Colloquium on Optimisation and Control University of Sheffield Monday April 24 th 2006 Sensitivity Analysis and Optimal Control Richard Vinter Imperial.

Slides:



Advertisements
Similar presentations
The Maximum Principle: Continuous Time Main purpose: to introduce the maximum principle as a necessary condition that must be satisfied by any optimal.
Advertisements

What is Optimal Control Theory? Dynamic Systems: Evolving over time. Time: Discrete or continuous. Optimal way to control a dynamic system. Prerequisites:
Chapter 11-Functions of Several Variables
State Variables.
Lect.3 Modeling in The Time Domain Basil Hamed
Linear Programming. Introduction: Linear Programming deals with the optimization (max. or min.) of a function of variables, known as ‘objective function’,
Chapter 9: Vector Differential Calculus Vector Functions of One Variable -- a vector, each component of which is a function of the same variable.
ESSENTIAL CALCULUS CH11 Partial derivatives
Visual Recognition Tutorial
D Nagesh Kumar, IIScOptimization Methods: M1L1 1 Introduction and Basic Concepts (i) Historical Development and Model Building.
Optimization in Engineering Design 1 Lagrange Multipliers.
Lagrangian and Hamiltonian Dynamics
Engineering Optimization
ERE5: Efficient and optimal use of environmental resources
Constrained Maximization
Economics 214 Lecture 37 Constrained Optimization.
Lecture 35 Constrained Optimization
Constrained Optimization Economics 214 Lecture 41.
Theoretical Mechanics - PHY6200 Chapter 6 Introduction to the calculus of variations Prof. Claude A Pruneau, Physics and Astronomy Department Wayne State.
INSTITUTO DE SISTEMAS E ROBÓTICA 1/31 Optimal Trajectory Planning of Formation Flying Spacecraft Dan Dumitriu Formation Estimation Methodologies for Distributed.
Dynamic Optimization Dr

Normalised Least Mean-Square Adaptive Filtering
Managerial Economics Managerial Economics = economic theory + mathematical eco + statistical analysis.
AUTOMATIC CONTROL THEORY II Slovak University of Technology Faculty of Material Science and Technology in Trnava.
GENERAL PRINCIPLES OF BRANE KINEMATICS AND DYNAMICS Introduction Strings, branes, geometric principle, background independence Brane space M (brane kinematics)
Revision Previous lecture was about Generating Function Approach Derivation of Conservation Laws via Lagrangian via Hamiltonian.
1 Adaptive, Optimal and Reconfigurable Nonlinear Control Design for Futuristic Flight Vehicles Radhakant Padhi Assistant Professor Dept. of Aerospace Engineering.
DSGE Models and Optimal Monetary Policy Andrew P. Blake.
ECE 8443 – Pattern Recognition ECE 8423 – Adaptive Signal Processing Objectives: Deterministic vs. Random Maximum A Posteriori Maximum Likelihood Minimum.
LAGRANGE mULTIPLIERS By Rohit Venkat.
Sec 15.6 Directional Derivatives and the Gradient Vector
Lecture 7 and 8 The efficient and optimal use of natural resources.
Ch. 3: Geometric Camera Calibration
Håkon Dahl-Olsen, Sridharakumar Narasimhan and Sigurd Skogestad Optimal output selection for batch processes.
Lecture 26 Molecular orbital theory II
Section 15.6 Directional Derivatives and the Gradient Vector.
Phy 303: Classical Mechanics (2) Chapter 3 Lagrangian and Hamiltonian Mechanics.
Functions of Several Variables Copyright © Cengage Learning. All rights reserved.
Optimal Path Planning Using the Minimum-Time Criterion by James Bobrow Guha Jayachandran April 29, 2002.
دانشگاه صنعتي اميركبير دانشكده مهندسي پزشكي استاد درس دكتر فرزاد توحيدخواه بهمن 1389 کنترل پيش بين-دکتر توحيدخواه MPC Stability-2.
Optimization in Engineering Design 1 Introduction to Non-Linear Optimization.
Review of PMP Derivation We want to find control u(t) which minimizes the function: x(t) = x*(t) +  x(t); u(t) = u*(t) +  u(t); (t) = *(t) +  (t);
Structure and Synthesis of Robot Motion Dynamics Subramanian Ramamoorthy School of Informatics 2 February, 2009.
By Verena Kain CERN BE-OP. In the next three lectures we will have a look at the different components of a synchrotron. Today: Controlling particle trajectories.
Economics 2301 Lecture 37 Constrained Optimization.
Copyright © Cengage Learning. All rights reserved. 14 Partial Derivatives.
Chapter 4 The Maximum Principle: General Inequality Constraints.
Section 15.3 Constrained Optimization: Lagrange Multipliers.
1 Introduction Optimization: Produce best quality of life with the available resources Engineering design optimization: Find the best system that satisfies.
Optimal Path Planning on Matrix Lie Groups Mechanical Engineering and Applied Mechanics Sung K. Koh U n i v e r s i t y o f P e n n s y l v a n i a.
Searching a Linear Subspace Lecture VI. Deriving Subspaces There are several ways to derive the nullspace matrix (or kernel matrix). ◦ The methodology.
Amir Yavariabdi Introduction to the Calculus of Variations and Optical Flow.
YLE13: Optimal control theory
Classical Mechanics Lagrangian Mechanics.
Chapter 14 Partial Derivatives.
YLE13: Hotelling model Marko Lindroos.
Optimal control T. F. Edgar Spring 2012.
Chap 9. General LP problems: Duality and Infeasibility
The Lagrange Multiplier Method
7.5 – Constrained Optimization: The Method of Lagrange Multipliers
13 Functions of Several Variables
Outline Unconstrained Optimization Functions of One Variable
Synthesis of SISO Controllers
EE 458 Introduction to Optimization
Chapter 4 . Trajectory planning and Inverse kinematics
Optimal Control of Systems
Analysis of Basic PMP Solution
Chapter 6. Large Scale Optimization
L8 Optimal Design concepts pt D
Presentation transcript:

Colloquium on Optimisation and Control University of Sheffield Monday April 24 th 2006 Sensitivity Analysis and Optimal Control Richard Vinter Imperial College

Sensitivity Analysis Sensitivity analysis: the effects of parameter changes on the solution of an optimisation problem: Practical Relevance: Resource economics (economic viability of optimal resource extraction in changing environment) Design (buildings to withstand earthquakes,..) Theoretical Relevance: Intimate links with theory of constrained optimization (Lagrange multipliers, etc.) Intermediate step in mini-max optimisation ‘parametric’ approaches to MPC

The Value Function Minimizeover s.t. and Data: m vector parameter Value function: (describes how minimum cost changes with ) (no constraints case)

Links With Lagrange Multipliers Minimize over s.t. (m vector parameter is value of equality constraint function ) Lagrange multiplier rule: Fix. Suppose is a minimiser for. Then for some m vector ‘Lagrange multiplier’ Special case:

Value function: Fact: The Lagrange multiplier has interpretation: ( is the gradient of the value function associated with perturbations of the constraint ) Show this: For any,, sofor By ‘minimality’: (since ) Hence, where (Caution: analysis not valid unless V is differentiable.)

Consider now the optimal control problem: Minimize s.t. Most significant value function is associated with perturbation of initial data: (data: ) sets and Minimize s.t. t x Domain of andand and

Pontryagin Maximum Principle Take a minimizer Define ‘Hamiltonian’: Then, for some ‘co-state arc’ where (adjoint equation) (max. of Hamiltonian cond.) (transversality cond.) (maximised Hamiltonian) is the normal cone at x C ‘normal vector’ at x

Sensitivity Relations in Optimal Control Gradients of value function w.r.t. ‘initial data’ are related to co-state variable What if V is not differentiable? Interpret sensitivity relation in terms of set valued ‘generalized gradients’: (definition for ‘Lipshitz functions’, these are ‘almost everywhere’ differentiable) +1 (Valid for non-differentiable value functions) For some choice of co-state p(.)

Generalizations Minimize s.t. Dynamics and cost depend on par. Obtain sensitivity relations (gradients of V’ ) by ‘state augmentation’. with extra state equationIntroduce expressible in terms of co-state arcs for state augmented problem, nominal value )( and

Application to ‘robust’ selection of feedback controls Classical tracker design: Step 1: Determine nominal trajectory using optimal control Step 2: Design f/b to track the nominal trajectory (widely used in space vehicle design) Can fail to address adequately conflicts between performance and robustness Alternatively, Integrate design steps 1 and 2 Append ‘sensitivity term’ in the optimal control cost to reduce effect of model inaccuracies This is ‘robust optimal control’

Robust Optimal Control 1) Model dynamics: 3) Model variables requiring de-sensitisation: 4) Feedback control law: Example: magnitude of deviation from desired terminal location : Objective: find sub-optimal control which reduces sensitivity of to deviation of from. 2) Model cost:

Sensitivity Relations For control, let be state trajectory for. The `sensitivity function’ has gradient: where the arc p (.) solves (Write

Optimal Control Problem with Sensitivity Term Minimize s.t. and Pink blocks indicate extra terms to reduce sensitivity is sensitivity tuning parameter 0 values sensitiveinsensitive

Trajectory Optimization for Air-to-Surface Missiles with Imaging Radars Researchers: Farooq, Limebeer and Vinter. Sponsors: MBDA, EPSRC ‘Terminal guidance strategies for air-to-surface missile using DBS radar seeker’. Specifications include: Stealthy terrain phase, followed by climb and dive phase (‘bunt’ trjectory) Sharpening radars impose azimuthal plane constraints on trajectory Stealth phase Bunt phase Six degree of freedom model of skid-to-turn missile (two controls: normal acceleration demains Select cost function to achieve motion, within constraints.

References: R B VINTER, Mini-Max Optimal Control, SIAM J. Control and Optim., 2004 V Papakos and R B Vinter, A Structured Robust Control Technique, CDC 2004 A Farooq and D J N Limebeer, Trajectory Optimization for Air-to-Surface Missiles with Imaging Radars, AIAA J., to appear