Modelling monthly-gridded carbon emissions based on nighttime light data.

Timely and accurate implementation of carbon emissions (CE) analysis and evaluation is necessary for policymaking and management. However, previous inventories, most of which are yearly, provincial or city, and incomplete, have failed to reflect the spatial variations and monthly trends of CE. Based on nighttime light (NTL) data, statistical data, and land use data, in this study, a high-resolution (1 km × 1 km) monthly inventory of CE was developed using back propagation neural network, and the spatiotemporal variations and impact factors of CE at multiple administrative levels was evaluated using spatial autocorrelation model and spatial econometric model. As a large province in terms of both economy and population, Guangdong is facing the severe emission reduction challenges. Therefore, in this study, Guangdong was taken as a case study to explain the method. The results revealed that CE increased unsteadily in Guangdong from 2013 to 2022. Spatially, the high CE areas were distributed in the Pearl River Delta region such as Guangzhou, Shenzhen, and Dongguan, while the low CE areas were distributed in West and East Guangdong. The Global Moran's I decreased from 2013 to 2022 at the city and county levels, suggesting that the inequality of CE in Guangdong steadily decreased at these two administrative levels. Specifically, at the city level, the Global Moran's I gradually decreased from 0.4067 in 2013 to 0.3531 in 2022. In comparison, at the county level, the trend exhibited a slower decline, from 0.3647 in 2013 to 0.3454 in 2022. Furthermore, the analysis of the impact factors revealed that the relationship between CE and gross domestic product was an inverted U-shaped, suggesting the existence of the inverted U-shaped Environmental Kuznets Curve for CE in Guangdong. In addition, the industrial structure had larger positive impact on CE at the different levels. The method developed in this study provides a perspective for establishing high spatiotemporal resolution CE evaluation through NTL data, and the improved inventory of CE could help understand the spatial-temporal variations of CE and formulate regional-monthly-specific emission reduction policies.

Formulation Optimization and Performance Prediction of Red Mud Particle Adsorbents Based on Neural Networks.

Red mud (RM), a bauxite residue, contains hazardous radioactive wastes and alkaline material and poses severe surface water and groundwater contamination risks, necessitating recycling. Pretreated RM can be used to make adsorbents for water treatment. However, its performance is affected by many factors, resulting in a nonlinear correlation and coupling relationship. This study aimed to identify the best formula for an RM adsorbent using a mathematical model that examines the relationship between 11 formulation types (e.g., pore-assisting agent, component modifier, and external binder) and 9 properties (e.g., specific surface area, wetting angle, and Zeta potential). This model was built using a back-propagation neural network (BP) based on single-factor experimental data and orthogonal experimental data. The model trained and predicted the established network structure to obtain the optimal adsorbent formula. The RM particle adsorbents had a pH of 10.16, specific surface area (BET) of 48.92 m2·g-1, pore volume of 2.10 cm3·g-1, compressive strength (ST) of 1.12 KPa, and 24 h immersion pulverization rate (ηm) of 3.72%. In the removal of total phosphorus in flotation tailings backwater, it exhibited a good adsorption capacity (Q) and total phosphorous removal rate (η) of 48.63 mg·g-1 and 95.13%, respectively.

An Adaptive Multi-Mode Navigation Method with Intelligent Virtual Sensor Based on Long Short-Term Memory in GNSS Restricted Environment.

Aiming at the problem of fast divergence of pure inertial navigation system without correction under the condition of GNSS restricted environment, this paper proposes a multi-mode navigation method with an intelligent virtual sensor based on long short-term memory (LSTM). The training mode, predicting mode, and validation mode for the intelligent virtual sensor are designed. The modes are switching flexibly according to GNSS rejecting situation and the status of the LSTM network of the intelligent virtual sensor. Then the inertial navigation system (INS) is corrected, and the availability of the LSTM network is also maintained. Meanwhile, the fireworks algorithm is adopted to optimize the learning rate and the number of hidden layers of LSTM hyperparameters to improve the estimation performance. The simulation results show that the proposed method can maintain the prediction accuracy of the intelligent virtual sensor online and shorten the training time according to the performance requirements adaptively. Under small sample conditions, the training efficiency and availability ratio of the proposed intelligent virtual sensor are improved significantly more than the neural network (BP) as well as the conventional LSTM network, improving the navigation performance in GNSS restricted environment effectively and efficiently.

Analysis of the Vibration Characteristics of a Leaf Spring System Using Artificial Neural Networks.

The real-time vibrations occurring in a leaf spring system may cause undesirable effects, such as stresses, strains, deflections, and surface deformations over the system. In order to detect the most appropriate working conditions in which the leaf spring system will work more stably and also to design optimized leaf spring systems, these external effects have to be detected with high accuracy. In this work, artificial neural network-based estimators have been proposed to analyze the vibration effects on leaf spring systems. In the experimental studies carried out, the vibration effects of low, medium, and high-pressure values applied by a hydraulic piston on a steel leaf spring system have been analyzed by a 3-axial accelerometer. After the experimental studies, the Radial Basis Artificial Neural Network (RBANN) and Cascade-Forward Back-Propagation Artificial Neural Network (CFBANN) based nonlinear artificial neural network structures have been proposed to analyze the vibration data measured from the leaf spring system under relevant working conditions. The simulation results represent that the RBANN structure can estimate the real-time vibrations occurring on the leaf spring system with higher accuracy and reaches lower RMS error values when compared to the CFBANN structure. In general, it can be concluded that the RBANN and CFBANN network structures can successfully be used in the estimation of real-time vibration data.

Strike Velocity Prediction of Stick Blunt Instruments Based on Backpropagation Neural Network.

OBJECTIVES: To analyze and predict the striking velocity range of stick blunt instruments in different populations, and to provide basic data for the biomechanical analysis of blunt force injuries in forensic identification. METHODS: Based on the Photron FASTCAM SA3 high-speed camera, Photron FASTCAM Viewer 4.0 and SPSS 26.0 software, the tester's maximum striking velocity of stick blunt instruments and related factors were calculated and analyzed, and inputed to the backpropagation (BP) neural network for training. The trained and verified BP neural network was used as the prediction model. RESULTS: A total of 180 cases were tested and 470 pieces of data were measured. The maximum striking velocity range was 11.30-35.99 m/s. Among them, there were 122 female data, the maximum striking velocity range was 11.63-29.14 m/s; there were 348 male data, the maximum striking velocity range was 20.11-35.99 m/s. The maximum striking velocity of stick blunt instruments increased with the increase of weight and height, but there was no obvious increase trend in the male group; the maximum striking velocity decreased with age, but there was no obvious downward trend in the female group. The maximum striking velocity of stick blunt instruments has no significant correlation with the material and strike posture. The root mean square error (RMSE), the mean absolute error (MAE) and the coefficient of determination (R2) of the prediction results by using BP neural network were 2.16, 1.63 and 0.92, respectively. CONCLUSIONS: The prediction model of BP neural network can meet the demand of predicting the maximum striking velocity of different populations.

Adaboost-Based Machine Learning Improved the Modeling Robust and Estimation Accuracy of Pear Leaf Nitrogen Concentration by In-Field VIS-NIR Spectroscopy.

Different cultivars of pear trees are often planted in one orchard to enhance yield for its gametophytic self-incompatibility. Therefore, an accurate and robust modelling method is needed for the non-destructive determination of leaf nitrogen (N) concentration in pear orchards with mixed cultivars. This study proposes a new technique based on in-field visible-near infrared (VIS-NIR) spectroscopy and the Adaboost algorithm initiated with machine learning methods. The performance was evaluated by estimating leaf N concentration for a total of 1285 samples from different cultivars, growth regions, and tree ages and compared with traditional techniques, including vegetation indices, partial least squares regression, singular support vector regression (SVR) and neural networks (NN). The results demonstrated that the leaf reflectance responded to the leaf nitrogen concentration were more sensitive to the types of cultivars than to the different growing regions and tree ages. Moreover, the AdaBoost.RT-BP had the best accuracy in both the training (R2 = 0.96, root mean relative error (RMSE) = 1.03 g kg-1) and the test datasets (R2 = 0.91, RMSE = 1.29 g kg-1), and was the most robust in repeated experiments. This study provides a new insight for monitoring the status of pear trees by the in-field VIS-NIR spectroscopy for better N managements in heterogeneous pear orchards.

Looseness Monitoring of Bolted Spherical Joint Connection Using Electro-Mechanical Impedance Technique and BP Neural Networks.

The bolted spherical joint (BSJ) has wide applications in various space grid structures. The bar and the bolted sphere are connected by the high-strength bolt inside the joint. High-strength bolt is invisible outside the joint, which causes the difficulty in monitoring the bolt looseness. Moreover, the bolt looseness leads to the reduction of the local stiffness and bearing capacity for the structure. In this regard, this study used the electro-mechanical impedance (EMI) technique and back propagation neural networks (BPNNs) to monitor the bolt looseness inside the BSJ. Therefore, a space grid specimen having bolted spherical joints and tubular bars was considered for experimental evaluation. Different torques levels were applied on the sleeve to represent different looseness degrees of joint connection. As the torque levels increased, the looseness degrees of joint connection increased correspondingly. The lead zirconate titanate (PZT) patch was used and integrated with the tubular bar due to its strong piezoelectric effect. The root-mean-square deviation (RMSD) of the conductance signatures for the PZT patch were used as the looseness-monitoring indexes. Taking RMSD values of sub-frequency bands and the looseness degrees as inputs and outputs respectively, the BPNNs were trained and tested in twenty repeated experiments. The experimental results show that the formation of the bolt looseness can be detected according to the changes of looseness-monitoring indexes, and the degree of bolt looseness by the trained BPNNs. Overall, this research demonstrates that the proposed structural health monitoring (SHM) technique is feasible for monitoring the looseness of bolted spherical connection in space grid structures.

Intelligent identification on cotton verticillium wilt based on spectral and image feature fusion.

BACKGROUND: Verticillium wilt is the major disease of cotton, which would cause serious yield reduction and economic losses, and the identification of cotton verticillium wilt is of great significance to cotton research. However, the traditional method is still manual, which is subjective, inefficient, and labor-intensive, and therefore, this study has proposed a novel method for cotton verticillium wilt identification based on spectral and image feature fusion. The cotton hyper-spectral images have been collected, while the regions of interest (ROI) have been extracted as samples including 499 healthy leaves and 498 diseased leaves, and the average spectral information and RGB image of each sample were obtained. In spectral feature processing, the preprocessing methods including Savitzky-Golay smoothing (SG), multiplicative scatter correction (MSC), de-trending (DT) and mean normalization (MN) algorithms have been adopted, while the feature band extraction methods have adopted principal component analysis (PCA) and successive projections algorithm (SPA). In RGB image feature processing, the EfficientNet was applied to build classification model and 16 image features have been extracted from the last convolutional layer. And then, the obtained spectral and image features were fused, while the classification model was established by support vector machine (SVM) and back propagation neural network (BPNN). Additionally, the spectral full bands and feature bands were used as comparison for SVM and BPNN classification respectively. RESULT: The results showed that the average accuracy of EfficientNet for cotton verticillium wilt identification was 93.00%. By spectral full bands, SG-MSC-BPNN model obtained the better performance with classification accuracy of 93.78%. By feature bands, SG-MN-SPA-BPNN model obtained the better performance with classification accuracy of 93.78%. By spectral and image fused features, SG-MN-SPA-FF-BPNN model obtained the best performance with classification accuracy of 98.99%. CONCLUSIONS: The study demonstrated that it was feasible and effective to use fused spectral and image features based on hyper-spectral imaging to improve identification accuracy of cotton verticillium wilt. The study provided theoretical basis and methods for non-destructive and accurate identification of cotton verticillium wilt.

Ratio Fluorescent Detecting of Tryptophan and Its Metabolite 5-Hydroxyindole-3-acetic Acid Relevant with Depression via Tb(III) Modified HOFs Hybrids: Further Designing Recyclable Molecular Logic Gate Connected by Back Propagation Neural Network.

Exploring intelligent fluorescent materials with high reliability and precision to diagnose diseases is significant but remains a great challenge. Herein, based on coordination post-synthetic modification, a Tb3+ functionalized ME-PA (Tb@1) is prepared, which can emit brilliant green fluorescence through ligand-to-mental charge transfer-assisted energy transfer (LMCT-ET) process from ME-PA to Tb3+ ions. Tb@1 can simultaneously distinguish Tryptophan (Try) and its metabolite 5-hydroxyindole-3-acetic acid (5-HIAA), two effective indicators for depression, in ratio and colorimetric mode. And this sensor behaves the advantages of high efficiency and sensitivity, as well as excellent reusability and anti-interference. The PET process from ME to Try and 5-HIAA, and the competitive absorption between analytes and Tb@1 may be relevant to sensing mechanism. In realistic serum or urine environment, the detection limits of Tb@1 for Try and 5-HIAA are 0.0183 and 0.0149 mg L-1 respectively. Moreover, in conjunction with back propagation neural network (BPNN), two dual-output molecular logic gates that can be calculated circularly are further designed, which realizes intelligent control of the electronic component to identify the existence of two biomarkers and judge their concentrations from fluorescence images. This work offers a novel approach to modulate logic circuits based on ML-assisted HOF fluorescent sensor, with promising application for a precise and pictorial depression diagnosis.

Exploring the relationships between perceived umami intensity, umami components and electronic tongue responses in food matrices.

Umami intensity promotes food flavor blending and food choice, while a universal quantification procedure is still lacking. To evaluate perceived umami intensity (PUI) in seven categories of foods, modified two-alternative forced choice (2-AFC) method with monosodium glutamate as reference was applied. Meanwhile, we explored whether equivalent umami concentration (EUC) by chemical analysis and electronic tongue (E-tongue) are applicable in PUI quantification. The results indicated that EUC was appropriate in quantifying PUI of samples from meat, dairy, vegetable and mushroom groups (r = 1.00, p < 0.05). Moreover, models with a good prediction capacity for PUI and EUC (R2 > 0.99) were established in separated food categories by back propagation neural networks, where E-tongue data were set as input. This study explored the effectiveness of the three methods in evaluating the PUIs of various foods, which provides multiple choices for the food industry.

Nondestructive prediction of freshness for bighead carp (Hypophthalmichthys nobilis) head by Excitation-Emission Matrix (EEM) analysis based on fish eye fluid: Comparison of BPNNs and RBFNNs.

This study established back-propagation neural networks (BPNNs) and radial basis function neural networks (RBFNNs) models for evaluating the freshness of bighead carp head storage at different temperatures via the characteristic components of Excitation-Emission Matrix (EEM). Two characteristic components of EEM data of fish eye fluid were extracted by parallel factor analysis (PARAFAC) and were the most efficient components to stimulate fluorophores responsible for fish freshness detection during variable temperatures. EEM-RBFNNs and EEM-BPNNs models based on characteristic components of EEM used to predict the fish freshness. The results demonstrated the relative errors of EEM-BPNNs models for hiobarbituric acid reactive substances (TBARS) and total viable bacteria count (TAC) prediction were less than 10% which were better than those of EEM-RBFNNs models. It indicated that EEM-BPNNs model of bighead carp eye fluid by PARAFAC has a high potential for predicting fish freshness under variable storage conditions.

Applying high-frequency surrogate measurements and a wavelet-ANN model to provide early warnings of rapid surface water quality anomalies.

It is critical for surface water management systems to provide early warnings of abrupt, large variations in water quality, which likely indicate the occurrence of spill incidents. In this study, a combined approach integrating a wavelet artificial neural network (wavelet-ANN) model and high-frequency surrogate measurements is proposed as a method of water quality anomaly detection and warning provision. High-frequency time series of major water quality indexes (TN, TP, COD, etc.) were produced via a regression-based surrogate model. After wavelet decomposition and denoising, a low-frequency signal was imported into a back-propagation neural network for one-step prediction to identify the major features of water quality variations. The precisely trained site-specific wavelet-ANN outputs the time series of residual errors. A warning is triggered when the actual residual error exceeds a given threshold, i.e., baseline pattern, estimated based on long-term water quality variations. A case study based on the monitoring program applied to the Potomac River Basin in Virginia, USA, was conducted. The integrated approach successfully identified two anomaly events of TP variations at a 15-minute scale from high-frequency online sensors. A storm event and point source inputs likely accounted for these events. The results show that the wavelet-ANN model is slightly more accurate than the ANN for high-frequency surface water quality prediction, and it meets the requirements of anomaly detection. Analyses of the performance at different stations and over different periods illustrated the stability of the proposed method. By combining monitoring instruments and surrogate measures, the presented approach can support timely anomaly identification and be applied to urban aquatic environments for watershed management.

Back propagation neural network modelling of biodegradation and fermentative biohydrogen production using distillery wastewater in a hybrid upflow anaerobic sludge blanket reactor.

In a hybrid upflow anaerobic sludge blanket (HUASB) reactor, biodegradation in association with biohydrogen production was studied using distillery wastewater as substrate. The experiments were carried out at ambient temperature (34±1°C) and acidophilic pH of 6.5 with constant hydraulic retention time (HRT) of 24h at various organic loading rates (OLRs) (1-10.2kgCODm(-3)d(-1)) in continuous mode. A maximum hydrogen production rate of 1300mLd(-1) was achieved. A back propagation neural network (BPNN) model with network topology of 4-20-1 using Levenberg-Marquardt (LM) algorithm was developed and validated. A total of 231 data points were studied to examine the performance of the HUASB reactor in acclimatisation and operation phase. The statistical qualities of BPNN models were significant due to the high correlation coefficient, R(2), and lower mean absolute error (MAE) between experimental and simulated data. From the results, it was concluded that BPNN modelling could be applied in HUASB reactor for predicting the biodegradation and biohydrogen production using distillery wastewater.