Search

Unsupervised Domain Adaptation (UDA) aims to learn a classifier for the unlabeled target domain by leveraging knowledge from a labeled source domain with a different but related distribution. Many existing approaches typically learn a domain-invariant representation space by directly matching the marginal distributions of the two domains. However, they ignore exploring the underlying discriminative features of the target data and align the cross-domain discriminative features, which may lead to suboptimal performance. To tackle these two issues simultaneously, this paper presents a Joint Clustering and Discriminative Feature Alignment (JCDFA) approach for UDA, which is capable of naturally unifying the mining of discriminative features and the alignment of class-discriminative features into one single framework. Specifically, in order to mine the intrinsic discriminative information of the unlabeled target data, JCDFA jointly learns a shared encoding representation for two tasks: supervised classification of labeled source data, and discriminative clustering of unlabeled target data, where the classification of the source domain can guide the clustering learning of the target domain to locate the object category. We then conduct the cross-domain discriminative feature alignment by separately optimizing two new metrics: 1) an extended supervised contrastive learning, i.e., semi-supervised contrastive learning 2) an extended Maximum Mean Discrepancy (MMD), i.e., conditional MMD, explicitly minimizing the intra-class dispersion and maximizing the inter-class compactness. When these two procedures, i.e., discriminative features mining and alignment are integrated into one framework, they tend to benefit from each other to enhance the final performance from a cooperative learning perspective. Experiments are conducted on four real-world benchmarks ( e.g., Office-31, ImageCLEF-DA, Office-Home and VisDA-C). All the results demonstrate that our JCDFA can obtain rem

Constrained by the expansion of the power grid, the development of offshore wind farms may be hindered and begin to experience severe curtailment or restriction. The combination of hydrogen production through electrolysis and hydrogen-to-power is considered to be a potential option to achieve the goal of low-carbon and energy security. This work investigates the competitiveness of different system configurations to export hydrogen and/or electricity from offshore plants, with particular emphasis on unloading the mixture of hydrogen and electricity to end-users on land. Including the levelized energy cost and net present value, a comprehensive techno-economic assessment method is proposed to analyze the offshore system for five scenarios. Assuming that the baseline distance is 10 km, the results show that exporting hydrogen to land through pipelines shows the best economic performance with the levelized energy cost of 3.40 $/kg. For every 10 km increase in offshore distance, the net present value of the project will be reduced by 5.69 MU$, and the project benefit will be positive only when the offshore distance is less than 53.5 km. An important finding is that the hybrid system under ship transportation mode is not greatly affected by the offshore distance. Every 10% increase in the proportion of hydrogen in the range of 70%–100% can increase the net present value by 1.43–1.70 MU$, which will increase by 7.36–7.37 MU$ under pipeline transportation mode. Finally, a sensitivity analysis was carried out to analyze the wind speed, electricity and hydrogen prices on the economic performance of these systems.

Offshore oil and gas field development consumes quantities of electricity, which is usually provided by gas turbines. In order to alleviate the emission reduction pressure and the increasing pressure of energy saving, governments of the world have been promoting the reform of oil and gas fields for years. Nowadays, environmentally friendly alternatives to provide electricity are hotspots, such as the integration of traditional energy and renewable energy. However, the determination of system with great environmental and economic benefits is still controversial. This paper proposed a wind–hydrogen–natural gas nexus (WHNGN) system for sustainable offshore oil and gas fields development. Combining the optimization model with the techno-economic evaluation model, a comprehensive evaluation framework is established for techno-economic feasibility analysis. In addition to WHNGN system, another two systems are designed for comparison, including the traditional energy supply (TES) system and wind–natural gas nexus (WNGN) system. An offshore production platforms in Bohai Bay in China is taken as a case, and the results indicate that: (i) WNGN and WHNGN systems have significant economic benefits, total investment is decreased by 5,190 and 5,020 million $ respectively, and the WHNGN system increases 4,174 million $ profit; (ii) WNGN and WHNGN systems have significant environmental benefits, annual carbon emission is decreased by 15 and 40.2 million kg respectively; (iii) the system can be ranked by economic benefits as follows: WHNGN>WNGN>TES; and (iV) the WHNGN system is more advantageous in areas with high hydrogen and natural gas sales prices, such as China, Kazakhstan, Turkey, India, Malaysia and Indonesia.

As climate issues gradually attract public attention worldwide, the operation and construction of oil product pipelines have been attached with new energy-saving and emission-reduction targets. Though previous studies concerning Life Cycle Assessment of oil and gas pipelines have estimated the carbon footprint to some extent, there is a lack of researches that take the characteristics of oil products pipelines into consideration. Oil products pipelines undertake the task of delivering various products to downstream demand locations, which differs greatly from other pipeline transportation systems as back-to-back sequential delivery is adopted. In this paper, a detailed Life Cycle Assessment model is established to analyze carbon emissions of oil products pipeline system from construction to disposal as well as its impact on soil environment. Data from practical pipes is adopted as the case study to reflect emissions produced in different stages, and the amount of total and unified emissions of different pipes provided through the proposed model is within the range of 2.78 to 4.70 tCO2e/t·km. Then, sensitivity analysis is carried out to identify the driving factors of emissions. According to the calculation results, pipe length, diameter and throughput turn out to be the dominating factors, and an empirical formula is derived for future planned pipes. Relevant recommendations are put forward based on the results to help reduce emissions from oil product pipe transportation.