Data-driven modeling of nano-nose gas sensor arrays

Alstrøm, Tommy Sonne; Larsen, Jan; Nielsen, Charlotte; Larsen, Niels Bent

doi:10.1117/12.850314

Cited by 6 publications

(5 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The literature presents several classification and quantification schemes to detect the gas and the concentration level. Alstrom et al . considered several methods for processing the signal of an array of quartz crystal microbalance (QCM) sensors able to detect six gas species: With regard to the classification, the authors compared ANNs, used jointly with principal component analysis (PCA), with non negative matrix factorization (NMF) and singular value decomposition (SVD), whereas to quantify the concentration level they compared principal component regression (PCR), Gaussian process regression (GPR), and ANNs, since PCR and GPR are linear and nonlinear methods, respectively, which perform well when the data set is limited.…”

Section: Literature Reviewmentioning

confidence: 99%

See 1 more Smart Citation

A Generalized Functional Network for a Classifier-Quantifiers Scheme in a Gas-Sensing System

Gaeta

Loia

Tomasiello

2013

Int. J. Intell. Syst.

View full text Add to dashboard Cite

This paper discusses a new computational scheme based on functional networks and applies it to the problem of classification and quantification of gas species in a mixture. A generalized functional network as a new classifier is proposed to improve the potentialities of the standard functional network classifier. Both methodology and learning algorithm are derived. The performance of this new classifier is examined by using experimental applications. A comparative study with the most common classification algorithms is carried out by showing the high‐quality performance of the proposed classifier. The classifier interacts with some quantifiers, again based on functional networks and finite differences. The scheme of the quantifiers was previously proposed for single gas exposure applications and is here extended to the multigas case. Numerical results show that our approach behaves quite satisfactorily.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

“…As done by other authors (e.g., Refs. ), the classification and the quantification are treated separately, following the scheme depicted in Figure . This section is devoted to introduce the GFN as a new classifier and the FN‐based quantifiers for the multigas detection.…”

Section: The Generalized Functional Network and The Classifier‐quantimentioning

confidence: 99%

A Generalized Functional Network for a Classifier-Quantifiers Scheme in a Gas-Sensing System

Gaeta

Loia

Tomasiello

2013

Int. J. Intell. Syst.

View full text Add to dashboard Cite

show abstract

“…To compute the fitness, apply all the obtained weights from the eqn (2) and the obtained inputs (concentration value and SMAC) in to eqn (1). Here error minimization is the fitness function which is shown below …”

Section: Fitness Computationmentioning

confidence: 99%

“…Especially, the electronic noses are used for process control, quality control in the food and beverage industry, pollution monitoring, and airport security. [1] Unpleasant odor or malodor has been regarded as an indicator of potential risks to human health but not necessarily the direct trigger of health effects. Thus, development of E-nose for medical diagnostics has become one of the important issues in the biomedical engineering research now days [2].…”

Section: Introductionmentioning

confidence: 99%

Formulating a Mathematical Model for the E-Nose Application through Genetic Algorithm (GA)

Nayak¹,

Nayak²,

Deccaraman³

2012

IJCA

View full text Add to dashboard Cite

From a technical and commercial point of view it is found that electronic sensing technologies have emerged significant progresses over the last few decades. The potentiality of reproducing human senses by sensor arrays and pattern recognition systems is termed as electronic sensing. E-Nose provides an industry-specific management resolution for the perpetual and real-time monitoring of environmental odor and air quality resulting in higher profit and improved community relations. The device constitutes arrays of effective and rapid acting chemical sensors, supplemented by patented electronics and software. Chemicals in the air are identified by the sensor arrays, registering complex odor images in real time. By means of wireless connection or lines a permanent record is sent to the computer, where it is detected, computed and alarms for inconsistent events were sent or else it can be indicated by some displacement. Electronic nose instruments are exploited by research and development laboratories, quality control laboratories, process & production department's of environmental protection, all these are done for the detection of volatile organic compounds in air, water and soil samples, and the measurement and comparison of the effects of manufacturing process on products are also determined. In this paper, An E-Nose is proposed to identify the gas component. For this process, the soft computing technique called Genetic algorithm is used. This provides an optimized weight to identify the gas component by means of the input concentration range and SMAC/hr (units in ppm). The intended Technique is evaluated with different training samples and results are produced.

show abstract

“…As the end-user is only interested in a 'yes' or 'no' signal, we have an entire work package dedicated to data analysis. 7 The challenge is to extract the significant signal features from each of the sensors while taking varying environmental conditions into consideration. To this end, machine learning will be employed to minimize the number of false positives and negatives.…”

mentioning

confidence: 99%