Optimal Trajectory for Network Establishment of Remote UAVs –1–1 Prachya Panyakeow, Ran Dai, and Mehran Mesbahi American Control Conference June 2013.

Slides:



Advertisements
Similar presentations
1/22 Worst and Best-Case Coverage in Sensor Networks Seapahn Meguerdichian, Farinaz Koushanfar, Miodrag Potkonjak, and Mani Srivastava IEEE TRANSACTIONS.
Advertisements

A. S. Morse Yale University University of Minnesota June 4, 2014 TexPoint fonts used in EMF. Read the TexPoint manual before you delete this box.: A A.
A Hierarchical Multiple Target Tracking Algorithm for Sensor Networks Songhwai Oh and Shankar Sastry EECS, Berkeley Nest Retreat, Jan
Principal Component Analysis Based on L1-Norm Maximization Nojun Kwak IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008.
22C:19 Discrete Math Graphs Fall 2014 Sukumar Ghosh.
Motion Planning for Point Robots CS 659 Kris Hauser.
Coverage by Directional Sensors Jing Ai and Alhussein A. Abouzeid Dept. of Electrical, Computer and Systems Engineering Rensselaer Polytechnic Institute.
What is Intractable? Some problems seem too hard to solve efficiently. Question 1: Does an efficient algorithm exist?  An O(a ) algorithm, where a > 1,
1 NP-Complete Problems. 2 We discuss some hard problems:  how hard? (computational complexity)  what makes them hard?  any solutions? Definitions 
Graphs Chapter 12. Chapter Objectives  To become familiar with graph terminology and the different types of graphs  To study a Graph ADT and different.
Management Science 461 Lecture 2b – Shortest Paths September 16, 2008.
Inexact SQP Methods for Equality Constrained Optimization Frank Edward Curtis Department of IE/MS, Northwestern University with Richard Byrd and Jorge.
Applications of Single and Multiple UAV for Patrol and Target Search. Pinsky Simyon. Supervisor: Dr. Mark Moulin.
1 Fast Primal-Dual Strategies for MRF Optimization (Fast PD) Robot Perception Lab Taha Hamedani Aug 2014.
Object Detection by Matching Longin Jan Latecki. Contour-based object detection Database shapes: …..
One-Shot Multi-Set Non-rigid Feature-Spatial Matching
CSE332: Data Abstractions Lecture 27: A Few Words on NP Dan Grossman Spring 2010.
Multiple Marine Vehile Deconflicted Path Planning with Currents and Communication Constraints A. Häusler 1, R. Ghabcheloo 2, A. Pascoal 1, A. Aguiar 1.
3 -1 Chapter 3 The Greedy Method 3 -2 The greedy method Suppose that a problem can be solved by a sequence of decisions. The greedy method has that each.
Network Coding Project presentation Communication Theory 16:332:545 Amith Vikram Atin Kumar Jasvinder Singh Vinoo Ganesan.
Graphs Chapter 12. Chapter 12: Graphs2 Chapter Objectives To become familiar with graph terminology and the different types of graphs To study a Graph.
1 Worst and Best-Case Coverage in Sensor Networks Seapahn Meguerdichian, Farinaz Koushanfar, Miodrag Potkonjak, Mani Srivastava IEEE TRANSACTIONS ON MOBILE.
Potential Fields for Maintaining Connectivity of Dynamic Graphs MEAM 620 Final Project Michael M. Zavlanos.
Spring 2010CS 2251 Graphs Chapter 10. Spring 2010CS 2252 Chapter Objectives To become familiar with graph terminology and the different types of graphs.
Computer Algorithms Integer Programming ECE 665 Professor Maciej Ciesielski By DFG.
© 2005, it - instituto de telecomunicações. Todos os direitos reservados. Gerhard Maierbacher Scalable Coding Solutions for Wireless Sensor Networks IT.
1 Efficient Placement and Dispatch of Sensors in a Wireless Sensor Network Prof. Yu-Chee Tseng Department of Computer Science National Chiao-Tung University.
Fall 2007CS 2251 Graphs Chapter 12. Fall 2007CS 2252 Chapter Objectives To become familiar with graph terminology and the different types of graphs To.
Two Discrete Optimization Problems Problem #2: The Minimum Cost Spanning Tree Problem.
Special Topics on Algorithmic Aspects of Wireless Networking Donghyun (David) Kim Department of Mathematics and Computer Science North Carolina Central.
Domain decomposition in parallel computing Ashok Srinivasan Florida State University COT 5410 – Spring 2004.
Dijkstra’s Algorithm and Heuristic Graph Search David Johnson.
© The McGraw-Hill Companies, Inc., Chapter 3 The Greedy Method.
The Coverage Problem in Wireless Ad Hoc Sensor Networks Supervisor: Prof. Sanjay Srivastava By, Rucha Kulkarni
CS774. Markov Random Field : Theory and Application Lecture 08 Kyomin Jung KAIST Sep
Structure Preserving Embedding Blake Shaw, Tony Jebara ICML 2009 (Best Student Paper nominee) Presented by Feng Chen.
Nattee Niparnan. Easy & Hard Problem What is “difficulty” of problem? Difficult for computer scientist to derive algorithm for the problem? Difficult.
A Framework for Distributed Model Predictive Control
Frank Edward Curtis Northwestern University Joint work with Richard Byrd and Jorge Nocedal February 12, 2007 Inexact Methods for PDE-Constrained Optimization.
Adaptive CSMA under the SINR Model: Fast convergence using the Bethe Approximation Krishna Jagannathan IIT Madras (Joint work with) Peruru Subrahmanya.
Week 10Complexity of Algorithms1 Hard Computational Problems Some computational problems are hard Despite a numerous attempts we do not know any efficient.
Frank Edward Curtis Northwestern University Joint work with Richard Byrd and Jorge Nocedal January 31, 2007 Inexact Methods for PDE-Constrained Optimization.
UNC Chapel Hill M. C. Lin Introduction to Motion Planning Applications Overview of the Problem Basics – Planning for Point Robot –Visibility Graphs –Roadmap.
Administration Feedback on assignment Late Policy
Presenter : Kuang-Jui Hsu Date : 2011/3/24(Thur.).
CSC 213 – Large Scale Programming. Today’s Goals  Discuss what is meant by weighted graphs  Where weights placed within Graph  How to use Graph ’s.
Decision Making Under Uncertainty PI Meeting - June 20, 2001 Distributed Control of Multiple Vehicle Systems Claire Tomlin and Gokhan Inalhan with Inseok.
State space representations and search strategies - 2 Spring 2007, Juris Vīksna.
Graphs Chapter 12. Chapter 12: Graphs2 Chapter Objectives To become familiar with graph terminology and the different types of graphs To study a Graph.
LIMITATIONS OF ALGORITHM POWER
Efficient Resource Allocation for Wireless Multicast De-Nian Yang, Member, IEEE Ming-Syan Chen, Fellow, IEEE IEEE Transactions on Mobile Computing, April.
A global approach Finding correspondence between a pair of epipolar lines for all pixels simultaneously Local method: no guarantee we will have one to.
Coverage Problems in Wireless Ad-hoc Sensor Networks Seapahn Meguerdichian 1 Farinaz Koushanfar 2 Miodrag Potkonjak 1 Mani Srivastava 2 University of California,
Mesh Segmentation via Spectral Embedding and Contour Analysis Speaker: Min Meng
Distributed cooperation and coordination using the Max-Sum algorithm
Searching a Linear Subspace Lecture VI. Deriving Subspaces There are several ways to derive the nullspace matrix (or kernel matrix). ◦ The methodology.
Solving a System of 3 Equations with 3 Unknowns. Breakdown Step 1 Labeling Step 2 Reduce to a 2 by 2 Step 3 Substitute Back In Step 4 Check Solution.
Camera Calibration Course web page: vision.cis.udel.edu/cv March 24, 2003  Lecture 17.
Construction of Optimal Data Aggregation Trees for Wireless Sensor Networks Deying Li, Jiannong Cao, Ming Liu, and Yuan Zheng Computer Communications and.
Spanning Trees Dijkstra (Unit 10) SOL: DM.2 Classwork worksheet Homework (day 70) Worksheet Quiz next block.
Discrete Optimization MA2827 Fondements de l’optimisation discrète Material from P. Van Hentenryck’s course.
Spectral Methods for Dimensionality
Prof. Yu-Chee Tseng Department of Computer Science
Optimal UAV Flight Path Selection
Outline Nonlinear Dimension Reduction Brief introduction Isomap LLE
Graph Searching.
All pairs shortest path problem
Survey on Coverage Problems in Wireless Sensor Networks - 2
“Traditional” image segmentation
Survey on Coverage Problems in Wireless Sensor Networks
Presentation transcript:

Optimal Trajectory for Network Establishment of Remote UAVs –1–1 Prachya Panyakeow, Ran Dai, and Mehran Mesbahi American Control Conference June 2013

Motivation Cooperative control of multi-vehicle systems –2–2

Motivation Cooperative control of multi-vehicle systems –3–3

Motivation Cooperative control of multi-vehicle systems –4–4

Motivation Reconnaissance, surveillance, monitoring, imaging, data processing –5–5

Motivation Why disperse? Coverage issue Limit field of view Why form a connected network? Energy efficiency of a formation Information sharing –6–6

Outline Background and Problem Formulation Optimal path planning for target tree-graph connectivity  Control law (PMP with end-point manifold)  Control sequence (Nonlinear Opt Control necessary conditions)  Eliminate candidates (Geometry Estimation Method)  Computational issues Nonlinear Programming Method Pros/Cons for each approach Future work –7–7

Background and Problem Formulation –8–8 Objective: Find paths that bring scattered UAVs into proximity to form a connected network at terminal time with minimum total control effort. Nonlinear Dynamics of Each UAVs: Assumptions: UAVs are first far from each other The connected network at final time is denoted as graph The mobile agents represent the vertex set in the final connected network The communication or relative sensing channel represents edge set The initial states are given as and terminal time is given as

Elements in Adjacency matrix A are determined by edges of Euclidean distance based connection: Laplacian Matrix, Network connectivity constraint: Review and Background –9–9 Related works:  Spanos and Murray, 2004, Robust connectivity of networked vehicles.  Zavlanos and Pappas, 2007, Maintaining connectivity of mobile networks  Kim and Mesbahi, 2006, Maximizing the second smallest eigenvalue of a state-dependent graph Laplacian.  Dai, Maximoff, and Mesbani, 2012, Formation of connected network for fractionated spacecraft Euclidean distance Based connection d 1 0 D 1010

UAV Network Establishment –10 Problem Formulation:

UAV Network Establishment –11 Problem Formulation: Logical Constraint

UAV Network Establishment –12 Problem Formulation: Logical Constraint Nonlinear Constraint

Optimal Target Tree-Graph Connectivity –13 Problem Formulation:

Optimal Target Tree-Graph Connectivity –14 Direct Method: Solve the Nonlinear Optimal Control Problem Hamiltonian: End-point Manifold:

Optimal Target Tree-Graph Connectivity –15 Direct Method: Solve the Nonlinear Optimal Control Problem PMP with End-Point Manifold Control Law

Dubin’s Problem –16

Optimal Target Tree-Graph Connectivity –17 Direct Method: Solve the Nonlinear Optimal Control Problem Control Sequence

Optimal Target Tree-Graph Connectivity –18 Direct Method: Solve the Nonlinear Optimal Control Problem Proposition 1: (Path-Synthesis) found from intermediate/final conditions Control law/sequence Necessary conditions 5n-2 unknowns Final Constraints Intermediate Constraints 2n-2 n-1 n 1 n 5n-2 nonlinear equations Substitute

Optimal Target Tree-Graph Connectivity –19 Direct Method: Solve the Nonlinear Optimal Control Problem Optimal Trajectories Candidates: Candidates RLL-RRLCandidates RRL-RLL(Global Sol.)

Optimal Target Tree-Graph Connectivity –20 Direct Method: Solve the Nonlinear Optimal Control Problem Geometry Estimation Method for eliminating the candidates

Optimal Target Tree-Graph Connectivity –21 Direct Method: Solve the Nonlinear Optimal Control Problem Geometry Estimation Method for eliminating the candidates Direction to turn Choose initial guess for Solve for Using proposition 1 Check result Terminate No Yes

Optimal Target Tree-Graph Connectivity –22 Nonlinear Optimal Control with Geometry Estimation Method Pros: Optimal Solution Provides the solution (switching time) for a given graph Cons: Computational issues (NP Hard) Number of final network configurations is exponential Cayley’s Theorem: Number of distinct labeled trees on n agents is Global search of all tree graphs with five agents: 1~2 minutes Global search of all tree graphs with eight agents: 1~2 days! An efficient algorithm is required to approach the problem

Nonlinear Programming Method –23 Parameterized Optimization Problem: Transform the original problem to Using the same 3 segment bang-bang control scheme,as unknown Same control law/sequence Relaxation of Logical On/Off Constraint Relaxation of Connectivity Nonlinear Constraint

Nonlinear Programming Method –24 Model of communication/relative sensing link: The entry of weight adjacency matrix is assigned as An exponential function for power of communication link Relaxation of the logical constraints The communication efficacy drops off continuously as the distance between the agents increases d 1 0 D

NLP Method (Parameter Optimization Problem) –25 Relaxation of connectivity constraints via Matrix similarity transformation: Original mixed integer problem with nonlinear constraints Semi-definite constraintNonlinear constraint Parameter Optimization Problem with semi-definite constraints = small positive number to guarantee The weight network is connected

–26 Optimal Target-Tree Method NLP Method VS Scalability of the two methods

–27 Optimal Target-Tree Method NLP Method VS Global Optimal Solution Exact results for a given graph Have to search for all possible graphs Not scalable for large-scale systems Sub-Optimal Solution (NP Hard) Base from the same bang-bang control law/sequence Faster Convergence without going through all possible graph Scalable for large-scale systems Scalability of the two methods

Future Work –28 Vary UAV speed between stall/max Optimal dispersion Considering scenarios that combine  Optimal path planning for network connectivity  Maintain the formation  Collision Avoidance  Optimal path planning for network dispersion