A product characteristic extraction method based on improved fruit fly algorithm under value chain synergy

Abstract

This invention discloses a product characteristic extraction method based on an improved fruit fly algorithm under value chain collaboration. First, a multi-objective optimization model for the quality characteristics of complex nuclear power equipment products is constructed. Next, the ranges of the first and second combined quality characteristics are determined. Then, the improved fruit fly algorithm is used to solve the multi-objective optimization model, obtaining a set of combined quality characteristics for the complex nuclear power equipment products. Finally, the combined quality characteristics in the set are transformed to obtain the characteristics of the complex nuclear power equipment products, and the most frequently occurring characteristics are selected as the key quality characteristics of the complex nuclear power equipment products. This invention establishes a multi-objective optimization model for extracting quality characteristics of complex nuclear power products, balancing the optimization of reliability, benefit value, and resource consumption value. An improved piecewise variable step-size fruit fly optimization algorithm is designed, achieving a highly efficient solution for the fruit fly optimization algorithm and improving the algorithm's running efficiency and global optimum.

Description

Technical Field

[0001] This invention relates to a method for extracting key quality characteristics of nuclear power equipment products, specifically a method for extracting key quality characteristics of nuclear power equipment products based on an improved fruit fly algorithm under value chain collaboration. Background Technology

[0002] In the product design industry, the weight of quality characteristics in the mapping process between customer needs and quality characteristics reflects the maximum extent to which quality characteristics can satisfy customer satisfaction. In the extraction of quality characteristics for nuclear power equipment products, the unity of corporate profit, social responsibility, and environmental impact should be emphasized to respond to the triple bottom line principle. However, it should also be recognized that the implementation of quality characteristics is constrained by many other factors in engineering practice, such as economic costs, technical difficulty, processing cycle, environmental restrictions, and ethical policies, laws, and regulations. As academic research on the application of QFD in the extraction of HOQ (Household Qualification) for nuclear power equipment product quality characteristics deepens, effectively addressing the above-mentioned issues is of great significance.

[0003] Every enterprise must prioritize customer needs to gain a competitive advantage; this is the focus of the modern market. With exciting technological advancements and fierce market competition, the ambiguity and invisibility of customer demand information have gradually become a primary challenge for manufacturing enterprises, including nuclear power equipment manufacturers. It is necessary to acquire, classify, and mine various types of customer demand information. Customer requirements can be translated into quality characteristics of nuclear power equipment products, guiding product design and thus developing high-quality, low-cost, and satisfactory products.

[0004] In traditional nuclear power equipment product design, it is assumed that the granularity of fuzzy language used by all customers is consistent. Then, the fuzzy evaluation language of customers is mapped to precise or fuzzy numbers, and the overall customer weight is calculated. Essentially, this means that a single customer has the same weight for all customer needs. However, customer credibility may differ under different customer requirements, making it unreasonable to directly assume that a customer has equal weight across all customer requirements. Furthermore, the granularity of fuzzy language evaluation of the target product often varies across different customer groups. Although customers' multi-granular fuzzy language evaluations of similar products may differ in granularity, they exhibit similarity at the logical order level. This similarity can be conveyed through relevant explicit granular structures and used as a basis for judging customer credibility. Therefore, introducing a dominant granular structure to calculate the fine-grained weight for each customer is beneficial for the accurate aggregation of customer demand information.

[0005] Advanced product optimization design methods can greatly enhance the market competitiveness of nuclear power equipment manufacturing enterprises and promote sustainable economic, social, and environmental development. Quality characteristics are the carriers of product quality and are closely related to sustainable product design. Value chain collaborative product characteristic extraction methods are a typical nonlinear multi-objective integer programming NP problem. If some traditional intelligent algorithms, such as genetic algorithms, machine learning algorithms, and neural network algorithms, are used to solve this model, it is inevitable to correctly determine the values ​​of many parameters. These have a direct impact on the accuracy of the solution, but it is difficult to guarantee their appropriateness in engineering practice. Fruit flies have better sensory perception than other species, especially in terms of smell and vision. Therefore, the fruit fly optimization algorithm is a new swarm intelligence algorithm based on fruit fly foraging behavior proposed in recent years. It inherits the characteristics of swarm intelligence algorithms such as group cooperation and information sharing, which can overcome the above difficulties. It is not dependent on the specific domain of the problem and has strong robustness to different problem types. However, the traditional fruit fly algorithm has a fixed search radius, a limited search range, and weak global search optimization ability, which leads to a decrease in population diversity and causes problems such as premature convergence, low accuracy, and instability in high-dimensional optimization problem environments. Summary of the Invention

[0006] To overcome the shortcomings of existing technologies, a method for extracting key quality characteristics of nuclear power equipment products under value chain collaboration is proposed. This invention fully retains the simplicity of rough set computation and the flexibility of interval numbers in expressing the correlation between customer requirements and the quality characteristics of nuclear power equipment products. A multi-objective optimization model for extracting quality characteristics of complex nuclear power products and nuclear power equipment products is established, achieving a balanced optimization of relevance reliability, resource consumption, and revenue. An improved piecewise variable-step-size fruit fly optimization algorithm is designed to achieve efficient solution of the model.

[0007] The technical solution adopted by this invention to solve its technical problem is:

[0008] 1) Based on the reliability, revenue, and resource consumption of complex nuclear power equipment products, construct a multi-objective optimization model for the quality characteristics of complex nuclear power equipment products;

[0009] 2) Determine the range of the first and second combination quality characteristics based on the total number of characteristics of complex nuclear power equipment products;

[0010] 3) Based on the range of the first and second combined quality characteristics, the improved fruit fly algorithm is used to solve the multi-objective optimization model and obtain the combined quality characteristic set of complex nuclear power equipment products.

[0011] 4) After converting each combined quality feature in the combined quality feature set, the characteristics of complex nuclear power equipment products are obtained, and the most frequently occurring characteristics of complex nuclear power equipment products are taken as the key quality characteristics of complex nuclear power equipment products.

[0012] In section 1), the formula for the multi-objective optimization model of the quality characteristics of complex nuclear power equipment products is as follows:

[0013]

[0014]

[0015] Among them, YX, ZY, and SY represent the reliability, resource consumption, and revenue of complex nuclear power equipment products, respectively. It is the j-th complex nuclear power equipment product characteristic x j Credibility weight, and These are the characteristics x of the j-th complex nuclear power equipment product. j The resource consumption value weight and revenue value weight; s represents the total number of characteristics of complex nuclear power equipment products, GY represents technical difficulty, SJ represents processing cycle, and CB represents cost; and These are the characteristics x of the j-th complex nuclear power equipment product. j The weights for technical difficulty, processing cycle, and cost;

[0016] The six constraints of the multi-objective optimization model are as follows:

[0017]

[0018] Among them, YX min It is the minimum value of credibility; ZY max It is the maximum value of resource consumption; SY min It is the minimum value of the profit; GY max This represents the maximum level of technical difficulty; SJ max This represents the maximum value of the processing cycle; CB max That is the maximum cost.

[0019] In 2), the range LR of the first combined quality characteristic x Satisfying LR x ∈[0,2 le -1], the range of the second combined quality characteristics LR Y Satisfying LR Y ∈[0,2 ri -1], where le represents the number of complex nuclear power equipment product characteristics in the first combination of quality characteristics, ri represents the number of complex nuclear power equipment product characteristics in the second combination of quality characteristics, s represents the total number of complex nuclear power equipment product characteristics, and round[] represents the rounding operator.

[0020] In the improved fruit fly algorithm described in 3), the position of the fruit fly in the x-direction is randomly generated within the range of the first combination of quality features, and the position of the fruit fly in the y-direction is randomly generated within the range of the second combination of quality features.

[0021] In the improved fruit fly algorithm described in 3), the random and amplified step sizes of the fruit fly in the x and y directions are both set to... The random and restore step sizes for fruit flies are set to... rand() represents the random number generation function, Maxgen is the maximum number of iterations, FR is the single flight range of the fruit fly, and t represents the number of iterations.

[0022] In step 4), each combined quality feature in the combined quality feature set is in decimal format. Each combined quality feature in decimal format is converted into binary format. Each bit of each combined quality feature in binary format represents a complex nuclear power equipment product characteristic. If the current bit is 0, it means that there is a corresponding complex nuclear power equipment product characteristic. If the current bit is 1, it means that there is no corresponding complex nuclear power equipment product characteristic.

[0023] The beneficial effects of this invention are as follows:

[0024] This invention fully preserves the simplicity of rough set calculation and the flexibility of interval numbers in expressing the correlation between customer requirements and the quality characteristics of nuclear power equipment products. It establishes a multi-objective optimization model for extracting the quality characteristics of complex nuclear power products and nuclear power equipment products, so as to optimize the relevant credibility, resource consumption and benefit value in a balanced way.

[0025] This invention proposes an improved piecewise variable step size fruit fly optimization algorithm. By flexibly changing the step size according to the iteration process, an efficient solution to the fruit fly optimization algorithm is achieved, improving the algorithm's running efficiency and global optimum. Attached Figure Description

[0026] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0027] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0028] like Figure 1 As shown, the present invention includes the following steps:

[0029] 1) Based on the reliability, revenue, and resource consumption of complex nuclear power equipment products, construct a multi-objective optimization model for the quality characteristics of complex nuclear power equipment products;

[0030] In section 1), the formula for the multi-objective optimization model of the quality characteristics of complex nuclear power equipment products is as follows:

[0031]

[0032] Among them, YX, ZY, and SY represent the reliability, resource consumption, and revenue of complex nuclear power equipment products, respectively. It is the j-th complex nuclear power equipment product characteristic x j Credibility weight, and These are the characteristics x of the j-th complex nuclear power equipment product. j The resource consumption value weight and revenue value weight; s represents the total number of characteristics of complex nuclear power equipment products, GY represents technical difficulty, SJ represents processing cycle, and CB represents cost; and These are the characteristics x of the j-th complex nuclear power equipment product. j The weights for technical difficulty, processing cycle, and cost;

[0033] Due to numerous limiting factors in actual production processes, it is impossible to achieve all ideal conditions. The six constraints of the multi-objective optimization model are as follows:

[0034]

[0035] Among them, YX min It is the minimum value of credibility; ZY max It is the maximum value of resource consumption; SY min It is the minimum value of the profit; GY max This represents the maximum level of technical difficulty; SJ max This represents the maximum value of the processing cycle; CB max That is the maximum cost.

[0036] 2) Determine the range of the first and second combination quality characteristics based on the total number of characteristics of complex nuclear power equipment products;

[0037] In 2), the range of the first combination of quality characteristics is LR. x Satisfying LR x ∈[0,2 le -1], the range of the second combined quality characteristics LR Y Satisfying LR Y ∈[0,2 ri -1], where le represents the number of complex nuclear power equipment product characteristics in the first combination of quality characteristics, ri represents the number of complex nuclear power equipment product characteristics in the second combination of quality characteristics, s represents the total number of complex nuclear power equipment product characteristics, and round[] represents the rounding operator.

[0038] 3) Based on the range of the first and second combined quality characteristics, the improved fruit fly algorithm is used to solve the multi-objective optimization model and obtain the combined quality characteristic set of complex nuclear power equipment products.

[0039] In the improved fruit fly algorithm (3), the position X of the fruit fly in the x-direction is randomly generated within the range of the first combination of quality features. _axis Within the range of the second combination of quality characteristics, randomly generate fruit flies at positions Y in the y-direction. _axis This forms the initial random coordinate position (X) of the fruit fly. _axis Y _axis The specific formula is as follows:

[0040] X _axis =round[rand(LR) X )]

[0041] Y _axis =round[rand(LR) Y )]

[0042] Here, rand() represents the random number generation function.

[0043] In the improved fruit fly algorithm, the random and amplified step sizes of the fruit fly in both the x and y directions are set to... The random and restore step sizes for fruit flies are set to... rand() represents the random number generation function, Maxgen is the maximum number of iterations, FR is the single flight range of the fruit fly, and t represents the number of iterations.

[0044] 4) After converting each combined quality feature in the combined quality feature set, the characteristics of complex nuclear power equipment products are obtained, and the most frequently occurring characteristics of complex nuclear power equipment products are taken as the key quality characteristics of complex nuclear power equipment products.

[0045] In section 4), each combined quality feature in the combined quality feature set is in decimal format. Each decimal combined quality feature is converted to binary format. Each bit of each binary combined quality feature represents a complex nuclear power equipment product characteristic. If the current bit is 0, it indicates the existence of a corresponding complex nuclear power equipment product characteristic; if the current bit is 1, it indicates the absence of a corresponding complex nuclear power equipment product characteristic. In specific implementation, the binary format combined quality feature is represented as (1, 1, 0, 1). The first bit indicates that the raw material origin is domestic; the second bit indicates that the product completion report is complete; the third bit indicates that the inspection and audit cycle is greater than 1 day but less than 2 days; and the fourth bit indicates that all equipment has a certified label. Its decimal format combined quality feature is 13, representing the aforementioned quality combination (1, 1, 0, 1).

[0046] The above embodiments do not limit the scope of protection of this invention. All equivalent modifications and variations made by those skilled in the art without departing from the overall concept of this invention are still within the scope of this invention.

Claims

1. A method for extracting product characteristics based on an improved fruit fly algorithm under value chain collaboration, characterized in that, Includes the following steps: 1) Based on the reliability, revenue, and resource consumption of complex nuclear power equipment products, construct a multi-objective optimization model for the quality characteristics of complex nuclear power equipment products; In section 1), the formula for the multi-objective optimization model of the quality characteristics of complex nuclear power equipment products is as follows: Among them, YX, ZY, and SY represent the reliability, resource consumption, and revenue of complex nuclear power equipment products, respectively. This is the j-th complex nuclear power equipment product characteristic. Credibility weight, and These are the characteristics of the j-th complex nuclear power equipment product. The weights of resource consumption and revenue; This represents the total number of characteristics of complex nuclear power equipment products. Indicating technical difficulty, Indicates the processing cycle. Indicate cost; , and These are the characteristics of the j-th complex nuclear power equipment product. The weights for technical difficulty, processing cycle, and cost; The six constraints of the multi-objective optimization model are as follows: in, It is the minimum value of credibility; It is the maximum value of resource consumption; It is the minimum value of the profit; This represents the maximum level of technical difficulty. This represents the maximum value of the processing cycle. It is the maximum cost; 2) Determine the range of the first and second combination quality characteristics based on the total number of characteristics of complex nuclear power equipment products; In 2), the range of the first combined quality characteristics ,satisfy ∈ The range of the second combination of quality characteristics ,satisfy ∈ ,in , , This indicates the number of complex nuclear power equipment product characteristics in the first set of quality characteristics. This indicates the number of complex nuclear power equipment product characteristics in the second set of quality characteristics. This represents the total number of characteristics of complex nuclear power equipment products. This represents the integer division operator; 3) Based on the range of the first and second combined quality characteristics, the improved fruit fly algorithm is used to solve the multi-objective optimization model and obtain the combined quality characteristic set of complex nuclear power equipment products; In the improved fruit fly algorithm described in 3), the position of the fruit fly in the x-direction is randomly generated within the range of the first combination of quality features, and the position of the fruit fly in the y-direction is randomly generated within the range of the second combination of quality features. In the improved fruit fly algorithm described in 3), the random and amplified step sizes of the fruit fly in the x and y directions are both set to... The random and restore step sizes for fruit flies are set to... , This represents the random number generation function, where Maxgen is the maximum number of iterations. This refers to the single flight range of a fruit fly. Indicates the number of iterations; 4) After converting each combined quality feature in the combined quality feature set, the characteristics of complex nuclear power equipment products are obtained, and the most frequently occurring characteristics of complex nuclear power equipment products are taken as the key quality characteristics of complex nuclear power equipment products. In step 4), each combined quality feature in the combined quality feature set is in decimal format. Each combined quality feature in decimal format is converted into binary format. Each bit of each combined quality feature in binary format represents a complex nuclear power equipment product characteristic. If the current bit is 0, it means that there is a corresponding complex nuclear power equipment product characteristic. If the current bit is 1, it means that there is no corresponding complex nuclear power equipment product characteristic.

Images