MODEL BASAD FAULT DETECTION FOR SAFETY MANAGEMENT OF THE INDUSTRIAL PROCESSES

Detection and diagnosis are important actions offered by the complex automations, that contribute to keeping the exploitation of the process safe. The paper presents a efficient detection solution concepts based on the dependency and the causal connections that exist between the process parameters and proposes techniques and algorithms for the safe evolution of the process. The variables of these connections are specified from the a-priori knowledge on the process. To detect the process faults, the paper proposes a method based on analytical redundance evaluation and on computing residues, using the measurements collected from the process and on the mathematical models associated to the process. This approach is recommended of the modern supervision and optimization systems, for the industrial processes.