The continuous growth of renewable energy sources and combined heat-power generation plants in public infrastructures allow them a certain degree of independence from the public power grid. Nevertheless, during the day, the power demand fluctuates, creating peak loads that cannot be fulfilled by the generated power and must be covered by the public power grid. Peak loads could be diminished by saving energy produced during low power demand periods and returning it during the high demand periods. The use of electric vehicles is expected to keep increasing in the next years and the use of vehicle-to-grid technology allows electric and hybrid cars to return stored energy to the power grid. An algorithm to evaluate the vehicle-to-grid technology as a solution for peak reduction, also called peak shaving, for public infrastructure was developed. The purpose of the algorithm is to predict the optimal charging and discharging schedule of the battery from electric vehicles in a parking space so that the peak load demands are met without sacrificing the driving demands of the electric vehicle (EV) users. This takes into consideration factors like the state of charge of the batteries and the mobility needs of the vehicle user. The algorithm takes the driving demands and mobility needs as constraints, and schedules the charging and discharging of the EVs during their stay in a parking place in order to reduce the peak load demands by supplying stored electrical power from EV batteries back to the grid. In order to validate the algorithm, three test case scenarios representing different parking patterns were generated using random and statistical distributed parameters. It is concluded that vehicle-to-grid technology can be used to reduce peak demands, where the number of EVs and their stay hours in a parking place represent the most critical parameters for the effectivity of the peak reduction. This implies that in the future, public infrastructure could profit better from renewable energy sources by adding loading stations for EVs that are vehicle-to-grid capable.

Fruit packaging is a time-consuming task due to its low automation level. The gentle handling required by some kinds of fruits and their natural variations complicates the implementation of automated quality controls and tray positioning for final packaging. In this article, we propose a method for the automatic localization and pose estimation of apples captured by a Red-Green-Blue (RGB) camera using convolutional neural networks. Our pose estimation algorithm uses a cascaded structure composed of two independent convolutional neural networks: one for the localization of apples within the images and a second for the estimation of the three-dimensional rotation of the localized and cropped image area containing an apple. We used a single shot multi-box detector to find the bounding boxes of the apples in the images. Lie algebra is used for the regression of the rotation, which represents an innovation in this kind of application. We compare the performances of four different network architectures and show that this kind of representation is more suitable than using state-of-the-art quaternions. By using this method, we achieved a promising accuracy for the rotation regression of 98.36%, considering an error range lower than 15 degrees, forming a base for the automation of fruit packing systems.

No abstract