A reliability optimization method for process parameters of key holes in marine diesel engine body

Abstract

The invention discloses a reliability optimization method for process parameters of key holes of a marine diesel engine fuselage. First, the main process parameters that affect the processing quality of the key holes of the diesel engine fuselage are determined, and then an approximate model between the main process parameters and the position accuracy of the key holes is established, the position accuracy of the key holes is output, and a nonlinear regression model between the main process parameters and the deformation of the holes is established. The deformation stability domain of the holes is analyzed, and the highest position accuracy of the key holes is taken as the optimization target. The probability of deformation exceeding the tolerance is less than the allowable value, the deformation stability domain, and the allowable range of the process parameters are taken as constraints to establish a reliability optimization model for the process parameters of the key holes of the fuselage. Finally, the reliability optimization model for the process parameters of the key holes of the fuselage is solved to obtain the process parameters after reliability optimization. The invention reduces the probability of uncertain defects in fuselage processing, and can quickly obtain the optimal process parameter combination for high-precision fuselage hole processing.

Description

Technical Field

[0001] The invention relates to a process parameter optimization method, in particular to a reliability optimization method for process parameters of a key hole system of a marine diesel engine body. Background Art

[0002] Marine diesel engines have an absolute leading position in marine engines due to their good economy, high thermal efficiency, good maneuverability, and wide application range. As the main propulsion device of ships, the cost of marine diesel engines generally accounts for 10% to 20% of the total ship price. Its structure is complex. Once a failure occurs during navigation, it is not only difficult to repair, but also may cause accidents and delay transportation. The quality of the marine diesel engine fuselage, as the assembly basis of the diesel engine, has a significant impact on the assembly and use of the diesel engine. The crankshaft hole, camshaft hole and cylinder hole are the key hole systems of the marine diesel engine fuselage. Their processing process is affected by the cutting process, and the position accuracy is difficult to guarantee. The processing quality of the key hole system of the fuselage directly affects the performance of the marine diesel engine. At present, the formulation of the processing technology of the fuselage hole system is often based on the personal experience of the process personnel. Generally, the process parameters are set based on the worst cutting conditions, which makes the hole system processing process in a conservative working state for a long time, limiting the performance of the processing equipment and causing substantial waste.

[0003] The existing technology is mainly based on the optimization of deterministic process parameters, without considering the uncertain defects in the processing process and lacking the support of reliability optimization of the processing process. The process parameters obtained by optimizing the process parameters may greatly increase the probability of defects in the processed object. Summary of the invention

[0004] Purpose of the invention: The present invention proposes a reliability optimization method for processing parameters for uncertain defects in the processing of key holes in a marine diesel engine body.

[0005] Technical solution: The reliability optimization method of the key hole system process parameters of a marine diesel engine body according to the present invention comprises:

[0006] (1) Analyze and determine the main process parameters that affect the machining quality of key holes in diesel engine bodies;

[0007] (2) establishing an approximate model between the main process parameters and the position accuracy of the key hole system, wherein the output of the approximate model is the position accuracy of the key hole system;

[0008] (3) Establish a nonlinear regression model between the main process parameters and the hole system deformation, calculate the probability of deformation exceeding the tolerance, use it as a reliability indicator, and analyze the deformation stability domain of the hole system;

[0009] (4) Taking the highest position accuracy of the key hole system as the optimization goal, the probability of deformation exceeding the tolerance less than the allowable value, the deformation stability domain, and the allowable range of process parameters as constraints to establish a reliability optimization model for the processing parameters of the key hole system of the fuselage;

[0010] (5) Solving the reliability optimization model of the processing parameters of the key holes of the fuselage to obtain the process parameters after reliability optimization.

[0011] Furthermore, the step (1) adopts an expert group judgment method, including:

[0012] For the i-th process parameter related to the processing of the fuselage hole system, when the expert judges that the process parameter has a significant relationship with the hole system position accuracy, x i =1; if there is no significant relationship, then x i =0, set weight ω for each expert according to their level i , calculate the weighted function value:

[0013]

[0014] When β i When ≥0.5, the i-th process parameter is determined as the main process parameter.

[0015] Furthermore, the process parameters related to the processing of the fuselage hole system include: cutting speed, cutting depth, feed rate, tool angle, and ambient temperature.

[0016] Furthermore, the step (2) comprises:

[0017] (21) sampling the main process parameter sample points to obtain an approximate model training sample set;

[0018] (22) A radial basis function neural network is used to establish an approximate model between the main process parameters and the position accuracy of the key hole system. The input of the approximate model is the main process parameters, and the output is the position accuracy of the key hole system. The approximate model training sample set is used to perform model training.

[0019] Preferably, the sample points in step (21) are sampled using the Latin hypercube method.

[0020] Furthermore, the step (3) comprises:

[0021] (31) The key hole system is simulated and a nonlinear exponential regression model between the main process parameters and the deformation of the key hole system is constructed, as shown in the following formula:

[0022]

[0023] Among them, θ is the hole position accuracy value, αi is the constant term coefficient, x i represents the main process parameters of hole system processing, i is the number of main process parameters;

[0024] (32) The sensitivity of the deformation of the key hole system is calculated according to the following formula:

[0025]

[0026] Among them, θ(x1,x2,x3,…,x i ) represents a nonlinear exponential regression model between the main process parameters and the deformation of the key hole system in step (31);

[0027] (33) The sensitivity variation ranges within different main process parameter ranges are calculated and compared, and the main process parameter range with the smallest sensitivity variation range is selected as the deformation stability domain.

[0028] Furthermore, the step (4) comprises:

[0029] Taking the established approximate model between the main process parameters and the key hole position accuracy as the objective function, taking the highest key hole position accuracy as the optimization target, taking the probability of deformation exceeding the tolerance less than the allowable probability as the reliability constraint, taking the deformation stability domain as the stability domain constraint, the process parameter reliability optimization model is established in the form of:

[0030] Find a, b, c…

[0031] Minθ ( a, b, c, ...)

[0032] st P(σ max ≤[σ])≥[P]

[0033] a min ≤a≤a max

[0034] b min ≤b≤b max

[0035] c min ≤c≤c max

[0036] Among them, a, b, c… are the main process parameters for hole system feature processing, θ is the position accuracy characteristic of the key hole system of the fuselage, P is the reliability of the hole system processing of the fuselage, σ max is the maximum deformation, [σ] is the allowable value of deformation, and [P] is the allowable reliability of fuselage hole processing.

[0037] Furthermore, the step (5) is solved by using a particle swarm algorithm and a Hooke-Jeeves algorithm, including the steps of:

[0038] First, the position and velocity of the particles in the particle swarm algorithm are initialized, the function value of each particle is calculated, and the individual extreme values ​​of the particles are stored to find the group extreme value. The particle swarm algorithm is used to locate the area where the target extreme value is located in the design space, and then the Hooke-Jeeves algorithm is used to accurately optimize the area. When the optimization accuracy reaches the target accuracy requirement, the best design result is obtained and the optimal solution, that is, the process parameters after reliability optimization, is output.

[0039] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: When optimizing the process parameters of the key hole system of the fuselage of a marine diesel engine, the present invention takes into account and limits the uncertain defects in the fuselage processing, while ensuring the accuracy of the key hole system of the fuselage, the probability of uncertain defects in the fuselage processing is reduced, and the optimal process parameter combination for the fuselage hole system processing is quickly obtained by solving the combined optimization algorithm, which can provide a theoretical basis for process personnel to formulate and improve the process parameters for the fuselage hole system processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a flow chart of the reliability optimization method of the key hole system process parameters of the marine diesel engine body of the present invention;

[0041] Figure 2 This is a structural diagram of a marine diesel engine fuselage according to an embodiment of the present invention;

[0042] Figure 3 It is a flow chart of a process for solving a reliability optimization model of processing parameters of a key hole system of a fuselage according to an embodiment of the present invention;

[0043] Figure 4 It is an optimization process diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0044] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0045] See also Figure 1 , which shows a flow chart of the reliability optimization method of the key hole system process parameters of a marine diesel engine body according to the present invention, taking the reliability optimization of the cylinder hole process parameters of a certain type of marine diesel engine body as an example, the schematic diagram of the marine diesel engine body is as follows Figure 2 As shown, A is the cylinder hole of the fuselage.

[0046] Step 1: Analyze and determine the main process parameters that affect the quality of diesel engine body cylinder bore machining

[0047] Five experts made significant judgments on the cutting speed, cutting depth, feed rate, ambient temperature, and tool angle in the machining of diesel engine cylinder holes, thereby determining the main process parameters for the machining of cylinder holes. When the experts judged that the process parameters had a significant relationship with the hole position accuracy, x i =1; if there is no significant relationship, then x i =0, set weight ω for each expert according to their level i , calculate the weighted function value:

[0048]

[0049] When β i When ≥0.5, the i-th process parameter is determined as the main process parameter.

[0050] The expert weights are shown in Table 1, and the judgment results are shown in Table 2.

[0051] Table 1

[0052] expert Weight 1 0.25 2 0.18 3 0.17 4 0.14 5 0.26

[0053] Table 2

[0054]

[0055] Through the expert group judgment method, it can be known that the main process parameters for cylinder bore machining are cutting speed, feed rate and ambient temperature, among which the cutting speed is the process parameter that has the greatest impact on cylinder bore machining.

[0056] Step 2: Establish an approximate model between the main process parameters and the verticality of the cylinder bore

[0057] (21) Latin hypercube was used to select the three process parameters of cutting speed, feed rate and ambient temperature, and the cylinder bore machining test was carried out based on the sample points designed by Latin hypercube experiment. The test results are shown in Table 3.

[0058] Table 3

[0059]

[0060]

[0061] (22) The radial basis function neural network is trained with the processing data as the training set, and the radial basis function is used as the "basis" of the hidden unit to construct the hidden layer space, and a radial basis function neural network model between the process parameters and the verticality of the cylinder bore is established.

[0062] Step 3: Establish a nonlinear regression model between the main process parameters and the cylinder bore deformation, calculate the probability of deformation exceeding the tolerance, use it as a reliability indicator, and analyze the stable domain of the cylinder bore deformation process parameters.

[0063] (31) Deform-3D software was used to calculate the cutting force and cutting temperature of the cylinder bore boring process under different cutting speeds, feed rates, and ambient temperatures.

[0064] (32) The diesel engine body model was meshed using Hypermesh software, and the meshed body was imported into ABAQUS and its material was defined as QT500. The clamping force and its own gravity were applied to the body, the cutting force was applied to the cylinder bore in the form of a static load, and the cutting temperature was applied to the key hole system in the form of a predefined temperature field.

[0065] (33) Based on the Taguchi design scheme, ABAQUS was used to solve the maximum deformation of the cylinder bore under different process parameters. The results are shown in Table 4.

[0066] serial number <![CDATA[v c (rpm)]]> f(mm / r) T(℃) Deformation 1 99.852 0.4 15 0.03140 2 99.852 0.5 20 0.03469 3 99.852 0.6 25 0.03653 4 133.136 0.4 20 0.04868 5 133.136 0.5 25 0.05058 6 133.136 0.6 15 0.04967 7 166.42 0.4 25 0.05592 8 166.42 0.5 15 0.05482 9 166.42 0.6 20 0.05513

[0067] (34) Taking the process parameters as variables and the deformation as the response, a nonlinear regression was performed on Table 1, and the nonlinear exponential regression model between the main process parameters and the deformation of the key hole system was obtained as follows:

[0068] Q=2.3*10 -4 *v c 0.89057 f 0.0922127 T 0.0974116

[0069] (35) From step (31), it can be seen that the cutting speed has the greatest impact on the cylinder bore deformation. Substituting the nonlinear regression model between the process parameters and the deformation into the sensitivity formula, the sensitivity model of the cylinder bore deformation to the cutting speed is obtained as follows:

[0070]

[0071] Based on the cutting speed intervals divided by Taguchi design in Table 4, the sensitivity amplitudes in different cutting speed intervals are calculated. It is calculated that the sensitivity change amplitude in the cutting speed interval of 99.852-133.136 m / min is 0.00000912, and the sensitivity change amplitude in the cutting speed interval of 133.16-166.42 m / min is 0.0000069. The interval of 133.16-166.42 m / min with a smaller sensitivity amplitude is selected as the sensitivity interval of the cutting speed.

[0072] Step 4: Taking the highest position accuracy of the key hole system as the optimization goal, the probability of deformation exceeding the tolerance less than the allowable value, the deformation stability domain, and the allowable range of process parameters as constraints to establish a reliability optimization model for the processing parameters of the key hole system of the fuselage.

[0073] (41) The cylinder bore machining process parameters are optimized, and the main focus is on minimizing the verticality of the cylinder bore relative to the crankshaft hole. Based on this and the obtained constraints, a cylinder bore process parameter optimization model for a marine diesel engine body can be constructed:

[0074] Find v c ,f,T

[0075] Minθ c( v c , f, T)

[0076] st P(Q max ≤0.05)≥0.9

[0077] 133.16m / min≤v c ≤166.42m / min

[0078] 0.4mm / r≤f≤0.6mm / r

[0079] 15℃≤T≤25℃

[0080] Among them, v c ,f,T are cutting speed, feed rate and ambient temperature respectively, θ c is the verticality of the cylinder bore relative to the crankshaft bore. Based on experience, the allowable deformation of the cylinder bore is set to 0.05 mm. max ≤0.05) indicates the probability that the maximum deformation of the cylinder bore is less than the allowable deformation.

[0081] Step 5: Solve the reliability optimization model of the processing parameters of the key holes of the fuselage

[0082] like Figure 3 As shown in the figure, the particle swarm algorithm is used to locate the area where the extreme value of cylinder bore verticality is located in the design space, and then the Hooke-Jeeves algorithm is used to accurately optimize the area, and finally the best design result is obtained. The particle swarm algorithm parameters are set as follows: inertia factor ω = 0.9, individual learning factor and environmental learning factor C1 = C2 = 2, number of particles is 40, maximum number of iterations is 200, and maximum flight speed is 100. The particle swarm algorithm parameters based on the Hooke-Jeeves algorithm are set as follows: initial step size δ = 0.5, reduction rate β = 0.5, acceleration factor α = 1, precision ε = 10 -6 , and the other parameters are the same as those of the particle swarm algorithm. The process of finding the optimal solution for cylinder bore verticality is shown in the attached figure. Figure 4 shown.

[0083] The optimal combination of process parameters for cylinder bore machining was found to be: cutting speed of 101.18 m / min, feed rate of 0.46 mm / r, and ambient temperature of 16.3 °C.

Claims

1. A reliability optimization method for process parameters of key holes of a marine diesel engine body, characterized in that: Includes steps: (1) Analyze and determine the main process parameters that affect the machining quality of key holes in diesel engine bodies; (2) establishing an approximate model between the main process parameters and the position accuracy of the key hole system, wherein the output of the approximate model is the position accuracy of the key hole system; (3) Establish a nonlinear exponential regression model between the main process parameters and the hole deformation, calculate the probability of deformation exceeding the tolerance, use it as a reliability indicator, and analyze the deformation stability domain of the hole system; including: (31) The key hole system is simulated and a nonlinear exponential regression model between the main process parameters and the deformation of the key hole system is constructed, as shown in the following formula: Among them, θ is the position accuracy value of the key hole system of the fuselage, α i is the constant term coefficient, x i represents the main process parameters of hole system processing, i is the number of main process parameters; (32) The sensitivity of the deformation of the key hole system is calculated according to the following formula: Among them, θ(x1,x2,x3,...,x i ) represents a nonlinear exponential regression model between the main process parameters and the deformation of the key hole system in step (31); (33) Calculate the sensitivity variation ranges within different main process parameter ranges for comparison, and select the main process parameter range with the smallest sensitivity variation range as the deformation stability domain; (4) Taking the highest position accuracy of the key hole system as the optimization goal, the probability of deformation exceeding the tolerance less than the allowable value, the deformation stability domain, and the allowable range of process parameters as constraints to establish a reliability optimization model for the processing parameters of the key hole system of the fuselage; including: Taking the established approximate model between the main process parameters and the key hole position accuracy as the objective function, taking the highest key hole position accuracy as the optimization target, taking the probability of deformation exceeding the tolerance less than the allowable probability as the reliability constraint, taking the deformation stability domain as the stability domain constraint, the process parameter reliability optimization model is established in the form of: Find a, b, c… Minθ ( a,b,c,…) st P(σ max ≤[σ])≥[P] a min ≤a≤a max b min ≤b≤b max c min ≤c≤c max Among them, a, b, c… are process parameters in the stable deformation domain, θ is the position accuracy value of the key hole system of the fuselage, P is the reliability of the fuselage hole system processing, σ max is the maximum deformation, [σ] is the allowable value of deformation, and [P] is the allowable reliability of fuselage hole processing); (5) Solving the reliability optimization model of the processing parameters of the key holes of the fuselage to obtain the process parameters after reliability optimization.

2. The reliability optimization method for process parameters of key holes of a marine diesel engine body according to claim 1 is characterized in that: The step (1) adopts an expert group judgment method, including: For the i-th process parameter related to the processing of the fuselage hole system, when the expert judges that the process parameter has a significant relationship with the hole system position accuracy, x i =1; if there is no significant relationship, then x i =0, set weight ω for each expert according to their level i , calculate the weighted function value: When β i When ≥0.5, the i-th process parameter is determined as the main process parameter.

3. The reliability optimization method for process parameters of key holes of a marine diesel engine body according to claim 2 is characterized in that: The process parameters related to the processing of the fuselage hole system include: cutting speed, cutting depth, feed rate, tool angle, and ambient temperature.

4. The reliability optimization method for process parameters of key holes of a marine diesel engine body according to claim 1 is characterized in that: The step (2) comprises: (21) sampling the main process parameter sample points to obtain an approximate model training sample set; (22) A radial basis function neural network is used to establish an approximate model between the main process parameters and the position accuracy of the key hole system. The input of the approximate model is the main process parameters, and the output is the position accuracy of the key hole system. The approximate model training sample set is used to perform model training.

5. The reliability optimization method for process parameters of key holes of a marine diesel engine body according to claim 4 is characterized in that: The sample points in step (21) are sampled using the Latin hypercube method.

6. The reliability optimization method for process parameters of key holes of a marine diesel engine body according to claim 1 is characterized in that: The step (5) is solved by using a particle swarm algorithm and a Hooke-Jeeves algorithm, and includes the following steps: First, the position and velocity of the particles in the particle swarm algorithm are initialized, the function value of each particle is calculated, and the individual extreme values ​​of the particles are stored to find the group extreme value. The particle swarm algorithm is used to locate the area where the target extreme value is located in the design space, and then the Hooke-Jeeves algorithm is used to accurately optimize the area. When the optimization accuracy reaches the target accuracy requirement, the best design result is obtained and the optimal solution, that is, the process parameters after reliability optimization, is output.

Images