Genetic algorithm (GA) is an effective method for path planning problems. As a powerful variant of GA, island genetic algorithm (IGA) has considerable improvement in performance. In this paper, a new island model of G...
Genetic algorithm (GA) is an effective method for path planning problems. As a powerful variant of GA, island genetic algorithm (IGA) has considerable improvement in performance. In this paper, a new island model of GA is proposed to avoid the premature phenomenon and achieve better efficiency. First, a new method of creating subpopulations is presented based on K-means to expand the searching area for the optima. Meanwhile, recombining subpopulations is proposed as a new strategy to improve the diversity of populations and save computational time. Moreover, a method is designed based on Monte Carlo sampling to handle the uncertainty of maps. Comparative experiments are presented to verify the efficiency of the proposed algorithm. Then, a proper number of samples is found by simulation to balance the accuracy and the time cost of Monte Carlo sampling.
After a large-scale grid failure, it is important for dispatchers to know the faulty components and causes and to develop the correct scheduling strategy. In this paper, a new fault diagnosis scheme is presented based...
After a large-scale grid failure, it is important for dispatchers to know the faulty components and causes and to develop the correct scheduling strategy. In this paper, a new fault diagnosis scheme is presented based on widely configured fault recorders. The proceeded algorithm using the wide-area recorder data to locate the fault components such as transmission lines, busbars and transformers. Different indicators are given for different components on the grid. A variety of criterion indexes are deduced for the transmission line using the distributed parameter model. The credibility of the judgment is improved by using multiple criteria. A search scheme based on the initial adjacency matrix is used to locate fault component on the grid. The fault area is defined according to the uploaded recorder data. The fault components are searched in the wide defined fault area according to the proposed scheme. The element with the greatest probability of failure is found. The three faulty elements simulation models based on PSCAD/EMDTC are established and studied, what's more, the power system model is also established to test the algorithm. The simulation results verify the effectiveness of the proposed method.
The distributed generation which is integrated in the active distribution network changes power flow, bringing new challenges to the voltage control. When voltage limit violation happens, in order to make the voltage ...
The distributed generation which is integrated in the active distribution network changes power flow, bringing new challenges to the voltage control. When voltage limit violation happens, in order to make the voltage return to normal range and improve the voltage quality, a novel voltage control strategy is proposed. Considering the voltage quality and node importance, the electrical closeness centrality and key node contribution degree are defined, and the key nodes are determined by the orders of the key node contribution degree. This paper uses the reactive power compensation devices which are installed at the key nodes coordinated with the reactive power output of the distributed generation to realize the voltage optimization control. The voltage optimization control model is established by taking the minimum power loss as an objective function. Using the particle swarm optimization algorithm solves the model. The simulation results of the improved IEEE-33 bus system verify the effectiveness of the proposed method.
Localization is of paramount importance for underwater wireless sensor networks (UWSNs). However, achieving accurate location is infeasible, especially in the highly dynamic underwater environment. The acoustic signal...
Localization is of paramount importance for underwater wireless sensor networks (UWSNs). However, achieving accurate location is infeasible, especially in the highly dynamic underwater environment. The acoustic signal may suffer hybrid loss, including path and absorption loss, which dramatically degrades the localization accuracy. Even though some localization methods have been proposed, the trade-off between accuracy and computational complexity cannot be well balanced. In this context, the paper proposes a computationally efficient method that investigates the localization problem in the alternating nonnegative constrained least squares (ANCLS) framework after linearization operation. The potential solutions are divided into two groups, wherein the optimal one is filtered under the constraint by exchanging the variables from one to another. A block principal pivoting-based localization (BPPL) method is then presented to estimate the target's location. Simulations reveal that the computational complexity and the localization accuracy of BPPL are competitive compared with the state-of-the-art methods in different scenarios.
Operating Statuses Identification (OSI) can help operators to find fault timely and minimize the loss. So it is of great significance for optimal operation of photovoltaic (PV) plants. The loss quantity of electricity...
详细信息
ISBN:
(纸本)9781510821415
Operating Statuses Identification (OSI) can help operators to find fault timely and minimize the loss. So it is of great significance for optimal operation of photovoltaic (PV) plants. The loss quantity of electricity (LQOE) is defined as that should have been generated but actually not, which is caused by inverter fault, PV modules fault, dust stratification or combinations of them. Firstly, mathematical models of PV cells are proposed to figure out the LQOE of each PV module. Secondly, after analysing the relation between LQOE and operating statuses, an index set of LQOE describing the distinction of different operating statuses are defined including four statistical feature parameters and a user-defined index. Thirdly, Support Vector Classification models for OSI (OSI-SVC) are built with input features extracted from the index set. Lastly, simulations are carried out to verify the effectiveness and evaluate the performance of the OSI-SVC models. The results indicated that the operating statuses can be effectively recognized by the proposed model.
Summer precipitation over the Tibetan Plateau (TP) sustains the "Water tower of Asia" and has notable influences on the climate and subsequently eco-environment of the Northern Hemisphere. Due to its high el...
详细信息
Although fantastic milestones of Gallium nitride (GaN)-based materials in optoelectronic devices had been reached, the focus on the optimization of their geometrical structure for gas-sensing applications was relative...
详细信息
In this paper, we address the distributed prescribed-time convex optimization (DPTCO) problem for a class of high-order nonlinear multi-agent systems (MASs) under undirected connected graphs. A cascade design framewor...
详细信息
The probability of instantaneous grounding fault during overhead line transmission is high, and it is necessary to configure an efficient restart plan to restore the system. The traditional reclosing strategy is easy ...
详细信息
Incremental few-shot semantic segmentation (IFSS) aims to incrementally expand a semantic segmentation model’s ability to identify new classes based on few samples. However, it grapples with the dual challenges of ca...
详细信息
ISBN:
(数字)9798350330991
ISBN:
(纸本)9798350331004
Incremental few-shot semantic segmentation (IFSS) aims to incrementally expand a semantic segmentation model’s ability to identify new classes based on few samples. However, it grapples with the dual challenges of catastrophic forgetting (due to feature drift in old classes) and overfitting (triggered by inadequate samples in new classes). To address these issues, a novel approach is proposed to integrate pixel-wise and region-wise contrastive learning, complemented by an optimized example and anchor sampling strategy. The proposed method incorporates a region memory and pixel memory designed to explore the high-dimensional embedding space more effectively. The memory, retaining the feature embeddings of known classes, facilitates the calibration and alignment of seen class features during the learning process of new classes. To further mitigate overfitting, the proposed approach implements an optimized example and anchor sampling strategy. Extensive experiments show the competitive performance of the proposed method. The source code of this work can be found in https://***.
暂无评论