Approach to Automated Visual Inspection of Objects Based on Artificial Intelligence

The article discusses the possibility of object detector usage in field of automated visual inspection for objects with specific parameters, specifically various types of defects occurring on the surface of a car tire. Due to the insufficient amount of input data, as well as the need to speed up the development process, the Transfer Learning principle was applied in a designed system. In this approach, the already pre-trained convolutional neural network AlexNet was used, subsequently modified in its last three layers, and again trained on a smaller sample of our own data. The detector used in the designed camera inspection system with the above architecture allowed us to achieve the accuracy and versatility needed to detect elements (defects) whose shape, dimensions and location change with each occurrence. The design of a test facility with the application of a 12-megapixel monochrome camera over the rotational table is briefly described, whose task is to ensure optimal conditions during the scanning process. The evaluation of the proposed control system with the quantification of the recognition capabilities in the individual defects is described at the end of the study. The implementation and verification of such an approach together with the proposed methodology of the visual inspection process of car tires to obtain better classification results for six different defect classes can be considered as the main novel feature of the presented research. Subsequent testing of the designed system on a selected batch of sample images (containing all six types of possible defect) proved the functionality of the entire system while the highest values of successful defect detection certainty were achieved from 85.15% to 99.34%.

Effect of quartz aperture covers on the fluid dynamics and thermal efficiency of falling particle receivers

Falling particle receivers are an emerging technology for use in concentrating solar power systems. In this work, quartz half-shells are investigated for use as full or partial aperture covers to reduce receiver thermal losses. A receiver subdomain and surrounding air are modeled using ANSYS® Fluent®. The model is used to simulate fluid dynamics and heat transfer for the following cases: (1) open aperture, (2), aperture fully covered by quartz half-shells, and (3) aperture partially covered by quartz half-shells. We compare the percentage of total incident solar power lost due to conduction through the receiver walls, advective losses through the aperture, and radiation exiting the aperture. Contrary to expected outcomes, results show that quartz aperture covers can increase radiative losses and result in modest to nonexistent reductions in advective losses. The increased radiative losses are driven by elevated quartz half-shell temperatures and have the potential to be mitigated by active cooling and/or material selection. Quartz half-shell total transmissivity was measured experimentally using a radiometer and the National Solar Thermal Test Facility heliostat field. Average measured total transmissivities are 0.97±0.01 and 0.94±0.02 for concave and convex side toward the heliostat field, respectively. Quartz half-shell aperture covers did not yield expected efficiency gains in numerical results due to increased radiative losses, but efficiency improvement in some numerical results and the performance of quartz half-shells subject to concentrated solar radiation suggest quartz half-shell aperture covers should be investigated further.

An in-depth Experimental Investigation of the Outdoor Performance of Wafer and Thin Film PV Technologies in a Tropical Climate

The photovoltaics (PV) industry is booming at an impressive rate. Knowledge of the outdoor perfor-mance of different PV technologies under different climatic conditions is becoming increasingly im-portant for all stakeholders. The aim of this research was to perform the outdoor characterisation of three PV technologies in a tropical climate and evaluate their performances with the aid of a set of key performance indicators. An innovative energy autonomous outdoor test facility has been used to measure the weather conditions and the IV curves of mono-Si, poly-Si and CIGS PV modules. Each IV curve was sampled within less than a second, for every 10 minutes, between sunrise and sunset for a whole year, representing a data set of around 28,000 IV curves of 240 points each. The variations of current, voltage and power were thoroughly studied for changes in temperature and irradiance. This paper reports the variations of temperature coefficients of current, voltage and power with the inten-sity of light. While PV module documentation only present the temperature coefficients of the short circuit current and open circuit voltage at Standard Test Conditions, this paper additionally provides highly valuable information to PV system designers on the variation of these coefficients in the field. The research is also the first to report the variations of the fill factor with temperature and irradi-ance. In general, the wafer technologies were found to have a better performance than the thin film technology. Moreover, the open-circuit temperature coefficient was found to improve for higher irra-diances only for the wafer technologies, while that for the thin-film technology experienced a degrada-tion. The temperature coefficient of current for the mono-Si module was found to be positive at low irradiance levels, but negative at higher irradiance levels.

DSMC simulations of neutral gas flow in the DTT particle exhaust system

Divertor Tokamak Test Facility (DTT) is a new European superconducting tokamak, currently under final design, addressed to investigate alternative power exhaust solutions for DEMO. Although the divertor system is not finalized yet, the machine and port geometry set limitations on the divertor pumping system operational space. A numerical study of neutral gas dynamics in the divertor region is performed based on the DSMC method by applying the DIVGAS code. The study includes both single-null (SN) and double-null (DN) divertor configurations. For both configurations, the SolEdge2D–EIRENE plasma simulations have been performed for a deuterium plasma with neon seeding and the extracted information about the neutral particles on the predefined interfaces is imposed as incoming boundary conditions for DIVGAS simulations. In the SN case, two plasma puffing scenarios and three candidate pumping port arrangements have been considered. The divertor dome influence on the pumped fluxes can reach 50%. An increase of the capture coefficient six times leads to a decrease in the pressure at the pumping openings by a factor of about 4.5–7. The influence of the size of the lower vertical opening has been studied showing that the enlarged vertical port may establish as the main pumping opening. In the DN case, when the pumping is performed from both lower and upper divertor the overall pumped fluxes at the upper divertor are always higher than the corresponding ones for the lower divertor by a factor of 2–2.5, mainly due to the difference in the pumping areas. In both SN and DN cases, the neutrals outflux toward the X-point dominates the particle transport in the private flux region. The operational space provided by this first assessment is relatively stable against modified classical divertor geometries and allows a more thorough assessment of the pumping technology of the DTT fusion device in the future.

Benchmarking 2D against 3D FDTD codes for the assessment of the measurement performance of a low field side plasma position reflectometer applicable to IDTT

O-mode reflectometry, a technique to diagnose fusion plasmas, is foreseen as a source of real-time (RT) plasma position and shape measurements for control purposes in the coming generation of machines such as DEMO. It is, thus, of paramount importance to predict the behavior and capabilities of these new reflectometry systems using synthetic diagnostics. Finite-difference time-domain (FDTD) time-dependent codes allow for a comprehensive description of reflectometry but are computationally demanding, especially when it comes to three-dimensional (3D) simulations, which requires access to High Performance Computing (HPC) facilities, making the use of two-dimensional (2D) codes much more common. It is important to understand the compromises made when using a 2D model in order to decide if it is applicable or if a 3D approach is required. This work attempts to answer this question by comparing simulations of a potential plasma position reflectometer (PPR) at the Low Field-Side (LFS) on the Italian Divertor Tokamak Test facility (IDTT) carried out using two full-wave FDTD codes, REFMULF (2D) and REFMUL3 (3D). In particular, the simulations consider one of IDTT’s foreseen plasma scenarios, namely, a Single Null (SN) configuration, at the Start Of Flat-top (SOF) of the plasma current.

Quantifying the Impact of Structural Fibers on the Performance of Concrete Overlays on Asphalt

Concrete overlays on asphalt pavement, also known as whitetopping, are growing in popularity as an option for the rehabilitation of distressed asphalt pavements. The performance of whitetoppings over the past several decades has shown that under heavy and frequent traffic loads, they can be susceptible to panel migration and faulting due to the lack of tie bars and dowel bars within the thin cross sections. One mitigation method to reduce panel migration and faulting is the inclusion of structural fibers into the concrete mix. While structural fibers have anecdotally been shown to contribute toward better performance in whitetoppings, few studies have quantified the benefits provided by the typical dosage of fibers used in recent specifications. Two sets of similarly designed experimental test sections constructed at the MnROAD test facility in 2004 and 2013, have provided the opportunity to evaluate and quantify the impact of structural fibers on whitetopping performance. This comparison of the performance between plain concrete and fiber-reinforced concrete overlay test sections includes analysis of material properties of the mixes, the difference in response to environmental and traffic loads, typical distresses, and ride quality. Based on the results of the analysis, recommendations were made with regards to whether the types and dosages of structural fibers used in the test sections made a sufficient impact on performance.