Game intelligent agent training method and device and equipment and storage medium

Abstract

The invention discloses a game agent training method. The method comprises the steps of receiving game scene information sent by a game server removing an NPC capable of generating a damage effect inthe game scene information, and modifying the game scene into a first path-finding following scene; controlling an intelligent agent to perform way-finding training according to a preset heuristic reward rule to obtain way-finding parameters; restoring an NPC capable of generating a damage effect in the game scene information, wherein the game scene is modified into an initial way-finding following scene; and controlling an intelligent agent to perform path-finding training according to the path-finding parameters until a preset training target is completed. The embodiment of the invention further discloses a game intelligent agent training device and equipment and a storage medium. By adopting the multiple embodiments, the problems that in the prior art, the reinforcement learning training efficiency is low, and the time period is long are effectively solved.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence, in particular to a training method, device, equipment and storage medium for a game agent. Background technique [0002] Traditional pathfinding and following algorithms include methods such as copying trajectories and re-planning routes. These methods can achieve good results in simple game scenes. However, as the environment of the game scene becomes more and more complex, these algorithms are required to respond quickly to changes in these complex environments, and these traditional pathfinding and following algorithms are increasingly unable to meet the corresponding requirements. [0003] With the explosion of deep learning technology, the reinforcement learning (Reinforcement Learning) technology based on deep learning has also developed rapidly. Deep reinforcement learning technology can use continuous trial and error and learning and imitation of human behavior to...

AI Technical Summary

Problems solved by technology

[0004] The embodiment of the present invention provides a training method, device, equipment and storage medium for a game agent, which can effectively solve the problems of low efficiency and long time period of reinforcement learning training in the prior art

Images