A product service system module configuration optimization method with multi-stakeholder participation

Abstract

The present invention discloses a method for optimizing the configuration of product service system modules with the participation of multiple stakeholders, belonging to the field of product design technology. The method comprises constructing a product service system module configuration optimization model as an upper-level model; constructing a stakeholder cost optimization model as a lower-level model, wherein the stakeholders include suppliers, manufacturers, and distributors; constructing a two-level model using the upper-level and lower-level models, and solving the two-level model based on a chaotic adaptive nested genetic algorithm. The present invention establishes a one-master-multiple-slave two-level planning model for stakeholder participation in PSS design, targeting the different optimization objectives and constraints of PSS design and stakeholders, and proposes a method for optimizing the configuration of PSS modules with the participation of multiple stakeholders.

Description

Technical Field

[0001] The present invention relates to the field of product design technology, and in particular to configuration optimization of a product service system with multi-stakeholder participation, and specifically to a method for module configuration optimization of a product service system with multi-stakeholder participation. Background Art

[0002] The design of a Product Service System (PSS) must not only meet the diverse and personalized needs of stakeholders but also consider the industrial efficiency of large-scale production. This requires optimizing the PSS's module configuration. Adopting appropriate management and manufacturing methods for different module categories, combined with module constraints for configuration optimization, can effectively resolve the conflict between personalization and economies of scale, enabling mass customization. PSS module configuration involves constructing a series of configuration options by selecting different module candidates based on a set of constraints, tailored to the needs of different stakeholders and the importance of technical features. Configuration options are crucial for meeting the needs of individual stakeholders and serve as a fundamental indicator of market satisfaction.

[0003] A PSS integrates both product and service modules into a single system. Each module is composed of several composite modules, which are further assembled from basic modules. Basic modules are the fundamental building blocks of a PSS and include both mandatory and optional modules. Mandatory modules are essential to the product, while optional modules provide additional functionality or features that can be selected as needed. Each basic module has several candidate modules, each offering different design, performance, or cost options, bringing differentiation and flexibility to the product. During the PSS configuration process, the selection of candidate modules, the assembly method of the composite modules, and the choice of manufacturers and distributors are all influenced by stakeholders. Different suppliers may offer module candidates with varying quality, price, and delivery times, while different manufacturers may have varying manufacturing capabilities and technological advantages. These factors all impact the cost, quality, and delivery time of the PSS. Therefore, when configuring a PSS, it is important to fully consider the opinions and needs of stakeholders to ensure that the final product-service system meets market and customer expectations. Furthermore, due to production capacity constraints, not all possible module combinations will be considered for production. This requires trade-offs and optimization during the module configuration process to select the combination with the greatest economic benefits and market potential.

[0004] In summary, considering the close interactions between physical products and services, and between services, as well as the complexity of the PSS, it is necessary to establish a systematic module configuration optimization plan. This plan, combining the importance of technical features, as well as constraints such as module relevance and mutual exclusivity, optimizes the module configuration of services and physical products while ensuring compliance with the modular division principle. Ultimately, a PSS configuration plan with multi-stakeholder participation is obtained. This approach aims to simultaneously ensure the configuration constraints of the PSS itself and the effective participation of stakeholders, thereby improving production and service efficiency and reducing resource consumption. Summary of the Invention

[0005] To solve the above problems, the main purpose of the present invention is to provide a product service system module configuration optimization method with the participation of multiple stakeholders.

[0006] The technical solutions of the present invention are as follows:

[0007] A method for optimizing product service system module configuration with multi-stakeholder participation includes the following steps:

[0008] S1. Construct a product service system module configuration optimization model as the upper model;

[0009] The upper-level model makes decisions about composite modules and the composition of modules;

[0010] The product configuration plan is as follows:

[0011] From the module candidate set Select different module candidates to form different basic modules m j , all basic modules form a basic module set {m1,…,m j ,…,m J}; Select different basic modules from the basic module set to form different composite modules M k , all composite modules form a composite module set {M1,…,M k ,…,M K}; Select different composite modules from the composite module set to form different product solutions PSS i , all product solutions constitute the product solution set PSS;

[0012] Basic modules are divided into two categories: optional modules and mandatory modules. The set of all optional modules is the optional module set May, and the set of all mandatory modules is the mandatory module set Must.

[0013] Represent the product solution in vector form, X i ={x ikjl}, X i represents the i-th product configuration scheme (PSS i)’s decision vector; x ikjl represents the selection variable of the kth composite module, jth basic module, and lth module candidate in the i-th product configuration scheme. When the product configuration scheme does not contain this selection variable, x ikjl =0, when the product configuration includes the selected variable, x ikjl =1;

[0014] The upper-level objective function forms the basis for product configuration decisions. It determines the number and composition of configurable composite modules, selecting from known mandatory and optional modules and their corresponding candidate modules to determine the modular structure of the configuration solution. The configuration process requires weighing multiple factors to reach the optimal decision, and the upper-level objective function integrates these factors to form a clear guiding principle for decision-making. Maximizing customer and business benefits is the most fundamental principle in PSS configuration decisions, as measured by the customer's perceived utility of the PSS.

[0015]

[0016] Where U i Represents the i-th product configuration plan PSS i The utility of μ jl is the utility of the lth module candidate of the jth basic module that the customer considers; ω ij The technical characteristics of the jth basic module in the product PSS i The weight in; K represents the total number of composite modules; J represents the total number of basic modules; L j represents the total number of candidate modules for the jth basic module;

[0017] The interests of other stakeholders are also a multi-dimensional, comprehensive indicator. For suppliers, manufacturers, and distributors, profitability is the core manifestation of their interests and the most important and objective reflection, directly reflecting the company's economic benefits and operational efficiency. Therefore, stakeholder costs are used as a representation of their respective interests.

[0018] The cost function of stakeholders is:

[0019]

[0020] In the two-level optimization model, the minimum total cost of the lower-level suppliers, manufacturers, and distributors (respectively expressed as ) is the key decision variable. The profit-cost ratio derived from the upper-level architecture's analysis of the lower-level architecture is denoted as α, where α∈(0,1). The lower level of the model involves a configuration plan that optimizes the selection of suppliers, manufacturers, and distributors in the supply, manufacturing, and distribution stages, aiming to optimize the overall cost of the three.

[0021] The maximum value of the ratio of customer utility to stakeholder cost is taken as the objective function, that is, the utility per unit cost is maximized, as shown in the following formula:

[0022]

[0023] The constraints are:

[0024]

[0025] x ikjl ∈{0,1}

[0026] α∈(0,1)

[0027] i=1,…,I; k=1,…,K; j=1,…,J; l=1,…,L j

[0028] I,K∈N +

[0029] Where m j represents the jth basic module; Must represents the set of mandatory modules; May represents the set of optional modules; I represents the total number of configuration schemes; h represents the total number of mandatory modules; a represents the total number of optional modules; N + represents a positive integer;

[0030] S2, build a stakeholder cost optimization model as the lower-level model;

[0031] The lower-level functions include manufacturers, suppliers, and distributors, stakeholders in the supply chain. The model costs compete with each other in the optimization process, and the maximum value of the utility of the upper-level customers is finally achieved to output the product service system configuration plan.

[0032] S21. The supplier cost function should comprehensively consider the raw material cost, ordering cost, inventory cost, and transportation cost corresponding to each module candidate. The corresponding objective function is as follows:

[0033]

[0034] The corresponding constraints are as follows:

[0035]

[0036]

[0037] Where, represents the material cost of the lth module candidate of the jth basic module in the sth supplier; d jl PSS iThe demand for the lth module candidate of the jth basic module; represents the unit ordering cost of the lth module candidate of the jth basic module from the sth supplier; represents the unit inventory cost of the lth module candidate of the jth basic module in the sth supplier; represents the unit transportation cost of the lth module candidate of the jth basic module in the sth supplier; represents the allocation function of the selected suppliers on the supply amount; Q is the total number of suppliers who can provide the lth module candidate for the jth basic module; j PSS i The demand for the jth basic module;

[0038] For the upper model, PSS i Select the corresponding basic module m j ,Right now Then any supplier s in the supplier set S corresponding to the module needs to make a decision for this situation, deciding whether to supply this basic module candidate. The decision variable is recorded as When the sth supplier provides the jth basic module and the lth module candidate, then otherwise

[0039] During the actual PSS design and production process, there may be compatibility and incompatibility constraints between modules. For incompatibility constraints, when two different optional modules cannot appear in the same product at the same time, it constitutes module incompatibility. If two modules cannot be selected at the same time, the constraints are as follows:

[0040]

[0041] Where, The selection variable representing the lth module candidate of the j1th basic module of the kth composite module of the i-th product configuration scheme. When the product configuration scheme does not contain this selection variable, When the product configuration contains this selection variable, The selection variable representing the lth module candidate of the j2th basic module of the k'th composite module of the i-th product configuration scheme. When the product configuration scheme does not contain this selection variable, When the product configuration contains this selection variable,

[0042] As for compatibility constraints, especially when two optional modules must appear at the same time, this usually means that the two modules are interdependent, or they together constitute a basic functional unit of the product. If two modules need to be selected at the same time, the module compatibility constraints are as follows:

[0043]

[0044] S22. The manufacturer's cost function should comprehensively consider the fixed costs of each composite module, as well as variable costs such as mechanical equipment and labor input. The corresponding objective function is as follows:

[0045]

[0046] The corresponding constraints are as follows:

[0047]

[0048] Where, R represents the operating fixed cost of the rth manufacturer of the kth composite module; k represents the total number of manufacturers; represents the unit operating variable cost of the rth manufacturer to assemble and manufacture the lth module candidate of the jth basic module; d k represents the demand for the kth composite module; It represents the process variation cost of the rth manufacturer to assemble the mandatory module j into the module candidate l;

[0049] For the upper model, PSS i Select the corresponding basic module m j ,Right now Then any manufacturer r in the manufacturer set R corresponding to the module needs to make a decision for this situation, deciding whether to manufacture the required or optional module candidate. The decision variable is recorded as When the rth manufacturer provides the kth composite module, then otherwise

[0050] represents the proportion of the demand for the kth composite module produced by the rth manufacturer; Select the total number of manufacturers for the kth composite module; is the production capacity range of the kth composite module by the rth manufacturer;

[0051] S23. The distributor's cost function should comprehensively consider the product's corresponding ordering cost, inventory cost, and shipping cost. The corresponding objective function is as follows:

[0052]

[0053] The corresponding constraints are as follows:

[0054]

[0055] y b ∈{0,1}

[0056] In the formula, for the upper model, the manufacturer selects the corresponding mandatory module m i or optional module m j To form a composite module and further assemble it into a product, any distributor b in the distributor set B corresponding to the product needs to make a decision for this situation, deciding whether to distribute the product assembled from this composite module. The decision variable is denoted as y b , when the bth distributor decides to distribute, then y b =1, otherwise y b =0;

[0057] N B Select the number of distributors for each product;

[0058] RC b represents the unit ordering cost of the b-th distributor;

[0059] HC b represents the unit inventory cost of the product in the b-th distributor;

[0060] TCP b represents the unit transportation cost of the product in the b-th distributor;

[0061] f(y b ) represents the proportion function of the b-th distributor in the distribution of products;

[0062] d represents the total demand for the product;

[0063] W b It represents the distribution capability range of the b-th distributor for the product;

[0064] The upper model constructed by S3 and S1 and the lower model constructed by S2 form a double-layer model, which is solved based on the chaotic adaptive nested genetic algorithm;

[0065] The specific steps of the chaotic adaptive nested genetic algorithm process are as follows:

[0066] S31. Parameter initialization: Determine the size of the initial population M, the genetic generation G, the chaotic state control parameter μ, the elite retention number m, and the upper and lower bounds p of the adaptive crossover and mutation operators. c0 、p c1 、p m0 and pm1 ;

[0067] S32, chaos initialization population: perform chaos optimization operations on the initial populations of the upper and lower layer models respectively;

[0068] The calculation formula for determining the chaotic variables in the upper model is as follows:

[0069] x k+1 =μx k (1-x k )

[0070] k=0,1,…0≤x k ≤1

[0071] Where x k is the result of k iterations of the chaotic variable x, μ is the chaotic state control parameter, μ∈[0,4], when μ=4, and When , the system is in a completely chaotic state, the chaotic variable traverses and searches within the range of (0,1), and the generated initial sequence value will have completely chaotic characteristics.

[0072] The calculation formula for determining the chaotic variables in the lower model is as follows:

[0073] t ik =Int(nμx k-1 (1-x k-1 ))i=0,1,…,m-1

[0074] t ik The value of the i-th optimization variable after the k-th iteration in the mathematical model is the number of candidate items. If there are n+1 candidate items, the value range is: ; n is the number of module candidates corresponding to each basic module;

[0075] S33. Assume that the population size in the algorithm is M, sort all individuals in the population according to their fitness values ​​from large to small, select the first m individuals with the best fitness values ​​and directly pass them on to the next generation; for the remaining Mm individuals, use a mixed ranking strategy to select individuals from these individuals to participate in crossover and mutation.

[0076] The calculation formula of the fitness value of the i-th individual is as follows:

[0077]

[0078] S34, combining the selection mechanism of elite retention and hybrid ranking to screen out the parent individuals for subsequent crossover and mutation operations;

[0079] S35, adaptive crossover and mutation operation: calculate the corresponding crossover and mutation probabilities based on the individual fitness value, and complete the crossover and mutation operations;

[0080] The sire is determined by roulette wheel selection, and the probability of an individual being selected as the sire is determined according to its fitness value. The probability of the i-th individual being selected can be calculated as:

[0081]

[0082] Adaptive crossover probability p c , adaptive mutation probability p m The calculation formulas are:

[0083]

[0084] Where p c0 、p c1 are the initial adjustment factors for the lower and upper bounds of the crossover probability, which together determine the fluctuation range of the crossover probability; p m0 、p m1 are the initial adjustment factors for the lower and upper bounds of the mutation probability, respectively, which limit the upper and lower limits of the mutation probability; define f as the individual fitness value; f max is the optimal fitness value of individuals in the current population; f min is the worst fitness value of individuals in the population; f avg is the average fitness value of individuals in the population;

[0085] S36. If the number of iterations reaches the upper limit of the genetic algebra G preset in step S31, the algorithm ends and outputs the optimal module configuration solution; otherwise, go to step S33.

[0086] Beneficial effects:

[0087] Aiming at the master-slave association optimization problem of PSS module configuration, the present invention establishes a one-master-multiple-slave two-layer configuration optimization model in which stakeholders participate in PSS design.

[0088] In order to effectively solve this model, a chaotic adaptive nested genetic algorithm was developed. By initializing the population through chaos theory, the advantages of chaos algorithm and genetic algorithm were integrated to form a complementary effect, and the algorithm's optimization ability was improved. As a result, the optimal solution for PSS module configuration with the participation of multiple stakeholders was obtained, helping enterprises design products with core competitiveness that meet the needs of stakeholders.

[0089] This method has the following main advantages:

[0090] (1) In view of the different optimization objectives and constraints of PSS design and stakeholders, this paper establishes a one-master-multiple-slave two-level planning model for stakeholders to participate in PSS design, and proposes a PSS module configuration optimization method with the participation of multiple stakeholders.

[0091] (2) The present invention proposes a chaotic adaptive nested genetic algorithm to effectively solve the one-master-multiple-slave two-level programming model in which stakeholders participate in the PSS design. The algorithm integrates the advantages of the chaotic algorithm and the genetic algorithm to form a complementary relationship, improves the algorithm's optimization ability, can avoid individuals from concentrating in a certain local area, and is more likely to contain individuals close to the global optimal solution, which helps to narrow the scope of subsequent searches; can dynamically adjust its parameters according to the characteristics of the problem and the search process, and conduct a fine search near the optimal solution, further improving the algorithm's convergence speed and global optimization ability; can maintain the diversity of the group without increasing the population size, which helps to reduce randomness and avoid premature convergence, and improve the probability of the algorithm finding the global optimal solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0092] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments.

[0093] Figure 1 This is a comparison chart of fitness values ​​of three composite module partitioning methods in an embodiment of the present invention;

[0094] Figure 2 It is the fitness value, customer utility and stakeholder cost change curve of the optimal composite module partitioning method in the case of the present invention;

[0095] Figure 3 This is a schematic diagram of a one-master-multiple-slave two-tier planning model for stakeholders in the present invention to participate in PSS design;

[0096] Figure 4 It is a schematic diagram of the chaotic adaptive nested genetic algorithm of the present invention. DETAILED DESCRIPTION

[0097] The present invention will be further described in detail below with reference to the embodiments and accompanying drawings.

[0098] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0099] Example

[0100] A method for optimizing product service system module configuration with multi-stakeholder participation includes the following steps:

[0101] S1. Construct a product service system module configuration optimization model as the upper model;

[0102] The product configuration scheme in the upper-level model is as follows: different module candidates are selected from the module candidate set to form different basic modules, and all basic modules form a basic module set; different basic modules are selected from the basic module set to form different composite modules, and all composite modules form a composite module set; different composite modules are selected from the composite module set to form different product solutions, and all product solutions form a product solution set; basic modules are divided into two categories, namely optional modules and mandatory modules, the set consisting of all optional modules is the optional module set, and the set consisting of all mandatory modules is the mandatory module set;

[0103] The objective function of the upper model is:

[0104]

[0105] in,

[0106] In the formula, F(x ikjl ) represents the cost-utility function; U i Represents the i-th product configuration plan PSS i The utility of TC i represents the cost of stakeholders in the i-th product configuration scheme; K represents the total number of composite modules; J represents the total number of basic modules; L j represents the total number of module candidates for the jth basic module; ω ij The technical characteristics of the jth basic module in the product PSS i The weight in μ jl x is the utility of the lth module candidate of the jth basic module that the customer considers; ikjl represents the selection variable of the lth module candidate of the jth basic module of the kth composite module of the ith product configuration scheme. When the product configuration scheme does not contain this selection variable, x ikjl =0, when the product configuration includes the selected variable, x ikjl =1; α is the profit cost ratio obtained by analyzing the benefits of the upper structure to the lower structure; denote the minimum total cost of lower-tier suppliers, manufacturers, and distributors respectively;

[0107] The constraints of the upper model include:

[0108]

[0109] x ikjl ∈{0,1}

[0110] α∈(0,1)

[0111] i=1,…,I; k=1,…,K; j=1,…,J; l=1,…,L j

[0112] I,K∈N +

[0113] Where m j represents the jth basic module; Must represents the set of mandatory modules; May represents the set of optional modules; h represents the total number of mandatory modules; a represents the total number of optional modules; I represents the total number of configuration options; N + represents a positive integer;

[0114] S2. Constructing a stakeholder cost optimization model as a lower-level model; the stakeholders include suppliers, manufacturers, and distributors;

[0115] The supplier's objective function in the lower model is:

[0116]

[0117] The constraints are:

[0118]

[0119] Where S jl is the total number of suppliers who provide candidate items, is a decision variable. When the sth supplier provides the jth basic module and the lth module candidate, otherwise represents the material cost of the lth module candidate of the jth basic module in the sth supplier; d jl PSS i The demand for the lth module candidate of the jth basic module; represents the unit ordering cost of the lth module candidate of the jth basic module from the sth supplier; represents the unit inventory cost of the lth module candidate of the jth basic module in the sth supplier; represents the unit transportation cost of the lth module candidate of the jth basic module in the sth supplier; represents the allocation function of the selected suppliers on the supply amount; Q is the total number of suppliers who can provide the lth module candidate of the jth basic module; j PSS i The demand for the jth basic module;

[0120] In addition, compatibility and incompatibility constraints are included;

[0121] The constraints for two modules to be incompatible (cannot be selected at the same time) are:

[0122]

[0123] Where, The selection variable representing the lth module candidate of the j1th basic module of the kth composite module of the i-th product configuration scheme. When the product configuration scheme does not contain this selection variable, When the product configuration contains this selection variable, The selection variable representing the lth module candidate of the j2th basic module of the k'th composite module of the i-th product configuration scheme. When the product configuration scheme does not contain this selection variable, When the product configuration contains this selection variable,

[0124] The constraints for two modules to be compatible (selected at the same time) are:

[0125]

[0126] Where, The selection variable representing the lth module candidate of the j3th basic module of the kth composite module of the i-th product configuration scheme. When the product configuration scheme does not contain this selection variable, When the product configuration contains this selection variable, The selection variable representing the lth module candidate of the j4th basic module of the k'th composite module of the i-th product configuration scheme. When the product configuration scheme does not contain this selection variable, When the product configuration contains this selection variable,

[0127] The manufacturer's objective function in the lower model is:

[0128]

[0129] The constraints are:

[0130]

[0131] Where R k represents the total number of manufacturers; represents the operating fixed cost of the rth manufacturer of the kth composite module; represents the unit operating variable cost of the rth manufacturer to assemble and manufacture the lth module candidate of the jth basic module; d k represents the demand for the kth composite module; It represents the process variation cost of the rth manufacturer to assemble the mandatory module j into the module candidate l; is a decision variable. When the rth manufacturer provides the kth composite module, otherwise represents the proportion of the demand for the kth composite module produced by the rth manufacturer; Select the total number of manufacturers for the kth composite module; is the production capacity range of the kth composite module by the rth manufacturer;

[0132] The objective function of the distributor in the lower model is:

[0133]

[0134] The constraints are:

[0135]

[0136] y b ∈{0,1}

[0137] Where B is the total number of distributors; N B Select the number of distributors for each product;

[0138] RC b represents the unit ordering cost of the b-th distributor;

[0139] HC b represents the unit inventory cost of the product in the b-th distributor;

[0140] TCD b represents the unit transportation cost of the product in the b-th distributor;

[0141] f(y b ) represents the proportion function of the b-th distributor in the distribution of products;

[0142] d represents the total demand for the product;

[0143] W b It represents the distribution capability range of the b-th distributor for the product;

[0144] The upper model constructed by S3 and S1 and the lower model constructed by S2 form a double-layer model, which is solved based on the chaotic adaptive nested genetic algorithm;

[0145] The specific steps of the chaotic adaptive nested genetic algorithm process are as follows:

[0146] S31. Parameter initialization: Determine the size of the initial population M, the genetic generation G, the chaotic state control parameter μ, the elite retention number m, and the upper and lower bounds p of the adaptive crossover and mutation operators. c0 、p c1 、pm0 and p m1 ;

[0147] S32, chaos initialization population: perform chaos optimization operations on the initial populations of the upper and lower layer models respectively;

[0148] The calculation formula for determining the chaotic variables in the upper model is as follows:

[0149] x k+1 =μx k (1-x k )

[0150] k=0,1,…0≤x k ≤1

[0151] Where x k is the result of k iterations of the chaotic variable x, μ is the chaotic state control parameter, μ∈[0,4], when μ=4, and When , the system is in a completely chaotic state, the chaotic variable traverses and searches within the range of (0,1), and the generated initial sequence value will have completely chaotic characteristics.

[0152] The calculation formula for determining the chaotic variables in the lower model is as follows:

[0153] t ik =Int(nμx k-1 (1-x k-1 ))i=0,1,…,m-1

[0154] t ik The value of the i-th optimization variable after the k-th iteration in the mathematical model is the number of module candidates. If there are n+1 module candidates, the value range is: 0, 1, ..., n; n is the number of module candidates corresponding to each basic module;

[0155] S33. The population size in the algorithm is M. All individuals in the population are sorted from large to small according to their fitness values, and the first m individuals with the best fitness values ​​are selected to be directly inherited to the next generation; for the remaining Mm individuals, a mixed ranking strategy is used to select individuals from these individuals to participate in crossover and mutation.

[0156] The calculation formula of the fitness value of the i-th individual is as follows:

[0157]

[0158] S34, combining the selection mechanism of elite retention and hybrid ranking to screen out the parent individuals for subsequent crossover and mutation operations;

[0159] S35, adaptive crossover and mutation operation: calculate the corresponding crossover and mutation probabilities based on the individual fitness value, and complete the crossover and mutation operations;

[0160] The sire is determined by roulette wheel selection, and the probability of an individual being selected as the sire is determined according to its fitness value. The probability of the i-th individual being selected can be calculated as:

[0161]

[0162] Adaptive crossover probability p c , adaptive mutation probability p m The calculation formulas are:

[0163]

[0164] Where p c0 、p c1 are the initial adjustment factors for the lower and upper bounds of the crossover probability, which together determine the fluctuation range of the crossover probability; p m0 、p m1 are the initial adjustment factors for the lower and upper bounds of the mutation probability, respectively, which limit the upper and lower limits of the mutation probability; define f as the individual fitness value; f max is the optimal fitness value of individuals in the current population; f min is the worst fitness value of individuals in the population; f avg is the average fitness value of individuals in the population;

[0165] S36. If the number of iterations reaches the upper limit of the genetic algebra G preset in step S31, the algorithm ends and outputs the optimal module configuration solution; otherwise, go to step S33.

[0166] Beneficial effects:

[0167] Aiming at the master-slave association optimization problem of PSS module configuration, the present invention establishes a one-master-multiple-slave two-layer configuration optimization model in which stakeholders participate in PSS design.

[0168] In order to effectively solve this model, a chaotic adaptive nested genetic algorithm was developed. By initializing the population through chaos theory, the advantages of chaos algorithm and genetic algorithm were integrated to form a complementary effect, and the algorithm's optimization ability was improved. As a result, the optimal solution for PSS module configuration with the participation of multiple stakeholders was obtained, helping enterprises design products with core competitiveness that meet the needs of stakeholders.

[0169] To illustrate the practical application of the proposed method, the present invention takes the configuration optimization of a smart speaker product service system as an example, establishes a one-master-multiple-slave two-layer configuration optimization model in which stakeholders participate in the design of the product service system, solves the model using the designed improved CANGA algorithm, and outputs the optimal smart speaker product service system module configuration solution.

[0170] The detailed steps are as follows:

[0171] Table 1 Classification of product service system modules

[0172]

[0173]

[0174] Three reasonable composite module partitioning schemes are selected to facilitate subsequent analysis, as shown in Tables 2, 3, and 4 respectively:

[0175] Table 2 Division of the first composite module

[0176] Composite Module Basic Module Security Module <![CDATA[m4、m5、m7、m 10 ]]> Personalization module <![CDATA[m6、m8、m9]]> Power Module <![CDATA[m1、m2、m3]]>

[0177] Table 3 Division of the second composite module

[0178] Composite Module Basic Module Structural Module <![CDATA[m5、m9]]> Communication sensor module <![CDATA[m4、m6、m7、m8、m 10 ]]> Power Module <![CDATA[m1、m2、m3]]>

[0179] Table 4 Division of the third composite module

[0180] Composite Module Basic Module Physical Product Module <![CDATA[m4、m5、m6、m7]]> Service Module <![CDATA[m8、m9、m 10 ]]> Power Module <![CDATA[m1、m2、m3]]>

[0181] By using SPSS software to perform joint analysis and calculation, the utility values ​​of customers for module candidates were obtained, as shown in Table 5.

[0182] Table 5 Utility values ​​of basic module candidates based on conjoint analysis

[0183]

[0184]

[0185] Table 6 Module Candidate-Supplier Information

[0186]

[0187]

[0188] Table 7 Composite Module - Manufacturer Information

[0189]

[0190]

[0191] Table 8 Product-Distributor Information

[0192]

[0193] Based on the chaotic adaptive nested genetic algorithm, the fitness values ​​of the three composite module division methods are calculated respectively, and the output images of the calculation results of the three division methods are compared. Figure 1 The optimal composite module partitioning method is selected as the second solution with the largest fitness value. After the composite module partitioning method is selected, the corresponding basic module and its module candidate configuration scheme are calculated.

[0194] Output the upper utility function, lower cost function and fitness value change trend of the second optimal composite module partitioning scheme as an image, such as Figure 2 As shown in Table 9, it can be seen that the upper-layer customer utility function shows an upward trend, while the lower-layer cost function shows a downward trend. Under the interaction between the two, the fitness value shows an upward trend until the optimal value of the PSS module configuration scheme is found, as shown in Table 9.

[0195] Table 9 Optimal configuration scheme

[0196]

[0197] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the embodiments of the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A method for optimizing product service system module configuration with multi-stakeholder participation, characterized in that: The following steps are involved: S1. Construct a product service system module configuration optimization model as the upper model; The product configuration scheme in the upper-level model is as follows: different module candidates are selected from the module candidate set to form different basic modules, and all basic modules form a basic module set; different basic modules are selected from the basic module set to form different composite modules, and all composite modules form a composite module set; different composite modules are selected from the composite module set to form different product solutions, and all product solutions form a product solution set; basic modules are divided into two categories, namely optional modules and mandatory modules, the set consisting of all optional modules is the optional module set, and the set consisting of all mandatory modules is the mandatory module set; The objective function of the upper model is: in, In the formula, F(x ikjl ) represents the cost-utility function; U i Represents the i-th product configuration plan PSS i The utility of TC i represents the cost of stakeholders in the i-th product configuration scheme; K represents the total number of composite modules; J represents the total number of basic modules; L j represents the total number of module candidates for the jth basic module; The technical characteristics corresponding to the jth basic module in the product PSS i The weight in μ jl x is the utility of the lth module candidate of the jth basic module that the customer considers; ikjl represents the selection variable of the lth module candidate of the jth basic module of the kth composite module of the ith product configuration scheme. When the product configuration scheme does not contain this selection variable, x ikjl =0, when the product configuration includes the selected variable, x ikjl =1; α is the profit cost ratio of the upper structure to the lower structure’s benefit analysis; TC i S TC i R TC i B denote the minimum total cost of lower-tier suppliers, manufacturers, and distributors respectively; S2. Constructing a stakeholder cost optimization model as a lower-level model, wherein the stakeholders include suppliers, manufacturers, and distributors; S3, a double-layer model is constructed with the upper model constructed by S1 and the lower model constructed by S2, and the double-layer model is solved based on the chaotic adaptive nested genetic algorithm.

2. The method for optimizing product service system module configuration with multi-stakeholder participation according to claim 1, characterized in that: The supplier's objective function in the lower model is: The constraints are: Where S jl The total number of suppliers who provide candidate items; is a decision variable. When the sth supplier provides the jth basic module and the lth module candidate, otherwise represents the material cost of the lth module candidate of the jth basic module in the sth supplier; d jl PSS i The demand for the lth module candidate of the jth basic module; represents the unit ordering cost of the lth module candidate of the jth basic module from the sth supplier; represents the unit inventory cost of the lth module candidate of the jth basic module in the sth supplier; represents the unit transportation cost of the lth module candidate of the jth basic module in the sth supplier; represents the allocation function of the selected suppliers on the supply amount; Q is the total number of suppliers who can provide the lth module candidate of the jth basic module; j PSS i The demand for the j-th basic module.

3. The method for optimizing product service system module configuration with multi-stakeholder participation according to claim 2, characterized in that: The supplier constraints in the lower model also include: The constraint that two modules cannot be selected at the same time is calculated as follows: Where, The selection variable representing the lth module candidate of the j1th basic module of the kth composite module of the i-th product configuration scheme. When the product configuration scheme does not contain this selection variable, When the product configuration contains this selection variable, The selection variable representing the lth module candidate of the j2th basic module of the k'th composite module of the i-th product configuration scheme. When the product configuration scheme does not contain this selection variable, When the product configuration contains this selection variable, The constraint conditions for selecting two modules at the same time are calculated as follows: Where, The selection variable representing the lth module candidate of the j3th basic module of the kth composite module of the i-th product configuration scheme. When the product configuration scheme does not contain this selection variable, When the product configuration contains this selection variable, The selection variable representing the lth module candidate of the j4th basic module of the k'th composite module of the i-th product configuration scheme. When the product configuration scheme does not contain this selection variable, When the product configuration contains this selection variable, May represents a collection of optional modules.

4. The method for optimizing product service system module configuration with multi-stakeholder participation according to claim 1, characterized in that: The manufacturer's objective function in the lower model is: The constraints are: Where R k represents the total number of manufacturers; represents the operating fixed cost of the rth manufacturer of the kth composite module; represents the unit operating variable cost of the rth manufacturer to assemble and manufacture the lth module candidate of the jth basic module; d k represents the demand for the kth composite module; It represents the process variation cost of the rth manufacturer to assemble the mandatory module j into the module candidate l; is a decision variable. When the rth manufacturer provides the kth composite module, otherwise represents the proportion of the demand for the kth composite module produced by the rth manufacturer; Select the total number of manufacturers for the kth composite module; is the production capacity range of the rth manufacturer for the kth composite module; Must represents the set of mandatory modules.

5. The method for optimizing product service system module configuration with multi-stakeholder participation according to claim 1, characterized in that: The objective function of the distributor in the lower model is: The objective function of the distributor in the lower model is: The constraints are: y b ∈{0,1} Where B is the total number of distributors for each product; N B Select the number of distributors for each product; RC b represents the unit ordering cost of the b-th distributor; HC b represents the unit inventory cost of the product in the b-th distributor; TCD b represents the unit transportation cost of the product in the b-th distributor; f(y b ) represents the proportion function of the b-th distributor in distributing the product; d represents the total demand for the product; W b represents the distribution capability range of the b-th distributor for the product; y b is a decision variable. When the bth distributor decides to distribute, y b =1, otherwise y b =0.

6. The method for optimizing product service system module configuration with multi-stakeholder participation according to claim 1, characterized in that: Step S3 includes: S31. Parameter initialization: Determine the size of the initial population M, the genetic generation number G, the chaotic state control parameter μ, the elite retention number m, and the upper and lower bounds of the adaptive crossover and mutation operators; S32, chaos initialization population: perform chaos optimization operations on the initial populations of the upper and lower layer models respectively; S33. Sort all individuals in the population according to their fitness values ​​from large to small, select the first m individuals with the best fitness values ​​and directly pass them on to the next generation; for the remaining Mm individuals, use a mixed ranking strategy to select individuals from these individuals to participate in crossover and mutation; The calculation formula of the fitness value of the i-th individual is as follows: S34, combining the selection mechanism of elite retention and hybrid ranking to screen out the parent individuals for subsequent crossover and mutation operations; S35, adaptive crossover and mutation operation: calculate the corresponding crossover and mutation probabilities based on the individual fitness value, and complete the crossover and mutation operations; The sire is determined by roulette wheel selection, and the probability of an individual being selected as the sire is determined according to its fitness value. The probability of the i-th individual being selected can be calculated as: Adaptive crossover probability p c , adaptive mutation probability p m The calculation formulas are: Where p c0 、p c1 are the initial adjustment factors for the lower and upper bounds of the crossover probability, which together determine the fluctuation range of the crossover probability; p m0 、p m1 are the initial adjustment factors for the lower and upper bounds of the mutation probability, respectively, which limit the upper and lower limits of the mutation probability; define f as the individual fitness value; f max is the optimal fitness value of individuals in the current population; f min is the worst fitness value of individuals in the population; f avg is the average fitness value of individuals in the population; S36. If the number of iterations reaches the upper limit of the genetic algebra G preset in step S31, the algorithm ends and outputs the optimal module configuration solution; otherwise, go to step S33.

7. The method for optimizing product service system module configuration with multi-stakeholder participation according to claim 5, characterized in that: The calculation formula for determining the chaotic variables in the upper model in step S32 is as follows: x k+1 =μx k (1-x k ) Where x k is the result of chaotic variable x after k iterations; x k+1 is the result of the chaotic variable x after iterating k+1 times; The calculation formula for determining the chaotic variables in the lower model is as follows: t ik =Int(nμx k-1 (1-x k-1 )) t ik is the value of the i-th optimization variable after the k-th iteration in the mathematical model; x k-1 is the result of the chaotic variable x after iterating k-1 times; n is the number of module candidates corresponding to each basic module.

Images