Automatic control is relatively artificial control terms ，which refers to correlate with no people directly involved in conditions, the additional equipments or devices to make the machine, equipment or the production process of a certain job of state or parameters automatically set to run the rule.
Automatic control technology research will benefit mankind from complicated, risky, tedious work environment free of control and greatly improve the efficiency. Automatic control is a branch of engineering science. It involves using a feedback of dynamic system of the principle of automatic influence, in order to make the output value close with that we set value. From the theory of methods, the mathematical theory is a foundation of Automatic control. As we known today, as automatic control is in the middle of the twentieth century from the control of a branch. The conclusion is based by Norbert wiener, Rudolf kalman proposes.
1. The first generation process Control System(PCS)
The first generation process Control System is based on 5-13 psi Pneumaticrebefo signal standard ( Pneumatic Control System) PCS before 150 years ,which includes simple on-site operation mode, control theory preliminary form. There does not has been the concept of the control room.
2. The second generation process Control System(ACS)
The second generation process Control System (ACS or Analog Control System) Analog to produce stats are based on 0-10 mA or 4-20 mA's current Analog signals, the obvious progress of the Control System is that to rule the whole firmly automatic Control field in the whole 25 years. It represents the arrival of the era of electrical automatic control. Control theory has the significant development, the establishment of the three big cybernetics laid the foundation of modern control; The establishment of control function, control room separate model has been used today.
3. The third generation of process Control System(CCS)
The third generation of process Control System, Computer Control System
(CCS) began in the 70’s ,the application of digital Computer has a great technical advantage, people in measurement, simulation and logic Control field, which
pioneered the use of the third generation process Control System, Computer Control System (CCS). This is called the third generation process control system is automatic control a revolution in the field, it give full play to the computer specialty, so it is widely acknowledged that computer can do all things, naturally produced is called \"the centralized control\" of the central control computer system, it should be pointed out that the signal transmission system is still the most used with 4-20 mA analog signals, but soon after that, as people focus and reliability of the control aspects of the problem, the out of control risk also focused on, a little wrong will make the whole system to paralysis. So it was quickly developed into the distributed control system (DCS).
4. The fourth generation process Control System(DCS)
With the rapid development of semiconductor manufacturing technology, the
microprocessor to the widespread use of computer technology greatly increased, the reliability of the currently used is the fourth generation process Control System (DCS, or Distributed digital Control System), it is the main features of the whole Control System that there is no longer only a computer, it is a Control System by a computer and some intelligent instruments and intelligent components comprise. So the way of distributed control became the most important characteristic. Another important development exception is the signal transmission among of them are not based on 4-20 mA analog signals, and gradually digital signal to replace analog signals.
5. The fifth generation process Control System (FCS)
The development of FCS from the DCS, like DCS from CCS over the development as, there is a qualitative leap, that is \"Distributed control\" developed to \" focused control\". The way of data transmission is \"bus\" way. The so-called field-bus is intelligent measurement and control equipment conect all the digital, two-way transmission, with many node of the structure of the branch communications link. Say simply traditional control is a loop, and FCS technology is modules such as controller, actuators, detector etc on a bus to hang up realize communication, of course, transmission is the digital signal. The main bus has Profibus, LonWorks, etc. But the real with DCS FCS difference is to have a more FCS wide development space. Because the traditional DCS technical level while continuously improved, but the
most low-end only communication network to the control station level, the control station and the field measurement instrument, the contact between the actuators are still used one-to-one transmission of 4-20 mA analog signals, high cost, low efficiency, maintenance difficulties, can't play the field instruments, to realize the potential of the intelligent field device the work state of comprehensive monitoring and deep management..
In the modern science and technology in many fields, automatic control
technology is playing a more and more important role. Automatic control theory is the study of the automatic control of common laws technology science. Its initial development stage, is based on the theory of the feedback of automatic adjustment principle, mainly for industrial control, during world war two, in order to design and manufacture the plane and Marine autopilot, gun positioning system, radar tracking system based on feedback and other military equipment, the principle of further promote and perfect the development of the theory of automatic control. After the war, in order to form the full automatic control theory system, which is the transfer function is the foundation of classical control theory, it mainly studies single input and single output, the linear system analysis and set constant design problem.
Automatic control of the development, from the start of the happen to form a
control theory, the whole that process. Speak Automatic control is refers to such feedback control system, this is a controller object with a control of the control object, the output signal get it back, after come back for measuring with the signal are compared. According to the error tell controller, which is within the machine work. Let the controller to complete the control effect, make the deviation eliminate or make the controlled objects output tracking what I need to be requirements of the signal. The controlled objects output in general is a physical quantities, for example say me to control a machine speed, is need to come out, to measure the speed control.
Saying to hear I have to mention is the \"engineering cybernetics qian xuesen\".
Qian xuesen, written in 1954, when his book in the United States, we wrote the book here also can't get in English, but the former Soviet union very seriously, the former Soviet union immediately translate it into Russian. We see of the former Soviet union was in 1956 put his Russian, translation come out, we see at that time is the Russian
version of, this is in about twenty s the formation process of the main process, the experience is concluded.
The robust leaves (Lurie) in the former Soviet union in 1944 about him out to everybody now make nonlinear may know, out to a robust leaves in question, this problem has been not solve, he later to write a book, is to keep the problem. The problem, start to when? I said something about this here two style, British and American of just everybody has to come out, make the person is engineering in make control, the former Soviet union is the application with mechanics in make home mathematician control, so two played the role is not the same. His 1951 book at that time is very hard to understand, hard to read. This work, he brought out the robust, Lurie), (until 1960, someone out to solve a solution, we may know, is that the absolute Popov) Popov (stability. Then out of the stability and is robust to solve this problem, Lurie (in). That is the book was 20 s, twenty century in the late '40 s, 50 s, some of the work has been influence of the twentieth century, but also affects the 60 s to present some nonlinear theory, is his work for the foundation. The twentieth century is in the 40 s, front of the development of the technology is a process, slowly forms a theory. I'm used to in such a schedule to said 50 s, most of the time, call classical control theory; 60 s call the state space method, is actually state space method, but you know, at that time, the name of the called it the modern control theory; Then in the 70 s is the modern the frequency domain method, such a process.
The next now is to explain the state space method, finished that modern control theory. The state space method who first brought out? The third book is just of the former Soviet union, these scholars. They make applied mathematics and mechanics, they never get the is used is the state space method. In 1960, it introduced to kalman English world, so the world it is spending in English, that means the country want English country, because it was all don't know, kalman is a Slavic name. In 1960, he, he put the state space method introduced to the United States. But add people to this hype, modern control theory to hype seem very much as gods, was also some people expected is relatively large. This is the development of the ten years later, he found too, like the expected only so! Some problem you didn't also solved. So the time, and some people back to the frequency domain method, is the earliest 50 s is the
frequency domain method, 60 s, 70 s state space and back to the frequency domain method.
Of course, this is the spiral, this time the frequency domain method and then add
a name, so that the modern the frequency domain method. Think frequency domain to consider the design problems or more appropriate, consider some of the design requirements, I just have this frequency domain method. Just in the modern the frequency domain method development on this momentum, 1981, and wrote you this no said robust, we now everyone make control theory know to robust design. Say you this modern the frequency domain method not robustness, when people do not believe, after the 80 s, the arguments that dispute slowly forms. By 1991, is now of course is that you may be someone the term does not necessarily unification, someone called it the modern control theory, Postmodern after keep control. We now have to go back to see, why do you say that no robustness? One will say, we from the many variable system, it is actually much more variable system into the system. Many variables don't appropriate, there is a lot of input, output a. Many avariable a problem, called the coupling, is between input and output coupling each other. Control, intuitive request was for decoupling control, decoupling control later! Is this the 1 to 1 can form output feedback system, the 2! with the 2 of output can form feedback systems, and the design will be easier.
自动控制
自动控制（automatic control）是相对人工控制概念而言的，指的是在没有人直接参与的情况下，利用外加的设备或装置，使机器、设备或生产过程的某个工作状态或参数自动地按照预定的规律运行。
自动控制技术的研究有利于将人类从复杂、危险、繁琐的劳动环境中解放出来并大大提高控制效率。 自动控制是工程科学的一个分支。它涉及利用反馈原理的对动态系统的自动影响，以使得输出值接近我们想要的值。从方法的角度看，它以数学的系统理论为基础。我们今天称作自动控制的是二十世纪中叶产生的控制论的一个分支。 基础的结论是由诺伯特·维纳，鲁道夫·卡尔曼提出的。
1. 第一代过程控制体系（PCS）
150多年前,第一代过程控制体系是基于5－13psi的气动信号标准（气动控制系统PCS，Pneumatic Control System）。简单的就地操作模式，控制理论初
步形成，尚未有控制室的概念。
2. 第二代过程控制体系(ACS)
第二代过程控制体系（模拟式或ACS,Analog Control System）是基于0－10mA或4－20mA的电流模拟信号，这一明显的进步，在整整25年内牢牢地统治了整个自动控制领域。它表征了电气自动控制时代的到来。控制理论有了重大发展，三大控制论的确立奠定了现代控制的基础；控制室的设立，控制功能分离的模式一直沿用至今。
3. 第三代过程控制体系(CCS)
70年代开始了数字计算机的应用，产生了巨大的技术优势，人们在测量，模拟和逻辑控制领域率先使用，从而产生了第三代过程控制体系（CCS，Computer Control System）。这个被称为第三代过程控制体系是自动控制领域的一次革命，它充分发挥了计算机的特长，于是人们普遍认为计算机能做好一切事情，自然而然地产生了被称为“集中控制”的中央控制计算机系统，需要指出的是系统的信号传输系统依然是大部分沿用4－20mA的模拟信号，但是时隔不久人们发现，随着控制的集中和可靠性方面的问题，失控的危险也集中了，稍有不慎就会使整个系统瘫痪。所以它很快被发展成分布式控制系统（DCS）。
4. 第四代过程控制体系(DCS)
随着半导体制造技术的飞速发展，微处理器的普遍使用，计算机技术可靠性的大幅度增加，目前普遍使用的是第四代过程控制体系（DCS，或分布式数字控制系统），它主要特点是整个控制系统不再是仅仅具有一台计算机，而是由几台计算机和一些智能仪表和智能部件构成一个了控制系统。于是分散控制成了最主要的特征。除外另一个重要的发展是它们之间的信号传递也不仅仅依赖于4－20mA的模拟信号，而逐渐地以数字信号来取代模拟信号。
5. 第五代过程控制体系(FCS)
FCS是从DCS发展而来，就像DCS从CCS发展过来一样，有了质的飞跃。“分散控制”发展到“现场控制”；数据的传输采用“总线”方式。所谓现场总线就是连接智能测量与控制设备的全数字式、双向传输、具有多节点分支结构的通信链路。简单地说传统的控制是一条回路，而FCS技术是各个模块如控制器、执行器、检测器等挂在一条总线上来实现通信，当然传输的也就是数字信号。主要的总线有Profibus，LonWorks等。但是FCS与DCS的真正的区别在于FCS有更广阔的发展空间。由于传统的DCS的技术水平虽然在不断提高，但通信网络最底端
只达到现场控制站一级，现场控制站与现场检测仪表、执行器之间的联系仍采用一对一传输的4-20mA模拟信号，成本高，效率低，维护困难，无法发挥现场仪表智能化的潜力，实现对现场设备工作状态的全面监控和深层次管理。
在现代科学技术的众多领域中，自动控制技术起着越来越重要的作用。自动控制理论是研究自动控制共同规律的技术科学。它的发展初期，是以反馈理论为基础的自动调节原理，主要用于工业控制，二战期间为了设计和制造飞机及船用自动驾驶仪，火炮定位系统，雷达跟踪系统以及其他基于反馈原理的军用设备，进一步促进并完善了自动控制理论的发展。到战后，以形成完整的自动控制理论体系，这就是以传递函数为基础的经典控制理论，它主要研究单输入-单输出，线形定常数系统的分析和设计问题。
从开始阶段的发生到形成一个控制理论，下面我们讲述自动控制系统的组成。自动控制系统就是指这样的反馈控制系统，系统由一个控制器一个控制对象组成，把这个控制对象的输出信号检测回来，检测所得的反馈信号跟所要求的信号进行比较。相比较后会产生一个误差，然后把这误差信号输入控制器，这就是机器内部的工作。让控制器根据相应的控制算法完成控制作用，使得这个偏差消除或者说使得控制对象的输出跟踪我们所设定的信号。控制对象的输出量一般来说都是一个物理量，比如说我控制一个机器的转速，就是需要把速度检测出来，从而才能进行控制。
说起控制理论在这里我们不得不提的是钱学森的《工程控制论》。这本书是钱学森学者1954年在美国写的书，但当时我们没有把那本书翻译为英文版。然而前苏联很重视，前苏联马上把它翻译为俄文。我们看到的最初版本是前苏联在1956年翻译的俄文版。书中的理论主要是在大概二十世纪50年代自动控制理论形成过程里边的几个主要的过程，总结出了很多经验。
还有一本书是鲁里叶（Lurie）的一个关于非线性的经典著作，这本著作是1951年出版的。1944年，鲁里叶（Lurie）在前苏联提出来了非线性问题，这个问题一直解决不了，他后来写成书了，仍然保留着这问题。这个问题，是到什么时候解决的呢？在这里我简要说两种不同的解决风格，英美的刚才大家已经听出来了，都是搞工程的人在搞控制，前苏联是应用数学家跟力学家在搞控制，所以两个起的作用都不一样。鲁里叶这本书理论1951年提出时是很难读懂的。这个鲁里叶（Lurie）问题，一直到1960年才有人解决。大家可能知道的，人们提出来的一个解决方案就是波波夫（Popov）的绝对稳定性。后来就提出来超稳定性，
主要是针对于这个鲁里叶（Lurie）问题的解决。就是说这本书是当时二十世纪40年代后期，50年代初的一些工作，却一直影响到了二十世纪60年代，而且目前的一些非线性的理论工作，都是以他的非线性理论为基础。二十世纪四十年代后逐渐形成一个系统的理论。我习惯上用这样一个时间表来表示。50年代的时候，一般都叫经典控制理论； 60年代叫状态空间法，实际上就是状态空间方法，而当时的名称，则是现代控制理论；后来70年代是现代频域法。总体来说是这么一个过程。
接下来现在就是要说明一下状态空间法，即我们最终称之的现代控制理论。这个状态空间法谁先提出来的呢？是上文提到的第三本书，即前苏联学者们研究的理论。他们在研究应用数学、力学的过程中一直用的就是状态空间法。1960年卡尔曼把它介绍到英语世界，这个世界用英语来说是English speaking country ，就是说英语的国家，因为本来大家都不知道，卡尔曼是个斯拉夫名字。他在1960年把状态空间法介绍到美国。但是由于人为的炒作，就把这个现代控制理论炒作得好像非常神一样，当时也有些人对此理论寄予了厚望。然而此控制理论发展了十年以后，他们就发现期望过大了，事实表明不过如此而已！有些问题仍然没有得到解决。所以那个时候，又有人回到频域法，就是最早50年代的频域法，60年代状态空间，70年代又回到频域法。
当然这个是螺旋上升的，就在这个时候的频域法就加了个修饰语，叫现代频域法。人们发现频域里边来考虑设计问题和设计要求还是比较恰当的。正好在现代频域法发展的这个势头上的时候，1981年有人写文章说你这个没有鲁里叶特性，而我们现在大家研究控制理论的都知道要考虑其鲁里叶特性设计。说你这个现代频域法没有鲁里叶特性，当时人家肯定不信、不服，经过80年代的论证，争议慢慢形成。直到1991年，有人觉得是可能这个术语不一定统一，有人把它叫做现代后控制理论，Postmodern control theory。我们现在就回过来看看，为什么说这个没有鲁里叶特性？对于这个问题的探讨我们从多变量系统来说，多变量系统实际上是多入多出系统。其实称之为多变量并不太恰当，其实质是输入有好多个，同时输出有好多个。多变量里有一个很突出的问题，叫做耦合，就是输入输出之间互相耦合。控制的时候，直观的要求就是要解耦控制，解耦控制以后就是这个变量1跟输出1可以组成反馈系统，这个变量2跟输出2可以组成反馈系统，这样设计控制系统的时候就比较容易了。
因篇幅问题不能全部显示，请点此查看更多更全内容