Ryerson University

About: Ryerson University is a education organization based out in Toronto, Ontario, Canada. It is known for research contribution in the topics: Population & Poison control. The organization has 7671 authors who have published 20164 publications receiving 394976 citations. The organization is also known as: Ryerson Polytechnical Institute & Ryerson Institute of Technology.

Segmentation-Aware Convolutional Networks Using Local Attention Masks

Abstract: We introduce an approach to integrate segmentation information within a convolutional neural network (CNN). This counter-acts the tendency of CNNs to smooth information across regions and increases their spatial precision. To obtain segmentation information, we set up a CNN to provide an embedding space where region co-membership can be estimated based on Euclidean distance. We use these embeddings to compute a local attention mask relative to every neuron position. We incorporate such masks in CNNs and replace the convolution operation with a “segmentation-aware” variant that allows a neuron to selectively attend to inputs coming from its own region. We call the resulting network a segmentation-aware CNN because it adapts its filters at each image point according to local segmentation cues, while at the same time remaining fully-convolutional. We demonstrate the merit of our method on two widely different dense prediction tasks, that involve classification (semantic segmentation) and regression (optical flow). Our results show that in semantic segmentation we can replace DenseCRF inference with a cascade of segmentation-aware filters, and in optical flow we obtain clearly sharper responses than the ones obtained with comparable networks that do not use segmentation. In both cases segmentation-aware convolution yields systematic improvements over strong baselines.

Optimal participation and risk mitigation of wind generators in an electricity market

Abstract: Forecasted output of wind electric generators (WEGs) in a 24-h horizon has large uncertainties. These uncertainties pose a challenge while computing optimal bids necessary for participating in the day-ahead unit commitment process (DACP) thus limiting their integration and success. This study proposes a new optimal participation strategy for a WEG that employs an energy storage device (ESD) for participating in the DACP. The WEG is modelled to function as a price-taker. The proposed formulation has two objectives: (a) maximise returns from the market considering the best forecast; and (b) minimise risks considering the forecast uncertainties. Risk in the participation strategy is quantified by computing expected energy not served (EENS). The multiobjective mixed integer linear programming formulation is transformed into a fuzzy optimisation problem and solved. Through suitable examples, the ESD is shown to play two important roles: (a) it helps to shift wind energy produced during hours with low marginal prices to those hours with higher marginal prices by appropriately storing and releasing it. This shift can be forward or backward in time. (b) The second crucial role played by ESD, upon minimising EENS, is to maintain an energy reserve akin to spinning reserve such that the risk of the optimal participation schedule is the least.

Improvement of Wind-Generator Stability by Fuzzy-Logic-Controlled SMES

Abstract: This paper proposes a fuzzy-logic-controlled superconducting magnetic energy storage (SMES) scheme for the stabilization of grid-connected wind-generator systems. The control scheme of SMES is based on a sinusoidal pulsewidth-modulation voltage-source converter and a two-quadrant dc-dc chopper using an insulated-gate bipolar transistor. A comparative study is carried out between the proposed fuzzy-logic-controlled SMES and the fuzzy-logic-based pitch controller to improve the wind- generator stability. Simulation results demonstrate that the performance of the proposed fuzzy-logic-controlled SMES is better than that of the fuzzy-logic-based pitch controller in order to stabilize the wind generator.