Jump to content
  • Advertisement
Sign in to follow this  
Zachesa

Smart Artificial Intelligence Game

This topic is 2988 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

On respawn player and his property, in case of death, seemed to be clear, everything depends on the concept of the game. Another thing when the question is about the break in communication. Each game is done (and defeating the screenplay) in its, it is simply just a matter of fact, so as always to strike a balance. The fact is that if a long break of connection the player loses too much more than simply lost in this moment of battle, the game will be too heavy and can not become popular by definition. At the same time lower the value of the life of the character and the lack of fines in the breakage of communication produces a specific image of the player's behavior. As noted the disappearance of a player from the battlefield can be attributed to magic in fantasy worlds, but totally illogical in the world close to a real or science fiction. Then the opportunity to slip away safely from the battlefield for many can become the basis of behavior, up to the fact that the player himself will chop off the Internet channel, to play again or not meddle where he snitched on the horns. In my project (UEF) put emphasis on indirect control character at the same time, at any point the player can take control your character entirely up to the management of a single hand to manipulate objects. The basis of this complex game AI characters, which manages and NPC and the player's character when he can not or will not control myself. As a result, the problem with a break in connection leveled so that the character begins to operate a computer and at least he will not stand there like a stone idol, or fall asleep in substitutes, or Avatar. And will do what is necessary in the current situation, in accordance with the settings given the nature of players with skills, and current challenges and genereriruemyh their action list. The same applies to associations of the characters, I think also to focus on the management of large characters associations managed game AI, with the help of the designer to create an association set out the basic goals and tasks, as assigned leaders to follow the global objectives, possibly changing the current challenges in the development of the situation. http://zp3.biz/universe-extend-frontier

Share this post


Link to post
Share on other sites
Advertisement
Quote:
Original post by alvaro
Errr... What are you talking about?

Suppose, I need to get a system could work in different applications, not only for management "mob" whose purpose is "to resist for decency and give the loot."
First what I read earlier, the ability to manage the player's character (autopilot) for breakage of communication.
This is further independent action of the same character when the player wants it and only gave orders to his character as a list of goals and priorities, so that the game AI is not only able to select actions to perform tasks, but also to change priorities according to changing situations.
Managing more than one character at a time, friends or servants (depending on concept), or both are taken together, until their own associations, which later will include other players.
The complex pattern of behavior of computer characters - the leaders of organizations (corporations, clans, guilds and so on) resulting from self-motivation and self-study to autogeneration plot.

Share this post


Link to post
Share on other sites
The structure of the server part of the project UEF

1. In a distributed database stores all data without exception, are constants, formulas, and simple algorithms (sub, function, procedure), as well as static and dynamic data reflecting the current state of the objects registered in the system, the standard definition of the essence of the virtual world UEF. Objects have a set of constant and variable properties, the possibility of applying the function to an object (or entity) depends on the particular set of these properties and their range states.
2. Septic system management event handler, run instances of event handlers in line with the needs of the system (occurrence of new events) cease their activity when idle or hang (infinite loop), and compares the versions of the event handlers on the stability and efficiency.
3. Event handlers do not contain functions and algorithms, and therefore have a small amount of code handling the event is due to sampling of suitable functions in the database along with other data and combining them together, depending on the task at hand. Employ a combination of functions and use the data cached in memory for faster access to permanently used data and functions.
4. The system will not contain large programs. Each program (system manager, the event handler) can be easily modified. Any function can be changed on the fly, may have variants (versions) for speed and accuracy. Data is available for any event handler that allows you to change the logic of the system elements on the fly without editing the code at all! Added a new algorithm (function) and a given event is handled by the new algorithm.
5. To improve the performance of the system to exert its utmost efforts to optimize code for execution of parallel computations. Main CPU power goes to the acceleration of output and input data in the database. To improve the performance and stability of the system a virtual world is divided into logically separate groups of objects (stellar systems, large groups of individuals, perhaps even some highly developed world) are serviced by its set of management system.
6. Every action on objects easily documented means DBMS that allows you to log on the basis of database and modify the values of viewing history. Creating new objects can easily controllable, there is no need for their destruction. With the physical destruction of an object instance, recording about it being destroyed, but simply writes that the object was destroyed when and for what reason.
7. Game AI characters, and control systems of vehicles, autonomous robots and industrial complexes built on the three-tier basis.

* Lower level - an action will work on the principle of finite automaton, for example moving from one point to another character (atonomny robot) stumbles upon a passive, no nothing notable obstacles, avoids boulders and trees down into a valley, up the hill. By analogy with the human elements of the selected automatically at the level of the spinal cord.
* Basic level - class. Analyzed the available objectives, form the list of tasks on the basis of these goals (perhaps the most algorithmically complex modulus) and task-based solutions are chosen representing a list of specific actions. Moreover, the steps set priorities, some actions can be performed simultaneously, while others are just consistently. To be implemented by a software mechanism for determining the possibility of pairing with one another on the experience of the characters and the analysis of physical models of objects.
* Higher level - the formation of targets based on global goals of life, the preferred methods to achieve them, the psychological features of character, his professional skills and knowledge in the light of fame and wealth, depending on the outcome of the set of unavoidable contact with other characters. In fact, the most important module (or set of software modules), since it was he who imitates the artificial life of the characters and the way entirely unnecessary atonomnym robots, intelligent systems, facilities and vehicles.

In carrying out activities and tasks using the usual algorithm that compares the current problem with the database problems and solutions on the basis of similarity, using the weights of solutions to the task, plus a random selection when there is a solution with similar coefficients, with it, the accuracy of selection decisions will depend on the particular nature of the character. For example, experienced a serious individual will choose the most important decisions, or produce a new, inexperienced and reckless chooses randomly from several suitable only superficially. If there is no meaningful solutions, and module design solutions are not available to a character is a random act, but the result is evaluated on performance, with the creation of new records in the database, if necessary, the analysis starts from the beginning. Special program module will adjust these rates based on the result of the task.

For each character gets on his kit modules. What has in the process in RAM, the server on each kit to create a cache queries and answers to the database, to expedite the work. A separate module will periodically check the relevance of the data cache, if necessary, clean the old data.

Share this post


Link to post
Share on other sites
Quote:
Original post by phresnel
TLDR.

Maybe you should ask an actual question, instead of just explaining the situation.


Question one, at the moment, looking for programmers who are familiar with Erlang, and other functional languages, and most importantly has experience in implementing a similar architecture or the desire to try to implement it.

Share this post


Link to post
Share on other sites
Quote:
Original post by VanKurt
So if your question is not AI related: Why not post it in the "help wanted" section?


I believe it is AI-related... I just can't tell what the problem is.

Share this post


Link to post
Share on other sites
http://zp3.biz/universe-extend-frontier

http://www.gamedev.net/community/forums/topic.asp?topic_id=568175

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

Participate in the game development conversation and more when you create an account on GameDev.net!

Sign me up!