本文選題：非玩家角色　切入點(diǎn)：自適應(yīng)　出處：《山東師范大學(xué)》2015年碩士論文　論文類型：學(xué)位論文

【摘要】：非玩家角色是指游戲中為了游戲進(jìn)行而設(shè)置的由程序控制的角色，是虛擬游戲世界的重要組成部分，一直貫穿游戲情節(jié)的始終。所以，非玩家角色一直以來(lái)都是玩家關(guān)注的焦點(diǎn)，其設(shè)計(jì)的好壞在很大程度上影響著游戲的成功與否。為了呈現(xiàn)完整的游戲交互敘事，喚起玩家對(duì)于非玩家角色的認(rèn)知和情感態(tài)度，設(shè)計(jì)者經(jīng)常利用非玩家角色行為機(jī)制的控制來(lái)實(shí)現(xiàn)游戲角色之間的信息交互。因此，非玩家角色的行為研究受到了游戲設(shè)計(jì)者的高度重視。 本研究通過(guò)相關(guān)文獻(xiàn)的檢索和分析得知，目前數(shù)字化游戲研究中非玩家角色的行為大部分是基于隨機(jī)決策或者傳統(tǒng)的預(yù)定義行為決策決定的，致使隨著玩家體驗(yàn)游戲次數(shù)的增多，就可能清晰地認(rèn)識(shí)到游戲中非玩家角色的大致行為模式，這樣極大地降低了玩家的參與性，使玩家喪失繼續(xù)體驗(yàn)游戲的興趣。因此，本研究結(jié)合相關(guān)的理論研究，提出了研究的重點(diǎn)，即能否使非玩家角色擁有人類的思維模式，使其通過(guò)感知虛擬游戲環(huán)境的變化與自身狀態(tài)的詳細(xì)信息，制定行為決策和規(guī)劃，指導(dǎo)自身的自適應(yīng)行為，進(jìn)而增加游戲的真實(shí)體驗(yàn)，牢牢地吸引玩家的注意力。 有鑒于此，本研究采用文獻(xiàn)分析法、案例分析法、系統(tǒng)開發(fā)研究法等研究方法，利用現(xiàn)有的游戲開發(fā)工具，設(shè)計(jì)與開發(fā)一款擁有非玩家角色自適應(yīng)行為的數(shù)字化游戲。其主要研究工作有以下四個(gè)方面。 （1）通過(guò)查閱與分析國(guó)內(nèi)外的相關(guān)研究文獻(xiàn)，對(duì)本研究中所涉及的概念以及理論進(jìn)行詳細(xì)的研究與闡述，為后續(xù)理論模型的構(gòu)建和游戲?qū)嵗拈_發(fā)奠定理論基礎(chǔ)。 （2）參照國(guó)內(nèi)外相關(guān)游戲案例的設(shè)計(jì)與分析，總結(jié)理論模型的構(gòu)建思路和原則，并提出該理論模型的構(gòu)建要素，以及各要素之間的相互聯(lián)系，最終構(gòu)建出具有非玩家角色自適應(yīng)行為的游戲整體理論模型，進(jìn)行非玩家角色自適應(yīng)行為系統(tǒng)結(jié)構(gòu)的細(xì)化設(shè)計(jì)。 （3）結(jié)合具體的學(xué)科知識(shí)體系，進(jìn)行詳細(xì)的教學(xué)設(shè)計(jì)。在此基礎(chǔ)上，利用“強(qiáng)化學(xué)習(xí)”技術(shù)實(shí)現(xiàn)非玩家角色自適應(yīng)行為的控制，制定詳細(xì)的各功能模塊實(shí)現(xiàn)思路。其中包括各角色狀態(tài)空間和行為空間的表示、不同狀態(tài)——?jiǎng)幼鲗?duì)值的修正，以及如何依據(jù)這些值選擇合理的行為等。 （4）鑒于以上提出的解決思路，利用相關(guān)游戲軟件開發(fā)平臺(tái)，開發(fā)設(shè)計(jì)出具體的游戲案例；并根據(jù)游戲運(yùn)行效果，做出強(qiáng)化學(xué)習(xí)過(guò)之后的非玩家角色與非強(qiáng)化學(xué)習(xí)和傳統(tǒng)強(qiáng)化學(xué)習(xí)訓(xùn)練的非玩家角色智能效果的相關(guān)實(shí)驗(yàn)驗(yàn)證。 隨著上述各項(xiàng)研究工作的逐步展開，本研究取得了以下三點(diǎn)研究成果。 （1）本研究構(gòu)建出具有非玩家角色自適應(yīng)行為的游戲整體理論模型，并結(jié)合初中《思想品德》學(xué)科進(jìn)行詳細(xì)的教學(xué)設(shè)計(jì)。在此基礎(chǔ)上分析了如何實(shí)現(xiàn)數(shù)字化游戲中非玩家角色的自適應(yīng)行為，其中包括：在程序中各角色狀態(tài)空間和動(dòng)作空間的表示、獎(jiǎng)懲反饋信號(hào)的制定、不同狀態(tài)—?jiǎng)幼鲗?duì)值的修正、根據(jù)這些值如何選擇適合的行為以及這些行為應(yīng)如何執(zhí)行等。 （2）本研究利用Eclipse軟件開發(fā)平臺(tái)實(shí)現(xiàn)了感知子系統(tǒng)、行為選擇子系統(tǒng)以及動(dòng)作執(zhí)行子系統(tǒng)的設(shè)計(jì)，最終開發(fā)出一款名為《警察快跑》的數(shù)字化模擬游戲，且在數(shù)字化游戲網(wǎng)站這一載體上呈現(xiàn)出來(lái)。 （3）結(jié)合非玩家角色自適應(yīng)行為的游戲?qū)嵗\(yùn)行效果，做出實(shí)驗(yàn)研究結(jié)果的分析，比較強(qiáng)化學(xué)習(xí)過(guò)之后的非玩家角色與非強(qiáng)化學(xué)習(xí)和傳統(tǒng)強(qiáng)化學(xué)習(xí)訓(xùn)練的非玩家角色之間的智能效果差異。 由于研究者的時(shí)間及技術(shù)開發(fā)能力有限，，數(shù)字化游戲中非玩家角色自適應(yīng)行為的設(shè)計(jì)雖能夠體現(xiàn)最初的研究意圖和思想，但是仍然存在許多問(wèn)題和不足，未來(lái)的研究工作將圍繞如何能夠更好的設(shè)計(jì)開發(fā)出非玩家角色的自適應(yīng)行為。
[Abstract]:Non game player role is that the game for the game set by the program control, is an important part of the virtual game world, has been through the plot of the game all the time. So, the non game player role has long been the focus of attention of the game player, its design quality has a great influence on the success of the game in order to present a complete interactive narrative game, game player game player role for non arousing cognitive and affective attitude control, designers often use non game player role behavior mechanism to realize the information interaction between the role of the game. Therefore, study on the behavior of non game player attaches great importance to the role of the game designer.
In this study, through literature search and analysis, the non digital game of game player role behavior is mostly based on stochastic decision decision or decision of the traditional deterministic behavior, resulting in increasing the number of game player game experience, can clearly recognize the general behavior pattern non game player characters in the game, thus greatly reduce the participation of the game player, the game player to lose interest in the game experience. Therefore, this study combined with relevant theoretical research, put forward the focus of the study, namely whether the non game player role owner class mode of thinking, the detailed information by sensing changes in virtual game environment and its status, making decision and behavior planning, guidance of adaptive behavior itself, and then increase the game experience, firmly attract game player's attention.
In view of this, this study adopts the method of literature analysis, case analysis, research methods of system development. On the other hand, the use of existing game development tools, design and development of a non adaptive behavior of digital game game player role. The main research work in the following four aspects.
(1) through consulting and analyzing related literatures at home and abroad, the concepts and theories involved in this study are studied and elaborated in detail, which will lay a theoretical foundation for the construction of subsequent theoretical models and the development of game instances.
(2) according to the design and analysis of the domestic and foreign related game case, construction ideas and principles summed up the theoretical model, and proposes the construction elements of the theoretical model, and relations among various elements, finally constructed with non adaptive behavior game game player role of the whole model, detailed design of non adaptive behavior of game player role system structure.
(3) combined with the knowledge system the specific teaching design in detail. On this basis, the use of "non adaptive behavior control game player role reinforcement learning technology, develop the function module detailed implementation method. It includes the role of the state space and behavior space, different values of state action on the basis of the value of correction, and how to choose reasonable behavior.
(4) given the solutions mentioned above, using the related game software development platform, development and design a specific game case; and according to the game running effect, to strengthen the relevant experimental study after the non game player role and non reinforcement learning and traditional reinforcement learning training non intelligent character verification. The effect of game player
With the gradual expansion of the above research work, the following three research achievements have been obtained in this study.
(1) this research constructs with non adaptive behavior game game player role of the whole model, and combined with the junior middle school ideological and moral teaching subject < > Design in detail. Based on the analysis of how to realize the adaptive behavior in non digital game game player role, including: said in the program of the role of the state space and action the reward feedback signal formulation, different state action value of the correction, based on these values to select the appropriate behavior and how to perform these behaviors.
(2) using Eclipse software development platform, we realized the design of perception subsystem, behavior selection subsystem and action execution subsystem. Finally, we developed a digital simulation game called "police run up", and presented it on the carrier of digital game website.
(3) combined with the effect of the game instances of the non player role adaptive behavior, we made the analysis of the experimental results, and compared the difference of the intelligence between the non player role and the non player role after reinforcement learning.
Due to the limited time and technology development, the design of non adaptive behavior in the digital game game player role can reflect the original intention and thought, but there are still many problems and shortcomings, future research will focus on how to design and develop the adaptive behavior of non game player role better.

【學(xué)位授予單位】：山東師范大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：G899

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 徐美娜;;“最近發(fā)展區(qū)”理論及對(duì)教育的影響與啟示[J];教育與教學(xué)研究;2010年05期

2 杜友福;李晨暉;;游戲引擎中智能角色的研究[J];長(zhǎng)江大學(xué)學(xué)報(bào)(自然科學(xué)版)理工卷;2008年04期

3 馬穎峰;隋志華;;教育網(wǎng)絡(luò)游戲中NPC與學(xué)習(xí)者置入感的關(guān)系研究——由Games-to-Teach項(xiàng)目想到的[J];電化教育研究;2009年06期

4 馬穎峰;隋志華;;基于Flow理論的教育游戲沉浸性設(shè)計(jì)策略研究——教育游戲活動(dòng)難度動(dòng)態(tài)調(diào)控研究[J];電化教育研究;2010年03期

5 賈曉峰;余雪麗;劉燕平;;Quake2中NPC智能結(jié)構(gòu)分析及行為改進(jìn)[J];電腦開發(fā)與應(yīng)用;2009年05期

6 戴曉林;陳澤琳;吳靜純;;動(dòng)畫自動(dòng)生成系統(tǒng)中的智能虛擬角色的研究[J];電腦與信息技術(shù);2009年04期

7 陳功;周誼成;王輝;;基于強(qiáng)化學(xué)習(xí)TD算法的乒乓游戲擊球策略優(yōu)化[J];電腦知識(shí)與技術(shù);2011年28期

8 閆振中;;沉浸理論研究現(xiàn)狀及其對(duì)遠(yuǎn)程教育的啟示[J];廣州廣播電視大學(xué)學(xué)報(bào);2010年01期

9 姜濤;;基于動(dòng)態(tài)腳本自適應(yīng)規(guī)則庫(kù)的智能角色戰(zhàn)術(shù)生成[J];福建電腦;2012年08期

10 畢靜;;一種適合網(wǎng)絡(luò)游戲的多NPC協(xié)同運(yùn)動(dòng)策略[J];計(jì)算機(jī)工程;2011年07期

本文編號(hào)：1573352