Filtern
Dokumenttyp
- Zeitschriftenartikel (8)
- Beitrag zu einem Tagungsband (4)
- Vortrag (3)
- Forschungsdatensatz (2)
- Buchkapitel (1)
- Posterpräsentation (1)
Sprache
- Englisch (19) (entfernen)
Schlagworte
- Artificial neural networks (7)
- Optical time domain reflectometry (5)
- Rayleigh scattering (5)
- Optical fiber sensor (4)
- Automation (3)
- Fiber optics sensors (3)
- Humidity (3)
- Online NMR spectroscopy (3)
- Polymer optical fibre (3)
- Process industry (3)
Organisationseinheit der BAM
- 8.6 Faseroptische Sensorik (19) (entfernen)
Eingeladener Vortrag
- nein (3)
Mass Spectrometry (MS) and Nuclear Magnetic Resonance Spectroscopy (NMR) are critical components of every industrial chemical process as they provide information on the concentrations of individual compounds and by-products. These processes are carried out manually and by a specialist, which takes a substantial amount of time and prevents their utilization for real-time closed-loop process control. This paper presents recent advances from two projects that use Artificial Neural Networks (ANNs) to address the challenges of automation and performance-efficient realizations of MS and NMR. In the first part, a complete toolchain has been developed to develop simulated spectra and train ANNs to identify compounds in MS. In the second part, a limited number of experimental NMR spectra have been augmented by simulated spectra to train an ANN with better prediction performance and speed than state-of-theart analysis. These results suggest that, in the context of the digital transformation of the process industry, we are now on the threshold of a possible strongly simplified use of MS and MRS and the accompanying data evaluation by machine-supported procedures, and can utilize both methods much wider for reaction and process monitoring or quality control.
Mass Spectrometry (MS) and Nuclear Magnetic Resonance Spectroscopy (NMR) are valuable analytical and quality control methods for most industrial chemical processes as they provide information on the concentrations of individual compounds and by-products. These processes are traditionally carried out manually and by a specialist, which takes a substantial amount of time and prevents their utilization for real-time closed-loop process control. This paper presents recent advances from two projects that use Artificial Neural Networks (ANNs) to address the challenges of automation and performance-efficient realizations of MS and NMR. In the first part, a complete toolchain has been realized to develop simulated spectra and train ANNs to identify compounds in MS. In the second part, a limited number of experimental NMR spectra have been augmented by simulated spectra, to train an ANN with better prediction performance and speed than state-of-the-art analysis. These results suggest that, in the context of the digital transformation of the process industry, we are now on the threshold of a strongly simplified use of MS and MRS and the accompanying data evaluation by machine-supported procedures, and can utilize both methods much wider for reaction and process monitoring or quality control.
Industry 4.0 is all about interconnectivity, sensor-enhanced process control, and data-driven systems. Process analytical technology (PAT) such as online nuclear magnetic resonance (NMR) spectroscopy is gaining in importance, as it increasingly contributes to automation and digitalization in production. In many cases up to now, however, a classical evaluation of process data and their transformation into knowledge is not possible or not economical due to the insufficiently large datasets available. When developing an automated method applicable in process control, sometimes only the basic data of a limited number of batch tests from typical product and process development campaigns are available. However, these datasets are not large enough for training machine-supported procedures. In this work, to overcome this limitation, a new procedure was developed, which allows physically motivated multiplication of the available reference data in order to obtain a sufficiently large dataset for training machine learning algorithms. The underlying example chemical synthesis was measured and analyzed with both application-relevant low-field NMR and high-field NMR spectroscopy as reference method. Artificial neural networks (ANNs) have the potential to infer valuable process information already from relatively limited input data. However, in order to predict the concentration at complex conditions (many reactants and wide concentration ranges), larger ANNs and, therefore, a larger Training dataset are required. We demonstrate that a moderately complex problem with four reactants can be addressed using ANNs in combination with the presented PAT method (low-field NMR) and with the proposed approach to generate meaningful training data.
We compare four different sensing solutions suitable for distributed fiber optic humidity sensing in per uorinated graded-index polymer optical fibers (PFGI-POFs). Compared to silica fbers, polymer optical fibers over advantageous beneffits including signifficantly higher break down strain, fracture toughness and humidity sensitivity. Various humidity-related effects in PFGI-POFs have been reported in the last years including measured attenuation and length changes as well as Brillouin frequency and Bragg wavelength shifts. The four aforementioned methods could serve as a basis for distributed and quasi-distributed humidity sensing and are described here closely with an emphasis on plausible cross effects to temperature and strain. The main focus of this paper lies on the comparison of four approaches with regard to method complexity, sensitivity to humidity, spatial resolution, real-time capability and effort to compensate for cross effects.
The distributed measurement of relative humidity is a sought-after capability for a wide range of applications in civil engineering and structural health monitoring. We show that polymethyl methacrylate (PMMA) optical fi-bers can be employed as a sensor medium to conduct distributed humidity measurement by analyzing Rayleigh backscattering traces obtained by OTDR. We make use of the effect that water penetrates the fiber core and directly influences the local fiber attenuation and Rayleigh backscatter coefficient. We conducted distributed backscattering analysis for two different pulse wavelengths: 500 nm and 650 nm. The 650 nm results are susceptible to both, attenuation changes and backscatter changes, whereas backscatter results at 500 nm are not affected by humidity-induced attenuation and only exhibit a change of Rayleigh backscattered power as a function of humidity. The combined measurement and analysis of both parameters at these two wavelengths has the advantage that cross-sensitivities on backscatter change and attenuation, such as strain and tempera-ture changes, could be separated from the humidity response of the fiber. We present laboratory results for a humidity range between 30% and 90% for both pulse wavelengths: including step responses, humidity cycles and hysteresis analysis. In addition to the attenuation and backscatter coefficient dependence, we also analyze optical runtime changes as a function of humidity. POFs have the advantage that they can be directly embed-ded into materials such as concrete or soil to measure water content or localize water ingress. Standard step-index PMMA POFs can be used as a distributed relative humidity sensor up to 200 m distance.
X-ray reflectivity (XRR) is a powerful and popular scattering technique that can give valuable insight into the growth behavior of thin films. This study Shows how a simple artificial neural network model can be used to determine the thickness, roughness and density of thin films of different organic semiconductors [diindenoperylene, copper(II) phthalocyanine and alpha-sexithiophene] on silica from their XRR data with millisecond computation time and with minimal user input or a priori knowledge. For a large experimental data set of 372 XRR curves, it is shown that a simple fully connected model can provide good results with a mean absolute percentage error of 8–18% when compared with the results obtained by a genetic least mean squares fit using the classical Parratt formalism. Furthermore, current drawbacks and prospects for improvement are discussed.
We report, to our knowledge, for the first time on humidity-induced Brillouin frequency shifts in perfluorinated graded index polymer optical fibers. A linear relation between Brillouin frequency shift and humidity was observed. Furthermore, the humidity coefficient of the Brillouin frequency shift is demonstrated to be a function of temperature (-107 to -64 kHz/%r.h. or -426 to -49 kHz m3/g in the range of 20 to 60 °C). An analytical description proves temperature and humidity as two mutually independent effects on the Brillouin frequency shift.
In this paper perfluorinated graded-index polymer optical fibers are characterized with respect to the influence of relative humidity changes on spectral transmission absorption and Rayleigh backscattering. The hygroscopic and thermal expansion coefficient of the fiber are determined to be CHE = (7.4 +/- 0.1) 10^-6 %r.h.^-1 and CTE = (22.7 +/- 0.3) 10^-6 K^-1, respectively. The influence of humidity on the Brillouin backscattering power and linewidth are presented for the first time to our knowledge. The Brillouin backscattering power at a pump wavelength of 1319 nm is affected by temperature and humidity. The Brillouin linewidth is observed to be a function of temperature but not of humidity. The strain coefficient of the BFS is determined to be CS = (146.5 +/- 0.9) MHz/% for a wavelength of 1319 nm within a strain range from 0.1 % to 1.5 %. The obtained results demonstrate that the humidity-induced Brillouin frequency shift is predominantly caused by the swelling of the fiber over-cladding that leads to fiber straining.
A long distance range over tens of kilometers is a prerequisite for a wide range of distributed fiber optic vibration sensing applications. We significantly extend the attenuation-limited distance range by making use of the multidimensionality of distributed Rayleigh backscatter data: Using the wavelength-scanning coherent optical time domain reflectometry (WS-COTDR) technique, backscatter data is measured along the distance and optical frequency dimensions. In this work, we develop, train, and test deep convolutional neural networks (CNNs) for fast denoising of these two-dimensional backscattering results. The very compact and efficient CNN denoiser “DnOTDR” outperforms state-of-the-art image denoising algorithms for this task and enables denoising data rates of 1.2 GB/s in real time. We demonstrate that, using the CNN denoiser, the quantitative strain measurement with nm/m resolution can be conducted with up to 100 km distance without the use of backscatter-enhanced fibers or distributed Raman or Brillouin amplification.