Search

Direct encounters, in which two or more individuals are physically close to one another, are a topic of increasing interest as more and better movement data become available. Recent progress, including the development of statistical tools for estimating robust measures of changes in animals’ space use over time, facilitates opportunities to link direct encounters between individuals with the long-term consequences of those encounters. Working with movement data for coyotes (Canis latrans) and grizzly bears (Ursus arctos horribilis), we investigate whether close intraspecific encounters were associated with spatial shifts in the animals’ range distributions, as might be expected if one or both of the individuals involved in an encounter were seeking to reduce or avoid conflict over space. We analyze the movement data of a pair of coyotes in detail, identifying how a shift in home range location resulting from altered movement behavior was apparently a consequence of a close intraspecific encounter. With grizzly bear movement data, we approach the problem from the perspective of a set of encounter pairs within a population. We find support for the hypotheses that 1) close intraspecific encounters between bears are, on average, associated with subsequent shifts in range distributions and 2) encounters defined at finer spatial scales are followed by greater changes in space use. Our results suggest that animals can undertake long-term, large-scale spatial shifts in response to close intraspecific encounters that have the potential for conflict. These results lend support for existing theory on the evolution of territories and space use (e.g., Maynard-Smith’s bourgeois strategy regarding low-conflict coexistence). Overall, we find that analyses of movement data in a pairwise context can 1) identify distances at which individuals’ proximity to one another may alter behavior and 2) facilitate testing of population-level hypotheses concerning the potential for direct encoun

Pipeline transportation is a low-energy and economical mode of transporting petroleum products in the downstream supply chain, however, there is almost no theoretical research on its pricing strategies. The unreasonable pricing strategy has resulted in low utilization of multi-product pipeline capacity as well as high energy consumption of petroleum products transportation. Therefore, this paper aims to improve pipeline turnover and promote the low-carbon transportation market from the perspective of pipeline pricing optimization. We propose an integrated framework for multi-product pipelines that couples pricing strategy and logistics optimization model. This framework simulates the pricing behavior of the pipeline carrier and the corresponding logistics planning behavior of the oil shipper. We apply the framwork to 10 pipeline pricing schemes for two regions in China with different logistics structures, and analyze the economic and environmental benefits of the new strategy. The results show that the well-performing scheme can increase pipeline carriers' revenue by 11.41 million CNY per month, significantly improve the competitive advantage of long-distance pipelines, and reduce energy consumption by 272 tce. Based on these findings, we provide recommendations for policymakers at four levels. In conclusion, the new pricing strategy will help reverse the disadvantageous situation of the pipeline in the competitive market and promote energy conservation in the petroleum products logistics industry.

Fine-Grained Visual Categorization (FGVC) aims to distinguish between extremely similar subordinate-level categories within the same basic-level category. Existing research has proven the great importance of the discriminative features in FGVC but ignored the contributions for correct classification from other features, and the extracted features always contain more information about the obvious regions but less about subtle regions. In this paper, firstly, a novel module named forcing module is proposed to force the network to extract more diverse features for FGVC, which generates a suppression mask based on the class activation maps to suppress the most distinguishable regions, so as to force the network to extract other secondary distinguishable features as the final features. The forcing module consists of the original branch and the forcing branch. The original branch focuses on the primary discriminative regions while the forcing branch focuses on secondary discriminative regions. Secondly, in order to solve the problem that information of small-scale distinguishable features is lost seriously after multi-layer down-sampling, according to the class activation maps of the first prediction, the object is cropped and scaled as the second input. To reduce the prediction error, the first and second prediction probabilities are fused as the final prediction result. Experimental results indicate that the proposed method not only outperforms the baseline model by a large margin (3.7%, 5.9%, 3.1% respectively) on CUB-200-2011, Stanford-Cars, and FGVC-Aircraft, but also achieves state-of-the-art performance on FGVC-Aircraft.

Recently, clean solar energy has aroused wide attention due to its excellent potential for electricity production. A highly accurate prediction of photovoltaic power generation (PVPG) is the basis of the production and transmission of electricity. However, the current works neglect the regional correlation characteristics of PVPG and few studies propose an effective framework by incorporating prior knowledge for more physically reasonable results. In this work, a hybrid deep learning framework is proposed for simultaneously capturing the spatial correlations among different regions and temporal dependency patterns with various importance. The scientific theory and domain knowledge are incorporated into the deep learning model to make the predicted results possess physical reasonability. Subsequently, the theory-guided and attention-based CNN-LSTM (TG-A-CNN-LSTM) is constructed for PVPG prediction. In the training process, data mismatch and boundary constraint are incorporated into the loss function, and the positive constraint is utilized to restrict the output of the model. After receiving the parameters of the neural network, a TG-A-CNN-LSTM model, whose predicted results obey the physical law, is constructed. A real energy system in five regions is used to verify the accuracy of the proposed model. The predicted results indicate that TG-A-CNN-LSTM can achieve higher precision of PVPG prediction than other prediction models, with RMSE being 11.07, MAE being 4.98, and R2 being 0.94, respectively. Moreover, the performance of prediction models with sparse data is tested to illustrate the stability and robustness of TG-A-CNN-LSTM.

Direct encounters, in which two or more individuals are physically close to one another, are a topic of increasing interest as more and better movement data become available. Recent progress, including the development of statistical tools for estimating robust measures of changes in animals’ space use over time, facilitates opportunities to link direct encounters between individuals with the long-term consequences of those encounters. Working with movement data for coyotes (Canis latrans) and grizzly bears (Ursus arctos horribilis), we investigate whether close intraspecific encounters were associated with spatial shifts in the animals’ range distributions, as might be expected if one or both of the individuals involved in an encounter were seeking to reduce or avoid conflict over space. We analyze the movement data of a pair of coyotes in detail, identifying how a shift in home range location resulting from altered movement behavior was apparently a consequence of a close intraspecific encounter. With grizzly bear 50 movement data, we approach the problem from the perspective of a set of encounter pairs within a population. We find support for the hypotheses that 1) close intraspecific encounters between bears are, on average, associated with subsequent shifts in range distributions and 2) encounters defined at finer spatial scales are followed by greater changes in space use. Our results suggest that animals can undertake long-term, large-scale spatial shifts in response to close intraspecific encounters that have the potential for conflict. These results lend support for existing theory on the evolution of territories and space use (e.g., Maynard-Smith’s bourgeois strategy regarding low-conflict coexistence). Overall, we find that analyses of movement data in a pairwise context can 1) identify distances at which individuals’ proximity to one another may alter behavior and 2) facilitate testing of population-level hypotheses concerning the potential for direct enc

Analyzing usability test videos is arduous. Although recent research showed the promise of AI in assisting with such tasks, it remains largely unknown how AI should be designed to facilitate effective collaboration between user experience (UX) evaluators and AI. Inspired by the concepts of agency and work context in human and AI collaboration literature, we studied two corresponding design factors for AI-assisted UX evaluation: explanations and synchronization. Explanations allow AI to further inform humans how it identifies UX problems from a usability test session; synchronization refers to the two ways humans and AI collaborate: synchronously and asynchronously. We iteratively designed a tool-AI Assistant-with four versions of UIs corresponding to the two levels of explanations (with/without) and synchronization (sync/async). By adopting a hybrid wizard-of-oz approach to simulating an AI with reasonable performance, we conducted a mixed-method study with 24 UX evaluators identifying UX problems from usability test videos using AI Assistant. Our quantitative and qualitative results show that AI with explanations, regardless of being presented synchronously or asynchronously, provided better support for UX evaluators' analysis and was perceived more positively; when without explanations, synchronous AI better improved UX evaluators' performance and engagement compared to the asynchronous AI. Lastly, we present the design implications for AI-assisted UX evaluation and facilitating more effective human-AI collaboration. © 2022 ACM.

This paper presents a model developed to numerically determine the radiation pattern presented by a given geodesic lens. A simplified model describing the trajectory of a TEM-mode wave through the lens was required as a time-efficient alternative to commercial software. Geometrical optics have been used to approximate the ray trajectory from the source to the aperture of the lens. The power conservation in ray tubes is then used to evaluate the amplitude of the electric field in the aperture, knowing the distribution at the source. The Kirchhoff diffraction formula is finally numerically approximated to evaluate the farfield. Three examples have been simulated with our algorithm and compared with full-wave simulations. We demonstrate that our assumptions provide a good agreement with full-wave simulation while reducing significantly the computation time, thus providing an efficient method for design optimization. Most notably, this more general formulation may be extended to non-rotationally symmetric lenses., QC 20220812Part of proceedings: ISBN 978-88-31299-04-6

Unsupervised Domain Adaptation (UDA) aims at learning a classifier for an unlabeled target domain by transferring knowledge from a labeled source domain with a related but different distribution. The strategy of aligning the two domains in latent feature space via metric discrepancy or adversarial learning has achieved considerable progress. However, these existing approaches mainly focus on adapting the entire image and ignore the bottleneck that occurs when forced adaptation of uninformative domain-specific variations undermines the effectiveness of learned features. To address this problem, we propose a novel component called Informative Feature Disentanglement (IFD), which is equipped with the adversarial network or the metric discrepancy model, respectively. Accordingly, the new network architectures, named IFDAN and IFDMN, enable informative feature refinement before the adaptation. The proposed IFD is designed to disentangle informative features from the uninformative domain-specific variations, which are produced by a Variational Autoencoder (VAE) with lateral connections from the encoder to the decoder. We cooperatively apply the IFD to conduct supervised disentanglement for the source domain and unsupervised disentanglement for the target domain. In this way, informative features are disentangled from the domain-specific details before the adaptation. Extensive experimental results on three gold-standard domain adaptation datasets, e.g., Office31, Office-Home and VisDA-C, demonstrate the effectiveness of the proposed IFDAN and IFDMN models for UDA.

Few-Shot Learning (FSL) is a challenging and practical learning pattern, aiming to solve a target task which has only a few labeled examples. Currently, the field of FSL has made great progress, but largely in the supervised setting, where a large auxiliary labeled dataset is required for offline training. However, the unsupervised FSL (UFSL) problem where the auxiliary dataset is fully unlabeled has been seldom investigated despite of its significant value. This paper focuses on the more general and challenging UFSL problem and presents a novel method named Coarse-to-Fine Pseudo Supervision-guided Meta-Learning (C2FPS-ML) for unsupervised few-shot object classification. It first obtains prior knowledge from an unlabeled auxiliary dataset during unsupervised meta-training, and then use the prior knowledge to assist the downstream few-shot classification task. Coarse-to-Fine Pseudo Supervisions in C2FPS-ML aim to optimize meta-task sampling process in unsupervised meta-training stage which is one of the dominant factors for improving the performance of meta-learning based FSL algorithms. Human can learn new concepts progressively or hierarchically following the coarse-to-fine manners. By simulating this human’s behaviour, we develop two versions of C2FPS-ML for two different scenarios: one is natural object dataset and another one is other kinds of dataset (e.g., handwritten character dataset). For natural object dataset scenario, we propose to exploit the potential hierarchical semantics of the unlabeled auxiliary dataset to build a tree-like structure of visual concepts. For another scenario, progressive pseudo supervision is obtained by forming clusters in different similarity aspects and is represented by a pyramid-like structure. The obtained structure is applied as the supervision to construct meta-tasks in meta-training stage, and prior knowledge from the unlabeled auxiliary dataset is learned from the coarse-grained level to the fine-grained level. T

Fine-Grained Visual Categorization (FGVC) aims at distinguishing subordinate-level categories with subtle interclass differences. Although previous research shows the impressive effectiveness of the recurrent multi-attention models and the second-order feature encoding, they often require an enormous amount of both computation and memory space, making them inadequate for mobile applications. This paper proposed a Category Attention Transfer CNN (CAT-CNN) to address the efficiency issue in solving FGVC problems. We transfer part attention knowledge from a very large-scale FGVC network to a small but efficient network to significantly improve its presentation ability. Using the proposed CAT-CNN, the accuracy of the efficient networks, such as ShuffleNet, MobilieNet, and EfficientNet, can be improved by up to 5.7% on the CUB-2011-200 dataset without increasing computation complexity or memory cost. Our experiments show that the proposed CAT-CNN can be applied to multiple structures to enhance their performance. With a single efficient network structure and single inference, the proposed CAT-MobileNet-large-1.0 and the CAT-EfficientNet-b0 can achieve accuracies of 86.5% and 86.7%, respectively, on the CUB-2011-200 dataset, which is close to or better than the results from state-of-the-art methods using large scale networks and multiple inferences, and make FGVC feasible on mobile devices.

The integrated energy system which coordinates natural gas, renewable energy, and other energy subsystems is an effective way to promote a low-carbon economy. An effective framework for system assessment and optimisation is a critical issue. This paper takes a natural gas-wind-photovoltaic integrated energy system as the research object and uses the simulation software to analyse its techno-enviro-economic feasibility. Firstly, a mathematical model is customised to optimise the system installation and operation plans. Renewable electricity replaces some natural gas, resulting in pipeline pressure fluctuation. Here, the Stoner Pipeline Simulator software is used to simulate pipeline network operation to quantify the aforementioned pressure fluctuations. The proportion of renewable energy is gradually reduced until the network pressure fluctuation is less than 20% to ensure the stability of pipeline operation. Then, the optimal operation scheme can be determined. Taking three cities in Shandong, China, as cases, the results show that the proposed system is beneficial for urban energy internet development: (i) the total net present cost is reduced by 19.7%, 19.8%, and 20.8%, (ii) annual CO2 emission is reduced by 23.7%, 18.4%, and 12.2%; (iii) the levelised cost of energy is 0.142 $/kWh, 0.143$/kWh, and 0.153$/kWh.

Logistics planning is regarded as the most complex part of supply chain management for refined products. A vital knowledge gap still exists in understanding the trade-offs between the economy and the practicability of logistics schemes. Focus on this issue, this paper proposes a model-experience-driven method for the planning of refined product primary logistics. The method couples three sub-modules: (1) use coordinator's preference information and convex function interpolation to construct satisfaction indicator; (2) set up a multi-objective model for logistics coordination and optimization considering supply adjustment and secondary delivery; (3) adopt the augmented ɛ-constraint method to obtain the Pareto solutions and balance the economy and satisfaction indicators. The method is verified by a small-scale system, where the satisfaction degree increases by 77% while the logistics cost remains unchanged. The method is also successfully applied to a large-scale system with 29 refineries and 196 market depots, where Pareto logistics schemes are obtained and the supply–demand imbalance is greatly eased. The proposed method can help provide theoretical guidance for real-world logistics planning.

The pipeline network integration reform enables unified management of pipelines from different entities. For refined oil logistics, this paper proposes a framework based on the MILP optimization model to quantify its flexibility. Considering the uncertainty, three disturbances occur in the logistics concurrently, and 10,000 simulations are performed to obtain the turnover cost. The ratio of pipeline transportation cost to the calculated average turnover cost is defined as the flexibility indicator. Taking China's largest refined oil pipeline network as an example, the results show that the flexibility rises 8.9% after the reform. The paper also quantifies the impact of the reform on logistics flexibility in South China, which is embodied in achieving lower freights and GHG emissions, lower impact by fluctuations, higher pipeline utilization, more efficient oil product turnover, and the avoiding of depot shortages when facing logistical disturbances. The underlying reasons for the results and 3E analysis are analyzed.

Electrolysis hydrogen generation technology is one of the feasible ways to alleviate the problem of wind electricity curtailment. One promising hydrogen value-added application is to blend hydrogen into the natural gas grid and sell it as the heat energy carrier. This paper aims to discuss the feasibility of a roadmap to urban energy internet with wind electricity-natural gas nexus. Firstly, a framework is raised to integrate wind electricity generation, electrolysis hydrogen generation, and hydrogen-natural gas blending systems. Secondly, a series of reasonable hydrogen supply profiles are provided based on annual electricity curtailment and realistic natural gas scheduling. Then, an energy optimisation model and a techno-economic model are applied to simulate the generation of electricity and hydrogen, as well as determine the most economical hydrogen supply scheme. Finally, a case study in the Beijing-Tianjin-Hebei region of China is taken to validate the benefits of the proposed roadmap. The preferred scheme is worked out with the net present value of 88.8 M$, including the economy configurations of the electricity-hydrogen hybrid generation system, as well as the hydrogen-natural gas blending plan. The results also indicate that annual electricity curtailment and annual carbon emission are decreased by 204 GWh (48.8%) and 40.2 kt (49.9%).