查看更多>>摘要：For the short-term crude oil scheduling problem, it is difficult to guarantee the feasibility of a schedule due to complicated constraints. Meanwhile, uncertainty is a very important concern in refineries, such as unexpected breakdown of charging tanks. Therefore, it is a great challenge to make a schedule feasible. Most existing works on multi-objective opti-mization of short-term crude oil scheduling are developed for refineries processing low-fusion-point oil (L-oil) only and little is done for the case with dual pipelines for processing both L-oil and high-fusion-point oil (H-oil). With five objectives and many constraints, it is challenging for a metaheuristic algorithm to find a feasible schedule. To solve this problem, in this work, constraint violation is used to describe the degree of constraint violation. Thus, an adaptive enhanced selection pressure algorithm based on NSGA-II-CDP (NSGA-II-APE) is proposed to efficiently solve the problem for processing both L-oil and H-oil. This algorithm can effectively enhance the selection pressure in the later iterations. Industrial case problems are used to test the proposed method and compare its perfor-mance with 11 state-of-the-art constrained multi-objective evolution algorithms (CMOEAs). Results show its superiority over the existing ones in terms of convergence, solution diversity, and time efficiency.(c) 2021 Elsevier Inc. All rights reserved.
查看更多>>摘要：Great research efforts have been devoted to exploiting deep neural networks in stock prediction. However, long-term dependencies and chaotic properties are still two major issues that lower the performance of state-of-the-art deep learning models in forecasting future price trends. In this study, we propose a novel framework to address both issues. Specifically, in terms of transforming time series into complex networks, we convert market price series into graphs. Then, structural information, referring to temporal point associations and node weights, is extracted from the mapped graphs to resolve the problems regarding long-term dependencies and chaotic properties. We take graph embeddings to represent the associations among temporal points as the prediction model inputs. Node weights are used as a priori knowledge to enhance the learning of temporal attention. The effectiveness of our proposed framework is validated using real-world stock data, and our approach obtains the best performance among several state-of-the-art benchmarks. Moreover, in the conducted trading simulations, our framework further obtains the highest cumulative profits. Our results supplement the existing applications of complex network methods in the financial realm and provide insightful implications for investment applications regarding decision support in financial markets.(c) 2021 Elsevier Inc. All rights reserved.
查看更多>>摘要：Distributed system has been widely used to solve massive data analysis tasks. This article targets on quantile regression on distributed system with non-randomly distributed mas-sive data, and proposes a new communication-efficient surrogate quantile regression. Specifically, based on a small size random pilot sample collected from different worker machines, we approximate the global quantile regression as a surrogate one on the master machine, which relates to the local datasets only through their gradient vectors, and can overcome the non-randomly distributed nature. Then the resulting estimator can be obtained on the master, and the communication cost is greatly reduced, since the pilot sample and local gradients can be transferred conveniently. In theory, without any restric-tive assumption about randomness, the established asymptotical results show that the proposed method works beautifully just as the data were stored on one single machine. Synthetic data and real world data evaluations are also used to illustrate the proposed method.(c) 2021 Elsevier Inc. All rights reserved.
查看更多>>摘要：In this paper, the quantization level based event-triggered control algorithm is novelly proposed and stability of the considered system could be guaranteed on account of hysteresis quantizer under measurement uncertainties. By carefully exploring the characteristics of nonlinear networked control systems, hysteresis quantizer is selected to avoid potential chatterings caused by frequent quantization level transitions. Two kinds of situations, one is networked control system without considering measurement uncertainties, and the other takes measurement uncertainties into consideration, are separately discussed under the quantization level based event-triggered control framework. By thoroughly digging out the relationship between quantizer structure and event-triggered control mechanism, two quantization level based event-triggered control mechanisms are put forward under different situations. Moreover, no matter whether measurement uncertainties are taken into account, system stability could always be guaranteed by our proposed criteria and Zeno behavior would never happen with our method. Finally, simulations show the feasibility and superiority of our proposed algorithms.(c) 2021 Published by Elsevier Inc.
NSTL
Elsevier