2024年1月12日发(作者：)

自动控制论文

作者 洪劲松

专业 电气工程及其自动化

学号 *********

指导教师 赵国新

Automatic control is when no one is directly involved in the case, the use of

additional equipment or control device, the machine, device, or a working state of the

control object or parameters (charged) automatically according to the predetermined rules.

The traditional industrial production process using dynamic control technology, can

effectively improve the quality of the products and the enterprise economic benefit. In

today's rapid development of science and technology, automatic control technology in the

field of industrial and agricultural production, national defense and science and technology,

has a very important role. In a short span of one hundred years, the development of

automatic control theory has been surprising, has a huge impact on human society.

Automatic control theory is the study of automatic control common law science and

technology. It is both an ancient and has become a mature discipline, another door is

developing, the strong vitality of the emerging disciplines. From 1868 maxwell J.C.M

axwell low order system stability criterion is put forward to date more than one hundred

years, the development of automatic control theory can be divided into four main stages:

the first stage: the classical control theory (or) classical control theory of the formation,

development and maturity; The second stage: the rise of modern control theory and

development; The third stage: big system control the rise and development stage; The

fourth stage: intelligent control stage of development.

The basic characteristics of the first stage of the classical control theory is mainly

used for linear time-invariant systems research, namely for describing the system of linear

differential equation with constant coefficients of analysis and synthesis; It is used only for

single input and single output feedback control system; Only discuss the relationship

between the system input and output, and ignore the internal state of the system, is a

method of external description of the system. The basic method used: root locus method,

frequency method, PID regulator (frequency domain). Control theory in the early stage of

development, the automatic adjustment principle is based on the feedback theory, mainly

used in industrial control. Feedback theory for feedback control. Feedback control is one

of the most basic is the most important control mode, after the introduction of feedback

signal, system response to come from the external and internal interference become very

dull, so as to improve the anti-interference ability and the control precision of the system.

Feedback effects, meanwhile, brings the problem of system stability, which was once the

system stability problem in people inspired people to conduct the thorough research to the

feedback control system in the enthusiasm, promote the development of the theory of

automatic control and improvement. So in a sense, the classical feedback control theory is

accompanied by the emergence and development of control technology and gradually

improve and mature. During the second world war, in order to design and manufacture of

aircraft and Marine autopilot, artillery positioning system, radar tracking system based on

feedback principle of military equipment, to further promote and perfect the development

of automatic control theory. In 1868, maxwell (J.C.M axwell) lower order algebraic criterion

of the stability of the system are put forward. In 1875 and 1896, mathematicians rous

(Routh) and hull weitz (Hurwitz) respectively independently the stability criterion of high

order system was put forward, namely the Routh Hurwitz criterion. During the second

world war (1938-1945), Nyquist (H.N yquist) in 1948, proposed the theory of frequency

response Evans (W.R.E vans) root locus method is proposed. At this point, the control

theory basic to complete the first stage of development, formed the frequency method and

the root locus method as the main method of classical control theory.

The second stage: along with the development of the space industry and computer,

in the early 1960 s, on the basis of the classical control theory, on the basis of the linear

algebra theory and state space analysis of the rapid development of modern control

theory. Behrman, 1954 (r. elman) ministry in 1956 has put forward the dynamic

programming theory (L.S.P ontryagin) maximum principle is put forward in 1960 kalman

(R.K.K alman) put forward multivariable optimal control and optimal filtering theory. In

mathematical tool and theoretical basis can provide the information within the system

state variables. It no matter for linear system and nonlinear system, time-invariant

systems or time-varying system, single variable system or a multivariable system, is a

kind of effective analysis method. Its basic method is used in the equation of state (the

time)

In the 1970 s, the development of modern control theory, continue to the depth and

breadth, there have been some new control methods and theories. As modern frequency

domain methods: mathematical model based on transfer function matrix, the linear

time-invariant multivariable system; Adaptive control theory and methods: on the basis of

system identification and parameter estimation, on the basis of real-time identification

online to determine the optimal control law; Robust control methods: on the basis of

guarantee the system stability and other performance, robust controller design is constant,

the uncertainty of the mathematical model to deal with. With the enlargement of the

control theory, application scope, from individual small control systems, to large system

composed of several correlated subsystems for the overall control, from the traditional

field of engineering control are generalized to include economic management, biological

engineering, energy, transportation, environment and other large systems, and social

sciences. Large system theory is the combination of process control and information

processing system engineering theory, a large scale, complicated structure, functional

diversity of comprehensive, objective, many factors, etc. It is a multiple input, multiple

output, more interference, multivariable system. Large system theory is still in the stage of

development and pioneering.

Intelligent control is a newly developed a kind of control technology in recent years,

is the application of artificial intelligence in the control. Concept and principle of intelligent

control is mainly in view of the controlled object, the environment, the complexity of the

control objectives or tasks, its guiding ideology is based on people's way of thinking and

skills to deal with problems, solve the intelligence of those who currently need to solve the

control problem of complex. The complexity of the controlled object is as follows: the

uncertainty of model, highly nonlinear, a distributed sensor and actuator, dynamic

mutation, multiple time scale, the complex mode of information, large amount of data, as

well as the characteristic parameters of strictly and so on. Intelligent control is driving

intelligent machines independently the realization of the goal. Starting from the concept of

"humanoid" intelligent control. Its methods include learning control, fuzzy control, neural

network control and expert control method.

Automatic control system is to point to to the working state of the controlled object of

the automatic control system. It is to control the objects and devices involved in

implementing its automatic control by control quantity or the combination of components,

generally consists of a control device and the controlled object. Generally includes three

types of institutions: measurement institutions, institutions and actuators. Automatic

control system of function and form is varied, its structure simple and complex. It can only

control a physical quantities, can also control multiple physical quantities and even an

enterprise organization all the production and management process; It can be a specific

engineering system, also can be more abstract social system, ecological or economic

systems. Control system can be divided into constant value system and servo system; A

linear system and nonlinear system; Continuous system and discrete system.

Automatic control system in the process of production need to design first, and then

analysis can be used in the production. Control system design: according to the

performance of the control object and a given system, the determination of reasonable

structural parameters of the control device, called the control system design. The

structure of control system analysis: the known system parameters, the stability of the

system is analyzed, the calculating system of dynamic and static performance index, and

evaluation system on the basis of analysis of the process is referred to as control system.

Analysis and design of control system of the first task is to build a system

mathematical model. Once get the reasonable mathematical model, different analysis

methods can be used to analyze the performance of the system. Classical control theory,

engineering method is commonly used in time domain analysis, frequency response

method, the root locus method. The time domain analysis method in time domain

research system under the action of the typical input signal, the output response variation

law with time. For any a stable control system, the output response contains the transient

and steady-state components.

Design of an automatic control system generally through the following steps: 1

according to the mission requirements, selected control objects;

2 according to the requirements of performance indicators to determine the system

control rule, and design the controller, and satisfy the control law for the preliminary

selected controller components; 3 of the selected control objects and the control system, if

a system can't satisfy all or can't meet the design requirements of performance indicators,

also must add the appropriate components, in a certain way connected to the original

system, the combined system fully meet the design requirements.

Automatic control system design and analysis of mathematical model usually need

to use. Mathematical model is to describe system mathematical expression of the internal

relations between physical quantities or variable. The purpose of the mathematical model

is set up is a priority for the analysis and design of control system and basic work. The

composition of automatic control system can be electrical, mechanical, hydraulic or

pneumatic, etc., described the system development model, however, can be the same.

Through mathematical models to study automatic control system, can get rid of the

external characteristics of different type system, studies its inherent common motion law.

To establish the mathematical model of both accuracy and simplify the sex. General

should be based on the actual structure of the system parameters and the calculation

precision of requirements, omit some minor factors, the model can accurately reflect the

dynamic nature of the system, and can simplify the work of analysis and calculation.

Unless the system contains strong nonlinear or parameters changing with time is larger,

general as far as possible by linear time-invariant mathematical model describes the

automatic control system. Able to use linear mathematical model to describe system,

known as a linear system. Linear mathematical model such as linear algebraic equation,

differential equation and difference equation, etc. Basic characteristics of this kind of

system, namely the output response characteristics and response characteristics, state

transfer characteristics meet the linear relationship. For the control system, composed of

linear element system as a linear system, its equation of motion for general linear

differential equation. If the various coefficient is constant, it is called a linear time-invariant

system. In the dynamic study, if the system under the action of multiple input single output

is equal to the input output under the action of the sum of the (additivity), and when the

input increase ratio, increase output corresponding to the same multiples (uniformity),

satisfies superposition principle, the system can be seen as a linear system. Does not

meet the superposition principle of the system, is a nonlinear system. Therefore the

response of the nonlinear system of two inputs cannot be calculated separately, the

system analysis will be more difficult, it is difficult to find a generic method. But in the

actual system, there is no such thing as absolute linear system, usually called a linear

system is within the scope of work to ensure the linear. Mathematical model of nonlinear

system is to point to describe system is a nonlinear system of differential equations, the

characteristic is not applied superposition principle.

Use mathematical model to describe system features have limitations. Because

such as the mathematical model and the differential equation, transfer function and so on,

are made with pure mathematics expressions to describe the system characteristics, can

not reflect the components of the system affect the performance of the whole system. So

other measures are required to make up for the defect of mathematical model. Structure is

also known as block diagram, block diagram, etc., which can describe the quantitative

relationship between the variables in the system, but also clearly says the system

components will affect the performance of the system. System structure composed of a

certain functional relation link, and indicate the signal flow of the system block diagram.

Chart is a block diagram with the combination of differential equation (transfer function).

On the one hand, it directly reflects the principle of the whole system structure

(advantages) block diagram, on the other hand, the system has carried on the accurate

quantitative description of each signal lines are sure signal function calculated. On the

research method can not only provide the external information system, but also can

describe the whole system inherent relations among the various components and

dynamic performance under zero initial conditions, but it can't reflect the dynamic

performance of the non-zero conditions. Structure: the most important feature of the

calculating the transfer function of the whole system on the same system, its structure has

the uniqueness; Simplified also has the uniqueness. But to get the system transfer

function is determined only. Chart China indicates the actual components, because box

can represent the combination of multiple components, and even the whole system. In

order to facilitate system analysis and design, often need to the complex structure of

system equivalence transformation, or by varying the system structure simplification,

calculate the total transfer function of system. Therefore, the structure transformation is

the basic content of control theory.

For the control precision of system is not possible for us to design a perfect. It

needs to be corrected so that make the system better. Can make the control performance

of the system satisfies the requirement of control and purposefully components referred to

as the calibration of the control system components to be added or correcting device. It

must be pointed out that not all through design system is through a comprehensive and

correct this step, the higher requirements for control precision and stability of system,

often need to introduce correction device can improve the performance of original system

are fully and compensation. On the other hand, if the original structure of the system itself

is simple and control law and the performance index requirements is not high, again by

adjusting its controller of amplification coefficient can make the system meet the actual

requirements of performance indicators. Control system includes two parts: 1, immutable

parts: executive components and measuring element once selected, the structure and its

parameters is fixed. Part 2, variable: when the system by adjusting the parameters of the

amplifier still cannot satisfy the performance index system, we need to add additional

device to improve the performance of the system, called calibration device. The essence

of the correction through the system is zero pole to change the system performance.

System analysis of task is based on the known system, and the performance of the

system and analyze the performance index and the relationship between the system

parameters, the results of the analysis are rather unique. System of comprehensive and

correction task is according to the performance of the control system should possess, and

the defect of the original system in performance indicators to determine the correction

device (components) of the structure, parameters and connection mode. Logically,

comprehensive and correction of the system is system analysis of inverse problem. At the

same time, meet the system performance index of the calibration device structure,

parameter and the connection is not the only way, to all aspects of performance, cost,

volume, weight, and the feasibility of comprehensive consideration, choose the best

solution.

Ordinary system collected signals are continuous. As the electronic computers into

the field of automatic control, a digital computer control system. The combination of

computer and automatic control, so to speak, make automation technology has entered a

new stage of unprecedented development. But in and out of the digital computer signals

are intermittent digital signal, so must be the original continuous signal into discontinuous

signal, the sampling signal. In a sense, the sampling signal has the nature of human. Such

a control system must be in a place or several pulse signal or digital signal, often called

the sampling control system.

The sampling control system due to its control object itself is continuous signal

components, thus it differ with the discrete system; Again because of its output signal and

the control effect of a given are in digital form, so it is with continuous system. Overall, the

analysis and design of the sampling system is according to the method of discrete system

to deal with, so often comes down to it the discrete system. But strictly speaking, there is a

difference between the two, mainly on the sampling signal and the description of the

discrete signal. Sampling signal (or function) is in the entire history of the real number line

values, is a dimension of its domain, and the discrete signal (or function) is to take positive

integer, real number line is isolated point set its domain. Discrete signal is a kind of

objective existence of signals, such as pulse sequence signal in the radar system, digital

system in which the binary digital and telegraph signals, and the sampling signal is a

continuous signal by the sampler sampling for got to posterity, the periodic visual actual

needs. Actual sampling device is varied, but whatever the specific implementation, the

basic function can be represented with a switch, often referred to as sampling switch. In

continuous signal sampling switch side, sampling switch open and close to certain rule,

the other end of the discrete signal. Every time sampling switch closure time very short,

can be considered to be instantaneous. This switch is closed, at a certain moment can

think the continuous signal values. This sampling switch called ideal sampling switch, said

after the sampling switch are refer to the ideal sampling switch, referred to as sampling

switch. If the switch is to sample interval sampling, it is called a normal sampling, sampling,

sampling, etc. Sampling time interval is called sampling cycle, commonly used T said. If

the sampling time interval is time-varying, then called aperiodic sample, non-uniform

sampling, etc. If the sampling switch sampling interval is random, it is called a random

sampling. A discrete systems tend to have multiple sampling switch. If the system in all

sampling switch sampling at the same time, it is called a synchronous sampling, otherwise

known as non synchronous sampling. If all sampling switch is uniform sampling, the

sampling period, it is called a multi-speed sampling.

Review system, the process of building dynamic equation both starting from the

actual physical system, or from the system block diagram, the differential equation or

transfer function or from the system, in the selection of state variables with a lot of human

conceptions, thus obtained the equation of state of the system is also has a lot of

man-made factors, great randomness, thus leads to different system state equation. So

said system dynamic equation is the only one. Although the same physical system, or the

same block diagram, or the same transfer function produced by the various dynamic

equation, but the independent of the number of state variables are the same, and a variety

of different also has a certain relationship between the dynamic equation, the link is linear

transformation relationship between variables.

Study of the principle of the automatic control can see automation is a promising

and applied widely. It makes people free from heavy work, better enjoy the pleasure of

science and technology bring us. Development of automatic control theory, will be to

improve the social productivity, improve people's living standards, promote the

development of human beings.