V-Formation via Model Predictive Control

We present recent results that demonstrate the power of viewing the problem of V-formation in a flock of birds as one of Model Predictive Control (MPC). The V-formation-MPC marriage can be understood in terms of the problem of synthesizing an optimal plan for a continuous-space and continuous-time Markov decision process (MDP), where the goal is to reach a target state that minimizes a given cost function. First, we consider ARES, an approximation algorithm for generating optimal plans (action sequences) that take an initial state of an MDP to a state whose cost is below a specified (convergence) threshold. ARES uses Particle Swarm Optimization, with adaptive sizing for both the receding horizon and the particle swarm. Inspired by Importance Splitting, the length of the horizon and the number of particles are chosen such that at least one particle reaches a next-level state. ARES can alternatively be viewed as a model-predictive control (MPC) algorithm that utilizes an adaptive receding horizon, aka Adaptive MPC (AMPC). We next present Distributed AMPC (DAMPC), a distributed version of AMPC that works with local neighborhoods. We introduce adaptive neighborhood resizing, whereby the neighborhood size is determined by the cost-based Lyapunov function evaluated over a global system state. Our experiments show that DAMPC can perform almost as well as centralized AMPC, while using only local information and a form of distributed consensus in each time step. Finally, inspired by security attacks on cyber-physical systems, we introduce controller-attacker games (CAG), where two players, a controller and an attacker, have antagonistic objectives. We formulate a special case of CAG called V-formation games, where the attacker's goal is to prevent the controller from attaining V-formation. We demonstrate how adaptation in the design of the controller helps in overcoming certain attacks.

[1]  Geir E. Dullerud,et al.  Distributed control design for spatially interconnected systems , 2003, IEEE Trans. Autom. Control..

[2]  Michael S. Selig,et al.  The aerodynamic benefits of self-organization in bird flocks , 2003 .

[3]  Iztok Lebar Bajec,et al.  Organized flight in birds , 2009, Animal Behaviour.

[4]  Olivier Sigaud,et al.  Policy Improvement Methods: Between Black-Box Optimization and Episodic Reinforcement Learning , 2012 .

[5]  Martin Guay,et al.  Adaptive Model Predictive Control for Constrained Nonlinear Systems , 2008 .

[6]  Kumpati S. Narendra,et al.  Adaptive control using neural networks , 1990 .

[7]  B. Womack,et al.  Adaptive Control Using Neural Networks , 1991, 1991 American Control Conference.

[8]  Han-Lim Choi,et al.  An Unknown-Input-Observer Based Approach for Cyber Attack Detection in Formation Flying UAVs , 2016 .

[9]  Shie Mannor,et al.  The Cross Entropy Method for Fast Policy Search , 2003, ICML.

[10]  Cédric Pralet,et al.  Synthesis of plans or policies for controlling dynamic systems , 2012 .

[11]  Paulo Tabuada,et al.  Secure Estimation and Control for Cyber-Physical Systems Under Adversarial Attacks , 2012, IEEE Transactions on Automatic Control.

[12]  Gary William Flake,et al.  The Computational Beauty of Nature: Computer Explorations of Fractals, Chaos, Complex Systems and Adaptation , 1998 .

[13]  Valmir Carneiro Barbosa,et al.  V-like Formations in Flocks of Artificial Birds , 2006, Artificial Life.

[14]  Marko Bacic,et al.  Model predictive control , 2003 .

[15]  Florian Dörfler,et al.  Attack Detection and Identification in Cyber-Physical Systems -- Part II: Centralized and Distributed Monitor Design , 2012, ArXiv.

[16]  Ezio Bartocci,et al.  Policy Learning for Time-Bounded Reachability in Continuous-Time Markov Decision Processes via Doubly-Stochastic Gradient Ascent , 2016, QEST.

[17]  Hovav Shacham,et al.  Comprehensive Experimental Analyses of Automotive Attack Surfaces , 2011, USENIX Security Symposium.

[18]  Insup Lee,et al.  Sensor attack detection in the presence of transient faults , 2015, ICCPS.

[19]  Radu Grosu,et al.  Attacking the V: On the Resiliency of Adaptive-Horizon MPC , 2017, ATVA.

[20]  Riccardo Poli,et al.  Particle swarm optimization , 1995, Swarm Intelligence.

[21]  Forrest Stonedahl,et al.  Finding Forms of Flocking: Evolutionary Search in ABM Parameter-Spaces , 2010, MABS.

[22]  Sean R Eddy,et al.  What is dynamic programming? , 2004, Nature Biotechnology.

[23]  Paulo Tabuada,et al.  Robustness of attack-resilient state estimators , 2014, 2014 ACM/IEEE International Conference on Cyber-Physical Systems (ICCPS).

[24]  Hao Wu,et al.  Controlling UAVs with Sensor Input Spoofing Attacks , 2016, WOOT.

[25]  Cutts,et al.  ENERGY SAVINGS IN FORMATION FLIGHT OF PINK-FOOTED GEESE , 1994, The Journal of experimental biology.

[26]  P. Lissaman,et al.  Formation Flight of Birds , 1970, Science.

[27]  Cyrille Jégourel,et al.  Feedback Control for Statistical Model Checking of Cyber-Physical Systems , 2016, ISoLA.

[28]  Ali H. Sayed,et al.  Modeling Bird Flight Formations Using Diffusion Adaptation , 2011, IEEE Transactions on Signal Processing.

[29]  Vijay Kumar,et al.  Resilient Flocking for Mobile Robot Teams , 2017, IEEE Robotics and Automation Letters.

[30]  Craig W. Reynolds Flocks, herds, and schools: a distributed behavioral model , 1987, SIGGRAPH.

[31]  Bin Wu,et al.  Fast Particle Filters and Their Applications to Adaptive Control in Change-Point ARX Models and Robotics , 2009 .

[32]  Jingyuan Zhan,et al.  Flocking of Multi-Agent Systems Via Model Predictive Control Based on Position-Only Measurements , 2013, IEEE Transactions on Industrial Informatics.

[33]  Ashish Tiwari,et al.  ARES: Adaptive Receding-Horizon Synthesis of Optimal Plans , 2016, TACAS.

[34]  Olivier Sigaud,et al.  Path Integral Policy Improvement with Covariance Matrix Adaptation , 2012, ICML.

[35]  Christel Baier,et al.  Tools and Algorithms for the Construction and Analysis of Systems , 2015, Lecture Notes in Computer Science.

[36]  Zhaowei Sun,et al.  Extended state observer–based finite-time controller design for coupled spacecraft formation with actuator saturation , 2017 .

[37]  Arthur J. Krener,et al.  Adaptive Horizon Model Predictive Control , 2016, 1602.08619.

[38]  Anh Duc Dang,et al.  Formation control of autonomous robots following desired formation during tracking a moving target , 2015, 2015 IEEE 2nd International Conference on Cybernetics (CYBCONF).

[39]  Bernard Chazelle,et al.  The Convergence of Bird Flocking , 2009, JACM.

[40]  H. Weimerskirch,et al.  Energy saving in flight formation , 2001, Nature.

[41]  Peter J Seiler,et al.  Analysis of bird formations , 2002, Proceedings of the 41st IEEE Conference on Decision and Control, 2002..

[42]  Edmund M. Clarke,et al.  Statistical Model Checking for Markov Decision Processes , 2012, 2012 Ninth International Conference on Quantitative Evaluation of Systems.

[43]  F. Vasca,et al.  Formation Control and Collision Avoidance in Mobile Agent Systems , 2005, Proceedings of the 2005 IEEE International Symposium on, Mediterrean Conference on Control and Automation Intelligent Control, 2005..

[44]  Magnus Egerstedt,et al.  Adaptive time horizon optimization in model predictive control , 2011, Proceedings of the 2011 American Control Conference.

[45]  Peter L. Bartlett,et al.  Experiments with Infinite-Horizon, Policy-Gradient Estimation , 2001, J. Artif. Intell. Res..

[46]  Ashish Tiwari,et al.  Love Thy Neighbor: V-Formation as a Problem of Model Predictive Control , 2016, CONCUR.

[47]  R. D'Andrea,et al.  Distributed control of close formation flight , 2002, Proceedings of the 41st IEEE Conference on Decision and Control, 2002..

[48]  C. R. Ramakrishnan,et al.  Using Statistical Model Checking for Measuring Systems , 2014, ISoLA.

[49]  Thomas Hérault,et al.  Approximate Probabilistic Model Checking , 2004, VMCAI.