Persistent Monitoring

Multi-Agent Persistent Monitoring on Networks Simulator

᠎

Click here! to read the technical manual.

Developped by: Shirantha Welikala.

Supervised by: Prof. Christos G. Cassandras.

Problem Configuration:

Targets:

Selected target:

Uncertainty Rate:

Max. Path Length: Make fully connected:

Agents:

Selected agent:

Sensing Rate:

Run Simulations:

Time (t):

Cost (J):

t:

J_e:

J_s:

J_T:

Δt= Frame rate =

T = Multiplier =

K = Step Size =

Receding Horizon Control:

Control Method: (Add noise: )

Datapoints/Iterations:

α = ;(Override: )

β =

α_u = ;(Fix: )

v_max = u_max =

Normli.:

Discount:

Step Val.:

Type:

Ran. Ex.:

Real. Fr.:

Depth (Num. of Hops): Width (Storage Size):

Plan. Horizon H =

Act.v_max:

Act.u_max:

Parameter:

Start:

Resol.:

End:

Noise in A_i(t):

Max. % Devia.: %

Noise in V_ij:

Max. % Devia.: %

Noise in Y_i(t):

Boundary:

Magnitude:

Noise in R_i(t_event):

Mean Int.:

Magnitude:

Noise in R_j:

Magnitude:

Spectral Graph Clustering Parameters:

Similarity measure based on:

Neighborhood width (σ):

Spectral clustering method:

K-M. Steps:

Greedy TCP Construction:

1. Cluster Targets:

2. Generate Cycles: Allow Multiple Visits to a Target on the Cycle:

3. Auto-Adjust Cycles:

4. Compute Thresholds:

Theo.:

Boosting Menu:

Boosting method:

Step sizes:

Boosting coefficient α_B:

Switching threshold α_S:

Best cost found so far:

J_min:

Randomly perturb thresholds:

Noise Level:

Uncertainty Rates:

Sensing Rates:

Target Uncertainties R_i(t):

Target Prioritization Policy (Based on):

Minimum Distance:

Maximum Uncertainty:

Threshold generation (from cycles) parameter:

Blocking Threshold:

More detailed data plots:

Generate Additional Data plots:

Multi-Agent Persistent Monitoring on Networks Simulator

᠎

Click here! to read the technical manual.

Developped by: Shirantha Welikala.

Supervised by: Prof. Christos G. Cassandras.

Problem Configuration:

Targets:

Selected target:

Uncertainty Rate:

Max. Path Length: Make fully connected:

Agents:

Selected agent:

Sensing Rate:

Run Simulations:

Time (t):

Cost (J):

t:

J_e:

J_s:

J_T:

Receding Horizon Control:

Noise in A_i(t):

Noise in V_ij:

Noise in Y_i(t):

Noise in R_i(t_event):

Noise in R_j:

Spectral Graph Clustering Parameters:

Greedy TCP Construction:

Boosting Menu:

Best cost found so far:

J_min:

Randomly perturb thresholds:

Uncertainty Rates:

Sensing Rates:

Target Uncertainties R_i(t):

Target Prioritization Policy (Based on):

Threshold generation (from cycles) parameter:

More detailed data plots:

Agent threshold values θ_ij^(a):

Agent threshold sensitivity values: J w.r.t. θ_ij^(a)

Multi-Agent Persistent Monitoring on Networks Simulator

᠎

Click here! to read the technical manual.

Developped by: Shirantha Welikala.

Supervised by: Prof. Christos G. Cassandras.

Problem Configuration:

Targets:

Selected target:

Uncertainty Rate:

Max. Path Length: Apply Make fully connected:

Agents:

Selected agent:

Sensing Rate:

Run Simulations: Θ

Time (t):

Cost (J):

t:

Je:

Js:

JT:

Receding Horizon Control:

Noise in A_i(t):

Noise in V_ij:

Noise in Y_i(t):

Noise in R_i(t_event):

Noise in R_j:

Spectral Graph Clustering Parameters:

Greedy TCP Construction:

Boosting Menu:

Best cost found so far:

Jmin:

Randomly perturb thresholds:

Uncertainty Rates:

Sensing Rates:

Target Uncertainties Ri(t): Rand. 0 Update

Target Prioritization Policy (Based on):

Threshold generation (from cycles) parameter:

More detailed data plots:

Agent threshold values θij(a): Rand. Update

Agent threshold sensitivity values: J w.r.t. θij(a)

Max. Path Length: Make fully connected:

Run Simulations:

J_e:

J_s:

J_T:

J_min:

Target Uncertainties R_i(t):

Agent threshold values θ_ij^(a):

Agent threshold sensitivity values: J w.r.t. θ_ij^(a)