Least Squares Support Vector Machine Research Articles

A Hybrid Soft Sensor Model for Measuring the Oxygen Content in Boiler Flue Gas.

As an indispensable component of coal-fired power plants, boilers play a crucial role in converting water into high-pressure steam. The oxygen content in the flue gas is a crucial indicator, which indicates the state of combustion within the boiler. The oxygen content not only affects the thermal efficiency of the boiler and the energy utilization of the generator unit, but also has adverse impacts on the environment. Therefore, accurate measurement of the flue gas's oxygen content is of paramount importance in enhancing the energy utilization efficiency of coal-fired power plants and reducing the emissions of waste gas and pollutants. This study proposes a prediction model for the oxygen content in the flue gas that combines the whale optimization algorithm (WOA) and long short-term memory (LSTM) networks. Among them, the whale optimization algorithm (WOA) was used to optimize the learning rate, the number of hidden layers, and the regularization coefficients of the long short-term memory (LSTM). The data used in this study were obtained from a 350 MW power generation unit in a coal-fired power plant to validate the practicality and effectiveness of the proposed hybrid model. The simulation results demonstrated that the whale optimization algorithm-long short-term memory (WOA-LSTM) model achieved an MAE of 0.16493, an RMSE of 0.12712, an MAPE of 2.2254%, and an R2 value of 0.98664. The whale optimization algorithm-long short-term memory (WOA-LSTM) model demonstrated enhancements in accuracy compared with the least squares support vector machine (LSSVM), long short-term memory (LSTM), particle swarm optimization-least squares support vector machine (PSO-LSSVM), and particle swarm optimization-long short-term memory (PSO-LSTM), with improvements of 4.93%, 4.03%, 1.35%, and 0.49%, respectively. These results indicated that the proposed soft sensor model exhibited more accurate performance, which can meet practical requirements of coal-fired power plants.

Biochar-enhanced concrete mixes: Pioneering multi-objective optimization

The incorporation of biochar (BC) as a partial substitute for cement in concrete formulations provides a promising pathway for mitigating the environmental impacts associated with carbon dioxide emissions during cement production. To achieve optimal composition of BC-enhanced concrete, multiple objectives including mechanical strength, economic factors, and embodied CO2 must be balanced while considering a multitude of variables constrained by complex non-linear relationships. Here, we introduced an intelligent hybrid optimization algorithm that combines Particle Swarm Optimization (PSO), Least Squares Support Vector Machine (LSSVM), and Non-dominated Sorting Genetic Algorithm II (NSGAII) to predict the performance of BC-enhanced concrete and optimize its mix proportions for multiple objectives. Our results demonstrate that the PSO optimization algorithm for searching LSSVM hyper parameters outperforms other optimization algorithms, exhibiting higher generalization performance and improved overall accuracy (R2 = 0.95). Furthermore, the proposed model framework effectively presents a complete Pareto front for the BC-enhanced concrete mix ratio. This triangular model system for concrete mix ratio can determine the optimal solution, tailored to the preferences of the owner's unit. Ultimately, PSO-LSSVM-NSGAII intelligent hybrid optimization algorithm enhances the efficiency of mix proportion design for BC-enhanced concrete.

Crater-Spectrum Feature Fusion Method for Panax notoginseng Cadmium Detection Using Laser-Induced Breakdown Spectroscopy.

Panax notoginseng (P. notoginseng) is a valuable herbal medicine, as well as a dietary food supplement known for its satisfactory clinical efficacy in alleviating blood stasis, reducing swelling, and relieving pain. However, the ability of P. notoginseng to absorb and accumulate cadmium (Cd) poses a significant environmental pollution risk and potential health hazards to humans. In this study, we employed laser-induced breakdown spectroscopy (LIBS) for the rapid detection of Cd. It is important to note that signal uncertainty can impact the quantification performance of LIBS. Hence, we proposed the crater-spectrum feature fusion method, which comprises ablation crater morphology compensation and characteristic peak ratio correction (CPRC), to explore the feasibility of signal uncertainty reduction. The crater morphology compensation method, namely, adding variables using multiple linear regression (MLR) analysis, decreased the root-mean-square error of the prediction set (RMSEP) from 7.0233 μg/g to 5.4043 μg/g. The prediction results were achieved after CPRC pretreatment using the calibration curve model with an RMSEP of 3.4980 μg/g, a limit of detection of 1.92 μg/g, and a limit of quantification of 6.41 μg/g. The crater-spectrum feature fusion method reached the lowest RMSEP of 2.8556 μg/g, based on a least-squares support vector machine (LSSVM) model. The preliminary results suggest the effectiveness of the crater-spectrum feature fusion method for detecting Cd. Furthermore, this method has the potential to be extended to detect other toxic metals in addition to Cd, which significantly contributes to ensuring the quality and safety of agricultural production.

Estimating the natural gas compressibility factor using a statistical correlations and machine learning approaches

Gas compressibility factor plays an critical role in petroleum engineering applications such as gas metering, pipeline design, reserve estimation, gas flow rate, material balance calculations, and many other significant tasks. Therefore, it is crucial to accurately estimate the gas compressibility factor. There have been a lot of studies on calculating the gas compressibility factor from laboratory data, which can be summarized into two main approaches: statistical correlations and machine learning algorithms. In this study, on statistical correlations the authors implement explicit and implicit method while on machine learning algorithms, we use Artificial Neural Network (ANN) and Least-Squares Support Vector Machine (LS-SVM). The data was collected from open literature. Implementing the two approaches mentioned above and comparing statistical parameters such as Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Coefficient of Determination (R2 ) found that machine learning algorithms give much more accurate estimation results than statistical correlations, and besides, the ANN algorithm has the most accurate prediction results with the lowest MSE and RMSE (0.000002 and 0.0016) and the highest R2 (0.9999). The high-precision calculation results show that the ANN algorithm mentioned above can be applied to estimate other real gas compressibility factor data sets. On the other hand, this study can be extended to another subset of machine learning algorithms, such as deep learning and ensemble learning.

Study on detecting main ingredients of silicone rubber based on terahertz spectrum

AbstractThe authors investigated the ingredient detection technique of silicone rubber based on the Terahertz spectrum. For this purpose, 18 diverse high‐temperature vulcanised silicone rubber (HTVSR) formulations were customised, 8 of which are used as calibration set while the rest 10 as prediction set. Based on the Beer‐Lambert Law, the partial‐least‐square (PLS) regression model and the least‐squares support‐vector machines (LS‐SVM) regression model were used to yield the relationships between the absorption spectrums and the content percentages of polydimethylsiloxane (PDMS), alumina trihydrate (ATH), and silica in HTVSR. The results showed that for the formulations tested, the prediction accuracy of all three main ingredients by the PLS regression model could be improved by changing the spectrum range from 0.2–4 to 0.5–2 THz. If the data were pre‐processed by the Savitzky–Golay smoothing method or multiplicative scatter correction method, the prediction accuracy of PDMS could be further enhanced. However, this would lead to a slight decrease in the prediction accuracy of ATH. For the LS‐SVM regression model, the radial basis function (RBF) kernel and the linear kernel were studied. It was found that the prediction accuracy of both kernels was better than that of the PLS regression model. With the LS‐SVM regression model using the RBF kernel, the correlated coefficients of PDMS and ATH in the prediction set could be up to 0.9915 and 0.9742, respectively.

New insights into culvert scour depth calculation by soft computing models using multi-model ensemble approach and uncertainty analysis

ABSTRACT The accurate prediction of scour depth in culverts is crucial to ensure public safety and due to the uncertainties in classical empirical/deterministic equations, it remains challenging. This study presents new ensemble multi-models to predict scour depth and quantify model bias and uncertainty. In this study, culvert scour was predicted by gene expression programming (GEP) and least square support vector machine (LSSVM), firstly based on dimensionless parameters. In the multi-model ensemble strategy, four ensemble approaches, namely SM, WM, LSSVM, and BMA were applied. The LSSVM showed overall best skill for the multi-model approach, that reduced the error by 61.3% and increased the coefficient of determination by 52.1%. The results showed the proposed approach can satisfy the requirements of simplicity, applicability, and accuracy at the same time. For application purposes, we provided a simple code to obtain scour depth with great accuracy using LSSVM. BMA and LSSVM were used to quantify the scour culvert uncertainty. The results of the uncertainty analysis showed that the LSSVM predicts scour depth quite well and generates high-quality prediction intervals. The results revealed that a multi-model ensemble strategy improved accuracy, certainty and reliability for culvert scour compared to empirical equations.

Underwater Sound Speed Field Forecasting Based on the Least Square Support Vector Machine

Underwater sound speed is one of the most significant factors that affects high-accuracy underwater acoustic positioning and navigation. Due to its complex temporal variation, the forecasting of the underwater sound speed field (SSF) becomes a challenging task. Taking advantage of machine learning methods, we propose a new method for SSF forecasting based on the least square support vector machine (LSSVM) and a multi-parameter model, aiming to enhance the forecasting accuracy of underwater SSF with hourly resolution. We first use a matching extension method to standardize profile data and train the LSSVM with the parameters of observation time, temperature, salinity, and depth. We then employ radial basis function kernels to construct the forecasting model of SSF. We validate the feasibility and effectiveness of the LSSVM model by comparing it with the polynomial fitting (PF) and back propagation neural network (BPNN) methods, using hourly data obtained from the measured data and open data. The results show that the means of the root mean square for the LSSVM based on the observation time parameter and the LSSVM based on the multi-parameter model achieve 0.51 m/s and 0.45 m/s, respectively, presenting a significant improvement compared with the PF (0.82 m/s) and BPNN (0.76 m/s) methods.

Hybrid Metaheuristic Optimization Algorithms with Least-Squares Support Vector Machine and Boosted Regression Tree Models for Prediction of Air-Blast Due to Mine Blasting

Hybrid Metaheuristic Optimization Algorithms with Least-Squares Support Vector Machine and Boosted Regression Tree Models for Prediction of Air-Blast Due to Mine Blasting

Enhanced complex wire fault diagnosis via integration of time domain reflectometry and particle swarm optimization with least square support vector machine

AbstractUrban power systems rely on intricate wire networks, known as the power grid, which form the essential electric infrastructure in cities. While these networks transmit electricity from power plants to consumers, they are vulnerable to faults caused by manufacturing errors and improper installation, posing risks to system integrity. Thus, accurate identification and assessment of these faults are crucial to prevent damage and maintain system reliability. The objective of this research is to present an innovative and efficient methodology for diagnosing complex wire networks through the application of time domain reflectometry (TDR) combined with the particle swarm optimization (PSO) and least squares support vector machine (LSSVM) algorithm. This research addresses the imperative need to accurately locate and assess breakage faults within wire networks, emphasizing their role in both power transmission and communication infrastructure. To model the TDR answer of a specific complex wire network, a forward model is established utilizing resistance, inductance, capacitance and conductance (RLCG) parameters and the finite difference time domain (FDTD) method. Subsequently, the PSO‐LSSVM approach is used to solve the inverse problem of localizing faults in complex wire networks. The experimental results validate the practicality of this approach in real‐world systems.

Research on SPAD Estimation Model for Spring Wheat Booting Stage Based on Hyperspectral Analysis.

With the rapid progression of agricultural informatization technology, the methodologies of crop monitoring based on spectral technology are constantly upgraded. In order to carry out the efficient, precise and nondestructive detection of relative chlorophyll (SPAD) during the booting stage, we acquired hyperspectral reflectance data about spring wheat vertical distribution and adopted the fractional-order differential to transform the raw spectral data. After that, based on correlation analysis, fractional differential spectra and fractional differential spectral indices with strong correlation with SPAD were screened and fused. Then, the least-squares support vector machine (LSSSVM) and the least-squares support vector machine (SMA-LSSSVM) optimized on the slime mold algorithm were applied to construct the estimation models of SPAD, and the model accuracy was assessed to screen the optimal estimation models. The results showed that the 0.4 order fractional-order differential spectra had the highest correlation with SPAD, which was 9.3% higher than the maximum correlation coefficient of the original spectra; the constructed two-band differential spectral indices were more sensitive to SPAD than the single differential spectra, in which the correlation reached the highest level of 0.724. The SMA-LSSSVM model constructed based on the two-band fractional-order differential spectral indices was better than the single differential spectra and the integration of both, which realized the assessment of wheat SPAD.

Ultra-Wideband Ranging Error Mitigation with Novel Channel Impulse Response Feature Parameters and Two-Step Non-Line-of-Sight Identification.

The effective identification and mitigation of non-line-of-sight (NLOS) ranging errors are essential for achieving high-precision positioning and navigation with ultra-wideband (UWB) technology in harsh indoor environments. In this paper, an efficient UWB ranging-error mitigation strategy that uses novel channel impulse response parameters based on the results of a two-step NLOS identification, composed of a decision tree and feedforward neural network, is proposed to realize indoor locations. NLOS ranging errors are classified into three types, and corresponding mitigation strategies and recall mechanisms are developed, which are also extended to partial line-of-sight (LOS) errors. Extensive experiments involving three obstacles (humans, walls, and glass) and two sites show an average NLOS identification accuracy of 95.05%, with LOS/NLOS recall rates of 95.72%/94.15%. The mitigated LOS errors are reduced by 50.4%, while the average improvement in the accuracy of the three types of NLOS ranging errors is 61.8%, reaching up to 76.84%. Overall, this method achieves a reduction in LOS and NLOS ranging errors of 25.19% and 69.85%, respectively, resulting in a 54.46% enhancement in positioning accuracy. This performance surpasses that of state-of-the-art techniques, such as the convolutional neural network (CNN), long short-term memory-extended Kalman filter (LSTM-EKF), least-squares-support vector machine (LS-SVM), and k-nearest neighbor (K-NN) algorithms.

Sulphate resistance prediction of RAC considering the coupling effect of RCA and water-binder ratio based on experimental analysis and LSSVM

ABSTRACT Waste concrete removed from old pavement can be crushed and sieved as recycled aggregate to prepare concrete, and the application of recycled aggregate concrete (RAC) can reduce the occupation of land resources by the waste concrete and save natural aggregate. In this paper, the influence of proportion of recycled coarse aggregate (RCA) replacing natural coarse aggregate (NCA) and water-binder ratio on the sulphate resistance of RAC was studied, and the compressive strength degradation rate (Df ) was introduced to evaluate the sulphate resistance of RAC. The experimental results showed that the compressive strength of RAC before and after sulphate attacks linearly decreased with the growth of the replacement rate of RCA, water-binder ratio, and the number of sulphate wet-dry cycles, respectively; while Df of RAC increased with the rise of the replacement rate of RCA, water-binder ratio, and sulphate wet-dry cycles, respectively. The results also indicated that Df of all groups of RAC was less than 25%, which can meet the engineering requirements. In addition, the Df model of RAC under a sulphate attack based on experimental data and that based on Least-Squares Support Vector Machine (LSSVM) was built and compared, and both models predict the sulphate resistance of RAC well. The prediction model based on LSSVM has higher prediction accuracy.

Improved prediction of groundwater depth in urban areas using the EWT-S–G-GRU and LSSVM model: a case study of Xinxiang City

ABSTRACT The improvement in predicting groundwater depth accuracy has significant implications for the management, ecological environment protection, and economic and social development of regional water resources. Employing empirical wavelet transform (EWT) for nonlinear processing, Savitzky–Golay (S–G) filtering to reduce high-frequency noise, gate recurrent unit (GRU) neural network for linear feature signal processing, and least squares support vector machine (LSSVM) for nonlinear signal handling, we established a comprehensive model combining EWT-S–G-GRU and LSSVM. The results demonstrate that the proposed model exhibits superior accuracy in groundwater depth prediction, with an average relative error of only 2.14% and Nash–Sutcliffe efficiency (NSE) of 0.93. This outperforms the other four models, with average relative errors of 12.88%, 11.90%, 7.07%, and 11.10%, and NSE values of 0.58, 0.56, 0.71, and 0.73, respectively. The superiority of the model established in this study is attributed to its effective handling of both nonlinear and linear features of groundwater, thereby enhancing predictive accuracy. The EWT-S–G-GRU and LSSVM model proves to be more reliable in revealing the spatial distribution and dynamic changes of future groundwater, providing a robust reference for the rational development and utilization of groundwater in the urban area of Xinxiang City.

Method for Fault Diagnosis of Track Circuits Based on a Time–Frequency Intelligent Network

In response to the limitations posed by noise interference in complex environments and the narrow focus of existing diagnosis methods for jointless track circuit faults, an innovative approach is put forward in this study. It involves the application of the continuous wavelet transform (CWT) for signal preprocessing, along with the integration of a deep belief network (DBN) and a genetic algorithm (GA) to improve the least-squares support vector machine (LSSVM) model for intelligent time–frequency fault diagnosis. Initially, the raw induced voltage signals are transformed using continuous wavelet transformation resulting in wavelet time–frequency representations that combine temporal and spectral information. Subsequently, these time–frequency representations are fed into the deep belief networks, which perform semi-supervised dimensionality reduction and feature extraction, thereby uncovering distinct fault characteristics in the track circuit. Finally, the genetic algorithms are employed to improve the kernel function and penalty factor parameters of the least-squares support vector machine, thus establishing an optimal DBN-GA-LSSVM diagnostic model. Experimental validation demonstrates the effectiveness of the proposed time–frequency intelligent network model by leveraging the advantages of deep belief networks in hierarchical feature extraction and the superior performance of the least-squares support vector machine in addressing high-dimensional pattern recognition problems with limited samples. The achieved accuracy rate on the testing dataset reaches an impressive 99.6%. Consequently, this comprehensive approach provides a viable solution for data-driven track circuit fault diagnosis.

A novel robust intelligent fault diagnosis method for rolling bearings based on SPAVMD and WOA-LSSVM under noisy conditions

Aiming at the problem that the fault information of rolling bearings under harsh operation environment is easily submerged by strong noise interference, which causes the traditional method to be difficult to identify bearing faults effectively, this paper proposes a novel robust intelligent fault diagnosis method for rolling bearings based on sparsity-assisted parameter adjustable variational mode decomposition (VMD) and whale optimization algorithm-based optimized least-squares support vector machine (WOA-LSSVM). Firstly, a sparsity measurement named the improved Gini index is introduced as the fitness function of grid search algorithm to adaptively adjust and search for the optimal decomposed mode number K and penalty factor α of VMD. Additionally, VMD containing the optimal parameters is adopted to decompose the original bearing vibration signal into several intrinsic mode function (IMF), and the effective signal reconstruction is performed by screening the sensitive IMF components according to the effective weighted kurtosis Gini index criterion. Subsequently, the refine composite multi-scale dispersion entropy of the reconstructed signal is further calculated to establish a multi-dimensional feature vector set. Finally, the constructed feature vector set is fed into the WOA-LSSVM to achieve automatic fault identification of rolling bearings. The effectiveness of the proposed method is verified by two experimental examples. Experimental results show that the proposed method has higher fault recognition accuracy and better robustness against noise than other homologous methods in noisy conditions. This study provides a new perspective for the developing of robust diagnosis methods.

Crashworthiness Analysis and Optimization of Thin-Walled Structures Based on the TSSA-LSSVM Surrogate Model

Thin-Walled Structures (TWS) play a crucial role in enhancing both the collision safety and lightweight characteristics of vehicles. Structural design and optimization have demonstrated significant potential in improving their crashworthiness and overall weight reduction. Several interconnected factors, including the cross-sectional shape and parameters of distinctive structural characteristics (known as induction grooves), impact the TWS’ resistance to collisions. Another critical aspect in the optimization of TWS design is striking the delicate balance between the surrogate model’s effectiveness and precision. Taking all these factors into account, this paper conducts an in-depth investigation into the cross-sectional shape, grooves, and other geometric attributes of various TWSs through a comparative analysis. It reveals that the thin-walled structure with a Double-Ribbed Rectangle cross-section (DR-TWS) proves to be the optimal choice, particularly in low-speed axial impact scenarios. Furthermore, we present an enhanced Least Squares Support Vector Machine (LSSVM) method, augmented with an adaptive [Formula: see text]-distribution Sparrow Search Algorithm (TSSA), referred to as the TSSA-LSSVM surrogate model. The optimized design of the DR-TWS is realized through the integration of the TSSA-LSSVM surrogate model with the adaptive hybrid multi-objective particle swarm optimization (AHMOPSO) algorithm. This paper not only provides substantial theoretical support but also offers valuable guidelines for the application of TWSs in the automotive industry.

Feasibility of using colorimetric devices for whole and ground coffee roasting degrees prediction.

Coffee roasting is one of the crucial steps in obtaining a high-quality product as it forms the product's color and flavor characteristics. Roast control is made by visual inspection or traditional instruments such as the Agtron spectrophotometer, which can have high implementation costs. Therefore, the present study evaluated colorimetric approaches (a bench colorimeter, smartphone digital images, and a colorimetric sensor) to predict the Agtron roasting degrees of whole and ground coffee. Two calibration approaches were assessed, that is, multiple linear regression and least-squares support vector machine. For that, 70 samples of whole and ground roasted coffees comprising the Agtron roasting range were prepared. The results showed that all three colorimetric acquisition types were efficient for the model building, but the bench colorimeter and the smartphone digital images generally performed with good determination coefficients and low errors as measured by external validation. For the whole bean coffee, the best model presented a determination coefficient (R2) of 0.99 and a root-mean-squared error (RMSE) of 1.91%, while R2 of 0.99 and RMSE of 0.87% was obtained for ground coffee, both using the colorimeter. The obtained models presented good prediction capability, as assessed by external validation and randomization tests. The obtained findings point to an alternative for coffee roasting monitoring that can lead to higher digitalization and local control of the process, even for smaller producers, due to its lower costs. © 2024 Society of Chemical Industry.

Prediction of Energy Consumption in a Coal-Fired Boiler Based on MIV-ISAO-LSSVM

Aiming at the problem that the energy consumption of the boiler system varies greatly under the flexible peaking requirements of coal-fired units, an energy consumption prediction model for the boiler system is established based on a Least-Squares Support Vector Machine (LSSVM). First, the Mean Impact Value (MIV) algorithm is used to simplify the input characteristics of the model and determine the key operating parameters that affect energy consumption. Secondly, the Snow Ablation Optimizer (SAO) with tent map, adaptive t-distribution, and the opposites learning mechanism is introduced to determine the parameters in the prediction model. On this basis, based on the operation data of an ultra-supercritical coal-fired unit in Xinjiang, China, the boiler energy consumption dataset under variable load is established based on the theory of fuel specific consumption. The proposed prediction model is used to predict and analyze the boiler energy consumption, and a comparison is made with other common prediction methods. The results show that compared with the LSSVM, BP, and ELM prediction models, the average Relative Root Mean Squared Errors (aRRMSE) of the LSSVM model using ISAO are reduced by 2.13%, 18.12%, and 40.3%, respectively. The prediction model established in this paper has good accuracy. It can predict the energy consumption distribution of the boiler system of the ultra-supercritical coal-fired unit under variable load more accurately.

Power Transformer Classification through Dissolved Gas Analysis Utilizing Least-Square Support Vector Machine

Power Transformer Classification through Dissolved Gas Analysis Utilizing Least-Square Support Vector Machine

Rolling bearing fault diagnosis based on improved whale-optimization- algorithm–variational-mode-decomposition method

Rolling bearings are a key component of rotating machinery and their health directly affects the safe operation of mechanical equipment. Therefore, fault diagnose for rolling bearings is very important. The fault diagnosis process of rolling bearings includes three stages: signal decomposition, feature extraction, and pattern recognition. Variational mode decomposition (VMD) can suppress end effects, but improper parameter settings will cause information losses or excessive decomposition. In this work, an improved whale optimization algorithm (IWOA) is applied to parameter settings of VMD. Correspondingly, an IWOA-VMD signal decomposition method is proposed. The decomposed signal is combined with a Laplace score method and classifier to remove the redundancy and noise in the feature set and obtain a low-dimensional sensitive feature subset. Then, aiming at the problem of the parameter settings of a least squares support vector machine (LSSVM) affecting the recognition performance and accuracy, a salp swarm algorithm (SSA) is used to globally optimize the penalty parameter and kernel width in the LSSVM to establish an SSA-LSSVM fault recognition model. This model is applied to the fault diagnosis of rolling bearings. In particular, rolling bearing fault samples at Case Western Reserve University are used to verify the method. The results indicate that the proposed method is effective and improves the speed and accuracy of fault diagnosis.