Pre-trained encoder-decoder models are widely applied in Task-Oriented Dialog (TOD) systems on the session level, mainly focusing on modeling the dialog semantic information. Dialogs imply structural information indicating the interaction among user utterances, belief states, database search results, system acts and responses, which is also crucial for TOD systems. In addition, for the system acts, additional pre-training and datasets are considered to improve their accuracies, undoubtedly introducing a burden. Therefore, a novel end-to-end TOD system named Winnie is proposed in this paper to improve the TOD performance. First, to make full use of the intrinsic structural information, supervised contrastive learning is adopted to narrow the gap in the representation space between text representations of the same category and enlarge the overall continuous representation margin between text representations of different categories in dialog context. Then, a system act classification task is introduced for policy optimization during fine-tuning. Empirical results show that Winnie substantially improves the performance of the TOD system. By introducing the supervised contrastive and system act classification losses, Winnie achieves state-of-the-art results on benchmark datasets, including MultiWOZ2.2, In-Car, and Camrest676. Their end-to-end combined scores are improved by 3.2, 1.9, and 1.1 points, respectively.

Load frequency control (LFC) of modern power systems depends on open communication networks to transmit measurement/control signals, which may make the LFC scheme vulnerable to cyberattacks. Therefore, it is necessary to ensure the resilience of LFC systems against cyberattacks so as to detect the attack and maintain the stable operation of the system during the attack. This article aims to investigate the attack detection and mitigation scheme of LFC systems against false data injection attacks. A model-based unknown input observer (UIO) is designed to simultaneously estimate the system state and attack signal based on load prediction, where the attack estimation is utilized to achieve real-time attack detection. Moreover, we present a sufficient and necessary condition for the existence of the UIO and a design method of the UIO based on linear matrix inequalities. Based on the attack estimation, a resilient LFC scheme is proposed to mitigate the attack impact and restore the frequency and tie-line power within an accepted range during the attack. Finally, case studies are undertaken by the IEEE 9- and 39-bus test systems to show the effectiveness of the proposed scheme.

Acting as a key to future environmentally friendly transportation systems, electric vehicles (EVs) have attached importance to develop fast charging technologies to accomplish the requirement of vehicle users. However, fast charging behaviors would cause degradations in EVs' batteries, as well as negative effects like new demand peak and feeder overloads to the connected distribution network, especially when plugging in large scale EVs. Decentralized coordination is encouraged and our goal is to achieve an optimal strategy profile for EVs in a decentralized way considering both the need of fast charging and reducing degradations in batteries and the distribution network. In this article, we innovatively model the EV fast charging problem as an optimization coordination problem subject to the coupled feeder capacity constraints in the distribution network. The need of fast charging is expressed by the total charging time, and the relative tendency to fully charge within the desired time period. We introduce a ℓ0-norm of the charging strategy which is non-convex to represent the total charging time, and apply the ℓ1-norm minimization to approximate the sparse solution of ℓ0-norm minimization. The shorter the charging horizon is the stronger willing of fast charging the user has. The objective of the optimization problem tradeoffs the EVs' battery degradation cost, the load regulation in the distribution network, the satisfaction of charging and the total charging time, which is non-separable among individual charging behaviors. Even though alternating direction method of multipliers (ADMM) has been widely applied in distributed optimization with separable objective and coupled constraints, its decentralized scheme cannot be applied directly to the underlying non-separable EV charging coordination problem. Hence, a hierarchical algorithm based on ADMM is proposed such that the convergence to the optimal strategies is guaranteed under certain step-size parameter. Furthermore, a receding horizon based algorithm is proposed considering the forecast errors on the base demand and the EV arrival distribution. The results are demonstrated via some simulation results.

This article studies the problem of the fuzzy H∞filter design for nonlinear networked control systems through eventtriggered communication (ETC) scheme. First, a novel adaptive ETC scheme is given to determine whether the sampled measurement output should be released to communication network or not. Consequently, less communication resources are occupied under the desired H∞performance. Second, augmented fuzzy lineintegral Lyapunov function is introduced in the H∞performance analysis of filter error systems, such that the information of time derivative of membership functions are fully considered to reduce the conservativeness of networked fuzzy filter design. Different from the existing results, the upper bounds of time derivative of membership functions need not to be known prior. Third, the resulting filter error system is modeled as time-delay system under ETC mechanism and asynchronous premise in a unified framework. As a result, applying Lyapunov theory and inequality technique, new sufficient condition is obtained to meet the H∞performance for the filter error systems. Further, the corresponding filter and event-triggering parameters are codesigned and solved by a set of linear matrix inequalities. Finally, two examples are offered to demonstrate the advantage of the proposed method.

We focus on the management of power electronic loads (PELs) in direct current (DC) distribution networks due to their promising development. We first build a voltage model of PELs based on which we formulate a demand side management optimization problem that coordinates the bus voltage by minimizing the system operation cost. However, the underlying optimization problem is non-convex. In order to solve this non-convex problem, we first reformulate it as a difference of convex programming problem, and then propose a novel algorithm based on branch and bound to implement the optimal solution. It is shown that the system is guaranteed to converge to the global optimum under the proposed method.

We study how to efficiently allocate the infinitesimal divisible resource under auction mechanism in a dynamic way. We propose a Vickrey-Clarke-Groves-type auction mechanism with a 2-D bid which specifies a per unit price and a maximum of the demand. Due to the absence of enough information related to the infinite dimensional valuations of individual players in a single-bid strategy, it is challenging to implement the efficient Nash equilibrium (NE) in a dynamic way. In this paper, we introduce a pair of parameters related to players' valuations, and design a decentralized dynamic process assisted with this pair of values, such that at each iteration, a single player updates its best bid under a constrained set of demand. Under the proposed auction mechanism, we show the incentive compatibility, efficiency, and uniqueness of the NE. Furthermore, our method is guaranteed to converge to the efficient NE, and it presents the enhanced convergence performance compared with those methods proposed in the literature. Case studies are given to demonstrate the results developed in this paper.

This work is concerned with large-scale incorporation of inter-temporal shifting loads into an electricity markets over multi time intervals. In the context of deregulated market structures with deep concerns for the social efficiency, Demand Response (DR) is promising to enable interaction of end-users with the grid. We design a decentralized dynamic pricing mechanism for DR based on auction rules, to induce users to behave efficiently and obtain system wide benefit. The system implements an optimal allocation with respect to the submitted bid profile, and broadcasts an associated system price to all the users. The expected system price dictates users' best response, on the other hand, the resulting load pattern will alter this price. Moreover, we propose a dynamic algorithm to implement the efficient results such that all the users simultaneously update their best responses with respect to the threshold price set in last step. We specify the sufficient conditions for the proposed method, under which the system converges to the efficient solution.