A diesel engine abnormal wear diagnosis system

Abstract

The application discloses a diesel engine abnormal wear diagnosis system, and belongs to the technical field of diesel engine wear detection.The system comprises a data acquisition and processing module, a model calculation module, a fault diagnosis module and a man-machine interaction module.The data acquisition and processing module is used for collecting diesel engine lubricating oil temperature and metal abrasive particle concentration data in the lubricating oil.The model calculation module is used for running a diesel engine operation state diagnosis model and fault prediction.The fault diagnosis module has a fault rule base, and according to the fault prediction result of the model calculation module, the fault rule base, the fusion of real-time working conditions of the diesel engine and diesel engine simulation model data, the final diagnosis result is output.The application realizes the purpose of early identification of abnormal trends of the diesel engine and avoidance of sudden engine faults.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of diesel engine wear detection, and particularly relates to a diesel engine abnormal wear diagnosis system. BACKGROUND

[0002] The existing diesel engine is a compression ignition internal combustion engine, which works based on a four-stroke cycle: in the intake stroke, the piston goes down to form a negative pressure, and filtered air enters the cylinder; in the compression stroke, the piston goes up to compress the air to a high-temperature and high-pressure state (compression ratio is 16-22) of 500-700 DEG C; in the power stroke, the high-pressure atomized diesel oil is injected into the oil nozzle, and the high-temperature and high-pressure gas generated after self-ignition pushes the piston to go down, and the mechanical energy is output through the connecting rod and the crankshaft; in the exhaust stroke, the piston goes up to exhaust the exhaust gas through the exhaust valve, and the cycle is completed; in the whole working process, the engine relies on the cooperation of multiple systems such as crank connecting rod mechanism, valve train, fuel supply system, lubrication system, cooling system, etc.; among them, the lubrication system delivers oil to the key friction pair surfaces such as crankshaft bearing, connecting rod bearing, camshaft, piston ring through the oil pump to form an oil film to reduce friction and wear, and the cooling system removes the heat generated by combustion and friction through the cooling liquid circulation to maintain the engine in the appropriate working temperature range, usually 80-95 DEG C.

[0003] The diesel engine is applied to different fields such as transportation, agriculture, industry, etc. due to its characteristics of large torque, low fuel consumption and high reliability. The transportation generally has long distance demand, and the agricultural engine generally needs long time operation, so the diesel engine will have a series of problems under long time working condition, such as wear of key parts, e.g. wear of crankshaft and bearing, wear of piston and cylinder liner, wear of camshaft and tappet; performance degradation of lubrication system, e.g. oil pollution, oil filter blockage; failure of auxiliary system, e.g. poor heat dissipation of cooling system, cylinder seizure, cylinder sticking.

[0004] Currently, the detection of diesel engine wear is mainly divided into offline detection and online detection. The specific offline detection methods include physical and chemical performance detection of engine oil, wear particle analysis, and disassembly detection method. The online detection methods include vibration detection, temperature and pressure detection, and acoustic detection method. However, the existing wear detection technology has some shortcomings. The offline detection has hysteresis and limitation, long detection cycle, and cannot detect the current wear state of the engine in real time. The disassembly detection method needs to disassemble the engine, which is time-consuming and labor-intensive, has high maintenance cost, and may cause secondary damage to other parts during disassembly. The accuracy and adaptability of the online detection method are insufficient. The vibration detection method is easily affected by engine working conditions (such as load and speed change) and external environmental vibration interference, which makes it difficult to extract characteristic frequency and reduces the detection accuracy, and it is difficult to distinguish between slight wear and normal vibration difference. The temperature and pressure detection is an indirect detection method, which cannot directly locate the wear position and wear degree, and is prone to false alarm or missed alarm. The acoustic detection is seriously affected by environmental noise and has large differences between different models. The existing detection methods are mostly aimed at a single wear indicator or a single part, and lack comprehensive evaluation ability of the overall wear state of the engine, and cannot realize synchronous monitoring and fault warning of multiple friction pairs. SUMMARY

[0005] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present application provides an abnormal wear diagnosis system for diesel engine, which takes the metal particle concentration in the lubricating oil of the diesel engine as the detection object, compares the fault through the set fault rule base, calculates the final diagnosis result by fusing the fault prediction result of the fusion model calculation module, the fault rule base, the real-time working condition of the diesel engine and the simulation model data of the diesel engine, realizes the purpose of identifying the abnormal trend of the diesel engine in advance and avoiding sudden engine failure.

[0006] In order to achieve the above-mentioned purpose, the main technical solutions adopted by the present application include:

[0007] An abnormal wear diagnosis system for diesel engine, comprising a data acquisition and processing module, a model calculation module, a fault diagnosis module and a man-machine interaction module. The data acquisition and processing module is used for acquiring the temperature of the lubricating oil of the diesel engine and the metal particle concentration in the lubricating oil, and sending the preprocessed data to the model calculation module.

[0008] The model calculation module has a diesel engine simulation model and a diagnosis model of the running state of the diesel engine. The model calculation module is used for running the diagnosis model of the running state of the diesel engine and fault prediction. The diesel engine simulation model has three-dimensional geometric data of key parts of the diesel engine. The diagnosis model of the running state of the diesel engine is established by detecting the dynamic change of the metal particle content in the lubricating oil sample, establishing the corresponding relationship between the engine wear state and the metal particle concentration, and the metal particle is Fe, Cu and Al particle. The diagnosis model of the running state of the diesel engine includes the following models:

[0009] Metal Fe and metal Cu, Al abrasive particle concentration model:

[0010]

[0011]

[0012] Wherein, C Fe (t) is the concentration of Fe element at time t; C Fe0 0 is the initial concentration of Fe element; k Fe is the wear rate coefficient of Fe element; C (Cu / Al) (t) is the concentration of Cu and Al element at time t; C (Cu / Al)0 0 is the initial concentration of Cu and Al element; k (Cu / Al) is the wear rate coefficient of Cu and Al element;

[0013] Wear feature vector Model:

[0014]

[0015] Wear particle concentration model:

[0016]

[0017] Where: C i (t) is the concentration of metal abrasive element i at time t, C i 0 is the initial concentration of metal abrasive element i, V0 is the volume of lubricating oil, alpha i is the material wear coefficient, N(τ) is the diesel engine speed, P(τ) is the diesel engine load, beta i is the sedimentation coefficient, C i (τ) is the concentration-dependent nonlinear sedimentation.

[0018] The fault diagnosis module has a fault rule base, the fault diagnosis module according to the fault prediction result of the model calculation module, the fault rule base, the fusion diesel engine real-time working condition and the diesel engine simulation model data, output the final diagnosis result, send to the man-machine interaction module, the man-machine interaction module to the fault prediction result of the model calculation module and the diagnosis result of the fault diagnosis module is displayed.

[0019] Further, the data acquisition and processing module pre-processes the collected data, including denoising, filtering and correcting the collected original signal, and time aligning the collected data.

[0020] ​​​​Further, the data acquisition and processing module pre-processes the collected data, including extracting real-time concentration values as key features for the abrasive particle concentration data, and performing sliding average processing on continuous multiple measurement results.

[0021] Further, the diesel engine simulation model of the model calculation module is based on the inherent design parameters of the diesel engine to be tested, a diesel engine model is established in the AVL-EXCITE platform, and the inherent design parameters of the diesel engine are input to the database module of the AVL-EXCITE platform; the three-dimensional geometric data of the key components are imported through the geometric modeling interface of AVL-EXCITE.

[0022] Further, the key components of the diesel engine include pistons, connecting rods, crankshafts, and cylinders.

[0023] Further, the fault prediction of the model calculation module on the operating state of the diesel engine includes calculating the current metal abrasive particle concentration theoretical prediction value based on the real-time collected metal abrasive particle concentration data, using the diagnostic model of the operating state of the diesel engine, comparing the metal abrasive particle concentration measured value with the theoretical prediction value, triggering a warning for the metal abrasive particles exceeding the threshold concentration, and triggering and grading the warning according to the degree of the metal abrasive particle concentration exceeding the threshold.

[0024] Further, the fault rule base includes the fault component name, fault mode, and fault reason corresponding to the metal element exceeding the threshold.

[0025] Further, when the metal abrasive particle concentration exceeds the threshold, the fault diagnosis module calls the fault rule base, matches the current wear feature vector with the fault mode in the fault rule base, and generates a fault candidate set.

[0026] Further, the fault diagnosis module sets the confidence of the fault prediction result of the model calculation module, the matching result of the fault rule base, and the simulation verification result of the diesel engine simulation model, respectively, and obtains a comprehensive confidence by weighted averaging the confidence of the fault prediction result of the model calculation module, the matching result of the fault rule base, and the simulation verification result of the diesel engine simulation model, and outputs the final diagnosis result and the corresponding fault mode according to the information of the comprehensive confidence.

[0027] Further, the system displays the comprehensive confidence in grades according to the set comprehensive confidence threshold range.

[0028] The beneficial effects of this invention are as follows: The diesel engine abnormal wear diagnosis system of this invention incorporates a diagnostic model of the diesel engine's operating status. The system compares the measured value of the concentration of metal abrasive particles in the lubricating oil with the predicted value calculated by the model, triggering an early warning for metal abrasive particles exceeding a threshold concentration. It then matches the fault type according to the fault rule base built into the fault diagnosis module, thus achieving fault monitoring. Furthermore, this invention employs a three-layer judgment mechanism combining feature threshold determination, fault matching rule determination, and simulation verification determination, thereby improving the accuracy of the system's diagnosis of abnormal wear in diesel engines and reducing the false alarm rate. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the diesel engine abnormal wear diagnosis system of the present invention. Detailed Implementation

[0030] To better explain and facilitate understanding of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0031] like Figure 1 As shown, the present invention provides a diesel engine abnormal wear diagnosis system, including a data acquisition and processing module, a model calculation module, a fault diagnosis module, and a human-machine interaction module. The human-machine interaction module uses the QtCreator platform to design a user interface to display the fault prediction results of the model calculation module and the diagnosis results of the fault diagnosis module. The system of the present invention may also include a recording module connected to the data acquisition and processing module, the model calculation module, the fault diagnosis module, and the human-machine interaction module to record the data in the system.

[0032] The data acquisition and processing module is used to collect data on the temperature of the diesel engine lubricating oil and the concentration of metal abrasive particles in the lubricating oil. After preprocessing the collected data, it is sent to the model calculation module. Specifically, the data acquisition and processing module is connected to the abrasive particle sensor and temperature sensor installed on the main channel of the diesel engine lubricating oil to obtain data on the temperature of the diesel engine lubricating oil and the concentration of metal abrasive particles in the lubricating oil. The concentration of metal abrasive particles in the lubricating oil is used as the main monitoring data, while the temperature of the diesel engine lubricating oil and the size of the metal abrasive particles in the lubricating oil are recorded as supplementary reference data.

[0033] Preferably, after the diesel engine runs normally for a period of time, when the diesel engine reaches a stable working temperature (about 75-95℃), the oil sample is collected from the oil drain port of the oil circulation system, and the oil sample is detected by the abrasive particle sensor. The detection accuracy of the abrasive particle sensor can be the ferromagnetic abrasive particle (Fe) detection and recognition ability (equivalent spherical shape): 0.050mm; the non-ferromagnetic abrasive particle (Cu, Al) recognition ability: 0.150mm, the oil inlet and outlet hole thread hole diameter Φ8±0.5mm; quantitative differentiation of 10μm; the detection rate is not less than 80%.

[0034] Specifically, the data acquisition and processing module pre-processes the collected data, including denoising, filtering and correction of the collected original signal, time alignment processing of the collected data, data synchronization, and subsequent data processing. The data acquisition and processing module extracts real-time concentration values as key features for abrasive particle concentration data, performs sliding average processing on continuous multiple measurement results, forms more stable and reliable comprehensive concentration values and trend data, and provides high-quality input for model calculation.

[0035] The model calculation module has a diesel engine simulation model and a diesel engine operating state diagnosis model. The model calculation module is used to run the diesel engine operating state diagnosis model and fault prediction. The diesel engine simulation model has three-dimensional geometric data of key components of the diesel engine. Specifically, the diesel engine simulation model of the model calculation module is based on the inherent design parameters of the diesel engine to be tested. The diesel engine model is established in the AVL-EXCITE platform, and the inherent design parameters of the diesel engine are input into the database module of the AVL-EXCITE platform. The three-dimensional geometric data of the key components are imported through the geometric modeling interface of AVL-EXCITE. Specifically, the core friction components of the diesel engine include pistons, connecting rods, crankshafts and cylinders.

[0036] The diesel engine operating state diagnosis model is established by detecting the dynamic change of the metal abrasive particle content in the lubricating oil sample, and establishing the corresponding relationship between the engine wear state and the metal abrasive particle concentration. The metal abrasive particles are Fe, Cu and Al particles. Specifically, the diesel engine operating state diagnosis model is based on the principle of combining SOAP (Spectrometric Oil Analysis Program) with engine wear mechanism. The dynamic change of the metal element content in the lubricating oil sample is used to establish the corresponding relationship between the engine wear state and the metal element concentration. The model mainly targets Fe, Cu and Al particles generated by friction pair wear. Through quantitative analysis of Fe, Cu and Al metal elements, the engine operating state diagnosis model is established.

[0037] The diesel engine operating state diagnosis model includes the following models:

[0038] (1) Metal Fe and metal Cu, Al abrasive particle concentration model:

[0039] ;

[0040] ;

[0041] Wherein, C Fe (t) is the concentration of Fe element at time t; C Fe0 is the initial concentration of Fe element; k Fe is the wear rate coefficient of Fe element; C (Cu / Al) (t) is the concentration of Cu and Al elements at time t; C (Cu / Al)0 is the initial concentration of Cu and Al elements; k (Cu / Al) is the wear rate coefficient of Cu and Al elements.

[0042] The metal Fe and metal Cu, Al abrasive particle concentration model is established by the following principles:

[0043] Assume that the volume of lubricating oil in the lubricating oil system is fixed V O , no significant loss occurs in the sampling period, the distribution of metal abrasive particles in the lubricating oil is uniform, the detection sampling can represent the overall concentration level, the generation rate of metal abrasive particles is a function of the load, speed and temperature of the worn parts; the concentration changes in accordance with the quasi-linear growth or exponential growth law, then the basic formula is:

[0044] ;

[0045] Wherein, C i (t) is the concentration of metal abrasive particle i element at time t, in ppm, i.e. mg / L; C i0 is the initial concentration, in ppm, i.e. mg / L; V O is the volume of lubricating oil, in L; G i (τ) is the metal abrasive particle generation rate, in mg / h; L i (τ) is the metal abrasive particle loss rate, in mg / h; i is the element category of metal abrasive particles.L i(τ) metal particles including filter interception and sedimentation part; filter interception part metal particles represent the mass of metal particles filtered by the filter per unit time, with units of mg / h. The lubricating oil system is generally provided with a lubricating oil filter to maintain the cleanliness of the lubricating oil. When the lubricating oil carrying metal particles flows through the filter, the metal particles will be intercepted and captured by the filter element, thereby being physically removed from the circulating engine oil. The sedimentation part metal particles represent the mass of metal particles that disappear per unit time due to sedimentation, with units of mg / h. Due to the action of gravity, heavier metal particles in the lubricating oil will gradually sink out of the flowing oil and eventually settle in the bottom of the oil pan, the inner wall of the crankcase, the oil passage and other low-flow parts. These settled metal particles no longer participate in circulation and therefore cannot be detected by sensors installed in the circulating oil circuit. The model introduces the filter interception and sedimentation part metal particles into the model, improving the accuracy of the model prediction results.

[0046] Assuming that the filtering effect is constant and weak, the model can be approximated as:

[0047] ;

[0048] where k i is the wear rate coefficient, with units of ppm / h, and the model can reflect the process of gradually stabilizing the element concentration over time.

[0049] (2) Wear feature vector Model:

[0050] .

[0051] (3) Wear particle concentration model:

[0052] ;

[0053] where: C i (t) is the concentration of metal particles element i at time t, with units of ppm, i.e. mg / L; C i 0 is the initial concentration of metal particles element i, with units of ppm, i.e. mg / L; V0 is the volume of lubricating oil, with units of L; α i is the material wear coefficient, with units of ; N(τ) is the speed of the diesel engine, with units of rpm; P(τ) is the load of the diesel engine, representing the percentage of rated load; β i is the sedimentation coefficient, C i (τ) is the concentration-dependent nonlinear sedimentation, with units of ppm, i.e. mg / L.

[0054] Specifically, the model calculation module includes the following steps for fault prediction of the diesel engine operating state: based on the real-time collected metal particle concentration data, a theoretical prediction value of the current metal particle concentration is calculated by using a diagnosis model of the diesel engine operating state, the measured value of the metal particle concentration is compared with the theoretical prediction value, a warning trigger is performed on the metal particles exceeding the threshold concentration, and the warning trigger and warning classification are performed according to the degree of the metal particle concentration exceeding the threshold.

[0055] The fault diagnosis module has a fault rule base, which is embedded in the diesel engine abnormal wear diagnosis system in the form of an SQLite database. The fault rule base includes a wear safety baseline and a wear grade table and fault matching rules. Specifically, the fault rule base includes the fault component name, fault mode, and fault cause corresponding to the metal element exceeding the threshold.

[0056] Table 1 is a wear safety baseline and wear grade table (unit: ppm):

[0057] .

[0058] The system gives the following maintenance suggestions according to the diagnosis results:

[0059] Mild wear: increase the oil monitoring frequency, sample every 100 h; use anti-wear additives to reduce the wear rate growth; moderate wear: locally overhaul the wear parts and replace the lubricating oil; adjust the engine operating parameters to reduce the load and wear rate; severe wear: immediately stop, overhaul the key parts and replace the lubricating oil; after maintenance, retest the three metal concentrations and vibration amplitude to ensure that the wear particle detection value returns to normal or mild level.

[0060] The fault matching rules are as follows:

[0061] When only the Fe concentration in the lubricating oil exceeds the threshold, and the Fe concentration C Fe > 15 ppm, it is determined that the fault component is the cylinder, the fault mode is cylinder pulling, and the fault reason is that the engine is run at high load without warming up after cold start; the gap between the piston and the cylinder sleeve is too small to damage the oil film; overloading operation causes local overheating.

[0062] When only the Al concentration in the lubricating oil exceeds the threshold, and the Al concentration C Al > 16 ppm, it is determined that the fault component is the piston connecting rod group; further calculate the change rate K Al of the Al concentration, if K Al< 0.5 ppm / h, it is determined that the big end bearing of the connecting rod is worn, and the cause of the fault is that the oil contains impurities, the oil pressure is insufficient, the bearing assembly gap is abnormal, the bearing material is fatigued and worn, the lubricating oil channel is blocked, the connecting rod bolt pre-tightening force is insufficient, or the crankshaft and connecting rod are seriously faulty; if KAI≥0.5 ppm / h, it is determined that the small end bearing of the connecting rod is worn, and the cause of the fault is that the oil quality is poor or the oil supply is insufficient, resulting in poor lubrication; or an accelerated fatigue crack caused by overload or improper assembly gap.

[0063] When the Fe and Cu concentrations in the lubricating oil exceed the threshold value, and the Fe concentration C Fe > 30 ppm and C Cu > 10 ppm, it is determined that the fault component is the crankshaft flywheel set; further calculation of C Cu / C Fe , if C Cu / C Fe ≤ 0.33, it is determined that the main bearing is worn, and the cause of the fault is poor lubrication, improper bearing installation, or shaft misalignment; if C Cu / C Fe > 0.33, it is determined that the gear is broken, and the cause of the fault is poor lubrication, improper assembly, or abnormal gear meshing.

[0064] Specifically, when the metal abrasive particle concentration exceeds the threshold value, the fault diagnosis module calls the fault rule library, matches the current wear characteristic vector with the fault patterns in the fault rule library, and generates a fault candidate set.

[0065] The fault diagnosis module outputs the final diagnosis result based on the fault prediction result of the model calculation module, the fault rule library, the real-time working condition of the diesel engine, and the diesel engine simulation model data, and sends it to the human-computer interaction module. The human-computer interaction module displays the fault prediction result of the model calculation module and the diagnosis result of the fault diagnosis module for the user to view.

[0066] More specifically, the model calculation module compares the real-time collected metal abrasive particle concentration data with the current metal abrasive particle concentration theoretical prediction value calculated by the diagnosis model of the diesel engine operating state, triggers a warning for metal abrasive particles exceeding the safety baseline in Table 1, and classifies the warning according to the range of the metal abrasive particle concentration exceeding the threshold value in Table 1, i.e. preliminary determination of the wear level. At the same time, according to the fault matching rules in the fault rule library, the fault patterns are matched to generate a fault candidate set.

[0067] The fault diagnosis module sets confidence levels for the fault prediction results, fault rule base matching results, and diesel engine simulation model verification results from the model calculation module. The module then obtains a comprehensive confidence level by weighted averaging these confidence levels. The system categorizes and displays the comprehensive confidence level into different levels based on a set threshold range. Preferably, the system can output different levels of certainty based on the confidence level, such as "Confirmed," "Suspected," or "Abnormal and pending investigation."

[0068] More specifically, the confidence level for matching according to Table 1 is defined as the feature threshold determination confidence level C. 阈值 When the concentration of metal abrasive particles is within a safe baseline range, C 阈值 =0; When the concentration of metal abrasive particles is within the range of mild wear, C 阈值 =0.6; When the concentration of metal abrasive particles is within the range of moderate wear, C 阈值 =0.8; When the concentration of metal abrasive particles is within the range of severe wear, C 阈值 =0.95.

[0069] The confidence level of matching based on fault matching rules is defined as the rule base matching confidence level C. 规则 If the system determines that the metal element exceeds the threshold, then C 规则 =0.5; If the system calculates and compares the concentration ratios of the metal elements that simultaneously exceed the standard, then C 规则 =0.3; If the system calculates the rate of change of metal concentration with time, then C 规则 =0.2.

[0070] The deviation between the measured frictional work loss value and the frictional work loss value in the AVL-EXCITE simulation is defined as the simulation verification confidence level C. 仿真 If the deviation is less than 10%, then C 仿真 =0.9; if 10% ≤ deviation < 20%, then the confidence level is C. 仿真 =0.7; if the deviation is ≥20%, then it is C. 仿真 =0.3.

[0071] The fault diagnosis module will perform C 阈值 C 规则 C 仿真 We perform weighted fusion to obtain the final overall confidence level C. 综合 C 综合 =W1×C 阈值 +W2×C 规则 +W3×C 仿真 The weighting coefficients are W1=0.2, W2=0.6, and W3=0.2.

[0072] Comprehensive confidence level classification: when 0.8≤C 综合 ≤1.0, it belongs to high confidence, and the specific diagnosis result is directly output, and "confirmed diagnosis" is displayed; when 0.5≤C 综合 ≤0.8, it belongs to medium confidence, the diagnosis result is output, and "suspected" is displayed; when C 综合 <0.5, it belongs to low confidence, and no specific diagnosis result is displayed, only "abnormal to be checked" is displayed, that is, it is warned that the engine has a wear state.

[0073] The system adopts a combination mode of three-layer judgment mechanisms of feature threshold judgment, fault matching rule judgment and simulation verification judgment, and improves the accuracy of the system on the diagnosis result of the abnormal wear of the diesel engine.

[0074] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application, and the modifications, modifications, replacements and variations of the above embodiments by those skilled in the art are within the scope of the present application.

Claims

1. A diesel engine abnormal wear diagnosis system characterized by comprising: The system comprises a data acquisition and processing module, a model calculation module, a fault diagnosis module, and a human-computer interaction module. The data acquisition and processing module is used for collecting the temperature of the lubricating oil of the diesel engine and the concentration of metal abrasive particles in the lubricating oil, and sending the preprocessed data to the model calculation module. The model calculation module has a diesel engine simulation model and a diagnosis model of the operating state of the diesel engine, and is used for running the diagnosis model of the operating state of the diesel engine and fault prediction. The diesel engine simulation model has three-dimensional geometric data of key components of the diesel engine. The diagnosis model of the operating state of the diesel engine is established by detecting the dynamic change of the content of metal abrasive particles in the lubricating oil sample, establishing the corresponding relationship between the engine wear state and the concentration of metal abrasive particles, and the metal abrasive particles are Fe, Cu and Al particles. ； ； wherein C Fe (t) is the concentration of Fe element at time t; C Fe0 is the initial concentration of Fe element; k Fe is the wear rate coefficient of Fe element; C (Cu / Al) (t) is the concentration of Cu and Al elements at time t; C (Cu / Al)0 is the initial concentration of Cu and Al elements; k (Cu / Al) is the wear rate coefficient of Cu and Al elements; Wear feature vector Model: ； The concentration model of metal Fe, Cu and Al abrasive particles: ； where: C i (t) is the concentration of the metal particulate element i at time t, C i 0 is the initial concentration of the metal particulate element i, V0 is the volume of the lubricating oil, a i is the material wear coefficient, N(τ) is the diesel engine speed, P(τ) is the diesel engine load, β i is the settling coefficient, C i (τ) is the concentration-dependent non-linear settling; The concentration model of wear particles:

2. A diesel engine abnormal wear diagnostic system according to claim 1, characterised in that: The fault diagnosis module has a fault rule base, and outputs the final diagnosis result according to the fault prediction result of the model calculation module, the fault rule base, the real-time working condition of the diesel engine and the data of the diesel engine simulation model, and sends the result to the human-computer interaction module.

3. A diesel engine abnormal wear diagnostic system according to claim 1 or 2, characterised in that: The preprocessing of the collected data by the data acquisition and processing module includes denoising, filtering and correction of the original signals, and time alignment of the collected data.

4. A diesel engine abnormal wear diagnostic system as set forth in claim 1 wherein: The preprocessing of the collected data by the data acquisition and processing module includes extracting real-time concentration values as key features for the abrasive particle concentration data, and performing sliding average processing on the continuous measurement results. The diesel engine simulation model of the model calculation module is based on the inherent design parameters of the diesel engine to be tested, and a diesel engine model is established in the AVL-EXCITE platform.

5. A diesel engine abnormal wear diagnostic system as set forth in claim 1 wherein: The three-dimensional geometric data of the key components are imported through the geometric modeling interface of AVL-EXCITE.

6. A diesel engine abnormal wear diagnostic system as set forth in claim 1 wherein: The key components of the diesel engine include pistons, connecting rods, crankshafts and cylinders.

7. A diesel engine abnormal wear diagnostic system as set forth in claim 1 characterized by: The fault prediction of the model calculation module for the operating state of the diesel engine includes calculating the theoretical prediction value of the current metal abrasive particle concentration based on the real-time collected metal abrasive particle concentration data, comparing the measured value with the theoretical prediction value, triggering a warning for the metal abrasive particles exceeding the threshold concentration, and triggering a warning and grading according to the degree of the metal abrasive particle concentration exceeding the threshold.

8. A diesel engine abnormal wear diagnostic system as set forth in claim 1 wherein: The fault rule base includes the fault component name, fault mode and fault reason corresponding to the metal element exceeding the threshold. When the metal abrasive particle concentration exceeds the threshold, the fault diagnosis module calls the fault rule base, matches the current wear feature vector with the fault mode in the fault rule base, and generates a fault candidate set.

9. A diesel engine abnormal wear diagnostic system as set forth in claim 1 wherein: The fault diagnosis module sets confidence degrees for the fault prediction result of the model calculation module, the fault rule base matching result and the simulation verification result of the diesel engine simulation model respectively, and obtains a comprehensive confidence degree by weighted average of the confidence degrees of the fault prediction result of the model calculation module, the fault rule base matching result and the simulation verification result of the diesel engine simulation model, and outputs a final diagnosis result and a corresponding fault mode according to information of the comprehensive confidence degree.

10. A diesel engine abnormal wear diagnostic system according to claim 9, characterised in that: The system displays the comprehensive confidence degree in grades according to a set comprehensive confidence degree threshold range.

Images