虚拟环境下农业移动机器人行为及其仿真建模

由于农业移动机器人的作业环境与对象的复杂性和不确定性,使机器人在作业时不能有效地获取导航知识以便对其行为进行决策。利用人工智能和虚拟现实技术,研究了移动机器人的复杂行为,构建虚拟环境和基于规则和案例的知识库,对机器人复杂行为分类,把复杂行为分解为简单、特定和已知的行为。对于不确定的和没有前驱知识的行为,利用相似案例和不确定推理为行为进行导航。然后,建立了多Agent行为决策系统,定义了该结构的六元函数,建立了消息和事件驱动机制,实现了虚拟环境下移动式机器人对水果的采摘行为及其仿真建模。

Modeling for Behavior and Simulation of Agriculture Mobile Robot in Virtual Environment

The agricultural mobile robots are unable to gain navigation knowledge effectively to make decision for their behaviors during the performance due to complexity and uncertainty of operating environment and object. Complex behaviors of the mobile robots were studied based on artificial intelligence and virtual reality technology. The virtual environment, rule-base and case-base knowledge were built, and the complex behavior of the mobile robots was classified into simple, specific and foregone behaviors. The robot was navigated using similar cases and uncertain reasoning for the behaviors which is uncertain and has no priori knowledge. Then a multi-agent behavior-decision system was built and sextuple function of the structure was defined. News and events-driven mechanism were set up to achieve robots behavior of fruit picking and simulation modeling in virtual environment.