire data processing corresponds to random misfire and / or continuous misfire; the random misfire and continuous misfire correspond to different misfire mode flag signals evz_austot; its second data acquisition step 200 can be completed under free operating conditions, which traverses the engine speed nmot range and the load midmd range.

[0103] Furthermore, such as Figure 7 The experimental apparatus 600 shown includes a first initialization unit 610 and a second data acquisition unit 620; wherein, the first initialization unit 610 has a preset self-learning region center point coordinate data structure; as shown Figure 2 As shown, the row coordinates 011, ii, mm of this data structure correspond to different rotational speeds nmot, and the column coordinates 111, jjj, nnn correspond to different load states midmd. The coordinates of the center point of the self-learning region are numbered. The coordinates of the center point of the self-learning region correspond to a preset operating point, which corresponds to a specific rotational speed nmot and / or a specific load state midmd.

[0104] Furthermore, the second data acquisition unit 620 acquires and / or stores engine data at the engine ignition time to form each data sample; starting from a preset first speed, a misfire process or misfire state is generated in the central region Zone(i,j) of each set of feature points defined by the coordinates of the center point of the self-learning region; in the misfire process and misfire state, misfire occurs in the cylinder of the engine; the engine load state midmd is adjusted so that the load state midmd covers all regions from idle state to full load state; each data sample includes at least one of the following: speed nmot, load state midmd, misfire characteristic signal evz_austot, and misfire mode flag signal evz_austot.

[0105] Furthermore, the second data acquisition unit 620 can receive adaptive instruction codes and / or activate a matching quantity self-learning process. This matching quantity self-learning process selects valid data and discards invalid data from the data samples according to a preset standard. This standard includes the values ​​or combinations of values ​​of the data samples during and / or under fire conditions. The combinations of values ​​include at least two of the following: rotational speed nmot, load state midmd, fire characteristic signal evz_austot, and fire mode flag signal evz_austot. The processing procedure of the second data acquisition unit 620 is repeatedly executed, and its valid data is stored or transmitted.

[0106] Furthermore, such as Figure 7 As shown, the device may further include a third data filtering unit 630; the third data filtering unit 630 detects the fire process and / or fire state; if the fire process and / or fire state exist, the data sample at this time is recorded or stored as valid data; the valid data is then used for subsequent processing.

[0107] Among them, the central region of the feature point, Zone(i,j), is determined according to the preset speed and load. i is the number of the speed zone in the central region of the feature point, m is the number of speed zones, and i varies from 1 to m; j is the number of the load zone in the central region of the feature point, n is the number of load zones, and j varies from 1 to n.

[0108] Specifically, its third data filtering unit 630 can classify each group of data points flutsk(nmot, midmd) of valid data into the corresponding self-learning region Zone(i, j) according to the rotation speed nmot and load state midmd, and store them by number according to the rotation speed nmot and load state midmd; if the fire process and / or fire state do not exist, the data samples obtained in the second data acquisition unit 620 are discarded.

[0109] Furthermore, its third data filtering unit 630 can scan and continuously acquire k groups of the aforementioned valid data, where k is a positive integer; for each newly added group of the aforementioned valid data, the random fire feature signal luts(nmot,midmd) and the continuous fire feature signal flutsk(nmot,midmd) are summed and counted in the feature point center region Zone(i,j) (999) to which the valid data belongs, respectively, to obtain the region and SZone. k，evz_austot (i, j) and the corresponding sample count value CountZone k，evz_austot (i, j), where the subscript k represents the cumulative number of fires in the region; MI represents the fire fault characteristic signal, including the random fire fault characteristic signal luts(nmot, midmd) and the continuous fire fault characteristic signal flutsk(nmot, midmd)), and the subscript evz_austot distinguishes the fire fault mode, such that:

[0110] SZone k，evz_austot (i, j) = SZone k-1，evz_austot (i,j)+MI k，evz_austot (i, j), SZone k-1，evz_austot (i, j) represents the value of the previous summation;

[0111] CountZone k，evz_austot (i, j) = CountZone k-1，evz_austot (i, j) + 1, CountZone k-1，evz_austot (i, j) represents the previous count value;

[0112] When the number of valid data points reaches N (where N is a positive integer), the learning process corresponding to the feature point center region Zone(i,j)(999) ends, and the average value SZoneAve of the feature parameters corresponding to the feature point center region Zone(i,j)(999) is output. N，evz_austot (i, j), such that:

[0113] SZoneAve N，evz_austot (i, j) = SZone N，evz_austot (i, j) / N; and such that:

[0114] The learning attribute A(i,j) corresponding to the central region Zone(i,j) (999) of the feature point is set from 0 to 1.

[0115] Furthermore, the experimental apparatus 600 may also include a fourth pulse spectrum generation unit 640; the fourth pulse spectrum generation unit 640 acquires the average value SZoneAve. N，evz_austot The fire threshold TAve corresponding to (i, j) evz_austot (i, j), such that:

[0116] TAve evz_austot (i, j) = factor * SZoneAve N，evz_austot (i, j), where factor is the weighting factor.

[0117] Correct SZoneAve(i,j) with a scaling factor and fill it into the pulse spectrum; the pulse spectrum is used to describe the corresponding operating point and related feature data when the fire occurs; the parameters of the operating point include the rotational speed nmot parameter and / or the load state midmd parameter; obtain the operating points and / or related feature data of all self-learning regions Zone(i,j) and complete the process of filling in its pulse spectrum.

[0118] If there are operating areas that the engine has not reached or cannot reach, or areas that do not require diagnosis, then the pulse spectrum is filled with interpolated values; if there are areas that do not require diagnosis, then the pulse spectrum of that point is filled with the maximum value.

[0119] Specifically, its data samples can be extracted in real time from the vehicle and / or engine control unit and / or microprocessor, and its control unit can be the vehicle electronic control unit (ECU); its misfire or misfire data processing corresponds to random misfire and / or continuous misfire; random misfire and continuous misfire correspond to different misfire mode flag signals evz_austot; its second data acquisition unit 620 can be completed under free operating conditions, which traverses the engine speed nmot range and the load midmd range.

[0120] like Figure 6 As shown, the test device 600 may also include a fifth human-computer interaction unit 666, which receives or displays a pulse spectrum to provide the test subjects, intermediate results or related content that the test device 600 is currently conducting or planning to conduct.

[0121] In addition, such as Figure 6 The computer storage medium 903 shown may include a storage medium body for storing a computer program; when the computer program is executed by a microprocessor, it can implement any of the above fire data processing methods.

[0122] Similarly, as Figure 6 As shown, its engine calibration equipment 800 may include any of the test devices 600 as described above; and / or computer storage media 903; or may include as Figure 6 The engine test module 700 shown is an integrated unit or assembly system of a standardized test device 600 and / or computer storage medium 903; the engine test module 700 can exchange information with the vehicle or test equipment via the vehicle bus or communication line.

[0123] Among them, such as Figure 4 As shown, its operating condition data may include speed operating condition data 2480 and load operating condition data 1465; each operating condition data 2480 and 1465 is used to provide initialization conditions for the self-learning process.

[0124] Taking the misfire detection of a two-cylinder engine as an example, and selecting the continuous misfire detection threshold calibration process, its automatic data processing process includes the division of the self-learning area, the prior assignment and matching of adaptive functions, free operating condition testing, data filtering and partition fitting process.

[0125] Specifically, such as Figure 2 As shown, the self-learning region is divided based on rotational speed and load, and the coordinates of the center point of the learning region are set. Each group of learning regions is numbered and identified based on rows and columns.

[0126] The matching quantity learning process is then activated using the following adaptive function code:

[0127] Function disabled: CWAF_Flutskzy=0;

[0128] Function activation: CWAF_Flutskzy=1.

[0129] Furthermore, such as Figure 3 As shown, taking the continuous misfire characteristic signal flutsk(nmot,midmd) as an example, while driving on the road or turning the wheel, the ECU is responsible for collecting and identifying the speed and load. Once it enters the legal diagnostic range, such as 8000 rpm or other speed ranges, it generates continuous misfires in the center area of ​​the speed nmot characteristic point. At the same time, it continuously adjusts the engine load state midmd sequentially from small to large or from large to small to cover the range from idle speed to full load.

[0130] The process parameters (characteristic signals such as speed, load, and continuous misfire, such as flutsk) do not need to be considered, as these parameters already exist in the ECU memory.

[0131] Within the self-learning region (defined by a given rotational speed and load), data accompanied by continuous misfires (evz_austot=1) constitute valid data. Specifically, in this example, the characteristic signal is flutsk.

[0132] Furthermore, for each group of valid data points, flutsk(nmot, midmd), based on its rotational speed nmot and load midmd, it is assigned to the nearest learning region Zone(i,j) and identified accordingly. This yields:

[0133] flutsk(i,j) = flutsk(nmot,midmd) (1)

[0134] Furthermore, the number of samples N in each learning region is set; for each newly added valid sample, the sum is calculated according to the partition to which it belongs, and the region and Szone are then combined. k，1 Refresh (i, j)

[0135] At the same time, the partition sample counter CountZone(i,j) is incremented by 1 to obtain:

[0136] SZone k，1 (i, j) = SZone k-1，1 (i,j)+flutsk k (i, j) (2)

[0137] SZone k-1，1 (i, j) represents the value of the previous summation;

[0138] CountZone k，1 (i, j) = CountZone k-1，1 (i, j) + 1 (3)

[0139] CountZone k-1，1 (i, j) represents the previous count value;

[0140] The subscript k represents the cumulative number of times in the region, and the subscript 1 represents the fire mode flag signal evz_austot=1;

[0141] When the partition counter reaches the set number of samples N, i.e., CountZone N When (i, j) = N, the average value of the partition characteristic parameters can be obtained:

[0142] SZoneAve N，1 (i, j) = SZone N，1 (i,j) / N (4)

[0143] At the same time, the learning attribute A(i,j) of this partition is set to "learned", resulting in:

[0144] Initial value: A(i,j)=0 (5)

[0145] Regional learning: A(i,j)=1 (6)

[0146] This leads to the following: Figure 5 The processing results are shown.

[0147] The blank areas are inaccessible working conditions or areas not covered during test sampling.

[0148] Furthermore, the threshold self-learning value is multiplied by the coefficient factor to generate the target data as follows:

[0149] TAve1(i,j)=factor* SZoneAve N，1 (i, j), factor is the scaling factor (7)

[0150] Once all learnable regions have been learned, the partition threshold is filled into the pulse spectrum, and the function is learned. If there are operating regions that the engine has not reached or cannot reach, or regions that do not require diagnosis, the pulse spectrum is filled with interpolated values. If there are regions that do not require diagnosis, the pulse spectrum of that point is filled with the maximum value.

[0151] The effective data depends on the given rotational speed and load area, and is filtered along with the fire state data; element 999 of the test point sequence is constructed from rotational speed data 003 and load data 004.

[0152] like Figure 3 and Figure 1 As shown, the sample data includes test data from the preset operating point from idling condition to full load condition; the misfire test process includes a first initialization step 100 and a second data acquisition step 200; it may further include a third data filtering step 300 and a fourth pulse spectrum generation step 400.

[0153] The system can also set a first test flag to distinguish different misfire test processes; the system can use first operating condition data, namely speed operating condition data 2480 and load operating condition data 1465, to initialize the system processing process.

[0154] Furthermore, such as Figure 2 As shown, the coordinates of the first center point of the self-learning process data area are obtained; under the preset speed and load conditions, each learning area is numbered and identified based on the row 11, ii, mm data structure and the column 111, jjj, nnn data structure.

[0155] The matching quantity learning process obtains intermediate variables from the learning process to perform matching operations or attribute assignments; and then obtains self-learning sample data; this sample data records the process parameters corresponding to the test flags.

[0156] Specifically, such as Figure 7 The system shown can significantly shorten the time required for the traditional hub operation test, which takes several days. The traditional method involves collecting data from steady-state conditions point by point for offline analysis, which takes several days.

[0157] Furthermore, this method and related products can reduce the workload of data analysis, which can take several days, to zero. Relevant experimental data can be read online and analyzed in real time during the test, significantly improving efficiency.

[0158] 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 fire detection data, characterized in that, include: The process includes a first initialization step (100), a second data acquisition step (200), a third data filtering step (300), and a fourth pulse spectrum generation step (400). The first initialization step (100) divides the target operating condition area for diagnosis into zones based on engine speed nmot (003) and load state midmd (004). It sets the coordinates of the center point of the self-learning area and fills in the data using the center point as the area representative. The row coordinates of the center point are used to express different engine speeds nmot (003), and the column coordinates are used to express different load states midmd (004). The coordinates of the center point of the self-learning area are numbered. The coordinates of the center point of the self-learning area correspond to preset operating conditions, and the preset operating conditions correspond to specific engine speeds nmot (003) and specific load states midmd (004). The second data acquisition step (200) involves receiving an adaptive instruction code to activate the matching quantity self-learning process, cutting off the ignition or injection process, and implanting or simulating random and continuous misfire processes: so that during the combustion power stroke corresponding to each ignition phase, the engine misfires during the power stroke of the cut-off ignition or injection process; simultaneously acquiring and calculating the engine speed nmot (003), load state midmd (004), random misfire characteristic signal luts, continuous misfire characteristic signal flutsk (002), and misfire mode flag signal evz_austot (001) corresponding to each ignition during the random and continuous misfire processes, based on... Engine data is collected and stored at the engine ignition phase to form each set of data samples; for a preset speed, the center region of the feature point Zone(i,j) (999) is defined by the coordinates of the center point of the self-learning region; the load state midmd (004) of the engine is changed so that the load state midmd (004) covers all regions from idle state to full load state; each set of data samples includes the speed nmot (003), the load state midmd (004), the random misfire feature signal luts, the continuous misfire feature signal flutsk (002) and the misfire mode flag signal evz_austot (001). The third data filtering step (300) detects the fire failure process. If the fire failure process exists, the data sample at this time is recorded and stored as valid data. The average value of the feature parameters corresponding to the central region Zone(i,j) (999) of the feature point is output according to the valid data. The central region Zone(i,j) (999) of the feature point is determined according to the preset rotational speed and load. i is the number of the rotational speed partition in the central region Zone(i,j) (999), m is the number of the rotational speed partition m, and i varies from 1 to m. j is the number of the load partition in the central region Zone(i,j) (999), n is the number of the load partition, and j varies from 1 to n. The fourth pulse spectrum generation step (400) obtains the fire threshold corresponding to the average value, such that the fire threshold is equal to the scaling factor multiplied by the average value; the pulse spectrum is used to describe the fire threshold of the operating point corresponding to the fire occurrence. When the fire feature value exceeds the fire threshold, it is determined that a corresponding fire has occurred. By traversing the rotation speed nmot (003) and the load state midmd (004), the feature data of all the operating points corresponding to the central region Zone (i, j) (999) of the feature points are obtained, and the process of filling the pulse spectrum is completed.

2. The misfire detection data processing method as described in claim 1, wherein the second data acquisition step (200) further calculates the self-learning region to which each operating point belongs in real time and marks it with the feature point center region Zone(i,j) (999); the calculation results are assigned to the self-learning region corresponding one-to-one with the engine speed nmot (003) and the load state midmd (004); during this period, engine operating condition and state data are automatically acquired according to the ignition synchronization phase, and the data of the experienced operating points are calculated and analyzed in real time, and the data are filtered, classified and marked by partition; The matching quantity self-learning process selects data with implanted fire state from each group of data samples according to preset standards, that is, data group with fire mode flag signal evz_austot=1, to form valid data for calculating fire threshold, and discards invalid data; the standard includes the value or combination of values ​​of the data sample during the fire process; the combination of values ​​includes the rotation speed nmot (003), the load state midmd (004), the random fire characteristic signal luts, the continuous fire characteristic signal flutsk (002), and the fire mode flag signal evz_austot (001); the throttle opening is adjusted according to preset process, the processing of the second data acquisition step (200) is repeatedly executed and the valid data is stored or transmitted.

3. The fire detection data processing method as described in claim 1, wherein: The third data filtering step (300) also assigns the random fire feature signal luts(nmot,midmd) or continuous fire feature signal flutsk(nmot,midmd) of each group of data points of the valid data to the corresponding feature point center region Zone(i,j) (999) according to the rotation speed nmot (003) and the load state midmd (004), and stores them with numbers according to the rotation speed nmot (003) and the load state midmd (004); if the fire process does not exist, the data sample obtained in the second data acquisition step (200) is discarded.

4. The fire detection data processing method as described in claim 3, wherein: The third data filtering step (300) scans and continuously acquires k groups of valid data, where k is a positive integer; for each newly added group of valid data, the random fire feature signal luts(nmot,midmd) and the continuous fire feature signal flutsk(nmot,midmd) are summed and counted in the central region Zone(i,j) (999) to which the valid data belongs, respectively, to obtain the region and SZone. k，evz_austot (i, j) and the corresponding sample count value CountZone k，evz_austot (i, j); the subscript k represents the cumulative count of the region; MI represents the fire fault characteristic signal, including the random fire fault characteristic signal luts(nmot, midmd) and the continuous fire fault characteristic signal flutsk(nmot, midmd)), and the subscript evz_austot distinguishes the fire fault mode, such that: SZone k，evz_austot (i, j) = SZone k-1，evz_austot (i,j)+MI k，evz_austot (i, j), SZone k-1，evz_austot (i, j) represents the value of the previous summation; CountZone k，evz_austot (i, j) = CountZone k-1，evz_austot (i, j) + 1, CountZone k-1，evz_austot (i, j) represents the previous count value; When the number of valid data points reaches N (a positive integer), the learning process for the central region Zone(i,j)(999) of that feature point ends, and the average value of the feature parameters corresponding to the central region Zone(i,j)(999) of that feature point is output, denoted as SZoneAve. N，evz_austot (i, j), such that: SZoneAve N，evz_austot (i, j) = SZone N，evz_austot (i, j) / N; and such that: The learning attribute A(i,j) corresponding to the central region Zone(i,j) (999) of the feature point is set from 0 to 1.

5. The fire detection data processing method as described in claim 1, wherein, Data samples are extracted in real time from the engine control unit (ECU); the misfire detection data processing corresponds to the random misfire and / or the continuous misfire; the random misfire and the continuous misfire correspond to different misfire mode flag signals evz_austot (001); the second data acquisition step (200) is completed under free operating conditions, which traverse the effective range of the engine speed nmot (003) and the effective range of the load state midmd (004).

6. A test apparatus (600), comprising a first initialization unit (610), a second data acquisition unit (620), a third data filtering unit (630), and a fourth pulse spectrum generation unit (640); wherein, The first initialization unit (610) has a preset self-learning region center point coordinate data structure; the row coordinates of the data structure are used to express different rotational speeds nmot (003), and the column coordinates are used to express different load states midmd (004); the center point coordinates of the self-learning region are numbered; the center point coordinates of the self-learning region correspond to preset operating points, and the preset operating points correspond to specific rotational speeds nmot (003) and specific load states midmd (004); The second data acquisition unit (620) acquires and stores engine data based on the engine ignition phase to form each set of data samples; starting from the preset first speed, a misfire process is generated in the central region Zone (i, j) (999) of the feature points defined by the center point coordinates of the self-learning region; when the engine is in normal working condition, the controller actively cuts off the ignition or fuel injection process to implant or simulate random misfire and continuous misfire, and the engine misfires at the set time of cutting off the ignition or fuel injection. The load state midmd (004) of the engine is adjusted so that the load state midmd (004) covers all regions from idle state to full load state; each set of data samples includes engine speed nmot (003), load state midmd (004), random misfire characteristic signal luts, continuous misfire characteristic signal flutsk (002) and misfire mode flag signal evz_austot (001); The third data filtering unit (630) detects the fire failure process. If the fire failure process exists, it records and stores the data sample at this time as valid data, and outputs the average value of the feature parameters corresponding to the feature point center region Zone(i,j) (999) based on the valid data. The feature point center region Zone(i,j) (999) is determined according to the preset rotational speed and load. i is the number of the rotational speed partition in the feature point center region Zone(i,j) (999), m is the number of the rotational speed partition m, and i varies from 1 to m. j is the number of the load partition in the feature point center region Zone(i,j) (999), n is the number of the load partition, and j varies from 1 to n. The fourth pulse spectrum generation unit (640) obtains the fire threshold corresponding to the average value, such that the fire threshold is equal to the scaling factor multiplied by the average value; constructs a pulse spectrum to describe the fire threshold of the operating point corresponding to the fire occurrence. When the fire feature value exceeds the fire threshold, it is determined that a corresponding fire has occurred. By traversing the rotation speed nmot (003) and the load state midmd (004), it obtains the feature data of all the operating points corresponding to the central region Zone (i, j) (999) of the feature points, and completes the filling process of the pulse spectrum.

7. The test apparatus (600) as described in claim 6; wherein, The second data acquisition unit (620) also receives an adaptive instruction code to activate the matching quantity self-learning process. The matching quantity self-learning process selects valid data and discards invalid data in the data sample according to a preset standard. The standard includes the value or combination of values ​​of the data sample during the fire failure process. The combination of values ​​includes at least two of the rotational speed nmot (003), the load state midmd (004), the random fire failure characteristic signal luts, the continuous fire failure characteristic signal flutsk (002), and the fire failure mode flag signal evz_austot (001). The processing procedure of the second data acquisition unit (620) is repeatedly executed and valid data is stored or transmitted.

8. The test apparatus (600) as claimed in claim 6, wherein, The third data filtering unit (630) also assigns the random fire feature signal luts(nmot,midmd) or continuous fire feature signal flutsk(nmot,midmd) of each group of data points of the valid data to the corresponding feature point center region Zone(i,j) (999) according to the rotation speed nmot (003) and the load state midmd (004), and stores them with numbers according to the rotation speed nmot (003) and the load state midmd (004); if the fire process does not exist, the data sample obtained in the second data acquisition unit (620) is discarded.

9. The test apparatus (600) as claimed in claim 8, wherein, The third data filtering unit (630) scans and continuously accumulates k sets of valid data by region, where k is a positive integer; for each newly added set of valid data, the random fire feature signal luts(nmot,midmd) and the continuous fire feature signal flutsk(nmot,midmd) are summed and counted in the central region Zone(i,j) (999) to which the valid data belongs, respectively, to obtain the region and SZone. k，evz_austot (i, j) and the corresponding sample count value CountZone k，evz_austot (i, j), where the subscript k represents the cumulative number of fires in the region; MI represents the fire fault characteristic signal, including the random fire fault characteristic signal luts(nmot, midmd) and the continuous fire fault characteristic signal flutsk(nmot, midmd)), and the subscript evz_austot distinguishes the fire fault mode, such that: SZone k，evz_austot (i, j) = SZone k-1，evz_austot (i,j)+MI k，evz_austot (i, j), SZone k-1，evz_austot (i, j) represents the value of the previous summation; CountZone k，evz_austot (i, j) = CountZone k-1，evz_austot (i, j) + 1, CountZone k-1，evz_austot (i, j) represents the previous count value; When the number of valid data points reaches N (where N is a positive integer), the learning process for the feature point center region Zone(i,j)(999) ends, and the average value of the feature parameters corresponding to the feature point center region Zone(i,j)(999) is output, denoted as SZoneAve. N，evz_austot (i, j), such that: SZoneAve N，evz_austot (i, j) = SZone N，evz_austot (i, j) / N; and such that: The learning attribute A(i,j) corresponding to the central region Zone(i,j) (999) of the feature point is set from 0 to 1.

10. The test apparatus (600) as claimed in claim 6, wherein, Data samples are extracted in real time from the engine control unit (ECU); the random misfire and the continuous misfire correspond to different misfire mode flag signals evz_austot (001); the second data acquisition unit (620) is completed under free operating conditions, which traverse the engine speed range and load range; the test device (600) also includes a fifth human-machine interaction unit (666), which receives or displays the pulse spectrum to provide the test subjects or intermediate results that the test device (600) is currently or plans to perform.

11. 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 fire detection data processing method as described in claim 1.

12. An engine calibration device (800) comprising the test apparatus (600) as claimed in claim 6, and / or the computer storage medium (903) as claimed in claim 11, and / or an engine test module (700), wherein the engine test module (700) is an integrated unit or assembly system of the standardized test apparatus (600) and / or the computer storage medium (903); the engine test module (700) exchanges information with a vehicle or test equipment via an onboard bus or communication line.

Images