Complete diesel engine reliability prediction method based on mechanism and data fusion

Abstract

The invention relates to a mechanism and data fused whole diesel engine reliability prediction method in the technical field of internal combustion engines. The method comprises the following steps: basic modeling parameter input, engine performance model construction, dynamic model construction, wear and fatigue calculation of multiple samples of key parts, reliability analysis of a mechanism model, and prediction of reliability of a mechanism model. And analyzing the reliability of the whole machine after fusing the statistical model. According to the method, the mechanism model and statistical data are fused to predict the reliability of the whole diesel engine, so that the reliability calculation confidence of the diesel engine is remarkably improved; moreover, the method not only can obtain the reliability of the parts and the whole machine of the diesel engine mechanism model, but also can obtain the reliability of the whole machine of the diesel engine fused with statistical data.

Description

Technical Field

[0001] The invention relates to the field of diesel engine reliability, and in particular to a mechanism capable of improving diesel engine reliability calculation confidence and a data fusion diesel engine complete machine reliability prediction method. Background Art

[0002] Traditional diesel engine reliability prediction methods are based on mathematical statistics or fault physics. The mathematical statistics-based method requires a large amount of test and operation data, which is difficult to implement for high-power marine engines with a small number of samples. The method based on fault physics uses mathematical models to establish cause-and-effect relationships and reflect objective physical laws. However, some faults have many influencing factors and it is difficult to establish an accurate mechanism model. The application of this method is also limited. Only by combining reliability statistical theory with failure physics model research can we better understand the complexity and regularity of the uncertain world. Summary of the invention

[0003] In order to overcome the shortcomings of the prior art, the present invention discloses a diesel engine whole machine reliability prediction method by integrating mechanism and data. The present invention integrates the mechanism model and statistical data to predict the whole diesel engine reliability, which significantly improves the confidence of the diesel engine reliability calculation. Moreover, this method can not only obtain the reliability of the components and the whole machine of the diesel engine mechanism model, but also obtain the reliability of the whole diesel engine integrated with the statistical data.

[0004] The present invention is achieved by the following technical solution, which comprises the following steps:

[0005] Step 1: Input basic modeling parameters, including material property parameters and engine structure and performance parameters;

[0006] Step 2: Engine performance model construction. To solve the change of state parameters in each subsystem with the crankshaft angle, the differential equations consisting of the energy conservation equation, mass conservation equation, and ideal gas state equation are solved using the prediction and correction method.

[0007] Step 3: Dynamic model construction, using the cylinder pressure, temperature and other parameters obtained from the engine performance model as boundary conditions, and the loads of each component are calculated based on the kinematics and dynamics of the crank-connecting rod mechanism;

[0008] Step 4: Wear and fatigue calculation of multiple samples of key components. Multiple samples are generated based on the Monte Carlo algorithm. The wear of the friction pair is calculated based on the Archard model. The fatigue of the components is calculated based on the SN curve of the material and the fatigue cumulative damage theory.

[0009] Step 5: Reliability analysis of the mechanism model: statistical analysis of the wear and fatigue of multiple samples of components obtained to obtain the reliability of each component; further, the reliability of the whole machine mechanism model is obtained;

[0010] Step 6: After integrating the statistical model, analyze the reliability of the whole machine, calculate the reliability of the statistical component failure rate data, and integrate it with the reliability calculated by the mechanism model to obtain the integrated whole machine reliability.

[0011] Further, in step one of the present invention, the material property parameters include: Poisson's ratio of key component materials; elastic modulus of materials; hardness of materials; SN curve of crankshaft materials; SN curve of connecting rod materials; tensile strength of crankshaft materials; tensile strength of connecting rod materials; engine structure and performance parameters include heat release rate, valve timing, stroke, cylinder diameter, compression ratio, number of cylinders, number of volumes, number of valves, ignition order, speed, power, size of each volume, connecting rod ratio, piston mass, connecting rod mass, crank unbalanced mass, connecting rod length, distance from the center of mass of the connecting rod to the center of the big end of the connecting rod, crank rotation angular velocity, crank radius, piston diameter, distance from the middle section of the main journal to the middle section of the crank arm on the nearest side, crank arm width, crank arm thickness, crank pin cross-sectional diameter, main journal cross-sectional diameter, rod cross-sectional area, key friction pair contact length, key friction pair component wear coefficient, key friction pair assembly clearance, and key friction pair replacement limit clearance.

[0012] Furthermore, in step 2 of the present invention, the energy conservation equation is:

[0013]

[0014] Where: Indicates the crankshaft angle, in degrees; m z Indicates the mass of the working fluid, in kg; u z is the specific internal energy of the working medium in the cylinder, in J / kg; m in is the intake air mass, in kg; h i is the specific enthalpy of the intake air, in J / kg; Q f The heat brought by the fuel injected into the cylinder, the unit is J; Q w is the energy lost through heat transfer, in J; m o is the exhaust mass, in kg; h o is the specific enthalpy of the exhaust gas, in J / kg; p z is the pressure in the subsystem, in Pa; V z is the volume of the subsystem, in m 3 ;

[0015] The mass conservation equation is:

[0016]

[0017] Where: Indicates the crankshaft angle, in degrees; m z Indicates the mass of the working fluid, in kg; m in is the intake air mass, in kg; m o is the exhaust mass, in kg; g f is the fuel mass in kg; x is the percentage of fuel combustion in the cylinder;

[0018] The ideal gas state equation is expressed as:

[0019] p z V z =m z R z T z

[0020] Where: p z is the pressure in the subsystem, in Pa; V z is the volume of the subsystem, in m 3 ;m z Indicates the mass of the working fluid, in kg; T z is the temperature in the cylinder, in K; R z is the gas constant of the working fluid, with the unit of J / (kg·K).

[0021] Furthermore, in step three of the present invention, the load of each component can be expressed as:

[0022]

[0023]

[0024]

[0025] Where: F ∑ is the force on the small end of the connecting rod along the center line of the cylinder, which can be decomposed into the piston lateral force F NG and connecting rod thrust F C , the unit is N; connecting rod thrust F C It can be further decomposed into the normal force F along the crank direction R and the tangential force F perpendicular to the crank direction T , the unit is N; F P is the force acting on the crank pin, in N; F K The force acting on the main shaft neck, unit is N; m p , m c and m KThey are respectively expressed as piston mass, connecting rod mass and crank unbalanced mass, all in kg; the connecting rod mass is often equivalent to two parts, namely the reciprocating motion part m CA And rotary motion part m CB , the unit is kg; F g It is expressed as the gas pressure that the piston is subjected to when it reciprocates in the cylinder, and the unit is N; L cb is the distance from the center of mass of the connecting rod to the center of the big end of the connecting rod, in m; L is the distance between the centers of the big and small ends of the connecting rod, in m; ω is the angular velocity, in rad / s; R is the crank radius, in m; β is the swing angle of the connecting rod, is the crank angle, the unit is deg; λ is the ratio of crank radius to connecting rod length; D P is the piston diameter, in m; m r It represents the total mass concentrated at the crank pin and rotating with the crank, in kg, which can be expressed as:

[0026] m r =m K +m CB

[0027] Where: m K and m CB They are the unbalanced mass of the crank and the mass of the rotating part of the connecting rod, both in kg.

[0028] Furthermore, in step four of the present invention, the Monte Carlo algorithm is used to generate samples by considering the influence of uncertainty factors such as material properties, manufacturing assembly, external loads, etc. on the calculated diesel engine damage, determining the distribution function of these uncertainty parameters, generating a series of random numbers, transforming these random numbers to obtain samples of the corresponding probability distribution function, substituting them into the corresponding uncertainty parameters, and generating a large number of random samples of the corresponding distribution function for each uncertainty parameter.

[0029] Furthermore, in step 4 of the present invention, the wear depth of key friction pair components in the Archard wear model can be expressed as:

[0030]

[0031] Where: h is the wear depth, in m; l is the sliding distance of the friction pair, in m; K is the dimensionless wear coefficient; H is the Brinell hardness of the material, in HBW; p is the contact stress, in Pa, which can be expressed as:

[0032]

[0033] Where: F is the normal contact pressure, unit is N; S is the contact area, unit is m 2 .

[0034] Wear damage is expressed as:

[0035]

[0036] Where: D w represents wear damage, the unit is 1; t represents time, the unit is h; h and h c They are wear depth and critical wear depth respectively, both in mm.

[0037] When the wear damage reaches 1, the component fails.

[0038] Among them, the wear of the cylinder liner-piston ring friction pair leads to a decrease in cylinder sealing and thus causes medium leakage in the cylinder, which is fed back into the calculation and analysis of the performance model.

[0039] The number of cycles under the symmetrical cyclic stress amplitude of the fatigue damage is expressed as:

[0040] lgN=a+blgσ -1

[0041] Where: N is the number of failure cycles corresponding to the stress amplitude, unit is 1; σ -1 It indicates the stress amplitude of the material under symmetrical loading with a stress ratio of -1, in MPa; a and b are material constants, obtained from the SN curve of the material. Since the force on the component is asymmetrical loading force, when calculating the fatigue life using the SN curve, the Goodman formula is usually used to correct the influence of the average stress. The Goodman formula is expressed as:

[0042]

[0043] Where: b is the tensile strength limit of the material, σ m With σ a are the stress mean and stress amplitude under actual working conditions, σ -1 is the stress amplitude when the stress ratio is -1, and the unit is MPa.

[0044] Fatigue damage is the cumulative damage at all stress levels and is expressed as:

[0045]

[0046] Where: D j represents fatigue damage, the unit is 1; t represents time, the unit is h; n i Indicates the stress amplitude σ of the component i The following has been completed i Cycles, unit is 1; N iIt represents the total number of cycles that the component can withstand under this stress amplitude, with the unit of 1; m represents the number of different stress amplitudes, with the unit of 1; when the fatigue damage reaches 1, the component fails.

[0047] Furthermore, the reliability of each component is obtained by statistical analysis in step 5 of the present invention, and the calculation principle is expressed as follows:

[0048]

[0049] Where: t represents time, the unit is h; n represents the total number of samples, D i is represented as the damage of the i-th sample, It is expressed as the reliability of the jth component at time t, and the unit is 1; I(x) is the indicator function, which is expressed as:

[0050]

[0051] The reliability of the whole machine mechanism model can be expressed as:

[0052]

[0053] Where: t represents time, the unit is h; P{d p (t)<1} is the reliability of the whole machine mechanism model at time t, j represents the jth component, It is expressed as the reliability of the j-th component at time t, and the unit is 1.

[0054] Furthermore, in step 6 of the present invention, the reliability of the component failure rate data calculated and statistically determined, for common components whose failure rates are determined by reliability tests and data statistics, such as some seals and common parts, it can be considered that the failure data of these components are obtained under the condition of normal operation of key components, and assuming that the failures of these components are independent of each other, the reliability of these components is expressed as:

[0055]

[0056] Where: t represents time, the unit is h; d p (t) and d d (t) are the norms of the damage vectors DP(t) and DD(t), respectively, which can be expressed as:

[0057] d p (t)=||DP(t)||

[0058] d d (t)=||DD(t)||

[0059] Where: DP(t) represents the damage of a part of the components calculated by the mechanism model, which is a random vector that varies with time; DD(t) represents the damage of another part of the components obtained by data statistics, which is also a random vector that varies with time; P{DD k (t)<1} is the reliability of the kth component of the data statistics, with the unit of 1, which can be expressed as:

[0060]

[0061] Where: k represents the kth component counted by the data, and the unit is 1; t represents the t time, η represents the η time, ξ represents the ξ time, and the unit is h; λ k (η) is the instantaneous failure rate of the kth component at time η, in units of h -1 .

[0062] The calculation principle of the integrated whole machine reliability is expressed as:

[0063] R(t)=P{d p (t)<1,d d (t)<1}=P{d p (t)<1}P{d d (t)<1|d p (t)<1}

[0064] Where: R(t) is the reliability of the whole machine after the fusion of mechanism and data, the unit is 1; t represents the time t, the unit is h; d p (t) and d d (t) are the norms of the damage vectors DP(t) and DD(t), respectively, which can be expressed as:

[0065] d p (t)=||DP(t)||

[0066] d d (t)=||DD(t)||

[0067] Where: DP(t) represents the damage of a part of the components calculated by the mechanism model, which is a random vector that varies with time; DD(t) represents the damage of another part of the components obtained by data statistics, which is also a random vector that varies with time.

[0068] In step 2 of the present invention, the energy conservation equation, mass conservation equation, and ideal gas state equation are established for the volume of each subsystem, taking into account the changes in state parameters at each crankshaft angle, and are solved using the prediction and correction method. In step 3, a dynamic model is constructed based on the kinematics and dynamics of the crank-connecting rod mechanism and the loads of each component are solved. In step 4, the component load obtained based on the dynamics is used as a boundary condition, the wear amount of the friction pair is calculated based on the Archard model, the fatigue of the component is calculated based on the SN curve of the material and the fatigue cumulative damage theory, and the Monte Carlo algorithm is used to generate samples of corresponding distributions for the uncertainty parameters in the wear model and the fatigue model. In step 5 of the present invention, the wear and fatigue damage of multiple samples are statistically analyzed to obtain the reliability of the component and the whole machine mechanism model. In step 6 of the present invention, the reliability of the component statistical model is obtained from the statistical prior data, and the reliability of the whole machine is obtained after being fused with the mechanism model.

[0069] Compared with the prior art, the beneficial effects of the present invention are as follows: compared with the traditional diesel engine reliability prediction method, the present method has two significant characteristics. First, the present method models the real-time reliability of the diesel engine and can comprehensively consider the coupling relationship between the real-time wear amount of the piston ring and the performance. Second, the present method not only focuses on the mechanism model of the diesel engine components, but also integrates the statistical model of the prior statistical data. The present invention integrates the mechanism model and the statistical data to predict the reliability of the diesel engine as a whole, which significantly improves the confidence of the diesel engine reliability calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 is a flow chart of the present invention;

[0071] Figure 2 A real-time pressure diagram in a cylinder according to an embodiment of the present invention;

[0072] Figure 3 This is a real-time temperature diagram of the cylinder in an embodiment of the present invention;

[0073] Figure 4 This is a piston ring damage distribution diagram after the embodiment of the present invention has been running for 8000 hours;

[0074] Figure 5 This is a cylinder liner damage distribution diagram after the embodiment of the present invention has been running for 8000 hours;

[0075] Figure 6 It is a reliability diagram of the whole machine mechanism model of the embodiment of the present invention;

[0076] Figure 7 This is a complete reliability diagram of a diesel engine with the mechanism and data fusion of an embodiment of the present invention. DETAILED DESCRIPTION

[0077] In order to make the content of the present invention easier to understand, the technical scheme of the present invention is further explained below in conjunction with specific embodiments. The examples are only used to illustrate the present invention, but the present invention is not limited to this content. The operating methods without specifying specific conditions in the following examples are usually carried out under conventional conditions or according to the product instructions.

[0078] Example

[0079] like Figures 1 to 6 As shown, the present invention is achieved through the following technical solutions:

[0080] Step 1: Obtain the real input parameters of a certain model. The required material property parameters and engine performance parameters are shown in Tables 1 and 2.

[0081] Table 1 Engine structural parameters

[0082]

[0083] Table 2 Material property parameters

[0084]

[0085]

[0086] Step 2: Construct a performance model. Assume that the pressure, temperature and other parameters in each subsystem are uniformly distributed in space, and the state parameters change with the crankshaft angle. The differential equation group consisting of the three basic equations is used to solve the change of the state parameters in each subsystem with the crankshaft angle using the prediction and correction method.

[0087] The energy conservation equation is expressed as:

[0088]

[0089] The mass conservation equation is expressed as:

[0090]

[0091] The ideal gas state equation is expressed as:

[0092] p z V z =m z RT z

[0093] Where: Indicates the crankshaft angle, in degrees; m z Indicates the mass of the working fluid, in kg; u z is the specific internal energy of the working medium in the cylinder, in J / kg; m inis the intake air mass, in kg; h i is the specific enthalpy of the intake air, in J / kg; Q f The heat brought by the fuel injected into the cylinder, the unit is J; Q w is the energy lost through heat transfer, in J; m o is the exhaust mass, in kg; h o is the specific enthalpy of the exhaust gas, in J / kg; p z and V z are the pressure and volume in the subsystem, in Pa and m respectively 3 ; g f is the fuel mass in kg; x is the percentage of fuel combustion in the cylinder; T z is the temperature in the cylinder, in K; R z is the gas constant of the working fluid, in J / (kg·K); It can be expressed as

[0094]

[0095]

[0096] Where: m p and m d are the quality coefficients of premixed combustion and diffusion combustion respectively; Q d The fuel fraction for diffusion combustion; and are the combustion duration angles of premixed combustion and diffusion combustion, respectively. is the combustion starting angle, τ is the premixed combustion leading angle, both in degrees.

[0097] Step 3: Construct a dynamic model. The cylinder pressure and cylinder temperature (such as Figure 2 and Figure 3 As shown in ) as the boundary condition, according to the dynamic analysis of the crank-connecting rod mechanism, the forces of each component can be expressed as:

[0098]

[0099]

[0100]

[0101] Where: F ∑ is the force on the small end of the connecting rod along the center line of the cylinder, which can be decomposed into the piston lateral force F NG and connecting rod thrust F C , the unit is N; connecting rod thrust F C It can be further decomposed into the normal force F along the crank directionR and the tangential force F perpendicular to the crank direction T , the unit is N; F P is the force acting on the crank pin, in N; F K The force acting on the main shaft neck, unit is N; m p , m c and m K They are respectively expressed as piston mass, connecting rod mass and crank unbalanced mass, all in kg; the connecting rod mass is often equivalent to two parts, namely the reciprocating motion part m CA And rotary motion part m CB , the unit is kg; F g It is expressed as the gas pressure that the piston is subjected to when it reciprocates in the cylinder, and the unit is N; L cb is the distance from the center of mass of the connecting rod to the center of the big end of the connecting rod, in m; L is the distance between the centers of the big and small ends of the connecting rod, in m; ω is the angular velocity, in rad / s; R is the crank radius, in m; β is the swing angle of the connecting rod, is the crank angle, the unit is deg; λ is the ratio of crank radius to connecting rod length; D P is the piston diameter, in m; m r It represents the total mass concentrated at the crank pin and rotating with the crank, in kg, which can be expressed as:

[0102] m r =m K +m CB

[0103] Where: m K and m CB They are the unbalanced mass of the crank and the mass of the rotating part of the connecting rod, both in kg.

[0104] Step 5: Calculate the wear and fatigue of multiple samples of key components. According to the Archard wear model, the wear depth of key friction pair components can be expressed as:

[0105]

[0106] Where: h is the wear depth, in m; l is the sliding distance of the friction pair, in m; K is the dimensionless wear coefficient; H is the Brinell hardness of the material, in HBW; p is the contact stress, in Pa, which can be expressed as:

[0107]

[0108] Where: F is the normal contact pressure, unit is N; S is the contact area, unit is m 2 .

[0109] Wear damage is expressed as:

[0110]

[0111] Where: D w Indicates wear damage (such as Figure 4 and Figure 5 ), the unit is 1; t represents time, the unit is h; h and h c They are wear depth and critical wear depth respectively, both in mm.

[0112] When the wear damage reaches 1, the component fails.

[0113] Among them, the wear of the cylinder liner-piston ring friction pair leads to a decrease in cylinder sealing and thus causes medium leakage in the cylinder, which is fed back into the calculation and analysis of the performance model.

[0114] The fatigue of the component is calculated based on the SN curve of the material, and the number of cycles under the symmetrical cyclic stress amplitude is expressed as:

[0115] lgN=a+blgσ -1

[0116] Where: N is the number of failure cycles corresponding to the stress amplitude, unit is 1; σ -1 It indicates the stress amplitude of the material under symmetrical loading with a stress ratio of -1, in MPa; a and b are material constants, obtained from the SN curve of the material. Since the force on the component is asymmetrical loading force, when calculating the fatigue life using the SN curve, the Goodman formula is usually used to correct the influence of the average stress. The Goodman formula is expressed as:

[0117]

[0118] Where: b is the tensile strength limit of the material, σ m With σ a are the stress mean and stress amplitude under actual working conditions, σ -1 is the stress amplitude when the stress ratio is -1, and the unit is MPa.

[0119] Fatigue damage is the cumulative damage at all stress levels and is expressed as:

[0120]

[0121] Where: D j represents fatigue damage, the unit is 1; t represents time, the unit is h; n i Indicates the stress amplitude σ of the component i The following has been completed i Cycles, unit is 1; N iIt represents the total number of cycles that the component can withstand under this stress amplitude, with the unit of 1; m represents the number of different stress amplitudes, with the unit of 1; when the fatigue damage reaches 1, the component fails.

[0122] Step 6: Calculate the reliability of the theoretical model and integrate it with the statistical model. Statistical analysis of the damage of each component is used to obtain the reliability of each component. The calculation principle is expressed as:

[0123]

[0124] Where: t represents time, the unit is h; n represents the total number of samples, D i is represented as the damage of the i-th sample, It is expressed as the reliability of the jth component at time t, and the unit is 1; I(x) is the indicator function, which is expressed as:

[0125]

[0126] Furthermore, the reliability of the whole machine mechanism model is obtained (such as Figure 6 ), can be expressed as:

[0127]

[0128] Where: t represents time, the unit is h; P{d p (t)<1} is the reliability of the whole machine mechanism model at time t, j represents the jth component, It is expressed as the reliability of the j-th component at time t, and the unit is 1.

[0129] Step 7: Calculate the reliability of the statistical component failure rate data. For common components whose failure rates are determined by reliability tests and data statistics, such as some seals and common parts, it can be assumed that the failure data of these components are obtained under the condition of normal operation of key components, and it is assumed that the failures of these components are independent of each other. After merging with the mechanism model, the reliability of the whole machine is obtained (such as Figure 7 The calculation principle is shown as:

[0130] R(t)=P{d p (t)<1,d d (t)<1}=P{d p (t)<1}P{d d (t)<1|d p (t)<1}

[0131] Where: R(t) is the reliability of the whole machine after the fusion of mechanism and data, unit is 1; t represents the time t, unit is h; where P{d d (t)<1|d p (t)<1} can be expressed as:

[0132]

[0133] Where: P{DD k (t)<1} is the reliability of the kth component of the data statistics, with the unit of 1, which can be expressed as:

[0134]

[0135] Where: k represents the kth component counted by the data, and the unit is 1; t represents the t time, η represents the η time, ξ represents the ξ time, and the unit is h; λ k (η) is the instantaneous failure rate of the kth component at time η, in units of h -1 .d p (t) and d d (t) are the norms of the damage vectors DP(t) and DD(t), respectively, which can be expressed as:

[0136] d p (t)=||DP(t)||

[0137] d d (t)=||DD(t)||

[0138] Where: DP(t) represents the damage of a part of the components calculated by the mechanism model, which is a random vector that varies with time; DD(t) represents the damage of another part of the components obtained by data statistics, which is also a random vector that varies with time.

[0139] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A diesel engine reliability prediction method based on mechanism and data fusion, characterized in that: The following steps are involved: Step 1: Input basic modeling parameters: including material property parameters and engine structure and performance parameters; Step 2: Engine performance model construction: To solve the change of state parameters in each subsystem with the crankshaft angle, the differential equations consisting of the simultaneous energy conservation equation, mass conservation equation, and ideal gas state equation are solved using the prediction and correction method; Step 3: Dynamic model construction: The cylinder pressure, temperature and other parameters obtained from the engine performance model are used as boundary conditions, and the loads of each component are calculated based on the kinematics and dynamics of the crank-connecting rod mechanism; Step 4: Wear and fatigue calculation of multiple samples of key components: Multiple samples are generated based on the Monte Carlo algorithm, the wear of the friction pair is calculated based on the Archard model, and the fatigue of the components is calculated based on the SN curve of the material and the fatigue cumulative damage theory; Step 5: Reliability analysis of the mechanism model: Statistical analysis is performed on the wear and fatigue of multiple samples of components to obtain the reliability of each component; further, the reliability of the whole machine mechanism model is obtained; Step 6: Analysis of whole machine reliability after integrating statistical model: Calculate the reliability of statistical component failure rate data and integrate it with the reliability calculated by mechanism model to obtain the integrated whole machine reliability.

2. The diesel engine reliability prediction method based on mechanism and data fusion according to claim 1 is characterized in that: The material property parameters described in step 1 include: Poisson's ratio of key component materials, elastic modulus of materials, hardness of materials, SN curve of crankshaft materials, SN curve of connecting rod materials, tensile strength of crankshaft materials, tensile strength of connecting rod materials; engine structure and performance parameters include heat release rate, valve timing, stroke, cylinder diameter, compression ratio, number of cylinders, number of volumes, number of valves, firing order, speed, power, size of each volume, connecting rod ratio, piston mass, connecting rod mass, crank unbalanced mass, connecting rod length, distance from connecting rod mass center to connecting rod big end center, crank rotation angular velocity, crank radius, piston diameter, distance from main journal mid-section to the mid-section of the nearest crank arm, crank arm width, crank arm thickness, crank pin cross-section diameter, main journal cross-section diameter, rod cross-sectional area, key friction pair contact length, key friction pair component wear coefficient, key friction pair assembly clearance, and key friction pair replacement limit clearance.

3. The diesel engine reliability prediction method based on mechanism and data fusion according to claim 1 is characterized in that: In step 2, the energy conservation equation is: Where: Indicates the crankshaft angle, in degrees; m z Indicates the mass of the working fluid, in kg; u z is the specific internal energy of the working medium in the cylinder, in J / kg; m in is the intake air mass, in kg; h i is the specific enthalpy of the intake air, in J / kg; Q f The heat brought by the fuel injected into the cylinder, the unit is J; Q w is the energy lost through heat transfer, in J; m o is the exhaust mass, in kg; h o is the specific enthalpy of the exhaust gas, in J / kg; p z is the pressure in the subsystem, in Pa; V z is the volume of the subsystem, in m 3 ; The mass conservation equation is: Where: Indicates the crankshaft angle, in degrees; m z Indicates the mass of the working fluid, in kg; m in is the intake air mass, in kg; m o is the exhaust mass, in kg; g f is the fuel mass in kg; x is the percentage of fuel combustion in the cylinder; The ideal gas state equation is expressed as: p z V z =m z R z T z Where: p z is the pressure in the subsystem, in Pa; V z is the volume of the subsystem, in m 3 ; m z Indicates the mass of the working fluid, in kg; T z is the temperature in the cylinder, in K; R z is the gas constant of the working fluid, with the unit being J / (kg·K).

4. The diesel engine reliability prediction method based on mechanism and data fusion according to claim 1 is characterized in that: In step three, the loads of the components can be expressed as: Where: F Σ is the force on the small end of the connecting rod along the center line of the cylinder, which can be decomposed into the piston lateral force F NG and connecting rod thrust F C , all in N; connecting rod thrust F C It can be further decomposed into the normal force F along the crank direction R and the tangential force F perpendicular to the crank direction T , the unit is N; F P is the force acting on the crank pin, in N; F K The force acting on the main shaft neck, unit is N; m p , m c and m K They are respectively expressed as piston mass, connecting rod mass and crank unbalanced mass, all in kg; the connecting rod mass is often equivalent to two parts, namely the reciprocating motion part m CA And rotary motion part m CB , the unit is kg; F g It is expressed as the gas pressure that the piston is subjected to when it reciprocates in the cylinder, and the unit is N; L cb is the distance from the center of mass of the connecting rod to the center of the big end of the connecting rod, in m; L is the distance between the centers of the big and small ends of the connecting rod, in m; ω is the angular velocity, in rad / s; R is the crank radius, in m; β is the connecting rod swing angle, is the crank angle, the unit is deg; λ is the ratio of crank radius to connecting rod length; D P is the piston diameter, in m; m r It represents the total mass concentrated at the crank pin and rotating with the crank, in kg, which can be expressed as: m r =m K +m CB Where: m K and m CB They are the unbalanced mass of the crank and the mass of the rotating part of the connecting rod, both in kg.

5. The diesel engine reliability prediction method based on mechanism and data fusion according to claim 1 is characterized in that: In step 4, the Monte Carlo algorithm is used to generate samples by considering the influence of uncertainty factors such as material properties, manufacturing assembly, external loads, etc. on the damage of the diesel engine, determining the distribution function of these uncertainty parameters, generating a series of random numbers, transforming these random numbers to obtain samples of the corresponding probability distribution function, substituting them into the corresponding uncertainty parameters, and generating a large number of random samples of the corresponding distribution function for each uncertainty parameter.

6. The diesel engine reliability prediction method based on mechanism and data fusion according to claim 1 is characterized in that: In the Archard wear model described in step 4, The wear depth of key friction pair components can be expressed as: Where: h is the wear depth, in m; l is the sliding distance of the friction pair, in m; K is the dimensionless wear coefficient; H is the Brinell hardness of the material, in HBW; p is the contact stress, in Pa, which can be expressed as: Where: F is the normal contact pressure, unit is N; S is the contact area, unit is m 2 ; Wear damage can be expressed as: Where: D w represents wear damage, the unit is 1; t represents time, the unit is h; h and h c are wear depth and critical wear depth, both in mm; when the wear damage reaches 1, the component fails; The number of cycles under the symmetrical cyclic stress amplitude of fatigue damage can be expressed as: lgN=a+blgσ -1 Where: N is the number of failure cycles corresponding to the stress amplitude, unit is 1; σ -1 It indicates the stress amplitude of the material under symmetrical loading with a stress ratio of -1, in MPa; a and b are material constants, obtained from the SN curve of the material. Since the force on the component is asymmetrical loading force, when calculating the fatigue life using the SN curve, the Goodman formula is usually used to correct the influence of the average stress. The Goodman formula can be expressed as: Where: b is the tensile strength limit of the material, σ m With σ a are the stress mean and stress amplitude under actual working conditions, σ -1 is the stress amplitude when the stress ratio is -1, the unit is MPa; Fatigue damage is the cumulative damage at all stress levels and can be expressed as: Where: D j represents fatigue damage, the unit is 1; t represents time, the unit is h; n i Indicates the stress amplitude σ of the component i The following has been completed i Cycles, unit is 1; N i It indicates the total number of cycles that the component can withstand under this stress amplitude, and the unit is 1; m indicates the number of different stress amplitudes, and the unit is 1; When the fatigue damage reaches 1, the component fails.

7. The diesel engine reliability prediction method based on mechanism and data fusion according to claim 1 is characterized in that: The reliability of each component is obtained by statistical analysis in step 5, and the calculation principle is expressed as follows: Where: t represents time, the unit is h; n represents the total number of samples, D i is represented as the damage of the i-th sample, It is expressed as the reliability of the jth component at time t, and the unit is 1; I(x) is the indicator function, which is expressed as: The reliability of the whole machine mechanism model can be expressed as: Where: t represents time, the unit is h; P{d p (t)<1} is the reliability of the whole machine mechanism model at time t, j represents the jth component, It is expressed as the reliability of the j-th component at time t, and the unit is 1.

8. The diesel engine reliability prediction method based on mechanism and data fusion according to claim 1 is characterized in that: The reliability of the component failure rate data calculated and statistically analyzed in step 6, for common components whose failure rates are determined by reliability tests and data statistics, can be considered that the failure data of these components are obtained under the condition of normal operation of key components, and assuming that the failures of these components are independent of each other, the reliability of these components is expressed as: Where: t represents time, the unit is h; d p (t) and d d (t) are the norms of the damage vectors DP(t) and DD(t), respectively, which can be expressed as: d p (t)=||DP(t)|| d d (t)=||DD(t)|| Where: DP(t) represents the damage of a part of the components calculated by the mechanism model, which is a random vector that varies with time; DD(t) represents the damage of another part of the components obtained by data statistics, which is also a random vector that varies with time; P{DD k (t)<1} is the reliability of the kth component of the data statistics, with the unit of 1, which can be expressed as: Where: k represents the kth component counted by the data, and the unit is 1; t represents the t time, η represents the η time, ξ represents the ξ time, and the unit is h; λ k (η) is the instantaneous failure rate of the kth component at time η, in units of h -1 ; The calculation principle of the integrated whole machine reliability is expressed as: R(t)=P{d p (t)<1,d d (t)<1}=P{d p (t)<1}P{d d (t)<1|d p (t)<1} Where: R(t) is the reliability of the whole machine after the fusion of mechanism and data, the unit is 1; t represents the time t, the unit is h; d p (t) and d d (t) are the norms of the damage vectors DP(t) and DD(t), respectively, which can be expressed as: d p (t)=||DP(t)|| d d (t)=||DD(t)|| Where: DP(t) represents the damage of a part of the components calculated by the mechanism model, which is a random vector that varies with time; DD(t) represents the damage of another part of the components obtained by data statistics, which is also a random vector that varies with time.

Images