- author = "Douglas {Mota Dias} and Marco Aur\'{e}lio C. Pacheco and Jos\'{e} F. M. Amaral",
- title = "Automatic Synthesis of Microcontroller Assembly Code Through Linear Genetic Programming",
- year = "2006",
- booktitle = "Genetic Systems Programming: Theory and Experiences",
- pages = "193--227",
- volume = "13",
- series = "Studies in Computational Intelligence",
- editor = "Nadia Nedjah and Ajith Abraham and Luiza {de Macedo Mourelle}",
- publisher = "Springer",
- address = "Germany",
- keywords = "genetic algorithms, genetic programming, control",
- ISBN = "3-540-29849-5",
-
DOI = "
doi:10.1007/3-540-32498-4_9",
- abstract = "This chapter considers the application of linear genetic programming in the automatic synthesis of microcontroller assembly language programs that implement strategies for time optimal or sub-optimal control of the system to be controlled, based on mathematical modelling through dynamic equations. One of the difficulties presented by the conventional design of optimal control systems lies in the fact that solutions to problems of this type normally involve a highly non-linear function of the system s state variables. As a result, it is often not possible to find an exact mathematical solution. As for the implementation of the controller, there arises the difficulty of programming the microcontroller manually in order to execute the desired control. The research that has been done in the area of automatic synthesis of assembly language programs for microcontrollers through genetic programming is surveyed in this chapter and a novel methodology in which assembly language programs are automatically synthesised, based on mathematical modeling through dynamic plant equations, is introduced. The methodology is evaluated in two case studies: the cart-centering problem and the inverted pendulum problem. The control performance of the synthesised programs is compared with that of the systems obtained by means of a tree-based genetic programming method. The synthesised programs proved to perform at least as well, but they had the additional advantage of supplying the solution already in the final format of the implementation platform selected: a microcontroller.",
-
notes = "http://www.springer.com/sgw/cda/frontpage/0,11855,5-146-22-92733168-0,00.html
Not Turing complete; no loops limited use of memory???
Subset 22 (of 75) PIC18F452 implemented. MPLAB simulator.",
- size = "35 pages",
