With the expansion of power grids and incorporation of new energy sources, fault and attack detection in power monitoring systems has become increasingly vital for maintaining grid stability and safety. A significant ...
详细信息
ISBN:
(纸本)9798350387780;9798350387797
With the expansion of power grids and incorporation of new energy sources, fault and attack detection in power monitoring systems has become increasingly vital for maintaining grid stability and safety. A significant challenge for data-driven detection methods is the scarcity of labeled measurement data, which critically hampers their detection performance. This paper proposes a novel data-driven framework, SuPer, for detecting faults and attacks in power monitoring systems. SuPer employs a novel combination of self-adaptive thresholding-based pseudolabeling and prototype-based data purification strategies, which have superior detection performance under conditions of scarcity labeled data. SuPer generates reliable pseudo-labels for massive unlabeled measurements and selects the high-confidence ones to participate in the training process, which enhances the machine learning model's ability to identify anomalies accurately. Extensive experiments are conducted in a power system framework to demonstrate SuPer's effectiveness, particularly highlighting its improved performance in scenarios with limited labeled data.
This paper explores the problem of quantitative trading based on stock data from a data-driven perspective and develops a new model that relies deeply on buy, sell and hold behaviour to quantify and optimise investmen...
详细信息
This paper addresses the problem of low fault estimation accuracy in Lipschitz nonlinear systems with external disturbances, presenting an innovative strategy based on a Reinforcement Iterative learning (RIL) framewor...
详细信息
An improved path planning algorithm based on RRT∗ is proposed for the manipulator to generate a collision-free path for end-effector in this paper, which maximizes the manipulator's manipulability throughout the e...
详细信息
Wind power generation technology is one of the research hotspots of renewable energy nowadays. In order to ensure the stable and reliable operation of wind power generation equipment, wind speed prediction is very imp...
详细信息
ISBN:
(纸本)9798350321050
Wind power generation technology is one of the research hotspots of renewable energy nowadays. In order to ensure the stable and reliable operation of wind power generation equipment, wind speed prediction is very important. This paper provides a new idea for the wind speed prediction based on Autoregressive Moving Average (ARMA) and Support Vector Regression (SVR). First, to reasonably divide the original data into multiple historical data with strong correlation as features to predict the future wind speed, the ARMA model is employed and its partial autocorrelation coefficient is calculated. By this means, the input features can be optimally selected and the training set of the prediction model can be constructed. Further, SVR model is used to build the nonlinear relationship between the input features and future wind speed. Finally, through simulation, it proves that this method saves more time than try and error method in selecting input features, and through comparison with Backpropagation Neural Network (BPNN), it proves that this method can achieves higher wind speed prediction accuracy.
Industrial aerodynamic system (IAS) involves a series of compressors to provide the compressed air for consumer of steel industrial park, which consumes a substantial amount of electrical energy and attracts a conside...
详细信息
A communication fault diagnosis and positioning scheme is designed according to the characteristics of the distributed control system of offshore wind turbine. Firstly, to simplify the communication link between distr...
详细信息
Advances in the field of reinforcement learning (RL)-based drive control allow formulation of holistic optimization goals for the data-driven training phase. The resulting controllers feature efficient drive operation...
详细信息
Advances in the field of reinforcement learning (RL)-based drive control allow formulation of holistic optimization goals for the data-driven training phase. The resulting controllers feature efficient drive operation without the necessity of an a priori known plant model but, so far, conduction of the corresponding training phase in real-world drive systems has been applied only sparsely due to safety concerns. This contribution targets the challenging problem of self-learning torque control for a permanent-magnet synchronous motor assuming a finite control set, i.e., the direct selection of switching actions instead of a modulator-based setup. In order to allow a secure and effective online training with real-world drive systems, the RL controller is monitored by a safeguarding algorithm that prevents application of unsafe switching actions, e.g., such that result in overcurrent. The accruing amount of measurement data is handled with the use of an edge-computing pipeline to outsource the RL training from the embedded control hardware. The inference of the utilized artificial neural network in hard real time is realized with the use of a reconfigurable field-programmable gate array architecture. The resulting RL-based algorithm is able to learn a torque control policy in just 10 min, which has been validated during comprehensive real-world experiments.
Teaching system identification for control to aerospace engineering students can significantly benefit from actual experiential learning activities in a laboratory environment. In this paper, we present the experience...
详细信息
ISBN:
(纸本)9798350370959;9798350370942
Teaching system identification for control to aerospace engineering students can significantly benefit from actual experiential learning activities in a laboratory environment. In this paper, we present the experience we gathered performing hands-on project-based experimental activities conducted with a multirotor Unmanned Aerial Vehicle (UAV) within a graduate-level course teaching how to learn linear models of dynamical systems from data. We developed the activities to complement face-to-face lectures to illustrate the steps involved in the process of identifying data-driven models in a hands-on manner, and implemented them in the form of a project during the last 8 weeks of the course, to consolidate the knowledge acquired in the theoretical lectures through practical experiences. We present the lecture material we prepared, provide an overview of the obtained results, discuss the pedagogical value of the experiential learning activities, summarize student's feedback and give an outline for further improvements.
Time delays exist in a large number of practical systems, and it is meaningful to study the robust performance improvement of time-delay systems. This paper proposes an additive-state-decomposition-based robust perfor...
详细信息
暂无评论