In order to better predict the education cost and improve the utilization rate of university funds, an education cost prediction model based on K-means clustering improved PSO-BP neural network was proposed. Firstly, ...
详细信息
ISBN:
(纸本)9781665454889
In order to better predict the education cost and improve the utilization rate of university funds, an education cost prediction model based on K-means clustering improved PSO-BP neural network was proposed. Firstly, aiming at the problem that BP neural network is prone to trap into local optimization and its convergence speed is slow, PSO was proposed to optimize the parameters of BP. Considering the limitations of PSO, the K-means clustering algorithm was added to cluster particleswarms, thereby enriching the diversity of the population. Finally, the improved particle swarm optimization algorithm was used for the education cost prediction of BP neural network. The results showed that the optimization performance of the improved particle swarm optimization algorithm is more stable and the convergence speed is faster. At the same time, in terms of prediction accuracy, the education cost prediction accuracy of the improved PSO-BP is higher.
The performance of piezoelectric ultrasonic transducer (PUT) is obviously affected by its matching layer. An intelligent optimization method for matching layer of PUT is proposed to fabricate PUT with high performance...
详细信息
The performance of piezoelectric ultrasonic transducer (PUT) is obviously affected by its matching layer. An intelligent optimization method for matching layer of PUT is proposed to fabricate PUT with high performance. The original data, including matching layer and performance parameters of PUT, is obtained by PiezoCAD software. Then the neural network models are established to mathematically characterize the effect of matching layer on the performance of PUT. The optimization criteria for PUT is constructed according to its performance requirements, such as center frequency (CF), bandwidth (BW) and pulse width (PW). In order to decrease the material cost, the thicknesses of matching layers are also considered in the optimization criteria. The modified particle swarm optimization algorithm is used to optimize the thicknesses of matching layers (Ag-epoxy and Parylene C). According to the designed performance, the optimized thicknesses of Ag-epoxy and Parylene C are about 54 mu m and 21 mu m, respectively. Based on the optimized thicknesses of matching layers, the simulated and experimental performances of PUT well agree with the designed ones. In addition, compared to the traditional quarter-wavelength theory, the proposed method can effectively decrease the materials cost. The -6dB BW and PW of PUT fabricated according to the optimized parameters are about 68.5% and 0.123 mu s, which are better than that fabricated according to the quarter-wavelength theory (about 50% and 0.147 mu s).
Considering the shortage of per capita water resources in China, the paper established a fractional order accumulated grey prediction model (FGM(1,1)) to predict per capita water consumption of 31 regions (provinces, ...
详细信息
Considering the shortage of per capita water resources in China, the paper established a fractional order accumulated grey prediction model (FGM(1,1)) to predict per capita water consumption of 31 regions (provinces, municipalities, and autonomous regions) in China from 2019 to 2024. The results show that per capita water consumption varies greatly across the different regions. Among them, per capita water consumption of nine regions (i.e., Beijing, Tianjin, Inner Mongolia, Jiangsu, Henan, Hubei, Guizhou, Yunnan, and Shaanxi) shows an increasing trend, whereas per capita water consumption in other 22 regions shows a downward trend. The predictive results can provide a basis for water resource management in China.
In this paper, a new adjacent non-homogeneous grey model was proposed to predict renewable energy consumption in Europe. Based on the principle of adjacent accumulation, the proposed model emphasizes the weight relati...
详细信息
In this paper, a new adjacent non-homogeneous grey model was proposed to predict renewable energy consumption in Europe. Based on the principle of adjacent accumulation, the proposed model emphasizes the weight relationship between the latest value and the historical data. Meanwhile, the heuristic algorithm is used to optimize the parameters of the proposed model. The model's performance was tested by the examples from four central European countries, including Austria, the Czech Republic, Hungary and Poland. In addition, the consumption of renewable energy in Europe and the world was predicted respectively, which showed that Europe's proportion of the world's renewable energy consumption will gradually decline. (C) 2020 Elsevier Inc. All rights reserved.
Rhizoma Coptidis is a Chinese herbal medicine with antibacterial and anti-inflammatory properties. It has extensive applications in modern medicine. The content of berberine in Rhizoma Coptidis directly determines its...
详细信息
Rhizoma Coptidis is a Chinese herbal medicine with antibacterial and anti-inflammatory properties. It has extensive applications in modern medicine. The content of berberine in Rhizoma Coptidis directly determines its quality. Fourier transforms near-infrared (FT-NIR) spectroscopy is a commonly used non-destructive method for rapidly detecting berberine content. In contrast to single-supervised learning algorithms in machine learning, ensemble learning combines individual learning algorithms to create a stable and better-performing strong-supervised model. This study collected spectral data of Rhizoma Coptidis using FT-NIR spectroscopy technology and established a chemometric model using a stacking ensemble approach with multiple models. Partial Least Squares (PLS), Adaptive Boosting (AdaBoost), Gradient boosting decision trees (GBDT), random forest (RF), and extreme gradient boosting (XGBoost) regression models were chosen as alternative base models, different Stacking models were established by random combinations. To fully leverage the strengths of each model and enhance predictive capability, an adaptive inertia weight particle swarm optimization algorithm (AWPSO) was used to search for the optimal parameters. The correlation coefficient of the test (RT) and the root mean square error of the test (RMSET) systematically evaluated the model performance. Finally, AWPSO-RF, AWPSOXGBoost, and AWPSO-AdaBoost were selected as the base models. The RMSET and RT for RF, XGBoost, and AdaBoost were 0.226, 0.250, 0.195, and 0.871, 0.830, 0.927. After optimizing with the AWPSO algorithm, the RMSET and RT for AWPSO-RF, AWPSO-XGBoost, and AWPSO-AdaBoost were 0.226, 0.245, 0.194, and 0.871, 0.843, 0.922, respectively. The RMSET and RT values for the stacking ensemble were 0.174 and 0.932. The prediction accuracy and generalization ability of multi -model fusion stacking ensemble learning are superior to the single -model regression methods. Therefore, the stacking ensemble learn
Data-driven soft sensor technology is an effective way to realize the online prediction of those important yet difficultto-measure quality and component variables in chemical *** the characteristics of strong nonlinea...
详细信息
ISBN:
(数字)9789887581536
ISBN:
(纸本)9781665482561
Data-driven soft sensor technology is an effective way to realize the online prediction of those important yet difficultto-measure quality and component variables in chemical *** the characteristics of strong nonlinearity,timevariability and multi-working conditions,the soft sensor models based on a single network are difficult to guarantee global-scale prediction accuracy,while the traditional integration methods usually lead to a decrease in prediction speed and an increase in storage *** cope with these issues,this paper proposes a novel ensemble soft sensor method using genetic algorithm to selectively integrate an extreme learning machine optimized by particleswarmalgorithm,named GASEN-PSOELM in ***,the sampling data are used to generate the training samples of the sub-models by Bootstrap ***,the candidate sub-models are established via the extreme learning machine optimized by particleswarmalgorithm,and then selected by using genetic ***,the prediction outputs are obtained by simple averaging *** validity and effectiveness of the proposed method is verified by the chemical equipment and chemical process,*** experimental results show that the proposed GASEN-PSOELM method can not only improve the prediction speed,but also ensure the good prediction accuracy and generalization performance.
The moisture content identification of a water drive oilfield is related to how to formulate and adjust the development plan of the oil field. However, traditional prediction methods for the water cut usually have pro...
详细信息
ISBN:
(纸本)9789811911668;9789811911651
The moisture content identification of a water drive oilfield is related to how to formulate and adjust the development plan of the oil field. However, traditional prediction methods for the water cut usually have problems such as slow recognition speed and limited by the specific conditions of the oil well. Therefore, in order to avoid the influence of the above problems, a machine algorithm model of water cut based on hybrid optimization is proposed, which is based on the support vector regression (SVR) model. First, the data is constructed by time sliding window;secondly, on the basis of the fundamental SVR model, this paper combines the particleswarmoptimization (PSO) and the Artificial Fish swarmalgorithm (AFSA) to optimize the hyperparameters of the SVR prediction model to achieve better experimental results;finally, if the hybrid model proposed in this article has some good experimental results, then it can be applied to the actual water cut prediction of the oilfield. After comparing four different models, the prediction model based on the hybrid optimizationalgorithm proposed in this paper has some good experimental results. The prediction curve and the real curve have the same trend as a whole, and the subtle errors are also the smallest. Thus, it performs better than the SVR prediction model optimized by the differential evolution algorithm, the SVR prediction model optimized by the genetic algorithm, and the SVR prediction model optimized by the PSO.
The active tuned mass damper (ATMD) is a reliable energy-dissipating device to effectively protect structures from serious damages due to earthquake excitations. This study proposes the optimal design of sliding secto...
详细信息
The active tuned mass damper (ATMD) is a reliable energy-dissipating device to effectively protect structures from serious damages due to earthquake excitations. This study proposes the optimal design of sliding sector control (SSC) for the seismic protection of an 11-story shear building structure equipped with ATMD. First, the SSC controller is optimally designed for the seismic control of the structure subjected to an artificial earthquake. Then, the effectiveness of the optimized SSC (OSSC) is assessed in reducing the seismic responses of the structure subjected to four near- and far-fault earthquake excitations. The efficient performance of the OSSC technique is also validated and compared with that of a number of the control techniques such as linear quadratic regulator (LQR), fuzzy logic control (FLC), proportional-integral-derivative (PID), and optimal sliding mode control (OSMC). Comparative results demonstrate the efficiency and robustness of the proposed OSSC in comparison with those of the other controllers.
Endohedrally doped clusters have received much attention because they can act as superatoms and have great potential as the building blocks for cluster-assembled materials. We have carried out a comprehensive study ba...
详细信息
Endohedrally doped clusters have received much attention because they can act as superatoms and have great potential as the building blocks for cluster-assembled materials. We have carried out a comprehensive study based on the combination of the particleswarming optimizationalgorithm and the relativistic density functional theory, to obtain geometric structures and stabilities of lead (Pb) clusters doped with one uranium (U) atom. A gradual evolution pattern was observed with the increasing number of Pb atoms, from exohedrally doped structures to quasi-endohedral structures, and finally to endohedrally doped structures. At least 12 and 11 Pb atoms were necessary to encapsulate the U atom completely in the calculation without and with the spin-orbit coupling (SOC) effect respectively. The UPb16 cluster was represented as a highly stable endohedral cage structure with the U atom around its center. In addition, the crucial role of the SOC effect was emphasized. We hope that our findings will provide a fundamental understanding of actinide-doped lead clusters, which may be quite different from their light element analogues.
This paper proposes a personalized learning resource recommendation method based on dynamic collaborative filtering algorithm. Pearson correlation coefficient is used to calculate the data similarity between learning ...
详细信息
This paper proposes a personalized learning resource recommendation method based on dynamic collaborative filtering algorithm. Pearson correlation coefficient is used to calculate the data similarity between learning users or project resources in the network, and the unscored value is obtained. In order to solve the problems of sparse data and poor scalability in collaborative filtering algorithm, dynamic k-nearest-neighbor and Slope One algorithm are used to optimize it, and the sparsity of learning resource data in the network is analyzed according to the result of neighbor selection. The bidirectional self-equalization of stage evolution is used to improve the personalized recommendation of resource push, and the fuzzy adaptive binary particle swarm optimization algorithm based on the evolution state judgment is used to solve the problem of the optimal sequence recommendation, so as to realize the personalized learning resource recommendation. The experimental results show that the proposed method has higher matching degree and faster recommendation speed.
暂无评论