Scribd
Upload
78 views

Advanced Flight Mechanics Optimal Control: Manuel Sanjurjo Rivo

The document discusses optimal control theory and its application to aircraft flight mechanics. It introduces concepts like the cost function, necessary conditions for optimality using calculus of variations, and generating neighboring optimal control laws through linearization. The document outlines these topics over several sections.

Uploaded by

Fernando Aspás Venegas
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
78 views

Advanced Flight Mechanics Optimal Control: Manuel Sanjurjo Rivo

The document discusses optimal control theory and its application to aircraft flight mechanics. It introduces concepts like the cost function, necessary conditions for optimality using calculus of variations, and generating neighboring optimal control laws through linearization. The document outlines these topics over several sections.

Uploaded by

Fernando Aspás Venegas
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 10

Introduction Statement Conditions Neighbouring

Advanced Flight Mechanics


Optimal Control

Manuel Sanjurjo Rivo

Degree in Aerospace Engineering


Escuela Politecnica Superior
Area de Ingeniera Aeroespacial

UC3M - Degree in Aerospace Engineering - Manuel Sanjurjo Advanced Flight Mechanics - Opt Contron
Introduction Statement Conditions Neighbouring

Outline

1 Introduction

2 Statement of the Problem

3 Conditions for Optimality

4 Continuous Neighbouring Optimal Control

UC3M - Degree in Aerospace Engineering - Manuel Sanjurjo Advanced Flight Mechanics - Opt Contron
Introduction Statement Conditions Neighbouring

Introduction

Controls considered only to come back to equilibrium


More general approach. No restricted to linear approximation
Accomplish a mission
Performance
Subject to optimization
Time
Fuel
Also come back to equilibrium efficiently
Instead of trial and error
Optimal control
Other ways of tuning gains

UC3M - Degree in Aerospace Engineering - Manuel Sanjurjo Advanced Flight Mechanics - Opt Contron
Introduction Statement Conditions Neighbouring

Definitions

u(t) control vector


Elevator, aileron, ...
x(t) state vector

Cost function, J
Any function of state and control
Time, mass, normal acceleration, ...
Z tf
J = (x(tf ), tf ) + L (x(t), u(t), t)dt
t0

UC3M - Degree in Aerospace Engineering - Manuel Sanjurjo Advanced Flight Mechanics - Opt Contron
Introduction Statement Conditions Neighbouring

Problem Statement

Types of cost function: Meyer, Lagrange, Bolza.


Control acts on dynamical system

x = f (x, u, t) x(t0 ) = x0

Statement of the Problem


Find the control history u (t), t0 t tf that minimizes a cost
function J, producing an optimal trajectory x (t)

UC3M - Degree in Aerospace Engineering - Manuel Sanjurjo Advanced Flight Mechanics - Opt Contron
Introduction Statement Conditions Neighbouring

Necessary Conditions

Optimal trajectories solution of equations stating necessary


and sufficient conditions for optimality
Necessary Condition
Control history that causes the cost function to be stationary with
respect to control u and satisfies dynamics of the system

Derivation based on Calculus of Variations.


Definition of Hamiltonian

H (x, u, , t) = L (x(t), u(t), t) T f (x, u, t)

UC3M - Degree in Aerospace Engineering - Manuel Sanjurjo Advanced Flight Mechanics - Opt Contron
Introduction Statement Conditions Neighbouring

Euler-Lagrange Equations

H (x, u, , t)
xT =

H (x, u, , t)
T =
x
H (x, u, , t)
0=
u
subject to

T (x(t), t)
x(t0 ) = x0 (tf ) =
x
t=tf

Two Point Boundary Value Problem

UC3M - Degree in Aerospace Engineering - Manuel Sanjurjo Advanced Flight Mechanics - Opt Contron
Introduction Statement Conditions Neighbouring

Sufficient Conditions

In many practical problems, stationary = optimal


Tests to assure optimality
Legendre-Clebsch condition (convexity) in Hessian matrix

2H
>0
u2
Existence of neighbouring-optimal solutions
Jacobi condition: no points of intersection of optimal
trajectories

UC3M - Degree in Aerospace Engineering - Manuel Sanjurjo Advanced Flight Mechanics - Opt Contron
Introduction Statement Conditions Neighbouring

Motivation

Generation of optimal paths close to nominal


Changes due to initial conditions, end conditions, parameters,
disturbances, ...
Based on linearization of cost function (and dynamics)
Same approach as in AFM
J(x + x) = J(x ) + 2 J(x)
x = F(t)x + G(t)u
Linear Euler-Lagrange equations
1 T
H = x Qx + 2xT M u + uT Ru +

2
+ T [Fx + Gu]

UC3M - Degree in Aerospace Engineering - Manuel Sanjurjo Advanced Flight Mechanics - Opt Contron
Introduction Statement Conditions Neighbouring

Optimal Feedback Law

Defined as:

u = R1 GT P(t) + MT x = C(t)x
 

P(t) is defined as = P(t)u and unknown


Riccati matrix
LQ (Linear Quadratic control law)
Solve for Riccati matrix
T
P = F GR1 MT P P F GR1 MT +


+ PGR1 GT P Q + MR1 MT

P(tf ) = xx (tf )

UC3M - Degree in Aerospace Engineering - Manuel Sanjurjo Advanced Flight Mechanics - Opt Contron

You might also like