鲸鱼优化支持向量机的短期风电功率预测

为提高风电预测的精度,提出一种鲸鱼优化支持向量机SVM(support vector machine)的组合预测模型。该模型针对风电序列的非平稳波动特性,首先应用集合经验模态分解技术EEMD(ensemble empirical mode de?composition)将原始风电序列分解为一系列不同特征尺度的子序列;并引入鲸鱼优化算法WOA(whales optimiza?tion algorithm)解决SVM中学习参数选择难的问题,进而对各子序列建立WOA_SVM预测模型;最后,叠加各子序列的预测值以得到最终预测值。仿真表明,所提EEMD_WOA_SVM模型具有较高的风电预测精度,显著优于其他基本模型。     

园区多能互补供能系统设计方法研究

为提高园区多能互补工程的能源综合利用效率及整体效益，对供能系统设计中的负荷统计、逐时负荷匹配、供能系统、设备选型、容量配置及运行策略等问题进行了分析和总结，提出园区多能互补供能系统设计的基本流程与方法，可为园区多能互补供能系统的设计提供参考。     

An improved iterative stochastic multi-objective acceptability analysis method for robust alternative selection in new product development

Alternative selection in new product development (NPD) is a multi-criteria decision-making (MCDM) problem. It usually starts with incomplete, imprecise or even partially missing information. Currently, most existing methods in dealing with this problem cannot work well if required information is incomplete or missing. It is acknowledged that stochastic multi-objective acceptability analysis (SMAA) can be applied to address MCDM problem with incomplete preference information and uncertain criteria measurements. In SMAA, alternatives are evaluated based on SMAA measurements (acceptability index, central weight vector and confidence factor). The discriminability of SMAA for the optimum alternative heavily depends on differences of SMAA measurements among different alternatives. Usually, a large number of alternatives and high level of uncertainty are involved in alternative selection in NPD. In this situation, the differences among SMAA measurements are not obvious, and therefore SMAA cannot deal with such problem very well. To this end, this paper proposes an improved SMAA method called Iterative-SMAA (I-SMAA) for alternative selection in NPD. In the I-SMAA, an iterative multi-step decision-making process is suggested to improve differences of SMAA measurements among different alternatives, and thus assist decision makers (DMs) to positively discern from the most preferred alternative. To enhance the decision-making efficiency, sensitive criteria are acquired in each iteration by ranking sensitivity analysis. DMs are guided to provide partial preference information and give more accurate criteria measurements for sensitive criteria rather than all criteria. Eventually, to verify the proposed method, a numerical example of the existing literature is solved with the method, and the results are compared. And then, a practical example of a preparation equipment for coal samples is further employed to verify the practicability of the proposed I-SMAA.     

5 000 t公务船调距桨桨叶数选型研究

针对国内外调距桨桨叶数选型依靠经验为主，对不同桨叶数对推进器整体性能的影响始终缺乏定量分析的现状，以某5 000 t公务船为目标船型，通过调距桨分别配置不同的桨叶数，分析不同桨叶数对目标船型综合性能的影响。总结了公务船调距桨桨叶数选型的方法，归纳了不同叶数调距桨的优劣，可为后续新造船桨叶数的选取提供参考。     

Hybrid self-inertia weight adaptive particle swarm optimisation with local search using C4.5 decision tree classifier for feature selection problems

ABSTRACT

Feature selection is an important task to improve the classifier’s accuracy and to decrease the problem size. A number of methodologies have been presented for feature selection problems using metaheuristic algorithms. In this paper, an improved self-adaptive inertia weight particle swarm optimisation with local search and combined with C4.5 classifiers for feature selection algorithm is proposed. In this proposed algorithm, the gradient base local search with its capacity of helping to explore the feature space and an improved self-adaptive inertia weight particle swarm optimisation with its ability to converge a best global solution in the search space. Experimental results have verified that the SIW-APSO-LS performed well compared with other state of art feature selection techniques on a suit of 16 standard data sets.     

Applying Feature-Weighted Gradient Decent K-Nearest Neighbor to Select Promising Projects for Scientific Funding

Due to its outstanding ability in processing large quantity and high-dimensional data, machine learning models have been used in many cases, such as pattern recognition, classification, spam filtering, data mining and forecasting. As an outstanding machine learning algorithm, K-Nearest Neighbor (KNN) has been widely used in different situations, yet in selecting qualified applicants for winning a funding is almost new. The major problem lies in how to accurately determine the importance of attributes. In this paper, we propose a Feature-weighted Gradient Decent K-Nearest Neighbor (FGDKNN) method to classify funding applicants in to two types: approved ones or not approved ones. The FGDKNN is based on a gradient decent learning algorithm to update weight. It updatesthe weight of labels by minimizing error ratio iteratively, so that the importance of attributes can be described better. We investigate the performance of FGDKNN with Beijing Innofund. The results show that FGDKNN performs about 23%, 20%, 18%, 15% better than KNN, SVM, DT and ANN, respectively. Moreover, the FGDKNN has fast convergence time under different training scales, and has good performance under different settings.     

MOOC Learner’s Final Grade Prediction Based on an Improved Random Forests Method

Massive Open Online Course (MOOC) has become a popular way of online learning used across the world by millions of people. Meanwhile, a vast amount of information has been collected from the MOOC learners and institutions. Based on the educational data, a lot of researches have been investigated for the prediction of the MOOC learner’s final grade. However, there are still two problems in this research field. The first problem is how to select the most proper features to improve the prediction accuracy, and the second problem is how to use or modify the data mining algorithms for a better analysis of the MOOC data. In order to solve these two problems, an improved random forests method is proposed in this paper. First, a hybrid indicator is defined to measure the importance of the features, and a rule is further established for the feature selection; then, a Clustering-Synthetic Minority Over-sampling Technique (SMOTE) is embedded into the traditional random forests algorithm to solve the class imbalance problem. In experiment part, we verify the performance of the proposed method by using the Canvas Network Person-Course (CNPC) dataset. Furthermore, four well-known prediction methods have been applied for comparison, where the superiority of our method has been proved.     

Geospatial analysis for utilizing the marginal land in regional biofuel industry: A case study in Guangdong Province,China

Due to the intensive and exhaustive land use in China, the so-called marginal land is expected to play a major role in supporting the biofuel industry of the country. We developed a regional-level framework of using geospatial information technologies to achieve an optimal utilization of the marginal land for biofuel production. The framework includes identifying marginal lands, evaluating optimality of the land for growing certain bioenergy crops, estimating local potential feedstock production, and finally selecting optimal sites for biofuel factories. We present a case study of farming Jatropha (Jatropha curcas L.) and Cassava (Manihot esculenta Crantz) in Guangdong, China. The marginal land was identified from satellite imageries at a 30-m resolution. The optimality for growing the two species was evaluated using the Ecological Niche Models (ENMs), which incorporates local temperature, precipitation, soil, and terrain. The optimality value was then converted into potential feedstock production using a conversion model. The site selection for the factories incorporated the local potential feedstock production, the transportation cost measured by road distance, and the farming cost related to the land patch geometry. Each candidate site received an overall optimality score derived based on those factors. We identified five sites that have high scores and also minimal or none spatial overlaps of their supporting areas. Three of them (Zhanjiang, Yunfu, and Jieyang) are for Cassava, located on in southern Guangdong. Two (Qingyuan and Meizhou) are for Jatropha in northern Guangdong.     

Dynamic facility layout problem in footwear industry

Dynamic facility layout problem (DFLP) deals with the arrangement of machines in a site as to minimize the sum of materials handling and re-layout costs by considering multi periods. The DFLP studies in the literature provide several different algorithms and utilize the well known test problems to assess their performance. However, real life applications are overlooked. The industries such as footwear and clothing are prone to seasonal demand changes. Therefore, time horizons and layout/re-layout of the machines within the facility should be studied carefully. This study considers a footwear facility and several scenarios are generated by using the real life data. A clonal selection based algorithm is proposed to solve the real life DFLP. The performance of the algorithm, further the effect of time periods on solution quality and applicability of the results are tested and promising results are obtained.