Scribd
Upload
2 views

chayma

The document discusses Linear Quadratic Regulators (LQR), a method in optimal control theory that minimizes costs in dynamic systems through a mathematical algorithm. It outlines the formulation of the LQR, including the cost function and the state feedback control law, while also addressing the determination of parameters Q and R. Additionally, it highlights the advantages of LQR, such as optimal control and state feedback, alongside its disadvantages, including complexity and limitations in non-linear systems.

Uploaded by

ali.belaid.tt
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
2 views

chayma

The document discusses Linear Quadratic Regulators (LQR), a method in optimal control theory that minimizes costs in dynamic systems through a mathematical algorithm. It outlines the formulation of the LQR, including the cost function and the state feedback control law, while also addressing the determination of parameters Q and R. Additionally, it highlights the advantages of LQR, such as optimal control and state feedback, alongside its disadvantages, including complexity and limitations in non-linear systems.

Uploaded by

ali.belaid.tt
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

I.1.

Linear Quadratic Regulators (LQR)


The goal of optimal control theory is to run a dynamic system as cheaply as possible.
The LQ problem is the situation in which a quadratic functional describes the cost and a set of
linear differential equations describes the system dynamics. The linear-quadratic regulator
(LQR), a feedback controller whose equations are supplied below, provides the solution,
according to one of the theory's primary findings [3].

I.1.1. General description of LQR


In layman's terms, Linear Quadratic Regulator (LQR) refers to the settings of a
(regulating) controller governing a machine or process (such as an aeroplane or chemical
reactor) that are determined by a mathematical algorithm that minimises a cost function with
weighting factors supplied by a human (engineer). The "cost" (function) is frequently defined
as the sum of important measurement deviations from their expected values. In effect, this
method identifies controller settings that minimise unwanted deviations, such as departures
from the target altitude or process temperature. The magnitude of the control action is
sometimes included in this amount in order to minimise the energy wasted by the control
action itself [32].

Figure 29: LQr block diagram [21]

The linear quadratic regulator (LQR) is a popular design method that offers useful
feedback benefits. To derive the linear quadratic regulator, assume the plant is written in
state-space form as follows:[3]

For a continuous linear system that is represented a state space:[34]

ẋ= Ax+ Bu (Ⅱ.44)

The cost function is defined as follows:


∞
J=∫ (x Qx+u Ru) dt
T T
(Ⅱ.45)
0
where Q is a Hermitian matrix or real positive definite symmetric matrix (or real
positive definite symmetric matrix), and R is a real symmetric matrix or real positive definite
symmetric matrix. The matrices Q and R determine the relative importance of the error and
the consumption of that energy and these matrices contribute to the function j.

The state feedback control law that minimises the cost function is as follows:

u=−Kx (Ⅱ.46)

When K is defined as follows:


−1
K=R B P
T
(Ⅱ.47)

P can be determined by solving the following algebraic equation, commonly known as


the Riccati equation:
T −1 T
A P+ PA−PB R B P+Q=0 (Ⅱ.48)

The optimal control law is given by equation (II.46)

I.2. Determine the Q and R coefficients [5]


The aim is to create adequate control laws according to the parameters Q and R.

To determine these parameters, we can be guided by knowledge of some limiting


properties when Q or R tend towards the end of the range:

Q 0:
The control is minimum energy since only 𝑢 is penalised. An unstable system requires a
fast closed loop if we want to make the minimum effort, for example, we cannot restore the
equilibrium of an inverted pendulum without a minimum of energy and this case is not
suitable for the control of the pendulum.

Q ∞ (or R  0):
The control is in minimum time, since the control has just been completely freed up.
The foreseeable trend is for response times to accelerate.

𝐾 has so far been determined by assuming that the state vector is available.

State feedback requires knowledge of the state 𝑋, but this is not always possible
because sometimes sensors are not fitted for economic reasons and sometimes the state
variables have no physical meaning. In this case, a state reconstruction is carried out.

I.3. Advantages and disadvantages of LQR Regulators


This section aims to explore the advantages and disadvantages of using Linear quadratic
Regulator [10].
I.3.1. Advantages
1. Optimum control: LQR controllers provide optimal control by minimising a quadratic
performance index, resulting in optimal steady-state and transient responses under specific
conditions.
2. LQR control is based on strong mathematical concepts that provide a clear basis for
analysis and design.
3. LQR control is often outperforms classic PID control in terms of settling time, overshoot
and disturbance rejection.
4. State feedback: Because LQR control provides direct feedback of system states, it can
more accurately control multiple variables simultaneously, making it suitable for multi-
variable systems.
I.3.2. Disadvantages
1. Complexity: Designing and implementing LQR Regulators can be difficult, requiring
extensive system modelling and determination of optimal control gains using system
matrices.
2. Implementation difficulties: Practical application of LQR control may require
sophisticated knowledge of control theory, numerical methods and system identification
techniques.
3. Non-linear systems: LQR control is inherently designed for linear systems and may not
work effectively in non-linear systems unless linearization or adaptation is used.
4. State feedback: Because LQR control provides direct feedback of system states, it can
more accurately control multiple variables simultaneously, making it suitable for multi-
variable systems.

You might also like