Optimization method for complex equipment forward design considering multi-source uncertainties

Abstract

The present invention discloses a method for optimizing complex equipment forward design schemes considering multi-source uncertainty. The method comprises the following steps: establishing three objective functions of the design scheme, then constructing a feasible complex equipment forward design scheme using the principle solution, and mapping with the three objective functions to generate a complex equipment design scheme morphological matrix under multi-source uncertainty, thereby establishing an original model; then transforming the three objective functions to obtain a deterministic model, using an improved multi-source uncertainty non-dominated solution genetic algorithm to solve the model and obtain an optimal design scheme set; optimizing using an intuitive fuzzy λ-Shapely Choquet integral method to obtain an optimal complex equipment forward design scheme. The present invention can not only efficiently obtain a complex equipment forward design scheme set, but also solve the problem that the existing design scheme optimization model has not yet simultaneously considered the influence of random, interval and fuzzy multi-source uncertainties, resulting in a lack of rationality in the design scheme.

Description

Technical Field

[0001] The present invention relates to a method for optimizing a complex equipment forward design scheme, and in particular to a method for optimizing a complex equipment forward design scheme taking multi-source uncertainties into consideration. Background Art

[0002] Complex equipment specifically refers to systems or products with a large product structure, multidisciplinary knowledge, a large number of parts, high technical content, and high integration, such as machine tools, ships, gas turbines, etc. Therefore, the research and development of complex equipment has great design difficulty, complex and diverse design requirements, designers' thinking is often characterized by jumping and discontinuity, and design information is often vague, abstract, and incomplete, resulting in product development requiring the search for feasible design solutions that meet technical performance in a highly nonlinear and high-dimensional design space. The entire process consumes a lot of resources and time. In order to meet the requirements for the size and precision of the products processed by the polishing machine and reduce the propagation of uncertainty in the design, people have proposed a series of methods. For example, Jiang used the FBS model in the paper "Research on the Propagation Path of Function Change in Product Conceptual Design" in "Lancaster: Destech Publications" in 2016 to determine the functional changes and propagation paths of the product during the design process, and quickly solve the optimal design solution. Huang applied a variety of computational intelligence methods to study the generation technology of product design solutions that meet customer needs in his paper "New evaluation methods for conceptual design selection using computational intelligence techniques" in Journal of Mechanical Science and Technology (vol. 27, no. 3, pp. 733-746) in 2013. Ma proposed a systematic decision making approach for product conceptual design based on fuzzy morphological matrix in his paper "A systematic decision making approach for product conceptual design based on fuzzy morphological matrix" in Expert Systems with Applications (vol. 81, pp. 444-456) in 2017. The method uses the subjective uncertainty knowledge and preferences of engineers and customers to quantitatively evaluate the functional solution principle. Summary of the invention

[0003] In order to solve the problem of design reciprocating cycles caused by multi-source uncertainties faced in the design stage of complex equipment, the purpose of the present invention is to provide a method for optimizing the forward design scheme of complex equipment taking into account multi-source uncertainties.

[0004] The technical solution adopted by the present invention comprises the following steps:

[0005] 1) Considering the uncertainty in the optimization process of complex equipment, the three uncertainty variables of each functional module in the forward design scheme of complex equipment are determined. The three uncertainty variables are module performance parameter P, configuration cost C and confidence S. Then, based on the three uncertainty variables of each functional module in the design scheme, the performance parameter responsiveness function, interval cost function and fuzzy confidence function of the design scheme are established respectively;

[0006] 2) Use principle analysis to construct feasible forward design solutions for complex equipment PS s The three uncertain variables of the functional module design scheme and the feasible complex equipment forward design scheme PS are stored in the computer through feasible media. s Mapping is performed to generate a morphological matrix of complex equipment design solutions under multi-source uncertainty;

[0007] 3) Comprehensively considering the three objectives of the design scheme, namely, the minimum total configuration cost, the maximum total confidence, and the optimal total module performance parameters, an optimization model of the original complex equipment forward design scheme principle is established according to the three objective functions of the design scheme in step 1) and the morphological matrix of the complex equipment design scheme under multi-source uncertainty;

[0008] 4) The interval cost function and the fuzzy confidence function are respectively converted into a deterministic cost function and a deterministic confidence function, and the original complex equipment forward design scheme principle solution optimization model is updated based on the deterministic cost function and the deterministic confidence function to obtain the deterministic complex equipment forward design scheme principle solution optimization model, and the deterministic complex equipment forward design scheme principle solution optimization model is solved using an improved multi-source uncertain non-dominated solution genetic algorithm to obtain the optimal complex equipment forward design scheme data set;

[0009] 5) The intuitive fuzzy λ-Shapely Choquet integral method is used to optimize the optimal complex equipment forward design scheme data set to obtain the optimal complex equipment forward design scheme.

[0010] The formula of the performance parameter responsiveness function of the design scheme in step 1) is as follows:

[0011]

[0012]

[0013] Where G(P) represents the performance parameter responsiveness function value of the design solution, N is the number of functional modules, J is the number of performance credibility of each functional module, and w ik is the kth performance credibility weight of the ith functional module, p ik is the reliability of the performance parameters of the random solution satisfying the kth performance credibility of the ith functional module, X = (X1, X2, ..., X m ) T is an m×1-dimensional random solution performance parameter vector, X∈f X (x), f X (x) is the joint probability density function, μ i , σ i are the i-th performance parameter variables X of the random scheme respectively. i The mean and standard deviation of

[0014] The formula of the interval cost function of the design scheme is as follows:

[0015]

[0016] in, represents the interval cost function value of the design solution, M is the number of class modules of the i-th functional module, ε il It is a binary decision variable indicating whether the lth class module of the i-th functional module exists during the design scheme configuration, represented by 0 or 1, m il is the number of uses of the lth class module of the ith functional module, is the configuration cost of the lth class module of the ith functional module, expressed in intervals express, is the minimum configuration cost of the lth class module of the ith functional module, is the maximum configuration cost of the lth class module of the i-th functional module;

[0017] The formula of the fuzzy confidence function of the design scheme is as follows:

[0018]

[0019]

[0020]

[0021]

[0022] in, represents the fuzzy confidence function value of the design solution. The confidence S is a triangular fuzzy set based on the language S. ilis the confidence of the lth class module of the ith functional module, expressed as a triangular fuzzy number, satisfying represents the lower bound of the confidence of the lth class module of the i-th functional module, represents the median confidence value of the lth class module of the i-th functional module, It represents the upper bound of the confidence of the lth class module of the i-th functional module, max() represents the maximum value operation, and min() represents the minimum value operation.

[0023] The feasible complex equipment forward design scheme PS in step 2) s The formula is as follows:

[0024]

[0025] Among them, n i The number of function module primitives that meet the function of the i-th function module.

[0026] The calculation formula of the complex equipment forward design scheme principle optimization model established in step 3) is as follows:

[0027]

[0028]

[0029]

[0030]

[0031] Among them, G represents the original solution of the forward design scheme of the original complex equipment, maxG(P) represents the optimal total module performance parameter of the design scheme, represents the original maximum total confidence of the design solution, represents the original minimum total configuration cost of the design solution, B max is the highest total configuration cost of the design.

[0032] In step 4), the formula of the deterministic cost function of the design solution is as follows:

[0033]

[0034]

[0035] Among them, G(C) represents the deterministic cost function value, ρ AB is the configuration cost judgment coefficient between the B-th class module and the A-th class module in the j-th functional module, indicating that the configuration cost of the B-th class module in the j-th functional module is better than the configuration cost of the A-th class module in the j-th functional module.L is the minimum configuration cost of the A-th class module, A U is the highest configuration cost of the A-th class module, B L represents the minimum configuration cost of the Bth class module, B U represents the maximum configuration cost of the Bth class module; M is the number of class modules of the jth functional module;

[0036] The formula for the deterministic confidence function of the design solution is as follows:

[0037]

[0038]

[0039] Among them, G(S) represents the confidence response function value, E η represents the expected function;

[0040] The formula of the principle optimization model of the deterministic complex equipment forward design scheme is as follows:

[0041] G′=(max(G(P)),max(G(S)),min(G(C)))

[0042] Among them, G′ represents the principle solution of the forward design scheme of deterministic complex equipment, maxG(P) represents the optimal total module performance parameter of the design scheme, max(G(S)) represents the maximum total confidence of the design scheme, and min(G(C)) represents the minimum total configuration cost of the design scheme.

[0043] In step 4), in the improved multi-source uncertain non-dominated solution genetic algorithm, the population is initialized as a set of the number of solution schemes of the class modules of each functional module, and the dimension d of each chromosome is expanded to 3. The three dimensions of each chromosome represent the module performance parameter P, configuration cost C and confidence S of each functional module, respectively. The coding structure of the multi-source uncertain non-dominated solution genetic algorithm is established as a coding structure between chromosomes, functional modules and uncertainty variables, so as to solve the principle optimization model of the forward design scheme of deterministic complex equipment.

[0044] The present invention has the following beneficial effects:

[0045] 1) Based on triangular fuzzy numbers, interval numbers and probabilities, the fuzzy, interval and random uncertainties in the complex product design process are quantified, and the multi-source uncertainty measurement in the complex product design process is effectively realized.

[0046] 2) The improved multi-source uncertain non-dominated solution genetic algorithm can effectively prevent the combinatorial explosion problem generated by morphological matrices and the difficulty of solving uncertain expressions.

[0047] 3) The intuitive fuzzy λ-Shapely Choquet integral method is used to optimize the obtained fuzzy non-dominated solutions to obtain the optimal product design solution. It minimizes the uncertainty caused by insufficient knowledge and has great practical significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a flow chart of the present invention.

[0049] Figure 2 This is the schematic diagram of the aspheric optical polishing machine.

[0050] Figure 3 It is to solve the evolutionary iterative process and the Pareto optimal solution diagram.

[0051] Figure 4 It is a design sketch of optical polishing machine.

[0052] Figure 5 It is the optimal optical grinding and polishing machine product design result diagram. DETAILED DESCRIPTION

[0053] The present invention will be further described below by taking an aspheric optical polishing machine among complex equipment as an example.

[0054] like Figure 1 As shown, the present invention comprises the following steps:

[0055] 1) Considering the uncertainty in the optimization process of complex equipment, the three uncertainty variables of each functional module in the forward design scheme of complex equipment are determined. The three uncertainty variables are module performance parameter P, configuration cost C and confidence S. Then, based on the three uncertainty variables of each functional module in the design scheme, the performance parameter responsiveness function, interval cost function and fuzzy confidence function of the design scheme are established respectively;

[0056] The formula of the performance parameter responsiveness function of the design scheme in step 1) is as follows:

[0057]

[0058]

[0059] Where G(P) represents the performance parameter responsiveness function value of the design solution, N is the number of functional modules, J is the number of performance credibility of each functional module, and w ik is the kth performance credibility weight of the ith functional module, p ik is the reliability of the performance parameters of the random solution satisfying the kth performance credibility of the ith functional module, X = (X1, X2, ..., X m ) Tis an m×1-dimensional random solution performance parameter vector, X∈f X (x), f X (x) is the joint probability density function, μ i , σ i are the i-th performance parameter variables X of the random scheme respectively. i The mean and standard deviation of

[0060] The formula for the interval cost function of the design solution is as follows:

[0061]

[0062] in, represents the interval cost function value of the design solution, M is the number of class modules of the i-th functional module, ε il It is a binary decision variable indicating whether the lth class module of the i-th functional module exists during the design scheme configuration, represented by 0 or 1, m il is the number of uses of the lth class module of the ith functional module, is the configuration cost of the lth class module of the ith functional module, expressed in intervals Indicates that the attribute value of the configuration cost C is an interval number, is the minimum configuration cost of the lth class module of the ith functional module, is the maximum configuration cost of the lth class module of the i-th functional module;

[0063] The formula of the fuzzy confidence function of the design solution is as follows:

[0064]

[0065]

[0066]

[0067]

[0068] in, represents the fuzzy confidence function value of the design solution, and the confidence S satisfies S=(s1,s2,…,s n ), is a triangular fuzzy set based on language S, n is the granularity of the triangular fuzzy set based on language S, S il is the confidence of the lth class module of the ith functional module, expressed as a triangular fuzzy number, satisfying represents the lower bound of the confidence of the lth class module of the i-th functional module, represents the median confidence value of the lth class module of the i-th functional module, It represents the upper bound of the confidence of the lth class module of the i-th functional module, max() represents the maximum value operation, and min() represents the minimum value operation.

[0069] 2) Use principle analysis to construct feasible forward design solutions for complex equipment PS s The three uncertain variables of the functional module design scheme and the feasible complex equipment forward design scheme PS are stored in the computer through feasible media. s Mapping is performed to generate a morphological matrix of complex equipment design solutions under multi-source uncertainty;

[0070] The feasible complex equipment forward design scheme PS in step 2) s The formula is as follows:

[0071]

[0072] Among them, n i The number of function module primitives that meet the function of the i-th function module.

[0073] 3) Comprehensively considering the three objectives of the design scheme, namely, the minimum total configuration cost, the maximum total confidence, and the optimal total module performance parameters, an optimization model of the original complex equipment forward design scheme principle is established according to the three objective functions of the design scheme in step 1) and the morphological matrix of the complex equipment design scheme under multi-source uncertainty;

[0074] The calculation formula of the complex equipment forward design scheme principle optimization model established in step 3) is as follows:

[0075]

[0076]

[0077]

[0078]

[0079] Among them, G represents the original solution of the forward design scheme of the original complex equipment, maxG(P) represents the optimal total module performance parameter of the design scheme, represents the original maximum total confidence of the design solution, represents the original minimum total configuration cost of the design solution, B max is the highest total configuration cost of the design.

[0080] 4) The interval cost function and the fuzzy confidence function are converted into deterministic cost function and deterministic confidence function respectively. Based on the deterministic cost function and the deterministic confidence function, the original complex equipment forward design scheme principle solution optimization model is updated to obtain the deterministic complex equipment forward design scheme principle solution optimization model. The deterministic complex equipment forward design scheme principle solution optimization model is solved using the improved multi-source uncertain non-dominated solution genetic algorithm in MATLAB to find the global optimal solution of the deterministic design scheme principle solution optimization model. Finally, the Pareto frontier solution set is obtained and used as the optimal complex equipment forward design scheme set based on Pareto, thereby obtaining the optimal complex equipment forward design scheme data set;

[0081] The module performance parameters are probability values, so the performance parameter responsiveness function does not need to be converted. The configuration cost C and confidence are interval numbers and triangular fuzzy numbers, respectively, so the interval cost function and fuzzy confidence function need to be converted into attribute values. In step 4), the formula of the deterministic cost function is as follows:

[0082]

[0083]

[0084] Among them, G(C) represents the deterministic cost function value, ρ AB is the configuration cost judgment coefficient between the B-th class module and the A-th class module in the j-th functional module, indicating that the configuration cost of the B-th class module in the j-th functional module is better than the configuration cost of the A-th class module in the j-th functional module. L is the minimum configuration cost of the A-th class module, A U is the highest configuration cost of the A-th class module, B L represents the minimum configuration cost of the Bth class module, B U represents the maximum configuration cost of the Bth class module; M is the number of class modules of the jth functional module;

[0085] The formula for the confidence response function is as follows:

[0086]

[0087]

[0088] Among them, G(S) represents the confidence response function value, E η represents the expected function;

[0089] The formula of the principle optimization model for the forward design of deterministic complex equipment is as follows:

[0090] G′=(max(G(P)),max(G(S)),min(G(C)))

[0091] Among them, G′ represents the principle solution of the forward design scheme of deterministic complex equipment, maxG(P) represents the optimal total module performance parameter of the design scheme, max(G(S)) represents the maximum total confidence of the design scheme, and min(G(C)) represents the minimum total configuration cost of the design scheme.

[0092] In step 4), in the improved multi-source uncertain non-dominated solution genetic algorithm, the population is initialized as a set of the number of solution schemes of the class modules of each functional module, and the dimension d of each chromosome is expanded to 3. The three dimensions of each chromosome represent the module performance parameter P, configuration cost C and confidence S of each functional module, respectively. The coding structure of the multi-source uncertain non-dominated solution genetic algorithm is established as a coding structure between chromosomes, functional modules and uncertainty variables, so as to solve the principle optimization model of the forward design scheme of deterministic complex equipment.

[0093] 5) The optimal complex equipment forward design scheme data set is optimized using the intuitionistic fuzzy λ-Shapely Choquet integral method to obtain the optimal complex equipment forward design scheme; in the specific implementation, the complex equipment is manufactured according to the optimal complex equipment forward design scheme.

[0094] The mechanical system components of the optical polishing machine design stage are as follows: Figure 2 As shown in Table 1, the distribution of uncertainty variables (P, S, C) in each system is shown in Table 1. Figure 2 The specific functional analysis results of the system composition shown in Table 2 are shown in Table 2. The performance credibility c corresponding to each functional module is obtained by the hierarchical analysis method σi and weight w σi As shown in Table 3. The uncertainty analysis of the morphological matrix of the optical polishing machine drive unit is carried out, and the uncertainty variable results are shown in Table 4. The design scheme principle optimization model is solved by the improved multi-source uncertain non-dominated solution genetic algorithm in matlab2019b. Assume that the internal population size N = 100 and the maximum evolutionary generation G max =200, the fitness function is the set objective function, the crossover probability P c =0.9, mutation probability P m =0.05,ε ik =m ik =1, the termination condition is the maximum evolutionary generation, and the optimal optical polishing machine design scheme set is obtained according to the parameters in Table 3 and Table 4. The evolutionary iteration process and Pareto optimal solution are solved as follows: Figure 3As shown in (a) and (b), the corresponding set of eight Pareto optimal solution representative results are shown in Table 5, and the corresponding optical polishing machine design sketch is shown in Figure 4 As shown, Figure 4 (a), (b) and (c) are the first scheme U1, the second scheme U2 and the eighth scheme U8 respectively.

[0095] Table 1

[0096]

[0097] Table 2

[0098]

[0099] Table 3

[0100]

[0101] Table 4

[0102]

[0103]

[0104] Table 5

[0105] Serial number plan combination 1 <![CDATA[U1]]> <![CDATA[FS A3 +FS B4 +FS C6 +FS D1 +FS E3 +FS F3 ]]> 2 <![CDATA[U2]]> <![CDATA[FS A3 +FS B2 +FS C1 +FS D4 +FS E3 +FS F3 ]]> 3 <![CDATA[U3]]> <![CDATA[FS A6 +FS B2 +FS C6 +FS D1 +FS E3 +FS F3 ]]> 4 <![CDATA[U4]]> <![CDATA[FS A3 +FS B4 +FS C2 +FS D4 +FS E3 +FS F2 ]]> 5 <![CDATA[U5]]> <![CDATA[FS A3 +FS B4 +FS C5 +FS D1 +FS E1 +FS F3 ]]> 6 <![CDATA[U6]]> <![CDATA[FS A6 +FS B2 +FS C1 +FS D2 +FS E3 +FS F1 ]]> 7 <![CDATA[U7]]> <![CDATA[FS A5 +FS B4 +FS C1 +FS D2 +FS E2 +FS F3 ]]> 8 <![CDATA[U8]]> <![CDATA[FS A3 +FS B3 +FS C4 +FS D2 +FS E3 +FS F1 ]]>

[0106] The intuitionistic fuzzy λ-Shapely Choquet integral method is used to optimize the optimal complex equipment forward design scheme set of the optical polishing machine, and the intuitionistic fuzzy decision matrix A of the optical polishing machine is obtained as follows:

[0107]

[0108] Based on the intuitionistic fuzzy decision matrix A of the optical polishing machine, the optimal design scheme of the optical polishing machine is obtained. The results are shown in Table 8. It can be seen from the table Therefore, the optimal design of the drive mechanism unit of the optical polishing machine is U2. The corresponding design results are as follows Figure 5 shown.

[0109] Table 8

[0110] <![CDATA[d i (A i ,b + )]]> <![CDATA[d i (A i ,b - )]]> <![CDATA[U - ]]> <![CDATA[U1]]> 0.339 0.751 0.6890 <![CDATA[U2]]> 0.341 0.757 0.6894 <![CDATA[U3]]> 0.510 0.645 0.5584 <![CDATA[U4]]> 0.492 0.740 0.6006 <![CDATA[U5]]> 0.385 0.699 0.6448 <![CDATA[U6]]> 0.437 0.649 0.5976 <![CDATA[U7]]> 0.544 0.610 0.5286 <![CDATA[U8]]> 0.431 0.688 0.6148

Claims

1. A method for optimizing complex equipment forward design schemes considering multi-source uncertainties, characterized in that: The following steps are involved: 1) Considering the uncertainty in the optimization process of complex equipment, the three uncertainty variables of each functional module in the forward design scheme of complex equipment are determined. The three uncertainty variables are module performance parameter P, configuration cost C and confidence S. Then, based on the three uncertainty variables of each functional module in the design scheme, the performance parameter responsiveness function, interval cost function and fuzzy confidence function of the design scheme are established respectively; 2) Use principle analysis to construct feasible forward design solutions for complex equipment PS s , the three uncertain variables of the functional module design scheme are combined with the feasible complex equipment forward design scheme PS s Mapping is performed to generate a morphological matrix of complex equipment design solutions under multi-source uncertainty; 3) Comprehensively considering the three objectives of the design scheme, namely, the minimum total configuration cost, the maximum total confidence, and the optimal total module performance parameters, an optimization model of the original complex equipment forward design scheme principle is established according to the three objective functions of the design scheme in step 1) and the morphological matrix of the complex equipment design scheme under multi-source uncertainty; 4) The interval cost function and the fuzzy confidence function are respectively converted into a deterministic cost function and a deterministic confidence function, and the original complex equipment forward design scheme principle solution optimization model is updated based on the deterministic cost function and the deterministic confidence function to obtain the deterministic complex equipment forward design scheme principle solution optimization model, and the deterministic complex equipment forward design scheme principle solution optimization model is solved using an improved multi-source uncertain non-dominated solution genetic algorithm to obtain the optimal complex equipment forward design scheme data set; 5) The intuitive fuzzy λ-Shapely Choquet integral method is used to optimize the optimal complex equipment forward design scheme data set to obtain the optimal complex equipment forward design scheme.

2. The method for optimizing complex equipment forward design considering multi-source uncertainty according to claim 1, characterized in that: The formula of the performance parameter responsiveness function of the design scheme in step 1) is as follows: Where G(P) represents the performance parameter responsiveness function value of the design solution, N is the number of functional modules, J is the number of performance credibility of each functional module, and w ik is the kth performance credibility weight of the ith functional module, p ik is the reliability of the performance parameters of the random solution satisfying the kth performance credibility of the ith functional module, X = (X1, X2, ..., X m ) T is an m×1-dimensional random solution performance parameter vector, X∈f X (x), f X (x) is the joint probability density function, μ i , σ i are the i-th performance parameter variables X of the random scheme respectively. i The mean and standard deviation of The formula of the interval cost function of the design scheme is as follows: in, represents the interval cost function value of the design solution, M is the number of class modules of the i-th functional module, ε il It is a binary decision variable indicating whether the lth class module of the i-th functional module exists during the design scheme configuration, represented by 0 or 1, m il is the number of uses of the lth class module of the ith functional module, is the configuration cost of the lth class module of the ith functional module, expressed in intervals express, is the minimum configuration cost of the lth class module of the ith functional module, is the maximum configuration cost of the lth class module of the i-th functional module; The formula of the fuzzy confidence function of the design scheme is as follows: in, represents the fuzzy confidence function value of the design solution. The confidence S is a triangular fuzzy set based on the language S. il is the confidence of the lth class module of the ith functional module, expressed as a triangular fuzzy number, satisfying represents the lower bound of the confidence of the lth class module of the i-th functional module, represents the median confidence value of the lth class module of the i-th functional module, It represents the upper bound of the confidence of the lth class module of the i-th functional module, max() represents the maximum value operation, and min() represents the minimum value operation.

3. The method for optimizing complex equipment forward design considering multi-source uncertainty according to claim 1, characterized in that: The feasible complex equipment forward design scheme PS in step 2) s The formula is as follows: Among them, n i The number of function module primitives that meet the function of the i-th function module.

4. The method for optimizing complex equipment forward design considering multi-source uncertainty according to claim 1, characterized in that: The calculation formula of the complex equipment forward design scheme principle optimization model established in step 3) is as follows: Among them, G represents the original solution of the forward design scheme of the original complex equipment, maxG(P) represents the optimal total module performance parameter of the design scheme, represents the original maximum total confidence of the design solution, represents the original minimum total configuration cost of the design solution, B max is the highest total configuration cost of the design; w ik is the kth performance credibility weight of the ith functional module, p ik is the reliability of the performance parameter of the random solution satisfying the kth performance credibility of the i-th functional module, N is the number of functional modules, J is the number of performance credibility of each functional module, M is the number of class modules of the i-th functional module, and ε il It is a binary decision variable indicating whether the lth class module of the i-th functional module exists during the design scheme configuration, represented by 0 or 1, m il is the number of uses of the lth class module of the ith functional module, is the configuration cost of the lth class module of the ith functional module, expressed in intervals express, is the minimum configuration cost of the lth class module of the ith functional module, is the maximum configuration cost of the lth class module of the ith functional module.

5. The method for optimizing complex equipment forward design considering multi-source uncertainty according to claim 1, characterized in that: In step 4), the formula of the deterministic cost function of the design solution is as follows: Among them, G(C) represents the deterministic cost function value, ρ AB is the configuration cost judgment coefficient between the B-th class module and the A-th class module in the j-th functional module, indicating that the configuration cost of the B-th class module in the j-th functional module is better than the configuration cost of the A-th class module in the j-th functional module. L is the minimum configuration cost of the A-th class module, A U is the highest configuration cost of the A-th class module, B L represents the minimum configuration cost of the Bth class module, B U represents the maximum configuration cost of the Bth class module; M is the number of class modules of the jth functional module; The formula for the deterministic confidence function of the design solution is as follows: Among them, G(S) represents the confidence response function value, E η represents the expected function; N is the number of functional modules; represents the lower bound of the confidence of the lth class module of the i-th functional module, represents the median confidence value of the lth class module of the i-th functional module, Represents the upper bound of the confidence of the lth class module of the i-th functional module; The formula of the principle optimization model of the deterministic complex equipment forward design scheme is as follows: G′=(max(G(P)),max(G(S)),min(G(C))) Among them, G′ represents the principle solution of the forward design scheme of deterministic complex equipment, maxG(P) represents the optimal total module performance parameter of the design scheme, max(G(S)) represents the maximum total confidence of the design scheme, and min(G(C)) represents the minimum total configuration cost of the design scheme.

6. The method for optimizing complex equipment forward design considering multi-source uncertainty according to claim 1, characterized in that: In the step 4), in the improved multi-source uncertain non-dominated solution genetic algorithm, the population is initialized as a set of solution numbers of class modules of each functional module, and the dimension d of each chromosome is expanded to 3. The three dimensions of each chromosome represent the module performance parameter P, configuration cost C and confidence S of each functional module respectively. The coding structure of the multi-source uncertain non-dominated solution genetic algorithm is established as a coding structure between chromosomes, functional modules and uncertainty variables, so as to solve the principle optimization model of the forward design solution of deterministic complex equipment.

Images