Electric power marketization is the product of the continuous development of the current electric power system. Day-ahead electricity price plays an important role in power market, so the prediction of day-ahead elect...
详细信息
ISBN:
(纸本)9781665479141
Electric power marketization is the product of the continuous development of the current electric power system. Day-ahead electricity price plays an important role in power market, so the prediction of day-ahead electricity price becomes very important. In this paper, the load rate of the system is introduced, and the machine learning algorithm is adopted. In addition, the improvement and expansion of the machine learning algorithm are realized by combining the tuning advantage of the bionic optimization algorithm, and then the accurate prediction of day-ahead electricity price is realized. The results show that this strategy algorithm can greatly improve the accuracy of day-ahead electricity price forecasting.
As a new branch of natural computing, membrane computing has received increasing attention. The hierarchical and parallel structure of P system provides benefits for the resolving of optimization problems. In this pap...
详细信息
ISBN:
(纸本)9783642342882
As a new branch of natural computing, membrane computing has received increasing attention. The hierarchical and parallel structure of P system provides benefits for the resolving of optimization problems. In this paper, we combined membrane computing and evolutionary algorithms, and proposed an optimization algorithm to resolve the multi-variable optimization problems with constraints. The two standard testing functions were adopted to evaluate the proposed optimization algorithm. The results of the experiments showed the effectiveness of the proposed method.
This paper presents an algorithm for the optimization of magnetic components in power electronics which includes a comprehensive analysis of the power losses and the temperature rise using dimensionless numbers for na...
详细信息
ISBN:
(纸本)9789075815412
This paper presents an algorithm for the optimization of magnetic components in power electronics which includes a comprehensive analysis of the power losses and the temperature rise using dimensionless numbers for natural and forced convection. The results show that this approach is highly efficient and comparatively accurate.
In the wireless ubiquitous environment, this paper analyses and models the dynamic volume of access to VOD business. This paper also proposes a wireless business optimization algorithm that enables the CDN edge server...
详细信息
ISBN:
(纸本)9781728121840
In the wireless ubiquitous environment, this paper analyses and models the dynamic volume of access to VOD business. This paper also proposes a wireless business optimization algorithm that enables the CDN edge server to help the core network to divert pressure of user's access best. Firstly, a probability model of the daily volume of access is defined, based on video's score and video's online time. Further, a function relationship between the parameter of the model and the joint conditions above is established to accurately reflect the average value of the daily volume of access. On this basis, the predicted value of the daily volume of access to any video under a certain guarantee probability is obtained. Then, an optimization model for distribution and download of videos is built for the edge server of CDN with limited storage space. A distribution and download scheme of videos based on 0-1 knapsack algorithm is proposed. The simulation results verify the correctness of the model of volume of access and the effectiveness of the optimization algorithm.
At present, GIM models face problems in the field of 3D modeling, including the complex hierarchical architecture of the model, the incomplete model classification information, many imaging topology errors of the mode...
详细信息
ISBN:
(纸本)9798350349047;9798350349030
At present, GIM models face problems in the field of 3D modeling, including the complex hierarchical architecture of the model, the incomplete model classification information, many imaging topology errors of the model, and the non-standardized phenomenon of attribute storage. These problems hinder the effective application and widespread dissemination of GIM models. This makes the fusion of GIM model and geographic information data has not yet reached the ideal state, which makes its digital application scope relatively limited. To solve the above problems, the 3D modeling and optimization method of power grid information model are studied deeply. Therefore, a data storage method based on B-tree structure is proposed, aiming to realize the unified storage of GIM data and the restoration of hierarchy structure. This not only standardizes the storage mode of attribute data, restores the complex hierarchical architecture of the model, but also facilitates the data exchange between different systems, and further promotes the application and development of GIM model. Moreover, because the pole-tower 3D model provided by GIM has some point accuracy problems in the original measurement, this study proposed an optimization algorithm based on adjacent point merging. It improves the accuracy and efficiency of the model data processing, and significantly enhances the visual effect of the model and provides a more solid and comprehensive technical support for the research and application of related fields.
In the past few years, deep learning has been used widely in bioinformatics area to solve common problems such as protein sequence prediction, phylogenic inferences, multiple sequence alignment and many more. It has b...
详细信息
ISBN:
(纸本)9781467377911
In the past few years, deep learning has been used widely in bioinformatics area to solve common problems such as protein sequence prediction, phylogenic inferences, multiple sequence alignment and many more. It has been in the spotlight as a powerful approach which makes significant advances in taking care of the issues that haunt artificial intelligence community for many years. However, several weaknesses such as trap at local minima, lower performance and high computational time still occur in deep learning. Therefore, global optimization technique such as differential search algorithm can be used to assist deep learning method in order to get best finding result and data. This review will cover fundamental of deep learning and their involvement in bioinformatics field as well as implementation of differential search algorithm and their involvement in bioinformatics field.
A new structure of hydraulic hybrid vehicle (HHV) with hydraulic transformer (HT) was built and the working principle of the new hydraulic hybrid vehicle was described. According to the operating characteristics of HT...
详细信息
ISBN:
(纸本)9783037850398
A new structure of hydraulic hybrid vehicle (HHV) with hydraulic transformer (HT) was built and the working principle of the new hydraulic hybrid vehicle was described. According to the operating characteristics of HT and energy-saving optimization conditions of accumulator used the HHV;Energy-saving optimization control algorithm with various operation conditions in different working conditions of vehicle was established. Then, simulation analysis to control performance of energy-saving algorithm was carried out using PID, Fuzzy Logic Controller (FLC) and Fuzzy-PID control strategy. Results show that Fuzzy-PID controller has a small influence on the parameters of energy-saving optimization algorithm of the hydraulic hybrid vehicle and maximizing energy recovery can be achieved in different energy states by Fuzzy-PID controller.
For the irregular nesting problem widely existing in modern manufacturing industry, this paper makes a research on it and presents an optimization algorithm based on no-fit polygon（NFP） method and hybrid heuristic s...
详细信息
ISBN:
(纸本)9781538629185
For the irregular nesting problem widely existing in modern manufacturing industry, this paper makes a research on it and presents an optimization algorithm based on no-fit polygon（NFP） method and hybrid heuristic strategy to solve it. The proposed algorithm first uses the composition method of trace line segment to calculate no-fit polygons（NFPs） between every two pieces in piece set, and extracts the candidate placement points for a candidate piece to be placed by using generated NFPs in combination with internal no-fit polygon（INFP） between the piece and the material plate. Then, the algorithm designs and applies three hybrid heuristic strategies to evaluate all candidate pieces and choose the best piece to place next and determine the best placement point for the selected piece. Experimental test has been performed to verify feasibility and effectiveness of the proposed algorithm. The test results show that the algorithm can solve the irregular nesting problem effectively, and improve the utilization of material to a certain extent.
In recent years, with the increasing complexity of system structure and refinement of management, power system optimization computation is facing the double pressure of computational accuracy and computational time. M...
详细信息
ISBN:
(纸本)9798350390780;9798350379228
In recent years, with the increasing complexity of system structure and refinement of management, power system optimization computation is facing the double pressure of computational accuracy and computational time. Machine learning (M) techniques can effectively utilize vast historical data to provide new theoretical basis for optimizing stable and fast solutions. This paper is aimed to summarize the optimization algorithm frameworks, which is based on the interaction between M and optimization computing for general power system optimization problems. The framework empowered by M mainly includes end-to-end optimization, ML-enhanced optimization, and joint-driven optimization. By utilizing the physical characteristics of power system, these frameworks not only provide a more efficient and reliable calculation scheme for the optimal operation of power system, but also provide innovative solutions for other complex optimization problems.
The past decades have seen an extensive investigation of evolutionary algorithms. The recombination operators of most evolutionary algorithms are either single-point crossover or multi-point crossover for solving 0-1 ...
详细信息
ISBN:
(纸本)9781728160924
The past decades have seen an extensive investigation of evolutionary algorithms. The recombination operators of most evolutionary algorithms are either single-point crossover or multi-point crossover for solving 0-1 combinatorial optimization problems. There is a little studies on excavating the variable relationship to improve the efficiency of the recombination operator. Hence, we propose an optimization algorithm based on excavating variable relationship for solving 0-1 combinational problems. The aim of the recombination operator is fully utilizing the inherent information from every slice component of all individuals. We compared the proposed algorithm with the classic evolutional algorithm with single-point crossover operation on several 0-1 knapsack problems, which is a classical combinatorial optimization problem. The simulation results show the convergence efficiency of the proposed algorithm.
暂无评论