A Negative Selection Immune System Inspired Methodology for Fault Diagnosis of Wind Turbines

Abstract

High operational and maintenance costs represent as major economic constraints in the wind turbine (WT) industry. These concerns have made investigation into fault diagnosis of WT systems an extremely important and active area of research. In this paper, an immune system (IS) inspired methodology for performing fault detection and isolation (FDI) of a WT system is proposed and developed. The proposed scheme is based on a self nonself discrimination paradigm of a biological IS. Specifically, the negative selection mechanism [negative selection algorithm (NSA)] of the human body is utilized. In this paper, a hierarchical bank of NSAs are designed to detect and isolate both individual as well as simultaneously occurring faults common to the WTs. A smoothing moving window filter is then utilized to further improve the reliability and performance of the FDI scheme. Moreover, the performance of our proposed scheme is compared with another state-of-the-art data-driven technique, namely the support vector machines (SVMs) to demonstrate and illustrate the superiority and advantages of our proposed NSA-based FDI scheme. Finally, a nonparametric statistical comparison test is implemented to evaluate our proposed methodology with that of the SVM under various fault severities.

A Data-Driven Fault Detection Framework Using Mahalanobis Distance Based Dynamic Time Warping

“…Fig. 4(d) gives the detection results on the IDV (15). The last row of Table 1 illustrates the FARs of faults on the fault free signal IDV(0), where the proposed method achieves the lowest FAR on IDV(0), which is also shown in Fig.…”

“…In this database, there are 21 different kinds of faults, named as IDV(1), IDV(2), • • • , IDV(21) [32], [33]. IDV (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20) are process faults while IDV(21) is an additional value fault. There are 22 training sets and 22 testing sets in the database, including 21 faulty data files and one normal data file.…”

An Improved Mixture of Probabilistic PCA for Nonlinear Data-Driven Process Monitoring

“…By means of defining a prior distribution over x and of ξ, a related distribution is induced in the data space according to (2). Then, maximum-likelihood approach is utilized to determine the model parameters given a set observational data.…”

“…For the sake of system reliability and operational safety, large-scale industrial systems and applications increasingly demand improved process monitoring technologies, which have been extensively researched in recent decades [1][2][3][4]. However, often quantitative models are difficult to establish due to lack of prior knowledge.…”

Runtime Analysis of Somatic Contiguous Hypermutation Operators in MOEA/D Framework

“…AISs with hypermutation operators have achieved great success in solving complex optimization problems, including single-objective and multi-objective optimization, for example, (Shang et al 2011;Yao et al 2016;Alizadeh, Meskin, and Khorasani 2017;Dudek 2017;Lin et al 2018;Xu et al 2018;Geng et al 2019). Compared to successes in application, the theoretical analysis on understanding the optimization behavior of hypermutation operators is underdeveloped, particularly on optimizing multi-objective optimization problems (MOPs).…”