A low-cost, IoT-enabled GNSS-based ionospheric monitoring system with real-time data access

The ionosphere is the active, complex, and ionized part of the atmosphere, and it ranges from about 50 km to more than 1000 km from the earth’s surface. Because the ionosphere is dispersive, ionosphere remote sensing using dual-frequency observations from the Global Navigation Satellite System (GNSS) signals is recognized as a cost-effective approach to monitor the ionosphere. However, the GNSS-derived ionospheric observables are biased due to Differential Code Biases (DCBs) (Xiang et al. 2017). The way to separate ionosphere from the satellite- and receiver-related DCBs is through ionospheric modeling. Precise ionospheric modeling and DCB determination are required for mitigating ionospheric corrections for single-frequency users , for reconstructing ionospheric tomography with precise slant Total Electron Content (TEC), and for shortening the convergence time for PPP when adding ionospheric constraints (Shi et al. 2012). In the ionospheric modeling, in order to separate the ionosphere from the DCBs, mapping function (MF) plays a critical role in that it projects the line-of-sight slant TEC (STEC) to vertical TEC (VTEC). The MF is defined as the ratio between STEC and VTEC based on the assumption of a single-layer model (SLM). The SLM assumes that the ionosphere is condensed into an infinitesimal thickness layer. From the definition, we can see that the MF is larger than 1 and usually smaller than 3. Based on the SLM assumption, the MF depends on the observing elevation and height of the SLM. The MF decreases with the increase of elevation and height. The height is designed to be established at the average altitude of electron density profile peaks (Lanyi and Roth 1988). If this chosen height is lower than the optimal height, the MF tends to be overestimated. Besides, the height also affects the pierce point position that is the interaction between the line of sights and the fixed-height layer. The height typically ranges from 300 km to 500 km (Hernández-Pajares et al. 2005), but for the convenience of the global ionospheric modeling, the height is usually fixed, regardless of daytime or nighttime, at 450 km (Hernández-Pajares et al. 2011; Hernandez-Pajares et al. 2009). However, because the ionosphere is inhomogeneous, the effective heights for SLM vary in location and time, depending on the condition of space weather. Many researchers showed that the MF at a given fixed height has significant errors at the low elevation and the large ionosphere gradient with a fixed height (Birch et al. 2002; Conker and El-Arini 2002; Nava et al. 2007; Palamartchouk 2010; Wang et al. 2016; Zus et al. 2016). When horizontal ionospheric gradients exist, azimuth effects along the line-of-sight are consequential (Conker and El-Arini 2002). Besides, the contributions of topside plasma to the mapping function are also substantial, especially at night when the electron density decreases compared to the topside plasma (Birch et al. 2002). Accordingly, there is a need to investigate the impacts of existing mapping functions on ionospheric modeling and derive improved MF for more accurate ionospheric modeling. There are six general kinds of MFs: (1) the single-layer MF. The simplest and most commonly used, it includes the broadcast model (Klobuchar 1987) and the modified SLM MF (Schaer 1999); (2) the MF that assumes the ionosphere is a spherical shell or flat-plane with thickness. The spherical shell model without thickness is almost the same as the flat-plane model, whereas regardless of the thickness, the MF can be overestimated up to 15% at the pierce point zenith angle of 700 and the height of 200 km (Smith et al. 2008); (3) the two-layer or multi-layer MF. It takes into account the vertical structure of the ionosphere electron density instead of assuming the ionosphere being fixed at an layer with infinitesimal thickness (Hernández-Pajares et al. 1999; Smith et al. 2008). This model is advantageous when there are implicit horizontal gradients and low-elevation observations; (4) the GNSS-data derived MF. Jin et al. (2010) and Birch et al. (2002) proposed to reproduce the height using GNSS data. They also pointed out the topside plasma has a significant effect on the MF, and the MF varies with location and seasons; (5) the MF based on empirical models like IRI and Chapman profile. These models are applied to calculate the average height of electron density or the integral height along the trace (Conker and El-Arini 2002; Zhong et al. 2016); (6) the Potsdam MF. It is proposed to consider the location, elevation, azimuth, and even bending effects (Zus et al. 2016). Some of these mapping functions are compared with each other in Schaer (1999), and the author showed that there is a relative error of 10% with reference to the height of 350 km. Based on these six current methods, the mapping errors to the ionospheric modeling will be investigated. Furthermore, because a unique height for the daytime and nighttime to cancel the mapping errors does not exist, a mapping function with variable heights from the latest IRI 2016 will be developed, aiming to reduce the mapping errors for the ionospheric modeling so that we can achieve more precise DCBs and STEC. The carrier-phase derived ionospheric observables using PPP are used to reduce the leveling errors from the smoothed code measurements. We anticipate smaller ionospheric modeling errors and more stable DCBs occur with the developed mapping function.

Global ionosphere models may not be accurate enough in regions where there are insufficient observations. Therefore, there is strong interest in regional ionosphere modeling using observations from dual-frequency Global Navigation Satellite System (GNSS) receivers in such regions. This paper presents a data assimilation method for hourly total electron content (TEC) modeling along with estimation of the receivers’ differential code biases (DCBs). For this purpose, first, initial values for slant TEC (STEC) and the receivers’ DCBs are estimated utilizing global ionosphere maps (GIMs). Second, STEC is expanded to a linear combination of spherical radial basis functions with time-dependent coefficients to reduce the number of unknowns, and a constrained least-squares adjustment is applied to resolve the corrected STEC and DCB values in the study area using observations from dual-frequency GNSS receivers. Third, the resolved STEC at ionospheric pierce points (IPPs) are converted to vertical TEC (VTEC) using a mapping function and assimilated into the background provided by spatial and temporal interpolation of GIM. Retrieved VTECs are weighted according to the satellites’ elevation angles, and an objective function is defined to determine the background’s covariance matrix. In this study, the Iranian permanent Global Positioning System (GPS) network is used to construct the regional hourly VTEC model and calculate hourly DCB variations. The proposed method is validated in two ways. First, some of the VTEC values are excluded from the dataset for use as test data and compared with modeling results. The root mean square (RMS) of the constructed regional assimilative model on test data is significantly better than GIM. Second, the RINEX file of Tehran station, whose observations did not play a role in modeling, is corrected using the regional model. The single-frequency positioning results with corrected observations show much better accuracy than GIMs.

As part of the ongoing system modernization, the Russian navigation satellite system GLONASS has specified a dedicated electron density model supporting ionospheric path delay corrections for single-frequency navigation users. Solar-geophysical parameters for use with this model are made available through the code division multiple access (CDMA) signals, transmitted by selected GLONASS-K1, -K2 and -M+ satellites on the L3 and L1 frequencies. As a notable feature, the GLONASS ionosphere model can be used to predict the slant total electron content (STEC) through numerical integration of the 3-dimensional electron density along the signal path or a single-layer approximation of the 2-dimensional vertical total electron content (VTEC). Based on reference TEC values provided by global ionosphere maps, the performance of the GLONASS ionosphere model is assessed over an 11-year period using measured solar flux and geomagnetic activity values and compared with correction models of the GPS and Galileo constellations. Furthermore, the quality of solar-geophysical parameters made available in the CDMA navigation message over 1 year after launch of the first GLONASS-K2 satellite is evaluated. Compared to global ionosphere maps of the International GNSS Service, the GLONASS model exhibits VTEC biases in the range of roughly $${\pm 1}\,\textrm{TECU}$$ ± 1 TECU . Mean absolute errors (MAE) range from about $${5}\,\textrm{TECU}$$ 5 TECU in quiet years to 16 TECU at high solar activity. The corresponding mean absolute percentage errors (MAPE) range from roughly 50% (high activity) to 60% (low activity). Only minor performance differences were observed when comparing predictions based on broadcast values of solar flux and geomagnetic activity with observed values from space weather centers. On the other hand, a clear reduction of both the mean absolute (3–14 TECU) and mean absolute percentage errors (41–45%) is achieved when adjusting the adaptation coefficient of the GLONASS model based on the daily mean ratio of predicted and observed VTEC values. Irrespective of this, major VTEC modeling problems at very high solar activity could be identified. Overall, the GLONASS model outperforms the Klobuchar model but does not reach the prediction performance of the Galileo NeQuick-G and NTCM-G models, which exhibit errors of about 2–8 TECU (MAE) and 26–37% (MAPE).

Global navigation satellite systems (GNSS) allow estimating total electron content (TEC). However, it is still a problem to calculate absolute ionosphere parameters from GNSS data: negative TEC values could appear, and most of existing algorithms does not enable to estimate TEC spatial gradients and TEC time derivatives. We developed an algorithm to recover the absolute non-negative vertical and slant TEC, its derivatives and its gradients, as well as the GNSS equipment differential code biases (DCBs) by using the Taylor series expansion and bounded-variable least-squares. We termed this algorithm TuRBOTEC. Bounded-variable least-squares fitting ensures non-negative values of both slant TEC and vertical TEC. The second order Taylor series expansion could provide a relevant TEC spatial gradients and TEC time derivatives. The technique validation was performed by using independent experimental data over 2014 and the IRI-2012 and IRI-plas models. As a TEC source we used Madrigal maps, CODE (the Center for Orbit Determination in Europe) global ionosphere maps (GIM), the IONOLAB software, and the SEEMALA-TEC software developed by Dr. Seemala. For the Asian mid-latitudes TuRBOTEC results agree with the GIM and IONOLAB data (root-mean-square was < 3 TECU), but they disagree with the SEEMALA-TEC and Madrigal data (root-mean-square was >10 TECU). About 9% of vertical TECs from the TuRBOTEC estimates exceed (by more than 1 TECU) those from the same algorithm but without constraints. The analysis of TEC spatial gradients showed that as far as 10–15° on latitude, TEC estimation error exceeds 10 TECU. Longitudinal gradients produce smaller error for the same distance. Experimental GLObal Navigation Satellite System (GLONASS) DCB from TuRBOTEC and CODE peaked 15 TECU difference, while GPS DCB agrees. Slant TEC series indicate that the TuRBOTEC data for GLONASS are physically more plausible.

Total electron content (TEC) is a important parameter in the domains of space weather studies and Global Navigation Satellite System (GNSS)-based navigation and communication applications. Conventional linear forecasting models face difficulties in effectively representing the complex nonlinear behaviors of the ionospheric dynamics. On the other hand, nonlinear approaches derived from advanced learning methods offer higher accuracy, but they necessitate substantial computational resources, building them impractical for real-time use in resource-constrained IoT environments. The emergence of Internet of Things (IoT) technology has facilitated the accessibility of affordable GNSS data associated through cloud platforms, allowing for ongoing and instantaneous collection of TEC data. In this paper, an efficient Successive Variational Mode Decomposition (SVMD) and Random Vector Functional Link (RVFL) framework is implemented to predict TEC via cloud platforms through Think Speak channels. The TEC observations from the year 2018 at Bengaluru (Geographic: 13.02° N, 77.57° E) is consider for analysis. The SVMD adaptively decomposes the TEC signal without requiring predefined mode selection, while RVFL enables fast training using random weights, direct connections, and universal approximation capabilities. The proposed model was evaluated using GNSS data from Bengaluru (13.02° N, 77.57° E). The results demonstrate that the SVMD–RVFL has an accuracy of 0.55 TECU for Root Mean Square Error (RMSE), 0.61 TECU for Mean Absolute Error (MAE), 7.64 % for Mean Absolute Percentage Error (MAPE), a correlation coefficient of 99.32 % and a training time of 3.82 s. The proposed approach demonstrates high precision and a low computational load, making it suitable for real-time ionospheric monitoring systems and IoT technologies.

The Global Navigation Satellite System (GNSS) allows for the cost-effective estimation of the ionospheric total electron content (TEC). However, research on error characteristics of the derived TEC is scarce, which provides insights into the quality of the GNSS ionospheric observation. We investigate characteristics of errors in the derived TEC with data from ~260 GNSS dual-frequency receivers of the Crustal Movement Observation Network of China (CMONOC). The slant TEC is calculated from carrier phase measurements and the vertical TEC over China is fitted with a spatial resolution of 1° by 1° in latitude and longitude in four seasons of 2014. It is found that the errors of both the slant TEC and the derived TEC follow Laplace distribution rather than Gaussian distribution in all seasons. The errors of the slant TEC have sharper peaks than those of the derived TEC. The Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) of the slant TEC are typically 0.04 TECU and 0.2 TECU, while the MAE and RMSE of the fitting residuals for the derived TEC are typically 1 TECU and under 2 TECU, respectively. Both MAEs and RMSEs of the derived TEC have the largest value in spring and the smallest value in summer, while the seasonal dependence is only observed in RMSE of the slant TEC.

Prediction of ozone hourly concentrations by support vector machine and kernel extreme learning machine using wavelet transformation and partial least squares methods

In India, the first confirmed case of novel corona virus (COVID-19) was discovered on January 30, 2020. The number of confirmed cases is increasing day by day, and it crossed 21,53,010 on August 9, 2020. In this paper, a hybrid forecasting model has been proposed to determine the number of confirmed cases for upcoming 10 days based on the earlier confirmed cases found in India. The proposed model is based on adaptive neuro-fuzzy inference system (ANFIS) and mutation-based Bees Algorithm (mBA). The meta-heuristic Bees Algorithm (BA) has been modified applying 4 types of mutation, and mutation-based Bees Algorithm (mBA) is applied to enhance the performance of ANFIS by optimizing its parameters. Proposed mBA-ANFIS model has been assessed using COVID-19 outbreak dataset for India and USA, and the number of confirmed cases in the next 10 days in India has been forecasted. Proposed mBA-ANFIS model has been compared to standard ANFIS model as well as other hybrid models such as GA-ANFIS, DE-ANFIS, HS-ANFIS, TLBO-ANFIS, FF-ANFIS, PSO-ANFIS and BA-ANFIS. All these models have been implemented using Matlab 2015 with 10 iterations each. Experimental results show that the proposed model has achieved better performance in terms of Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), Mean absolute error (MAE) and Normalized Root Mean Square Error (NRMSE). It has obtained RMSE of 1280.24, MAE of 685.68, MAPE of 6.24 and NRMSE of 0.000673 for India Data. Similarly, for USA the values are 4468.72, 3082.07, 6.1, and 0.000952 for RMSE, MAE, MAPE, and NRMSE, respectively.

The growing interest in using plant-based materials as substitutes for synthetic ones in the polymer composites industry is primarily driven by the negative health and environmental impacts of synthetic materials. However, the effectiveness and application of these composites are often hindered by poor material design. This research focused on extracting and processing stem fibers from the Newbouldia laevis plant into particles, followed by the creation of composites with varying particulate weight contents (wt). Mechanical testing showed that the composites had a maximum tensile strength of 31.3496 N/mm² for those containing 10 wt. %, a maximum compression strength of 59.8716 N/mm² for those with 40 wt. %, and a maximum flexural strength of 36.5808 N/mm² for those with 10 wt. % particle content. The experimental results were analyzed using artificial neural networks (ANN) to create a precise predictive model aimed at improving material design. The reliability of the ANN models was assessed using various performance metrics such as mean square error (MSE), mean absolute error (MAE), mean absolute percentage error (MAPE), normalized root mean square error (NRMSE), and the coefficient of determination (R²). The results showed low error rates for MSE, MAE, MAPE, and NRMSE, and a high R² value exceeding 0.9, indicating that the model predictions are highly accurate. These results demonstrate that ANN can be an effective mathematical tool for modeling and predicting the mechanical properties of reinforced polymer composites, ultimately saving both time and resources in material design research.

The 3-D computerized ionospheric tomography (CIT) based on Global Navigation Satellite System (GNSS) data is a classic ill-posed inverse problem. This study proposes an algorithm based on the semi-parametric model, which leverages the nonparametric component in the semi-parametric model to address systematic errors in CIT, thus improving the accuracy and effectiveness of reconstructed ionospheric electron density (IED). The feasibility and effectiveness of the proposed algorithm in processing systematic errors and reconstructing IED values are validated through simulation and real data experiments. In the simulation experiment, the proposed algorithm can separate systematic errors effectively. Compared with the Tikhonov regularization algorithm, the proposed algorithm offers improvements of 50.2% and 50.3% in root mean square error (RMSE) and the mean absolute error (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\Delta E$ </tex-math></inline-formula>) for the reconstructed IEDs. The reconstructed 3-D ionospheric structure based on real data is consistent with the real spatiotemporal variation characteristics of the ionosphere. The average RMSE and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\Delta E$ </tex-math></inline-formula> of the reconstructed slant total electron content (STEC) using the proposed algorithm are 30.2% and 32.1% higher than those of the Tikhonov regularization algorithm, respectively. The proposed algorithm demonstrates superior reconstruction performance in several aspects.

To evaluate continuous and non-invasive sweat-inferred blood lactate monitoring for determining the exercise intensity at the second metabolic threshold (MT2) and maximal lactate steady state (MLSS), compared to capillary blood lactate and gas exchange analysis. 17 physically active individuals (11 males and 6 females) and 19 endurance athletes (14 males and 5 females) completed a maximal graded exercise test (GXT) to quantify oxygen uptake (VO2), percentage of peak oxygen uptake (%VO2peak) and power output (watts) at MT2, using sweat-inferred blood lactate and capillary blood lactate for the second lactate threshold (LT2), and gas exchange analysis for the second ventilatory threshold (VT2). Participants also completed a MLSS test to quantify the power output (watts) at MLSS. Between-methods agreement was assessed using linear mixed models with mean differences (MD), mean absolute error (MAE), mean absolute percentage error (MAPE), Lin's concordance correlation coefficient (CCC) and Bland-Altman analysis. Compared with VT2, LT2 derived from sweat-inferred blood lactate showed no significant differences when using the Log-Exp-Mod-Dmax method, yielding the strongest agreement across VO2 (MD = 0.3 mL·min -1·kg -1, p = 0.999; MAE = 2.1 mL·min -1·kg -1, MAPE = 5.2%), %VO2peak (MD = 0.4%, p = 0.999; MAE = 4.1%, MAPE = 5.2%), and power output (MD = 1.9W, p = 0.999; MAE = 9.1W, MAPE = 5.0%). Direct comparison of capillary blood and sweat-inferred blood lactate using the Log-Exp-Mod-Dmax method showed high agreement for VO2 (MD = 0.2 mL·min -1·kg -1, p = 0.999; MAE = 2.4 mL·min -1·kg -1, MAPE = 6.4%), %VO2peak (MD = -0.1%, p = 0.999; MAE = 4.9%, MAPE = 6.4%), and power output (MD = 0.0W, p = 0.999; MAE = 10.6W, MAPE = 6.1%). Capillary blood and sweat-inferred blood lactate also showed excellent agreement for MLSS-associated power output (MD = -2.0 W, p = 0.089; MAE = 3.1 W, MAPE = 2.5%). Continuous sweat-inferred blood lactate can be used to determine MT2 and MLSS intensities, providing an alternative to capillary blood lactate.

Uniaxial compressive strength (UCS) is one of the most important geomechanical parameters for evaluating rock strength along the wellbore. It is essential for ensuring drilling safety and optimizing drilling operations, particularly in determining the appropriate mud weight window to maintain well stability and improve the rate of penetration. Additionally, the UCS parameter plays a significant role in predicting sand production. Typically, UCS is determined through core sample tests in the laboratory, but this method is costly and time-consuming and provides either scattered data nor continuous log profile. Therefore, many studies around the world have proposed correlation between cored UCS and well log. However, these models often exhibit significant errors when applied to field X, Block Y in the Northern area of Cuu Long basin. In this paper, we propose a correlation between P-wave modulus (M) and Uniaxial compressive strength (UCS) derived from hydraulic flow units. Consequently, the correlation coefficient between cored UCS and log-derived UCS values is very high, with an overall value of 0.82 for all wells. In particular, the mean absolute error (MAE), mean absolute percentage error (MAPE), and root mean square error (RMSE) are as follows: for well A-1X — MAE: 426.3 psi, MAPE: 10.4%, RMSE: 582.5 psi; for well A-2X — MAE: 843.17 psi, MAPE: 16%, RMSE: 1115.5 psi; for well A-3X — MAE: 321.9 psi, MAPE: 7%, RMSE: 385.6 psi; and for well A-4X — MAE: 286.46 psi, MAPE: 10%, RMSE: 438.76 psi. The blind-test well B-1X shows acceptable errors, with MAE: 335.5 psi, MAPE: 14%, and RMSE: 383.37 psi. This model enhances the efficiency and safety of drilling and production operations in Field X, Block Y, located in the northeastern part of the Cuu Long Basin.

Ionosphere research using the Global Navigation Satellite Systems (GNSS) techniques is a hot topic, with their unprecedented high temporal and spatial sampling rate. We introduced a new GNSS Ionosphere Monitoring and Analysis Software (GIMAS) in order to model the global ionosphere vertical total electron content (VTEC) maps and to estimate the GPS and GLObalnaya NAvigatsionnaya Sputnikovaya Sistema (GLONASS) satellite and receiver differential code biases (DCBs). The GIMAS-based Global Ionosphere Map (GIM) products during low (day of year from 202 to 231, in 2008) and high (day of year from 050 to 079, in 2014) solar activity periods were investigated and assessed. The results showed that the biases of the GIMAS-based VTEC maps relative to the International GNSS Service (IGS) Ionosphere Associate Analysis Centers (IAACs) VTEC maps ranged from −3.0 to 1.0 TECU (TEC unit) (1 TECU = 1 × 1016 electrons/m2). The standard deviations (STDs) ranged from 0.7 to 1.9 TECU in 2008, and from 2.0 to 8.0 TECU in 2014. The STDs at a low latitude were significantly larger than those at middle and high latitudes, as a result of the ionospheric latitudinal gradients. When compared with the Jason-2 VTEC measurements, the GIMAS-based VTEC maps showed a negative systematic bias of about −1.8 TECU in 2008, and a positive systematic bias of about +2.2 TECU in 2014. The STDs were about 2.0 TECU in 2008, and ranged from 2.2 to 8.5 TECU in 2014. Furthermore, the aforementioned characteristics were strongly related to the conditions of the ionosphere variation and the geographic latitude. The GPS and GLONASS satellite and receiver P1-P2 DCBs were compared with the IAACs DCBs. The root mean squares (RMSs) were 0.16–0.20 ns in 2008 and 0.13–0.25 ns in 2014 for the GPS satellites and 0.26–0.31 ns in 2014 for the GLONASS satellites. The RMSs of receiver DCBs were 0.21–0.42 ns in 2008 and 0.33–1.47 ns in 2014 for GPS and 0.67–0.96 ns in 2014 for GLONASS. The monthly stability of the GPS satellite DCBs was about 0.04 ns (0.07 ns) in 2008 (2014) and that for the GLONASS satellite DCBs was about 0.09 ns in 2014. The receiver DCBs were less stable than the satellite DCBs, with a mean value of about 0.16 ns (0.47 ns) in 2008 (2014) for GPS, and 0.48 ns in 2014 for GLONASS. It can be demonstrated that the GIMAS software had a high accuracy and reliability for the global ionosphere monitoring and analysis.

Globally, wind energy if properly harnessed, could serve as a source of energy generation in Africa. This study compared the performance of two Machine Learning (ML) algorithms (Linear regression and Random Forest) in predicting wind speed in five major cities in Africa (Yaoundé, Pretoria, Nairobi, Cairo and Abuja). Wind data were collected between January 1, 2000, and December 31, 2022, using the Solar Radiation Data Archive. The data preprocessing was carried out with 80% of the data used for training and 20% for validation. The performance of these ML algorithms was evaluated using Mean Square Error (MSE), Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE) and coefficient of determination (R2). The result shows that Nairobi (3.814795 m/s) closely followed by Cairo (3.606453 m/s) has the highest mean wind speed while Yaoundé (1.090512 m/s) has the lowest. Based on the performance metrics used, the two Machine Learning algorithms were competitive. Still, the Linear Regression (LR) algorithm outperformed the Random Forest Algorithm in predicting wind speed in all the selected major African cities. In Yaoundé (RMSE = 0.3892, MAE= 0.3001, MAPE =0.5030), Pretoria (RMSE=1.2339, MAE=0.9480, MAPE=0.7450) Nairobi (RMSE= 0.4223, MAE =0.6499, MAPE =0.1872), Nairobi (RMSE=0.6499, MAE=0.5171, MAPE =0.1872), Cairo (RMSE =1.0909, MAE =0.8544, MAPE =0.3541) and Abuja (RMSE = 0.70245, MAE =0.5441, MAPE= 0.4515) the Linear regression algorithms was found to outperformed Random Forest Regression. Therefore, the Linear regression algorithm is more reliable in predicting wind speed compared with the Random Forest regression.

To prevent in a short time the rapid increase of tourists and corresponding traffic restriction measures’ lack in scenic areas, this study established a prediction model based on an improved convolutional neural network (CNN) and long- and short-term memory (LSTM) combined neural network. The study used this to predict the inflow and outflow of tourists in scenic areas. The model uses a residual unit, batch normalization, and principal component analysis to improve the CNN. The experimental results show that the model works best when batches’ quantity is 10, neurons’ quantity in the LSTM layer is 50, and the number of iterations is 50 on a workday; on non-working days, it is best to choose 10, 100, or 50. Using root mean square error (RMSE), normalized root mean square error (NRMSE), mean absolute percentage error (MAPE), and mean absolute error (MAE) as evaluation indicators, the inflow and outflow RMSEs of this study model are 82.51 and 89.80, MAEs are 26.92 and 30.91, NRMSEs are 3.99 and 3.94, and MAPEs are 1.55 and 1.53. Among the various models, this research model possesses the best prediction function. This provides a more accurate prediction method for the prediction of visitors’ flow rate in scenic spots. Meanwhile, the research model is also conducive to making corresponding flow-limiting measures to protect the ecology of the scenic area.