Artificial Intelligence Implementation Essay

The design and development of Artificial Intelligence further improves the shortcomings of traditional AI in real-time strategy games such as DOTA by designing AI for all level levels of play from the player's perspective. The following is my carefully organized thesis on artificial intelligence implementation of the relevant information, I hope it is helpful to you!

Thesis on Artificial Intelligence Implementation Part I

Design and Implementation of Artificial Intelligence in DOTA Game

Introduction

DOTA game, as well as all the real-time strategy games, the artificial intelligence (AI) is a major indispensable tool. Novice players initially understand the rules of the whole game, the way of fighting, the characteristics of the heroes, etc. by playing against the AI. In non-networked situations, the AI is also the player's only opponent.

Traditional AI: Single combat mode, slow reaction speed, dull action commands, unable to better analyze the battlefield situation and combat situation. Low level, easy to be recognized by the player designed commands, which leads to easy kills, affecting the entertainment of the game.

Innovative AI: simulates human thinking with a high IQ. Novice players are able to gradually learn how to play the game and improve their knowledge of the game by playing against the AI, unlike the previous direct confrontation with humans leading to being ravaged by high-end players.

In addition, innovative AI is not only for novice players, based on the AI has a high level, a wide range of routines, know how to improvise and so on, and at the same time can make high-level players from the battle with the AI can also be fun, to further enhance the user experience.

1. Core functions of the AI algorithm

The core functions of the AI include the control center, inspection system, and command center. The inspection system is like the eyes and brain of the AI, the AI obtains the game data through the inspection system and analyzes the data; the results of the inspection system analysis are conveyed to the command center, which is processed and transformed into command signals and conveyed to the control center; the control center transforms the signals into the specific behaviors of the AI [1]. Specific functions are as follows:

2. AI Core Functions Implementation

2.1 Inspection System

Set a character's current life value is H, attack power is A, defense power is D, magic value is M, the four skills are A1, A2, A3, A4, and corresponding to the strength of the four skills respectively, P1, P2, P3, P4, the skill's cooldown degree is C1, C2, C3, C4.

If for any skill An knows that its remaining cooldown time is Yn, and the total cooldown time is Zn, then there must exist a linear function fn that makes the skill cooldown Cn:

Cn=fn(Yn, Zn), Cn?[0, 1]

Calculating skill cooldown is a very important part of the skill cooldown in the high-end game. The degree of mastery of skill cooling largely reflects the level of a player's game, when the degree of skill cooling is about 0 on behalf of the skill just entered the cooling time, for some heroes rely on the skills of the main representative of the loss of combat power; when the degree of skill cooling is about to reach 1, the hero is about to restore the combat power and in the moment it is equal to 1 to restore a large number of combat power. And this recovery process is often unexpected. In this recovery process that is the test of the player's technical content of the process, such as retreat, normal attacks, positioning or articulation of other strategic solutions, etc. [2].

Then for any skill An the skill strength Pn corresponds to the cooldown Cn and other parameters X as follows:

Pn=fc(Cn, X)

2.2 Self-learning Function

Self-learning of the AI allows the AI to continuously improve itself in the actual battle. In order to realize this piece of functionality, we change the traditional programming directly to the AI fixed set, so that the AI knows what kind of skills, attributes, and tell the AI various behaviors will produce the results, so that the AI itself calculates the most effective set in the current situation, so that the design allows the AI in the complexity of the actual game combat can have excellent performance [3].

2.3 Command Center

The Command Center is the functional module of the system that translates signals into actual actions.

For example, if the command center receives a ?ATTACK? command from the control center, then the command center will give an attack command to the hero, and feed back to the control center the result of this command, such as the hero is knocked out, then the command will not be able to be executed, and then it will be fed back to the control center with a signal of being knocked out, and the control center will immediately recalculate the countermeasures.

2.4 Control Center

Each subsystem is connected to a complete AI system through the control center, which accepts the data and analysis results of each subsystem, and then releases commands to the command center.

For example, in the actual battle, a high explosive mage with control skills, first he saw an enemy directly in front of him, through the Know Thyself and Know Your Enemy system, the AI learns that the target enemy's combat power is lower than his own, and can be killed. Then the AI starts to think about the killing strategy, through the self-learning system, the AI calculates the optimal plan: first, approach the target enemy through positioning, then try to subdue it with the control skill, during the control skill period the AI performs normal attacks on the target enemy, and when the control skill is about to end, the AI releases a big move to kill it. The results calculated by the Know Your Enemy and Self Learning System are communicated to the control center, which issues instructions to the command center, so the AI acts.

3. Conclusion

The design and development of Artificial Intelligence further improves the inadequacy of traditional AI in real-time strategy games such as DOTA, and from the player's point of view, the design of AI for all levels of hierarchy. The player's needs are the root of the game design, the player's experience is the goal of the game design, to ensure that the player's excellent gaming experience, in order to make the game have a greater market competitiveness.