本帖最后由 微笑的猫 于 2023-12-6 15:12 编辑
Behavior 系列索引
Behavior 0.2.3 下载链接
Behavior 2 Game Creator 2 by Catsoft Works 2.1.3 下载链接
This module requires Game Creator 2 in order to work.
Create your own behavioral AI systems using 4 industry-standard techniques.
State Machines are the easiest and most intuitive tools to use. At any given time an agent can be executing a single state, and has a list of transitions to other ones.
Behavior Trees offer a way to keep systems more organized. It comes with all the common types of nodes, such as Tasks and Sub Graphs.
Composites nodes allow to branch the logic into multiple paths, using Selectors and Sequence types, among others.
Decorator nodes allow to transform the result of executing its child nodes. For example, using the Inverse decorator turns a Failure result into a Success.
GOAP or Goal Oriented Action Planning is a novel technique in which you define multiple tasks of things to do, without any order and the behavior brain will figure out the best possible combination at any given time in order to fulfill a list of Goals.
Utility AI (also known as Needs-based AI) is an advanced behavioral AI technique that allows to define how much a Task node should be prioritized over others using a dynamic numeric value and an easing curve.
This is extremely useful to create games such as The Sims, where each character has different needs based on their personality traits and current context.
Includes
Source code
Examples & Templates
本模块需要 Game Creator 2 才能运行。
使用 4 种行业标准技术创建自己的行为人工智能系统。
状态机是最简单、最直观的工具。在任何给定时间内，代理都可以执行单个状态，并有一个过渡到其他状态的列表。
行为树提供了一种让系统更有条理的方法。它包含所有常见的节点类型，如任务（Tasks）和子图（Sub Graphs）。
复合节点可以使用选择器和序列类型等将逻辑分支为多条路径。
装饰节点可以转换执行其子节点的结果。例如，使用反向装饰器可将 "失败 "结果转化为 "成功"。
GOAP 或以目标为导向的行动规划是一种新颖的技术，在这种技术中，你可以定义多个任务，不分先后顺序，行为大脑会在任何特定时间找出最佳组合，以实现一系列目标。
实用性人工智能（又称基于需求的人工智能）是一种先进的行为人工智能技术，它允许使用动态数值和缓和曲线来定义任务节点相对于其他节点的优先级。
这对于制作《模拟人生》等游戏非常有用，在这些游戏中，每个角色都会根据自己的个性特征和当前环境有不同的需求。
包括
源代码
示例和模板
|
