In power markets, SO can maintain system security by preventively restraining the energy transactions or, if having procured sufficient reserves, by dispatching the reserves correctively after any contingency happens. The traditional security constrained unit commitment (SCUC) or economic dispatch (SCED), being with all the contingency constraints, probably leads to lower market efficiency. In this paper, a security coordinated economic dispatch (SCoED) model is presented for day-ahead energy and reserve joint market, which is formulated as a Stackelberg leader-follower problem, the upper level determining the energy and reserve procurement schedules subjecting to only the normal constraints while in the lower level, economic dispatch of the procured reserves is executed for every possible N-1 contingency to maintain system security. A numerical example is carried out on a 6-bus system to compare the energy market efficiency and reserve procurement cost under the traditional SCED and the SCoED.

A health assessment and fault prediction method for wind turbine generators is proposed in this article. In health assessment module, considering generator status transferring along with environment and wind turbine–self operating, variables under wind turbine normal working are divided into two parameter spaces and recognized, namely operating conditions and status parameters. Then generator health benchmark models based on Gaussian mixture model are established in different operating condition sub-spaces after the data imbalance problem solved. For online health assessment, health deterioration index based on condition recognition models is calculated and a dual-threshold alarm scheme is proposed. When an alarm is raised by degraded health deterioration index, the program could access fault prediction module, where the generator rear bearing temperature trend and fault remaining time can be predicted through weights redistribution and hyper-parameter optimized support vector regression. In experiments, the proposed health assessment and fault prediction was verified in a real wind farm, and results showed this method could assess generator condition accurately and improve special fault prediction performance.

Deep reinforcement learning (DRL) has been widely applied in autonomous exploration and mapping tasks, but often struggles with the challenges of sampling efficiency, poor adaptability to unknown map sizes, and slow simulation speed. To speed up convergence, we combine curriculum learning (CL) with DRL, and first propose a Cumulative Curriculum Reinforcement Learning (CCRL) training framework to alleviate the issue of catastrophic forgetting faced by general CL. Besides, we present a novel state representation, which considers a local egocentric map and a global exploration map resized to the fixed dimension, so as to flexibly adapt to environments with various sizes and shapes. Additionally, for facilitating the fast training of DRL models, we develop a lightweight grid-based simulator, which can substantially accelerate simulation compared to popular robot simulation platforms such as Gazebo. Based on the customized simulator, comprehensive experiments have been conducted, and the results show that the CCRL framework not only mitigates the catastrophic forgetting problem, but also improves the sample efficiency and generalization of DRL models, compared to general CL as well as without a curriculum. Our code is available at https://github.com/BeamanLi/CCRL_Exploration.