Control algorithms for networked control and communication systems.

摘要

NC architectures that utilize processing power at each node allow for the modularization of functionality and standard interfaces for interchangeability and interoperability. With independent sensors and actuators, it is possible then to create a network of such nodes to exchange information like reference input, plant output, control effort, etc., among the control system components.; Sensors and sensor networks are fast gaining popularity amongst research circles. In recent years, sensor research has been undergoing a quiet revolution, promising to have significant impact on a broad range of applications relating to national security, health care, the environment, energy, food safety, and manufacturing. The convergence of the Internet, communications, and information technologies with techniques for miniaturization has placed sensor technology at the threshold of a period of major growth. Large numbers of sensors may be integrated into systems to improve performance and lifetime, and decrease life-cycle costs. Communications networks for sensors need to provide rapid access to information and computing, eliminating the barriers of distance and time for telemedicine, transportation, tracking endangered species, detecting toxic agents, and monitoring the security of civil and engineering infrastructures. Sensor networking research needs to include: enabling networking technologies for distributed wireless and wired sensor networks; scalable and robust architectures; design, automated tasking, and querying techniques; adaptive management and control of sensor nodes; design tradeoffs and performance optimization in resource-constrained sensor networks; design of ultra-low power processing nodes for local information management; investigation of localized versus distributed versus centralized processing of sensor data; common building blocks and interfaces for sensor networking; strategies for using heterogeneous sensor and network nodes to enhance performance and host of other issues.; This dissertation attempts to demonstrate the use of existing network structures for advanced control system implementation and to suggest new network structures and algorithms for rapid deployment of sensors and sensor networks. The use of Local Area Networks and the Internet for implementation of intelligent control algorithms is discussed. A novel hierarchical network structure is proposed, where the nodes could be sensors, actuators or controllers. The network could be either a control network or a data network. Routing algorithms for message routing in the proposed structure are suggested. The proposed structure is analyzed for problems such as link failures. Deadlock issues in the routing algorithms are discussed and a deadlock avoidance algorithm is proposed. Issues related to distributed implementation of the proposed algorithms for routing, failure recovery and deadlock avoidance are discussed and a few distributed algorithms are suggested.