A5101896167- Power System Optimization and Stability
- Smart Grid and Power Systems
- Power Systems and Renewable Energy
- Advanced Measurement and Detection Methods
- Power Systems and Technologies
- Recommender Systems and Techniques
- Optimal Power Flow Distribution
- Artificial Intelligence in Healthcare
- Robotic Path Planning Algorithms
- Sparse and Compressive Sensing Techniques
- Smart Grid Energy Management
- Data Management and Algorithms
- Target Tracking and Data Fusion in Sensor Networks
- Energy Load and Power Forecasting
- Multimodal Machine Learning Applications
- Globalization, Economics, and Policies
- Thermal Analysis in Power Transmission
- Robotics and Sensor-Based Localization
- Evaluation and Optimization Models
- Machine Learning and ELM
- Enhanced Oil Recovery Techniques
- IoT-based Smart Home Systems
- Advanced Research in Science and Engineering
- Optical Systems and Laser Technology
- Advanced Multi-Objective Optimization Algorithms
State Grid Corporation of China (China)
2024
University of Manchester
2023
Second Affiliated Hospital of Nanjing Medical University
2022
Tsinghua University
2016-2019
Stanford University
2019
Xidian University
2016-2017
University of Maryland, College Park
2002
We present a navigation system that combines ideas from hierarchical planning and machine learning. The uses traditional global planner to compute optimal paths towards goal, deep local trajectory velocity controller motion commands. latter components of the adjust behavior robot through attention mechanisms such it moves avoids obstacles, respects space nearby pedestrians. Both structure proposed models use make system's execution interpretable. Our simulation experiments suggest...
Abstract The 3D physical properties of porous rocks directly determine the subsurface flow and modeling. However, predicting a wide range remains formidable challenge requires large amount input. Reliable microstructure‐property correlations can accurately predict properties, avoiding time‐consuming experimental testing. Here, we propose new dimensionless convolutional neural network (CNN)‐based method to find inherent for using only single 2D slice or series slices. Training testing were...
With the rapid development of new power systems, usage stations are becoming more diverse and complex. Fine-grained management demand-side load has become increasingly crucial. To address accurate forecasting needs for various consumption types provide data support in stations, this study proposes a sequence noise reduction method. Initially, wavelet is performed on multiple sequences collected by system. Subsequently, northern goshawk optimization employed to optimize parameters variational...
The current statistical model is one of most effective method for maneuvering target tracking. It's the mean acceleration and it reflects variation target's rang intensity more practically . However, frequency α , which represents maneuvering, needs to be predefined won't changed with targets maneuvering. That makes effect tracking system drop dramatically. Based on analysis kinetic characteristic target, a adaptive algorithm was proposed. built parameter adjusted adaptively At last,...
Location recommendation has attracted increasing attention in recent years. This paper proposes a novel multi-objective framework for location based on user preference. Under this framework, preference can be separated into common and individual Then two contradictory objective functions are designed to describe these kinds of preferences. It is difficult optimize simultaneously. In paper, evolutionary algorithm proposed functions. The make good balance between Experiments real application...
This paper introduces a robust sparse recovery model for compressing bad data and state estimation (SE), based on revised multi-stage convex relaxation (R-Capped-L1) model. To improve the calculation efficiency, fast decoupled solution is adopted. The proposed method can be used three-phase unbalanced distribution networks with both phasor measurement unit remote terminal measurements. robustness computational efficiency of R-Capped-L1 are compared some popular SE methods by numerical tests...
This paper proposes a robust three-phase state-estimation method that is applicable to both balanced transmission networks and unbalanced distribution with nonlinear measurements. We extend the recently introduced bilinear formulation of state estimation problems model. The original solved by resorting two auxiliary linear model transformation in between. Bad measurements are further suppressed introducing Numerical tests on IEEE benchmark systems, including network network, demonstrate...
This paper investigates the sparse recovery models for bad data detection and state estimation in power networks. Two models, L1-relaxation model (L1-R) multi-stage convex relaxation (Capped-L1), are compared with weighted least absolute value (WLAV) aspects of processing capacity computational efficiency. Numerical tests conducted on systems linear nonlinear measurements. Based numerical tests, evaluates performance these robust models. Furthermore, suggestion how to select parameter is...
This paper investigates the sparse recovery models for bad data detection and state estimation in power networks. Two models, L1-relaxation model (L1-R) multi-stage convex relaxation (Capped-L1), are compared with weighted least absolute value (WLAV) aspects of processing capacity computational efficiency. Numerical tests conducted on systems linear nonlinear measurements. Based numerical tests, evaluates performance these robust mod-els. Furthermore, suggestion how to select parameter is...
The distribution network line loss rate is a crucial factor in improving the economic efficiency of power grids. However, traditional prediction model has low accuracy. This study proposes predictive method based on data preprocessing and integration to improve Data employs dynamic cleaning technology with machine learning enhance quality. Model combines long short-term memory (LSTM), linear regression, extreme gradient boosting (XGBoost) models achieve multi-angle modeling. regression...
Factory and facility automation has been widely implemented since the revolution of computers. A variety automatic material processing machines handling equipment have designed manufactured for various applications. The layout optimization is one major tasks in designing automated systems. This paper proposes a generic methodology that provides systematic way to design an optimal all types facilities. quantitative criteria are addressed first. An connectivity among then solved through...
In recent years, China's electricity market reform has continued to deepen, and the construction of been accelerating. According national power pilot policy, southern regional will start with Guangdong spot market, then gradually develop market. this paper, realization method integration is analysed, several paths for provincial into are put forward according characteristics Southern system.
The theory of component value analysis and lines were introduced in this paper, we can get a comprehensive description the product model as system by identify inadequacies system‥ We weakest link components analysis, so to provide direction solve problem improving design. application innovation design was illustrated with an engineering instance future
The affect of Jerk model in maneuvering target tracking is discussed this paper. In order to deal with these problems, an improved algorithm combining the Model Error Predictive filter proposed. This can avoid limitation that assumes process noise as Gaussian white noise. Meanwhile, estimates error online and then amends it. caused mismatches real motion decreased obviously. Extended Kalman Filter has problems complicated calculation, low accuracy convergence estimating nonlinear system. To...
To realize the grid security checking of large consumer direct electric quantity purchasing plan, a monthly method based on partitions is presented in paper. In method, power divided into containing certain number units and loads, decomposition daily models with constraints partition being considered are built. That flow tie line less than allowed stability limit value one trading different partitions. The deviation reflected form slack variables constraint conditions, minimum variable used...
To achieve the grid-connected operation performances evaluation of photovoltaic power stations, a performance index system and comprehensive method on station is presented. The rank sum ratio each to reflect calculated by using rank-sum method. According values, evaluated. actual monitoring data stations are used calculate indexes, verified. obtained results show that can be correctly evaluated presented