A5026840397- Microstructure and Mechanical Properties of Steels
- Metal Alloys Wear and Properties
- Adversarial Robustness in Machine Learning
- Hydrogen embrittlement and corrosion behaviors in metals
- Smart Grid Energy Management
- Metallurgy and Material Forming
- Tuberculosis Research and Epidemiology
- Anomaly Detection Techniques and Applications
- Privacy-Preserving Technologies in Data
- Mycobacterium research and diagnosis
- Energy Load and Power Forecasting
- Microstructure and mechanical properties
- Microgrid Control and Optimization
- Advanced Neural Network Applications
- Bacillus and Francisella bacterial research
- Building Energy and Comfort Optimization
- High Temperature Alloys and Creep
- Optimal Power Flow Distribution
- Electric Vehicles and Infrastructure
- Power System Optimization and Stability
- Membrane Separation Technologies
- Energy Efficient Wireless Sensor Networks
- Underwater Vehicles and Communication Systems
- Material Properties and Failure Mechanisms
- Speech and Audio Processing
China Iron and Steel Research Institute Group
2013-2025
Shandong University
2025
Yantai Academy of Agricultural Sciences
2023-2024
Institute of Plant Protection
2024
Xinjiang University
2024
Harbin Institute of Technology
2024
Shanxi Agricultural University
2024
Beijing Information Science & Technology University
2019-2024
Chongqing University
2021-2023
NARI Group (China)
2018-2023
Volt-VAR control is critical to keeping distribution network voltages within allowable range, minimizing losses, and reducing wear tear of voltage regulating devices. To deal with incomplete inaccurate models, we propose a safe off-policy deep reinforcement learning algorithm solve problems in model-free manner. The problem formulated as constrained Markov decision process discrete action space, solved by our proposed soft actor-critic algorithm. Our achieves scalability, sample efficiency,...
Dynamic distribution network reconfiguration (DNR) algorithms perform hourly status changes of remotely controllable switches to improve system performance. The problem is typically solved by physical model-based control algorithms, which not only rely on accurate parameters but also lack scalability. To address these limitations, this paper develops a data-driven batch-constrained reinforcement learning (RL) algorithm for the dynamic DNR problem. proposed RL learns policy from finite...
The microstructural evolution of Fe–0.2C–5Mn steel during intercritical annealing with holding time for up to 144 hours was examined by TEM and STEM. It demonstrated that the martensite lath structure gradually transformed into a lamellar ferrite austenite duplex structure. partitioning manganese from found Typical Kurdjumov-Sachs orientation relationship between observed electron back scattered diffraction (EBSD). Based on analysis thckening behavior, it proposed Mn-partitioning in...
Providing ride-hailing services with electric vehicles can help reduce greenhouse gas emissions and solve the last mile problem. This paper develops a reinforcement learning based algorithm to operate community owned vehicle fleet, which provides local residents. The goals of operating fleet are minimize customer waiting time, electricity cost, operational costs vehicles. A novel framework characterized by decentralized centralized decision making is proposed dispatch process allows...
The hydrogen embrittlement susceptibility of a newly developed 1700 MPa-grade ultra-high-strength steel with primary austenite grain size 4 μm was studied and the mechanical properties microstructure were characterized. results show that content in increases extension charging time: value reached 0.35 wppm time 96 h. On contrary, fracture mode experimental remained ductile after charging, elongation section shrinkage showed little difference, indicating an excellent resistance to...
A large amount of energy is wasted through inefficient operation heating, ventilation, and air conditioning (HVAC) system due to the lack reliable building occupancy measurement prediction. To mitigate this problem, an innovative change-point logistic regression model developed provide accurate forecast occupancy. novel HVAC control algorithm then by embedding prediction into predictive (MPC) framework. The occupancy-based MPC tries minimize electricity consumption maximize occupants’...
Pre-trained general-purpose language models have been a dominating component in enabling real-world natural processing (NLP) applications. However, pre-trained model with backdoor can be severe threat to the Most existing attacks NLP are conducted fine-tuning phase by introducing malicious triggers targeted class, thus relying greatly on prior knowledge of task. In this paper, we propose new approach map inputs containing directly predefined output representation models, e.g., for...
Federated Learning (FL), the de-facto distributed machine learning paradigm that locally trains datasets at individual devices, is vulnerable to backdoor model poisoning attacks. By compromising or impersonating those an attacker can upload crafted malicious updates manipulate global with behavior upon attacker-specified triggers. However, existing attacks require more information on victim FL system beyond a practical black-box setting. Furthermore, they are often specialized optimize for...
Online power system event identification and classification are crucial to enhancing the reliability of transmission systems. In this paper, we develop a deep neural network (DNN) based approach identify classify events by leveraging real-world measurements from hundreds phasor measurement units (PMUs) labels thousands events. Two innovative designs embedded into baseline model built on convolutional networks (CNNs) improve accuracy. First, propose graph signal processing PMU sorting...
ObjectivesWe carried out a randomized multicentre study in China to investigate whether the clofazimine would improve efficacy of standardized regimen patients with multidrug-resistant tuberculosis (MDR-TB).MethodsPatients MDR-TB managed 17 TB specialist hospitals between September 2009 and 2011 were randomly assigned treatment groups at enrolment. In intervention group, 100 mg per day was added regimen. The primary outcome proportion successful outcomes.ResultsFrom 156 that screened, 74...
Smoke produced by wildfires is usually visible much earlier than flames. Hence, early detection of wildfire smoke essential to prevent severe property losses and heavy casualties from catastrophic wildfires. Camera networks are being built expanded achieve timely detection. To the best camera coverage accuracy with limited budget, an intelligent video algorithm optimal placement strategy in a critical need. In this paper, we propose efficient framework designed for embedded applications on...
A key factor in big data analytics and artificial intelligence is the collection of user from a large population. However, comes at price privacy risks, not only for users but also businesses who are vulnerable to internal external breaches. To address issues, local differential (LDP) has been proposed enable an untrusted collector obtain accurate statistical estimation on sensitive (e.g., location, health, financial data) without actually accessing true records. As key-value extremely...
The 2200 MPa low-alloy ultra-high-strength steels microalloyed with Ti (Ti-steel) and Nb (TiNb-steel) were quenched at 860 °C tempered 180 °C, the mechanical properties, microstructure, precipitated phase studied by SEM, TEM, physicochemical analysis. results of properties showed that TiNb-steel had higher strength than Ti-steel, a yield 1746 1802 tensile 2198 2232 MPa, respectively. finer structure effective grain sizes being 0.86 μm 1 μm, more MC-type carbides carbide contents 0.24 wt.%...
Acute pesticide poisoning is a major public health concern. The relationship between family function and depression in caregivers of patients with acute unclear. Therefore, this study aimed to explore the function, coping style, among provide theoretical basis intervention targets for future research. A cross-sectional was conducted Department Toxicology Occupational Disease Grade hospital Jinan, Shandong Province, from November 2022 June 2023. general data questionnaire, caring index scale,...
Large Language Models (LLMs) have achieved significant performance in various natural language processing tasks but also pose safety and ethical threats, thus requiring red teaming alignment processes to bolster their safety. To effectively exploit these aligned LLMs, recent studies introduced jailbreak attacks based on multi-turn dialogues. These aim prompt LLMs generate harmful or biased content by guiding them through contextual content. However, the underlying reasons for effectiveness...