Previous cross-lingual knowledge graph (KG) alignment studies rely on entity embeddings derived only from monolingual KG structural information, which may fail at matching entities that have different facts in two KGs. In this paper, we introduce the topic entity graph, a local sub-graph of an entity, to represent entities with their contextual information in KG. From this view, the KB-alignment task can be formulated as a graph matching problem; and we further propose a graph-attention based solution, which first matches all entities in two topic entity graphs, and then jointly model the local matching information to derive a graphlevel matching vector. Experiments show that our model outperforms previous state-of-theart methods by a large margin.
It is a challenging task to automatically compose poems with not only fluent expressions but also aesthetic wording. Although much attention has been paid to this task and promising progress is made, there exist notable gaps between automatically generated ones with those created by humans, especially on the aspects of term novelty and thematic consistency. Towards filling the gap, in this paper, we propose a conditional variational autoencoder with adversarial training for classical Chinese poem generation, where the autoencoder part generates poems with novel terms and a discriminator is applied to adversarially learn their thematic consistency with their titles. Experimental results on a large poetry corpus confirm the validity and effectiveness of our model, where its automatic and human evaluation scores outperform existing models.
In this paper, a distributed output feedback model predictive control (OFMPC) algorithm is presented for the polytopic uncertain system subject to randomly occurring actuator saturation and packet loss. Compared with the intensively applied state feedback control in MPC, the OFMPC is more feasible to the real world because the system states are often unmeasurable. With taking both actuator saturation and packet loss into account, the presented OFMPC algorithm is more practical. Moreover, by splitting the controller inputs into two independent parts, the presented dynamic output feedback control (DOFC) strategy provides more freedom to the controller design. With the global system decomposed into some subsystems, the computation complexity is reduced, thus the online designing time can be saved. By defining the estimation error function and forming an augmented system to handle the DOFC and by transforming the nonlinear feedback law into a convex hull of linear feedback laws, the distributed controllers are obtained by solving a linear matrix inequality (LMI) optimization problem. Finally, some simulation examples are employed to show the effectiveness of the techniques proposed in this paper.
3037Note that the results by OFMPC strategy provided earlier are mostly on centralized MPC. However, with the rapid development of the science technology and the network application, the systems to be handled are becoming more and more complex. It is often infeasible to apply the classical centralized MPC control solution to the practical systems accounting for the computational complexity and communication bandwidth limitations. Thus, a distributed MPC strategy has attracted much attention to reduce the computation time and the network transmission burden recently [11,12]. In a distributed MPC framework, each subsystem is controlled by an independent controller, all subsystems communicate with each other via network, and the technical target is to achieve the global performance of entire system. In this way, the system with distributed MPC has better error-tolerance and robustness compared with the traditional centralized MPC. Mercangöz et al. proposed an iterative implementation of a distributed MPC scheme and applied it to a four-tank system [13]. In [14], a class of linear systems coupled with neighbors through inputs was discussed and the corresponding distributed MPC algorithms were presented for the cooperative subsystems. A robust distributed MPC problem was discussed for a system with polytopic uncertainty in [15] and [16], where the model uncertain problems were casted into an LMI optimization problem.In [17], Zhang et al. presented a distributed MPC algorithm for the polytopic system subject to inputs saturation.On the other hand, the obstacles in delivering are often owing to the physical limitations of system components, of which the most commonly encountered one stems from the saturation that occurs in actuators. Saturation brings in nonlinear characteristics that can severely restrict the amount of transmitted...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.