1988-2013年基于BP神经网络的植被叶面积指数遥感定量反演

Aiming at the shortcomings of the low efficiency optimization algorithm and the tendency to fall into the local optimal solution in the existing shallow sea geoacoustic parameters inversion research, a shallow sea geoacoustic parameters inversion method based on the Back Propagation (BP) neural network model is proposed. First, the theoretical prediction value of the shallow sea sound pressure field is obtained through the Fast Field Method (FFM). Then the mapping relationship between the predicted sound pressure field and the geoacoustic parameters value to be inverted is established according to the BP neural network model. Finally, the measured sound pressure field data is brought into the neural network to obtain the inversion results. The experimental results show that, compared with the classic simulated annealing (Simulated Annealing, SA) algorithm, the BP neural network model converts the data matching process of the optimization algorithm into the construction of a relational model, avoiding multiple matching optimization processes for the inversion results, and the inversion results can be output directly, accurately, and efficiently. Under the premise of setting the same accuracy, the calculation time of the BP neural network model for the inversion result is shortened, and the number of iterations is reduced to 5% of the SA algorithm.

In this paper, the effectiveness of the genetic algorithm-based back propagation (GABP) neural network model and its application to the breakout prediction in the continuous casting process are investigated. The formation of the sticking-type breakouts and the prediction principle of thermocouple thermometry method are analysed firstly. Then the genetic algorithm-based back propagation neural network model is proposed by fusing genetic algorithm (GA), and error back propagation neural network to offset the demerits of one paradigm by the merits of another. Finally, the GABP neural network model is applied to the breakout prediction in the continuous casting process; and the feasibility of the model is verified by the testing result with the accuracy rate of 97.56% and the prediction rate of 100%.

In view of emerging applications of alpha+beta titanium alloys in aerospace and defense, we have aimed to develop a Back propagation neural network (BPNN) model capable of predicting the properties of these alloys as functions of alloy composition and/or thermomechanical processing parameters. The optimized BPNN model architecture was based on the sigmoid transfer function and has one hidden layer with ten nodes. The BPNN model showed excellent predictability of five properties: Tensile strength (r: 0.96), yield strength (r: 0.93), beta transus (r: 0.96), specific heat capacity (r: 1.00) and density (r: 0.99). The developed BPNN model was in agreement with the experimental data in demonstrating the individual effects of alloying elements in modulating the above properties. This model can serve as the platform for the design and development of new alpha+beta titanium alloys in order to attain desired strength, density and specific heat capacity.

Quantitative analysis of fatty acid value during rice storage based on olfactory visualization sensor technology

Improving the resilience of enterprise safety production is one of the important ways to deal with the frequency of safety accidents. Based on the definition of enterprise safety production resilience, we fully consider the impacts of recovery resilience, self-organizing resilience, and learning resilience as the three dimensions of enterprise safety production resilience. We build a back propagation (BP) neural network model that analyzes the main factors of enterprise safety production resilience using the results of gray relational analysis as an input that can assess the resilience of enterprise safety production and provide a valuable reference for the improvement of an enterprise’s safety production level. The results show that the resilience of production safety obviously increased after the Chinese enterprises with low resilience (as predicted by the model) adopted the corresponding early warning methods. The gray relational degree analysis method can incorporate well the variables for the establishment of the BP neural network prediction model.

To study drivers’ driving behaviour in dilemma zones at signalized intersections, dilemma zones at signalized intersections are taken as the research objects. Based on analyzing the driving characteristics of motor vehicles. And collect the driving behavior parameters of video observations and road parameters of field observations, take these parameters as the factors which influence the drivers’ decision, conduct the significance analysis with SPSS(Statistical Product Service Solutions) for significant influencing factors, and then take these factors as the input value of BP(Back Propagation) neural network model, BP neural network model is set up and tested based on TensorFlow in Python. The prediction model of drivers’ decision-making in dilemma zone of signalized intersections under different speed limits is obtained, the research shows that the lower the speed limit, the better the accuracy of the intersection prediction. Further verified by comparison with the prediction model of drivers’ decision-making based on binary logistic, the accuracy of different prediction models is analyzed based on actual driver decisions.

Sea surface temperature (SST) is one of the most important parameters in air–sea interaction, and its accurate prediction is of great significance in the study of global climate change. However, SST is affected by heat flux, ocean dynamic processes, cloud coverage, and other factors, which means it contains linear and nonlinear components. Existing prediction models, especially single prediction models, cannot effectively handle these linear and nonlinear components in the meantime, degrading their accuracy concerning the prediction of SST. To remedy this weakness, this paper proposes a novel prediction model by the Lagrange multiplier method to combine the auto-regressive integrated moving average (ARIMA) model and the back propagation (BP) neural network model, where these two models have superior prediction performance for linear and nonlinear components, respectively. Moreover, the genetic algorithm is exploited to construct the genetic algorithm BP (GABP) neural network to further improve the performance of the proposed model. To verify the effectiveness of the proposed model, experiments predicting the SST based on historic time-series data are performed. The experiment results indicate that the mean absolute error (MAE) of the ARIMA-GABP model is only 0.3033 °C and the root mean square error (RMSE) is 0.3970 °C, which is better than the ARIMA model, BP neural network model, long short-term memory (LSTM) model, GABP neural network model, and ensemble empirical model decomposition BP model among various datasets. Therefore, the proposed model has superior and robust performance concerning predicting SST.

Mining-induced ground fissures are the main type of geological disasters found on the Loess Plateau, China, and cause great impacts on the soil properties around ground fissures. However, little research has been conducted on the quantitative relationship between ground fissures and changes in soil properties. To address this, 40 ground fissures in the Yungang mining area, Datong City, Shanxi Province, China, were investigated, and changes in soil properties (soil organic matter, soil moisture, field capacity, bulk density, soil porosity, and grain compositions) were revealed by the difference in soil properties between the edge and contrast points around ground fissures. Redundancy analyses were used to illustrate the relationships between the value (Si_DV) and percentage (Si_DP) of the difference in soil properties between the edge and contrast points, as well as the ground fissures. The characteristics of ground fissures that had a significant correlation according to Pearson correlation analysis with Si_DP were selected and analyzed via multivariate linear fitting model, random forest model, and Back Propagation (BP) neural network model, respectively. Results show that soil organic matter, soil moisture content, bulk density, field capacity, and the content of clay at the edge points were significantly less than those at the contrast points; conversely, soil porosity at the edge points was significantly greater. The average percentage of the difference between the edge points and contrast points of ground fissures in these six properties was 15.27%, while soil moisture content showed the greatest change (20.65%). The Si_DP was significantly correlated with the width, slope, and vegetation coverage of ground fissures; however, the vegetation coverage was the determining factor. BP neural network model had the greatest performance in revealing the relationships between ground fissures and changes in soil properties. The model for soil organic matter had the highest accuracy (R2 = 0.89), and all others were above 0.5. This research provides insights into the quantitative relationship between ground fissures and their impacts on soil physical properties, which can be used in conjunction with remote sensing images to rapidly assess soil erosion risks caused by mining on a large scale, given that soil physical properties are closely related to topsoil stability.

With the rapid development of communication technology as well as a rapid rise in the usage of electronic devices, a growth of concerns over unintentional electromagnetic interference emitted by these devices has been witnessed. Pioneer researchers have deeply studied the relationship between the shielding effectiveness and a few mixed design parameters for cementitious composites incoporating carbon fibres by conducting physical experiments. This paper, therefore, aims to develop and propose a series of prediction models for the shielding effectiveness of cementitious composites involving carbon fibres using frequency and mixed design parameters, such as the water-to-cement ratio, fibre content, sand-to-cement ratio and aspect ratio of the fibres. A multi-variable non-linear regression model and a backpropagation neural network (BPNN) model were developed to meet the different accuracy requirements as well as the complexity requirements. The results showed that the regression model reached an R2 of 0.88 with a root mean squared error (RMSE) of 2.3 dB for the testing set while the BPNN model had an R2 of 0.96 with an RMSE of 2.64 dB. Both models exhibited a sufficient prediction accuracy, and the results also supported that both the regression and the BPNN model are reasonable for such estimation.

In this study, a back-propagation neural network (BPNN) model is developed to optimize forming parameters of aluminum alloy 5052-H32 sheets subjected to incremental sheet forming (ISF). Process parameters including tool diameter, step size in depth, tool feed rate, and tool spindle speed are varied to investigate the formability of the test, which is quantified by a multi-objective fitness function to reach the maximum forming angle and the minimum thickness reduction. A series of experimental tests have been conducted with different values of the above-mentioned process parameters. A BPNN model was then developed to predict material responses during ISF process. This BPNN model is employed to search for Pareto optimal solutions of formability of ISF. The derived Pareto optimal front appears to be a useful design guide for forming of aluminum alloy 5052 sheets in the ISF process.KeywordsIncremental sheet formingParameter optimizationBack-propagation neural networkMulti-objective optimalGenetic algorithm

To reduce the impact of wildfires on the operation of power systems, a back-propagation neural network (BPNN) model is used to evaluate the wildfire risk distribution after feature selection. Data from 14 types of wildfire-related features, including anthropogenic, geographical, and meteorological factors, were collected from public data websites and local departments. The weight ranking was calculated using filtering and wrapper methods to form five feature subsets. These are used as the input sets of the BPNN model training, and network parameters are optimized by genetic algorithm (GA). Finally, the optimal feature subset is chosen to establish the optimal BPNN model. With the optimal model, the prediction results are graded to draw a wildfire risk distribution map. Situated in medium-, high-, and very-high-risk zones are 90.26% of new fire incidents, indicating the applicability of the proposed BPNN model.

Leaf area index (LAI) serves as a critical indicator for evaluating crop growth and guiding field management practices. While spectral information (vegetation indices and texture features) extracted from multispectral sensors mounted on unmanned aerial vehicles (UAVs) holds promise for LAI estimation, the limitations of single-texture features necessitate further exploration. Therefore, this study conducted field experiments over two consecutive years (2021-2022) to collect winter oilseed rape LAI ground truth data and corresponding UAV multispectral imagery. Vegetation indices were constructed, and canopy texture features were extracted. Subsequently, a correlation matrix method was employed to establish novel randomized combinations of three-dimensional texture indices. By analyzing the correlations between these parameters and winter oilseed rape LAI, variables with significant correlations (p < 0.05) were selected as model inputs. These variables were then partitioned into distinct combinations and input into three machine learning models-Support Vector Machine (SVM), Backpropagation Neural Network (BPNN), and Extreme Gradient Boosting (XGBoost)-to estimate winter oilseed rape LAI. The results demonstrated that the majority of vegetation indices and texture features exhibited significant correlations with LAI (p < 0.05). All randomized texture index combinations also showed strong correlations with LAI (p < 0.05). Notably, the three-dimensional texture index NDTTI exhibited the highest correlation with LAI (R = 0.725), derived from the spatial combination of DIS5, VAR5, and VAR3. Integrating vegetation indices, texture features, and three-dimensional texture indices as inputs into the XGBoost model yielded the highest estimation accuracy. The validation set achieved a determination coefficient (R2) of 0.882, a root mean square error (RMSE) of 0.204 cm2cm-2, and a mean relative error (MRE) of 6.498%. This study provides an effective methodology for UAV-based multispectral monitoring of winter oilseed rape LAI and offers scientific and technical support for precision agriculture management practices.

The Leaf Area Index (LAI) is a critical parameter that sheds light on the composition and function of forest ecosystems. Its efficient and rapid measurement is essential for simulating and estimating ecological activities such as vegetation productivity, water cycle, and carbon balance. In this study, we propose to combine high-resolution GF-6 2 m satellite images with the LESS three-dimensional RTM and employ different machine learning algorithms, including Random Forest, BP Neural Network, and XGBoost, to achieve LAI inversion for forest stands. By reconstructing real forest stand scenarios in the LESS model, we simulated reflectance data in blue, green, red, and near-infrared bands, as well as LAI data, and fused some real data as inputs to train the machine learning models. Subsequently, we used the remaining measured LAI data for validation and prediction to achieve LAI inversion. Among the three machine learning algorithms, Random Forest gave the highest performance, with an R2 of 0.6164 and an RMSE of 0.4109, while the BP Neural Network performed inefficiently (R2 = 0.4022, RMSE = 0.5407). Therefore, we ultimately employed the Random Forest algorithm to perform LAI inversion and generated LAI inversion spatial distribution maps, achieving an innovative, efficient, and reliable method for forest stand LAI inversion.

Disassembly is a crucial step in the remanufacturing of end-of-life (EoL) electronic products. Disassembly depth refers to the disassembly stop point determined by the disassembly sequence. For the disassembly depth optimization of EoL electronic products, a feasibility model with a fast convergence and low mean squared error (MSE) is needed to improve optimization accuracy. However, the use of a backpropagation neural network (BPNN) model or mathematical model often results in a slow convergence and high MSE due to the randomness of the initial weights and biases. In this study, an improved method for the disassembly depth optimization of smartphones based on a Particle Swarm Optimization-BPNN (PSO-BPNN) predictive model is proposed. Compared with the traditional BPNN optimization method, the proposed method in this study is that the BPNN predictive model is optimized by using PSO, which shows a superior predictive performance and reduces the MSE. The case of ‘Huawei P7’ is used to verify the feasibility of the method. The results show that the method maintains disassembly profit while reducing the disassembly time and carbon emissions by 17.1% and 7.8%, respectively. Compared with the BPNN model, the PSO-BPNN model converges 18.6%, 32.8%, and 16.6% faster in predicting the disassembly time, profit, and carbon emissions, respectively, with MSE reductions of 92.95%, 96.51%, and 92.74%, respectively.

A back-propagation neural network (BPNN) model is developed to predict the punching shear strength of square ferrocement slabs. The experimental data used for training and testing the neural network model, are collected from several sources.They are arranged in a format of seven input parameters (the effective span, slab thickness, yield tensile strength of wire mesh, volume fraction of wire mesh, mortar compressive strength, width of square loaded area,boundary condition of the supported slabs) and one output parameter (punching shear strength). A parametric study is carried out using BPNN to study the influence of each parameter affecting the punching shear strength of ferrocement slabs. A comparison with the experimental results and those from other existing empirical equations demonstrates that the predictions from BPNN are indeed better. We conclude that the BPNN model may serve as a good tool for predicting the punching shear strength.