A comparison study of basic data-driven fault diagnosis and process monitoring methods on the benchmark Tennessee Eastman process

Fault diagnosis of downhole drilling incidents using adaptive observers and statistical change detection

“…Fault diagnosis methods can be divided into model-based methods using mathematical models of the system [19,20,21], and data-driven methods that only are dependent on measurements, which can be beneficial for large systems [22,23,24]. Data-driven methods for multivariate statistical fault diagnosis are presented and discussed in [22,23,25].…”

“…Data-driven methods for multivariate statistical fault diagnosis are presented and discussed in [22,23,25]. This paper presents a model-based multivariate statistical fault diagnosis method to detect and isolate the possible incidents.…”

The application of big data-driven prognostic and health management on complex equipment based on internet of things

“…By this new technology and application, Great savings in equipment maintenance have been provided, delays in equipment maintenance have been reducing, what's more, operational safety has been improved. The understanding of the current and the prediction of the future health condition of equipment components may provide guidelines to the maintenance activities and spare components logistics [1].…”

Improved Many Machine Excitation Nonlinear Robust Controller Design

“…In order to obtain faster convergence speed, the K class function is referred in the selection of virtual stabilization function. When time tends to go infinity, and the controller gain also tends to be traditional design of the adaptive backstepping method [2], in this way, the transient response speed would be improved significantly and the controller gain doesn't need to increase too much, The adjustable parameters in the K class function keeps the balance between the controller gain and transient response.…”