CN104504177A - Method for quickly configuring and designing large container crane - Google Patents

Abstract

The invention discloses a large-scale container crane product rapid configuration design method. The method includes: establishing a rapid design-oriented large-scale container crane product configuration framework; Carry out analysis and calculation, and build a constrained 0-1 programming gene evolution model; in combination with the large-scale characteristics of the large-scale container crane product configuration optimization problem, the present invention redesigns the pseudo-random proportional rule of the ant colony algorithm; and for the candidate gene For the problem that the number of instances is often large, a mechanism for reducing the number of candidate gene instances is proposed; an x_Sigmoid function is designed, and the transition probability is adjusted according to this function to speed up the algorithm search; the predatory strategy is implemented to strengthen the development of better regions, Enhanced search capabilities. By comprehensively using the model and algorithm proposed by the invention, a satisfactory preliminary design scheme of a large container crane can be obtained in a relatively short period of time, and has better flexibility.

Description

Design method for rapid configuration of large container cranes

technical field

The invention relates to the technical field of product configuration theory and method research, in particular to a large-scale container crane rapid configuration design method.

Background technique

Large-scale container crane is an evolutionary serialized and personalized complex mechanical product under the single-piece and small-batch production mode, and its design process has certain complexity. On the one hand, the conditions of each container terminal, the type of ship being loaded and unloaded, the water depth, and hydrological conditions will all lead to the diversification of large container cranes. For example, the reach, lifting height, and lifting capacity of large container cranes are different, depending on the type of ship being loaded and unloaded; the maximum wheel pressure and wheel base of the crane are different, depending on the geological conditions of the wharf, allowable wheel pressure, etc. In fact, the container cranes in major ports are not standard, and each container terminal has individual requirements for cranes. Therefore, each new type of container crane needs to select the standard parts of the mechanism and design the non-standard parts, thus requiring modularization and configuration design.

On the other hand, large-scale container cranes are evolutionary series mechanical products with multi-level product structure trees. There are different degrees of similarity and evolution relationship between old and new products, and between the same type of series products, making product configuration design unique. High-dimensional configuration space data. Determining the key characteristic parameters of product components through repeated checks and calculations is the basic feature of large container crane design. The analogy design and empirical design methods mainly used at present fail to make scientific use of product historical design information, and there is a lot of duplication of labor and huge waste of resources. Therefore, there is an urgent need to develop methods to support rapid product configuration design and improve design flexibility.

Contents of the invention

The invention provides a rapid configuration design method for a large container crane, which can obtain a satisfactory preliminary design scheme in a relatively short period of time, improve the efficiency and quality of product configuration design work, and reduce design costs.

In order to achieve the above object, the present invention provides a method for quickly configuring and designing a large container crane, which is characterized in that the method includes:

The above-mentioned large-scale container crane product configuration framework for rapid design includes:

Analyze and calculate the local similarity between the example template and customer needs, the satisfaction of local green characteristics, and the affinity for cooperation, etc., and construct a constrained 0-1 planning gene evolution model;

Combined with the large-scale characteristics of the large-scale container crane product configuration optimization problem, the present invention redesigns the pseudo-random proportional rule of the ant colony algorithm; Mechanism; a x_Sigmoid function is designed, and the transition probability is adjusted according to this function to speed up the search speed of the algorithm; the predation strategy is implemented to strengthen the development of better areas and enhance the search ability.

The above-mentioned rapid configuration design framework for large container crane products includes:

Change the traditional analogy design and experience design mode to improve the efficiency of large container crane product design.

First, construct the product configuration gene evolution model and initialize the basic parameters of the model.

Then, based on the improved ant colony, a rapid product configuration design algorithm is established, and the preliminary design scheme is obtained by using this algorithm.

Finally, on the basis of the preliminary design scheme, a three-dimensional model of the non-standard parts of the traveling mechanism of the cart is established. According to the user's requirements, such as the strength and stiffness of the cart structure, check the relevant constraints. If it is satisfied, the design of the non-standard part is successful; if it is not satisfied, the non-standard part is designed with variable parameters and features, such as thinning the plate thickness and reducing the height, etc., and re-checking until the conditions are met , and add it to the instance library.

The above analyzes and calculations are carried out on the local similarity between the example template and customer needs, the satisfaction of local green characteristics, and the affinity for cooperation, etc., and the construction of a constrained 0-1 planning gene evolution model includes:

Use the formulas of local similarity, local green feature satisfaction, and coordination affinity to analyze and calculate product configuration schemes, and establish the selection, coordination, correlation, and cost of product virtual feature genes with the goal of maximizing the adaptability of product configuration schemes. and a weight-constrained 0-1 programming gene evolution model.

The above redesigned pseudo-random ratio rules of the ant colony algorithm include:

Each ant establishes a configuration design scheme by repeatedly applying the pseudo-random proportional rule of formula (1). The pseudo-random scaling rule is designed to allow more selection opportunities for gene instances with high fitness and a large amount of pheromone. Indicates the probability that ant l selects gene instance _Xij .

in, The local similarity describes the degree of similarity between the gene instance X _i,j in the instance template and the customer requirement R _i,k . is the intensity of pheromone track on gene instance Xi _,j . beta, is two parameters, β reflects the relative importance of local similarity information in the optional gene instance, Reflects the relative importance of the information accumulated in the ant movement process in the optional gene instance. The size of the parameter _θ0 determines the relative importance between exploiting prior knowledge and exploring new gene instances. Whenever an ant is ready to select a gene instance X _i,j , the algorithm generates a random number 0≤θ≤1. If θ≤θ ₀ , the best gene instance is selected according to prior knowledge, otherwise it is selected by probability.

In view of the above-mentioned problem that the number of candidate gene instances is often large, a mechanism for reducing the number of candidate gene instances is proposed, including:

In the large-scale product allocation problem, there are many gene examples for some gene positions, which will reduce the search speed of the algorithm. In order to reduce the number of candidate gene instances, the present invention sets the number of selectable gene instances of the i-th gene position (n gene positions in total) to M _i , and adjusts M _i to M _i according to the following mechanism for reducing the number of candidate gene instances '.

Else, M _i '＝M _i

Among them, int[] is rounded down; ξ is a positive integer, which is the threshold of the number of optional genes; is the difference ratio of gene instances, The proportion of candidate gene instances that are reduced is then:

1. Redesigned the pseudo-random ratio rule of the ant colony algorithm;

Each ant establishes a configuration design scheme by repeatedly applying the pseudo-random proportional rule of formula (1). The pseudo-random scaling rule is designed to allow more selection opportunities for gene instances with high fitness and a large amount of pheromone. Indicates the probability that ant l selects gene instance _Xij .

in, The local similarity describes the degree of similarity between the gene instance X _i,j in the instance template and the customer requirement R _i,k . is the intensity of pheromone track on gene instance Xi _,j . beta, is two parameters, β reflects the relative importance of local similarity information in the optional gene instance, Reflects the relative importance of the information accumulated in the ant movement process in the optional gene instance. The size of the parameter _θ0 determines the relative importance between exploiting prior knowledge and exploring new gene instances. Whenever an ant is ready to select a gene instance X _i,j , the algorithm generates a random number 0≤θ≤1. If θ≤θ ₀ , the best gene instance is selected according to prior knowledge, otherwise it is selected by probability.

2. The rapid configuration design method for large container cranes according to claim 1, characterized in that, in view of the problem that the number of candidate gene instances is often large, a mechanism for reducing the number of candidate gene instances is proposed.

In the large-scale product allocation problem, there are many gene examples for some gene positions, which will reduce the search speed of the algorithm. In order to reduce the number of candidate gene instances, the present invention sets the number of selectable gene instances of the i-th gene position (n gene positions in total) to M _i , and adjusts M _i to M _i according to the following mechanism for reducing the number of candidate gene instances '.

Else, M _i '＝M _i

Among them, int[] is rounded down; ξ is a positive integer, which is the threshold of the number of optional genes; is the difference ratio of gene instances, The proportion of candidate gene instances that are reduced is then:

The proportion of reduced candidate gene instances mainly depends on the proportion of M _i >ξ, the relative size of M _i and ξ, , the higher the ratio of _Mi >ξ, the larger _Mi is relative to ξ, The smaller is, the higher the proportion of candidate gene instances is reduced, that is, the greater the amount of calculation is reduced.

The proportion of reduced candidate gene instances mainly depends on the proportion of M _i >ξ, the relative size of M _i and ξ, , the higher the ratio of _Mi >ξ, the larger _Mi is relative to ξ, The smaller is, the higher the proportion of candidate gene instances is reduced, that is, the greater the amount of calculation is reduced.

A x_Sigmoid function is designed above, and the transition probability is adjusted according to this function to speed up the algorithm search speed, including:

Theorem: For a mapping Let T _i,j , The corresponding transition probabilities are P _i,j , like For any T _i,j1 < T _i,j2 has but $\frac{P_{i,j1}}{P_{i,j2}}>\frac{P_{i,j1}^{\mathrm{new}}}{P_{i,j2}^{\mathrm{new}}}.$

Basic idea: make From formula (1), we have $P_{x_{\mathrm{ij}}}^l=\frac{T_{i,j}}{\underset{i=1}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}{T}_{i,j}}$

In the ant colony algorithm, the choice of transition probability has a great influence on the algorithm search process. Aiming at the optimization problem of large-scale product configuration, in order to further speed up the algorithm search speed, the present invention adjusts T _i,j . Among the M _i T _i,j , the smaller T _i,j should be lowered; the larger T _i,j should not be greatly changed. In this way, the transition probability corresponding to the smaller T _i,j is reduced, and the transition probability corresponding to the larger T _i,j is increased; but several larger T _i,j correspond to the transition probability (that is, the probability of being selected for the corresponding gene instance) The relative size did not change much. This adjustment is conducive to speeding up the search speed of the algorithm, and is also conducive to avoiding the algorithm from falling into local extremum.

Design of x_Sigmoid function: when T _i,j is small, should be small; when T _i,j is large, should remain essentially unchanged.

However, after the Sigmoid function shifts to the right x ₁ > 0, when the independent variable is smaller than x ₁ , the dependent variable is smaller; when the independent variable is greater than x ₁ , the dependent variable is relatively large, and the larger the independent variable, the closer the dependent variable is to 1. That is, after the sigmoid function is shifted to the right by x ₁ , the same as unanimous.

Therefore, first linearly convert T _i,j to [0,2x ₁ ] then let Varies according to the Sigmoid function.

First perform the conversion, let T _max ＝maxT _i,j , T _min ＝minT _i,j , convert the midpoint (T _max +T _min )/2 of T _max and T _min to x ₁ , then T _i,j is converted for:

${\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}}}{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}}{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

Right now in the range of $[\frac{T_{\min }}{T_{\max }+T_{\min }}{2x}_1,\frac{T_{\max }}{T_{\max }+T_{\min }}{2x}_1]\&Subset;[0,{2x}_1].$

Then, take $T_{i,j}^{\mathrm{new}}/T_{i,j}^{*}=\frac{1}{1+\exp (x_1-T_{i,j}^{*})}---\left(4\right)$

Right now $T_{i,j}^{\mathrm{new}}=\frac{T_{i,j}^{*}}{1+\exp (x_1-T_{i,j}^{*})}---\left(5\right)$

The present invention refers to this function as x_Sigmoid function.

Transition probability adjustment strategy: For any T _i,j1 <T _i,j2 , since T _i,j1 ,T _i,j2 , T _max ,T _min ,x ₁ >0, after conversion according to formula (3), we have

The Sigmoid function is a monotonically increasing function, then the functional formula (4) after the Sigmoid function is shifted to the right by x ₁ is also monotonically increasing, so

According to Theorem 1, if T _{i, j} is mapped according to equations (3) and (5) as And recalculate the transition probability, the relatively small original transition probability will be smaller, and the relatively large original transition probability will be greater.

The above-mentioned implementation of predation strategies, strengthening the development of better areas, and enhancing search capabilities include:

At the end of each generation, the top χ ants with the best search results are used as escaped ants, and then η predator ants are used to hunt them down. For predatory ants, temporarily update the pheromone of each gene instance, as shown in formula (6).

${\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&rho;</span>}<span\; class="notranslate">\; \&times;</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&Delta;\&tau;</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; +</span>{{\mathrm{<span\; class="notranslate">\; \&Delta;\&tau;</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

in, Indicates the amount of pheromone left after the l-th ant selects the gene instance X _{i, j} , the higher the similarity, the more pheromone is released, which is attractive to ants; Indicates the pheromone increment of the gene instance Xi _{, j} in this cycle;

${{\mathrm{\&Delta;\&tau;}}^{\&prime;}}_{x_{i,j}}=\underset{\mathrm{\&gamma;}=1}{\overset{\mathrm{\&chi;}}{\mathrm{\&Sigma;}}}{\mathrm{\&Delta;\&tau;}}_{x_{i,j}}^{\mathrm{\&gamma;}},{\mathrm{\&Delta;\&tau;}}_{x_{i,j}}^{\mathrm{\&gamma;}}=\left\{\begin{array}{ccc}{Q}^{\&prime;}\&times;S_{(x_{i,j},R_{i,k})},& \mathrm{if}& G^{\mathrm{\&gamma;}}=x_{i,j}\\ 0,& \mathrm{if}& G^{\mathrm{\&gamma;}}\&NotEqual;x_{i,j}\end{array}\right.,$ Among them, Q'>Q, this kind of pheromone is temporarily updated, which is only applicable to the current generation of predatory ants, and not applicable to the search of future generations. The purpose is to strengthen the search in the neighborhood of escaped ants and speed up the search speed of the algorithm.

Compared with the analog design and empirical design methods currently used, the rapid configuration design method for large container crane products of the present invention has the advantage that it has high-dimensional configuration space data for the product configuration design of large container cranes. The present invention proposes an alternative gene Mechanism for reducing the number of examples; a kind of x_Sigmoid function is designed, and the transition probability is adjusted according to this function, so as to speed up the algorithm search speed; the strategy of predation is implemented, the development of better areas is strengthened, and the search ability is enhanced; the model and algorithm proposed by the present invention can be used comprehensively Obtain a satisfactory preliminary design scheme for large container cranes in a short period of time, with better flexibility.

After the preliminary design plan obtained by using the method of the invention is checked, parameter-varied and feature-variable designed, the configuration design result of the large-scale container crane can be viewed. Compared with the design results obtained by the traditional design mode, the method of the present invention can quickly obtain the preliminary design scheme close to the customer's demand, and then through less variable parameter and variable feature design work, the design that meets the customer's demand can be realized, compared with the traditional design mode The design time is greatly reduced and the design cost is saved.

Description of drawings

Fig. 1 is the frame diagram of the rapid configuration design method for large-scale container crane products of the present invention

Fig. 2 is the flowchart of the rapid configuration design method for large container crane products of the present invention

Fig. 3 is the method transfer strategy parameter of the present invention Variation with T _i,j

Fig. 4 is the method transfer strategy parameter of the present invention Variation with p _i,j

Fig. 5 is the box diagram comparison of the inventive method and other four kinds of method optimization results

Detailed ways

Specific embodiments of the present invention will be further described below in conjunction with the accompanying drawings.

As shown in Fig. 1, the present invention discloses a fast configuration design method for large-scale container crane products, and the method includes the following steps:

The traditional design method mainly adopts the method of analogy design or empirical design. According to the function and parameter requirements of the customer order, the designer borrows or appropriately modifies the similar design examples in the past based on experience to generate a new design scheme. This design method fails to make scientific use of product historical design information, which is time-consuming and labor-intensive, and the efficiency is relatively low.

In order to improve the efficiency of product design of large container cranes, the invention changes the traditional design mode and proposes a rapid design-oriented large container crane product configuration framework:

First, construct the product configuration gene evolution model and initialize the basic parameters of the model.

Then, a rapid product configuration design algorithm is established based on the improved ant colony, and the preliminary design scheme is obtained by using the algorithm.

Finally, on the basis of the preliminary design scheme, a three-dimensional model of the non-standard parts of the traveling mechanism of the cart is established. According to the user's requirements, such as the strength and stiffness of the cart structure, check the relevant constraints. If it is satisfied, the design of the non-standard part is successful; if it is not satisfied, the non-standard part is designed with variable parameters and features, such as thinning the plate thickness and reducing the height, etc., and re-checking until the conditions are met , and add it to the instance library.

As shown in Figure 2, the flow chart of the rapid configuration and design method for large-scale container crane products of the present invention specifically includes the following steps:

Step 1. Initialize basic parameters, including α, θ ₀ ,β, ρ, Q, etc.

According to the initialization rules, the amount of pheromone on each gene instance is equal, and set (0)=ψ (ψ is a constant). α ants are randomly placed on n product virtual feature genes in the initial stage.

Step 2. Adjust the transition probability based on the x_Sigmoid function.

Step 3. Ants select parts according to the pseudo-random proportional rule.

Ant l (l=1,2,...,α) is affected by local similarity information (they tend to choose gene instances with high similarity) and pheromone trajectory intensity information (the pheromone intensity is high Instances of genes that are more attractive to ants) guide.

Step 4: After the ant completes a cycle, it modifies the pheromone quantity of the gene instance according to the local update rule.

Step 5, implement the predation strategy, strengthen the search in the neighborhood of escaped ants, and speed up the search speed of the algorithm.

Step 6. Calculate the fitness with the formula (5), and determine whether the configuration design scheme is a preliminary design scheme for non-standard parts according to the fitness threshold and the iteration number threshold. "Yes" executes Step8, "no" executes Step7.

Step 7. Once all ants have completed their configuration design search, modify the pheromone quantity of the gene instance again according to the global update rule, and then return to step 2.

Step 8, output the result. the

For the image description, if x ₁ =5, then The variation with T _i,j is shown in Fig. 3. It can be seen from Figure 3 that when T _i,j is small, is very small; when T _i,j is large, The change is small, close to 1.

If M _i =10, T _i,j =2.6604,0.6472,1.6992,1.3608,2.4956,2.1339,1.2781,0.0518,2.2999,1.2452, The variation with P _i,j is shown in Fig. 4. It can be seen from Figure 4 that the transition probability that was originally small is now even smaller; the transition probability that was originally large is now increased, and the multiples of the increase are similar, that is, their relative sizes do not change much. That is to say, if the transition probability is originally small, it will decrease approximately exponentially, and if the transition probability is originally large, it will increase approximately linearly.

Below in conjunction with Fig. 5, illustrate the embodiment of rapid configuration design:

Adopt method of the present invention, basic ant colony algorithm, ant colony algorithm based on evolution strength, improved ant colony algorithm based on particle swarm parameter optimization, ant colony algorithm based on quantum space to carry out fast configuration optimization calculation, the ant structure of these four kinds of algorithms simultaneously , population size, number of iterations are the same as the method of the present invention. Use the product configuration design system (developed using Pro/E 3D modeling software, VB language, Access database, etc.) to carry out product configuration. The computer configuration is: Pentium Dual 2.00GHZ, 2GHz memory, Windows XP. According to the constrained 0-1 programming gene evolution model, these five algorithms were run continuously for 50 times, and the box plot comparisons of fitness, convergence algebra, and convergence time are shown in Figure 5. The coordinates 1, 2, 3, 4, and 5 on the horizontal axis in Figure 5 correspond to the basic ant colony algorithm, the ant colony algorithm based on evolutionary strength, the improved ant colony algorithm based on particle swarm parameter optimization, the ant colony algorithm based on quantum space, and the original ant colony algorithm, respectively. Invention method. The average convergence result of the method of the invention is superior to other methods, and the convergence speed is further improved compared with the other four methods. The method of the invention can quickly obtain a preliminary design scheme close to the customer's needs, and then through less variable parameter and variable feature design work, the design that meets the customer's needs can be realized, which greatly reduces the design time and saves the design cost compared with the traditional design mode .

Although the content of the present invention has been described in detail through the above preferred embodiments, it should be understood that the above description should not be considered as limiting the present invention. Various modifications and alterations to the present invention will become apparent to those skilled in the art upon reading the above disclosure. Therefore, the protection scope of the present invention should be defined by the appended claims.

Claims (7)

1. A rapid configuration design method for a large container crane, characterized in that the method comprises: Establish a rapid design-oriented product configuration framework for large container cranes; Analyze and calculate the local similarity between the example template and customer needs, the satisfaction of local green characteristics, and the affinity for cooperation, etc., and construct a constrained 0-1 planning gene evolution model; Combined with the large-scale characteristics of the large-scale container crane product configuration optimization problem, the present invention redesigns the pseudo-random proportional rule of the ant colony algorithm; Mechanism; a x_Sigmoid function is designed, and the transition probability is adjusted according to this function to speed up the search speed of the algorithm; the predation strategy is implemented to strengthen the development of better areas and enhance the search ability. 2. The large-scale container crane rapid configuration design method according to claim 1, wherein said establishment of a large-scale container crane product rapid configuration design framework includes: Change the traditional analogy design and experience design mode to improve the efficiency of large container crane product design. First, construct the product configuration gene evolution model and initialize the basic parameters of the model. Then, based on the improved ant colony, a rapid product configuration design algorithm is established, and the preliminary design scheme is obtained by using this algorithm. Finally, on the basis of the preliminary design scheme, a three-dimensional model of the non-standard parts of the traveling mechanism of the cart is established. According to the user's requirements, such as the strength and stiffness of the cart structure, check the relevant constraints. If it is satisfied, the design of the non-standard part is successful; if it is not satisfied, the non-standard part is designed with variable parameters and features, such as thinning the plate thickness and reducing the height, etc., and re-checking until the conditions are met , and add it to the instance library. 3. The large-scale container crane rapid configuration design method according to claim 1, characterized in that, analyzing and calculating the local similarity between the example template and the customer's needs, the satisfaction of local green characteristics, and the cooperation affinity, etc., to construct a belt The constrained 0-1 programming gene evolution model includes: Use the formulas of local similarity, local green feature satisfaction, and coordination affinity to analyze and calculate product configuration schemes, and establish the selection, coordination, correlation, and cost of product virtual feature genes with the goal of maximizing the adaptability of product configuration schemes. and a weight-constrained 0-1 programming gene evolution model. 4. the large-scale container crane quick disposition design method as claimed in claim 3, is characterized in that, redesigned the pseudo-random ratio rule of ant colony algorithm; Each ant establishes a configuration design scheme by repeatedly applying the pseudo-random proportional rule of formula (1). The pseudo-random scaling rule is designed to allow more selection opportunities for gene instances with high fitness and a large amount of pheromone. Indicates the probability that ant l selects gene instance _Xij . in, The local similarity describes the degree of similarity between the gene instance X _i,j in the instance template and the customer requirement R _i,k . is the intensity of pheromone track on gene instance Xi _,j . beta, is two parameters, β reflects the relative importance of local similarity information in the optional gene instance, Reflects the relative importance of the information accumulated in the ant movement process in the optional gene instance. parameter The size of σ determines the relative importance between exploiting prior knowledge and exploring new gene instances. Whenever an ant is about to select a gene instance X _i,j , the algorithm generates a random number if Then choose the best gene instance according to prior knowledge, otherwise choose according to probability. 5. The rapid configuration design method for large container cranes according to claim 1, characterized in that, for the problem that the number of candidate gene instances is often large, a mechanism for reducing the number of candidate gene instances is proposed. In the large-scale product allocation problem, there are many gene examples for some gene positions, which will reduce the search speed of the algorithm. In order to reduce the number of candidate gene instances, the present invention sets the number of selectable gene instances of the i gene position (n gene positions in total) as M _i , and adjusts M _i to M _i according to the following mechanism for reducing the number of candidate gene instances '. If M _i >ξ, Else,M _i '=M _i Among them, int[] is rounded down; ξ is a positive integer, which is the threshold of the number of optional genes; is the difference ratio of gene instances, The proportion of candidate gene instances that are reduced is then: The proportion of reduced candidate gene instances mainly depends on the proportion of M _i >ξ, the relative size of M _i and ξ, , the higher the ratio of _Mi >ξ, the larger _Mi is relative to ξ, The smaller is, the higher the proportion of candidate gene instances is reduced, that is, the greater the amount of calculation is reduced. 6. the large-scale container crane quick disposition design method as claimed in claim 1, is characterized in that, designed a kind of x_Sigmoid function, and adjust transition probability according to this function, accelerate algorithm search speed; Theorem: For a mapping make The corresponding transition probabilities are P _i,j , like For any T _i,j1 < T _i,j2 has but Basic idea: make From formula (1), we have In the ant colony algorithm, the choice of transition probability has a great influence on the algorithm search process. Aiming at the optimization problem of large-scale product configuration, in order to further speed up the algorithm search speed, the present invention adjusts T _i,j . Among the M _i T _i,j , the smaller T _i,j should be lowered; the larger T _i,j should not be greatly changed. In this way, the transition probability corresponding to the smaller T _i,j is reduced, and the transition probability corresponding to the larger T _i,j is increased; but several larger T _i,j correspond to the transition probability (that is, the probability of being selected for the corresponding gene instance) The relative size did not change much. This adjustment is conducive to speeding up the search speed of the algorithm, and is also conducive to avoiding the algorithm from falling into local extremum. Design of x_Sigmoid function: when T _i,j is small, should be small; when T _i,j is large, should remain essentially unchanged. However, after the Sigmoid function shifts to the right x ₁ > 0, when the independent variable is smaller than x ₁ , the dependent variable is smaller; when the independent variable is greater than x ₁ , the dependent variable is relatively large, and the larger the independent variable, the closer the dependent variable is to 1. That is, after the sigmoid function is shifted to the right by x ₁ , the same as unanimous. Therefore, first linearly convert T _{i, j} to [0,2x ₁ ] interval then let Varies according to the Sigmoid function. First perform the conversion, let T _max ＝maxT _i,j , T _min ＝minT _i,j , convert the midpoint (T _max +T _min )/2 of T _max and T _min to x ₁ , then T _i,j is converted for: Right now in the range of Then, take Right now The present invention refers to this function as x_Sigmoid function. Transition probability adjustment strategy: For any T _i,j1 <T _i,j2 , since T _i,j1 ,T _i,j2 , T _max ,T _min ,x ₁ >0, after conversion according to formula (3), we have The Sigmoid function is a monotonically increasing function, then the functional formula (4) after the Sigmoid function is shifted to the right by x ₁ is also monotonically increasing, so According to Theorem 1, if T _{i, j} is mapped according to equations (3) and (5) as And recalculate the transition probability, the relatively small original transition probability will be smaller, and the relatively large original transition probability will be greater. For the image description, if x ₁ =5, then The variation with T _i,j is that when T _i,j is small, is very small; when T _i,j is large, The change is small, close to 1. If M _i =10, T _i,j =2.6604,0.6472,1.6992,1.3608,2.4956,2.1339,1.2781,0.0518,2.2999,1.2452, The change with P _i,j is that the transition probability that was originally small is now smaller; the transition probability that was originally large is now increased, and the multiples of the increase are similar, that is, their relative size changes little. That is to say, if the transition probability is originally small, it will decrease approximately exponentially, and if the transition probability is originally large, it will increase approximately linearly. 7. The design method for rapid configuration of large container cranes according to claim 1, characterized in that, the strategy of predation is implemented, the development of better areas is strengthened, and the search ability is enhanced. At the end of each generation, the top χ ants with the best search results are used as escaped ants, and then η predator ants are used to hunt them down. For predatory ants, temporarily update the pheromone of each gene instance, as shown in formula (6). in, Indicates the amount of pheromone left after the l-th ant selects the gene instance X _{i, j} , the higher the similarity, the more pheromone is released, which is attractive to ants; Indicates the pheromone increment of the gene instance Xi _{, j} in this cycle; Among them, Q'>Q, this kind of pheromone is temporarily updated, which is only applicable to the current generation of predatory ants, and not applicable to the search of future generations. The purpose is to strengthen the search in the neighborhood of escaped ants and speed up the search speed of the algorithm.