博文
《控制论与信息论融合之初探》一书 Introduction 节选
|||
Towards Integrating Control and Information Theories: From Information-Theoretic Measures to Control Performance Limitations
控制论与信息论融合之初探：从信息论测度到控制性能极限
Song Fang, Jie Chen, Hideaki Ishii
http://www.springer.com/cn/book/9783319492889
Introduction 节选:
1.1 A Word on Control versus Communication
The attempt to merge control and communication theories dates back to the early works of Hendrik Bode, Claude Shannon, and Norbert Wiener. In [166], observations on similarities and relations between control and communication were made in many places, proclaiming that “the theory of control in engineering, whether human or animal or mechanical, is a chapter in the theory of messages,” and that “it is my book that the physical functioning of the living individual and the operation of some of the newer communication machines are precisely parallel in their analogous attempts to control entropy through feedback.” Indeed, the book [167] is themed on the merging of “control and communication in the animal and the machine.” As Norbert Wiener put it, “(we) become aware of the essential unity of the set of problems centring about communication, control, and statistical mechanics, whether in the machine or living tissue... We have decided to call the entire field of control and communication theory, whether in the machine or the animal, by the same Cybernetics.” On the other hand, as early as in [12], it was observed that “there is an obvious analogy between the problem of smoothing the data to eliminate or reduce the effect of tracking errors and the problem of separating a signal from interfering noise in communications systems,” (see, e.g., [14, 165]) while noting that “this analogy... must of course not be carried too far.”
It was not carried too far. Since then, control and communication have grown into two almost independent fields (see, e.g., [7, 52], respectively), with different focuses and even different mathematical tools (control/estimation theories and information/coding theories, respectively) for modeling, analysis, and design. In a way, control mainly concerns utilization of information, while communication is more about information transmission.
In control, properties of systems are almost always of first priority and should be soundly modeled where possible, based on which stability and performance analysis can then be carried out. On the other hand, in communication, properties of signals are most emphasized, whereas models of communication systems are often not easily obtained, and certainly not describable in ways typical for dynamic processes. For instance, what is the input-output “dynamic model” of the source coding layer/process using Hamming code [52, 62]?
In addition, in control, the element of “time” is always explicit in the analysis of system dynamics, while in communication, at least in classical information theory, “time” is not directly involved. Time delays are tolerable in communication and information theory. Indeed, to achieve a tight bound of channel capacity usually requires coding with arbitrarily long block lengths [34, 52, 142, 143]. On the other hand, time delays severely deteriorate stability and performance as well as robustness of the closed-loop systems in control theory [59], and must be taken good care of in the analysis and design.
In spite of these differences, and just as an old Chinese saying goes, “that which is long divided must unify; that which is long unified must divide,” the interplay of control and communication (information theory) is once again a heated topic in recent years.
参考文献详见 http://link.springer.com/book/10.1007%2F978-3-319-49289-6 的 Back Matter。
https://blog.sciencenet.cn/blog-286797-1023180.html
上一篇：《控制论与信息论融合之初探》一书中所引的 quotes (续)
下一篇：《控制论与信息论融合之初探》一书中名词的翻译
