1、Boiler is a work under pressure energy conversion devices. Its role is to effectively put fuel chemical energy is converted into heat, or thermal energy through the appropriate equipment production and converted to other forms of energy needed in life, for a long time and living in industrial produc
2、tion have played an extremely important role.The boiler is one of the main production equipment of thermal power plants. Many different types of boilers, by type of fuel, the most used are coal, oil and gas boilers. Most boilers currently in manual control. Manual control not only increase the labor
3、 intensity of workers, and the consumption of steam production depends mainly on the ratio of fuel technology and operations workers sense of responsibility, it is difficult to make the boiler is in good working condition, increased fuel consumption, reduces the thermal efficiency, increased environ
4、mental pollution.Since the 1970s the worlds energy crisis has accelerated the development of various energy-saving technologies, economic combustion boiler has become a very important issue. Our scientific and technical personnel, computer controlled boiler made of extensive research, has made grati
5、fying achievements. Due to the use of computer control, the boiler can be improved in good running condition, the thermal efficiency of the boiler, steam production and quality, to the city, so that the secondary product quality, quantity and safety are to improve economic efficiency is very impress
6、ive. Because of this, the boiler microcomputer control, has become an important task of transformation.Traditional control methods include classical control and modern control, is based on the precise control of the controlled object model, lack of flexibility and adaptability, suitable to solve the
7、 invariance problem is relatively simple control of linear time.The traditional way of intelligent control is a control or a control system, if it can effectively overcome the controlled object （process） and the environment with a high degree of complexity and uncertainty, and to achieve the desired
8、 goal, then said this control method for the intelligent control, saying such a control system for the intelligent control problems. In the production practice, complex control problems can be experienced and skilled operators control theory combined to solve the resulting intelligent control. The m
9、ethod for controlling intelligent control theory and artificial intelligence techniques to combine a flexible, object other control methods to adapt the complexity and uncertainty.Intelligent control is the control theory advanced stage of development , it is mainly used to solve the complex system
10、of control problems with traditional control methods are difficult to solve. Intelligent control study has the following features:（ 1 ） uncertainty in the model . Intelligent control adapted to control the object of uncertainty , the uncertainties include two meanings : First , the model is unknown
11、or poorly understood ; Second, the structure and parameters of the model may vary over a wide range .（ 2 ） a high degree of non-linearity . Intelligent control methods can solve the problem of nonlinear control systems .（ 3 ） complex mission requirements.Intelligent control has three important branc
12、h , namely fuzzy control, neural network control and genetic algorithms.（ 1 ） Fuzzy ControlVarious conventional methods are conventional control is based on the mathematical model of the controlled object on the basis of precise , but as the complexity of the system increases, the system will be dif
13、ficult to establish a precise mathematical model .In engineering practice , it was found that a complex control system operator by a person with a wealth of practical experience satisfactory control effect. This shows that, if the human brain through the analog controller design thinking can achieve
14、 control of complex systems , resulting fuzzy control .Development of fuzzy control can be divided into three phases: 1965-1974 , the first phase of the development of fuzzy control , fuzzy math development and formation stage ; 1974-1979 , the fuzzy control of the second phase of development , resu
15、lting in a simple fuzzy controller ; 1979 to now , fuzzy control third stage of development, namely high performance fuzzy control stage.（ 2 ） neural network controlNeural networks have decades of history. In 1943 , McCulloch and Pitts neuron proposed mathematical model ; 1950-1980 for the formation
16、 of neural networks , with a small amount of the outcome ; since 1980 for the development of neural networks .The neural network is introduced to form a control field neural network control . Neural network control from the physiological mechanism of the human brain simulation system structure is si
17、mple an emerging intelligent control methods. Neural network has the characteristics of a parallel mechanism , pattern recognition , memory and learning ability , it can fully approximate any complex nonlinear systems , able to learn and adapt to the dynamic characteristics of uncertain systems with
18、 robustness and fault tolerance . Neural Network Control in the control field has a wide range of applications .（ 3 ） genetic algorithmThe genetic algorithm is an important branch of artificial intelligence , is the search algorithm based on the principle of natural selection and genetics , and is b
19、ased on Darwins theory of evolution , simulated life on a computer evolution mechanism and developed a science . Currently Genetic algorithms have been widely used in many practical problems , new ideas and new methods become hard to solve highly complex problems.Genetic algorithms can be used to op
20、timize learning and fuzzy neural network parameters and weight control rules in the field of intelligent control of a wide range of applications.Intelligent control is the latest stage of development of automatic control is mainly used to solve complex control problems difficult to solve traditional
21、 control .As an advanced stage of development of intelligent control , intelligent control mainly to solve the control problem that complex systems are difficult to resolve with traditional control methods , including the intelligent robot control, computer integrated manufacturing systems, industri
22、al process control , aerospace control , socio-economic management system, transportation systems, environmental protection and energy systems.（ 1 ） Application in Robot Control : intelligent robot research currently the hot topic. Currently used in over 90% of the industrial robot is not intelligen
23、t . With the rapid development of robot technology , with varying degrees of intelligence requires a variety of robots.（ 2 ） application in process control : process control and automation technology is an important aspect . Intelligent control in process control has been widely used .Since the 1960
24、s , due to the technology, computer technology and artificial intelligence development, research scholars in the control community self-organizing , self-learning control based on self-learning ability in order to improve control systems , began to pay attention to artificial intelligence techniques
25、 and methods applied to the control .In recent years, neural networks, fuzzy math development , expert systems , evolution and other disciplines to intelligent control into a tremendous vitality , thereby producing a variety of intelligent control methods.火电厂锅炉汽包水位的智能控制锅炉是一种承受一定工作压力的能量转换设备。其作用就是有效地把
26、燃料中的化学能转换为热能，或再通过相应设备将热能转化为其它生产和生活中所需的能量形式，长期以来在工业生产和居民生活中都扮演着极其重要的角色。锅炉是火电厂的主要生产设备之一。锅炉种类很多，按燃料分类，用的最多的有燃煤、燃油和燃气锅炉。多数锅炉目前处于人工控制状态。人工控制不仅加大了操作工人的劳动强度，而且燃料的消耗量与蒸汽生产量的比值主要取决于操作工人的技术水平和工作责任心，难以使锅炉处于良好的工况，增加了燃料消耗，降低了锅炉热效率，增加了环境污染。 自70年代世界发生能源危机以来，加速了各种节能技术的发展，锅炉的经济燃烧也成了十分重要的课题。我国的科技人员，对锅炉的微机控制作了大量研究，取得了
27、可喜的成果。由于采用微机控制后，锅炉可以在良好的工况下运行，锅炉的热效率，蒸汽生产量和品质，都会的到改善，从而使二次产品质量，数量和安全性都的到提高，经济效益是十分可观的。正因为如此，锅炉的微机控制，已成为重要的技术改造任务。传统控制方法包括经典控制和现代控制，是基于被控对象精确模型的控制方式，缺乏灵活性和应变能力，适于解决线性时不变性等相对简单的控制问题。传统的智能控制就是一种控制方式或一种控制系统，如果它能够有效地克服被控制对象（过程）和环境所具有的高度复杂性和不确定性，并且能够达到所期望的目标，那么称这种控制方式为智能控制，称这样的控制系统为智能控制问题。在生产实践中，复杂控制问题可通过
28、熟练操作人员的经验和控制理论相结合去解决，由此产生了智能控制。智能控制将控制理论的方法和人工智能技术灵活地结合起来，其他控制方法适应对象的复杂性和不确定性。智能控制是控制理论发展的高级阶段，它主要用来解决那些用传统控制方法难以解决的复杂系统的控制问题。智能控制研究对象具备以下一些特点：（1） 不确定性的模型。智能控制适合于不确定性对象的控制，其不确定性包括两层意思：一是，模型未知或知之甚少；二是，模型的结构和参数可能在很大范围内变化。（2） 高度的非线性化。采用智能控制方法可以较好地解决非线性系统的控制问题。（3） 复杂的任务要求。智能控制有3个重要分支，分别为模糊控制，神经网络控制以及遗传算
29、法。（1） 模糊控制以往的各种传统控制方法均是建立在被控对象精确数学模型的基础上，然而，随着系统复杂程度的提高，将难以建立系统的精确数学模型。在工程实践中，人们发现，一个复杂的控制系统可由一个人操作人员凭借着丰富的实践经验得到满意的控制效果。这说明，如果通过模拟人脑的思维方法设计控制器，可实现复杂系统的控制，由此产生了模糊控制。模糊控制的发展可分为3个阶段： 19651974年，模糊控制发展的第一阶段，模糊数学发展和形成阶段； 19741979年，模糊控制发展的第二阶段，产生了简单的模糊控制器； 1979年至现在，模糊控制发展的第三阶段，即高性能模糊控制阶段。（2） 神经网络控制神经网络的研究
30、已经有几十年的历史了。1943年，McCulloch和Pitts提出了神经元数学模型；19501980年为神经网络的形成期，有了少量成果；1980年以后为神经网络的发展期。将神经网络引入控制领域就形成了神经网络控制。神经网络控制是从机理上对人脑生理系统进行简单结构模拟的一种新兴智能控制方法。神经网络具有并行机制、模式识别、记忆和自学能力的特点，它能充分逼近任意复杂的非线性系统，能够学习与适应不确定系统的动态特性，有很强的鲁棒性和容错性。神经网络控制在控制领域有着广泛的应用。（3） 遗传算法遗传算法是人工智能的一个重要分支，是基于自然选择和基因遗传学原理的搜索算法，是基于达尔文进化论，在计算机上
31、模拟生命进化论机制而发展起来的一门科学。目前遗传算法已经被广泛应用于许多实际问题，成为用力解决高度复杂问题的新思路和新方法。遗传算法可用于模糊控制规则的优化及神经网络参数及权值的学习，在智能控制领域有广泛的应用。智能控制是自动控制发展的最新阶段，主要用于解决传统控制难以解决的复杂的控制问题。作为智能控制发展的高级阶段，智能控制主要解决那些用传统控制方法难以解决的复杂系统的控制问题，其中包括智能机器人控制、计算机集成制造系统、工业过程控制、航空航天控制、社会经济管理系统、交通运输系统、环保及能源系统等。（1） 在机器人控制中的应用：智能机器人是目前机器人研究中的热门课题。目前工业上用的90以上的机器人都不具有智能。随着机器人技术的迅速发展，需要各种具有不同程度智能的机器人。（2） 在过程控制中的应用：过程控制是自动化技术的一个及其重要的方面。智能控制在过程控制中有着广泛的应用。从20世纪60年代起，由于空间技术、计算机技术及人工智能的发展，控制界学者在研究自组织、自学习控制的基础上，为了提高控制系统的自学能力，开始注意将人工智能技术与方法应用于控制中。近年来神经网络、模糊数学、专家系统、进化论等各门学科的发展给智能控制注入了巨大的活力，由此产生了各种智能控制方法。
