Distributed Control with Networked Embedded Systems

1
Distributed Control with Networked Embedded
Systems

• Objectives
• Implementation of distributed, cooperative
  control using a networked embedded system.
• Of particular interest are the observation and
  control of distributed parameter systems
  vibrations, acoustics, environmental etc.
• Critical issues
• Maximum sampling frequency is limited by network
  communication
• Efficient distributed control design
• Scalable control laws
• Performance limited by partial system information

2
Distributed Control with Networked Embedded
Systems

• Computational Platform
• Prometheus PC/104 from Diamond Systems
  Corporation
• Linux or QNX OS
• Experimental Platforms
• Smart Structure Experiments of increasing
  complexity
• Simply supported beam for basic algorithm
  development
• Simply supported plate for more complex dynamics
• Launch vehicle fairing for unknown dynamics and
  distributed system identification

3
Distributed Control with Networked Embedded
Systems

• Distributed Adaptive Algorithm Development
• Distributed adaptive system identification
• Distributed adaptive controller
• Scalable control system design
• Account for unknown dynamics and time-varying
  environment

• Results
• Minimum distributed controller processing time of
  around 1ms permits maximum sampling rate of 600Hz
• Scalable control laws that approach global
  control performance even with limited sensor
  information