CN110559664A - game hero outgoing recommendation method and system based on multi-objective optimization - Google Patents

Abstract

The invention discloses a game hero outfit recommendation method and system based on multi-objective optimization, and relates to the technical field of mobile games and optimization algorithms, including generating an iterative population including M individuals according to the acquired game hero attribute table and equipment attribute table. Individual characterizes the game hero outfit recommendation scheme; calculates the fitness value of each individual in the iterative population according to the constructed multi-objective fitness function, which is used to calculate the sum of the attack value and defense value of the game hero; Determine whether the current number of iterations is less than the total number of preset iterations. If so, determine the individual with the largest fitness value as the recommended outfit for the game hero. Otherwise, perform cross-mutation selection operations on the individuals in the iterative population to update the iterative population, and return Steps to calculate individual fitness values until the end of the iteration. The application of the present invention can provide users with an optimal outfit recommendation scheme for game heroes, maximizing the attack and defense capabilities of game heroes.

Description

A method and system for recommending game hero outfits based on multi-objective optimization

technical field

The invention relates to the technical field of mobile games and optimization algorithms, in particular to a multi-objective optimization-based game hero outfit recommendation method and system.

Background technique

Dota mobile games have always been loved by gamers. For example, the dota-like mobile game "Glory of the King" launched by Tianmei Studio under Tencent, the gameplay of this game is mainly competitive. There are 53 kinds of third-level equipment (top equipment) in the game. The attributes of each equipment are different, and the outfits of game heroes with different attributes are also different in the game. In the game, each game hero can purchase 6 types of equipment, and there are a total of 22,957,480 types of equipment combinations. At present, the in-game system provides 3 outfit recommendations for each game hero, but for different game lineups, these three outfit recommendations cannot meet the needs of players.

Contents of the invention

The purpose of the present invention is to provide a game hero outfit recommendation method and system based on multi-objective optimization, so as to maximize the attack and defense capabilities of the game hero.

To achieve the above object, the present invention provides the following scheme:

A game hero outfit recommendation method based on multi-objective optimization, including:

Obtain a game hero attribute table and an equipment attribute table; both the game hero attribute table and the equipment attribute table are in tabular form; the first column of the game hero attribute table is the game hero category, and the first row of the game hero attribute table is The second column is the physical attack value of the game hero, the third column of the game hero attribute table is the magic attack value of the game hero, the fourth column of the game hero attribute table is the physical defense value of the game hero, and the game hero attribute table is listed as the physical defense value of the game hero. The fifth column of the attribute table is the spell defense value of the game hero; the first column of the equipment attribute table is the equipment number, the second column of the equipment attribute table is the equipment name, and the third column of the equipment attribute table is The physical attack value of the equipment, the fourth column of the equipment attribute table is the magic attack value of the equipment, the fifth column of the equipment attribute table is the physical defense value of the equipment, and the sixth column of the equipment attribute table is the equipment's magic defense value;

Generate an iterative population according to the game hero attribute table and the equipment attribute table; the iterative population includes M individuals who characterize the game hero outfit recommendation scheme, and the game heroes represented by all of the individuals are of the same category; each Each of the individuals includes 6 sets of different equipment, and the structure of each individual is in the form of an array of 6 rows and 5 columns; wherein, a row of the individual represents the number and attribute of an equipment, and the individual Each row of contains 5 elements, which are the equipment number, the physical attack value of the equipment, the magic attack value of the equipment, the physical defense value of the equipment, and the magic defense value of the equipment;

Calculate the fitness value of each individual in the iterative population according to the constructed multi-objective fitness function, and record the current number of iterations; the multi-objective fitness function is used to calculate the sum of the attack value and defense value of the game hero ;

Judging whether the current number of iterations is less than the total number of preset iterations to obtain a first judgment result;

If the first judgment result indicates that the current number of iterations is not less than the total number of preset iterations, then the individual with the largest fitness value is determined as the outfit recommendation plan for the game hero;

If the first judgment result indicates that the current number of iterations is less than the total number of preset iterations, then add 1 to the current number of iterations, and then perform a crossover mutation operation on the individuals in the iterative population corresponding to the current number of iterations to generate new individuals, and calculate the fitness value of each of the new individuals according to the multi-objective fitness function; the total number of the new individuals is M;

Add all the new individuals to the iterative population corresponding to the current iteration number to update the iterative population, and sort the individuals in the updated iterative population according to the fitness value of each individual in the updated iterative population, select The first M individuals are used as the iterative population of the next generation, and then return to the step of calculating the fitness value of each individual in the iterative population according to the constructed multi-objective fitness function, and recording the current iteration number.

Optionally, the method for recommending game hero outfits also includes:

Reserving the iteration population corresponding to the current number of iterations not less than the preset total number of iterations;

Calculating the first target value and the second target value of each individual in the iterative population retained; the first target value is the attack value of the game hero, and the second target value is the defense value of the game hero;

Calculate the third target value of each individual in the retained iterative population; the third target value is the ratio of the first target value to the second target value;

The individuals in the retained iterative population are sorted in descending order based on the third target value of each individual as the sorting standard.

Optionally, the formula for calculating the first target value is The calculation formula of the second target value is X _m represents the mth individual, i represents the number of rows of the m-th individual; w ₁ , w ₂ , w ₃ , and w ₄ represent the physical attack value, magic attack value, physical defense value, and magic defense value of the game hero represented by the m-th individual, respectively.

Optionally, the multi-objective fitness function is F(X _m )=f ₁ (X _m )+f ₂ (X _m ).

Optionally, the generating iterative population according to the game hero attribute table and the equipment attribute table specifically includes:

According to the game hero attribute table and the equipment attribute table, an individual is generated and initialized randomly by means of encoding, and an iterative population including M individuals is obtained.

Optionally, adding all the new individuals to the iterative population corresponding to the current number of iterations to update the iterative population, and updating the updated iterative population according to the fitness value of each individual in the updated iterative population Individuals are sorted, and the top M individuals are selected as the iterative population of the next generation, including:

All the new individuals are added to the iterative population corresponding to the current number of iterations to obtain an updated iterative population; the number of individuals in the updated iterative population is 2M;

According to the fitness value of each individual in the updated iterative population, the Pareto dominance algorithm is used to select the individual in the updated iterative population;

Sort the individuals that have been selected by dominance, and select the top M individuals as the iterative population of the next generation.

A game hero outfit recommendation system based on multi-objective optimization, including:

An information acquisition module, configured to obtain a game hero attribute table and an equipment attribute table; both the game hero attribute table and the equipment attribute table are in the form of tables; the first column of the game hero attribute table is a game hero category, and the The second column of the game hero attribute table is the physical attack value of the game hero, the third column of the game hero attribute table is the game hero's magic attack value, and the fourth column of the game hero attribute table is the game hero's physical defense value, the fifth column of the game hero attribute table is the spell defense value of the game hero; the first column of the equipment attribute table is the equipment number, the second column of the equipment attribute table is the equipment name, and the equipment attribute The third column of the table is the physical attack value of the equipment, the fourth column of the equipment attribute table is the magic attack value of the equipment, the fifth column of the equipment attribute table is the physical defense value of the equipment, and the The sixth column is the spell defense value of the equipment;

The iterative population generation module is used to generate an iterative population according to the game hero attribute table and the equipment attribute table; the iterative population includes M individuals who characterize the game hero outfit recommendation scheme, and the game heroes represented by all the individuals They are all of the same category; each of the individuals includes 6 sets of equipment that are different from each other, and the structure of each of the individuals is in the form of an array of 6 rows and 5 columns; wherein, a row of the individual represents a piece of equipment Number and attributes, each row of the individual includes 5 elements, which are respectively the equipment number, the physical attack value of the equipment, the magic attack value of the equipment, the physical defense value of the equipment, and the magic defense value of the equipment;

The fitness value calculation module is used to calculate the fitness value of each individual in the iterative population according to the constructed multi-objective fitness function, and record the current number of iterations; the multi-objective fitness function is used to calculate the game hero The sum of attack value and defense value;

A judging module, configured to judge whether the current number of iterations is less than the total number of preset iterations to obtain a first judgment result;

The game hero outfit recommendation scheme determination module is used to determine the individual with the largest fitness value as the game hero outfit recommendation scheme when the first judgment result indicates that the current iteration number is not less than the preset iteration total number ;

A cross mutation operation module, configured to add 1 to the current number of iterations when the first judgment result indicates that the current number of iterations is less than the total number of preset iterations, and then add 1 to the iteration population corresponding to the current number of iterations The individual of the individual carries out the cross mutation operation to generate new individual, and calculates the fitness value of each described new individual according to described multi-objective fitness function; The total number of described new individual is M;

The iterative population update module is used to add all the new individuals to the iterative population corresponding to the current number of iterations to update the iterative population, and update the updated iterative population according to the fitness value of each individual in the updated iterative population The individuals in are sorted, and the top M individuals are selected as the iterative population of the next generation, and then returned to the fitness value calculation module.

Optionally, the game hero outfit recommendation system also includes:

A retaining module, configured to retain the iterative population corresponding to the current number of iterations that is not less than the preset total number of iterations;

The first target value and the second target value calculation module are used to calculate the first target value and the second target value of each individual in the iterative population retained; the first target value is the attack value of the game hero, so The second target value is the defense value of the game hero;

The third target value calculation module is used to calculate the third target value of each individual in the retained iterative population; the third target value is the ratio of the first target value to the second target value;

The descending order module is used to arrange the individuals in the retained iterative population in descending order by using the third target value of each individual as a sorting standard.

Optionally, the iterative population generation module specifically includes:

The iterative population generation unit is used to generate an individual and randomly initialize it according to the game hero attribute table and the equipment attribute table in a coding manner, so as to obtain an iterative population including M individuals.

Optionally, the iterative population update module specifically includes:

An update unit, configured to add all the new individuals to the iterative population corresponding to the current number of iterations to obtain an updated iterative population; the number of individuals in the updated iterative population is 2M;

The dominant selection unit is used to perform dominant selection on the individuals in the updated iterative population by using the Pareto dominance algorithm according to the fitness value of each individual in the updated iterative population;

The individual selection unit is used to sort the individuals selected by the dominant, select the first M individuals as the iterative population of the next generation, and jump to the fitness value calculation module.

According to the specific embodiments provided by the invention, the invention discloses the following technical effects:

The present invention provides a game hero outfit recommendation method and system based on multi-objective optimization. The multi-objective optimization algorithm is used to establish the goal of maximizing the attack ability and defense ability of game heroes, and combined with genetic algorithm for iterative optimization, the obtained The optimal game hero outfit recommendation scheme meets user needs and maximizes the attack and defense capabilities of game heroes.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the accompanying drawings required in the embodiments. Obviously, the accompanying drawings in the following description are only some of the present invention. Embodiments, for those of ordinary skill in the art, other drawings can also be obtained according to these drawings without paying creative labor.

FIG. 1 is a schematic flow diagram of a game hero outfit recommendation method based on multi-objective optimization according to an embodiment of the present invention;

FIG. 2 is a schematic structural diagram of a game hero outfit recommendation system based on multi-objective optimization according to an embodiment of the present invention;

FIG. 3 is a schematic flowchart of a method for recommending outfits for heroes in the game "Honor of Kings" based on multi-objective optimization according to an embodiment of the present invention;

Fig. 4 is a schematic structural diagram of an individual solution according to an embodiment of the present invention;

5 is a schematic diagram of a genetic algorithm crossover operation according to an embodiment of the present invention;

Fig. 6 is a schematic diagram of a mutation operation of a genetic algorithm according to an embodiment of the present invention.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

The purpose of the present invention is to provide a game hero outfit recommendation method and system based on multi-objective optimization, so as to maximize the attack and defense capabilities of the game hero.

In order to make the above objects, features and advantages of the present invention more comprehensible, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

Embodiment one

As shown in FIG. 1 , the multi-objective optimization-based outfit recommendation method for game heroes provided in this embodiment includes the following steps.

Step 101: Obtain a game hero attribute table and an equipment attribute table; both the game hero attribute table and the equipment attribute table are in the form of tables; the first column of the game hero attribute table is the game hero category, and the game hero attribute The second column of the table is the physical attack value of the game hero, the third column of the game hero attribute table is the magic attack value of the game hero, and the fourth column of the game hero attribute table is the physical defense value of the game hero. The fifth column of the game hero attribute table is the spell defense value of the game hero; the first column of the equipment attribute table is the equipment number, the second column of the equipment attribute table is the equipment name, and the first column of the equipment attribute table is the equipment name. The third column is the physical attack value of the equipment, the fourth column of the equipment attribute table is the magic attack value of the equipment, the fifth column of the equipment attribute table is the physical defense value of the equipment, and the sixth column of the equipment attribute table It is the spell defense value of the equipment.

The specific process of step 101 is: currently there are 93 game heroes and 53 kinds of third-level equipment (top equipment) in the game. In order to give full play to the maximum performance of each equipment, this embodiment only retains 4 attributes of game heroes and equipment. They are physical attack, magic attack, physical defense and magic defense. The remaining attributes of the equipment are converted into physical attack, magic attack, physical defense and magic defense. The converted equipment attributes are shown in Table 1, and the game hero attributes are shown in Table 2.

Table 1 Equipment attributes after conversion

Table 2 Game hero attributes

Step 102: Generate an iterative population according to the game hero attribute table and the equipment attribute table; the iterative population includes M individuals representing the game hero outfit recommendation scheme, and the game heroes represented by all the individuals are of the same category ; Each individual includes 6 sets of equipment that are different from each other, and the structure of each individual is an array of 6 rows and 5 columns; wherein, one row of the individual represents the serial number and attribute of an equipment, Each row of the individual includes 5 elements, which are the equipment number, the physical attack value of the equipment, the magic attack value of the equipment, the physical defense value of the equipment, and the magic defense value of the equipment.

Step 102 specifically includes:

According to the game hero attribute table and the equipment attribute table, an individual is generated by encoding, and then randomly initialized to obtain an iterative population including M individuals.

Each game hero in the game can purchase 6 pieces of equipment. Since the active or passive skills of the equipment are unique, it is wasteful for the same equipment to appear multiple times. Therefore, in this embodiment, a symbol code with a length of 6 is used. There are already 53 equipment numbers in Table 1, and an effective code is a combination of non-repeated numbers from 1-53, such as 2-17-18-44-36 -45 (recommended outfits are Boots of Resistance, Shadow Battleaxe, Armor Breaker, Grip of Ice Scars, Anti-Injury Stab Armor, and Sage's Protection) is an effective code, 2-2-2-2-2-2 (The outfit recommendation is Boots of Resistance, Boots of Resistance, Boots of Resistance, Boots of Resistance, Boots of Resistance, Boots of Resistance) is an invalid code. The structural form of an individual adopts the form of a 6*5 two-dimensional array, which is X _m represents the mth individual; among them, the first column of X _m represents the equipment number, and the remaining columns represent the physical attack, magic attack, physical defense, and magic defense value in Table 1 in turn, that is, each row of X _m is the equipment for details.

Then use the above-mentioned individual encoding method to randomly initialize and generate an iterative population with a population size of M.

Step 103: Calculate the fitness value of each individual in the iterative population according to the constructed multi-objective fitness function, and record the current number of iterations; the multi-objective fitness function is used to calculate the attack value and defense of the game hero sum of values.

The game heroes in the game can choose different equipment according to their own and friendly lineups. There are two considerations, one is the attack value of the game hero, and the other is the defense value of the game hero. The invention maximizes the attack value and defense value of game heroes through a multi-objective optimization algorithm.

Step 104: Judging whether the current number of iterations is less than the preset total number of iterations, and obtaining a first judgment result; if the first judgment result indicates that the current number of iterations is not less than the preset total number of iterations, then perform step 105; if If the first judgment result indicates that the current number of iterations is less than the preset total number of iterations, step 106 is executed.

Step 105: Determine the individual with the largest fitness value as the recommended outfit for the game hero.

Step 106: Add 1 to the current number of iterations, then perform a cross mutation operation on the individuals in the iterative population corresponding to the current number of iterations to generate new individuals, and calculate each of the new individuals according to the multi-objective fitness function The fitness value of ; the total number of new individuals is M.

Step 107: Add all the new individuals to the iterative population corresponding to the current iteration number to update the iterative population, and perform an individual in the updated iterative population according to the fitness value of each individual in the updated iterative population Sorting, select the first M individuals as the iterative population of the next generation, and then return to step 103.

Step 107 specifically includes:

All the new individuals are added to the iterative population corresponding to the current iteration number to obtain an updated iterative population; the number of individuals in the updated iterative population is 2M.

According to the fitness value of each individual in the updated iterative population, the Pareto dominance algorithm is used to select the individual in the updated iterative population.

Sort the individuals that have been selected by dominance, select the top M individuals as the iterative population of the next generation, and then return to step 103 .

In order to clarify the characteristics of the outfits recommended by this embodiment, this embodiment, on the basis of combining the characteristics of game hero attributes and equipment attributes, reprocesses the iterative population of M individuals after evolution, so as to provide players with more Outfit recommendations for game heroes.

The game hero outfit recommendation method also includes:

An iteration population corresponding to the current number of iterations not less than the preset total number of iterations is retained.

Calculate the first target value and the second target value of each individual in the retained iterative population; the first target value is the attack value of the game hero, and the second target value is the defense value of the game hero. The calculation formula of the first target value is The calculation formula of the second target value is X _m represents the mth individual, i represents the number of rows of the m-th individual; w ₁ , w ₂ , w ₃ , and w ₄ represent the physical attack value, magic attack value, physical defense value, and magic defense value of the game hero represented by the m-th individual, respectively.

Calculate a third target value of each individual in the retained iterative population; the third target value is a ratio of the first target value to the second target value. Among them, the larger the ratio, the higher the attack attribute, and the smaller the ratio, the better the defense attribute.

Based on the third target value of each individual as the sorting standard, the individuals in the retained iterative population are arranged in descending order, and players can freely choose the outfits and equipment of game heroes according to the individuals arranged in descending order.

The multi-objective fitness function in step 103 is F(X _m )=f ₁ (X _m )+f ₂ (X _m ).

Embodiment two

As shown in Figure 2, this embodiment provides a game hero outfit recommendation system based on multi-objective optimization, including:

The information acquisition module 201 is used to obtain the game hero attribute table and the equipment attribute table; the game hero attribute table and the equipment attribute table are both in the form of tables; the first column of the game hero attribute table is the game hero category, so The second column of the game hero attribute table is the physical attack value of the game hero, the third column of the game hero attribute table is the magic attack value of the game hero, and the fourth column of the game hero attribute table is the physical attack value of the game hero. Defense value, the fifth column of the game hero attribute table is the spell defense value of the game hero; the first column of the equipment attribute table is the equipment number, the second column of the equipment attribute table is the equipment name, and the equipment The third column of the attribute table is the physical attack value of the equipment, the fourth column of the equipment attribute table is the magic attack value of the equipment, the fifth column of the equipment attribute table is the physical defense value of the equipment, and the equipment attribute table The sixth column of is the spell defense value of the equipment.

The iterative population generation module 202 is used to generate an iterative population according to the game hero attribute table and the equipment attribute table; the iterative population includes M individuals who characterize the game hero outfit recommendation scheme, and the game represented by all the individuals The heroes are all of the same category; each individual includes 6 sets of different equipment, and the structure of each individual is an array of 6 rows and 5 columns; wherein, a row of the individual represents a piece of equipment Each line of the individual includes 5 elements, which are the equipment number, the physical attack value of the equipment, the magic attack value of the equipment, the physical defense value of the equipment, and the magic defense value of the equipment.

The fitness value calculation module 203 is used to calculate the fitness value of each individual in the iterative population according to the constructed multi-objective fitness function, and record the current number of iterations; the multi-objective fitness function is used to calculate the game The sum of the hero's attack value and defense value.

A judging module 204, configured to judge whether the current number of iterations is less than the preset total number of iterations, and obtain a first judgment result.

The game hero outfit recommendation scheme determination module 205 is used to determine the individual with the largest fitness value as the game hero outfit recommendation when the first judgment result indicates that the current iteration number is not less than the preset iteration total number Program.

Cross mutation operation module 206, used to add 1 to the current number of iterations when the first judgment result indicates that the current number of iterations is less than the total number of preset iterations, and then perform an iteration population corresponding to the current number of iterations The individuals in are subjected to a cross mutation operation to generate new individuals, and the fitness value of each new individual is calculated according to the multi-objective fitness function; the total number of the new individuals is M.

The iterative population update module 207 is used to add all the new individuals to the iterative population corresponding to the current number of iterations to update the iterative population, and to update the updated iterative population according to the fitness value of each individual in the updated iterative population The individuals in the population are sorted, the top M individuals are selected as the iterative population of the next generation, and then the fitness value calculation module 203 is returned.

In order to clarify the characteristics of the outfits recommended by this embodiment, this embodiment, on the basis of combining the characteristics of game hero attributes and equipment attributes, reprocesses the iterative population of M individuals after evolution, so as to provide players with more Therefore, the game hero outfit recommendation system provided in this embodiment also includes:

A retaining module, configured to retain the iteration population corresponding to the current number of iterations not less than the preset total number of iterations.

The first target value and the second target value calculation module are used to calculate the first target value and the second target value of each individual in the iterative population retained; the first target value is the attack value of the game hero, so The second target value is the defense value of the game hero.

A third target value calculation module, configured to calculate a third target value of each individual in the retained iterative population; the third target value is a ratio of the first target value to the second target value. The larger the ratio, the higher the attack attribute, and the smaller the ratio, the better the defense attribute.

The descending order module is used to arrange the individuals in the retained iterative population in descending order by using the third target value of each individual as a sorting criterion.

Wherein, the calculation formula of the first target value is The calculation formula of the second target value is X _m represents the mth individual, i represents the number of rows of the m-th individual; w ₁ , w ₂ , w ₃ , and w ₄ represent the physical attack value, magic attack value, physical defense value, and magic defense value of the game hero represented by the m-th individual, respectively.

The multi-objective fitness function in the fitness value calculation module 203 is F(X _m )=f ₁ (X _m )+f ₂ (X _m ).

The iterative population generation module 201 specifically includes:

The iterative population generation unit is used to generate an individual and randomly initialize it according to the game hero attribute table and the equipment attribute table in a coding manner, so as to obtain an iterative population including M individuals.

The iterative population update module 207 specifically includes:

The update unit is configured to add all the new individuals to the iterative population corresponding to the current iteration number to obtain an updated iterative population; the number of individuals in the updated iterative population is 2M.

The dominant selection unit is configured to perform dominant selection on the individuals in the updated iterative population by using the Pareto dominance algorithm according to the fitness value of each individual in the updated iterative population.

The individual selection unit is used to sort the individuals selected by dominance, select the top M individuals as the iterative population of the next generation, and jump to the fitness value calculation module 203 .

Embodiment three

As shown in Figure 3, a method for recommending outfits for heroes in the game "Glory of the King" based on multi-objective optimization provided by this embodiment includes the following steps:

Step 1, convert hero and equipment attributes; the conversion process is the same as step 101 in Embodiment 1, and will not be repeated here.

Step 2, randomly initialize the population; according to the fact that each game hero in the game "Glory of the King" can only purchase 6 pieces of equipment and the uniqueness of the active or passive skills of each equipment, the effective code of an individual should be a non-repeating length of 6 Symbol encoding, considering the equipment attributes, this embodiment adopts the encoding form in step 102 of the first embodiment, and randomly generates 100 symbol encodings with a length of 6 and a range of 1-53 to obtain an initial population. For example, in the initial population, individual 1 is 1-45-34-25-50-3, individual 2 is 3-43-51-40-4-37, and individual 99 is 49-34-45-6-39-5 , the individual 100 is 23-43-50-3-53-5, and the corresponding solution structure is shown in FIG. 4 .

Step 3, crossover and mutation to generate offspring; use genetic algorithm to perform crossover mutation operation on the individuals in the initial population to generate M new individuals.

In the initial population, the outfit recommendations represented by Individual 1, Individual 2, Individual 99, and Individual 100 are (1) Foot of Shadow Ninja - Sanctuary of the Sage - Book of the Sage - Wand of Void - Devour of Greed - Boots of Calmness, (2) Boots of Calmness - Extreme Cold Storm - Wings of Redemption - Boots of Mystery - Wrath of the Gorefiend, (99) Glory of the Guards - Book of Sages - Protection of Sages - Boots of Quick Walk - Overlord Heavy Equipment - Boots of Rapidity and (100) Blazing Dominator - Extreme Cold Storm - Bite of Greed - Boots of Calm - Emblem of the Wolf - Boots of Rapidity. If the currently selected game hero is set as a mage hero, then the attribute of the game hero is W=(0.1, 0.8, 0.05, 0.05). Bring the individual into the constructed multi-objective fitness function to calculate the fitness value (target value) of each individual. The calculated fitness values of individual 1, individual 2, individual 99 and individual 100 are (1.019, 0.207), (0.463, 0.262), (0.57, 0.26) and (0.324, 0.25), respectively. Next, randomly select two individuals in the initial population to crossover and mutate until 100 new individuals are generated, and calculate the fitness value of the new individual.

The crossover operation is an operation in which two individuals have data interleaving or exchanging positions at one or several positions. The mutation operation is to randomly select a row of data from an individual and replace it with some data in Table 1.

Assuming that individual 1 and individual 99 are selected as parent individuals, and the single-point crossover method is adopted, the crossover result is shown in Figure 5.

In this example, the individual has a certain probability of mutation under natural circumstances, and the rule is to randomly select a certain position and replace it with the data that did not appear in the original individual in Table 1; The individual must mutate, and its operation process is shown in Figure 6.

In crossover and mutation, both the crossover position and the mutation position are randomly generated, the crossover probability is 100%, and the mutation probability is 5%.

Step 4, merging populations and calculating the fitness value of individuals; adding new individuals to the initial population to obtain a merged population, the number of individuals in the merged population is 200.

Step 5, Pareto algorithm sorting, the top 100 individuals are used as the initial population of the next generation; after the cross-mutation operation, two excellent individuals have a great probability to produce two excellent new individuals. The purpose of the mutation operation is to make the individual jump out of the local optimum. The new individual may be worse than the parent individual, and the new individual is brought into the constructed multi-objective fitness function to calculate its fitness value and then added to the initial population. After using Pareto to dominate the selection and sorting, the non-excellent individuals will be discarded. Only the top 100 outstanding individuals are kept as the initial population of the next generation, and the initial population of the last generation is repeatedly iteratively output.

Step 6, judging whether the termination condition is satisfied; the termination condition is whether the current number of iterations is the total number of preset iterations. If yes, go to step 7, otherwise return to step 3.

Step 7, calculate the ratio and sort; in the final output initial population, calculate the first target value f ₁ and the second target value f ₂ of each individual, and then calculate the first target value f ₁ and the second target value f 2 for each individual The ratio of the two target values f ₂ , the larger the ratio, the higher the attack capability of the outfit recommendation scheme, and the smaller the ratio, the stronger the defense capability of the outfit recommendation scheme.

Step 8, end. Table 3 is part of the data of the outfit recommendation scheme for mage heroes optimized in this embodiment.

Table 3: Outfit recommendations for mage heroes

Based on the combination of the characteristics of game hero attributes and equipment attributes, the present invention can recommend more outfit options for players.

Each embodiment in this specification is described in a progressive manner, each embodiment focuses on the difference from other embodiments, and the same and similar parts of each embodiment can be referred to each other. As for the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and for relevant details, please refer to the description of the method part.

In this paper, specific examples have been used to illustrate the principle and implementation of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; meanwhile, for those of ordinary skill in the art, according to the present invention Thoughts, there will be changes in specific implementation methods and application ranges. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims (10)

1. A game hero outfit recommendation method based on multi-objective optimization, characterized in that, the game hero outfit recommendation method comprises: Obtain a game hero attribute table and an equipment attribute table; both the game hero attribute table and the equipment attribute table are in tabular form; the first column of the game hero attribute table is the game hero category, and the first row of the game hero attribute table is The second column is the physical attack value of the game hero, the third column of the game hero attribute table is the magic attack value of the game hero, the fourth column of the game hero attribute table is the physical defense value of the game hero, and the game hero attribute table is listed as the physical defense value of the game hero. The fifth column of the attribute table is the spell defense value of the game hero; the first column of the equipment attribute table is the equipment number, the second column of the equipment attribute table is the equipment name, and the third column of the equipment attribute table is The physical attack value of the equipment, the fourth column of the equipment attribute table is the magic attack value of the equipment, the fifth column of the equipment attribute table is the physical defense value of the equipment, and the sixth column of the equipment attribute table is the equipment's magic defense value; Generate an iterative population according to the game hero attribute table and the equipment attribute table; the iterative population includes M individuals who characterize the game hero outfit recommendation scheme, and the game heroes represented by all of the individuals are of the same category; each Each of the individuals includes 6 sets of different equipment, and the structure of each individual is in the form of an array of 6 rows and 5 columns; wherein, a row of the individual represents the number and attribute of an equipment, and the individual Each row of contains 5 elements, which are the equipment number, the physical attack value of the equipment, the magic attack value of the equipment, the physical defense value of the equipment, and the magic defense value of the equipment; Calculate the fitness value of each individual in the iterative population according to the constructed multi-objective fitness function, and record the current number of iterations; the multi-objective fitness function is used to calculate the sum of the attack value and defense value of the game hero ; Judging whether the current number of iterations is less than the total number of preset iterations to obtain a first judgment result; If the first judgment result indicates that the current number of iterations is not less than the total number of preset iterations, then the individual with the largest fitness value is determined as the outfit recommendation plan for the game hero; If the first judgment result indicates that the current number of iterations is less than the total number of preset iterations, then add 1 to the current number of iterations, and then perform a crossover mutation operation on the individuals in the iterative population corresponding to the current number of iterations to generate new individuals, and calculate the fitness value of each of the new individuals according to the multi-objective fitness function; the total number of the new individuals is M; Add all the new individuals to the iterative population corresponding to the current iteration number to update the iterative population, and sort the individuals in the updated iterative population according to the fitness value of each individual in the updated iterative population, select The first M individuals are used as the iterative population of the next generation, and then return to the step of calculating the fitness value of each individual in the iterative population according to the constructed multi-objective fitness function, and recording the current iteration number. 2. a kind of game hero outfit recommendation method based on multi-objective optimization according to claim 1, is characterized in that, described game hero outfit recommendation method also comprises: Reserving the iteration population corresponding to the current number of iterations not less than the preset total number of iterations; Calculating the first target value and the second target value of each individual in the iterative population retained; the first target value is the attack value of the game hero, and the second target value is the defense value of the game hero; Calculate the third target value of each individual in the retained iterative population; the third target value is the ratio of the first target value to the second target value; The individuals in the retained iterative population are sorted in descending order based on the third target value of each individual as the sorting standard. 3. a kind of game hero outfit recommendation method based on multi-objective optimization according to claim 2, characterized in that, the calculation formula of the first target value is The calculation formula of the second target value is X _m represents the mth individual, i represents the number of rows of the m-th individual; w ₁ , w ₂ , w ₃ , and w ₄ represent the physical attack value, magic attack value, physical defense value, and magic defense value of the game hero represented by the m-th individual, respectively. 4. A method for recommending outfits for game heroes based on multi-objective optimization according to claim 3, wherein the multi-objective fitness function is F(X _m )=f ₁ (X _m )+f ₂ (X _m ). 5. A method for recommending outfits for game heroes based on multi-objective optimization according to claim 1, wherein said generating an iterative population according to said game hero attribute table and said equipment attribute table specifically comprises: According to the game hero attribute table and the equipment attribute table, an individual is generated and initialized randomly by means of encoding, and an iterative population including M individuals is obtained. 6. A method for recommending outfits for game heroes based on multi-objective optimization according to claim 1, wherein all the new individuals are added to the iterative population corresponding to the current number of iterations to update the iterative population , and sort the individuals in the updated iterative population according to the fitness value of each individual in the updated iterative population, and select the first M individuals as the next generation iterative population, specifically including: All the new individuals are added to the iterative population corresponding to the current number of iterations to obtain an updated iterative population; the number of individuals in the updated iterative population is 2M; According to the fitness value of each individual in the updated iterative population, the Pareto dominance algorithm is used to select the individual in the updated iterative population; Sort the individuals that have been selected by dominance, and select the top M individuals as the iterative population of the next generation. 7. A game hero outfit recommendation system based on multi-objective optimization, characterized in that the game hero outfit recommendation system includes: An information acquisition module, configured to obtain a game hero attribute table and an equipment attribute table; both the game hero attribute table and the equipment attribute table are in the form of tables; the first column of the game hero attribute table is a game hero category, and the The second column of the game hero attribute table is the physical attack value of the game hero, the third column of the game hero attribute table is the game hero's magic attack value, and the fourth column of the game hero attribute table is the game hero's physical defense value, the fifth column of the game hero attribute table is the spell defense value of the game hero; the first column of the equipment attribute table is the equipment number, the second column of the equipment attribute table is the equipment name, and the equipment attribute The third column of the table is the physical attack value of the equipment, the fourth column of the equipment attribute table is the magic attack value of the equipment, the fifth column of the equipment attribute table is the physical defense value of the equipment, and the The sixth column is the spell defense value of the equipment; The iterative population generation module is used to generate an iterative population according to the game hero attribute table and the equipment attribute table; the iterative population includes M individuals who characterize the game hero outfit recommendation scheme, and the game heroes represented by all the individuals They are all of the same category; each of the individuals includes 6 sets of equipment that are different from each other, and the structure of each of the individuals is in the form of an array of 6 rows and 5 columns; wherein, a row of the individual represents a piece of equipment Number and attributes, each row of the individual includes 5 elements, which are respectively the equipment number, the physical attack value of the equipment, the magic attack value of the equipment, the physical defense value of the equipment, and the magic defense value of the equipment; The fitness value calculation module is used to calculate the fitness value of each individual in the iterative population according to the constructed multi-objective fitness function, and record the current number of iterations; the multi-objective fitness function is used to calculate the game hero The sum of attack value and defense value; A judging module, configured to judge whether the current number of iterations is less than the total number of preset iterations to obtain a first judgment result; The game hero outfit recommendation scheme determination module is used to determine the individual with the largest fitness value as the game hero outfit recommendation scheme when the first judgment result indicates that the current iteration number is not less than the preset iteration total number ; A cross mutation operation module, configured to add 1 to the current number of iterations when the first judgment result indicates that the current number of iterations is less than the total number of preset iterations, and then add 1 to the iteration population corresponding to the current number of iterations The individual of the individual carries out the cross mutation operation to generate new individual, and calculates the fitness value of each described new individual according to described multi-objective fitness function; The total number of described new individual is M; The iterative population update module is used to add all the new individuals to the iterative population corresponding to the current number of iterations to update the iterative population, and update the updated iterative population according to the fitness value of each individual in the updated iterative population The individuals in are sorted, and the top M individuals are selected as the iterative population of the next generation, and then returned to the fitness value calculation module. 8. A kind of game hero outfit recommendation system based on multi-objective optimization according to claim 7, characterized in that, the game hero outfit recommendation system further comprises: A retaining module, configured to retain the iterative population corresponding to the current number of iterations that is not less than the preset total number of iterations; The first target value and the second target value calculation module are used to calculate the first target value and the second target value of each individual in the iterative population retained; the first target value is the attack value of the game hero, so The second target value is the defense value of the game hero; The third target value calculation module is used to calculate the third target value of each individual in the retained iterative population; the third target value is the ratio of the first target value to the second target value; The descending order module is used to arrange the individuals in the retained iterative population in descending order by using the third target value of each individual as a sorting criterion. 9. The game hero outfit recommendation system based on multi-objective optimization according to claim 7, wherein the iterative population generation module specifically includes: The iterative population generation unit is used to generate an individual and randomly initialize it according to the game hero attribute table and the equipment attribute table in a coding manner, so as to obtain an iterative population including M individuals. 10. The game hero outfit recommendation system based on multi-objective optimization according to claim 7, wherein the iterative population update module specifically includes: An update unit, configured to add all the new individuals to the iterative population corresponding to the current number of iterations to obtain an updated iterative population; the number of individuals in the updated iterative population is 2M; The dominant selection unit is used to perform dominant selection on the individuals in the updated iterative population by using the Pareto dominance algorithm according to the fitness value of each individual in the updated iterative population; The individual selection unit is used to sort the individuals selected by the dominant, select the first M individuals as the iterative population of the next generation, and jump to the fitness value calculation module.