Optimization Design And Implementation Of RTS Game Balance And Intelligent NPC

In recent years,real-time strategic mobile games are popular,such games have a strong sense of rhythm and have wonderful battle scenes,so the factors of fairness and performance in the game are important.A good real-time strategy game should not only make the investment return ratio of character value similar,but also keep the level of players similar.At the same time,the action decision-making and search of characters in the game should also be as efficient as possible.But now many games are lack of detailed polishing in this respect.Through requirement analysis,this thesis designs an optimization scheme for the balance of the game and the intelligent character of the game.The main work is summarized as follows:(1)In view of the imbalance in the game scene and the low efficiency of the action of the game characters,this thesis analyzes the requirements of the game system,including functional requirements and non functional requirements,and designs the specific optimization scheme for the existing problems.(2)For the imbalance problem in real-time strategic games,this thesis optimizes from two aspects: game character value and player matching.In the aspect of numerical balance,the attribute value model is proposed,which can make the arms with attributes be measured by unified attributes;in the aspect of player matching balance,the fluctuation of player's level is considered on the basis of ELO integral,and the matching algorithm based on Bayesian probability model is proposed,which can make the matching system more compatible with player's ability fluctuation,so as to balance more truly Measure players' game level to improve matching fairness.(3)For the problem of low efficiency of game character execution,according to its environment,dynamic search,A* iteration,convex polygon optimization and behavior tree premise node are proposed to improve the operation efficiency.In addition,for AI players assigned by the system,this thesis proposes a deep reinforcement learning method to improve the game ability of AI players,so as to approach the level of real players.Finally,the requirements proposed in the requirements analysis are tested.The test results show that the optimization scheme proposed in this thesis is effective,and the optimization of the game has been published in the enterprise online.