Prediction of Hot Deformation Behavior of 7Mo Super Austenitic Stainless Steel Based on Back Propagation Neural Network

The hot compression tests of 7Mo super austenitic stainless(SASS)were conducted to obtain flow curves at the temperature of 1000-1200℃and strain rate of 0.001 s^(-1)to 1 s^(-1).To predict the non-linear hot deformation behaviors of the steel,back propagation-artificial neural network(BP-ANN)with 16×8×8 hidden layer neurons was proposed.The predictability of the ANN model is evaluated according to the distribution of mean absolute error(MAE)and relative error.The relative error of 85%data for the BP-ANN model is among±5%while only 42.5%data predicted by the Arrhenius constitutive equation is in this range.Especially,at high strain rate and low temperature,the MAE of the ANN model is 2.49%,which has decreases for 18.78%,compared with conventional Arrhenius constitutive equation.

人工神经网络与计算机

探讨人工神经网络的计算机实现，从软件和硬件两方面阐述其实现过程，并从其数学算法中导出了计算机的实现算法。

Temperature compensation method of silicon microgyroscope based on BP neural network

The temperature characteristics of a silicon microgyroscope are studied, and the temperature compensation method of the silicon microgyroscope is proposed. First, an open-loop circuit is adopted to test the entire microgyroscope＇s resonant frequency and quality factor variations over temperature, and the zero bias changing trend over temperature is measured via a closed-loop circuit. Then, in order to alleviate the temperature effects on the performance of the microgyroscope, a kind of temperature compensated method based on the error back propagation（BP）neural network is proposed. By the Matlab simulation, the optimal temperature compensation model based on the BP neural network is well trained after four steps, and the objective error of the microgyroscope＇s zero bias can achieve 0.001 in full temperature range. By the experiment, the real time operation results of the compensation method demonstrate that the maximum zero bias of the microgyroscope can be decreased from 12.43 to 0.75（°）/s after compensation when the ambient temperature varies from -40 to 80℃, which greatly improves the zero bias stability performance of the microgyroscope.

Approximation Property of the Modified Elman Network

A new type of recurrent neural network is discussed, which provides the potential for modelling unknown nonlinear systems. The proposed network is a generalization of the network described by Elman, which has three layers including the input layer, the hidden layer and the output layer. The input layer is composed of two different groups of neurons, the group of external input neurons and the group of the internal context neurons. Since arbitrary connections can be allowed from the hidden layer to the context layer, the modified Elman network has more memory space to represent dynamic systems than the Elman network. In addition, it is proved that the proposed network with appropriate neurons in the context layer can approximate the trajectory of a given dynamical system for any fixed finite length of time. The dynamic backpropagation algorithm is used to estimate the weights of both the feedforward and feedback connections. The methods have been successfully applied to the modelling of nonlinear plants.

VIRTUAL TARGET DIFFERENTIAL GAME MIDCOURSE GUIDANCE LAW FOR HYPERSONIC CRUISE MISSILE BASED ON NEURAL NETWORK

For the high altitude cruising flight phase of a hypersonic cruise missile （HCM）, a relative motion mod- el between the missile and the target is established by defining virtual target and combining the theory of the dif- ferential geometry with missile motion equations. Based on the model, the motion between the missile and the tar- get is considered as a single target differential game problem, and a new open-loop differential game midcourse guidance law （DGMGL） is deduced by solving the corresponding Hamiltonian Function. Meanwhile, a new struc- ture of a closed-loop DGMGL is presented and the training data for back propagation neural network （BPNN） are designed. By combining the theory of BPNN with the open-loop DGMGL obtained above, the law intelligence is realized. Finally, simulation is carried out and the validity of the law is testified.

Improved BP Neural Network for Transformer Fault Diagnosis

The back propagation (BP)-based artificial neural nets (ANN) can identify complicated relationships among dissolved gas contents in transformer oil and corresponding fault types, using the highly nonlinear mapping nature of the neural nets. An efficient BP-ALM (BP with Adaptive Learning Rate and Momentum coefficient) algorithm is proposed to reduce the training time and avoid being trapped into local minima, where the learning rate and the momentum coefficient are altered at iterations. We developed a system of transformer fault diagnosis based on Dissolved Gases Analysis (DGA) with a BP-ALM algorithm. Training patterns were selected from the results of a Refined Three-Ratio method (RTR). Test results show that the system has a better ability of quick learning and global convergence than other methods and a superior performance in fault diagnosis compared to convectional BP-based neural networks and RTR.

Comprehensive Analysis and Artificial Intelligent Simulation of Land Subsidence of Beijing, China

Mechanism and modeling of the land subsidence are complex because of the complicate geological background in Beijing, China. This paper analyzed the spatial relationship between land subsidence and three factors, including the change of groundwater level, the thickness of compressible sediments and the building area by using remote sensing and GIS tools in the upper-middle part of alluvial-proluvial plain fan of the Chaobai River in Beijing. Based on the spatial analysis of the land subsidence and three factors, there exist significant non-linear relationship between the vertical displacement and three factors. The Back Propagation Neural Network (BPN) model combined with Genetic Algorithm (GA) was used to simulate regional distribution of the land subsidence. Results showed that at field scale, the groundwater level and land subsidence showed a significant linear relationship. However, at regional scale, the spatial distribution of groundwater depletion funnel did not overlap with the land subsidence funnel. As to the factor of compressible strata, the places with the biggest compressible strata thickness did not have the largest vertical displacement. The distributions of building area and land subsidence have no obvious spatial relationships. The BPN-GA model simulation results illustrated that the accuracy of the trained model during fifty years is acceptable with an error of 51% of verification data less than 20 mm and the average of the absolute error about 32 mm. The BPN model could be utilized to simulate the general distribution of land subsidence in the study area. Overall, this work contributes to better understand the complex relationship between the land subsidence and three influencing factors. And the distribution of the land subsidence can be simulated by the trained BPN-GA model with the limited available dada and acceptable accuracy.

Optimization of processing parameters for microwave drying of selenium-rich slag using incremental improved back-propagation neural network and response surface methodology

In the non-linear microwave drying process, the incremental improved back-propagation （BP） neural network and response surface methodology （RSM） were used to build a predictive model of the combined effects of independent variables （the microwave power, the acting time and the rotational frequency） for microwave drying of selenium-rich slag. The optimum operating conditions obtained from the quadratic form of the RSM are： the microwave power of 14.97 kW, the acting time of 89.58 min, the rotational frequency of 10.94 Hz, and the temperature of 136.407 ℃. The relative dehydration rate of 97.1895% is obtained. Under the optimum operating conditions, the incremental improved BP neural network prediction model can predict the drying process results and different effects on the results of the independent variables. The verification experiments demonstrate the prediction accuracy of the network, and the mean squared error is 0.16. The optimized results indicate that RSM can optimize the experimental conditions within much more broad range by considering the combination of factors and the neural network model can predict the results effectively and provide the theoretical guidance for the follow-up production process.

Comparison of flow behaviors of near beta Ti-55511 alloy during hot compression based on SCA and BPANN models

The flow behavior of Ti-55511 alloy was studied by hot compression tests at temperatures of 973−1123 K and strain rates of 0.01−10 s^(−1).Strain-compensated Arrhenius(SCA)and back-propagation artificial neural network(BPANN)methods were selected to model the constitutive relationship,and the models were further evaluated by statistical analysis and cross-validation.The stress−strain data extended by two models were implanted into finite element to simulate hot compression test.The results indicate that the flow stress is sensitive to deformation temperature and strain rate,and increases with increasing strain rate and decreasing temperature.Both the SCA model fitted by quintic polynomial and the BPANN model with 12 neurons can describe the flow behaviors,but the fitting accuracy of BPANN is higher than that of SCA.Sixteen cross-validation tests also confirm that the BPANN model has high prediction accuracy.Both models are effective and feasible in simulation,but BPANN model is superior in accuracy.

LEARNING ALGORITHM OF STAGE CONTROL NBP NETWORK

This letter analyzes the reasons why the known Neural Back Promulgation (NBP)network learning algorithm has slower speed and greater sample error. Based on the analysis and experiment, the training group descending Enhanced Combination Algorithm (ECA) is proposed.The analysis of the generalized property and sample error shows that the ECA can heighten the study speed and reduce individual error.

Application of neural network to prediction of plate finish cooling temperature

To improve the deficiency of the control system of finish cooling temperature (FCT), a new model developed from a combination of a multilayer perception neural network as the self-learning system and traditional mathematical model were brought forward to predict the plate FCT. The relationship between the self-learning factor of heat transfer coefficient and its influencing parameters such as plate thickness, start cooling temperature, was investigated. Simulative calculation indicates that the deficiency of FCT control system is overcome completely, the accuracy of FCT is obviously improved and the difference between the calculated and target FCT is controlled between -15 ℃ and 15 ℃.

Prostate Cancer Risk Prediction and Online Calculation Based on Machine Learning Algorithm

Objective To build a prostate cancer(PCa) risk prediction model based on common clinical indicators to provide a theoretical basis for the diagnosis and treatment of PCa and to evaluate the value of artificial intelligence(AI) technology under healthcare data platforms.Methods After preprocessing of the data from Population Health Data Archive,smuothly clipped absolute deviation(SCAD) was used to select features.Random forest(RF),support vector machine(SVM),back propagation neural network(BP),and convolutional neural network(CNN) were used to predict the risk of PCa,among which BP and CNN were used on the enhanced data by SMOTE.The performances of models were compared using area under the curve(AUC) of the receiving operating characteristic curve.After the optimal model was selected,we used the Shiny to develop an online calculator for PCa risk prediction based on predictive indicators.Results Inorganic phosphorus,triglycerides,and calcium were closely related to PCa in addition to the volume of fragmented tissue and free prostate-specific antigen(PSA).Among the four models,RF had the best performance in predicting PCa(accuracy:96.80%;AUC:0.975,95% CI:0.964-0.986).Followed by BP(accuracy:85.36%;AUC:0.892,95% CI:0.849-0.934) and SVM(accuracy:82.67%;AUC:0.824,95% CI:0.805-0.844).CNN performed worse(accuracy:72.37%;AUC:0.724,95% CI:0.670-0.779).An online platform for PCa risk prediction was developed based on the RF model and the predictive indicators.Conclusions This study revealed the application value of traditional machine learning and deep learning models in disease risk prediction under healthcare data platform,proposed new ideas for PCa risk prediction in patients suspected for PCa and had undergone core needle biopsy.Besides,the online calculation may enhance the practicability of AI prediction technology and facilitate medical diagnosis.

Simulation of phytoplankton biomass in Quanzhou Bay using a back propagation network model and sensitivity analysis for environmental variables

Prediction and sensitivity models,to elucidate the response of phytoplankton biomass to environmental factors in Quanzhou Bay,Fujian,China,were developed using a back propagation(BP) network.The environmental indicators of coastal phytoplankton biomass were determined and monitoring data for the bay from 2008 was used to train,test and build a three-layer BP artificial neural network with multi-input and single-output.Ten water quality parameters were used to forecast phytoplankton biomass(measured as chlorophyll-a concentration).Correlation coefficient between biomass values predicted by the model and those observed was 0.964,whilst the average relative error of the network was-3.46% and average absolute error was 10.53%.The model thus has high level of accuracy and is suitable for analysis of the influence of aquatic environmental factors on phytoplankton biomass.A global sensitivity analysis was performed to determine the influence of different environmental indicators on phytoplankton biomass.Indicators were classified according to the sensitivity of response and its risk degree.The results indicate that the parameters most relevant to phytoplankton biomass are estuary-related and include pH,sea surface temperature,sea surface salinity,chemical oxygen demand and ammonium.

Developing energy forecasting model using hybrid artificial intelligence method

An important problem in demand planning for energy consumption is developing an accurate energy forecasting model. In fact, it is not possible to allocate the energy resources in an optimal manner without having accurate demand value. A new energy forecasting model was proposed based on the back-propagation(BP) type neural network and imperialist competitive algorithm. The proposed method offers the advantage of local search ability of BP technique and global search ability of imperialist competitive algorithm. Two types of empirical data regarding the energy demand(gross domestic product(GDP), population, import, export and energy demand) in Turkey from 1979 to 2005 and electricity demand(population, GDP, total revenue from exporting industrial products and electricity consumption) in Thailand from 1986 to 2010 were investigated to demonstrate the applicability and merits of the present method. The performance of the proposed model is found to be better than that of conventional back-propagation neural network with low mean absolute error.

Auto recognition of carbonate microfacies based on an improved back propagation neural network

Though traditional methods could recognize some facies, e.g. lagoon facies, backshoal facies and foreshoal facies, they couldn't recognize reef facies and shoal facies well. To solve this problem, back propagation neural network(BP-ANN) and an improved BP-ANN with better stability and suitability, optimized by a particle swarm optimizer(PSO) algorithm(PSO-BP-ANN) were proposed to solve the microfacies' auto discrimination of M formation from the R oil field in Iraq. Fourteen wells with complete core, borehole and log data were chosen as the standard wells and 120 microfacies samples were inferred from these 14 wells. Besides, the average value of gamma, neutron and density logs as well as the sum of squares of deviations of gamma were extracted as key parameters to build log facies(facies from log measurements)-microfacies transforming model. The total 120 log facies samples were divided into 12 kinds of log facies and 6 kinds of microfacies, e.g. lagoon bioclasts micrite limestone microfacies, shoal bioclasts grainstone microfacies, backshoal bioclasts packstone microfacies, foreshoal bioclasts micrite limestone microfacies, shallow continental micrite limestone microfacies and reef limestone microfacies. Furthermore, 68 samples of these 120 log facies samples were chosen as training samples and another 52 samples were gotten as testing samples to test the predicting ability of the discrimination template. Compared with conventional methods, like Bayes stepwise discrimination, both the BP-ANN and PSO-BP-ANN can integrate more log details with a correct rate higher than 85%. Furthermore, PSO-BP-ANN has more simple structure, smaller amount of weight and threshold and less iteration time.

APPLICATIONOFNEURALNETWORKTOFLIGHTCONTROLSYSTEMDESIGN

Artificial neural network (ANN) has a great capability of self learning. The application of neural network to flight controller design can get good result. This paper studies the method of choosing controller parameters using neural network with Back Propagation (B P) algorithm. Design and simulation results show that this method can be used in flight control system design.

Underwater vehicle sonar self-noise prediction based on genetic algorithms and neural network

The factors that influence underwater vehicle sonar self-noise are analyzed, and genetic algorithms and a back propagation (BP) neural network are combined to predict underwater vehicle sonar self-noise. The experimental results demonstrate that underwater vehicle sonar self-noise can be predicted accurately by a GA-BP neural network that is based on actual underwater vehicle sonar data.

Classification and Identification of Nuclear, Biological or Chemical Agents Taken from Remote Sensing Image by Using Neural Network

In the context of new risks and threats associated to nuclear, biological and chemical （NBC） attacks, and given the shortcomings of certain analytical methods such as principal component analysis （PCA）, a neural network approach seems to be more accurate. PCA consists in projecting the spectrum of a gas collected from a remote sensing system in, firstly, a three-dimensional space, then in a two-dimensional one using a model of Multi-Layer Perceptron based neural network. It adopts during the learning process, the back propagation algorithm of the gradient, in which the mean square error output is continuously calculated and compared to the input until it reaches a minimal threshold value. This aims to correct the synaptic weights of the network. So, the Artificial Neural Network （ANN） tends to be more efficient in the classification process. This paper emphasizes the contribution of the ANN method in the spectral data processing, classification and identification and in addition, its fast convergence during the back propagation of the gradient.

Prediction of Wing Aerodynamic Performance in Rain Using Neural Net

A new method for prediction of wing aerodynamic performance in rain condition was presented.Three-and four-layer artificial neural networks based on improved algorithm for error Back Propagation(BP)network were respectively built.Detailed approaches to determine the optical parameters for network model were introduced and the specific steps for applying BP network model to predict wing aerodynamic performance in rain were given.On this basis,the established optimal three-and four-layer BP network model was used for this prediction.Results indicate that both of the network models are appropriate for predicting wing aerodynamic performance in rain.The sum of square error level produced by two models is less than 0.2%,and the prediction accuracy by four-layer network model is higher than that of three-layer network.

A biologically inspired model for pattern recognition

In this paper, a novel bionic model and its performance in pattern recognition are presented and discussed. The model is constructed from a bulb model and a three-layered cortical model, mimicking the main features of the olfactory system. The olfactory bulb and cortex models are connected by feedforward and feedback fibers with distributed delays. The Breast Cancer Wisconsin dataset consisting of data from 683 patients divided into benign and malignant classes is used to demonstrate the capacity of the model to learn and recognize patterns, even when these are deformed versions of the originally learned patterns. The performance of the novel model was compared with three artificial neural networks (ANNs), a back-propagation network, a support vector machine classifier, and a radial basis function classifier. All the ANNs and the olfactory bionic model were tested in a benchmark study of a standard dataset. Experimental results show that the bionic olfactory system model can learn and classify patterns based on a small training set and a few learning trials to reflect biological intelligence to some extent.