Potential applications for autonomous navigation of mobile robots include automatic driving, guiding the blind and disabled, exploring dangerous regions, transporting objects in factories or offices, collecting geographic information in unknown territories, such as unmanned exploration of a new planetary surface, etc.
Navigation strategies of mobile robots can be broadly divided into two categories: global trajectory planning and reactive navigation. In the first approach, a collision-free trajectory is designed to guide the robot to a given target through a known environment. But, the real world environment is subject to change over time. Thus, it is expected that an autonomous robot may encounter uncertain environmental situations, and reactive navigation capabilities are required. Especially during the initial exploration of an unknown environment to create a preliminary map of the environment, which can later be used to optimize the path. An autonomous robot must react to the surrounding situation in its immediate vicinity in such a way as to achieve the goal without colliding with obstacles.
This article discusses the application of the Steering Behaviours method in simulating natural robot movement. In modern computer graphics and animation, the concept of Steering Behaviours refers to a set of algorithms and methods used to create controlled behaviour for virtual objects. These algorithms allow for modelling various types of behaviour, such as flocking, obstacle avoidance, object following, and more. These behaviours allow virtual objects to exhibit intelligent and realistic movement in virtual environments.
In an attempt to simulate the process of avoiding obstacles when moving, a combination of basic reflex actions and higher-level logical decisions is implemented. It is shown that for reflexive navigation of autonomous mobile robots, the ability to reflexively avoid obstacles on only one side (left or right) is sufficient to avoid obstacles on both sides. The use of such a behavior model provides the basis for a compact representation of reflex behavior.
The goal is to first explore all the individual behaviors, moving from the truly simple to the more complex, and eventually combine and apply them to control robot motion.
key words: robot motion, robot contol, steering behaviours method, intelligent control
