PeMNet for Pectoral Muscle Segmentation

As an important imaging modality, mammography is considered to be the global gold standard for early detection of breast cancer. Computer-Aided (CAD) systems have played a crucial role in facilitating quicker diagnostic procedures, which otherwise could take weeks if only radiologists were involved. In some of these CAD systems, breast pectoral segmentation is required for breast region partition from breast pectoral muscle for specific analysis tasks. Therefore, accurate and efficient breast pectoral muscle segmentation frameworks are in high demand. Here, we proposed a novel deep learning framework, which we code-named PeMNet, for breast pectoral muscle segmentation in mammography images. In the proposed PeMNet, we integrated a novel attention module called the Global Channel Attention Module (GCAM), which can effectively improve the segmentation performance of Deeplabv3+ using minimal parameter overheads. In GCAM, channel attention maps (CAMs) are first extracted by concatenating feature maps after paralleled global average pooling and global maximum pooling operation. CAMs are then refined and scaled up by multi-layer perceptron (MLP) for elementwise multiplication with CAMs in next feature level. By iteratively repeating this procedure, the global CAMs (GCAMs) are then formed and multiplied elementwise with final feature maps to lead to final segmentation. By doing so, CAMs in early stages of a deep convolution network can be effectively passed on to later stages of the network and therefore leads to better information usage. The experiments on a merged dataset derived from two datasets, INbreast and OPTIMAM, showed that PeMNet greatly outperformed state-of-the-art methods by achieving an IoU of 97.46%, global pixel accuracy of 99.48%, Dice similarity coefficient of 96.30%, and Jaccard of 93.33%, respectively.

Biospeckle Activity of Highbush Blueberry Fruits Infested by Spotted Wing Drosophila (Drosophila suzukii Matsumura)

In this study, the potential of the biospeckle phenomenon for detecting fruit infestation by Drosophila suzukii was examined. We tested both graphical and analytical approaches to evaluate biospeckle activity of healthy and infested fruits. As a result of testing the qualitative approach, a generalized difference method proved to be better at identifying infested areas than Fujii’s method. Biospeckle activity of healthy fruits was low and increased with infestation development. It was found that the biospeckle activity index calculated from spatial-temporal speckle correlation of THSP was the best discriminant of healthy fruits and fruits in two different stages of infestation development irrespective of window size and pixel selection strategy adopted to create the THSP. Other numerical indicators of biospeckle activity (inertia moment, absolute value of differences, average differences) distinguished only fruits in later stage of infestation. Regular values of differences turned out to be of no use in detecting infested fruits. We found that to provide a good representation of activity it was necessary to use a strategy aimed at random selection of pixels gathered around the global maximum of biospeckle activity localized on the graphical outcome. The potential of biospeckle analysis for identification of highbush blueberry fruits infested by D. suzukii was confirmed.

On the environmental zero-point problem in microevolutionary climate response predictions

It is well documented that populations adapt to climate change by means of phenotypic plasticity, but few reports on adaptation by means of genetically based microevolution caused by selection. Disentanglement of these separate effects requires that the environmental zero-point is defined, and this should not be done arbitrarily. Together with parameter values, the zero-point can be estimated from environmental, phenotypic and fitness data. A prediction error method for this purpose is described, with the feasibility shown by simulations. An estimated environmental zero-point may have large errors, especially for small populations, but may still be a better choice than use of an initial environmental value in a recorded time series, or the mean value, which is often used. Another alternative may be to use the mean value of a past and stationary stochastic environment, which the population is judged to have been fully adapted to, in the sense that the mean fitness was at a global maximum. An exception is here cases with constant phenotypic plasticity, where the microevolutionary change per generation follows directly from phenotypic and environmental data, independent of the chosen environmental zero-point.

High-Flexibility MPPT Techniques with Communication Scan Network for PV Micro-Grid System

This paper proposes the design and implementation of inductor-inductor-capacitor (LLC) converters with modules connected in series with the power scan method and communication scan network (CSN) to achieve MPPT and regulate the output voltage for the PV micro-grid system. The Dc/Dc converters includes six isolated LLC modules in series to supply ±380 V output voltage and track the maximum power point of the PV system. The series LLC converters are adopted to achieve high efficiency and high flexibility for the PV micro-grid system. The proposed global maximum power scan technique is implemented to achieve global maximum power tracking by adjusting the switching frequency of the LLC converter. To improve the system flexibility and achieve system redundancy, module failure can be detected in real time with a communication scan network, and then the output voltage of other modules will be changed by adjusting the switching frequency to maintain the same voltage as before the failure. Additionally, the proposed communication scan network includes the RS-485 interface of the MPPT series module and the CAN BUS communication interface with other subsystems’ communication for the PV micro-grid application system. Finally, a 6 kW MPPT prototype with a communication scan network is implemented and the proposed control method is verified for the PV system.

Test Results of the Long Baseline Navigation Solutions under a Large a Priori Position Uncertainty

A method is proposed for long baseline navigation of autonomous underwater vehicles (AUV) to be used in the case of a large a priori position uncertainty. The new modified method is based on the iterated Kalman filter (IKF) working with different initial linearization points. The final solution is calculated by clustering and weighting the IKF results. This approach allows position estimates to be determined in accordance with the global maximum of posteriori probability density of coordinates. The test results obtained with the use of three beacons and an underwater vehicle are presented.

Improved Invasive Weed Optimization Algorithm for Global Maximum Power Point Tracking of PV Array Under Partial Shading Conditions

Photovoltaic (PV) array under partial shading conditions (PSCs) has several maximum power points (MPPs) on the power-voltage curve of the PV array. These points; have a unique global peak (GP) and the others are local peaks (LPs). This paper aims to study an improved version of a heuristic optimization technique namely, Invasive Weed Optimization (IWO) to track the global maximum power point (GMPP) of a PV array which is an important issue. The proposed improved IWO (IIWO) algorithm modifies IWO to speed up the convergence and make the system more efficient. In addition to study the effect of changing input parameters of IIWO on its performance. An overall statistical evaluation of IIWO, with standard IWO and Particle Swarm Optimization (PSO) is executed under different shading conditions. The simulation results show that IIWO has faster and better convergence as it can reach the GMPP in less time compared with other techniques.

Investigation of Partial Shading Scenarios on a Photovoltaic Array’s Characteristics

The purpose of this study is to investigate the impact of different partial shading scenarios on a PV array’s characteristics in order to develop a simple and easy-to-implement GMPP controller that tracks the PV array’s global maximum power point (GMPP). The P-V characteristic of the PV array becomes more complicated under partial shading, owing to the presence of many power peaks, as opposed to uniform irradiance conditions, when there is only one peak called the maximum power point. In fact, and according to an experiment conducted in this study, when a PV array is partially shaded, the P-V characteristic mostly presents two peaks, given the existence of only two levels of irradiance, one of which is called the global peak (i.e., the GMPP). Furthermore, the first peak is located at Vmpp1 (the PV array’s voltage corresponds to this peak), whereas the second is at Vmpp2. The proposed approach works by estimating the values of Vmpp1 and Vmpp2 using two equations in order to control the DC/DC converter of the PV system. The first equation is used when the GMPP is at the first peak, while the other is used when the GMPP is at the second peak. Several scenarios are simulated and presented in this paper to verify the accuracy of these equations. In addition, some conclusions are drawn to suggest a simple method for tracking the GMPP.

A maximum power point tracking strategy applied to building integrated photovoltaics

To solve the problem of permanent-shadow shading of photovoltaic buildings, a maximum power point tracking (MPPT) strategy to determine the search range by pre-delimiting area is proposed to improve MPPT efficiency. The single correspondence between the solar-cell current–voltage (I–V) curve and the illumination conditions was proved by using the single-diode model of photovoltaic cells, thus proving that a change in the illumination conditions corresponds to a unique maximum power point (MPP) search area. According to the approximate relationship between MPP voltage, current and open-circuit voltage and short-circuit current of a photovoltaic module, the voltage region where the MPP is located is determined and the global maximum power point is determined using the power operating triangle strategy in this region. Simulation carried out in MATLAB proves the correctness and feasibility of the theoretical research. Simulation results show that the MPPT strategy proposed in this paper can improve the average efficiency by 1.125% when applied in series as building integrated photovoltaics.

ON THE CAPACITY OF SOLAR CELLS UNDER PARTIAL SHADING CONDITIONS

A photovoltaic system under uniform illumination conditions has only one global maximum power point, but this same panel in the case of partial shading conditions can have more maxima on the power-voltage curve. The motivation of this work are two-fold: to show nonlinear capacitance of solar cell and to propose a determination technique for partial shading condition. The author shows the results of measurements on a photovoltaic system under different lighting conditions. Shows the effect of capacitance on the shape of the response to excitation in the form of a load. The paper also shows which photovoltaic cell parameters influence the shape of the C-V curve.

Uncertainty Quantification and Optimization of Deep Learning for Fracture Recognition

Automatic fracture recognition from borehole images or outcrops is applicable for the construction of fractured reservoir models. Deep learning for fracture recognition is subject to uncertainty due to sparse and imbalanced training set, and random initialization. We present a new workflow to optimize a deep learning model under uncertainty using U-Net. We consider both epistemic and aleatoric uncertainty of the model. We propose a U-Net architecture by inserting dropout layer after every "weighting" layer. We vary the dropout probability to investigate its impact on the uncertainty response. We build the training set and assign uniform distribution for each training parameter, such as the number of epochs, batch size, and learning rate. We then perform uncertainty quantification by running the model multiple times for each realization, where we capture the aleatoric response. In this approach, which is based on Monte Carlo Dropout, the variance map and F1-scores are utilized to evaluate the need to craft additional augmentations or stop the process. This work demonstrates the existence of uncertainty within the deep learning caused by sparse and imbalanced training sets. This issue leads to unstable predictions. The overall responses are accommodated in the form of aleatoric uncertainty. Our workflow utilizes the uncertainty response (variance map) as a measure to craft additional augmentations in the training set. High variance in certain features denotes the need to add new augmented images containing the features, either through affine transformation (rotation, translation, and scaling) or utilizing similar images. The augmentation improves the accuracy of the prediction, reduces the variance prediction, and stabilizes the output. Architecture, number of epochs, batch size, and learning rate are optimized under a fixed-uncertain training set. We perform the optimization by searching the global maximum of accuracy after running multiple realizations. Besides the quality of the training set, the learning rate is the heavy-hitter in the optimization process. The selected learning rate controls the diffusion of information in the model. Under the imbalanced condition, fast learning rates cause the model to miss the main features. The other challenge in fracture recognition on a real outcrop is to optimally pick the parental images to generate the initial training set. We suggest picking images from multiple sides of the outcrop, which shows significant variations of the features. This technique is needed to avoid long iteration within the workflow. We introduce a new approach to address the uncertainties associated with the training process and with the physical problem. The proposed approach is general in concept and can be applied to various deep-learning problems in geoscience.