作者：
Qing SongXiaofan WangDepartment of Automation
Shanghai Jiao Tong University and Key Laboratory of System Control and Information Processing Ministry of Education of China Shanghai 200240 China
Abstract The problem of efficient path computation arises in several applications such as intelligent transportation and network routing. Although various algorithms exist for computing shortest paths, their heavy pre...
Abstract The problem of efficient path computation arises in several applications such as intelligent transportation and network routing. Although various algorithms exist for computing shortest paths, their heavy precomputation/storage costs and/or query costs hinder their application to real-time routing. By detecting hierarchical community structure in road networks, we develop a community-based hierarchical graph model that supports efficient path computation on large road networks. We then propose a new hierarchical routing algorithm that can significantly reduce the search space over the conventional algorithms with acceptable loss of accuracy. Finally, we evaluate our approach experimentally using a large, real-world road network to show the gain in performance.
作者：
Zhi ZhangLi-Sheng HuDepartment of Automation
Shanghai Jiao Tong University and Key Laboratory of System Control and Information Processing Ministry of Education of China Shanghai 200240 China
Abstract During the last two decades, performance assessment for controlsystems has been receiving wide attention. The key step is to describe and estimate the benchmark. However, the issue for more complex control s...
Abstract During the last two decades, performance assessment for controlsystems has been receiving wide attention. The key step is to describe and estimate the benchmark. However, the issue for more complex controlsystems still remains open. This paper is concerned primarily with the estimation of the performance benchmark in the tensor space, and explores to extend traditional methods to the complex system. An estimation procedure based on higher-order singular value decomposition is proposed. Numerical examples shows the effectiveness of the proposed method.
Global Positioning system (GPS)-equipped probe vehicles have been one of the principal techniques in detection of traffic flow. In order to guide the application of the different methods using data from GPS probe vehi...
详细信息
In this paper, we address the design of decentralized controller for connectivity-preserving flocking without velocity measurement. An output vector based on neighbors' position information alone is constructed to...
详细信息
Market clearing price (MCP) forecasting techniques is very important for the development of the electricity market. A three-layered neural network is used to predict electricity prices. MCP is seen as a multi-input si...
详细信息
Market clearing price (MCP) forecasting techniques is very important for the development of the electricity market. A three-layered neural network is used to predict electricity prices. MCP is seen as a multi-input single-output system and the historical electricity price and load data is utilized in an electricity market. The neural network is based on Minimum Entropy Error (MEE) cost function and Batch-Sequential mode. Compared with other models, the proposed approach improves the prediction accuracy and speed.
Thermal power plants constitute the largest proportion of installed capability in global power generation system, which consumes large quantities of coal. Therefore achieving optimal operation of thermal power units, ...
详细信息
The problem of real-time rectangle detection on high-resolution image arises in actual panel production. This paper proposes a robust real-time method for panel rectangle detection. In line extraction part, a new tech...
详细信息
The problem of real-time rectangle detection on high-resolution image arises in actual panel production. This paper proposes a robust real-time method for panel rectangle detection. In line extraction part, a new technique called "Crossing Dilation" is proposed to improve the Progressive Probabilistic Hough Transform performance, to extract long line segments that are not perfectly straight in large sparse binary image. The rectangle is detected using the proposed rectangle fitting method. Experiments on a variety of real cases of large panels are executed, some of which are badly disturbed by the fake edges caused by the dust in the background region. The results show great real-time performance on all the cases, and the proposed method is robust to the disturbances and different panel sizes.
This paper examines robust partially mode delay dependent H ∞ output feedback controller design for discrete-time systems with random communication delays. A finite state Markov chain with partially known transition ...
详细信息
Finding the closest points in the two normal convex hulls (NCH) gives an intuitive geometry explanation to support vector machine (SVM) in hard-margin case. However, the reduced convex hull (RCH) introduced in soft-ma...
详细信息
Finding the closest points in the two normal convex hulls (NCH) gives an intuitive geometry explanation to support vector machine (SVM) in hard-margin case. However, the reduced convex hull (RCH) introduced in soft-margin case is difficult to understand and explain because it cuts off some boundary points of NCH, which probably become support vectors. This paper proposes a soft-margin SVM model still based on NCH, named as L2-SVC-NCH. L2-SVC-NCH can be interpreted as finding two closest points in the two NCHs on a feature space and the punishment parameter C for empirical risk can be viewed as a trade off between the ordinary kernel function and Kronecker delta kernel function, which makes that L2-SVC-NCH is more perfect and easier to understand and explain than the SVM model based on RCH. In addition, L2-SVC-NCH obtains many additional advantages, such as the good feasible region, the strictly convex objective function and many optional geometric algorithms. The relationships between L2-SVC-NCH and some commonly used SVM models are also illustrated. The comparative experiments on five benchmark datasets show that L2-SVC-NCH is effective and competitive.
Background reconstruction plays an important role in many applications like video surveillance, motion analysis. Traditional Adaptive Gaussian Mixture Model will lose target when deal with arbitrary-long stationary ob...
详细信息
Background reconstruction plays an important role in many applications like video surveillance, motion analysis. Traditional Adaptive Gaussian Mixture Model will lose target when deal with arbitrary-long stationary object. In this paper, a novel method for detecting this kind of object is proposed to improve the performance of Adaptive Gaussian Mixture Model. Parameter restoration is designed to deal with arbitrary-long stationary target and solve the short-comings of the latest algorithm. The parameters among the K distribution of each pixel covered by the object will be restored when the object stayed for over threshold frames. Then the target will not be updated as a part of background model. Experimental results show that the proposed algorithm proves to be a more robust method by detecting the stationary target in an arbitrary-long time.
暂无评论