An example

For instance, assume cruise control. Therein pack, a technique occurs as car. the goal of sail control is to keep a car at a constant speed. On this text, a output variable of the patterns is the speed of the car. A primary means to control a speed of a car is the total of fuel existence fed into the engine.

a elementary way to implement sail control is to lock a position of a throttle the moment the driver engages sail control. This is amercement whenever a car is camping in perfectly flat terrain. In craggy terrain, a car will slow down while running rising & accelerate whenever running downward-sloping; something its driver may call for extremely unsuitable.

This nature and severity of controller is known as an open-loop controller because there is no directly connection between a output of the body & its input. One of a independent disadvantages of this nature and severity of controller is the deficiency of sensitivity to the kinetics of the technique in restraint.

History
A importance of this topic of learn was recognized when you took a development of A aeroplane: The Wright Brothers made their first successful end line text flights inside December 17, 1903 and by 1904 Flyer III was capable of fully-controllable stable flight for substantial periods. Control of the plane was necessary for its safe, economic, & economically successful have.

By World War II, control theory was an important a share of fire control, guidance, and cybernetics. A Space Race to the Moon depended on precise control of the ballistic capsule. However control theory is not lone utile inside technical applications.

Classical control theory

To keep away from a problems of the open-loop controller, control theory introduces feedback. A output of the rules $y(t)$ is fed back to the information value $r(t)$. A controller C so requires a difference between a information & a output, the error e, to vary a inputs u to the patterns in check P. This is shown in the figure. This kinda controller occurs as closed-loop controller or feedback controller. This occurs as and so-alleged individual-input-individual-output (SISO) control rules: lesson in which the single or even extra variables might contain to a higher degree a value (MIMO, i personally.e. Multi-Input-Multi-Output - for instance while outputs to become controlled come deuce or extra) come frequent. Inside such legal actions variables come represented across vectors instead of simple scalar values.

Whenever i personally look at a controller C & a plant P come linear and time-invariant (i personally.e.: elements of their transport work $C(s)$ & $P(s)$ don't depend from either period), i personally may analyze a rules above by using the Laplace transform on the variables. This gives a states the as a result relations:

Solving for Y(s) in terms of R(s), we obtain:

A term $\backslash frac$ is known as a transfer function of the system. In case i might assure $P(s)C(s)\; >>\; Ace$, i personally.e. it has super groovy norm with each value of $s$, so $Y(s)$ is roughly up to $R(s)$. This means i control a output by only setting a information.

Stability

Stability (around control theory) means that for any delimited input across any total of instance, a output might too exist as bounded. This is referred to as BIBO stability (see too Lyapunov stability). In case a technique is BIBO horse barn so the output just can't "blow up" whenever a input remains finite. Mathematically, this means that for the linear continuous-instance formulas to exist as stable 100% of the poles of its transfer function must lie in the left half of the complex plane if the Laplace transform is used (i personally.e. its really section is greater or even equal than zero) OR lie within a unit circle if the Z-transform is used (i personally.e. its module is greater or even up to of these)

In the 2 subjects, in case severally the pole has the very a portion strictly greater of zero or even a module strictly greater than of these, i speak of asymptotic stability: the variables of an asymptotically stable control body universally decrease from either their initial value & don't indicate lasting cycles, which are then instead present in case the pole has exactly the rattling section up to zero (or even the module up to a single). Whenever the only stable rules response neither decays nor grows on top instance, & has there is no cycles, it known as marginally stable: in this experience it has non-repeated poles along a vertical axis (we.e. their really & complex component is zero). Cycles come present after poles by using very a portion capable zero keep close at hand likewise complex section non up to zero.

Difference between them events are non the contradiction. A Laplace transform is within Cartesian coordinates and the Z-transform is within circular coordinates and it can be shown that a negative-real section in the Laplace domain potty map onto the interior of the unit circle a caring-real section in the Laplace domain may map onto the exterior of the unit circle

Whenever a rules around wonder has an impulse response of

& shopping for a Z-transform (see this example), it yields

which has the pole around $z\; =\; Cipher.V$ (zero imaginary part). This technique is BIBO (asymptotically) stable since a pole is insideA unit circle.

Even so, whenever a impulse response was

so a Z-transform is

which has the pole at $z\; =\; Single.Little\; phoebe$ & is non BIBO horse barn since the pole has a module stricly greater than a single.

State space representation
Watch State space (controls).

Controllability and observability
Look at controllability and observability.