xjtu-rl / msn-flocking-formation-control

:flying_saucer: An implementation of multi-agent flocking formation control with specific formations that can follow a target without collision and can avoid obstacles.

MSN Flocking Formation Control

This is an implementation of the MSN Flocking Formation Control presented in the paper Flocking for Multi-Agent Dynamic Systems: Algorithms and Theory.

Getting started

$ git clone
$ cd 
$ pip install

Run the python files for each of the five cases explained below.

$ python src/msn_1.py

Project parameters:

• Number of sensor nodes: n =100.
• Space dimensions: m = 2.
• Desired distance among sensor node: d = 15.
• Scaling factor: k = 1.2 and interaction range r = k*d.
• Epsilon = 0.1 and Delta_t = 0.009 (These two parameters are optional and you can change them).

Variable names

For easier understanding the variables are named similar to the equations described below.

For example, is denoted as C1_ALPHA (all capital as it's a constant), is denoted as p_i_k, and so on.

Cases

Case 1 - MSN Fragmentation

Filename: msn_1.py

• Randomly generates a connected network of 100 nodes in the area of 50x50.
• Plots the initial deployment of the MSN of 100 nodes.
• Links the neighboringing nodes together by a line.

Plots

• Plots the fragmentation of the nodes.
• Plots the velocity.
• Plots the connectivity.
• Plots the trajectory.

Algorithm 1:

Result

Case 2 - Implements MSN Quasi-Lattice Formation with static target

Filename: msn_2.py

• Randomly generates a connected network of 100 nodes in the area of 50x50.
• Sets up a target (gamma agent) as static point with its coordinate (x = 150, y =150).
• Implements flocking behavior of the MSN.

Algorithm 2 for Case 2:

Plots

• Plots the flocking of the nodes.
• Plots the velocity.
• Plots the connectivity.
• Plots the trajectory.

Result

Case 3 - Implements MSN Quasi-Lattice Formation with dynamic target (Sine wave trajectory)

Filename: msn_3.py

• Randomly generates a connected network of 100 nodes in the area of 150x150.
• Sets up a target (gamma agent) moving in a sine wave trajectory.
• Implements flocking behavior of the MSN.

Algorithm 2 for Case 3:

Plots

• Plots the flocking of the nodes following the target (Sine wave trajectory).
• Plots the velocity.
• Plots the connectivity.
• Plots the trajectory.
• Plots the center of mass and target trajectory.

Result

Case 4 - Implements MSN Quasi-Lattice Formation with dynamic target (Circular trajectory)

Filename: msn_4.py

• Randomly generates a connected network of 100 nodes in the area of 150x150. In this case you
• Sets up a target (gamma agent) moving in a circular trajectory.
• Implements flocking behavior of the MSN.

Algorithm 2 for Case 3 (Same equation, change the values to make it a circular trajectory):

Plots

• Plots the flocking of the nodes following the target (Circular trajectory).
• Plots the velocity.
• Plots the connectivity.
• Plots the trajectory.
• Plots the center of mass and target trajectory.

Result

Case 5 - Implements MSN Quasi-Lattice Formation with obstacle avoidance

Filename: msn_5.py

• Randomly generates a connected network of 100 nodes in the area of 50x50.
• Sets up a target (gamma agent) at the location of (200, 25).
• Sets up an obstacle, circular in shape with a radius of 15 and its center location at (100,25).
• Implements flocking behavior of the MSN avoiding obstacles.

Algorithm 3:

Plots

• Plots the flocking of the nodes following the target, avoiding obstacles.
• Plots the velocity.
• Plots the connectivity.
• Plots the trajectory.
• Plots the center of mass and target trajectory.

Result