Integration of behavioral fault-detection models and an intelligent reactive scheduler

An integrated approach for monitoring large-scale distributed processes and modifying the control plan in real-time, in reaction to deviations from the planned production schedule, is described. The monitoring is accomplished via behavioral models that represent and manipulate both discrete and continuous process dynamics. Reactive scheduling and control is accomplished by combining artificial intelligence and numeric optimization techniques to minimize the cost of recovery. The feasibility of recovery strategies is checked online using the behavioral models in a simulation mode. A prototype systems has been implemented using CROPS5 (Concurrent real-time OPS5) and C++ and runs on the Mach operating systems. The system is intended for online monitoring and reactive scheduling of continuous-caster steel mills.<<ETX>>

[1]  Bruce H. Krogh,et al.  Monitoring Behavioral Evolution for On-Line Fault Detection , 1991 .

[2]  Hideyuki Tokuda,et al.  Real-Time Mach: Towards a Predictable Real-Time System , 1990, USENIX MACH Symposium.

[3]  Lawrence E. Holloway,et al.  Fault detection and diagnosis in manufacturing systems: a behavioral model approach , 1990, [1990] Proceedings. Rensselaer's Second International Conference on Computer Integrated Manufacturing.

[4]  Hideyuki Tokuda,et al.  ARTS: a distributed real-time kernel , 1989, OPSR.

[5]  Jay K. Strosnider,et al.  Reducing problem-solving variance to improve predictability , 1991, CACM.

[6]  J.J. Gertler,et al.  Survey of model-based failure detection and isolation in complex plants , 1988, IEEE Control Systems Magazine.

[7]  L. F. Pau,et al.  Applications of pattern recognition to the diagnosis of equipment failures , 1974, Pattern Recognit..

[8]  Pradeep K. Khosla,et al.  Implementing real-time robotic systems using CHIMERA II , 1990, Proceedings., IEEE International Conference on Robotics and Automation.