Development of an Approach to a Biologically Inspired Self-Organizing Control Structure

Abstract

In many fields of the industry, like biology, medicine, power generation, chemical engineering, etc., where dealing with complex systems, the design of the control structure is a key factor. Designing robust and fault-tolerant controllers for these sophisticated processes has always been a challenging task. This paper discusses the design of a fault tolerant multi-agent system using fractional order controllers for level control in a three-tank system, which is very close to the hydrodynamic evolution of the isotope separation columns cascade, the final goal of the research. The fractional order controllers are designed using the Particle Swarm Optimization (PSO) method. The proposed architecture is implemented in JADE (Java Agent Development Environment) . The control structure is tested with good results on the laboratory equipment simulating different faults in the system.