![Main steps of the AIRS algorithm [18].](https://www.researchgate.net/profile/Sulo-Lahdelma/publication/279189348/figure/fig3/AS:294329638768643@1447185062552/Main-steps-of-the-AIRS-algorithm-18_Q320.jpg)
![The test rig [20].](https://www.researchgate.net/profile/Sulo-Lahdelma/publication/279189348/figure/fig4/AS:294329638768644@1447185062615/The-test-rig-20_Q320.jpg)
Content uploaded by Sulo Lahdelma
Author content
All content in this area was uploaded by Sulo Lahdelma on Jun 26, 2015
Content may be subject to copyright.
A preview of the PDF is not available
Artificial Immune Systems in Condition Monitoring
Abstract and Figures
The immune system is a complex, robust, adaptive natural system that defends the body against foreign pathogens. It is capable of categorising all the cells within the body as self or nonself cells. There has been increasing interest in the last few decades to develop engineering tools called Artificial Immune Systems (AIS), which are inspired by the human immune system. The immune system, with its cell diversity and variety of information processing mechanisms, is a cognitive system comparable to the brain in terms of its complexity. Understanding the way how this organ solves its computational tasks can suggest new engineering solutions or new ways of looking at old problems. This paper discusses an ongoing investigation into the usefulness of two applications of AIS algorithms, the AbNET and the AIRIS (Artificial Immune Recognition System), in solving problems in vibration diagnostics. Fault detection based on the specific immunity response algorithms seems to be adequate for characterising the particular nature of normal conditions as well as reacting to new and unexpected anomaly situations.
The results of the investigation clearly indicate that raw data pre-processing and feature extraction are important parts in the development of automatic fault detection systems, and even an advanced AIS classifier cannot compensate for major deficits in feature generation. Generalised moments and norms obtained from higher order derivatives seem to be adequate methods for feature calculation.
Figures - uploaded by Sulo Lahdelma
Author content
All figure content in this area was uploaded by Sulo Lahdelma
Content may be subject to copyright.
ResearchGate has not been able to resolve any citations for this publication.
Advanced signal processing methods combined with automatic fault detection enable reliable condition monitoring for long periods of continuous operation. Any attempt to detect different types of machine faults reliably at an early stage requires the development of improved signal processing methods. Vibration measurements provide a good basis for condition monitoring. In some cases the simple calculation of root-mean-square and peak values obtained from vibration signals are useful features for detecting various faults. Unbalance, misalignment, bent shaft, mechanical looseness and some electrical faults, for example, can be detected using features of displacement and velocity. Higher order derivatives provide additional possibilities for detecting faults that introduce high-frequency vibrations or impacts. New generalised moments and norms related to lp space have been used for diagnosing faults in a roller contact on a rough surface. This paper extends the field of possible applications from roller bearing fault detection to more complex faults situations where different kinds of fault occur simultaneously. In consequence, feature calculation and signal processing have to be adopted and optimized for each fault type on the basis of one measured signal. The features of x(4) indicate well the intact case and the outer race fault. Velocity x(1) is needed for detecting unbalance. This approach also works for the combined case, outer race fault and unbalance. Derivation reduced the effect of noise by amplifying higher frequency components from bearing faults more than the added noise components.
During the last decades there has been an increasing interest to develop engineering tools inspired by the human immune system are called Artificial Immune Systems (AIS). The immune system, with its cell diversity and variety of information processing mechanisms, is a cognitive system of complexity comparable to the brain. Understanding the way this organ solves its computational tasks suggests new engineering solutions or new ways to look at old problems. This paper presents an application of a special AIS algorithm, the AbNET, to a problem of vibration diagnostics. A fault detection based on the specific immunity response algorithms seems to be an adequate device to characterize the particular nature of the normal conditions as well as to react to new and unexpected anomaly situations.
To design a pattern recognition engine based on concepts derived from mammalian immune systems.
A supervised learning system (Immunos-81) was created using software abstractions of T cells, B cells, antibodies, and their interactions. Artificial T cells control the creation of B-cell populations (clones), which compete for recognition of "unknowns." The B-cell clone with the "simple highest avidity" (SHA) or "relative highest avidity" (RHA) is considered to have successfully classified the unknown.
Two standard machine learning data sets, consisting of eight nominal and six continuous variables, were used to test the recognition capabilities of Immunos-81. The first set (Cleveland), consisting of 303 cases of patients with suspected coronary artery disease, was used to perform a ten-way cross-validation. After completing the validation runs, the Cleveland data set was used as a training set prior to presentation of the second data set, consisting of 200 unknown cases.
For cross-validation runs, correct recognition using SHA ranged from a high of 96 percent to a low of 63.2 percent. The average correct classification for all runs was 83.2 percent. Using the RHA metric, 11.2 percent were labeled "too close to determine" and no further attempt was made to classify them. Of the remaining cases, 85.5 percent were correctly classified. When the second data set was presented, correct classification occurred in 73.5 percent of cases when SHA was used and in 80.3 percent of cases when RHA was used.
The immune system offers a viable paradigm for the design of pattern recognition systems. Additional research is required to fully exploit the nuances of immune computation.
This paper presents a new tool for supervised learning, modeled on
resource limited Artificial Immune Systems. A supervised learning
system, it is self-regulatory, efficient, and stable under a wide range
of user-set parameters. Its performance is comparable to
well-established classifiers on a variety of testbeds, including the
iris data, the diabetes classification problem, the ionosphere problem,
and the rock/metal classification problem for mine detection
This paper presents the inception and subsequent revisions of an immune-inspired supervised learning algorithm, Artificial Immune Recognition System (AIRS). It presents the immunological components that inspired the algorithm and describes the initial algorithm in detail. The discussion then moves to revisions of the basic algorithm that remove certain unnecessary complications of the original version. Experimental results for both versions of the algorithm are discussed and these results indicate that the revisions to the algorithm do not sacrifice accuracy while increasing the data reduction capabilities of AIRS.
If the process of condition monitoring of rolling bearings should be executed automati-cally to a large extend, a couple of specifications have to be considered for a reliable and early fault diagnosis. These concern the measuring-technical part of the signal ex-traction and the part of a bearing condition statement on basis of those won signals. Be-cause of being proportional to the impact pulse during damage passing of the rolling elements the acceleration time signal seems to be suitable in terms of understandable measuring variable for diagnostics and should be the basic principle of this approach. Automated analyses can be performed with suitable experience by investigation of the time signal. This can help indicating future damages in case of unchanged surroundings since the time of the investigation. Here it should be tried to formalize the necessary decisions and to integrate them into a program sequence for bearing diagnostics. Such decisions concern the choice of the suitable frequency range, but also the parameters of other methods like the order of higher derivations and settings of a wavelet filter. Be-sides, an essential step of the formalization represents the deliberate use of Data Mining methods. Subsequently the rolling-bearing-diagnostic characteristics are considered with the crea-tion of signal processing, feature generation, classification, evaluation criteria and searching algorithms. The experimental proof occurs with the help of an artificial outer ring damage of known geometry. The geometry-caused impulse is made visible by mea-suring the vibration signals at different contact angle situations, not only to show this context but to illustrate the effects of the signal processing.
The natural immune system is an adaptive distributed pattern recognition system with several functional components designed for recognition, memory acquisition, diversity and self-regulation. In artificial immune systems, some of these characteristics are exploited in order to design computational systems capable of detecting novel patterns or the anomalous behaviour of a system in some sense. Despite their obvious promise in the application of fault diagnostic systems in process engineering, their potential remains largely unexplored in this regard. In this paper, the application of real-valued negative selection algorithms to simulated and real-world systems is considered. These algorithms deal with the self–nonself discrimination problem in immunocomputing, where normal process behaviour is coded as the self and any deviations from normal behaviour is encoded as nonself. The case studies have indicated that immunocomputing based on negative selection can provide competitive options for fault diagnosis in nonlinear process systems, but further work is required on large systems characterized by many variables.
- S Lahdelma
- V Kotila
S Lahdelma and V Kotila, 2005, Complex Derivatives A
Kunnossapito, Vol. 19, No. 4, pp. 39-46.
Machinery Failure Prevention Technologies CM
- S Lahdelma
- E Juuso
S Lahdelma and E Juuso, 2010,
Machinery Failure Prevention Technologies CM 2010 MFPT 2010, Stratford-upon-Avon,
UK, June 22-24. 13 p.