HYFA_CAT

Modern society relies heavily on the availability and proper functioning of systems (cars, airplanes, trains, etc.) and processes (water and energy distribution networks, chemical plants, etc.) with complex technological infrastructure. Due to the simultaneous increase in economic demands, along with ecological and safety restrictions that must be met, high safety in these systems and processes has become a top priority in recent years.

However, failures may occur, causing economic losses, breakdowns in operators and machines, and inconveniences for users, etc. In addition, although automation using response controllers has avoided manual system operations, it has not made them immune to malfunction. In general terms, a failure is any change in the behavior of a component of the process/system. Moreover, a closed-loop control system can both amplify and hide failures, making them undetectable until the system/process fails.

One way to increase the reliability of a system/process is by including fault-tolerant mechanisms, which allow systems/processes to operate even after a fault has occurred, potentially progressively. For this reason, there is a growing need and interest, especially in critical applications, to develop control systems that can ensure acceptable operation of faulty systems/processes. These control systems are known as fault-tolerant controllers (FTC).

A failure can be considered as a discrete event that affects the system by altering one of its particular properties (structure, parameters, or both). Then, in return, an active FTC should detect and isolate the fault and, if possible, calculate its magnitude (fault diagnosis) through the FDI module, and adapt the controller to the failure condition, ensuring that the control objective is still met even in the event of a failure (controller redesign). An FTC system can be considered as a hybrid system if we take into account the discrete nature of the fault processes and the control redesign actions, making its analysis and design non-trivial. So far, the hybrid nature has traditionally been rejected in favor of simpler designs, reliable implementation, and systematic testing. Therefore, the analysis and design of fault-tolerant controllers require techniques that are developed for hybrid systems.

Another way to increase system reliability is by using predictive maintenance. The goal of predictive maintenance is to predict the necessary maintenance schedule by analyzing signals that are sensitive to changes and failures in the system. The idea is to provide higher plant availability and economy, estimate the residual useful life of the installations, calculate the accumulated wear (for example, for mechanical / mechatronic systems) and detect incipient faults that are difficult to detect. To achieve these goals, fault diagnosis-based models will be used, capable of dealing with incipient faults and reliability analysis methods.

There are several reasons for using hybrid systems and methods in researching new techniques for diagnosis, fault-tolerant control, and predictive maintenance: