Game role behavior control method and device, storage medium and electronic equipment

Abstract

The invention provides a game role behavior control method and device, a computer storage medium and electronic equipment, and relates to the field of artificial intelligence. The method comprises thesteps: obtaining global current game state information of a game, and performing feature extraction on the current game state information to obtain first feature information; constructing second feature information corresponding to a target game role according to the first feature information and target state information corresponding to the target game role in the current game state information;performing classification processing on the second feature information to obtain action probability distribution corresponding to the target game role; and determining target behavior information corresponding to the target game role based on the action probability distribution and the target state information. According to the invention, the intelligent calling of the game role can be generatedbased on the reinforcement learning network model, the intelligent fighting level of the NPC is improved, the NPC is matched with the level of a human player, the playability of the game is enhanced,and the user experience is improved.

Description

technical field [0001] The present disclosure relates to the technical field of artificial intelligence, in particular, to a game character behavior control method, a game character behavior control device, a computer-readable storage medium, and electronic equipment. Background technique [0002] With the emergence and development of online games, there are more and more types of online games, and multi-role RPG (role-playing game) games are one of them. In all role-playing games, players have to play a fictional character and control Many activities for this character. In a multi-role RPG game, a turn-based method is usually adopted. After a turn-based game starts, multiple combat characters (including player characters and non-player characters in multiple combat characters) participating in this round of turn-based Take turns in sequence, and only when it is your turn can you perform operations. [0003] When the combat role includes a non-player character (Non-player ...

AI Technical Summary

Problems solved by technology

The currently used methods include the decision tree method and supervised learning method, but the moves generated by the decision tree method are relatively fixed and single, and it is difficult to match the level of human players, so it cannot give players enough experience. When it comes to constructing behavior trees for each character, it takes a lot of time; supervised learning methods are usually impossible because of the difficulty of obtaining real game data, or are limited to the existing data itself, and are prone to overfitting or cannot guarantee generalization character, it is difficult to make game characters have a high level

Images