An Improved Quadratic Spline Model Using Curvature Tip Compression - Particle Swarm Optimization to Forecast Accurately the Nonlinear Fluid Calibration Curve

Omar M., Yakub F., Abdullah S.S., Rahim M.S., Zuhairi A.H., Govindan N.

In the increasingly complex urban transportation landscape, the necessity for accurate, multi-step forecasting has never been more apparent to help plan for long-term, strategic transportation initiatives like Intelligent Transportation Systems. This work focuses on resolving uncertainties around the proper training step size for machine learning models performing multi-step traffic flow forecasting. Two strategies are considered: one-step and horizon-step training sizes, which extend from the single-step forecasting method. This paper compares and evaluates two machine learning models: the Support Vector Regression and Long Short-Term Memory (LSTM). Data from two country road locations, including urban roads in Kuala Lumpur and the I5-North freeway in California, are employed to forecast traffic flow in 1-hour increments, projecting from 2 h up to 24 h ahead. The results reveal a significant difference in performance between the two training step sizes, with the one-step training size emerging as the more consistent and optimal strategy for both Direct and Multi-Input Multi-Output forecasting strategies. The proposed model, the Direct Particle Swarm Optimization Grid Search Support Vector Regression (Direct-PSOGS-SVR) model, exhibited comparable or slightly better performance than the LSTM model in both environments. In the I5-North freeway dataset, the Direct-PSOGS-SVR achieved similar Root Mean Squared Error values across most forecasting tasks compared to the LSTM, indicating its effectiveness in stable traffic flow patterns. Notably, in the more dynamic urban environment of Kuala Lumpur, the Direct-PSOGS-SVR model also demonstrated stability and resilience in forecasting accuracy, effectively handling the inherent noise and fluctuations in traffic patterns. These findings serve as a valuable guide for practitioners in selecting the most efficacious combination of training strategies for specific time series forecasting tasks utilizing machine learning models, particularly in traffic flow forecasting. Introducing the Direct-PSOGS-SVR model enriches the landscape of machine learning solutions for traffic forecasting, underscoring the paper's contributions to the broader understanding of time series forecasting dynamics.

Song Y., Gao M., Wang J.

Soil salinization hinders sustainable agricultural development in coastal regions. Developing a multi-layer soil salinity inversion model and accurately predicting water demand for salt regulation are essential for improving soil salinity management. Wudi County in Shandong Province was selected as the research area, with 79 sampling sites chosen. Soil salinity was measured at the surface (0–20 cm), middle (20–40 cm), and bottom (40–60 cm) layers. Vegetation and salinity indices were extracted from Landsat 8 remote sensing imagery to estimate surface soil salinity. A correlation-based inversion method was developed to obtain multi-layer soil salinity data by leveraging the strong correlation between adjacent soil layers. The water requirement for salt regulation was optimized and predicted by integrating the results from multi-layer soil salinity estimation with Groundwater Management System (GMS) software. The results indicated that the surface layer soil salinity inversion model performed well, with an R2 > 0.75 and an RMSE < 0.43 g/kg for both the training and validation sets. Additionally, the prediction accuracy of the correlation-based inversion method exceeded that of the direct modeling approach, with the middle and bottom layer soil salinity models achieving an R2 > 0.6 and an RMSE < 1 g/kg. Soil salinization in the study area was more severe in the northeast than in the southwest, with both measured and estimated data showing similar spatial distributions. Over the past decade, the overall trend of soil salinization has shown a general decline with localized intensification. The salt distribution patterns in saline soil profiles were predominantly homogeneous and bottom-accumulated. The projected water demand for salt regulation calculated from the estimated data was slightly lower than the actual measurements, yet their spatial distribution was nearly identical. This study provides a scientific foundation for the dynamic monitoring and precise management of soil salinity in coastal regions.

Fang J., Wang Y.

High-precision waveform decomposition is crucial for LiDAR applications. Existing methods encounter challenges including poor target detection and low accuracy in extracting parameters of irregular components, especially in complex echoes. We introduce an adaptive B-spline-based decomposition (AdaptB-spline) method, which uses B-spline curves to adaptively adjust the shape and position of component through the particle swarm optimization (PSO); and proposes an initial parameter estimation method based on the B-spline and Richardson-Lucy (RL) deconvolution, which improves the noise immunity and component detection. Experiments were conducted on synthetic waveforms and satellite LiDAR waveforms by AdaptB-spline and other four methods (Gaussian (Gauss), B-spline-based (B-spline), skew-normal (SkewN), and multi-Gaussian (MultiGauss) decomposition). We concluded that AdaptB-spline exhibits superior performance in terms of component RMSE, CC, R2, component parameter error and range error metrics compared to the four methods. So AdaptB-spline can enhance component detection and accurately fit Gaussian or non-Gaussian waveforms, demonstrating outstanding target detection and ranging precision.

Zhang E., Yang S., Zhang L.

Romano D., Kovacevic-Badstuebner I., Antonini G., Grossner U.

Zhi Hui T., Syazwani Mohd Ali N., Sabri Minhat M., Zainal J., Arif Sazali M., Syahir Sarkawi M., Jamaluddin K., Afifah Basri N., Mohd Sies M., Khair Alang Md Rashid N.

One of the control rod calibration methods in research reactors is the doubling time. However, this method reduces the operation time and limits the number of research activities. A pre-calibration method is proposed in this study by utilizing an ANFIS method. Two data inputs based on the annual rod worth, and the worth drop of the Shim and Transient rods were collected to predict the Safety and Regulating rod's worth. The results showed that ANFIS can predict the Safety rod worth with the lowest MAE and RMSE errors of 0.0156 and 0.0204 while 0.0616 of the percent error. For the Regulating rod's worth, the predicted and actual output data were less accurate due to the composition of air in the Transient rod. Nevertheless, the utilization of ANFIS as the pre-calibration method for control rod calibration at research reactors could be implemented and optimized in future studies.

K. V., Gopi E.S., Agnibhoj T.

Camera-based object tracking systems in a given closed environment lack privacy and confidentiality. In this study, light detection and ranging (LiDAR) was applied to track objects similar to the camera tracking in a closed environment, guaranteeing privacy and confidentiality. The primary objective was to demonstrate the efficacy of the proposed technique through carefully designed experiments conducted using two scenarios. In Scenario I, the study illustrates the capability of the proposed technique to detect the locations of multiple objects positioned on a flat surface, achieved by analyzing LiDAR data collected from several locations within the closed environment. Scenario II demonstrates the effectiveness of the proposed technique in detecting multiple objects using LiDAR data obtained from a single, fixed location. Real-time experiments are conducted with human subjects navigating predefined paths. Three individuals move within an environment, while LiDAR, fixed at the center, dynamically tracks and identifies their locations at multiple instances. Results demonstrate that a single, strategically positioned LiDAR can adeptly detect objects in motion around it. Furthermore, this study provides a comparison of various regression techniques for predicting bounding box coordinates. Gaussian process regression (GPR), combined with particle swarm optimization (PSO) for prediction, achieves the lowest prediction mean square error of all the regression techniques examined at 0.01. Hyperparameter tuning of GPR using PSO significantly minimizes the regression error. Results of the experiment pave the way for its extension to various real-time applications such as crowd management in malls, surveillance systems, and various Internet of Things scenarios.

Zhai H., Song X., Wang X., Liu G.

Rochmanto B., Setiapraja H., Haryono I., Yubaidah S.

Calibrating a gas flowmeter requires a suitable working medium that represent its use in the real operating condition. However, factors like installation conditions, costs, and safety considerations may necessitate to utilize a different medium for calibration. The accuracy of measuring fluid flow by a flowmeter is greatly influenced by the kinematic viscosity property of the gas medium. In this particular study, a turbine flowmeter for compressed natural gas (CNG) application was calibrated using air as a substitute with an approach that simulated the kinematic viscosity property of CNG. Here, the pressure and temperature of the air were calculated to achieve the same kinematic viscosity value as CNG during the calibration process, which was accomplished by adjusting the setting of a regulating instrument in the calibration installation. Furthermore, argon (Ar) was also used for comparison with air, with a modification in working pressure to attain similar kinematic viscosity properties for both gases. The calculation results showed that the working pressure setting for the air used was 4.5 bars, while for argon was 3.38 bars to achieve similar kinematic viscosity. By utilizing air instead of CNG, a minimum and maximum volume flowrate difference of −0.0242 m3/h and 0.0016 m3/h, respectively, was observed, with an absolute average of 0.0086 m3/h or 0.8 %. This study reveals the effectiveness of the kinematic viscosity approach in calibrating a turbine flowmeter under working pressure of up to 4.5 barg. This process results in uncertainty measurement of under 1 % when calibrating gas flowmeters, using air as the medium.

Wang X., Bai X., Zhang M., Wu C.

Su J., Xiao Y., Chen S., Zhang C., Gao W., Fu Y., Bai X., Wu C.

Bashir R.R., Saeed Y., Ali A., Algarni A.D., Muthanna A., Hijjawi M., Alsboui T.

Shukla J.P., Singh B.K., Cattani C., Gupta M.

Zhao W., Wang L., Zhang Z., Mirjalili S., Khodadadi N., Ge Q.

An original math-inspired meta-heuristic algorithm, named quadratic interpolation optimization (QIO), is proposed to address numerical optimization and engineering issues. The main inspiration behind QIO is derived from mathematics, specifically the newly proposed generalized quadratic interpolation (GQI) method. This method overcomes the limitations of the traditional quadratic interpolation method to better find the minimizer of the quadratic function formed by any three points. The QIO utilizes the GQI method as a promising searching mechanism for tackling various types of optimization problems. This searching mechanism delivers exploration and exploitation strategies, in which the minimizer provided by the GQI method assists the QIO algorithm in exploring a promising region in unexplored areas and exploit the optimal solutions in promising regions. To evaluate QIO’s effectiveness, it is comprehensively compared with 12 other commonly used optimizers on 23 benchmark test functions and the CEC-2014 test suite. Ten engineering problems are also tested to assess QIO’s practicality. Eventually, a real-world application of QIO is presented in the operation management of a microgrid with an energy storage system. The results demonstrate that QIO is a promising alternative for addressing practical challenges. The source code of QIO is publicly available at https://ww2.mathworks.cn/matlabcentral/fileexchange/135627-quadratic-interpolation-optimization-qio.

Rustandi D., Prakosa J.A., Purwowibowo P., Wijornako S., Maftukhah T., Kurniawan E., Sirenden B.H., Mahmudi M.

Prakosa J.A., Alias N., Purwowibowo P., Algarni A.D., Soliman N.F.