- author = "Shaobo Li and Jianjun Hu",
- title = "Genetic Programming and Creative Design of Mechatronic Systems",
- publisher = "China Machine Press",
- year = "2009",
- email = "hujianju@gmail.com",
- keywords = "genetic algorithms, genetic programming, bond graphs, evolutionary design",
- isbn13 = "9787111254157",
-
URL = "
http://product.dangdang.com/product.aspx?product_id=20470877",
- abstract = "Mechatronic product design is a multi-domain design problem, which is different from the common design of mechanical, electronic or hydraulic systems in isolation. In a multi-domain system, there are many energy conversion behaviors among energy sources of different types. The designer often faces the challenge of proposing an innovative solution that satisfies a multitude of design objectives and constraints as perfectly as practical. Traditionally, innovative design of mechatronic products relies on engineers who explore and accumulate experience over a long period of time, and is incremental in nature. In 2005, the authors were funded by the National Natural Science of China under Grant 50575047 to investigate Creative Design of Mechatronic Systems Based on Genetic Programming. This book, which arises from that study and preceding work by the authors and their colleagues, covers several fields such as sustainable evolutionary algorithms, bond graph theory, modeling and simulation of bond graphs, genetic programming, scalable evolutionary synthesis of dynamic system, and related concepts. An innovative design method for mechatronic products, based on genetic programming and simulation of bond graphs, is presented. In this method, the model of a mechatronic product is expressed as a bond graph and genetic programming is used to search for the best individual (representing the best system) in the design space. A unique advantage of evolutionary synthesis is its capability to produce innovative designs, starting from a nearly blank sheet of paper, rather than relying solely on expert knowledge and thereby, sometimes, not discovering solutions that are radically different from those already existing. Evolutionary synthesis can find an innovative solution by searching in an open-ended design space, subject only to the constraints imposed by the designer based on the real requirements for the system. This book is the first monograph on creative design of mechatronic products using genetic programming and bond graphs published in China. It includes 11 chapters. Chapter 1 introduces the background of the research, methods and techniques for creative design of mechatronic systems, and the objectives of the research; Chapter 2 mainly describes two evolutionary algorithms GA and GP; Chapter 3 describes a sustainable GA; Chapter 4 presents sustainable GP, including a sustainable evolutionary model based on the HFC concept; Chapter 5 describes sustainable SA (Simulated Annealing) based on the HFC model; Chapter 6 introduces basic knowledge such as the theory of bond graphs and system modeling and simulation with bond graphs, including modular modeling of mechatronic systems using bond graphs; Chapter 7 concerns evolutionary synthesis based on bond graphs and GP, describing basic and advanced methods; Chapter 8 introduces GP-based design using Open BEAGLE, which includes basic grammar, a class library, and program design; Chapter 9 is an analysis of an example evolutionary synthesis of an analog circuit; Chapter 10 presents an example of evolutionary synthesis of a vibration absorber; and Chapter 11 concludes the book with an analysis of an example of evolutionary synthesis of a MEMS system.",
-
abstract = "The research in this book spans several fields,
including large-scale scientific computation,
simulation of mechatronic and control systems,
computational intelligence and genetic programming
technology, automated design, and parameter
optimization. By employing genetic programming and
simulation of dynamic systems in a bond graph
expression, this book provides a systematic exposition
of automated design of hybrid mechatronic systems that
include mechanical, electronic and control systems. The
research results in this book include a systematic
method for creative design of modern mechatronic
products. Based on GP, the hybrid topology algorithm,
mixing bond graphs with control block diagrams and
circuit diagrams, has many advantages, such as
searching structures in an open-ended way and
simultaneously searching for the optimal parameters of
the components of the structure. This hybrid searching
method breaks through the restrictions of the classical
parametric optimization of designs based on GA. And it
also implements automated design of complicated
systems. This hybrid search algorithm can evolve some
complicated mechatronic products with designs that are
superior to those done by human designers, in areas
such as circuits, controllers, vibration absorbers,
etc. The design theory, methods, and prototype systems
for evolutionary computation based on GP can also
potentially improve design practice for other types of
engineering systems. The important advances described
in this book give it profound academic value and
application significance.
The content of this book is easy to understand. It can be used as a guide for creative design theory and practice. It also will be a practical toolbox for students in mechanical engineering, computer science and related majors. Of course, it will also be a good choice as teaching reference book for postgraduate students or doctoral students. I enthusiastically recommend it for study by persons interested in studying the automated design of mechatronic systems.
Erik D. Goodman Professor and Design Coordinator, Electrical and Computer Engineering Professor, Mechanical Engineering Michigan State University Vice President for Technology, Red Cedar Technology, Inc. Founding Chair, ACM Special Interest Group on Genetic and Evolutionary Computation",
