The advent of deep learning and neural network has revolutionized the study of time series forecasting. Diverse forecasting networks seem to achieve more promising performances than traditional forecasting models especially when handling complicated and non-linear conditions. However, most of forecasting networks are built upon Seq2Seq model, which means that they pursue the unique forecasting result when given certain input sequence. However, due to the natural noises, tiny errors and distribution shifts in real-world time series, Seq2Seq models are vulnerable to over-fitting problem. Based on these observations, we propose Area2Area forecasting formula containing C ausal S equence-wise C ontrastive L earning ( CSCL ) and A rea L oss ( AL ) mechanisms to loosen the forecasting constraint and alleviate over-fitting problem. CSCL utilizes contrastive learning technique to transform the input sequence into an input Area while AL modifies the loss function to transform the prediction sequence into a prediction Area . We additionally propose a novel encoder network Temporal E fficient L ayer A ggregation N etwork ( Temporal ELAN ). Extensive experiments on six datasets and nine baselines demonstrate that Area2Area forecasting is literally capable of alleviating the over-fitting problem of existing Seq2Seq forecasting networks. The source code is released on https://github.com/OrigamiSL/A2A .

The utilization of multiple unmanned aerial vehicles (multi-UAVs) greatly enhances the quality of public services, whereas presents challenges to effective mission planning. Existing research predominantly concentrates on task allocation and path planning, with limited emphasis on integrating UAV deployment and task allocation. Despite the promise of learning-based approaches for joint optimization, their applicability to sim-to-real problems remains restricted. In this paper, we propose a hierarchical decision-making framework empowered by digital twin (DT) technology to enhance resource-efficient utilization and facilitate realtime mission planning. In the upper layer, a swarm of UAVs is divided into subgroups using a K-means method and strategically deployed to corresponding task regions. In the lower layer, a DT-enabled reinforcement learning approach is adopted to leverage real-world experiences. Specifically, we employ a multiagent deep Q learning algorithm to optimize the training process, where each agent interacts with the DT environment through original and target networks to acquire an optimal strategy. The proposed approach is evaluated using simulations that demonstrate its effectiveness in achieving high-quality plans.

This paper presents FDNet: a Focal Decomposed Network for efficient, robust and practical time series forecasting. We break away from conventional deep time series forecasting formulas which obtain prediction results from universal feature maps of input sequences. In contrary, FDNet neglects universal correlations of input elements and only extracts fine-grained local features from input sequence. We show that: (1) Deep time series forecasting with only fine-grained local feature maps of input sequence is feasible upon theoretical basis. (2) By abandoning global coarse-grained feature maps, FDNet overcomes distribution shift problem caused by changing dynamics of time series which is common in real-world applications. (3) FDNet is not dependent on any inductive bias of time series except basic auto-regression, making it general and practical. Moreover, we propose focal input sequence decomposition method which decomposes input sequence in a focal manner for efficient and robust forecasting when facing Long Sequence Time series Input (LSTI) problem. FDNet achieves competitive forecasting performances on six real-world benchmarks and reduces prediction MSE by 38.4% on average compared with other thirteen SOTA baselines. The source code is available at https://github.com/OrigamiSL/FDNet.

This paper develops a novel approach to addressing the distributed situation awareness (SA) consensus problem for unmanned aerial vehicle (UAV) swarm systems. SA consensus is an important condition for gaining decision-making superiority. However, because of the complexity and antagonism of the mission environment, the widely employed centralized architectures are unable to reach a distributed consensus. We address this urgent issue by proposing a systematic distributed SA consensus scheme, including a distributed optimization-based consensus reaching model, dual-loop decision-making framework, and novel coordination algorithm. Necessarily, we conduct convergence analysis on the proposed algorithm. The effectiveness and superiority of the proposed method are verified by comparative simulations.