A method and system for identifying pilot-combustion and diagnosing the proportion of compression ignition of a jet flow ignition engine based on cylinder pressure signals

Abstract

This invention discloses a method and system for identifying ignition-compression ignition modes and diagnosing compression ignition ratios in a jet ignition engine based on cylinder pressure signals. The method includes: acquiring the original in-cylinder pressure signal and crankshaft angle signal within a single engine cycle; reconstructing the apparent heat release rate curve and cumulative heat release curve for the single cycle based on the cylinder pressure signal, crankshaft angle signal, and instantaneous cylinder volume; extracting multi-dimensional candidate features; performing correlation analysis to form a key feature vector; inputting the key feature vector into a first-stage classification model to identify the combustion mode of the current cycle; if identified as TJI mode, outputting the current identification result; if identified as TJACI mode, calculating the ratio of compression ignition heat release to total heat release; outputting the combustion mode identification result and compression ignition ratio for the current cycle, and returning to execute the synchronous acquisition step to achieve cycle-by-cycle online diagnosis. This invention achieves online combustion mode identification and real-time compression ignition ratio diagnosis.

Description

Technical Field

[0001] This invention relates to the field of engine combustion diagnosis and control, and in particular to a method and system for identifying ignition-compression ignition modes and diagnosing compression ignition ratios in jet ignition engines based on cylinder pressure signals. Background Technology

[0002] With increasingly stringent emission regulations, pre-combustion chamber jet ignition technology has become an important development direction for efficient and clean engines. Especially for low-carbon / zero-carbon fuel engines such as those using natural gas, methanol, and ammonia, which are difficult to ignite and have slow combustion rates, jet ignition technology achieves ultra-lean combustion by generating a high-speed turbulent jet after the rich mixture is ignited by a spark plug in the pre-combustion chamber, which is then injected into the main combustion chamber to ignite the mixture. Depending on the combustion mode of the mixture in the main combustion chamber, the combustion process of a pre-combustion chamber jet ignition engine mainly exhibits two modes.

[0003] The first type is the pre-combustion chamber turbulent jet ignition (TJI) mode. In this mode, the pre-combustion chamber jet enters the main combustion chamber and initiates flame propagation. The air-fuel mixture in the main combustion chamber completes combustion through flame front propagation. The heat release process in the TJI mode exhibits a single-peak characteristic, with a smooth and continuous heat release rate curve, a relatively mild peak pressure rise rate, and low combustion noise.

[0004] The second type is the Turbulent Jet Assisted Compression Ignition (TJACI) mode. In this mode, the pre-combustion chamber jet not only initiates initial flame propagation but also triggers volumetric auto-ignition (compression ignition) of the unburned mixture in the far region of the main combustion chamber through the high-temperature active free radicals carried by the jet and the local temperature rise effect. The exothermic process of the TJACI mode exhibits a significant bimodal characteristic: the first exothermic peak corresponds to the flame propagation stage induced by the jet, and the second exothermic peak corresponds to the compression auto-ignition stage. The TJACI mode can further improve thermal efficiency, but there is a risk of excessive pressure rise leading to combustion noise and mechanical load.

[0005] In TJACI mode, if the in-cylinder temperature is too low or the air-fuel mixture is too lean, the end mixture may fail to auto-ignite, and the combustion mode may degenerate into TJI mode. Alternatively, under the target operating conditions of TJI mode, if the intake air temperature is too high, the end mixture may undergo unexpected auto-ignition, and the actual combustion mode may switch to TJACI mode, leading to risks of excessive pressure rise and worsened combustion noise. Furthermore, even if TJACI mode is successfully achieved, the compression ignition ratio (i.e., the proportion of auto-ignition heat release to total heat release) directly affects the improvement in thermal efficiency and the level of mechanical load, requiring precise quantitative control.

[0006] In actual operation, transient changes in engine operating conditions can easily lead to unexpected switching between the two combustion modes. For example, under conditions targeting TJACI mode, low in-cylinder temperature or excessively lean mixture concentration may cause the end-of-combustion mixture to fail to auto-ignite, causing the combustion mode to degenerate into TJI mode. Conversely, under conditions targeting TJI mode, high intake air temperature may cause unexpected auto-ignition of the end-of-combustion mixture, causing actual combustion to switch to TJACI mode, which carries the risk of excessive pressure rise rate and deteriorated combustion noise. Furthermore, even if TJACI mode is successfully achieved, the compression ignition ratio (i.e., the proportion of auto-ignition heat release to total heat release) directly affects the improvement in thermal efficiency and mechanical load level, requiring precise quantitative control. However, current technology lacks an effective method for online identification of the two combustion modes and real-time diagnosis of the compression ignition ratio using cylinder pressure signals during engine operation.

[0007] Specifically, the shortcomings of existing technologies are mainly reflected in the following three aspects: First, traditional cylinder pressure thermodynamic analysis methods are difficult to directly apply to the complex combustion process of jet ignition in the pre-combustion chamber of an ammonia engine. Traditional spark-ignition engine exothermic analysis typically focuses on a single flame core and a single-stage flame propagation process. However, in the jet ignition process of an ammonia engine's pre-combustion chamber, the ignition source in the main combustion chamber is not a single central flame core, but rather a distributed ignition formed by multiple high-speed turbulent flame jets. Under certain operating conditions, after the initial jet ignition, the final mixture may further trigger auto-ignition, making the overall exothermic process exhibit two-stage or even multi-stage characteristics. If a simplified understanding based on single flame propagation is still used, the concentrated auto-ignition exothermic heat release in the later stages is easily submerged in the overall exothermic curve, making it difficult to clearly reveal the essential differences between the two modes.

[0008] Second, traditional single-feature cylinder pressure analysis methods are difficult to reliably distinguish between the two combustion modes. Existing combustion diagnostic methods based on cylinder pressure signals mostly use single features such as peak pressure, maximum pressure rise rate, CA50, or a single knock index for discrimination. These methods are suitable for relatively simple diagnostic tasks such as knock, misfire, or combustion stability. When the compression ignition ratio is low, the TJACI cycle may highly overlap with the TJI cycle in terms of maximum cylinder pressure, overall heat release peak, and some combustion phase parameters. Relying solely on a single threshold feature is prone to misjudgment and missed judgment.

[0009] Third, there is a lack of online quantification methods for the compression ignition ratio. Even if compression ignition can be observed, existing methods rely on high-precision heat release rate analysis and manual calibration of the inflection point, resulting in high computational complexity and failing to meet the requirements of cycle-by-cycle real-time control. For ammonia engines, if we can only know that "compression ignition has occurred" but cannot further quantify the proportion of compression ignition in the total heat release, it is still difficult to provide sufficient direct control basis for ignition timing adjustment, pre-combustion chamber supply adjustment, intake air temperature adjustment, and abnormal combustion suppression. Summary of the Invention

[0010] To address the technical problems existing in the prior art, this invention proposes a method and system for identifying ignition-compression ignition modes and diagnosing compression ignition ratios in jet ignition engines based on cylinder pressure signals. This method enables automatic identification of the two combustion modes, TJI and TJACI, in jet ignition engines and accurately calculates the compression ignition ratio in TJACI, providing key technical support for efficient and clean combustion control of engines.

[0011] On the one hand, to achieve the above objectives, the present invention provides a method for identifying the ignition-compression ignition mode and diagnosing the compression ignition ratio of a jet ignition engine based on cylinder pressure signals, comprising: Within the current engine working cycle, the original cylinder pressure signal and crankshaft angle signal within a single engine cycle are simultaneously acquired. The original pressure signal is preprocessed to obtain the preprocessed cylinder pressure signal. Based on the preprocessed cylinder pressure signal, the crankshaft angle signal, and the instantaneous cylinder volume, the first law of thermodynamics is used to perform inverse calculations to reconstruct the apparent heat release rate curve and the cumulative heat release curve of a single cycle. Through the apparent heat release rate curve and the cumulative heat release curve, multi-dimensional candidate features are extracted from three dimensions: combustion phase, pressure evolution, and heat release mode. Correlation analysis is performed on the multidimensional candidate features to screen out key features with low redundancy and high distinguishability of combustion modes, forming a key feature vector. The key feature vector is then input into the first-stage classification model to identify the combustion mode of the current cycle. The combustion modes include the pre-combustion chamber turbulent jet ignition (TJI) mode and the pre-combustion chamber turbulent jet assisted compression ignition (TJACI) mode. If the TJI mode is identified, the combustion mode identification result of the current cycle is directly output; if the TJACI mode is identified, the second stage is entered, and the ratio of compression ignition heat release to total heat release is calculated based on the apparent heat release rate curve. Output the combustion mode identification result and compression ignition ratio of the current cycle, and return to execute the synchronous acquisition step to achieve online diagnosis cycle by cycle.

[0012] Preferably, the apparent heat release rate curve is reconstructed based on the instantaneous apparent heat release rate, wherein the instantaneous apparent heat release rate is: ; In the formula, Apparent heat release represents the total heat released during fuel combustion; The crankshaft rotation angle signal is represented by γ, which is the polytropic index. This refers to the instantaneous pressure inside the cylinder, which varies with the crankshaft angle. This represents the rate of change of cylinder volume with crankshaft angle. This refers to the instantaneous volume of the cylinder. This is the first derivative of cylinder pressure with respect to crankshaft rotation angle; This is a heat transfer correction term, representing the instantaneous heat loss rate per unit crankshaft angle; This refers to heat loss. This refers to the cylinder volume; This refers to the cylinder pressure.

[0013] Preferably, the instantaneous volume of the cylinder Calculated based on the following formula: ; In the formula, This refers to the clearance volume of the combustion chamber. The cylinder bore is... Where is the crankshaft radius. This is the length of the connecting rod.

[0014] Preferably, the multidimensional candidate features include combustion phase features, pressure evolution features, and exothermic morphology features; The combustion phase characteristics include: the crankshaft angle, combustion center of gravity, combustion endpoint, and combustion duration corresponding to the cumulative heat release reaching a preset proportion of the total heat release; The pressure evolution characteristics include: maximum cylinder pressure, crankshaft angle corresponding to maximum cylinder pressure, and maximum pressure rise rate; The heat release morphology features include: maximum apparent heat release rate, crankshaft angle corresponding to maximum apparent heat release rate, amplitude of the first heat release peak, amplitude of the second heat release peak, intensity ratio of the two peaks, interval between the two peaks, amplitude of the valley value between the peaks, and the heat release initiation angle of the second stage.

[0015] Preferably, correlation analysis is performed on the multidimensional candidate features, including: Calculate the Pearson correlation coefficient between each pair of the multidimensional candidate features, construct the correlation coefficient matrix, and draw the correlation heatmap; When the absolute value of the Pearson correlation coefficient between any two candidate features is greater than a preset threshold, the feature that is more sensitive to the distinction of combustion modes or has a clearer physical meaning is retained, and the other feature is deleted to form the key feature vector used for combustion mode identification in the first stage.

[0016] Preferably, the classification model in the first stage is a lightweight gradient boosting decision tree model.

[0017] Preferably, calculating the ratio of compression ignition heat release to total heat release based on the apparent heat release rate curve includes: Peak detection is performed on the apparent heat release rate curve to identify the first and second heat release peaks; Locate the local valley point between the first exothermic peak and the second exothermic peak, and use the crankshaft angle corresponding to the local valley point as the dividing point between ignition exothermic and compression exothermic. Based on the aforementioned dividing point, the positive heat release region of the compression ignition stage is integrated to obtain the heat release of compression ignition, and the positive heat release region of the entire combustion stage is integrated to obtain the total heat release. The compression ignition ratio is calculated by comparing the heat released from compression ignition with the total heat released.

[0018] Preferably, the compression ignition ratio is calculated as follows: η CI = Q CI / Q tot ×100%; ; In the formula, The crankshaft angle corresponding to the peak-to-valley value; The combustion initiation angle; The combustion end angle; The heat released by compression ignition; This represents the total heat release; This refers to the apparent heat release. This is the endpoint angle of the cylinder heat release integral.

[0019] Preferably, calculating the compression ignition ratio further includes: If the second exothermic peak does not reach the preset minimum peak value threshold, minimum peak interval threshold, or minimum area threshold, the combustion mode of the current cycle will be corrected to TJI mode.

[0020] On the other hand, to achieve the above objectives, the present invention also provides a jet ignition engine ignition-compression ignition mode recognition and compression ignition ratio diagnosis system based on cylinder pressure signals, comprising: The hardware acquisition and signal conditioning unit is used to simultaneously acquire the original in-cylinder pressure signal and crankshaft angle signal within a single engine cycle during the current engine working cycle. The diagnostic control unit is used to preprocess the original pressure signal to obtain the preprocessed cylinder pressure signal, reconstruct the apparent heat release rate curve and cumulative heat release curve of a single cycle, extract multi-dimensional candidate features through the apparent heat release rate curve and cumulative heat release curve, perform correlation analysis on the multi-dimensional candidate features, screen out key features with low redundancy and high distinguishability of combustion mode, form a key feature vector, and input the key feature vector into the first-stage classification model to identify the combustion mode of the current cycle. The control output and interaction unit is used to identify the combustion mode. If it is identified as TJI mode, it directly outputs the combustion mode identification result of the current cycle. If it is identified as TJACI mode, it enters the second stage and calculates the ratio of compression ignition heat release to total heat release based on the apparent heat release rate curve. The hardware acquisition and signal conditioning unit includes a cylinder pressure sensor for acquiring raw cylinder pressure signals, a crankshaft angle sensor for providing crankshaft angle phase reference, a combustion analyzer, and an engine electronic control module. The diagnostic control unit includes: The signal preprocessing module is used to preprocess the raw pressure signal; The thermodynamic inverse operation module is used to reconstruct the apparent heat release rate curve and cumulative heat release curve of a single cycle; The feature extraction and filtering module is used to extract multidimensional candidate features and perform correlation analysis to filter out key feature vectors. The combustion mode recognition module has a built-in first-stage classification model to identify the combustion mode of the current cycle. The control output and interaction unit includes: The compression ignition ratio diagnostic module is used to calculate the compression ignition ratio based on the apparent heat release rate curve when the TJACI mode is identified. The result output module is used to output the combustion mode recognition result and compression ignition ratio to the engine electronic control module to form a closed-loop control.

[0021] Compared with the prior art, the present invention has the following advantages and technical effects: (1) This invention realizes cycle-by-cycle online identification of TJI and TJACI modes under jet ignition conditions in the pre-combustion chamber of an ammonia engine, enabling combustion control to no longer rely solely on preset operating conditions and offline calibration parameters, but to make feedback judgments based on the actual exothermic process occurring in the cylinder. Existing research on pre-combustion chamber ignition in ammonia engines has shown that under different compression ratios and equivalence ratios, the main combustion chamber may change from a single-peak flame propagation to a double-peak exothermic process with a significant second peak, indicating that online mode identification has direct control significance.

[0022] (2) This invention improves the ability to distinguish between TJI and TJACI modes under low-pressure combustion ratio boundary conditions by combining thermodynamic inverse operation, key feature extraction and heat map screening, and overcomes the problems of easy misjudgment and easy omission of traditional single feature threshold method.

[0023] (3) The present invention uses a two-stage architecture to complete the combustion mode identification in the first stage and perform compression ignition ratio calculation on the cycle identified as TJACI in the second stage, which shortens the single cycle calculation chain and has the ability to output the combustion mode and compression ignition ratio in real time for each cycle.

[0024] (4) The present invention realizes the quantification of compression ignition ratio based on the stage decomposition of the apparent heat release rate curve, which can provide a more direct basis for ignition timing adjustment, pre-combustion chamber supply adjustment and abnormal combustion suppression. Attached Figure Description

[0025] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of a jet ignition engine ignition-compression ignition mode identification and compression ignition ratio diagnosis method based on cylinder pressure signal according to an embodiment of the present invention. Figure 2 This is a flowchart of a two-stage online diagnostic process based on cylinder pressure signals according to an embodiment of the present invention; Figure 3 This is a flowchart illustrating the workflow of a jet ignition engine ignition-compression ignition mode recognition and compression ignition ratio diagnosis system based on cylinder pressure signals, according to an embodiment of the present invention. Detailed Implementation

[0026] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0027] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0028] This embodiment proposes a method for identifying the ignition-compression ignition mode and diagnosing the compression ignition ratio in a jet ignition engine based on cylinder pressure signals. Figure 1 ,include: Within the current engine working cycle, the original cylinder pressure signal and crankshaft angle signal within a single engine cycle are simultaneously acquired. The original pressure signal is preprocessed to obtain the preprocessed cylinder pressure signal. Based on the preprocessed cylinder pressure signal, the crankshaft angle signal, and the instantaneous cylinder volume, the first law of thermodynamics is used to perform inverse calculations to reconstruct the apparent heat release rate curve and the cumulative heat release curve of a single cycle. Through the apparent heat release rate curve and the cumulative heat release curve, multi-dimensional candidate features are extracted from three dimensions: combustion phase, pressure evolution, and heat release mode. Correlation analysis is performed on the multidimensional candidate features to screen out key features with low redundancy and high distinguishability of combustion modes, forming a key feature vector. The key feature vector is then input into the first-stage classification model to identify the combustion mode of the current cycle. The combustion modes include the pre-combustion chamber turbulent jet ignition (TJI) mode and the pre-combustion chamber turbulent jet assisted compression ignition (TJACI) mode. If the TJI mode is identified, the combustion mode identification result of the current cycle is directly output; if the TJACI mode is identified, the second stage is entered, and the ratio of compression ignition heat release to total heat release is calculated based on the apparent heat release rate curve. Output the combustion mode identification result and compression ignition ratio of the current cycle, and return to execute the synchronous acquisition step to achieve online diagnosis cycle by cycle.

[0029] Specifically, within each engine working cycle, the raw in-cylinder pressure signal Praw(θ) and the crankshaft angle signal θ are synchronously acquired. First, the raw pressure signal is zero-drift corrected and reference pressure calibrated to eliminate the influence of sensor drift and installation errors; then, cyclic registration and smoothing filtering are performed to obtain the low-frequency indicated pressure component Plow(θ) for thermodynamic analysis.

[0030] Furthermore, the calculation of the apparent heat release rate (AHRR) is a core step based on the inverse operation of the in-cylinder combustion process using the first law of thermodynamics. Treating the working fluid in the cylinder as an ideal gas and ignoring the effects of in-cylinder gas leakage and heat transfer losses from the cylinder walls, the instantaneous apparent heat release rate is: ; In the formula, Apparent heat release represents the total heat released during fuel combustion; The crankshaft rotation angle signal is represented by γ, which is the polytropic index. This refers to the instantaneous pressure inside the cylinder, which varies with the crankshaft angle. This represents the rate of change of cylinder volume with crankshaft angle. This refers to the instantaneous volume of the cylinder. This is the first derivative of cylinder pressure with respect to crankshaft rotation angle; This is a heat transfer correction term, representing the instantaneous heat loss rate per unit crankshaft angle; This refers to heat loss. This refers to the cylinder volume; This refers to the cylinder pressure.

[0031] cylinder instantaneous volume The calculations are based on the geometric relationships of the crank-connecting rod mechanism: ; In the formula, This refers to the clearance volume of the combustion chamber. The cylinder bore is... Where is the crankshaft radius. This is the length of the connecting rod.

[0032] Furthermore, the multidimensional candidate features include combustion phase features, pressure evolution features, and exothermic morphology features; The combustion phase characteristics include: combustion start point (crankshaft angle corresponding to 10% of the total heat release), combustion center of gravity, combustion end point (characterizing the burnout rate in the later stage of combustion), and combustion duration (characterizing the overall combustion rate). The pressure evolution characteristics include: maximum cylinder pressure Pmax, crankshaft angle θPmax corresponding to maximum cylinder pressure, and maximum pressure rise rate (dp / dθ)max; these characteristics reflect the pressure build-up process and combustion intensity in the main combustion chamber. The heat release morphology features include: maximum apparent heat release rate HRRmax, crankshaft angle θHRR,max corresponding to the maximum apparent heat release rate, amplitude of the first heat release peak HRR1, amplitude of the second heat release peak HRR2, bipeak intensity ratio Rpeak=HRR2 / HRR1, interval between the two peaks Δθpeak, amplitude of the valley between the peaks HRRvalley, and the second-stage heat release initiation angle θCI,start; among them, the bipeak intensity ratio, the interval between the two peaks, and the second-stage heat release initiation angle are used to directly characterize whether there is a second-stage concentrated spontaneous combustion heat release, which are key features to distinguish TJI from TJACI.

[0033] Furthermore, correlation analysis is performed on the multidimensional candidate features, including: Calculate the Pearson correlation coefficient between each pair of the multidimensional candidate features, construct the correlation coefficient matrix, and draw the correlation heatmap; When the absolute value of the Pearson correlation coefficient between any two candidate features is greater than a preset threshold, the feature that is more sensitive to the distinction of combustion modes or has a clearer physical meaning is retained, and the other feature is deleted to form the key feature vector used for combustion mode identification in the first stage.

[0034] Specifically, heatmap analysis visualizes the linear correlation between features by calculating the Pearson correlation coefficient, thereby selecting key features with low redundancy and high discriminative power. The specific steps are as follows: ; In the formula, For the first i The variable and the first j Pearson correlation coefficients of the variables; n The number of samples; k For the first k One observation value; For the first i The first variable k One observation value; For the first i The average of the variables; For the first j The first variable k One observation value; For the first jThe average of the variables.

[0035] Create a correlation heatmap: use color intensity to represent the magnitude of the correlation coefficient, and label the specific correlation coefficient values ​​on the heatmap. If the absolute value of the correlation coefficient between two features is higher than a preset threshold, preferably 0.85 or 0.90, it is determined that there is significant collinearity, and only the feature with more explicit physical meaning or higher relevance to the target label is retained.

[0036] Furthermore, the two-stage identification architecture includes: Phase 1: Combustion Mode Identification The filtered core feature vectors are input into a lightweight classification model, preferably using a gradient boosting decision tree model, to determine whether the current cycle belongs to the pre-combustion chamber turbulent jet ignition mode or the pre-combustion chamber turbulent jet assisted compression ignition mode. The input of the first stage is the core feature vector F, and the output is the mode label Cycle_Type of the current cycle and the corresponding confidence score.

[0037] Phase Two, Compression Ignition Ratio Diagnosis: When the first stage output is TJI, the system directly outputs the current combustion mode.

[0038] When the output of the first stage is TJACI, the second stage compression ignition ratio diagnostic process begins. The second stage first performs peak detection on the apparent heat release rate curve to identify the ignition heat release peak in the first stage and the concentrated heat release peak in the second stage. Then, it locates the local valley point between the two peaks and uses this valley point as the boundary between ignition heat release and compression ignition heat release. Afterwards, it integrates the positive heat release regions of the compression ignition stage and the entire combustion stage to obtain the compression ignition heat release QCI and the total heat release Qtot.

[0039] Compression ignition ratio is calculated using the following formula: ηCI = QCI / Qtot × 100%, where, ; In the formula, The crankshaft angle corresponding to the peak-to-valley value; The combustion initiation angle; The combustion end angle; The heat released by compression ignition; This represents the total heat release; This refers to the apparent heat release. This is the endpoint angle of the cylinder heat release integral.

[0040] Furthermore, calculating the compression ignition ratio also includes: If the second exothermic peak does not reach the preset minimum peak value threshold, minimum peak interval threshold, or minimum area threshold, the combustion mode of the current cycle will be corrected to TJI mode.

[0041] The system outputs the combustion mode identification results and compression ignition ratio diagnosis results of the current cycle to the engine electronic control module in real time for closed-loop correction of subsequent ignition timing, pre-combustion chamber supply, injection parameters and intake conditions.

[0042] This embodiment also provides a jet ignition engine ignition-compression ignition mode recognition and compression ignition ratio diagnosis system based on cylinder pressure signals, the workflow of which is as follows: Figure 3 ,include: The hardware acquisition and signal conditioning unit is used to simultaneously acquire the original in-cylinder pressure signal and crankshaft angle signal within a single engine cycle during the current engine working cycle. The diagnostic control unit is used to preprocess the original pressure signal to obtain the preprocessed cylinder pressure signal, reconstruct the apparent heat release rate curve and cumulative heat release curve of a single cycle, extract multi-dimensional candidate features through the apparent heat release rate curve and cumulative heat release curve, perform correlation analysis on the multi-dimensional candidate features, screen out key features with low redundancy and high distinguishability of combustion mode, form a key feature vector, and input the key feature vector into the first-stage classification model to identify the combustion mode of the current cycle. The control output and interaction unit is used to identify the combustion mode. If it is identified as TJI mode, it directly outputs the combustion mode identification result of the current cycle. If it is identified as TJACI mode, it enters the second stage and calculates the ratio of compression ignition heat release to total heat release based on the apparent heat release rate curve. The hardware acquisition and signal conditioning unit includes a cylinder pressure sensor for acquiring raw cylinder pressure signals, a crankshaft angle sensor for providing crankshaft angle phase reference, a combustion analyzer, and an engine electronic control module. The diagnostic control unit includes: The signal preprocessing module is used to preprocess the raw pressure signal; The thermodynamic inverse operation module is used to reconstruct the apparent heat release rate curve and cumulative heat release curve of a single cycle; The feature extraction and filtering module is used to extract multidimensional candidate features and perform correlation analysis to filter out key feature vectors. The combustion mode recognition module has a built-in first-stage classification model to identify the combustion mode of the current cycle. The control output and interaction unit includes: The compression ignition ratio diagnostic module is used to calculate the compression ignition ratio based on the apparent heat release rate curve when the TJACI mode is identified. The result output module is used to output the combustion mode recognition result and compression ignition ratio to the engine electronic control module to form a closed-loop control.

[0043] Specifically, the cylinder pressure sensor is installed at the pressure measurement hole in the engine cylinder head to collect the original pressure signal in the cylinder during a single cycle in real time. A crankshaft angle sensor, installed at the front end of the engine crankshaft, is used to provide a crankshaft angle phase reference, and its resolution is preferably not less than 0.1°CA; Combustion analyzer: Electrically connected to cylinder pressure sensor and crankshaft angle sensor, used to amplify cylinder pressure signal, perform analog-to-digital conversion, synchronous triggering, and buffer raw data; The engine electronic control module (ECU) communicates with the combustion analyzer and diagnostic control unit to receive combustion mode identification results and compression ignition ratio results, and accordingly executes corrections to ignition, injection, and related control strategies.

[0044] To more clearly illustrate the technical solution of the present invention, specific embodiments are provided below for description: This embodiment details the online processing flow of the ignition-compression ignition mode recognition and compression ignition ratio diagnosis method, and its logic follows... Figure 2 The two-stage online diagnostic flowchart is explained in detail below: Step S101: Synchronous signal acquisition.

[0045] During each engine working cycle, the raw cylinder pressure signal P is synchronously acquired. raw (θ) and crankshaft angle signal θ. The cylinder pressure signal is acquired by a cylinder pressure sensor, and the crankshaft angle signal is acquired by a crankshaft angle sensor. The two are triggered synchronously to ensure that the cylinder pressure change corresponds one-to-one with the crankshaft phase.

[0046] Step S102: Preprocessing of cylinder pressure signal.

[0047] The raw cylinder pressure signal acquired in step S101 is preprocessed, including reference pressure calibration, cyclic registration, and smoothing filtering. If zero drift exists in the sensor, further zero drift correction is performed to obtain the preprocessed cylinder pressure signal for thermodynamic analysis. This step aims to reduce the impact of measurement errors and random noise on the subsequent thermodynamic inverse calculation results.

[0048] Step S103: Thermodynamic inverse operation and reconstruction of heat release rate.

[0049] Based on the preprocessed cylinder pressure signal obtained in step S102, and combined with the instantaneous cylinder volume and the first law of thermodynamics, the apparent heat release rate curve and cumulative heat release curve of a single cycle are calculated.

[0050] Preferably, combustion-related parameters are further calculated, including combustion start point, combustion center of gravity, combustion end point, combustion duration, and maximum cylinder pressure P. max The crankshaft rotation angle θ corresponding to the highest cylinder pressure Pmax Maximum pressure rise rate (dp / dθ)max and maximum apparent heat release rate HRR max .

[0051] Step S104: Candidate feature extraction.

[0052] Based on the apparent heat release rate curve, cumulative heat release curve and cylinder pressure curve obtained in step S103, candidate features are extracted from three dimensions: combustion phase, pressure evolution and heat release morphology.

[0053] Combustion phase characteristics include the combustion start point (crankshaft angle corresponding to 10% of the total heat release), combustion center of gravity, combustion end point (characterizing the burnout rate in the later stage of combustion), and combustion duration (characterizing the overall combustion rate). Pressure evolution characteristics include the highest cylinder pressure P max The crankshaft rotation angle θ corresponding to the highest cylinder pressure Pmax Maximum pressure rise rate (dp / dθ) max ; Exothermic morphological characteristics include maximum apparent heat release rate (HRR). max The maximum apparent heat release rate corresponds to the crankshaft angle θ HRR,max The amplitude of the first exothermic peak (HRR1), the amplitude of the second exothermic peak (HRR2), and the intensity ratio of the two peaks (R) peak =HRR2 / HRR1, the interval between the two peaks Δθ peak Peak-to-valley amplitude HRR valley and the second stage exothermic initiation angle θ CI,start .

[0054] Step S105: Key feature screening.

[0055] Calculate the Pearson correlation coefficient for each pair of candidate features extracted in step S104, construct a correlation coefficient matrix and draw a heatmap; when the absolute value of the correlation coefficient between any two candidate features is greater than a preset threshold, delete the features with high redundancy and retain the features with clearer physical meaning and more sensitive to the distinction of combustion modes to form the key feature vector F used for the first stage of identification.

[0056] The key feature vector F includes the combustion center of gravity, combustion duration, and maximum cylinder pressure P. max Maximum pressure rise rate (dp / dθ) max Maximum apparent heat release rate (HRR) max Bimodal intensity ratio R peak The interval between the two peaks Δθ peak and the second stage exothermic initiation angle θ CI,start .

[0057] Step S106: First stage combustion mode identification.

[0058] The key feature vector F obtained in step S105 is input into the first-stage classification model to determine whether the current cycle belongs to the pre-combustion chamber turbulent jet ignition mode or the pre-combustion chamber turbulent jet assisted compression ignition mode.

[0059] The classification model is a lightweight classification model. This stage outputs the current loop's pattern label Cycle_Type and its corresponding confidence score; when Cycle_Type is TJI, proceed to step S108; when Cycle_Type is TJACI, proceed to step S107.

[0060] Step S107: Second stage compression ignition ratio diagnosis.

[0061] For cycles identified as TJACI, bimodal detection is performed based on the apparent heat release rate curve obtained in step S103 to identify the first-stage ignition heat release peak and the second-stage concentrated heat release peak; subsequently, the local valley point θ between the two peaks is located. v The valley point is used as the boundary between ignition heat release and compression heat release; then, the positive heat release regions of the compression ignition stage and the entire combustion stage are integrated to obtain the compression ignition heat release Q. CI Total heat release Q tot And calculate the compression ignition ratio η according to the following formula. CI η CI =Q CI / Q tot ×100%.

[0062] Step S108: Result output and feedback.

[0063] Output the combustion mode identification result Cycle_Type and compression ignition ratio η for the current cycle. CI Specifically, when step S106 determines that the current cycle is TJI, the combustion mode is directly output; after step S107 completes the compression ignition ratio calculation, the corresponding η is output. CI The results can be sent to the engine electronic control module in real time for closed-loop correction of subsequent ignition timing, pre-combustion chamber supply parameters, and abnormal combustion suppression strategies.

[0064] Step S109: Execute repeatedly.

[0065] The control process returns to step S101, and continues to acquire cylinder pressure signals, extract features, recognize patterns, and calculate compression ignition ratios for the next working cycle, thereby achieving cycle-by-cycle online diagnostics.

[0066] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for identifying pilot- compression ignition mode and diagnosing compression ratio of a jet ignition engine based on cylinder pressure signal, characterized in that, include: Within the current engine working cycle, the original cylinder pressure signal and crankshaft angle signal within a single engine cycle are simultaneously acquired. The original pressure signal is preprocessed to obtain the preprocessed cylinder pressure signal. Based on the preprocessed cylinder pressure signal, the crankshaft angle signal, and the instantaneous cylinder volume, the first law of thermodynamics is used to perform inverse calculations to reconstruct the apparent heat release rate curve and the cumulative heat release curve of a single cycle. Through the apparent heat release rate curve and the cumulative heat release curve, multi-dimensional candidate features are extracted from three dimensions: combustion phase, pressure evolution, and heat release mode. Correlation analysis is performed on the multidimensional candidate features to screen out key features with low redundancy and high distinguishability of combustion modes, forming a key feature vector. The key feature vector is then input into the first-stage classification model to identify the combustion mode of the current cycle. The combustion modes include the pre-combustion chamber turbulent jet ignition (TJI) mode and the pre-combustion chamber turbulent jet assisted compression ignition (TJACI) mode. If it is identified as TJI mode, the combustion mode identification result of the current cycle will be output directly; If the TJACI mode is identified, the second stage is entered, and the ratio of the heat release from compression ignition to the total heat release is calculated based on the apparent heat release rate curve. Output the combustion mode identification result and compression ignition ratio of the current cycle, and return to execute the synchronous acquisition step to achieve online diagnosis cycle by cycle.

2. The method of claim 1, wherein, The apparent heat release rate curve is reconstructed based on the instantaneous apparent heat release rate, wherein the instantaneous apparent heat release rate is: ； In the formula, Apparent heat release represents the total heat released during fuel combustion; The crankshaft rotation angle signal is represented by γ, which is the polytropic index. This refers to the instantaneous pressure inside the cylinder, which varies with the crankshaft angle. This represents the rate of change of cylinder volume with crankshaft angle. This refers to the instantaneous volume of the cylinder. This is the first derivative of cylinder pressure with respect to crankshaft rotation angle; This is a heat transfer correction term, representing the instantaneous heat loss rate per unit crankshaft angle; This refers to heat loss. This refers to the cylinder volume; This refers to the cylinder pressure.

3. The method of claim 2, wherein, the instantaneous volume of the cylinder is calculated based on the following equation: ； In the formula, This refers to the clearance volume of the combustion chamber. The cylinder bore is... Where is the crankshaft radius. This is the length of the connecting rod.

4. The method of claim 1, wherein, The multidimensional candidate features include combustion phase features, pressure evolution features, and exothermic morphology features; The combustion phase characteristics include: the crankshaft angle, combustion center of gravity, combustion endpoint, and combustion duration corresponding to the cumulative heat release reaching a preset proportion of the total heat release; The pressure evolution characteristics include: maximum cylinder pressure, crankshaft angle corresponding to maximum cylinder pressure, and maximum pressure rise rate; The heat release morphology features include: maximum apparent heat release rate, crankshaft angle corresponding to maximum apparent heat release rate, amplitude of the first heat release peak, amplitude of the second heat release peak, intensity ratio of the two peaks, interval between the two peaks, amplitude of the valley value between the peaks, and the heat release initiation angle of the second stage.

5. The method of claim 4, wherein, The correlation analysis of the multidimensional candidate features includes: Calculate the Pearson correlation coefficient between each pair of the multidimensional candidate features, construct the correlation coefficient matrix, and draw the correlation heatmap; When the absolute value of the Pearson correlation coefficient between any two candidate features is greater than a preset threshold, the feature that is more sensitive to the distinction of combustion modes or has a clearer physical meaning is retained, and the other feature is deleted to form the key feature vector used for combustion mode identification in the first stage.

6. The method of claim 1, wherein, The first-stage classification model is a lightweight gradient boosting decision tree model.

7. The method of claim 1, wherein, The ratio of the heat released by compression ignition to the total heat released is calculated based on the apparent heat release rate curve, including: Peak detection is performed on the apparent heat release rate curve to identify the first and second heat release peaks; Locate the local valley point between the first exothermic peak and the second exothermic peak, and use the crankshaft angle corresponding to the local valley point as the dividing point between ignition exothermic and compression exothermic. Based on the aforementioned dividing point, the positive heat release region of the compression ignition stage is integrated to obtain the heat release of compression ignition, and the positive heat release region of the entire combustion stage is integrated to obtain the total heat release. The compression ignition ratio is calculated by comparing the heat released from compression ignition with the total heat released.

8. The method according to claim 7, characterized in that, The compression ignition ratio is calculated as follows: η CI = Q CI / Q tot × 100%; ； In the formula, The crankshaft angle corresponding to the peak-to-valley value; The combustion initiation angle; The combustion end angle; The heat released during compression ignition; This represents the total heat release; This refers to the apparent heat release. This is the endpoint angle of the cylinder heat release integral.

9. The method of claim 8, wherein, Calculating the compression ignition ratio also includes: If the second exothermic peak does not reach the preset minimum peak value threshold, minimum peak interval threshold, or minimum area threshold, the combustion mode of the current cycle will be corrected to TJI mode.

10. A jet ignition engine ignition-compression ignition mode recognition and compression ignition ratio diagnosis system based on cylinder pressure signal, used to execute the method according to any one of claims 1-9, characterized in that, include: The hardware acquisition and signal conditioning unit is used to simultaneously acquire the original in-cylinder pressure signal and crankshaft angle signal within a single engine cycle during the current engine working cycle. The diagnostic control unit is used to preprocess the original pressure signal to obtain the preprocessed cylinder pressure signal, reconstruct the apparent heat release rate curve and cumulative heat release curve of a single cycle, extract multi-dimensional candidate features through the apparent heat release rate curve and cumulative heat release curve, perform correlation analysis on the multi-dimensional candidate features, screen out key features with low redundancy and high distinguishability of combustion mode, form a key feature vector, and input the key feature vector into the first-stage classification model to identify the combustion mode of the current cycle. The control output and interaction unit is used to identify the combustion mode. If it is identified as TJI mode, it directly outputs the combustion mode identification result of the current cycle. If it is identified as TJACI mode, it enters the second stage and calculates the ratio of compression ignition heat release to total heat release based on the apparent heat release rate curve. The hardware acquisition and signal conditioning unit includes a cylinder pressure sensor for acquiring raw cylinder pressure signals, a crankshaft angle sensor for providing crankshaft angle phase reference, a combustion analyzer, and an engine electronic control module. The diagnostic control unit includes: The signal preprocessing module is used to preprocess the raw pressure signal; The thermodynamic inverse operation module is used to reconstruct the apparent heat release rate curve and cumulative heat release curve of a single cycle; The feature extraction and filtering module is used to extract multidimensional candidate features and perform correlation analysis to filter out key feature vectors. The combustion mode recognition module has a built-in first-stage classification model to identify the combustion mode of the current cycle. The control output and interaction unit includes: The compression ignition ratio diagnostic module is used to calculate the compression ignition ratio based on the apparent heat release rate curve when the TJACI mode is identified. The result output module is used to output the combustion mode recognition result and compression ignition ratio to the engine electronic control module to form a closed-loop control.

Images