In this paper, a novel dynamic navigational planning strategy is proposed for single as well as multiple humanoids in intricate environments on a glowworm-based optimization method. The sensory information regarding the obstacle distances and target information are supplied as inputs to the navigational model. The essential turning angle is generated as the output of the controller to avoid obstacles present in the environment and reach the target location with ease. The proposed model is certified in a V-REP simulation software, and the simulation results are authenticated in a real-time setup arranged under testing conditions.

This paper proposes a new dynamic planning navigation strategy for use with mobile terrestrial robots. The strategy was applied to situations in which the environment and obstacles were unknown. After each displacement event, the robot replanned its route using a control algorithm that minimized the distance to the target and maximized the distance between the obstacles. Using a spatial localization sensor and a set of distance sensors, the proposed navigation strategy was able to dynamically plan optimum routes that were free of collisions. Simulations performed using different types of environment demonstrated that the technique offers a high degree of flexibility and robustness, and validated its potential use in real applications involving mobile terrestrial robots.

In this paper we consider the problem of planning paths for articulated bodies operating in workplaces containing obstacles and regions with preferences expressed as degrees of desirability. Degrees of desirability could specify danger zones and desire zones. A planned path should not collide with the obstacles and should maximize the degrees of desirability. Region desirability can also convey search-control strategies guiding the exploration of the search space. To handle desirability specifications, we introduce the notion of flexible probabilistic roadmap (flexible PRM) as an extension of the traditional PRM. Each edge in a flexible PRM is assigned a desirability degree. We show that flexible PRM planning can be achieved very efficiently with a simple sampling strategy of the configuration space defined as a trade-off between a traditional sampling oriented toward coverage of the configuration space and a heuristic optimization of the path desirability degree. For path planning problems in dynamic environments, where obstacles and region desirability can change in real time, we use dynamic and anytime search exploration strategies. The dynamic strategy allows the planner to replan efficiently by exploiting results from previous planning phases. The anytime strategy starts with a quickly computed path with a potentially low desirability degree which is then incrementally improved depending on the available planning time.