Configuration method and device for game AI models of different styles, equipment and medium

Abstract

The invention provides a configuration method and device for game AI models of different styles, equipment and a medium, and relates to the technical field of software algorithms. The configuration method comprises the steps: determining a plurality of different optimization objectives, wherein the different optimization objectives correspond to different styles of the game AI models; determiningparameter values of a plurality of groups of optimization objectives of the n generation according to a multi-objective evolutionary algorithm; training the game AI models according to the parameter values of the plurality of groups of optimization objectives to determine multiple groups of corresponding optimization objective values; when at least one objective optimization objective value meeting the optimization threshold exists in the plurality of groups of optimization objective values, retaining a trained game AI model corresponding to at least one target optimization target value. By means of the configuration method and device for the game AI models of different styles, the process of manual parameter adjustment can be omitted, the output efficiency of the game AI models of different styles can be improved, and the diversity of the styles of the game AI models can be achieved.

Description

technical field [0001] The present application relates to the technical field of software algorithms, in particular to a configuration method, device, device and medium for AI models of different styles of games. Background technique [0002] There are a large number of diverse objects in online games, and non-player characters (Non-player Characters, NPCs) with different styles can greatly enhance the freshness of players, thereby improving game experience and entertainment. [0003] In the prior art, the game field is generally based on reinforcement learning (Reinforcement Learning, RL), combined with manual parameter adjustment to adjust the weight ratio of various incentive signal rewards to guide the training of NPCs, thereby training NPCs of various styles. [0004] However, in the existing methods for training NPCs, due to the need for manual parameter adjustment, there are problems of limited optimization capabilities and low output efficiency. Contents of the inv...

AI Technical Summary

Problems solved by technology

[0005] The purpose of this application is to provide a configuration method, device, equipment, and medium for AI models of different styles of games in view of the deficiencies in the above-mentioned prior art, which can solve the problem of limited optimization ability and low output of manual parameter adjustment in the prior art. The problem of inefficiency

Images