KR100913030B1 - Method and system for game strategy using cascade-fuzzy losic - Google Patents

Abstract

Disclosed are a game strategy method and a system using stepped fuzzy logic.

Game strategy system using the step-by-step fuzzy logic is a fuzzy logic unit for providing a fuzzy logic for fuzzy the strategy information according to the map (map) and agent (agent) information; And a strategic location generator that determines a strategic location of a user agent based on the strategic information by the fuzzy logic.

Online games, competitive games, AI characters, agents, phased fuzzy, fuzzy logic, fuzzification, fuzzy inference rules, defuzzification, strategic location

Description

Game strategy method and system using stepped fuzzy logic {METHOD AND SYSTEM FOR GAME STRATEGY USING CASCADE-FUZZY LOSIC}

The present invention relates to a game system, and more particularly, to a game strategy method using a step-by-step fuzzy logic for generating a high-level artificial intelligence character and progressing the game, and a system thereof.

Recently, the rapid development of the information and communication field not only dramatically improves the work efficiency, but also changes our daily life style. Due to the rapid spread of high speed networks, many people are using the Internet, and various online game services using a computer communication network are provided to Internet users.

In general, an online game is composed of a graphic image field for a background and a character used in the game, a sound field for a background music and a sound effect, a program field for calling the graphic image and sound for the progress of the game, and the like. The game production tool is intended to facilitate the user by effectively combining the above areas.

The game production tool includes a game modeling process, an analysis process of a modeled game, and a source code conversion process of a modeled game, and recently, artificial intelligence techniques have been attempted to be incorporated into game modeling methods.

Conventional artificial intelligence techniques are mainly used as a finite state machine (hereinafter referred to as FSM), artificial life, neural network, fuzzy logic (fuzzy logic).

FSM is a rule-based system in which finite states are connected in a graph controlled by transitions between states. FSM is widely used because it can be implemented with only if-else or switch-case statements. For example, if a character finds an enemy while on the move, it moves from the moving state to the tracking state and follows the found enemy. If the enemy is within a certain distance, it will change to attack and attack the enemy. Thus, using FSM has the advantage of easy implementation and precisely defined behavior, but has the disadvantage that the progress of the game operates only in a predefined manner. In other words, if the opponent of the game is implemented by the FSM, because the behavior of the opponent is constant, it can be easily predicted after playing the game for a certain time, which acts as a factor that halves the interest of the game.

Artificial life is the study of the general characteristics of life by reproducing phenomena represented by life on artificial media such as computers and robots. If biology has taken an analytical approach to biological phenomena, artificial life has a comprehensive approach. Applying the basic characteristics of artificial life, adaptability and emergence to the game, can increase the interest of the game by adapting to complex environment or user's operation or unexpected unexpected behavior. However, due to the unpredictable nature of artificial life, it has been used in a limited way and is still at the basic stage.

The neural network mimics the human nervous system, has a structure in which many components are connected, and produces an output based on pattern recognition of input data. The biggest advantage of the neural network applied to the game is that since the neural network has learning ability, the intelligence can be continuously improved as the game progresses. So many attempts have been made to apply neural networks to games. Most of them are board games such as concave and Tic-Tac-Toe. The reason is that the board game is easy to learn the expected pattern in advance because you have to learn in advance. However, in an action game such as a competitive game, if a user learns a situation pattern, it requires a great amount of time.

In the case of fuzzy logic, fuzzy functions are applied to the inputs and outputs so that they can be operated at random to some extent. Because it contains random elements, it is possible for opponents in the game to behave differently in the same situation, making it difficult to predict their behavior. In other words, it does not always produce the same pattern of behaviors, making it difficult for the other person to predict and making the situation judgment feel like a person.

Most of the existing AI techniques are for games that recognize and cope with the situation of the entire game, such as board games, and the game designer has to design and code every situation in advance. In addition, even a game made in this way has no ability to recognize and adapt to the environment by itself, so if the situation or the rules of the game change, it must be recoded.

In addition, if the character is given intelligence using an existing artificial intelligence technique, there is a problem in that a program must be newly created to improve or update the intelligence.

Recently, there have been attempts to impart intelligence to games by using artificial life or neural networks, but these attempts are mainly based on artificial life and are still at the basic stage of application to the game field.

In addition, the competitive game mainly uses the FSM to impart intelligence to the character, so that when the user competes with the character, the user may easily grasp the pattern of the character and feel tired of the game.

On the other hand, the present invention provides a game strategy method and system using stepped fuzzy logic for generating a character performing a game at a high level in a competitive game.

The present invention provides a game strategy method and a system that makes it easier to improve or update a character's intelligence.

The present invention provides a game strategy method and system using step-by-step fuzzy logic that maximizes the user's interest in the game and enables developers to easily implement game AI.

The present invention includes a fuzzy logic unit for providing a fuzzy logic for fuzzy strategy information according to map information and agent information; And a strategy position generating unit determining a strategic position of a user agent based on the strategy information by the fuzzy logic.

The present invention includes the steps of defining the strategy information according to the map information and the agent information; And determining the strategic position of the user agent by the stepped fuzzy logic based on the defined strategy information.

According to the present invention, there is provided a game strategy module that allows a user agent to preempt an optimal strategy position than an opponent in a competitive game between AI characters to maintain a game.

The game strategy method and system using the step-by-step fuzzy logic according to the present invention are variously applied to the fuzzy values for the rule elements to implement the AI character of various levels, and to predict by adjusting the inference rules according to the behavior of the character. AI characters moving on an impossible basis can be implemented.

Therefore, the present invention has an effect that the AI character performing a high level game can continuously maintain and maximize the user's interest in the game.

In addition, the present invention can easily program an artificial intelligence character performing a high level game by using a stepped fuzzy logic can improve the convenience and efficiency of development.

Hereinafter, a game strategy system and a game strategy method using stepped fuzzy logic according to the present invention will be described with reference to the accompanying drawings.

Referring to Figure 1 will be described a game strategy system using a stepped fuzzy logic of the present invention. 1 illustrates a configuration of a game strategy module using stepped fuzzy logic.

The game system for providing an online game via the Internet includes a game interface 110 that is a game driving device that connects to a server device (not shown) that provides a game and drives a game desired by a user.

The game system according to the present invention is a strategy for obtaining a strategic position value for each cell of a map (combat area) by using stepped fuzzy logic to provide an optimal strategy position for an agent in a competitive game between AI characters. Provide module 130.

In a match game, you need to find the optimal strategic position for the agent, where the strategic position must first avoid the attack of the opposing agent and satisfy the positional conditions that can be won when engaging the opponent. In consideration of these conditions, the strategy module of the present invention is an element for calculating a strategic position (hereinafter referred to as 'strategic information') as an obstacle-select, attack-first, and safety degree. Use safety.

In order to derive the strategy information, various rule factors for map information and agent information are used. The map information includes cell location information and topographic information for each cell (high-terrain or low-). Terrain, etc.), and the agent information includes energy, attack power, defense, range, field of view, moving speed, attack delay, current location, number of surviving allies, number of surviving enemies, Mokpo agent and You can use the distance, the number of nearby allies / enemies based on the target agent, the number of nearby allies / enemies based on the current location, and the number of allies / enemies within the field of view.

The strategy module 130 may include a memory 131 for storing the rule element for the competitive game driven through the game interface 110, and the strategy information based on the rule element (relationship with cover). And a fuzzy logic unit 135 that performs fuzzy logic about the attackable degree and the degree of safety, and calculates a strategic position value from the fuzzy value of each strategy information and calculates an optimal strategic position of the user agent. And a strategy location generator 137 for delivery to 110.

The apparatus further includes a rule selecting unit 133 for selectively inputting rule elements necessary for deriving each strategy information among the rule elements stored in the memory 131 to the fuzzy logic unit 135. The rule selector 133 is a process of allowing a result value to be selected according to the behavioral style when determining the inference rule in the process of purging the relationship with the cover, the attackable degree, and the safety degree according to the behavioral behavior of the agent. Do this.

The fuzzy logic unit 135 includes an obfuscation fuzzy logic for performing a purge of a relationship with a cover, an attack degree fuzzy logic for performing fuzzy for an attackable degree, and a safety degree for performing fuzzy for safety. It consists of fuzzy logic.

The fuzzy logic unit 135 diffuses fuzzy inference logic for performing an inference rule process based on the fuzzy strategy information, and a fuzzy value defined in the entire set of each strategy information by the inference rule process. Include fuzzy logic to fuzzy).

According to the above-described configuration, the fuzzy logic unit 135 includes a fuzzification for changing a value of an input variable measured by one crisp value into an appropriate fuzzy value, and a conditional sentence (if−) of a conditional part and a conclusion part. then fuzzy inference rules, and defuzzification that transforms the fuzzy values defined in the entire set according to the fuzzy inference rules into clear fuzzy values.

The cover fuzzy logic, the attack degree fuzzy logic, and the safety fuzzy logic perform fuzzy according to each fuzzy logic using a rule element input from the rule selecting unit 133 as an input value. At this time, the fuzzy inference rule is passed, but it is preferable to use a rule element selected by the rule selector 133 as a rule value and to use a min-max method as the rule value. The fuzzy value derived through the fuzzy inference rule is defusized. The non-fuge is possible by the maximum value method, the maximum average method, the center of gravity method (preferably the center of gravity method).

The strategic position generating unit 137 inputs a fuzzy value of the relationship with the cover, the attackable degree, and the safety degree derived from the fuzzy logic unit 135, and the strategic position for fuzzy strategic positions for each cell of the map. It consists of fuzzy logic. Similarly, the strategic position generation unit 137 includes fuzzy inference logic for performing an inference rule process based on the input value, and non-fuzzy logic for dispersing a fuzzy value for the strategic position.

Each fuzzy logic and the strategic position fuzzy logic for the strategy information will be described in detail as follows.

First, the cover purge logic will be described with reference to FIGS. 2 to 4.

The cover fuzzy logic is a fuzzy module that finds cover to avoid the attack of the opponent agent. In order to find the optimal cover in the cover fuzzy logic, the number of enemy groups around the cover and the distance between the cover is an input value, and a fuzzy value for the relationship with the cover is derived from the two rule elements. FIG. 2 shows the membership function for the number of enemies around the cover, and FIG. 3 shows the membership function for the distance to the cover.

The cover number fuzzy logic has a total of nine rule sets because the number of enemy units around the cover and the distance between the cover have three fuzzy values and the min-max method is used as the inference rule.

For example, assuming that rule values are equally applied in the rule-selection, the cover fuzzy logic is shown in FIG. 4.

(1) If <distance from cover is Close> and <number of enemies around cover is Normal>, <Obstacle-Select is Good>.

(2) If <distance from cover is Medium> and <number of enemies around cover is Little>, <Obstacle-Select is Normal>.

(3) If the distance to cover is Far and the number of enemies around cover is many, Obstacle-Select is No Good.

Rule can be defined. The fuzzy value of the relationship with the cover can be fuzzy by the center of gravity method to obtain a desirable value of the relationship with the cover.

5 to 8, the attack fuzzy logic will be described.

The attack degree fuzzy logic checks whether the attack can be performed at the current position. The attack value fuzzy logic is based on the current attack-delay, the terrain range of the terrain, and the relationship between the target agent and the range. From these three rule elements, the fuzzy value of the attackable degree is derived. FIG. 5 shows the membership function for the current attack delay state, FIG. 6 shows the membership function for the terrain coverage relationship, and FIG. 7 shows the membership function for the relationship between the target agent and the range.

The current attack delay state, the relationship between the target agent and the range has a fuzzy value of three each, the terrain range of the relationship has two fuzzy values and the attack degree fuzzy because the min-max method is used as the inference rule Logic has a total of 18 rule sets.

For example, the attack purge logic is as shown in Figure 8,

(1) If <Attack-Delay is Now>, <Relationship between target agent and range is Close>, and <Sight of terrain is not visible>, <Attack-First is attacking together>.

(2) If <Attack-Delay is Wait>, <Relationship between target agent and medium range is Medium>, and <Target field of view relationship is shown>, <Attack-First is attack first>.

(3) If <Attack-Delay is Long>, <Relationship between target agent and Far is Far>, and <Target field of view relationship is visible>, <Attack-First cannot attack>.

Rule can be defined. The fuzzy value for the attackable degree may be defusified by the center of gravity method to obtain a desirable value of the attackable degree.

The safety fuzzy logic will be described with reference to FIGS. 9 through 12.

The safety fuzzy logic determines the degree of danger of the location selected by the user. The safety fuzzy logic is based on the number of enemy soldiers in the field of view, the number of allies around him, and the energy state. To derive FIG. 9 shows the membership function for the number of enemies in the field of view, FIG. 10 shows the membership function for the number of allies around him, and FIG. 11 shows the membership function for the energy state.

The safety fuzzy logic has a total of 27 rule sets because the number of enemies in the field of view, the number of allies around it, and the energy state have three fuzzy values and the min-max method is used as the inference rule.

For example, the safety fuzzy logic is as shown in FIG.

(1) If the number of enemies in the field of sight is Little, the number of allies around you is many, and the energy status is Good, the safety is High.

(2) If the number of enemies in the field of view is Normal, the number of allies around you is Normal, and the energy state is Normal, then Safety is Normal.

(3) If <Number of enemy soldiers in the field of view>, <Number of allies around you>, and <Energy Status is No Good>, <Safety is Low>.

Rule can be defined. The fuzzy value for the safety can be defusified by the center of gravity method to derive the desirable value for the safety.

Finally, the strategic position fuzzy logic will be described with reference to FIG.

The strategic position fuzzy logic is based on the fuzzy values of the strategic information and the fuzzy values of the strategic information and the relations with the cover, the attackable degree, and the safety level. The strategic position values for each cell can be calculated. Figure 13 shows the membership function for strategic location.

Since the relationship with the cover, the attackable degree, and the safety degree each have three fuzzy values, the strategic position fuzzy logic has a total of 27 rule sets.

For example, the strategic position fuzzy logic is shown in FIG. 13,

(1) If <Obstacle-Select is Good>, <Attack-First cannot attack> and <Safety is High>, <Strategic Position is Good>.

(2) If <Obstacle-Select is Normal>, <Attack-First is Attacking Together> and <Safety is Normal>, <Strategic Position is No Good>.

(3) If <Obstacle-Select is No Good>, <Attack-First is First Attack> and <Safety is High>, <Strategic Position is Normal>.

Rule can be defined. The fuzzy values for these strategic positions can be defusified by the center of gravity method to obtain desirable values, which can be used as strategic position values for each cell.

Accordingly, the strategic position generation unit 137 derives a fuzzy value for the relationship with the cover, the attackable degree, and the safety level based on the position of each cell, and calculates the strategic position value of the corresponding cell based on the fuzzy value. .

The strategy position generator 137 derives strategic position values of each cell with respect to all of the cells forming the entire map or the predetermined competitive zone, and finally provides the game interface 110 with position information of the cell having the highest position value. To pass). At this time, the game interface 110 determines the cell received from the strategy position generation unit 137 as the optimal strategic position for the agent and controls the game driving based on this.

The game strategy method in the strategy module by the stepped fuzzy logic is as follows. 14 is a diagram illustrating a game strategy process using stepped fuzzy logic according to the present invention.

As described above, in the present invention, as the strategic information for determining the strategic position of the agent, the relationship with the cover, the attackable degree, and the degree of safety are used.

Rule elements required for deriving the strategic information are defined, and in order to derive a relationship with the cover, using a membership function of the distance to the cover and the number of enemies around the cover, and deriving the attackable degree. In order to derive the safety, the number of enemies in the field of view, the number of allies around them, and the energy state are used (S210). ).

In the competitive game in which the competition between agents is performed, fuzzy logic on the strategy information is performed on all cells that form the entire map or the predetermined competitive area based on the cell where the user agent is located (S220).

The fuzzy logic performs the fuzzy logic step by step in order of rule elements, strategy information, and strategy position.

In other words, the distance between the cover and the number of enemies around the cover is input to purge the relationship with the cover, and the current attack delay status, the relationship between the target agent and the range, and the terrain range of the terrain are input. It purges the attackable degree and performs the fuzzy fuzzy safety by inputting the number of enemy soldiers in the field of view, the number of allies around them, and the energy state.

Then, the fuzzy value for the strategic position is performed by inputting the fuzzy values for the relationship with the cover, the attackable degree, and the safety level. At this time, after the inference rule process on the relationship with the cover, the attackable degree, and the safety degree derived by the fuzzy logic, defuzzy is performed by the center of gravity method.

Through the above process, the strategic position value for each cell, that is, the relationship with the cover for the cell, the attackable degree, and the respective non-fuzzy values for the safety are calculated by adding up (S230).

When the calculation of the electric power position value is completed for all the cells of the entire map or the competitive zone, the cell corresponding to the highest position value is finally determined as the strategic position of the user agent (S240) (S250).

Accordingly, the present invention provides a strategy module that allows a user agent to preempt an optimal strategy position in a competitive game between AI characters. It is possible to generate various types of characters by differently applying the fuzzy value without being limited to the above-described rule elements or strategy information, and it is also possible to generate characters that perform a game based on completely different criteria by adjusting the inference rule. In addition, by using a strategy module using stepped fuzzy logic, it is possible to easily and conveniently develop a character performing a high level game.

The game strategy method using the step-by-step fuzzy logic according to the present invention is implemented in the form of program instructions that can be executed by various computer means can be recorded on a computer readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. Program instructions recorded on the media may be those specially designed and constructed for the purposes of the present invention, or they may be of the kind well-known and available to those having skill in the computer software arts. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tape, optical media such as CD-ROMs, DVDs, and magnetic disks, such as floppy disks. Magneto-optical media, and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like. The hardware device described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

As described above, the present invention has been described by way of limited embodiments and drawings, but the present invention is not limited to the above embodiments, and those skilled in the art to which the present invention pertains various modifications and variations from such descriptions. This is possible.

Therefore, the scope of the present invention should not be limited to the described embodiments, but should be determined not only by the claims below but also by the equivalents of the claims.

1 is a diagram showing the configuration of a game strategy system using stepped fuzzy logic according to the present invention.

FIG. 2 is a diagram showing the fuzzy values of the rule elements corresponding to the number of enemies around the cover to derive a relationship with the cover.

3 is a diagram showing a fuzzy value of a rule element corresponding to a distance from a cover to derive a relationship with the cover.

FIG. 4 is a diagram for explaining fuzzy logic of a relationship between a cover and a cover derived from the rule elements of FIGS. 2 and 3.

5 is a diagram illustrating a fuzzy value of a rule element corresponding to a current attack delay state for deriving an attackable degree.

FIG. 6 is a diagram illustrating a fuzzy value of a rule element corresponding to a view range relationship of a terrain for deriving an attackable degree.

FIG. 7 is a diagram illustrating fuzzy values of rule elements corresponding to a relationship between a target agent and a range for deriving an attackable degree.

FIG. 8 is a diagram for describing fuzzy logic for an attackable degree derived from the rule elements of FIGS. 5 to 7.

FIG. 9 is a diagram showing fuzzy values of rule elements corresponding to the number of enemy soldiers in the field of view for deriving safety. FIG.

10 is a diagram showing the purge value of the rule element corresponding to the number of allies around to derive the degree of safety.

FIG. 11 is a diagram showing a fuzzy value of a rule element corresponding to an energy state for deriving safety. FIG.

FIG. 12 is a diagram for describing fuzzy logic for a safety degree derived from the rule elements of FIGS. 9 to 11.

FIG. 13 is a diagram for explaining the fuzzy logic for the strategic position derived by the fuzzy logic of the relationship with the cover, the attackable degree, and the safety level.

14 is a diagram illustrating a game strategy process using stepped fuzzy logic according to the present invention.

<Explanation of symbols for the main parts of the drawings>

130: strategy module

131: memory

133: rule selector

135: fuzzy logic section

137: strategic position generator

Claims (13)

Map information corresponding to a rule factor of a competitive zone defined in a competitive game and agent information corresponding to a rule element for an agent, a game character defined in the competitive game, are input values. By using the fuzzy logic to the user agent, a user-specified game character, the relationship with the cover that can avoid the attack of the target agent, the target of the game, the degree of attack that the user agent can attack the target agent, the user A fuzzy logic unit for calculating a degree of safety indicating a degree of danger to the location of the agent; And, And a strategic location generating unit for determining a strategic location which is a location of the user agent to play against the target agent using the calculated relationship with the cover, the attackable degree, and the degree of safety. The fuzzy logic unit, A cover fuzzy logic that calculates a relationship with the cover using a fuzzy logic having at least one of a distance between the cover and the number of enemy units around the cover in the map information and the agent information; An attack degree fuzzy logic that calculates the attackable degree using fuzzy logic having at least one of a current attack delay state, a relationship between a target agent and a range range, and a field of view relationship in the map information and the agent information; A stepped fuzzy logic for calculating the safety degree using fuzzy logic having at least one of an enemy number within a field of view, an allied number within a field of view, and an energy state as input values in the map information and agent information. Game strategy system using logic. delete delete The method of claim 1, A rule selector for inputting the rule element necessary to calculate a relationship with the cover, an attackable degree, and a safety degree to the fuzzy logic unit; Game strategy system using stepped fuzzy logic further comprising. delete delete The method of claim 1, The fuzzy logic unit, Calculating the relationship with the cover, the attackable degree, and the safety level for all the cells constituting the charged area, The strategic position generating unit, After calculating the strategic position value for the case where the user agent is located in the cell by using the fuzzy logic including the relationship with the cover, the attackable degree, and the safety level calculated for each cell, the calculated position A game strategy system using stepped fuzzy logic, characterized in that the cell having the largest position value among the values is determined as the strategic position of the user agent. In the game strategy method of the game strategy system including a fuzzy logic unit and a strategy position generation unit, In the fuzzy logic unit, map information corresponding to a rule element of a competitive zone defined in the competitive game and agent information corresponding to a rule element for an agent which is a game character defined in the competitive game are inputted. By using fuzzy logic as a value, the user agent, which is a user-specified game character, has a relationship with cover to avoid the attack of the target agent, which is the competitive target, the degree of attack that the user agent can attack the target agent, Calculating a degree of safety indicating a degree of risk for the location of the user agent; And, Determining, by the strategic position generating unit, a strategic position which is a position of the user agent for playing against the target agent using the calculated relationship with the cover, an attackable degree, and a degree of safety; Computing the relationship with the cover, the attackable degree, the degree of safety, Calculating a relationship with the cover using the fuzzy logic having at least one of a distance between the cover and the number of enemies around the cover in the map information and the agent information; Calculating the attackable degree using fuzzy logic having at least one of a current attack delay state, a relationship between a target agent and a range range, and a viewing range relationship in the map information and the agent information; Calculating the degree of safety by using fuzzy logic having at least one of an enemy number within a viewing range, an ally number within a viewing range, and an energy state in the map information and the agent information. Game strategy method. delete delete The method of claim 8, Computing the relationship with the cover, the attackable degree, the degree of safety, Calculating the relationship with the cover, the attackable degree, and the safety level for all the cells constituting the charged area, Determining a strategic location of the user agent, Calculating a strategic position value for the case where the user agent is located in the cell by using fuzzy logic that uses the relation, covertability, and safety as an input value calculated for each cell; Determining a cell having the largest position value among the calculated strategic position values as the strategic position of the user agent. delete A computer-readable recording medium in which a program for executing the method of claim 8 or 11 is recorded.

Images