Spark plug replacement reminding system based on resistance value change

Abstract

The invention belongs to the technical field of automotive electronics and intelligent sensing, and particularly relates to a spark plug replacement reminding system based on resistance value change, which comprises a composite sensing module, the composite sensing module is in signal connection with an edge computing node, and the edge computing node is connected with an ECU (Electronic Control Unit) for real-time data interaction; the composite sensing module comprises a first part and a second part, the first part is arranged on the central electrode or the grounding electrode of the spark plug, and the second part is arranged on the opposite grounding electrode or the central electrode; and the edge computing node is wirelessly connected with a vehicle-mounted network for wireless diagnosis, and after acquiring the resistance value of the spark plug from the composite sensing module, the edge computing node generates a dynamic threshold value through the dynamic threshold value decision-making device in combination with ECU working condition data so as to perform early warning judgment, generate an early warning instruction and output the early warning instruction. The method is based on directly collected resistance value data, and through FTT and adaptive Kalman filtering preprocessing, the monitoring sensitivity is high, and the method can adapt to extreme working conditions such as a turbocharged engine.

Description

Technical Field

[0001] The present invention belongs to the field of automotive electronics and intelligent sensing technology, and particularly relates to a spark plug replacement reminder system based on resistance value changes. Background Art

[0002] The spark plug real-time monitoring system provides real-time health management for spark plugs in internal combustion engines, hybrid engines, and hydrogen fuel cells. Compared to traditional spark plugs that rely on periodic replacement, this system can detect electrode wear, carbon deposits, or misfire risks in real time. Spark plugs operate in high temperatures (>800°C) and high voltages (30kV). Conventional testing methods that directly collect spark plug resistance are difficult to achieve reliable data acquisition at such high temperatures and high pressures. Furthermore, they cannot ensure that the monitoring and discharge functions do not interfere with each other, resulting in indirect data collection for testing. Therefore, current monitoring methods mostly use indirect methods, calculating the spark plug's condition based on engine status parameters or setting up a circuit to sample the spark plug voltage. While these methods can effectively monitor the spark plug's condition, they cannot monitor actual changes in the spark plug's resistance, resulting in certain issues with accuracy and reliability. Summary of the Invention

[0003] The purpose of the present invention is to provide a spark plug replacement reminder system based on resistance value changes, which is used to solve the technical problem that the existing technology cannot directly detect and collect data on the resistance value of the spark plug, resulting in insufficient accuracy and reliability of the monitoring results.

[0004] The spark plug replacement reminder system based on resistance value changes includes a composite sensing module directly arranged on the spark plug, the composite sensing module is signal-connected to the edge computing node, and the edge computing node is connected to the ECU for real-time data exchange; the composite sensing module includes a first part and a second part, the first part is arranged on the center electrode or ground electrode of the spark plug, and the second part is arranged on the opposite ground electrode or center electrode; the first part and the second part both use a carbon nanotube-ceramic composite material as a sensor substrate, the first part provides a dynamic impedance layer on the sensor substrate, and the second part provides a 3D micro-resistance array on the sensor substrate, and the 3D micro-resistance array and the dynamic impedance layer are both located between the center electrode and the ground electrode; the edge computing node is provided with an ASIC chip with an integrated FFT accelerator and capable of executing adaptive Kalman filtering, the edge computing node is wirelessly connected to the vehicle network for wireless diagnosis, after obtaining the spark plug resistance from the composite sensing module, combined with the ECU operating condition data, generates a dynamic threshold through a dynamic threshold decision maker, and then makes a warning judgment, generates a warning instruction, and outputs it.

[0005] Preferably, the carbon nanotube-ceramic composite material is a CNT-Al2O3 material, and the CNT-Al2O3 material is deposited by the ALD method to form an ALD deposition layer, the ALD deposition layer includes a CNT-Al2O3 passivation layer and a 5nm Pt sensing layer, the Pt sensing layer serves as a resistance sensing material, the 3D microresistor array is a 3D interdigitated resistor array, the dynamic impedance layer uses a VO2 phase change material layer, the phase change is triggered by the spark discharge pulse width, and a femtosecond laser welding wire is used between the sensor and the external circuit.

[0006] Preferably, the CNT-Al2O3 passivation layer is 20 nm thick, the Pt sensing layer is 5 nm thick, the 3D interdigitated resistor array is prepared by deep ultraviolet lithography, with a line width of 5 μm, an array thickness of 50 μm, and a resistance of 10 MΩ±1%; the phase change response time of the VO2 phase change material layer is 1.2 μs; the diameter of the weld spot of the femtosecond laser welding is <30 μm, and the heat-affected zone is <5 μm.

[0007] Preferably, a multi-parameter diagnostic algorithm is used to achieve early warning, and the multi-parameter diagnostic algorithm includes the following steps:

[0008] S1. Data acquisition: The composite sensor module collects the spark plug resistance and obtains the corresponding temperature, speed, and noise energy through real-time interaction with the ECU.

[0009] S2. Perform signal preprocessing, which includes FFT analysis and Kalman filtering;

[0010] S3, calculating the dynamic threshold of resistance change and comparing;

[0011] S4, performing mutation detection and noise energy detection on the second-order derivative of the temperature coefficient of resistance TCR respectively;

[0012] S5. Based on the warning trigger conditions, a warning instruction of the corresponding level is issued and a warning action is executed.

[0013] Preferably, in step S1, the temperature is used to dynamically compensate the threshold and calculate the temperature change. When the temperature T>150°C, the dynamic threshold is relaxed by 10%. The speed is used to adjust the noise criterion weight. When the speed>5000rpm, spectrum-assisted analysis is enabled, and the temperature coefficient of resistance TCR is calculated based on the temperature change and resistance value.

[0014] Preferably, in step S2, the FFT analysis uses a Blackman-Harris window function to optimize the spectral resolution, and the spectrum range of the noise is 200kHz–500kHz; the state variables of the Kalman filter include the resistance change ΔR, the resistance temperature coefficient TCR and the noise energy, and the noise parameters include process noise and observation noise, thereby optimizing the parameters.

[0015] Preferably, in step S3, the calculation formula of the dynamic threshold is: R 阈值 =R 基础 ×(1+αΔT)×.(1+β(speed / 1000), where α is the temperature compensation coefficient, β is the speed compensation coefficient, R 基础 Indicates the basic threshold of resistance change, automatically correcting R based on historical diagnostic results 基础 , based on the dynamic threshold comparison judgment, if the resistance change ΔR>R 阈值 , then mark the carbon deposit warning and proceed to subsequent judgment, otherwise end.

[0016] Preferably, in step S4, the second-order derivative of the temperature coefficient of resistance TCR is used for mutation detection, and the formula is as follows:

[0017]

[0018] Where t is the temperature and Δt is the temperature change value; and the detection result is judged according to the corresponding threshold. The mutation judgment condition is: |d 2 TCR / dt 2 |>mutation condition threshold; perform noise energy detection and compare the noise energy with the noise energy threshold.

[0019] Preferably, step S5 uses decision-level fusion to perform early warning, and triggers the early warning instruction based on the early warning trigger logic. The early warning trigger logic is as follows:

[0020] Level 1 warning: trigger condition is ΔR>R 阈值 If the noise energy is greater than 3dB, the action to be performed is to light up the dashboard indicator light;

[0021] Level 2 warning: The trigger condition is that TCR meets the mutation judgment conditions and ΔR>R 阈值 , the execution action is to generate an OBD fault code;

[0022] Level 3 warning: The triggering conditions are that the three parameters of ΔR, TCR, and noise energy are abnormal at the same time, and the execution action is that the ECU limits the torque output.

[0023] Preferably, step S5 calculates a comprehensive outlier value by feature-level fusion, and then performs early warning judgment using a comprehensive threshold value calculated by a dynamic threshold value; the fusion algorithm for the comprehensive outlier value is as follows:

[0024] Comprehensive abnormal value = w1ΔR + w2TCR 突变量 +w3 noise energy,

[0025] The weight distribution is obtained after optimization based on historical data, TCR 突变量 is the result of the second-order derivative mutation detection of the temperature coefficient of resistance TCR, w1, w2, and w3 are the corresponding resistance changes ΔR and TCR respectively.突变量 The comprehensive threshold is generated based on the weight distribution and dynamic threshold, mutation detection threshold and noise energy threshold. The comprehensive outlier value is compared with the comprehensive threshold to make an early warning.

[0026] The advantages of the present invention are:

[0027] 1) Through the design of a new composite sensor module, the problem of sensor failure on spark plugs in high-temperature environments is solved, interference with the spark discharge path is avoided, non-invasive measurement is achieved, monitoring and discharge functions do not interfere with each other, energy loss is reduced, and the spark plug resistance data is directly obtained.

[0028] 2) Through the connection between edge computing nodes, composite sensor modules, and vehicle networks, on the one hand, real-time noise processing and high-speed processing of relevant data can be achieved through ASIC chips, and wireless diagnosis can be realized. Diagnostic and early warning functions are realized through multi-parameter diagnostic algorithms.

[0029] 3) Because the diagnostic algorithm of this invention is based on directly collected resistance data and pre-processed with FTT and adaptive Kalman filtering, it offers high monitoring sensitivity, eight times higher than traditional OBD systems, enabling submillimeter wear monitoring. Furthermore, the monitoring system is adaptable to extreme operating conditions, such as turbocharged engines. Furthermore, the addition of a composite sensor module has no substantial impact on the spark plug structure, making mass production and maintenance costs manageable without requiring significant production line modifications. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 The figure is a schematic diagram of the structure of the spark plug replacement reminder system based on resistance value change in the present invention.

[0031] Figure 2 for Figure 1 Schematic diagram of the composite sensing module in the shown structure.

[0032] Figure 3 This is a dynamic impedance switching waveform diagram of the spark plug voltage change in the present invention.

[0033] Figure 4 This is a dynamic impedance switching waveform diagram of the spark plug resistance change in the present invention.

[0034] Figure 5 This is a flow chart of the multi-parameter diagnostic algorithm used in the present invention. DETAILED DESCRIPTION

[0035] The specific implementation methods of the present invention will be further explained in detail below through the description of embodiments with reference to the accompanying drawings, so as to help those skilled in the art to have a more complete, accurate and in-depth understanding of the inventive concept and technical solution of the present invention.

[0036] like Figure 1-Figure 5 As shown, the present invention provides a spark plug replacement reminder system based on resistance value changes, including a composite sensor module directly arranged on the spark plug, the composite sensor module is signal-connected to the edge computing node, and the edge computing node is connected to the ECU for real-time data exchange; the composite sensor module includes a first part and a second part, the first part is arranged on the center electrode or ground electrode of the spark plug, and the second part is arranged on the opposite ground electrode or center electrode; the first part and the second part both use a carbon nanotube-ceramic composite material as a sensor substrate, the first part is provided with a dynamic impedance layer on the sensor substrate, and the second part is provided with a 3D micro-resistance array on the sensor substrate, and the 3D micro-resistance array and the dynamic impedance layer are both located between the center electrode and the ground electrode; the edge computing node is provided with an ASIC chip with an integrated FFT accelerator and capable of executing adaptive Kalman filtering, the edge computing node is wirelessly connected to the vehicle network for wireless diagnosis, after obtaining the spark plug resistance from the composite sensor module, the edge computing node generates a dynamic threshold through a dynamic threshold decision maker in combination with the ECU operating condition data, and then makes an early warning judgment, generates an early warning instruction, and outputs it.

[0037] The carbon nanotube-ceramic composite material is a CNT-Al2O3 material with a temperature resistance of >1000°C and a resistance linearity error of <±0.3%. The CNT-Al2O3 material is deposited using the ALD method to form an ALD deposition layer. The ALD deposition layer includes a 20nm CNT-Al2O3 passivation layer and a 5nm Pt sensing layer, with a total thickness of 25nm. The CNT-Al2O3 passivation layer is used to improve the sensor's high temperature resistance and chemical corrosion resistance; the Pt sensing layer serves as a resistance sensing material to ensure high sensitivity and stability. The ALD deposition layer forms a multi-layer composite structure, ensuring the reliable operation of the sensor in extreme environments (>800°C, 30kV).

[0038] The 3D micro-resistor array is a 3D interdigitated resistor array fabricated using deep ultraviolet lithography. Its line width is 5μm, with a line width accuracy of ±0.1μm, ensuring uniform resistance distribution. The array thickness is 50μm, and its resistance is 10MΩ ±1%. The 3D interdigitated resistor array enables non-invasive resistance measurement by connecting it in parallel with a ground electrode or a center electrode, avoiding interference with the spark discharge path.

[0039] The dynamic impedance layer utilizes a VO2 phase-change material layer with a phase change response time of 1.2μs, triggered by the spark discharge pulse width. The VO2 phase-change material layer automatically switches to a high-impedance state (>100MΩ) at the moment of spark discharge (pulse width <2μs), returning to a low-impedance state when static. This ensures that the monitoring and discharge functions do not interfere with each other, reducing energy loss to less than 0.1%.

[0040] A femtosecond laser (wavelength 1030nm, pulse width 500fs) is used to weld the wires between the sensor and the external circuitry. The weld diameter is less than 30μm, reducing mechanical stress on the sensor structure. The heat-affected zone is less than 5μm, preventing thermal damage and performance degradation. This ensures a high-precision connection between the sensor and the external circuitry while maintaining structural integrity.

[0041] In the edge computing node, the ASIC chip integrates an FFT accelerator (1024 points / 10μs), capable of real-time noise processing and adaptive Kalman filtering. The wireless module utilizes the LoRa protocol, is compatible with ISO 11565 in-vehicle networks, and supports wireless diagnostics. Its sleep power consumption is less than 10μW, and the transmission interval is adjustable from 1s to 10 minutes. The power management module supports 12V / 24V in-vehicle power supplies, with a peak current of less than 5mA. Heat dissipation is dissipated through the metal housing of the spark plug, eliminating the need for additional cooling.

[0042] Figure 3 、 Figure 4 The following are the dynamic impedance switching waveforms, corresponding to voltage and resistance changes. The corresponding key parameters in the figure have the following meanings:

[0043] t1: spark discharge triggering time (0μs);

[0044] t2: VO2 layer begins phase transition (0.3μs);

[0045] t3: Impedance switching completed (1.5μs);

[0046] Response time: Δt = t3 - t1 = 1.2 μs;

[0047] The resistance range includes:

[0048] Low resistance state (monitoring phase): 1kΩ (0-0.3μs);

[0049] High impedance state (discharge stage): 100MΩ (0.3-1.5μs).

[0050] It can be seen from the above figures that the spark plug can still work normally and collect corresponding data after adopting the composite sensor module.

[0051] The spark plug replacement reminder system uses a multi-parameter diagnostic algorithm to achieve early warning. The multi-parameter diagnostic algorithm includes the following steps:

[0052] S1. Data Acquisition: The composite sensor module collects the spark plug resistance ΔR and, through real-time interaction with the ECU, obtains the corresponding temperature, speed, and noise energy (noise spectrum is 200-500kHz). Temperature and speed are both provided by the ECU. The former is used to dynamically compensate for the threshold and calculate temperature changes. For example, when the temperature T>150°C, the dynamic threshold is relaxed by 10%. The latter is used to adjust the noise criterion weighting, and spectrum-assisted analysis is enabled when the speed is >5000rpm. The temperature coefficient of resistance (TCR) is calculated based on the temperature change and resistance value.

[0053] S2. Perform signal preprocessing.

[0054] Signal preprocessing includes FFT analysis and Kalman filtering.

[0055] FFT analysis uses a Blackman-Harris window function to optimize spectral resolution. The noise spectrum range is 200 kHz–500 kHz, with a bin bandwidth of 48.8 kHz. Execution time: 10 μs (accelerated by an ASIC chip).

[0056] The state variables of the Kalman filter include resistance change ΔR, resistance temperature coefficient TCR and noise energy. The noise parameters include: process noise Q = diag(0.01, 0.005, 0.1), observation noise R = diag(0.1, 0.05, 0.2). Through adaptive Kalman filtering, this step

[0057] Adaptive Kalman Filter:

[0058] By state estimation (x = [ΔR, TCR, noise energy] T ) and covariance matrix update to optimize the fusion results in real time:

[0059] Prediction equation

[0060]

[0061] Symbol definition:

[0062] The prior state estimate (predicted value) at the kth moment, that is, the state estimate without considering the current observation data.

[0063] A: State transition matrix, which describes how the system state evolves over time (dynamic relationship between ΔR and TCR as they change with temperature).

[0064] The posterior state estimate at the k-1th moment (the optimal estimate after correction by the previous round of observations).

[0065] B: Control input matrix, the impact of ignition timing adjustment instructions on the state.

[0066] u k : The control signal sent by the control input ECU at the kth moment.

[0067] Update equation

[0068]

[0069]

[0070] Symbol definition:

[0071] K k : Kalman gain, used to balance the weights of predicted values ​​and observed values.

[0072] Prior estimation error covariance matrix

[0073] H: Observation matrix, which maps the state space to the observation space (from ΔR, TCR to the actual measured values ​​of the resistance change).

[0074] R: Observation noise covariance matrix, which characterizes the statistical characteristics of the sensor measurement noise (i.e., R = diag(0.1, 0.05, 0.2)).

[0075] z k : The actual observed value at the kth moment (the resistance change measured by the sensor).

[0076] The posterior state estimate at the kth moment (the corrected optimal estimate).

[0077] P k : The posterior estimation error covariance matrix, the updated state uncertainty.

[0078] S3. Calculate the dynamic threshold of resistance change and compare.

[0079] The calculation formula of dynamic threshold is: R 阈值 =R 基础 ×(1+αΔT)×.(1+β(speed / 1000), where α is the temperature compensation coefficient, usually 0.005 / °C, β is the speed compensation coefficient, ΔT is the temperature change value, usually 0.03, R 基础 Indicates the basic threshold value of resistance change. Then, based on the dynamic threshold, the comparison is made and the judgment is made. If the resistance change ΔR>R 阈值 , then the carbon deposit warning is marked, and every 5% increase in ΔR corresponds to 0.1mm of carbon deposits, and then subsequent judgment is made, otherwise it ends.

[0080] It should be noted that this method requires dynamic value adjustment in combination with environmental parameters, including adjusting the temperature compensation coefficient and the speed compensation coefficient. The threshold is relaxed by 10% at high temperatures (T>150°C) and the noise spectrum auxiliary criterion is enabled at high speeds (speed>5000rpm). In addition, this method has a real-time feedback mechanism that automatically corrects R based on historical diagnostic results. 基础 , to avoid misjudgment due to sensor drift.

[0081] S4. Perform mutation detection and noise energy detection on the second-order derivative of the temperature coefficient of resistance (TCR).

[0082] The second-order derivative of the temperature coefficient of resistance TCR is used for mutation detection. The formula is as follows:

[0083]

[0084] Where t is the temperature and Δt is the temperature change value; and the detection result is judged according to the corresponding threshold. The mutation judgment condition is: |d 2 TCR / dt 2 |>0.15μΩ / ℃ / ms (mutation condition threshold).

[0085] Noise energy detection is performed, and the noise energy is compared with a noise energy threshold. The judgment condition is: noise energy>noise energy threshold (3dB in the embodiment).

[0086] S5. Based on the warning trigger conditions, a warning instruction of the corresponding level is issued and a warning action is executed.

[0087] One way to do this is to use decision-level fusion for early warning. This method triggers early warning instructions based on early warning trigger logic. The early warning trigger logic and early warning instructions are shown in Table 1:

[0088] Table 1: Warning trigger logic and warning instructions

[0089] Trigger Conditions Warning level Execute an action ΔR>threshold & noise energy>3dB Level 1 Light up the yellow indicator light TCR mutation &ΔR>threshold Level 2 Generate OBD fault code P030X Three parameters are abnormal at the same time Level 3 ECU limits torque output

[0090] As shown in Table 1:

[0091] Level 1 warning (warning information: carbon deposit warning): trigger condition is ΔR>R 阈值 If the noise energy is greater than 3dB, the action to be performed is to light up the indicator light on the instrument panel.

[0092] Secondary warning (warning information: electrode wear critical): trigger condition is that TCR meets the judgment mutation condition and ΔR>R 阈值 The execution action is to generate OBD fault code P030X.

[0093] Level 3 warning (warning information: replace immediately): The trigger condition is that the three parameters of ΔR, TCR, and noise energy are abnormal at the same time (ΔR>R 阈值 , TCR meets the judgment mutation conditions, noise energy>3dB), the execution action is to limit the torque output of the ECU.

[0094] Execution action feedback:

[0095] Closed-loop control is performed in the third-level warning: the ECU limits the torque output according to the warning level, and also adjusts the ignition timing by adjusting the instructions to prevent the engine from being seriously damaged due to spark plug failure.

[0096] Another way to do this step is to calculate the comprehensive anomaly value through feature-level fusion, and then make an early warning judgment using the comprehensive threshold calculated by the dynamic threshold.

[0097] The weighted fusion method is used to perform feature-level fusion on data, and the fusion algorithm for comprehensive outliers is as follows:

[0098] Comprehensive abnormal value = w1ΔR + w2TCR 突变量 +w3 noise energy,

[0099] The weight distribution is obtained after optimization based on historical data, TCR 突变量 is the result of the second-order derivative mutation detection of the temperature coefficient of resistance TCR, w1, w2, and w3 are the corresponding resistance changes ΔR and TCR respectively. 突变量 The weight of the noise energy; for example: w1 = 0.5, w2 = 0.3, w3 = 0.2. The comprehensive threshold is generated based on the weight distribution and each threshold (dynamic threshold, mutation detection threshold and noise energy threshold). The comprehensive abnormal value is compared with the comprehensive threshold to make an early warning. This early warning method also needs to dynamically modify the weights based on environmental parameters: w i '=w i ×(1+αΔT), where α is the temperature compensation coefficient, ΔT is the temperature change value, w i is the original value of the weight before correction, w i ' is the corrected value of the weight after correction. After collecting environmental parameters, the corrected value of the weight is selected as the new weight to calculate the comprehensive outlier value and comprehensive threshold.

[0100] The present invention is described above by way of example in conjunction with the accompanying drawings. It is obvious that the specific implementation of the present invention is not limited to the above-mentioned method. As long as various non-substantial improvements are made using the inventive concept and technical solution of the present invention, or the inventive concept and technical solution are directly applied to other occasions without improvement, they are all within the scope of protection of the present invention.

Claims

1. A spark plug replacement reminder system based on resistance value changes, characterized by: It includes a composite sensing module directly arranged on the spark plug, the composite sensing module is signal-connected to the edge computing node, and the edge computing node is connected to the ECU for real-time data interaction; the composite sensing module includes a first part and a second part, the first part is arranged on the center electrode or ground electrode of the spark plug, and the second part is arranged on the opposite ground electrode or center electrode; the first part and the second part both use carbon nanotube-ceramic composite materials as sensor substrates, the first part is provided with a dynamic impedance layer on the sensor substrate, and the second part is provided with a 3D micro-resistance array on the sensor substrate, and the 3D micro-resistance array and the dynamic impedance layer are both located between the center electrode and the ground electrode; the edge computing node is provided with an ASIC chip with an integrated FFT accelerator and capable of executing adaptive Kalman filtering, the edge computing node is wirelessly connected to the vehicle network for wireless diagnosis, after the edge computing node obtains the resistance value of the spark plug from the composite sensing module, combines the ECU operating condition data to generate a dynamic threshold through a dynamic threshold decision maker, and then makes an early warning judgment, generates an early warning instruction and outputs it.

2. The spark plug replacement reminder system based on resistance value change according to claim 1, characterized in that: The carbon nanotube-ceramic composite material is a CNT-Al2O3 material, which is deposited using an ALD method to form an ALD deposition layer. The ALD deposition layer includes a CNT-Al2O3 passivation layer and a 5nm Pt sensing layer. The Pt sensing layer serves as a resistance sensing material. The 3D microresistor array is a 3D interdigitated resistor array. The dynamic impedance layer uses a VO2 phase change material layer. The phase change is triggered by a spark discharge pulse width. Femtosecond laser welding wires are used between the sensor and the external circuit.

3. The spark plug replacement reminder system based on resistance value change according to claim 2, characterized in that: The CNT-Al2O3 passivation layer is 20nm thick, the Pt sensing layer is 5nm thick, and the 3D interdigitated resistor array is prepared by deep ultraviolet lithography, with a line width of 5μm, an array thickness of 50μm, and a resistance of 10MΩ±1%. The phase change response time of the VO2 phase change material layer is 1.2μs. The diameter of the weld spot of the femtosecond laser welding is <30μm, and the heat-affected zone is <5μm.

4. The spark plug replacement reminder system based on resistance value change according to any one of claims 1 to 3, characterized in that: A multi-parameter diagnostic algorithm is used to achieve early warning. The multi-parameter diagnostic algorithm includes the following steps: S1. Data acquisition: The composite sensor module collects the spark plug resistance and obtains the corresponding temperature, speed, and noise energy through real-time interaction with the ECU. S2. Perform signal preprocessing, which includes FFT analysis and Kalman filtering; S3, calculating the dynamic threshold of resistance change and comparing; S4, performing mutation detection and noise energy detection on the second-order derivative of the temperature coefficient of resistance TCR respectively; S5. Based on the warning trigger conditions, a warning instruction of the corresponding level is issued and a warning action is executed.

5. The spark plug replacement reminder system based on resistance value change according to claim 4, characterized in that: In step S1, the temperature is used to dynamically compensate the threshold and calculate the temperature change. When the temperature T is greater than 150°C, the dynamic threshold is relaxed by 10%. The speed is used to adjust the noise criterion weight. When the speed is greater than 5000 rpm, spectrum-assisted analysis is enabled to calculate the temperature coefficient of resistance (TCR) based on the temperature change and resistance value.

6. The spark plug replacement reminder system based on resistance value change according to claim 4, characterized in that: In step S2, the FFT analysis uses the Blackman-Harris window function to optimize the spectral resolution, and the noise spectrum range is 200kHz–500kHz. The state variables of the Kalman filter include the resistance change ΔR, the resistance temperature coefficient TCR, and the noise energy. The noise parameters include process noise and observation noise, thereby optimizing the parameters.

7. The spark plug replacement reminder system based on resistance value change according to claim 6, characterized in that: In step S3, the calculation formula of the dynamic threshold is: R 阈值 =R 基础 ×(1+αΔT)×.(1+β(speed / 1000), where α is the temperature compensation coefficient, β is the speed compensation coefficient, R 基础 Indicates the basic threshold of resistance change, automatically correcting R based on historical diagnostic results 基础 , based on the dynamic threshold comparison judgment, if the resistance change ΔR>R 阈值 , then mark the carbon deposit warning and proceed to subsequent judgment, otherwise end.

8. The spark plug replacement reminder system based on resistance value change according to claim 7, characterized in that: In step S4, the second-order derivative of the temperature coefficient of resistance (TCR) is used for mutation detection, and the formula is as follows: Where t is the temperature and Δt is the temperature change value; and the detection result is judged according to the corresponding threshold. The mutation judgment condition is: |d 2 TCR / dt 2 |>mutation condition threshold; perform noise energy detection and compare the noise energy with the noise energy threshold.

9. The spark plug replacement reminder system based on resistance value change according to claim 8, characterized in that: Step S5 uses decision-level fusion to perform early warning, and triggers the early warning instruction based on the early warning trigger logic. The early warning trigger logic is as follows: Level 1 warning: trigger condition is ΔR>R 阈值 If the noise energy is greater than 3dB, the action to be performed is to light up the dashboard indicator light; Level 2 warning: The trigger condition is that TCR meets the mutation judgment conditions and ΔR>R 阈值 , the execution action is to generate an OBD fault code; Level 3 warning: The triggering conditions are that the three parameters of ΔR, TCR, and noise energy are abnormal at the same time, and the execution action is that the ECU limits the torque output.

10. The spark plug replacement reminder system based on resistance value change according to claim 4, characterized in that: Step S5 calculates the comprehensive outlier value through feature-level fusion, and then makes an early warning judgment based on the comprehensive threshold value calculated by the dynamic threshold value; the fusion algorithm of the comprehensive outlier value is as follows: Comprehensive abnormal value = w1ΔR + w2TCR 突变量 +w3 noise energy, The weight distribution is obtained after optimization based on historical data, TCR 突变量 is the result of the second-order derivative mutation detection of the temperature coefficient of resistance TCR, w1, w2, and w3 are the corresponding resistance changes ΔR and TCR respectively. 突变量 The comprehensive threshold is generated based on the weight distribution and dynamic threshold, mutation detection threshold and noise energy threshold. The comprehensive outlier value is compared with the comprehensive threshold to make an early warning.

Images