Engine supercharging pressure frequency domain identification self-learning compensation and fault handling method

Abstract

The application discloses an engine supercharging pressure frequency domain identification self-learning compensation and fault processing method. By judging the steady state condition of the engine, transient interference is avoided from the source, the supercharging pressure difference signal is collected and frequency domain transformation is performed, the inherent frequency of the supercharging system and the corresponding fluctuation amplitude are accurately extracted, a compensation signal in the opposite phase is generated based on the identification result and is superimposed to the throttle basic control signal, and the supercharging pressure resonance fluctuation is inhibited from the root cause. Meanwhile, the compensation effect and the identification effect feedback parameters are collected, and the three core control parameters are self-learned and updated. The method does not require additional hardware cost, a closed-loop control system of "identification-compensation-feedback-self-learning" is constructed, the working condition change and component aging of the engine in the whole life cycle can be adaptively adapted, the optimal supercharging control precision and engine running stability can be long-term maintained, and the fuel economy is effectively improved.

Description

Technical Field

[0001] This application relates to the field of engine control technology, and more specifically, to a method for self-learning compensation and fault handling of engine boost pressure frequency domain identification. Background Technology

[0002] In response to engine intake boost and torque increase requests, the turbocharging system recovers exhaust energy to achieve intake boost; its control precision directly determines the engine's power and fuel economy. Current technologies for detecting and handling turbocharging system anomalies have the following significant shortcomings:

[0003] 1. Existing slow response diagnostic methods for actuators determine the actuator responsiveness under dynamic operating conditions by calculating the rate of change of time integral. They use a general fault detection threshold and do not make specific corrections for different operating conditions, resulting in low diagnostic accuracy and inability to accurately identify abnormal pressure fluctuations caused by the inherent resonance of the booster system.

[0004] 2. Existing diagnostic methods for valve components require the additional configuration of an electric supercharger to provide compressed air to simulate operating conditions, which increases hardware costs and system complexity, and cannot be directly applied to pressure fluctuation detection in conventional turbocharged systems.

[0005] 3. Although the existing turbocharger closed-loop control method can control the duty cycle signal by outputting the exhaust gas bypass valve to control the actual boost pressure to follow the target boost pressure, and update the integral term of the PID control through machine learning to ensure the stability of the basic control, it does not consider the inherent resonance fluctuations caused by the coupling of the turbocharger system pipelines and components. Moreover, the relevant parameters for fluctuation suppression are fixed, and it cannot adapt to the changes in operating conditions and component aging throughout the engine's entire life cycle. The control effect gradually decays with the use time, and it does not achieve integrated detection and post-processing of low boost pressure faults and abnormal resonance fluctuations. Summary of the Invention

[0006] The purpose of this application is to overcome the shortcomings of the prior art and provide a method for frequency domain identification, self-learning compensation, and fault handling of engine boost pressure. This method requires no additional hardware costs, accurately identifies the inherent resonance characteristics of the boost system through frequency domain analysis, achieves adaptive compensation control by combining full-parameter closed-loop self-learning, and integrates graded post-processing for low boost pressure faults, comprehensively improving the control accuracy, adaptability, and reliability of the boost system. To achieve the above objectives, this application adopts the following technical solution.

[0007] In a first aspect, embodiments of this application provide a method for self-learning compensation and fault handling of engine boost pressure frequency domain identification, comprising the following steps:

[0008] S1. Determine whether the engine is in a steady-state condition. If it is in a steady-state condition, start the turbocharger system characteristic identification process.

[0009] S2. Acquire multiple consecutive frames of boost pressure difference signals according to the preset sampling period;

[0010] S3. Perform frequency domain transformation on the pressure difference signal of each frame to obtain the corresponding frequency domain signal;

[0011] S4. Extract the inherent frequency and corresponding fluctuation amplitude of the booster system based on the frequency domain signal;

[0012] S5. Generate a compensation signal that is inversely related to the boost pressure fluctuation based on the inherent frequency, corresponding fluctuation amplitude, and self-learning updated control parameters.

[0013] S6. The inverted compensation signal is superimposed on the basic control signal of the throttle valve to drive the throttle valve to perform compensation action to suppress boost pressure fluctuations.

[0014] S7. Collect compensation effect feedback parameters and recognition effect feedback parameters, and update the control parameters based on the aforementioned feedback parameters through self-learning. The control parameters include at least one of compensation gain learning coefficient, spectral amplitude threshold learning coefficient, and consistency deviation threshold learning coefficient.

[0015] Furthermore, the self-learning update of the control parameters in step S7 specifically includes at least one of the following:

[0016] The ratio of the fluctuation amplitude of the boost pressure difference after compensation to that before compensation is used as the compensation deviation. When the compensation deviation is greater than the first preset threshold, the compensation gain learning coefficient is increased. When the compensation deviation is less than the second preset threshold, the compensation gain learning coefficient is decreased. When the compensation deviation is between the first and second preset thresholds, the compensation gain learning coefficient remains unchanged.

[0017] Collect the false recognition rate and recognition success rate of multiple consecutive sets of data. When the false recognition rate is greater than the preset false recognition threshold, increase the learning coefficient of the spectrum amplitude threshold. When the recognition success rate is less than the preset recognition threshold and the false recognition rate is not greater than the preset false recognition threshold, decrease the learning coefficient of the spectrum amplitude threshold. When both indicators meet the standard, keep the learning coefficient of the spectrum amplitude threshold unchanged.

[0018] Collect the false recognition rate and recognition success rate of multiple consecutive sets of data. When the recognition success rate is less than the preset recognition threshold and the false recognition rate is not greater than the preset false recognition threshold, increase the consistency deviation threshold learning coefficient. When the false recognition rate is greater than the preset false recognition threshold and the recognition success rate is not less than the preset recognition threshold, decrease the consistency deviation threshold learning coefficient. When both indicators meet the standard, keep the consistency deviation threshold learning coefficient unchanged.

[0019] Upper and lower limits are imposed on the value range of all the aforementioned control parameters.

[0020] Furthermore, a signal preprocessing step is included before step S3, specifically including:

[0021] The collected pressure difference signal of each frame is subjected to DC removal processing to eliminate the static pressure difference component and obtain a pure fluctuation signal.

[0022] The pure wave signal is then windowed to suppress spectral leakage during the frequency domain transformation process.

[0023] Furthermore, the generation of the inverted compensation signal in step S5 specifically includes:

[0024] Based on the identified inherent frequency and preset sampling period, the compensation phase angle is calculated using a discrete recursive method;

[0025] Select the corresponding compensation gain coefficient based on the current control mode of the throttle.

[0026] The fluctuation amplitude, the compensation phase angle, and the preset phase compensation angle are combined to generate an inverse compensation signal.

[0027] Furthermore, step S4 specifically includes:

[0028] In the frequency domain signal, filter the spectrum signal within the preset target frequency band, extract the signal with the maximum amplitude within the frequency band, and remove noise signals with amplitudes lower than the self-learned and updated spectrum amplitude threshold;

[0029] Peak frequencies are extracted from multiple consecutive frames of frequency domain signals, and the consistency of multiple sets of peak frequencies is verified based on the consistency deviation threshold updated by self-learning.

[0030] After successful verification, the average natural frequency and the corresponding average fluctuation amplitude are calculated.

[0031] Furthermore, the specific steps in step S6 involving driving the throttle valve to perform the compensation action include:

[0032] The carrier frequency of the throttle valve PWM drive is kept constant, and compensation control is achieved only by adjusting the PWM duty cycle;

[0033] The duty cycle of the final output after superposition is limited to keep it within a preset effective range.

[0034] Furthermore, it also includes a self-learning parameter management step, specifically including:

[0035] All the self-learned updated control parameters are persistently stored after the vehicle is powered off, and the previously learned values ​​are automatically loaded after the vehicle is powered on.

[0036] When an engine overhaul, replacement of a core component of the turbocharger system, or failure of the self-learning parameters to reach the preset target value for a consecutive preset period are detected, all the control parameters are reset to their initial values ​​and the self-learning process is restarted.

[0037] Furthermore, it also includes troubleshooting steps for low boost pressure, specifically including:

[0038] The target boost pressure and the actual boost pressure are collected in real time, and the deviation between the two is calculated.

[0039] When the deviation value is greater than the preset fault threshold and lasts for a preset time, it is determined to be a fault of low boost pressure;

[0040] A graded post-fault response strategy is implemented based on the severity of the fault, including at least one of limiting engine output power, illuminating the fault indicator light, and storing fault codes.

[0041] Secondly, embodiments of this application provide an electronic device, including: one or more processors;

[0042] A memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors are able to implement the steps in the self-learning compensation and fault handling method described in any of the preceding claims.

[0043] Thirdly, embodiments of this application provide a computer-readable medium storing a computer program, which, when executed by a processor, can implement the steps in the self-learning compensation and fault handling method described in any of the preceding claims.

[0044] This application discloses a method for engine boost pressure frequency domain identification, self-learning compensation, and fault handling. It avoids transient interference at the source by determining the engine's steady-state operating condition, collects the boost pressure difference signal and performs frequency domain transformation to accurately extract the inherent frequency and corresponding fluctuation amplitude of the boost system. Based on the identification result, an inverse compensation signal is generated and superimposed on the throttle body's basic control signal to suppress boost pressure resonance fluctuations at the source. Simultaneously, it collects feedback parameters of the compensation and identification effects and self-learns to update the three core control parameters. This method requires no additional hardware cost and constructs a closed-loop control system of "identification-compensation-feedback-self-learning," which can adapt to changes in engine operating conditions and component aging throughout the engine's entire life cycle, maintaining optimal boost control accuracy and engine operational stability over the long term, effectively improving fuel economy. Attached Figure Description

[0045] Figure 1 This application provides a core flowchart of a method for self-learning compensation and fault handling of engine boost pressure frequency domain identification in an embodiment of the present application.

[0046] Figure 2 This is a structural block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0047] To enable those skilled in the art to better understand the technical solutions of this application, exemplary embodiments of this application are described below with reference to the accompanying drawings, including various details of the embodiments of this application to aid understanding. These should be considered merely exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this application. Similarly, for clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description. Unless otherwise specified, the various embodiments of this application and the features within those embodiments can be combined with each other.

[0048] As used herein, the term "and / or" includes any and all combinations of one or more of the associated enumerated entries. The terminology used herein is for describing particular embodiments only and is not intended to limit the application. As used herein, the singular forms "a" and "the" are also intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that when the terms "comprising" and / or "made of" are used herein, the presence of the stated feature, integral, step, operation, element, and / or component is specified, but the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof is not excluded. Terms such as "connected" or "linked" are not limited to physical or mechanical connections but can include electrical connections, whether direct or indirect.

[0049] Unless otherwise specified, all terms used in this application (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art. It should also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and this application, and will not be interpreted as having an idealized or overly formal meaning, unless expressly so defined in this application.

[0050] All physical parameters involved in the embodiments of this application are uniformly defined as follows:

[0051] 1. General sampling parameters

[0052] T s The system sampling period is a fixed value of 10 ms, i.e., T s =0.01 s, which is the basic time unit for pressure difference acquisition, frequency identification, and compensation calculation.

[0053] F s System sampling frequency, F s =1 / T s =100 Hz, which means that the pressure difference signal is collected 100 times per second.

[0054] k: Discrete sampling time number, which is a non-negative integer (k=0,1,2...n), representing the kth sampling.

[0055] N: FFT operation data frame length, fixed at 1024 points, which is the standard length for spectrum analysis of automotive ECUs.

[0056] 2. Parameters related to boosting pressure difference

[0057] ΔP[k]: The original pressure difference signal collected at time k, ΔP = pressure after pressure - pressure before pressure, unit: kPa.

[0058] : Average pressure difference signal of a single frame, unit: kPa, representing the static pressure difference component.

[0059] : Pressure difference fluctuation signal after DC depressurization at time k, unit: kPa, which is the pure fluctuation component after removing the static pressure difference.

[0060] w[k]: Hanning window function sequence, the value of the Hanning window function at time k, used to suppress spectral leakage during frequency domain transformation.

[0061] x[k]: The pressure difference fluctuation signal after windowing at time k, in kPa.

[0062] X[m]: The complex frequency domain signal after FFT transformation, that is, the complex frequency domain signal corresponding to the m-th spectral line, where m is the spectral line number (m=0,1,2...1023).

[0063] A[m]: The signal amplitude corresponding to the m-th spectral line, in kPa.

[0064] Δf: FFT frequency resolution, unit: Hz.

[0065] f[m]: The frequency value corresponding to the m-th spectral line, in Hz.

[0066] A max Maximum spectral amplitude within the target frequency band, unit: kPa.

[0067] A th : Spectrum amplitude threshold, unit: kPa, used to filter low-amplitude noise signals.

[0068] f n : The inherent frequency of the boost system, in Hz, is the core output parameter for frequency identification.

[0069] A n : The average fluctuation amplitude of the pressure difference corresponding to the natural frequency, in kPa.

[0070] 3. Throttle PWM Compensation Control Parameters

[0071] θ[k]: Compensation phase angle at time k, unit: rad.

[0072] Phase compensation angle, unit: rad, is used to achieve phase reversal cancellation of pressure difference fluctuations. It is calibrated according to the length of the booster pipeline and the installation position of the sensor to ensure that the compensation signal is out of phase with the pressure fluctuation.

[0073] K: Compensation gain coefficient (compensation gain value), dimensionless, calibrated through engine bench testing.

[0074] d comp [k]: Throttle PWM discrete compensation duty cycle at time k, in units of %.

[0075] d base [k]: Throttle valve basic PWM duty cycle at time k, unit: %, calculated from the engine speed, load, and target boost pressure difference by looking up a table.

[0076] d out [k]: The duty cycle of the final PWM output of the throttle body at time k, in percentage.

[0077] f PWM The throttle PWM carrier frequency is set to 15 kHz in this embodiment and remains constant during operation, without changing with the boost pressure difference frequency.

[0078] 4. Self-learning related parameters

[0079] r K : Compensation gain learning coefficient, default value is 0, can be self-learned and updated and persistently stored after the vehicle is powered off.

[0080] r kth : Spectrum amplitude threshold learning coefficient, default value is 0, can be self-learned and updated and persistently stored after the vehicle is powered off.

[0081] r f : Consistency deviation threshold learning coefficient, with a default value of 0. It can be self-learned and updated and persistently stored after the vehicle is powered off.

[0082] ε: Compensation deviation, which is the ratio of the fluctuation amplitude of the boost pressure difference after compensation to that before compensation.

[0083] η: Recognition success rate, which is the percentage of valid recognition groups in a series of consecutive data sets.

[0084] ζ: False recognition rate, which is the proportion of falsely identified groups in a series of consecutive data sets.

[0085] M: Number of self-learning statistical data groups, a fixed value of 10, with each group containing 3 frames of valid recognition data.

[0086] refer to Figure 1 One embodiment of this application proposes a self-learning compensation and fault handling method for frequency domain identification of engine boost pressure, which may specifically include the following steps.

[0087] Step 1: Determine whether the engine is in a steady-state condition. If it is in a steady-state condition, start the turbocharger system characteristic identification process.

[0088] The turbocharger system characteristic identification process is initiated only under steady-state engine conditions to avoid interference from transient conditions on frequency identification. The following conditions must be met simultaneously:

[0089] 1. Engine speed fluctuation ≤ ±50 rpm.

[0090] 2. Throttle opening change rate ≤ ±2% / 100 ms.

[0091] 3. Engine load (intake density of fresh air entering the cylinder) fluctuation ≤ ±3%.

[0092] 4. The boost control enters the closed-loop control state, that is, the engine controller outputs the duty cycle signal of the exhaust gas bypass valve to control the actual boost pressure to follow the target boost pressure, and updates the PID control integral term at the target boost pressure ratio and engine speed through machine learning to ensure control stability.

[0093] If all the above conditions are met, proceed to step 2; otherwise, terminate the current recognition process and wait for the next 10 ms judgment cycle.

[0094] Step 2: Collect multiple consecutive frames of pressure difference signals according to the preset sampling period.

[0095] According to T s With a sampling period of 10 ms, N=1024 points of raw boost pressure difference signal ΔP[k] are continuously collected to form one frame of data. The sampling duration of a single frame is T. frame =N×T s =1024 × 0.01 s = 10.24 s. Three frames of data were continuously collected for subsequent consistency verification, with a total collection time of 30.72 s.

[0096] The core reasons for selecting 3 frames of data for subsequent processing are: on the one hand, using multiple frames of data to offset the impact of real vehicle noise and random disturbances on frequency recognition, avoiding misidentification caused by abnormal single-frame data; on the other hand, the 30.72 s update cycle can track the subtle changes in the inherent frequency of the turbocharger system in a timely manner, adapt to the slow fluctuations in engine operating conditions, and will not increase the ECU's computing load due to excessively frequent updates, meeting the performance requirements of automotive ECU engineering computing, and achieving a balance between recognition accuracy and computing efficiency.

[0097] Step 3: Signal preprocessing.

[0098] Each frame of the boost pressure difference signal is sequentially processed by DC removal and Hanning windowing to provide a high-quality signal for subsequent frequency domain analysis.

[0099] 1. DC removal processing: Calculate the arithmetic mean of the single-frame pressure difference signal, remove the static pressure difference component, and retain only the fluctuation signal.

[0100]

[0101] 2. Hanning Window Processing: The Hanning window is used to weight the fluctuating signal, reducing the discontinuity at the beginning and end of the signal and suppressing spectral leakage.

[0102]

[0103] Step 4: Perform frequency domain transformation on each frame of the boost pressure difference signal to obtain the corresponding frequency domain signal.

[0104] Perform a 1024-point Fast Fourier Transform (FFT) on the windowed signal x[k] to obtain the frequency domain signal X[m], and calculate the amplitude spectrum, frequency resolution, and corresponding spectral line frequencies:

[0105]

[0106] Δf=F s / N≈0.0977 Hz

[0107]

[0108] Step 5: Extract the inherent frequency and corresponding fluctuation amplitude of the booster system based on the frequency domain signal.

[0109] 1. Target frequency band peak extraction: The target frequency band is limited to 0.5 Hz to 20 Hz, based on the natural frequency of the turbocharged system. The maximum spectral amplitude A is searched within this band. max And calculate the spectral amplitude threshold A. th Filter low-amplitude noise.

[0110] A max=max{A[m]|f[m]∈(0.5,20] Hz}

[0111] A th =k r ×A max

[0112] Wherein, the amplitude threshold coefficient k r =k rDefault ×(1+r kth ), k rDefault The default amplitude threshold coefficient is set to 0.3. Only the spectral amplitude A[m]>A is retained. th The effective resonant spectral lines are used to eliminate low-amplitude random noise and sensor clutter interference. If no spectral line in the frequency band satisfies the condition A[m]>A th If it is determined that there is no effective pressure difference resonance fluctuation at present, the frequency extraction of this frame will be abandoned directly and will not proceed to the subsequent multi-frame verification process.

[0113] 2. Multi-frame consistency verification: Repeat steps 2-4 above for 3 consecutive frames of data to obtain 3 sets of peak frequencies f1, f2, and f3. The judgment rule is as follows:

[0114] |f 1- f2∣ <f lim And |f2-f3| < f lim

[0115] Wherein, the consistency deviation f lim =f limDefault ×(1+r f ), f limDefault The default consistency deviation is set to 0.2Hz.

[0116] 3. Parameter Calculation and Update: After verification, calculate the average natural frequency and average fluctuation amplitude.

[0117] f n =(f1+f2+f3) / 3

[0118] A n =(A 1max +A 2max +A 3max ) / 3

[0119] In the formula, A 1max A 2max A 3max : These are the maximum spectral amplitudes of the 1st, 2nd, and 3rd frames of data within the target frequency band, in kPa.

[0120] If the verification fails, it is considered a random disturbance, and the identification is abandoned. f is updated every 30.72 seconds (total duration of 3 frames of data). n and An It is used for subsequent throttle compensation control.

[0121] Step 6: Generate a compensation signal that is inversely related to the boost pressure fluctuation based on the inherent frequency, corresponding fluctuation amplitude, and self-learning updated control parameters.

[0122] 1. Discrete compensation phase recursive calculation: The compensation phase angle is recursively calculated based on a 10 ms sampling period to avoid phase accumulation error caused by continuous integration.

[0123] θ[k]=θ[k-1]+2πf n T s

[0124] 2. Dynamic adjustment of compensation gain: Select the corresponding compensation gain according to the current throttle control mode to balance the compensation effect and control stability.

[0125] If the throttle is in normal control mode: K=K Default ×(1+r K ), K Default The default compensation gain is set to 0.5.

[0126] If the throttle is in anti-overshoot control mode: C1 is the correction coefficient when the throttle is in anti-overshoot control mode, and its value is not less than 0. In this embodiment, it is taken as 0.2.

[0127] 3. Calculation of discrete compensation duty cycle: Generate a compensation duty cycle that is opposite to the pressure fluctuation to achieve fluctuation cancellation.

[0128]

[0129] Step 7: Superimpose the inverted compensation signal onto the basic control signal of the throttle valve to drive the throttle valve to perform compensation action to suppress boost pressure fluctuations.

[0130] The compensated duty cycle is superimposed on the base duty cycle to maintain the throttle PWM carrier frequency f. PWM =15 kHz constant, only update duty cycle:

[0131] d out [k] = d base [k]+ d comp [k]

[0132] If d out [k] If it exceeds the effective range of 0%~100%, it will be limited to the boundary value to ensure that the throttle valve can drive normally.

[0133] Step 8: Collect compensation effect feedback parameters and recognition effect feedback parameters, and update the control parameters based on the aforementioned feedback parameters through self-learning. The control parameters include at least one of the following: compensation gain learning coefficient, spectral amplitude threshold learning coefficient, and consistency deviation threshold learning coefficient.

[0134] With the goal of achieving "optimal compensation effect and highest recognition accuracy," the three core control parameters are self-learned and updated every 30.72 seconds (consistent with the natural frequency update cycle) based on feedback parameters from M=10 consecutive sets of data:

[0135] 1. Feedback parameter acquisition

[0136] Compensation deviation: ε=A n_comp / A n_original The target ε ≤ 0.3.

[0137] In the formula, A n_comp The average fluctuation amplitude of a single group (3 frames) after compensation, in kPa; A n_original The value represents the average fluctuation amplitude of the corresponding single group before compensation, in kPa. The smaller the value of ε, the better the compensation effect.

[0138] Recognition success rate η: The percentage of groups that are effectively recognized (passed the verification in step 5) out of M consecutive data sets, η = number of effectively recognized groups / M; the higher the η, the better the recognition stability, the target η ≥ 90%.

[0139] False recognition rate ζ: The percentage of groups that are falsely identified (noise signals are judged as valid resonant signals) in M ​​consecutive data sets. ζ = number of falsely identified groups / M. The lower the ζ, the better the noise filtering effect. The target ζ ≤ 5%.

[0140] 2. Compensation gain learning coefficient r K renew

[0141] When ε > 0.3 (compensation effect is poor, compensation is insufficient): r K =r K (z)+0.05 means that the compensation gain value of the previous cycle is increased by 5% to enhance the compensation strength.

[0142] When ε < 0.1 (overcompensation, easily leading to throttle control fluctuations): r K =r K (z)-0.05 means that the compensation gain value of the previous cycle is reduced by 5% to weaken the compensation intensity.

[0143] When 0.1≤ε≤0.3 (optimal compensation effect): r K =r K (z), keep the current K value unchanged.

[0144] Constraint: 0.02≤ ≤0.15, avoid An excessively large value leads to unstable throttle control, while an excessively small value renders compensation ineffective. In the formula, r... K (z) represents the learning value of the compensation gain learning coefficient from the previous cycle.

[0145] 3. Spectral amplitude threshold learning coefficient r kth renew

[0146] When ζ > 5% (false recognition rate is too high, noise filtering is insufficient): = +0.1 means increasing the spectral amplitude threshold coefficient value by 10% based on the previous cycle, in order to improve the noise screening threshold and reduce false identification.

[0147] When η < 90% and ζ ≤ 5% (low recognition success rate, valid signals are falsely filtered): = -0.1 means reducing the spectral amplitude threshold coefficient value of the previous cycle by 10% to lower the noise screening threshold and retain more effective signals.

[0148] When η≥90% and ζ≤5% (noise filtering effect and recognition success rate both meet the standards): = Keep the current value unchanged. In the formula, This is the learning value of the learning coefficient for the threshold of the spectrum amplitude in the previous cycle.

[0149] Constraints: The value range is limited to 0.2 ≤ ≤0.5, avoid Too large an amount will result in the loss of effective signals, while too small an amount will exacerbate noise interference.

[0150] 4. Consistency Deviation Threshold Learning Coefficient r f renew

[0151] When η < 90% and ζ ≤ 5% (low recognition success rate, effective frequency is misjudged as disturbance): = +0.1 means increasing the consistency deviation threshold by 10% based on the previous cycle, relaxing the consistency verification standard to improve the effective recognition rate.

[0152] When ζ>5% and η≥90% (high false recognition rate, the disturbance is misjudged as an effective frequency): = -0.1 means reducing the consistency deviation threshold by 10% from the previous cycle, strictly adhering to consistency verification standards to reduce false identification.

[0153] When η≥90% and ζ≤5% (both recognition stability and accuracy meet the standards): rf =r f (z), keeping the current value unchanged. In the formula, r f (z) is the learning value of the learning coefficient of the consistency deviation threshold of the previous period.

[0154] Constraints: The learning coefficient for the consistency deviation threshold is limited to a range of 0.1 Hz ≤ The frequency should be ≤0.3 Hz to avoid a decrease in recognition accuracy due to excessive frequency and a decrease in recognition success rate due to excessive frequency.

[0155] Step 9: Self-learning parameter management.

[0156] 1. Persistent storage: All self-learning parameters (r) K r kth r f The learned values ​​are written to the ECU's non-volatile memory after the vehicle is powered off, and automatically loaded when the vehicle is powered on, avoiding relearning every time the vehicle is powered on.

[0157] 2. Parameter Reset: When an engine overhaul, replacement of core components of the turbocharging system (turbo, throttle, turbocharger pipeline) is detected, or the self-learning parameters fail to reach the target values ​​(ε≤0.3, η≥90%, ζ≤5%) for 5 consecutive update cycles (approximately 2.56 min, 5×30.72 s≈153.6 s) simultaneously, all self-learning parameters are reset to their initial values ​​(0), and the self-learning process is restarted to ensure the rationality and reliability of parameter updates.

[0158] Step 10: Troubleshooting low boost pressure.

[0159] Real-time acquisition of target boost pressure and actual boost pressure; calculation of the deviation between the two; and performance of staged post-processing.

[0160] When the deviation value is >20 kPa and lasts for 5 seconds, it is determined to be a mild underpressure fault, the fault indicator light on the vehicle's instrument panel will be illuminated and the corresponding fault code will be stored.

[0161] When the deviation value is >50 kPa and lasts for 3 seconds, it is determined to be a severe underboost pressure fault, limiting the engine's maximum output power to 60% of the rated power, while illuminating the fault indicator light and storing the fault code.

[0162] Overall, the advantages of this application compared to the prior art include:

[0163] 1. Accuracy of inherent resonance identification

[0164] Existing technologies mostly employ time-domain analysis methods to detect turbocharger system anomalies, relying on general fault thresholds for judgment. This approach fails to accurately identify inherent resonant fluctuations generated by the coupling of turbocharger system piping and components, resulting in low diagnostic accuracy and susceptibility to transient operating conditions. This application filters transient interference through steady-state condition assessment, performs DC removal and Hanning window preprocessing on the turbocharger pressure difference signal, and then performs frequency domain transformation. Combined with multi-frame consistency verification to extract inherent frequencies, the recognition accuracy can be improved to within ±0.2Hz, completely resolving the misidentification problem caused by real-vehicle noise and random disturbances.

[0165] 2. Control parameter adaptability

[0166] Existing boost pressure compensation control systems mostly use fixed parameters, which cannot adapt to changes in engine operating conditions and component aging throughout the engine's entire life cycle. The control effect gradually decays with use, and consistency between different vehicles is poor. This application constructs a closed-loop self-learning system with three core parameters: compensation gain, spectral amplitude threshold, and consistency deviation threshold. Based on compensation deviation, recognition success rate, and false recognition rate, the parameters are dynamically adjusted in real time, enabling the system to maintain optimal fluctuation suppression effect and recognition accuracy over a long period without the need for manual recalibration.

[0167] 3. In terms of the stability of compensation control

[0168] Existing phase compensation technologies mostly use continuous integration to calculate the phase angle, which is prone to accumulated computational errors leading to inaccurate compensation phase. Furthermore, some solutions achieve compensation by changing the PWM carrier frequency, which can cause NVH problems such as throttle motor vibration and whistling. This application uses a discrete recursive method to calculate the compensation phase angle, completely avoiding accumulated integral errors while maintaining a constant throttle PWM carrier frequency. Compensation is achieved solely by adjusting the duty cycle, eliminating abnormal motor noise and meeting automotive functional safety requirements.

[0169] 4. Comprehensiveness of troubleshooting

[0170] Existing technologies often separate boost pressure fluctuation suppression from fault detection, providing only passive protection after a fault occurs. They fail to provide early warning of latent degradation in the boost system and do not offer tiered responses to low boost pressure faults. This application integrates natural frequency identification, anti-phase fluctuation suppression, and tiered post-processing for low boost pressure faults. Furthermore, it can detect early warnings of latent faults such as pipe leaks and turbine jamming through abnormal deviations in the natural frequency, comprehensively improving the operational safety and reliability of the boost system.

[0171] 5. In terms of engineering feasibility

[0172] Some existing diagnostic solutions for turbocharger systems require additional hardware such as electric turbochargers and dedicated sensors, increasing system complexity and vehicle cost, making large-scale application difficult. This application is entirely based on existing engine pressure sensors and ECU hardware, requiring no additional cost. All algorithms are compatible with the computing power of automotive ECUs, enabling direct engineering implementation and mass application.

[0173] Based on the same inventive concept, embodiments of this application also provide an electronic device. Figure 2 This is a structural block diagram of an electronic device provided in an embodiment of this application. Figure 2 As shown in the embodiments of this application, an electronic device includes: one or more processors 101, a memory 102, and one or more I / O interfaces 103. The memory 102 stores one or more programs, which, when executed by the one or more processors, enable the one or more processors to implement any of the self-learning compensation and fault handling methods described in the above embodiments; the one or more I / O interfaces 103 are connected between the processor and the memory, configured to enable information interaction between the processor and the memory.

[0174] The processor 101 is a device with data processing capabilities, including but not limited to a central processing unit (CPU); the memory 102 is a device with data storage capabilities, including but not limited to random access memory (RAM, more specifically SDRAM, DDR, etc.), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), and flash memory (FLASH); the I / O interface (read / write interface) 103 is connected between the processor 101 and the memory 102, and can realize information interaction between the processor 101 and the memory 102, including but not limited to a data bus (Bus).

[0175] In some embodiments, the processor 101, memory 102, and I / O interface 103 are interconnected via bus 104, and thus connected to other components of the computing device.

[0176] In some embodiments, the one or more processors 101 include a field-programmable gate array.

[0177] This application also provides a computer-readable medium. The computer-readable medium stores a computer program, which, when executed by a processor, implements the steps of any of the self-learning compensation and fault handling methods described in the above embodiments. The computer-readable storage medium may be volatile or non-volatile.

[0178] This application also provides a computer program product, including computer-readable code, or a non-volatile computer-readable storage medium carrying computer-readable code. When the computer-readable code is run in the processor of an electronic device, the processor in the electronic device executes the above-described self-learning compensation and fault handling method.

[0179] Those skilled in the art will understand that all or some of the steps, systems, and apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software can be distributed on a computer-readable storage medium, which may include computer storage media (or non-transitory media) and communication media (or transient media).

[0180] As is known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable program instructions, data structures, program modules, or other data. Computer storage media includes, but is not limited to, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), static random access memory (SRAM), flash memory or other memory technologies, portable compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, it is known to those skilled in the art that communication media typically contain computer-readable program instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0181] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0182] The computer program instructions used to perform the operations of this application may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuits, such as programmable logic circuits, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), are personalized by utilizing the status information of the computer-readable program instructions. These electronic circuits can execute the computer-readable program instructions to implement various aspects of this application.

[0183] The computer program product described herein can be implemented specifically through hardware, software, or a combination thereof. In one alternative embodiment, the computer program product is specifically embodied in a computer storage medium; in another alternative embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.

[0184] Various aspects of this application are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0185] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processor of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner; thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.

[0186] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0187] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0188] Exemplary embodiments have been disclosed in this application, and while specific terminology has been used, it is used only and should be interpreted in a general illustrative sense and is not intended to be limiting. In some embodiments, it will be apparent to those skilled in the art that features, characteristics, and / or elements described in conjunction with particular embodiments may be used alone, or in combination with features, characteristics, and / or elements described in conjunction with other embodiments, unless otherwise expressly indicated. Therefore, those skilled in the art will understand that various changes in form and detail may be made without departing from the scope of this application as set forth by the appended claims.

Claims

1. A method for self-learning compensation and fault handling of engine boost pressure frequency domain identification, characterized in that, Includes the following steps: S1. Determine whether the engine is in a steady-state condition. If it is in a steady-state condition, start the turbocharger system characteristic identification process. S2. Acquire multiple consecutive frames of boost pressure difference signals according to the preset sampling period; S3. Perform frequency domain transformation on the pressure difference signal of each frame to obtain the corresponding frequency domain signal; S4. Extract the inherent frequency and corresponding fluctuation amplitude of the booster system based on the frequency domain signal; S5. Generate a compensation signal that is inversely related to the boost pressure fluctuation based on the inherent frequency, corresponding fluctuation amplitude, and self-learning updated control parameters. S6. The inverted compensation signal is superimposed on the basic control signal of the throttle valve to drive the throttle valve to perform compensation action to suppress boost pressure fluctuations. S7. Collect compensation effect feedback parameters and recognition effect feedback parameters, and update the control parameters based on the aforementioned feedback parameters through self-learning. The control parameters include at least one of compensation gain learning coefficient, spectral amplitude threshold learning coefficient, and consistency deviation threshold learning coefficient.

2. The self-learning compensation and fault handling method according to claim 1, characterized in that, Step S7, which involves self-learning to update the control parameters, specifically includes at least one of the following: The ratio of the fluctuation amplitude of the boost pressure difference after compensation to that before compensation is used as the compensation deviation. When the compensation deviation is greater than the first preset threshold, the compensation gain learning coefficient is increased. When the compensation deviation is less than the second preset threshold, the compensation gain learning coefficient is decreased. When the compensation deviation is between the first and second preset thresholds, the compensation gain learning coefficient remains unchanged. Collect the false recognition rate and recognition success rate of multiple consecutive sets of data. When the false recognition rate is greater than the preset false recognition threshold, increase the learning coefficient of the spectrum amplitude threshold. When the recognition success rate is less than the preset recognition threshold and the false recognition rate is not greater than the preset false recognition threshold, decrease the learning coefficient of the spectrum amplitude threshold. When both indicators meet the standard, keep the learning coefficient of the spectrum amplitude threshold unchanged. Collect the false recognition rate and recognition success rate of multiple consecutive sets of data. When the recognition success rate is less than the preset recognition threshold and the false recognition rate is not greater than the preset false recognition threshold, increase the consistency deviation threshold learning coefficient. When the false recognition rate is greater than the preset false recognition threshold and the recognition success rate is not less than the preset recognition threshold, decrease the consistency deviation threshold learning coefficient. When both indicators meet the standard, keep the consistency deviation threshold learning coefficient unchanged. Upper and lower limits are imposed on the value range of all the aforementioned control parameters.

3. The self-learning compensation and fault handling method according to claim 1, characterized in that, Before step S3, there is also a signal preprocessing step, which specifically includes: The collected pressure difference signal of each frame is subjected to DC removal processing to eliminate the static pressure difference component and obtain a pure fluctuation signal. The pure wave signal is then windowed to suppress spectral leakage during the frequency domain transformation process.

4. The self-learning compensation and fault handling method according to claim 1, characterized in that, The specific steps in step S5 to generate the inverted compensation signal include: Based on the identified inherent frequency and preset sampling period, the compensation phase angle is calculated using a discrete recursive method; Select the corresponding compensation gain coefficient based on the current control mode of the throttle. The fluctuation amplitude, the compensation phase angle, and the preset phase compensation angle are combined to generate an inverse compensation signal.

5. The self-learning compensation and fault handling method according to claim 1, characterized in that, Step S4 specifically includes: In the frequency domain signal, filter the spectrum signal within the preset target frequency band, extract the signal with the maximum amplitude within the frequency band, and remove noise signals with amplitudes lower than the self-learned and updated spectrum amplitude threshold; Peak frequencies are extracted from multiple consecutive frames of frequency domain signals, and the consistency of multiple sets of peak frequencies is verified based on the consistency deviation threshold updated by self-learning. After successful verification, the average natural frequency and the corresponding average fluctuation amplitude are calculated.

6. The self-learning compensation and fault handling method according to claim 1, characterized in that, Step S6, which involves driving the throttle valve to perform compensation actions, specifically includes: The carrier frequency of the throttle valve PWM drive is kept constant, and compensation control is achieved only by adjusting the PWM duty cycle; The duty cycle of the final output after superposition is limited to keep it within a preset effective range.

7. The self-learning compensation and fault handling method according to claim 2, characterized in that, It also includes self-learning parameter management steps, specifically including: All the self-learned updated control parameters are persistently stored after the vehicle is powered off, and the previously learned values ​​are automatically loaded after the vehicle is powered on. When an engine overhaul, replacement of a core component of the turbocharger system, or failure of the self-learning parameters to reach the preset target value for a consecutive preset period are detected, all the control parameters are reset to their initial values ​​and the self-learning process is restarted.

8. The self-learning compensation and fault handling method according to claim 1, characterized in that, It also includes troubleshooting steps for low boost pressure, specifically: The target boost pressure and the actual boost pressure are collected in real time, and the deviation between the two is calculated. When the deviation value is greater than the preset fault threshold and lasts for a preset time, it is determined to be a fault of low boost pressure; A graded post-fault response strategy is implemented based on the severity of the fault, including at least one of limiting engine output power, illuminating the fault indicator light, and storing fault codes.

9. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors are enabled to implement the steps in the self-learning compensation and fault handling method as described in any one of claims 1 to 8.

10. A computer-readable medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it can implement the steps in the self-learning compensation and fault handling method as described in any one of claims 1 to 8.

Images