Limits...
Advanced Fuzzy Potential Field Method for Mobile Robot Obstacle Avoidance.

Park JW, Kwak HJ, Kang YC, Kim DW - Comput Intell Neurosci (2016)

Bottom Line: An advanced fuzzy potential field method for mobile robot obstacle avoidance is proposed.The potential field method primarily deals with the repulsive forces surrounding obstacles, while fuzzy control logic focuses on fuzzy rules that handle linguistic variables and describe the knowledge of experts.This study also examines the rule-explosion problem of conventional fuzzy logic and assesses the performance of our proposed AFPFM through simulations carried out using a mobile robot.

View Article: PubMed Central - PubMed

Affiliation: Department of Electronic Engineering, Incheon National University, Incheon 402-752, Republic of Korea.

ABSTRACT
An advanced fuzzy potential field method for mobile robot obstacle avoidance is proposed. The potential field method primarily deals with the repulsive forces surrounding obstacles, while fuzzy control logic focuses on fuzzy rules that handle linguistic variables and describe the knowledge of experts. The design of a fuzzy controller--advanced fuzzy potential field method (AFPFM)--that models and enhances the conventional potential field method is proposed and discussed. This study also examines the rule-explosion problem of conventional fuzzy logic and assesses the performance of our proposed AFPFM through simulations carried out using a mobile robot.

No MeSH data available.


The simulation for passing between closely placed obstacles. Left side: using the conventional PFM; right side: using the AFPFM.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4829793&req=5

fig12: The simulation for passing between closely placed obstacles. Left side: using the conventional PFM; right side: using the AFPFM.

Mentions: Figure 12 shows the trajectories of the mobile robot when it attempted to pass between two closely placed obstacles: O1 and O2. The left- and right-hand sides of Figure 12 show the trajectories of the mobile robot with the conventional PFM and the AFPFM, respectively. The two obstacles were placed at variable positions and orientations while keeping a regular interval (i.e., 1.0 m). Compared to the obstacles shown in Figure 12(a), the topmost obstacles shown in Figures 12(b) and 12(c) were placed 0.3 m and 0.6 m lower, respectively, and those shown in Figures 12(d) and 12(e) were oriented at angles of 25° and 50°, respectively. As illustrated on the left-hand side of Figure 12, the robot that used the conventional PFM was not able to easily and efficiently pass between the two closely placed obstacles. In fact, the trajectories of the robots using the conventional PFM and AFPFM were only similar when the robots approached both obstacles at almost equidistance. In other cases, the robot using the conventional PFM had a sharply curved trajectory. In contrast, the robot using the AFPFM not only passed through the opening between the closely placed obstacles but also had a smooth and short trajectory toward the goal. This minimized the time taken to reach the goal, and the robot did not collide with surrounding obstacles.


Advanced Fuzzy Potential Field Method for Mobile Robot Obstacle Avoidance.

Park JW, Kwak HJ, Kang YC, Kim DW - Comput Intell Neurosci (2016)

The simulation for passing between closely placed obstacles. Left side: using the conventional PFM; right side: using the AFPFM.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4829793&req=5

fig12: The simulation for passing between closely placed obstacles. Left side: using the conventional PFM; right side: using the AFPFM.
Mentions: Figure 12 shows the trajectories of the mobile robot when it attempted to pass between two closely placed obstacles: O1 and O2. The left- and right-hand sides of Figure 12 show the trajectories of the mobile robot with the conventional PFM and the AFPFM, respectively. The two obstacles were placed at variable positions and orientations while keeping a regular interval (i.e., 1.0 m). Compared to the obstacles shown in Figure 12(a), the topmost obstacles shown in Figures 12(b) and 12(c) were placed 0.3 m and 0.6 m lower, respectively, and those shown in Figures 12(d) and 12(e) were oriented at angles of 25° and 50°, respectively. As illustrated on the left-hand side of Figure 12, the robot that used the conventional PFM was not able to easily and efficiently pass between the two closely placed obstacles. In fact, the trajectories of the robots using the conventional PFM and AFPFM were only similar when the robots approached both obstacles at almost equidistance. In other cases, the robot using the conventional PFM had a sharply curved trajectory. In contrast, the robot using the AFPFM not only passed through the opening between the closely placed obstacles but also had a smooth and short trajectory toward the goal. This minimized the time taken to reach the goal, and the robot did not collide with surrounding obstacles.

Bottom Line: An advanced fuzzy potential field method for mobile robot obstacle avoidance is proposed.The potential field method primarily deals with the repulsive forces surrounding obstacles, while fuzzy control logic focuses on fuzzy rules that handle linguistic variables and describe the knowledge of experts.This study also examines the rule-explosion problem of conventional fuzzy logic and assesses the performance of our proposed AFPFM through simulations carried out using a mobile robot.

View Article: PubMed Central - PubMed

Affiliation: Department of Electronic Engineering, Incheon National University, Incheon 402-752, Republic of Korea.

ABSTRACT
An advanced fuzzy potential field method for mobile robot obstacle avoidance is proposed. The potential field method primarily deals with the repulsive forces surrounding obstacles, while fuzzy control logic focuses on fuzzy rules that handle linguistic variables and describe the knowledge of experts. The design of a fuzzy controller--advanced fuzzy potential field method (AFPFM)--that models and enhances the conventional potential field method is proposed and discussed. This study also examines the rule-explosion problem of conventional fuzzy logic and assesses the performance of our proposed AFPFM through simulations carried out using a mobile robot.

No MeSH data available.