APF-DWA Hybrid Local Planner

A ROS-based local path planning solution that combines Artificial Potential Fields (APF) with Dynamic Window Approach (DWA) fallback for robust autonomous navigation.

Overview

This project implements a hybrid local planner that addresses the classic local minima problem in APF-based navigation. The system intelligently switches between APF and DWA algorithms to ensure reliable path execution in cluttered or narrow environments.

Key Features

• Hybrid Architecture: Seamlessly combines APF efficiency with DWA robustness
• Local Minima Detection: Automatic detection of stuck states using movement thresholds
• Intelligent Fallback: Dynamic switching to DWA when APF encounters local minima
• Real-time Visualization: Interactive force field visualization in RViz
• ROS Integration: Full compatibility with ROS navigation stack

Hybrid Strategy

1. Primary Navigation: Uses APF for efficient goal-seeking
2. Stuck Detection: Monitors robot movement over time
3. Fallback Activation: Switches to DWA when local minimum detected
4. Recovery: Returns to APF after successful escape

Visual Results

The system provides real-time visualization of:

• 🟢 Green Arrows: Attractive forces toward goal
• 🔴 Red Arrows: Repulsive forces from obstacles

Obstacle Avoidance Demonstration

Prerequisites

• ROS Noetic

sudo apt-get update

sudo apt-get install ros-noetic-navigation

sudo apt-get install ros-noetic-amcl

Running Instructions

To run a demo of the robot implementing the custom planner, use the following commands:

Step 1: Clone the Project

git clone https://github.com/abdullaxahmed/APF_DWA_fallback.git

cd APF_DWA_fallback

Step 2: Build the Workspace

catkin_make

Step 3: Sourcing the Workspace

source devel/setup.bash

Step 4: Launching the Project

roslaunch mybot all.launch

Step 5: Select the following marker topic in RViz in the marker array field

/move_base/APFLocalPlannerROS/vector_field

APF Parameters

Key parameters in src/mybot/params/apf_local_planner_params.yaml:

# APF Parameters
k_att: 1.0              # Attractive gain
k_rep: 100.0            # Repulsive gain
influence_dist: 1.0     # Obstacle influence radius
robot_radius: 0.2       # Robot collision radius

# Velocity Limits
max_linear_vel: 0.3     # Maximum linear velocity
max_angular_vel: 1.0    # Maximum angular velocity

# Fallback Parameters
use_dwa_fallback: true  # Enable DWA fallback
stuck_dist_thresh: 0.01 # Movement threshold for stuck detection
fallback_duration: 5.0  # Duration to use DWA after stuck detection