Chapter 7 - Risk control and process safety management systems

Abstract

The flow of data obtained using various sensor systems is converted to digital form and can be used to solve problems of ensuring process safety. If it is necessary to use the experience of personnel and experts for effective actions while minimizing risks, then expert assessments are made of the nature of the processes that are described in qualitative categories, i.e., a scale of linguistic variables ordered by a given criterion is used. In this case, risk control systems based on fuzzy logic are used. If there is no description of actions to minimize risk, and there is an array of data accumulated over a certain period of time, including input data and output data, then neural network technology is used to build a risk control system. When solving distributed risk control problems, the problem arises of decomposing the management problem using technological agents that can collect information and exchange it at the request of other agents. This approach allows us to solve the optimization problem of minimizing risk. When building decision support systems, approaches based on fuzzy logic and neural networks are widely used. The use of big data technologies makes it possible, using intelligent methods, to provide training systems with the required information and to configure the fuzzy system.