Particle Swarm Optimization Support Vector Regression Research Articles

Prediction of Deformation in Expansive Soil Landslides Utilizing AMPSO-SVR

A non-periodic “step-like” variation in displacement is exhibited owing to the repeated instability of expansive soil landslides. The dynamic prediction of deformation for expansive soil landslides has become a challenge in actual engineering for disaster prevention and mitigation. Therefore, a support vector regression prediction (AMPSO-SVR) model based on adaptive mutation particle swarm optimization is proposed, which is suitable for small samples of data. The shallow displacement is decomposed into a trend component and fluctuating component by complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), and the trend displacement is predicted by cubic polynomial fitting. In this paper, the multiple disaster-inducing factors of expansive landslides and the time hysteresis effect between displacement and its influencing factors are fully considered, and the crucial influencing factors which eliminate the time lag effect and state factors are input into the model to predict the fluctuation displacement. Monitoring data in the Ningming area of China are employed for the model validation. The predicted results are compared with those of the traditional model. The model performance is evaluated through indicators such as the goodness of fit R2 and root mean square error RMSE. The results show that the prediction RMSE of the new model for three monitoring stations can reach 2.6 mm, 6.6 mm, and 2.5 mm, respectively. Compared with the common Grid search support vector regression (GS-SVR), the Particle Swarm Optimization Support Vector Regression (PSO-SVR) and Back Propagation Neural Network (BPNN) models have average improvements of 58.4%, 38.1%, and 25.2% respectively. The goodness of fit R2 is superior to 0.99 in the new method. The proposed model can effectively be deployed for the displacement prediction of non-periodic stepped expansive soil landslides driven by multiple influencing factors, providing a reference idea for the deformation prediction of expansive soil landslides.

Utilization of hyperspectral imaging for the analysis of aroma components of Soy Sauce-Aroma Type Baijiu

This study sought to develop a hyperspectral imaging (HSI)- and machine learning (mL)-based method to quantitatively analyze the composition of alcohols and esters of Soy Sauce-Aroma Type Baijiu (SSAB). First, preprocessed the spectral data in the 900–1700 nm NIR range using various algorithms. Then, particle swarm optimization-support vector regression (PSO-SVR) and random forest (RF) models were established based on the full wavelength and the feature wavelength spectral data. In the preprocessing method, the external parameter orthogonalization (EPO) method demonstrated superior modeling performance in most cases. Spearman correlation analysis revealed a significant positive correlation between the concentrations of ethanol and alcohols in the SSAB samples, with an r-value of 0.93. Ethyl acetate and ethyl lactate also showed a significant positive correlation with esters, with r-values of 0.87 and 0.88, respectively. Combining the relevant compounds with feature wavelengths as input yields a model with high accuracy, with an Rp2 of 0.997 and RMSEP of 0.118 mg/L for alcohols and an Rp2 of 0.996 and RMSEP of 0.017 mg/L for esters. The results indicate that HSI can achieve non-destructive and accurate detection of alcohols and esters in the aroma of SSAB, providing a new method for aroma analysis in Baijiu.

A framework for estimating the matric suction in unsaturated soils using multiple artificial intelligence techniques

AbstractImplementation of the state‐of‐the‐art understanding of the mechanics of unsaturated soils into geotechnical engineering practice is partly limited due to the lack of quick, reliable, and economical techniques for matric suction measurement. Matric suction is one of the key stress state variables that significantly influences the hydro‐mechanical behavior of unsaturated soils. In this paper, to address this objective, two artificial intelligence (AI) models were developed for estimating matric suction in unsaturated soils based on the particle swarm optimization support vector regression (PSO‐SVR) and multivariate adaptive regression spline (MARS) algorithms. The results suggest that both these models can reasonably estimate matric suction. Compared to the MARS model, the PSO‐SVR model can achieve higher accuracy. Nonetheless, the MARS model facilitates the sensitivity analysis and the selection of essential inputs. A novel integrated framework is proposed and validated, leveraging the strengths, and alleviating the limitations of the PSO‐SVR and MARS algorithms for reliable and rapid estimation of matric suction in the range of 0–1500 kPa for low plastic soils (0 < Ip ≤ 7). Six inputs are required to use this model successfully; some can be measured using conventional laboratory tests, and others can be calculated from mass‐volume relationships.

Prediction of flow stress in Mg-3Dy alloy based on constitutive equation and PSO-SVR model

This study conducted hot compression experiments on as-cast Mg-3Dy alloy under deformation parameters of 380 °C–470 °C and 0.001–1 s−1. The microstructure of the alloy was observed using EBSD, and the flow stress of the Mg-3Dy alloy was predicted using the Arrhenius model and the particle swarm optimization-support vector regression (PSO-SVR) model. The organizational analysis results showed that the main recrystallization mechanism in the alloy is the discontinuous dynamic recrystallization (DDRX) mechanism. The generation of twins in the alloy was mostly the result of local stress action. The optimal processing window for this alloy was determined to be 380 °C–470 °C and 0.001–0.01 s−1 through the thermal processing map. The prediction accuracies of the Arrhenius model and PSO-SVR model were evaluated using the correlation coefficient R2 and mean squared error MSE. The results showed that the PSO-SVR model significantly outperforms the Arrhenius model in prediction accuracy, with R2 value of 0.99982 and MSE of 0.074.

An interval-valued carbon price forecasting method based on web search data and social media sentiment.

Accurate carbon price prediction is a crucial task for the carbon trading market. Previous studies have ignored the impact of online data and are limited to point predictions, which brings challenges to the accurate forecasting of carbon prices. To address those issues, this paper proposes an interval-valued carbon price forecasting method based on web search data and social media sentiment. First, we collect web search data and social media sentiment to improve prediction performance by synthesizing multiple types of data information. Second, we employ principal component analysis (PCA) to preprocess high-dimensional web search data, and utilize BosonNLP for quantifying social media information, thereby enhancing the predictability of the dataset. Subsequently, a variational mode decomposition (VMD) is applied to the carbon price and online data, followed by utilizing particle swarm optimization support vector regression (PSO-SVR) to predict each sub-modes and summing them up to obtain the ultimate forecasting outcome. Finally, using carbon prices in Guangdong and Hubei provinces as case studies, the experimental results demonstrate that web search data and social media sentiment significantly enhance the predictive accuracy of interval-valued carbon prices. Furthermore, the proposed VMD-PSO-SVR outperforms other comparative models in the accuracy and reliability of interval-valued forecasting.

Prediction of chloride ion concentration distribution in basalt-polypropylene fiber reinforced concrete based on optimized machine learning algorithm

The distribution characteristics of chloride ion concentration in basalt-polypropylene fiber reinforced concrete (BPFRC-CC) are important for assessing the durability of concrete structures under marine environment. This study collected the distribution data of BPFRC-CC in three different environments: tidal, submersible, and salt spray. Several optimized machine learning models were developed to predict the distribution of BPFRC-CC. The performance of these models was evaluated based on a comparison between actual and predicted values, residual distribution histograms, and performance indicators. The research results of this article showed that the particle swarm optimization support vector regression (PSO-SVR) model exhibits significant accuracy and generalization ability, with closer proximity to the actual values in both the training and testing sets. The residual distribution histogram of the PSO-SVR model approximated a normal distribution with a smaller mean and standard deviation. Evaluation indicators, including R2, MSE, RMSE, MAE, and MAPE, for the PSO-SVR model were 0.994, 6.92e-5, 0.0083, 0.0062, and 0.0453, respectively. The SHAP values revealed that depth and age were the most critical influencing factors. The decrease in age, water-cement ratio, and water-binder ratio, the addition of basalt fibers and polypropylene fibers, and the increase in depth reduced the BPFRC-CC. In addition, BPFRC-CC in tidal, salt spray, and submersible areas decreased in that order. Finally, a graphical user interface for BPFRC-CC was developed, providing researchers and production personnel with an intuitive, efficient, and visual production experience. This study has important implications for the optimization of BPFRC formulations and the prediction and assessment of the life of BPFRC structures.

Hyperspectral Detection of Moisture Content in Rice Straw Nutrient Bowl Trays Based on PSO-SVR

In the process of rice straw nutrient bowl tray drying, real-time detection of changes in moisture content to achieve automatic adjustment of drying factors is one of the important means to ensure its drying quality. At present, the main method for measuring the moisture content of rice straw nutrient bowl trays is the drying and weighing method. This method is not only time consuming, labor intensive, and complex to operate, but also has poor real-time performance, which cannot meet the demand for real-time detection of the moisture content in the production process of rice straw nutrient bowl trays. In this regard, a real-time moisture content detection method for rice straw nutrient bowl trays based on hyperspectral imaging technology was studied. In this study we took the rice straw nutrient bowl tray during the drying process as the research object, adopted a single factor experiment, took microwave power, hot air temperature, and hot air speed as the drying factors, and took the moisture content of the rice straw nutrient bowl tray as the drying index. The rice straw nutrient bowl tray was dried according to the designed drying conditions. When drying, we removed the rice straw nutrient bowl tray every 5 min for weighing and collected hyperspectral image data within the wavelength range of 400~1000 nm until its quality remained unchanged. A total of 204 samples were collected. Using the average spectrum of the region of interest as the sample for effective spectral information, spectral preprocessing was performed using multivariate scattering correction (MSC), standardization normal variables (SNV), and Savitzky–Golay convolution smoothing (SG) methods. Principal component analysis (PCA) and competitive adaptive reweighting (CARS) methods were adopted for the dimensionality reduction of the spectral data. Three prediction models of rice straw nutrient bowl tray moisture content, namely random forest regression (RF), particle swarm optimization support vector regression (PSO-SVR), and XGBoost model were constructed using the reduced dimension spectral data. Finally, the performance of the model was compared using the coefficient of determination (R2) and mean square error (RMSE) as evaluation indicators. The research results indicate that the PCA-PSO-SVR model established based on SG method preprocessing has the best predictive performance, with a training set decision coefficient R2C of 0.984, a training set mean square error RMSE-C of 2.775, a testing set decision coefficient R2P of 0.971, and a testing set mean square error RMSE-P of 3.448. The model therefore has a high accuracy. This study achieved rapid detection of water content in rice straw nutrition trays. This method provides a reliable theoretical basis and technical support for the rapid detection of rice straw nutrient bowl tray moisture content, and is of great significance for improving the quality of rice straw nutrient bowl trays; promoting the popularization and application of raising rice straw nutrient bowl tray seedlings and whole process mechanized planting technology system; improving soil structure; and protecting the ecological environment.

Confocal Raman microspectroscopy combined with spectral screening algorithms for quantitative analysis of starch in rice

Starch is one of the main components of rice. The feasibility of confocal Raman microspectroscopy (CRM) combined with spectral screening algorithms for quantitative detection of starch in rice was investigated in this paper. Firstly, the Raman spectra of the samples were processed by pre-processing and spectral screening algorithm of the data. Secondly, the quantitative starch detection models were respectively established by correction set by partial least squares regression, particle swarm optimization-support vector regression, gray wolf optimizer-support vector regression (GWO-SVR) as well as back propagation neural network. After that, the quantitative starch detection models were validated using the prediction set. Eventually, the related coefficient of the prediction set, root mean square error of the prediction set, and mean relative error of GWO-SVR starch quantification detection model combined with normalization and IPLS screening were respectively 0.8915, 1.03%, and 1.08%. The experimental results manifested that the swift and precise detection of starch content in rice by CRM was workable after proper pre-processing and spectral screening algorithm.

Prediction Method for Sugarcane Syrup Brix Based on Improved Support Vector Regression

The brix of syrup is an important parameter in sugar production. To accurately measure syrup brix, a novel measurement method based on support vector regression (SVR) is presented. With the resonant frequency and quality factor as inputs and syrup brix as the output, a mathematical model of the relationship between the resonant frequency, quality factor, and syrup brix is established. Simultaneously, the particle swarm optimization (PSO) algorithm is used to optimize the penalty coefficient and radial basis kernel function of SVR to improve the performance of the model. The calculation model is trained and tested using the collected experimental data. The results show that the mean absolute error, mean absolute percentage error, and root mean square error of the syrup brix calculation model based on the improved SVR model can reach 0.74 °Bx, 2.24%, and 0.90 °Bx, respectively, while the determination coefficient can reach 0.9985. The simulation of the online measurement of syrup brix in the actual production process proves the excellent prediction performance of the syrup brix calculation model based on the improved PSO–SVR model, which can thus be used to predict the syrup brix.

Hydrocarbon Generation from Low-Mature Saline Lacustrine Sediments Studied Using Machine Learning and Chemometric Methods: The Succession of the Sikeshu Sag, Junggar Basin, NW China.

Significant attention has been given to the extensive development of saline environments in petroliferous basins. Further exploration and studies have discovered that saline environments, such as those for the deposition of source rocks in the Paleogene Anjihaihe (E2-3 a) Formation of the Sikeshu Sag, are ubiquitous in terrestrial lake basins. Previous studies have suggested that the oil reservoirs in the Sikeshu Sag and its peripheral regions are predominantly derived from the black mudstone and coal measures of the Lower Jurassic Badaowan (J1 b) Formation. However, with deeper exploration of the study area, a growing number of reservoirs with geochemical characteristics different from the J1 b oil source have been discovered, indicating that there are oil sources other than the J1 b source rocks. In this study, various machine learning algorithms were used (random forest, RF; convolutional neural networks, CNN; extreme gradient boosting, XGBoost; ElasticNetCV; Bayesian Ridge; and particle swarm optimization-support vector regression) to select the most suitable algorithm for predicting and comparing the quality of potential source rocks. A violin plot and Taylor diagram were applied to visually compare the reliability and application effectiveness of the models. The results demonstrated that XGBoost and RF can become essential tools for predicting the quality of potential source rocks. Moreover, the measured and predicted values of total organic carbon (TOC), hydrocarbon potential (S1 + S2), and hydrogen index indicate that there are three main source rocks: the E2-3 a, Lower Jurassic Sangonghe (J1 s), and J1 b formations. The thermal maturity of the E2-3 a source rocks is still early mature because of the saline-brackish water nature of these rocks, although large-scale hydrocarbon generation and expulsion can be achieved in the early mature stage. Based on their geochemical characteristics and stepwise discriminant analysis, the oils in the Sikeshu Sag and its peripheral regions can be categorized into two types: groups A and B. Comprehensive organic geochemical evidence suggests that genetically, group A oils are originated from E2-3 a less-mature saline lacustrine sedimentary rocks, while group B oils indicate similar affinity to the Jurassic source. Fluid inclusion microthermometry and one-dimensional basin modeling showed that the oil charging periods of group A and B oils were Middle-Late Miocene (13-8 Ma) and Late Oligocene (23-20 Ma), respectively. Quantitative grain fluorescence (QGF) analysis further propose that the hydrocarbon supply region of the E2-3 a sources is mainly located east of the Western Chepaizi Uplift and the interior area of the Sikeshu Sag, which breaks through the previous understanding that the Jurassic coal-derived oil source is the only main contributor in this study area. The research results can be widely applied to assess the petroleum resources of source rocks in similar areas worldwide.

Predicting physical fatigue in athletes in rope skipping training using ECG signals

Physical fatigue is a crucial factor that leads to a decrease in performance, especially in athletes. This study proposes effective physiological indicators and methods to predict physical fatigue. We designed a ”training-cognitive task” experiment using specialized electrocardiogram (ECG) acquisition devices (Shimmer3) to collect ECG data during the training process. The participants were required to complete a 2-back cognitive task before and after one-hour of rope skipping training. After the experiment, we analyzed the heart rate variability (HRV) signals of the subjects using the time domain, frequency domain, and nonlinear dynamics. Pearson’s correlation coefficient analysis and paired t-tests were applied to the measured indicators. Subsequently, selected ECG indicators with statistical significance, the particle swarm optimization-support vector regression (PSO-SVR), and extreme learning machine (ELM) were used to build regression models that predicted subjective and objective physical fatigue. Moreover, we have compared the performance of Gaussian process regression (GPR), random forest (RF), k-nearest neighbors (KNN), and adaptive boosting (AdaBoost) based machine learning model for prediction. The results indicated that PSO-SVR had a better performance in predicting subjective physical fatigue, with mean absolute percentage error (MAPE) of 11.53%, and ELM achieved competitive outcomes for objective physical fatigue prediction with MAPE of 19.77%. This study provides a reliable experimental basis for identifying effective physical fatigue.

Sea Surface Salinity Inversion Model for Changjiang Estuary and Adjoining Sea Area with SMAP and MODIS Data Based on Machine Learning and Preliminary Application

Sea surface salinity (SSS) is one of the most important basic parameters for studying the oceanographic processes and is of great significance in identifying oceanic currents. However, for a long time, the salinity observation in the estuary and coastal waters has not been well resolved due to the technology limitation. In this study, the SSS inversion models for the Changjiang Estuary and the adjacent sea waters were established based on machine learning methods, using SMAP (Soil Moisture Active and Passive) salinity data combined with the specific bands and bands ratios of MODIS (Moderate Resolution Imaging Spectroradiometer). The performance of the three machine learning methods (Random Forest, Particle Swarm Optimization Support Vector Regression (PSO-SVR) and Automatic Machine Learning (TPOT)) are compared with accuracy verification by the in-situ measured SSS. Random Forest is proven to be effective for the SSS inversion in flood season, whereas TPOP performs the best for the dry season. The machine learning-based models effectively solve the problem of insufficient time span of SSS observation from salinity satellites. At the same time, an empirical algorithm was established for the SSS inversion for the sea areas with low salinity (<30 psu) where the machine learning based model fails with great errors. The average deviation of the complex SSS inversion models is −0.86 psu, validated with Copernicus Global Ocean Reanalysis Data. The long term series SSS dataset of March and August from 2003 to 2020 was then constructed to observe the salinity distribution characteristics of the flood season and the dry season, respectively. It is indicated that the distribution pattern of CDW can be divided into three categories: northeast-oriented expansion pattern, multi direction isotropic expansion pattern, and a turn pattern of which CDW shows changing direction, namely the northeast-southeast expansion pattern. The pattern of CDW expansion is indicated to be the comprehensive effect of the interaction of different currents. In addition, it is noteworthy that CDW shows increasing expansion with decreasing SSS in the front plume, especially in the flood season. This study not only gives a feasible solution for effective SSS observation, but also provides a dataset of basic oceanographic parameters for studying the coastal biogeochemical processes, evolution of land-sea interaction, and changing trend of material and energy transport by the CDW in the west Pacific boundary.

Field experiment and numerical investigation on the mechanical response of buried pipeline under traffic load

The acceleration of global urbanization and an increasing intersection of roads and buried pipelines pose a certain threat to the operation safety of pipelines. In this paper, the mechanical response of the X65 pipeline under empty-load, half-load, and full-load traffic loads is investigated via field experiments. The numerical model considering the nonlinear interaction between the pipe and the soil is established. Moreover, an accurate simulation of dynamic vehicle load is achieved by Dload user-defined subroutine. The error between the numerical calculation results and the field experiment results is less than 5 %. On this basis, the influence of five important factors, such as vehicle mass and buried depth, on the axial stress of pipelines is further explored. The results show that the traffic load leads to vertical bending deformation of the pipeline and local deformation of the wheel rolling position. The axial tensile stress generated at the bottom of the pipe is superimposed on the axial tensile stress generated by the internal pressure Poisson effect. The vehicle mass most significantly affects the axial stress of the pipeline, which is generally less than 100 MPa. Furthermore, the operating conditions of the pipeline with large coverage in service are calculated based on Python and ABAQUS co-simulation calculation of 1280 working conditions. The particle swarm optimization-support vector regression (PSO-SVR) hybrid machine learning model is used to accurately predict the axial stress of the buried pipeline under traffic load, and the corresponding correlation coefficient reaches 99.77 %. The research results reveal the mechanical response law of buried pipeline under traffic load, which can provide a basis for the applicability evaluation of buried pipeline in service under traffic load, and accurately evaluate the safety state of buried pipeline.

Estimating Leaf Nitrogen Content in Wheat Using Multimodal Features Extracted from Canopy Spectra

The leaf nitrogen content (LNC) of wheat is one of key bases for wheat nitrogen fertilizer management and nutritional diagnosis, which is of great significance to the sustainable development of precision agriculture. The canopy spectrum provides an effective way to monitor the nitrogen content of wheat. Previous studies have shown that features extracted from the canopy spectrum, such as vegetation indices (VIs) and band positions (BPs), have successfully achieved the monitoring of crop nitrogen nutrition. However, the features mentioned above are spectral features extracted on the basis of linear or nonlinear combination models with a simple structure, which limits the general applicability of the model. In addition, models based on spectral features are prone to overfitting, which also reduces the accuracy of the model. Therefore, we propose an estimation model based on multimodal features (convolutional features and VIs, BPs) of the canopy spectrum, which aim to improve accuracy in estimating wheat LNC. Among these, the convolutional features (CFs) extracted by the designed convolutional neural network represent the deep semantic information of the canopy reflection spectrum, which can make up for the lack of robustness of the spectral features. The results showed that the accuracy of the model based on the fusion features (VIs + BPs + CFs) was higher than that of the feature of single modality. Moreover, the particle swarm optimization–support vector regression (PSO-SVR) model based on multimodal features had the best prediction effect (R2 = 0.896, RMSE = 0.188 for calibration, R2 = 0.793, RMSE = 0.408 for validation). Therefore, the method proposed in this study could improve performance in the estimation of wheat LNC, which provides technical support for wheat nitrogen nutrition monitoring.

On-line mechanical life prediction method for a conventional circuit breaker based on multi-parameter particle swarm optimization-support vector regression using vibration detection

For on-line mechanical condition monitoring of conventional circuit breakers, based on the analysis of vibration signal measurements and the generation principle in the circuit breaker breaking motion process, the time parameters for mechanism action with clear physical meaning are correlated with vibration events. Thus, an on-line mechanical life prediction method based on vibration detection and multi-parameter support vector regression (MP-SVR) optimized by particle swarm optimization is proposed. This method aims to use the action time parameter as the main monitoring parameter to reflect the degradation of mechanical conditions and the life failure criterion. First, the complete MPs of life degradation are obtained through the designed on-line measurement algorithm for the action time parameter and the assisted time-domain features of vibration signals in the corresponding time. Then, a quantitative life prediction model is constructed based on MP-SVR. Finally, the degraded MPs are used as input to the prediction model to predict the remaining mechanical life. Our experiment shows that this on-line model for life evaluation is definitely effective and practical in engineering applications.

Time Series Forecasting Based on Support Vector Machine Using Particle Swarm Optimization

In recent years, due to the non-linear nature, complexity, and irregularity of time series, especially in energy consumption and climate, studying this field has become very important. Therefore, this study aims to provide a high accuracy and efficiency hybrid approach to time series forecasting. The proposed model is called EDFPSO-SVR (Empirical Mode Decomposition- Discrete Wavelet Transform- Feature selection with Particle Swarm Optimization-Support Vector Regression). In the proposed hybrid approach, the first step is to decompose the signal into the Intrinsic Mode Functions (IMF) component using the Empirical Mode Decomposition (EMD) algorithm. In the second step, each component is transformed into subsequences of approximation properties and details by converting the Wavelets. In the third step, the best feature is extracted by the PSO algorithm. The purpose of using the PSO algorithm is feature extraction and error minimization of the proposed approach. The fourth step, using time vector regression, has dealt with time series forecasting. Four data sets in two different fields have been used to evaluate the proposed method. The two datasets are electric load of England and Poland, and the other two datasets are related to the temperature of Australia and Belgium. Evaluation criteria include MSE, RMSE, MAPE, and MAE. The evaluation results of the proposed method with other Principal component analysis (PCA) feature extraction algorithms, and comparisons with methods and studies in this field, indicate the proper performance of the proposed approach.

A state of health estimation method for electric vehicle Li-ion batteries using GA-PSO-SVR

State of health (SOH) is the ratio of the currently available maximum capacity of the battery to the rated capacity. It is an important index to describe the degradation state of a pure electric vehicle battery and has an important reference value in evaluating the health level of the retired battery and estimating the driving range. In this study, the random forest algorithm is first used to find the most important health factors to lithium-ion batteries based on the dataset released by National Aeronautics and Space Administration (NASA). Then the support vector regression (SVR) algorithm is developed to predict the SOH of a lithium-ion battery. The genetic algorithm-particle swarm optimization (GA-PSO) algorithm is brought forward to optimize the parameter values of the SVR, which could improve the estimation accuracy and convergence speed. The proposed SOH estimation method is applied to four batteries and gets a root mean square error (RMSE) of 0.40% and an average absolute percentage error (MAPE) of 0.56%. In addition, the method is also compared with genetic algorithm-support vector regression (GA-SVR) and particle swarm optimization-support vector regression (PSO-SVR), respectively. The results show that (i) compared with the PSO-SVR method, the proposed method can decrease the average RMSE by 0.10%, and the average MAPE by 0.17%; (ii) compared with the GA-PSO method, number of iterations under the proposed method can be reduced by 7 generations.

A Hybrid Machine Learning Model for Reliability Evaluation of the Reactor Protection System

Reliability evaluation is a vital issue for safety and continuous progress of critical safety system such as Reactor Protection System (RPS) in Nuclear Power Plants (NPPs). In this study, a hybrid reliability evaluation approach of the RPS based on machine learning algorithms is proposed. First, the significant reliability factors in RPS are classified to four sections: Interlocks, over-power DT-shutdown, over-temperature DT-shutdown, and logic circuit reactor shutdown. Each of these sections represents a subsystem of the integrated RPS safety system. Then, a feature selection strategy based on Principal Component Analysis (PCA) is applied to exclude low impact factors. Second, a hybrid Particle Swarm Optimization - Support Vector Regression (PSO-SVR) algorithm is proposed to evaluate the reliability for each subsystem. Third, a hybrid Particle Swarm Optimization – Artificial Neural Network (PSO-ANN) is developed to associate the reliability of the four subsystems with the reliability of the entire RPS. Finally, residual error correction of Markov chain is adapted to enhance the prediction performance of the proposed model. The proposed approach can estimate more accurate prediction error of 0.0892 compared to other the state-of-art models. The RPS unavailability measure using our approach is equal to 1.30E-06 compared to values estimated using other methods.

Short-Term Electric Load Forecasting Using Auto Regressive-Candidates Area Shifting Technique

Electric power load forecasting is needed to be used as a consideration in providing electricity in the future. A combination of the Auto Regressive model and the Candidates Area Shifting Technique is proposed to predict the demand for electrical loads. The results of the proposed method are compared with the ones of the hybrid Particle Swarm Optimization-Support Vector Regression and FCM Clustering Technique methods. The results show that the proposed method provides better performance than the other ones. The three methods provide mean absolute percentage error and maximum absolute percentage error respectively as follows: Particle Swarm Optimization-Support Vector Regression Method 2.859% and 9,516%, FCM Clustering 1.032% and 2.798%, and Auto Regressive-Candidates Area Shifting Technique 0.298% and 0.872%, respectively.

Maximum Power Point Tracking for a PV System Using Tuned Support Vector Regression by Particle Swarm Optimization

The photovoltaic (PV) system is always operated at the maximum power point (MPP) condition irrespective of the fluctuations in PV voltage. The maximum power point tracking (MPPT) employed in PV system is not effective during the presence of current ripple as normal tracking becomes increasingly complex during fluctuation in solar irradiation or due to change in MPP condition. This paper proposes a high-efficiency power point tracking algorithm to minimize the current ripple and power oscillation around the maximum power point. The developed algorithm is based on particle swarm optimization-support vector regression (PSO-SVR) technique. The proposed algorithm is implemented to select and tune the Support Vector Regression (SVR) parameters such as kernel parameters, variance, and the penalty factor for predicting the irradiation level as well as to determine the PV voltage corresponding of maximum power point. The PSO method is used to accelerate the process of optimizing the SVR parameters at different conditions and get knowledge about the corresponding global optimum. From the experimental results,the efficiency of maximum power point tracking is found to be 99.8%. The proposed algorithm PSO-SVR shows a better performance than using SVR alone. The stability and accuracy of MPPT have been validated during the rapid fluctuation of solar irradiation in the range of 25% to 100%.