Squared Research Articles

A Bayesian Shrinkage Approach for the Inverse Weibull Distribution under the Type-II Censoring Schemes

One main challenge in the application of the lifetime distribution models, such as inverse Weibull (IW) distribution is the need for an appropriate estimation method based on experimental conditions. When prior information and certain guessed values are available for model parameters the Bayesian shrinkage (BS) method becomes a valuable approach in this situation. This study considered the BS estimation method in the two-parameter IW distribution under the squared error loss function (SELF) and the type-II censored data. The maximum likelihood (ML), the least squares (LS), and Bayes estimation methods were also examined for a comparative study. Due to the complexity of calculations, the Lindley approach was utilized to approximate the Bayes estimates. The BS estimates were derived and a score test for the guessed value was presented. Additionally, a Monte Carlo simulation was conducted to evaluate the efficiency of all estimation methods. Furthermore, a real data set was implemented to illustrate and compare the BS estimates with the other estimates. The simulation study indicated the consistency of the estimators. The numerical studies also demonstrated that the BS estimators outperform the others.

Prediction of Summer Precipitation via Machine Learning with Key Climate Variables：A Case Study in Xinjiang, China

Study region: Xinjiang is located in the mid-latitude region of Eurasia in northwestern China. Precipitation is predominantly concentrated in northern Xinjiang, while southern Xinjiang remains comparatively arid. Summer precipitation accounts for 54.4% of the annual total. Study focus: This study aims to develop a machine learning model to predict summer precipitation (June–August) in XJ and explore the key variables contributing to summer precipitation in this region. The SHapley Additive exPlanations method was integrated with an extreme tree model to quantify the contributions of variables towards precipitation. Artificial neural networks, support vector machines, and extreme gradient boosting were considered to predict summer precipitation. To train the ML model, we used precipitation data from 1961 to 2012, whilst the forecast results from 2013 to 2017 were used for validation.New hydrological insights for the regions: The results demonstrated that the ANN model achieved robust performance during both the training and validation periods. For Northern and Southern XJ, the Mean Absolute Error and Root Mean Square Error of the ANN model were 15.34 (20.40) and 23.21 (30.01), respectively. The SHAP analysis showed that in the context of Northern Xinjiang, the Niño B Sea Surface Temperature Anomaly, Western Pacific Subtropical High Intensity, Pacific Subtropical High Intensity, and Multivariate ENSO Index play crucial roles in the prediction of summer precipitation. In Southern Xinjiang, the South China Sea Subtropical High Intensity, South China Sea Subtropical High Area, Western Pacific Warm Pool Strength, and Atlantic multidecadal oscillation have emerged as key variables affecting summer precipitation forecasting.

Prediction model for earthquake death toll based on PCA-BAS-ELM in mainland China

In recent years, China has experienced frequent catastrophic earthquakes, causing huge casualties. If the death toll can be quickly predicted after a disaster, then relief supplies can be delivered in a timely and reasonable manner, and the death toll and property losses can be minimized. Therefore, rapid and effective prediction of earthquake deaths plays a key role in guiding post-earthquake emergency rescue. However, there are many factors affecting the number of deaths in an earthquake. Aimed at this issue, a prediction model for earthquake deaths based on extreme learning machine (ELM) optimized by principal component analysis (PCA) and beetle antennae search (BAS) algorithm has been proposed in this study. Firstly, this study selected sample data of destructive earthquakes in mainland China in the past 50 years, then PCA was used to reduce the dimensionality of the factors affecting earthquake deaths, the principal components with lower contribution rates were removed, and the principal components with higher contribution rates were used as the input variables of ELM. Meanwhile, the earthquake deaths were used as the output variable, and the connection weights and thresholds of ELM was optimized using BAS. Finally, the prediction model for earthquake deaths based on PCA-BAS-ELM was established. The established model was used to predict the test samples. The results showed that the prediction results of PCA-BAS-ELM model had a higher fit with the actual values, and its mean square error, mean absolute percentage error and root mean square error were 2.433, 2.756% and 5.443, respectively, which suggested higher prediction accuracy.

Multi-model fusion method for predicting CO2 concentration in greenhouse tomatoes

With the rapid development of greenhouse agriculture, accurate prediction of environmental parameters such as temperature, humidity, and carbon dioxide concentration is crucial for optimal crop growth. Traditional forecasting models struggle with the nonlinear and complex nature of greenhouse data, leading to challenges in model robustness. This study addresses these issues by proposing a multi-model fusion strategy for predicting CO2 concentration in greenhouse tomatoes. The proposed method integrates wavelet denoising (WT), variational mode decomposition (VMD), and long short-term memory networks (LSTM). This innovative nonlinear ensemble model effectively extracts key time series features and removes noise, while an introduced attention mechanism enhances the model’s focus on essential time steps, improving prediction accuracy. Experimental results demonstrate that the multi-model fusion approach significantly outperforms single models in terms of accuracy and stability, achieving mean absolute error (MAE) and root mean square error (RMSE) of 0.0117 and 0.0194, respectively. The proposed method offers significant advantages for CO2 prediction in greenhouse crops, providing a theoretical basis and technical support for optimizing and controlling greenhouse parameters. This contributes to the advancement of smart agriculture by offering an efficient environmental monitoring and prediction tool. Additionally, the study presents new ideas and technical solutions for addressing similar agricultural environment prediction challenges, optimizing greenhouse environment control strategies, and improving crop production efficiency.

A novel hybrid deep learning model for real-time monitoring of water pollution using sensor data

This study developed a hybrid model for predicting dissolved oxygen (DO) using real-time sensor data for thirteen parameters. This novel hybrid model integrated one-dimensional convolutional neural networks (CNN) and long short-term memory (LSTM) to improve the accuracy of prediction for DO in water. The hybrid CNN-LSTM model predicted DO concentration in water using soft sensor data. The primary input parameters to the model were temperature, pH, specific conductivity, salinity, density, chlorophyll, and blue-green algae. The model used 38,681 water quality data for training and testing the hybrid deep learning network. The training procedure for the model was successful. The training and test losses were both nearly zero and within a similar range. With a coefficient of determination (R2) of 0.94 and a mean squared error (MSE) of 0.12, the hybrid model indicated higher performance compared to the classical models. The normal distribution of residual errors confirmed the reliability of the DO predictions by the hybrid CNN-LSTM model. Feature importance analysis indicated pH as the most significant predictor and temperature as the second important predictor. The feature importance scores based on extreme gradient boosting (XGBoost) for the pH and temperature were 0.76 and 0.12, respectively. This study indicated that the hybrid model can outperform the classical machine learning models in the real-time prediction of DO concentration.

Enhancing Estimation Accuracy of Prostate Cancer VMAT Planning: A Knowledge-based Approach Using Multiple Collimator Angles.

This study aimed to determine whether a knowledge-based planning model incorporating treatment plans with multiple collimator angles (Multi-coll. model) provides superior estimation accuracy and plan quality for a range of collimator angles compared with a typical institutional model (Inst. model). Five institutions using volumetric modulated arc therapy for prostate cancer participated in the study. The Inst. model comprised plans using a fixed collimator angle, whereas the Multi-coll. model registered plans using single arcs and collimator angles of 10°, 30°, 50°, and 70°. The coefficient of determination and mean squared error were calculated, and the estimation accuracy and dosimetric results for three validation cases at each institution were compared for each model. Dosimetric parameters included volumes receiving at least 20%, 50%, and 90% of the prescription dose (V20, V50, and V90, respectively) at each collimator angle. The Multi-coll. model yielded a higher coefficient of determination and lower mean squared error than the Inst. model, although overfitting of data was a concern at two institutions. The mean absolute errors between estimated and calculated values were (Inst. model vs. Multi-coll. model): 4.41% vs. 2.60% (V20), 4.03% vs. 2.27% (V50), and 2.07% vs. 1.09% (V90) for the rectum; 3.48% vs. 3.42% (V20), 2.24% vs. 3.65% (V50), and 1.11% vs. 0.73% (V90) for the bladder. The Multi-coll. model exhibited a lower rectal and vesical V20 than the Inst. The Multi-coll. model had a greater estimation accuracy and slightly higher plan quality than the Inst.

State of charge estimation method for lithium-ion batteries based on adaptive central difference particle filter with weight reconstruction

The particle filter (PF) has been widely used for state of charge (SOC) estimation. However, the particle degradation phenomenon will affect the estimation accuracy. In response to this issue, a state of charge estimation method for lithium-ion batteries based on adaptive central difference particle filter with weight reconstruction (ACDPF-WR) is designed in this paper. This method combines the preferred importance density function and the optimized resampling strategy. Firstly, the central difference Kalman filter (CDKF) is used to update the sampled particles to reduce the influence of particle degradation. Secondly, the Gaussian process regression (GPR) model of particles and weights is constructed by combining the offline learning experimental battery data set, and the GPR is used to generate the weight distribution of the particle filter. Then, an adaptive step size mechanism is introduced, which determines the optimal step size by calculating the mean square error of the weight distribution under different steps. Finally, the weight distribution generated by the central difference particle filter (CDPF) based on the optimal step size is combined with the typical resampling algorithm to select high-quality particles to achieve the optimal estimation. The adaptability and robustness of the algorithm are verified under Beijing Dynamic Stress Test (BJDST), US06 Highway Driving Schedule (US06), and Dynamic Stress Test (DST) conditions, and evaluated by mean absolute error (MAE) and root mean square error (RMSE) indicators. The average comparison results of the three working conditions show that the MAE of ACDPF-WR is 54.1 % higher than that of EPF and 24.3 % higher than that of CDPF, and the RMSE of ACDPF-WR is 64.6 % higher than that of EPF and 24.5 % higher than that of CDPF. The proposed algorithm achieves better performance and provides new insights and methods for the optimization and improvement of the battery management system.

Automatic detection and counting of wheat spike based on DMseg-Count

The automatic detection and counting of wheat spike images are of great significance for yield prediction and variety evaluation. Therefore, accurate and timely estimation of spike numbers is crucial for wheat production. However, in actual production, due to the susceptibility of wheat spike images to factors such as lighting conditions, shooting angles, occlusion, and overlap, the contour and features of wheat spike is unclear, which affects the accuracy of automatic detection and counting of wheat spike. In order to solve the above problems and further improve the accuracy of wheat spike counting, an improved wheat spike counting model DMseg-Count was proposed by enhancing local contextual supervision information based on existing target object counting model DM-Count. Firstly, wheat spike local segmentation branch was introduced to improve the network architecture of DM-Count, so as to extract the local contextual supervision information of wheat spike. Secondly, an element-by-element point multiplication mechanism was designed to fuse global and local contextual supervision information of wheat spike. Finally, the total loss function was constructed to optimize the model. The test results showed that the mean absolute error (MAE) and root mean square error (RMSE) of the proposed DMseg-Count model were 5.79 and 7.54, respectively, which were 9.76 and 10.91 higher than the standard distribution matching for crowd counting (DM-Count) model. Compared with other deep learning models, the proposed DMseg-Count model can detect wheat spike image in challenging situations, and has better computer vision processing capabilities and performance evaluation detection effect. In summary, the proposed DMseg-Count model can effectively detect wheat spike and has good counting performance, which provides a new method for automatic counting of wheat spike and yield prediction in complex field environments.

Bayesian approach in the nonparametric conditional density estimation

In the literature, it is well known that in kernel density estimation, plug-in and cross-validation techniques, for the selection of the smoothing parameter, tend to provide under-or over-smoothed estimators when the sample size is small or medium, or the function to be estimated is complex. To overcome this latest problem, recently, the Bayesian approach has been proposed as an alternative to these classical methods. In this paper, we restricted attention to extending the idea of the Bayes rule to estimate the smoothing parameter of the conditional density kernel estimation. We are interested in the impact of the use of the global Bayesian approach for the smoothing parameter selection on the performance (the average of the Integrated Squared Error (ISE) and the calculation time required for their implementation) of the kernel conditional density estimator compared to the classical method trough a Monté Carlo simulation. Moreover, based on simulated samples of different sizes from two different conditional models, we adopted the Bayesian approach compared with the classical smoothing parameter selection procedure (the cross-validation method), using the Gaussian kernel to construct the estimators in question.

Robust lithium-ion battery state of health estimation based on recursive feature elimination-deep Bidirectional long short-term memory model using partial charging data

Accurate perception of the state of health (SOH) of lithium-ion batteries is crucial for their safety and reliable operation. To meet this demand, a recursive feature elimination-deep bidirectional long short-term memory (RFE-DBiLSTM) model suitable for partial charging data is proposed to effectively estimate the SOH of lithium-ion batteries. In this study, the recursive feature elimination (RFE) method is used to screen multiple charging features for obtaining the key features that best represent the SOH under two scenarios with different charging segment lengths. Due to the robust noise-filtering capability and strong ability to capture complex and multi-level temporal dependencies, the deep bidirectional long short-term memory (DBiLSTM) model is used for time series data training, verification, and testing during aging. Experimental results show that compared with benchmark time series models such as long short-term memory (LSTM) and gated recurrent unit (GRU), the proposed method significantly reduces the estimated mean absolute error (MAE) and root mean square error (RMSE) on diverse batteries in the above scenarios. In the scenario for missing partial constant current (CC) charging data, the MAE and RMSE of B0005 cell are 0.0062 and 0.0094, the MAE and RMSE of B0006 cell are 0.0294 and 0.0314, the MAE and RMSE of CS2_36 cell are 0.0510 and 0.0601, the MAE and RMSE of B0029 cell are 0.0057 and 0.0072, and the MAE and RMSE of B0030 cell are 0.0088 and 0.0102. This research innovatively combines the RFE method with the DBiLSTM model to improve the accuracy and robustness of SOH estimation.

Error compensation for industrial robots combining kinematics characterized by conformal geometric algebra and fused observability-index-based measurement strategy

Abstract The geometric errors of industrial robots are key factors affecting positioning accuracy. A new compensation method for industrial robots is proposed based on the kinematics characterizing with Conformal Geometric Algebra (CGA) and measurement strategy with Double Ball Bar (DBB) path point optimization. Firstly, a kinematic error model for the industrial robot is established using CGA, the starting point of the CGA model is modified to simplify the modeling process and reduce computational complexity. Secondly, fused observability index is proposed and the relationship between the number of sampling points on the DBB path and the effectiveness of error parameter is obtained. Thirdly, the adaptive golden spiral optimization algorithm for error parameter identification is proposed, achieving efficient and stable identification of error parameters. Finally, a case study is carried out on a six-degree-of-freedom industrial robot. The validity of measurement strategy and error parameter identification algorithm are confirmed by comparing the residuals and uncertainties of predicted point positions in space with different methods. The spatial compensation results show that, after compensation, the average error and root mean square error of measurement paths are reduced by 36.76% and 33.96%, respectively.

AI-Driven Circular Economy of Enhancing Sustainability and Efficiency in Industrial Operations

This study investigates integrating circular economy principles—such as closed-loop systems and economic decoupling—into industrial sectors, including refining, clean energy, and electric vehicles. The primary objective is to quantify the impact of circular practices on resource efficiency and environmental sustainability. A mixed-methods approach combines qualitative case studies with quantitative modelling using the Brazilian Land-Use Model for Energy Scenarios (BLUES) and Autoregressive Integrated Moving Average (ARIMA). These models project long-term trends in emissions reduction and resource optimization. Significant findings include a 20–25% reduction in waste production and an improvement in recycling efficiency from 50% to 83% over a decade. Predictive models demonstrated high accuracy, with less than a 5% deviation from actual performance metrics, supported by error metrics such as Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE). Statistical validations confirm the reliability of these forecasts. The study highlights the potential for circular economy practices to reduce reliance on virgin materials and lower carbon emissions while emphasizing the critical role of policy support and technological innovation. This integrated approach offers actionable insights for industries seeking sustainable growth, providing a robust framework for future resource efficiency and environmental management applications.

State of Charge Prediction of Mine-Used LiFePO4 Battery Based on PSO-Catboost

The accurate prediction of battery state of charge (SOC) is one of the critical technologies for the safe operation of a power battery. Aiming at the problem of mine power battery SOC prediction, based on the comparative experiments and analysis of particle swarm optimization (PSO) and Categorical Boosting (Catboost) characteristics, the PSO-Catboost model is proposed to predict the SOC of a power lithium iron phosphate battery. Firstly, the classification model based on Catboost is constructed, and then the particle swarm algorithm is used to optimize the Catboost hyperparameters to build the optimal model. The experiment and comparison show that the optimized model’s prediction accuracy and average precision are superior to other comparative models. Compared with the Catboost model, the Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) values of the PSO-Catboost model decreased by 12.4% and 25.4% during charging and decreased by 5.5% and 12.2% during discharging. Finally, the Random Forest (RF) and Extreme Gradient Boosting (XGBoost) models, both ensemble learning models, are selected and compared with PSO-Catboost after being optimized via PSO. The experimental results show that the proposed model has a better performance. In this paper, experiments show that the optimization model can select parameters more intelligently, reduce the error caused by artificial experience to adjust parameters, and have a better theoretical value and practical significance.

Impact of WRF Model Parameterization Settings on the Quality of Short-Term Weather Forecasts over Poland

This research examines the impact of various parameterization settings within the Weather Research and Forecasting (WRF) model on the accuracy of short-term weather forecasts for Poland. The study focuses on the sensitivity of key meteorological variables—namely, air temperature, wind speed, relative humidity, and atmospheric pressure—to different combinations of physical parameterization schemes. Utilizing data from the Global Forecast System (GFS) spanning 2019 to 2022, a series of model simulations were conducted with support from the Poznań Supercomputing and Networking Center (PCSS). To assess the model’s performance across different weather stations, statistical metrics such as the mean absolute error (MAE) and root mean square error (RMSE) were employed. The findings indicate that the configuration labeled “p2” produced the most accurate forecasts for temperature, wind speed, and atmospheric pressure, achieving MAE values of 1.5 °C, 1.6 m/s, and 2 hPa, respectively. However, forecast inaccuracies were notably higher in mountainous regions, particularly regarding wind speed. These results underscore the importance of selecting appropriate parameterization settings tailored to regional characteristics, as different configurations can significantly impact the forecast accuracy, especially in complex terrains. This study contributes to the understanding of short-term weather forecasting models for Central Europe, offering potential pathways for improving localized forecast accuracy.

Estimating the Lifetime Parameters of the Odd-Generalized-Exponential–Inverse-Weibull Distribution Using Progressive First-Failure Censoring: A Methodology with an Application

This paper investigates statistical methods for estimating unknown lifetime parameters using a progressive first-failure censoring dataset. The failure mode’s lifetime distribution is modeled by the odd-generalized-exponential–inverse-Weibull distribution. Maximum-likelihood estimators for the model parameters, including the survival, hazard, and inverse hazard rate functions, are obtained, though they lack closed-form expressions. The Newton–Raphson method is used to compute these estimations. Confidence intervals for the parameters are approximated via the normal distribution of the maximum-likelihood estimation. The Fisher information matrix is derived using the missing information principle, and the delta method is applied to approximate the confidence intervals for the survival, hazard rate, and inverse hazard rate functions. Bayes estimators were calculated with the squared error, linear exponential, and general entropy loss functions, utilizing independent gamma distributions for informative priors. Markov-chain Monte Carlo sampling provides the highest-posterior-density credible intervals and Bayesian point estimates for the parameters and reliability characteristics. This study evaluates these methods through Monte Carlo simulations, comparing Bayes and maximum-likelihood estimates based on mean squared errors for point estimates, average interval widths, and coverage probabilities for interval estimators. A real dataset is also analyzed to illustrate the proposed methods.

Transient Stability-Based Fast Power System Contingency Screening and Ranking

Today’s power systems are operated closer to their stability limits due to the continuously growing load demands, interface to open markets, and integration of more renewable energies. In order to provide operators with clear insight on the current system situation, near real-time power systems dynamic security assessment tools are required. One of the core elements of near real-time dynamic security assessment tools is contingency screening and ranking. Most of the commercially available tools screen and rank contingencies by using the traditional numerical integration or Transient Energy Functions (TEFs) or hybrid methods. The traditional numerical integration method is accurate but computationally intensive and has a slow assessment speed which makes it difficult to identify any insecure contingency before it happens. Despite the TEF method of transient stability analysis being relatively fast, it develops less accurate results due to models simplification and assumptions. This paper introduces transient stability based on fast and robust contingency screening and ranking using an Adaptive step-size Differential Transformation (AsDTM) method. Based on the most current snapshot from Supervisory Control and Data Accusation (SCADA) data, the proposed method triggers AsDTM-based transient stability simulation for each credible contingency and evaluates Transient Stability Indices (TSI) as the normalized weighted sum of squares of errors derived from state variables and complex bus voltages at every simulation time step. Finally, contingencies are ranked based on these TSI and the worst contingency is identified for the next detail assessment. The method is tested on IEEE 9 bus and 39 bus test systems. Test results reveal that the proposed method is faster, robust, and can be used in near real-time dynamic security assessment sessions.

Significant wave height prediction in monsoon regions based on the VMD-CNN-BiLSTM model

A novel significant wave height prediction method for monsoon regions is proposed, utilizing the VMD-CNN-BiLSTM model to enhance prediction accuracy under complex meteorological conditions. Traditional numerical models exhibit limitations in managing extreme marine conditions and fail to fully integrate wind field information. Meanwhile, existing machine learning models demonstrate insufficient generalization and robustness for long-term predictions. To address these shortcomings, the predictive approach combines Variational Mode Decomposition (VMD) with a hybrid deep learning model (CNN-BiLSTM). VMD is employed to decompose the original wave height sequence and extract key features, while CNN captures the spatial features of wind field and wave height data. BiLSTM, in turn, models the temporal dependencies. Experimental results reveal that the VMD-CNN-BiLSTM model provides substantial advantages in prediction performance across all seasons, including the entire year. Compared to traditional models, the proposed method demonstrates significantly reduced Mean Absolute Error (MAE) and Root Mean Square Error (RMSE), alongside an improved coefficient of determination (R²). These findings confirm the effectiveness and reliability of the method under complex meteorological conditions such as monsoons and typhoons.

Retrieval of cloud fraction using machine learning algorithms based on FY-4A AGRI observations

Abstract. Cloud fraction as a vital component of meteorological satellite products plays an essential role in environmental monitoring, disaster detection, climate analysis and other research areas. Random forest (RF) and multilayer perceptron (MLP) algorithms were used in this paper to retrieve the cloud fraction of AGRI (Advanced Geosynchronous Radiation Imager) on board the Fengyun-4A (FY-4A) satellite based on its full-disk level-1 radiance observation. Corrections have been made subsequently to the retrieved cloud fraction in areas where solar glint occurs using a correction curve fitted with sunglint angle as weight. The algorithm includes two steps: the cloud detection is conducted firstly for each AGRI field of view to identify whether it is clear sky, partly cloudy or overcast within the observation field. Then, the cloud fraction is retrieved for the scene identified as partly cloudy. The 2B-CLDCLASS-lidar cloud fraction product from the CloudSat and CALIPSO active remote sensing satellite is employed as the truth to assess the accuracy of the retrieval algorithm. Comparison with the operational AGRI level-2 cloud fraction product is also conducted at the same time. The results indicate that both the RF and MLP cloud detection models achieved high accuracy, surpassing that of operational products. However, both algorithms demonstrated weaker discrimination capabilities for partly cloudy conditions compared to clear-sky and overcast situations. Specifically, they tended to misclassify fields of view with low cloud fractions (e.g., cloud fraction = 0.16) as clear sky and those with higher cloud fractions (e.g., cloud fraction = 0.83) as overcast. Between the two models, RF exhibited higher overall accuracy. Both RF and MLP models performed well in cloud fraction retrieval, showing lower mean error (ME), mean absolute error (MAE) and root mean square error (RMSE) compared to operational products. The ME for both RF and MLP cloud fraction retrieval models was close to zero, while RF had slightly lower MAE and RMSE than MLP. During daytime, the high reflectance in sunglint areas led to larger retrieval errors for both RF and MLP algorithms. However, after correction, the retrieval accuracy in these regions improved significantly. At night, the absence of visible light observations from the AGRI instrument resulted in lower classification accuracy compared to daytime, leading to higher cloud fraction retrieval errors during nighttime.

Microwave-vacuum drying behavior of rosehips: Experimental investigation and mathematical modeling

In this study, the microwave-vacuum drying of rosehips (Rosa canina) was investigated and the effect of different drying conditions during the microwave vacuum drying process on the drying kinetics of rosehips was determined. For this purpose, microwave vacuum drying experiments were carried out until a moisture content less than 10 % by weight was attained at three power (50, 100, and 150 W) and three pressure (40, 75, and 110 mBar (abs)) levels. The drying times ranged from 75 to 195 min depending on the power level and vacuum degree applied. Seven different mathematical models, which are most commonly used in the literature, were fitted to the experimental data, and among these models, the most suitable model was selected by considering the statistical criteria of the highest regression coefficient (R2), the lowest chi-square (χ2), total squared error (SSE) and root mean square error (RMSE). The drying rate (DR) increased with increasing microwave power levels under all pressure levels. The Midilli model was proposed as the best-fit model for predicting the drying behavior of rosehip plants under all drying process conditions. The effective moisture diffusion (Deff) of rosehips under different drying conditions during the MVD process ranged from 5.81⨮10−8-17.32⨮10−8 m2/s and increased with the increasing microwave power at constant pressure.

Short-term Passenger Flow Prediction of Subway based on Neural Network LSTM Algorithm

The present study investigates the application of the Long Short-Term Memory (LSTM) neural network algorithm for short-term passenger flow forecasting in subway systems. The dataset, sourced from Hangzhou subway stations, records passenger flow at 10-minute intervals over 30 days, distinguishing between weekdays and weekends. The LSTM model was trained using 80% of the available data, while the remaining 20% was reserved for testing purposes. Results showed the model achieved a mean squared error (MSE) of 19.33 on training data, indicating a good fit. However, the test data MSE was significantly higher at 460.78, suggesting overfitting. This implies the model captures training data patterns well but struggles with generalizing to new data. The findings underline the potential of LSTM networks in handling long-term dependencies and gradient vanishing issues inherent in traditional Recurrent Neural Networks (RNNs). Despite these advancements, the study highlights the need for model refinement to address overfitting and improve generalizability. Future research should focus on data augmentation, model enhancement, and regularization techniques to develop more robust prediction models. Accurate short-term predictions of passenger flow play a crucial role in optimizing train scheduling, resource allocation, and overall service quality within urban rail transit systems.