Intelligent Control Over Multiple Ovens by Connection Models

Abstract

<p>Intelligent control technology and intelligent information integration technology are the two core technologies of intelligent control system. In this paper, intelligent control and intelligent system are studied from two perspectives: multi-agent integration and neural network intelligent control technology. The multi-agent oriented research aims to provide an architecture suitable for information integration of distributed intelligent control systems. The research of neural network aims to deeply study the theory and technology of neural control, an important branch of intelligent control. The main conclusions of this paper are as follows: In the aspect of multi-layer forward neural network learning algorithm, the weight learning and structure learning methods of multi-layer forward neural network are studied. A heuristic genetic algorithm for sample division and a composite structure learning method are proposed. The sample division heuristic genetic algorithm uses the sample division method to construct the initial genetic population, and combines the subspace division for heuristic search. Several algebraic properties of sample set classes and their training are presented and proved.</p>