变截面门式刚架结构极限承载力分析

采用变截面薄腹构件是门式刚架结构轻型化的主要措施之一。如何利用变截面薄腹构件的腹板屈曲后强度并同时保证构件乃至结构的安全度始终是门式刚架结构工程设计的重要内容。根据有效宽度法的基本思想,提出了相关屈曲系数概念,在变截面H钢构件双重非线性数值分析基础上获得的局部屈曲荷载拟合公式,应用于变宽度楔形腹板的有效宽度简化计算公式当中。应用改进后的有效宽度法,进行了变截面门式刚架结构考虑构件局部屈曲影响的极限承载力分析,探讨了腹板平均宽厚比、腹板楔率对结构极限承载力的影响,提出了对变截面门式刚架结构优化设计的工程建议,并具有工程指导意义。     

The photocatalytic degradation of carbamazepine and prediction by artificial neural networks

The three layer artificial neural network model was applied to predict the degradation efficiency for carbamazepine in photocatalytic oxidation under UV radiation. Titania–zirconia was employed as a catalyst for the photooxidation. The catalyst was prepared using titanium isopropoxide and zirconium oxychloride by sol–gel method and characterized by transmission electron microscopy and BET analysis. Different process parameters such as, initial concentration of carbamazepine, pH of the solution, catalyst concentration and time of UV irradiation were employed as the input to the artificial neural network model and the output of the network was degradation efficiency of carbamazepine. The multilayer feed-forward networks with the Levenberg–Marquardt (trainlm) backpropagation training algorithm was used for the network training. The smallest mean square error was obtained for three-layer network with ‘logsig’ transfer function and five neurons in the hidden layer gave optimal results. A comparison between the predicted values and selective experimental data of degradation efficiency showed a high correlation coefficient (R²) of 0.997.     

Prediction of gas explosion pressures: A machine learning algorithm based on KPCA and an optimized LSSVM

A gas explosion, as a common accident in public life and industry, poses a great threat to the safety of life and property. The determination and prediction of gas explosion pressures are greatly important for safety issues and emergency rescue after an accident occurs. Compared with traditional empirical and numerical models, machine learning models are definitely a superior approach. However, the application of machine learning in gas explosion pressure prediction has not reached its full potential. In this study, a hybrid gas explosion pressure prediction model based on kernel principal component analysis (KPCA), a least square support vector machine (LSSVM), and a gray wolf optimization (GWO) algorithm is proposed. A dataset consisting of 12 influencing factors of gas explosion pressures and 317 groups of data is constructed for developing and evaluating the KPCA-GWO-LSSVM model. The results show that the correlations among the 12 influencing factors are eliminated and dimensioned down by the KPCA method, and 5 composite indicators are obtained. The proposed KPCA-GWO-LSSVM hybrid model performs well in predicting gas explosion pressures, with coefficient of determination (R²), root mean square error (RMSE), and mean absolute error (MAE) values of 0.928, 26.234, and 12.494, respectively, for the training set; and 0.826, 25.951, and 13.964, respectively, for the test set. The proposed model outperforms the LSSVM, GWO-LSSVM, KPCA-LSSVM, beetle antennae search improved BP neural network (BAS-BPNN) models and reported empirical models. In addition, the sensitivity of influencing factors to the model is evaluated based on the constructed database, and the geometric parameters X1 and X2 of the confined structure are the most critical variables for gas explosion pressure prediction. The findings of this study can help expand the application of machine learning in gas explosion prediction and can truly benefit the treatment of gas explosion accidents.     

A beetle antennae search improved BP neural network model for predicting multi-factor-based gas explosion pressures

A gas explosion in an underground structure may cause serious damage to the human body and ground buildings and may result in huge economic losses. The pressure of the gas explosion is an important parameter in determining its severity and designating an emergency plan. However, existing empirical and computational fluid dynamics (CFD) methods for pressure prediction are either inaccurate or inefficient when considering multiple influencing factors and their interrelationships. Therefore, for a more efficient and reliable prediction, the present study developed a multifactorial prediction model based on a beetle antennae search (BAS) algorithm improved back propagation (BP) neural network. A total of 317 sets of data which considered factors of geometry, gas, obstacle, vent, and ignition were collected from previous studies. The results showed that the established model can predict pressures accurately by low RMSE (43.4542 and 50.7176) and MAPE (3.9666% and 4.9605%) values and high R² (0.7696 and 0.7388) values for training and testing datasets, respectively. Meanwhile, the BAS algorithm was applied to improve both the calculation efficiency and the accuracy of the proposed model by enabling a more intelligent hyperparameter tuning method. Furthermore, the permutation importance of input variables was investigated, and the length (L) and the ratio of length and diameter (L/D) of geometry were found to be the most critical factors that affect the explosion pressure level.     

基于粗糙集-神经网络的矿山地质环境影响评价模型及应用

采用衡山白果地区石膏矿山的11个评价指标,综合运用粗糙集和神经网络理论,构建了基于粗糙集-神经网络(RS-ANN)的矿山地质环境影响评价模型,对RSES软件约简的数据和无约简的数据采用EasyNN-plus软件进行预测评价。神经网络模型的输入属性为8个,而粗糙集-神经网络模型的输入属性为6个,训练样本均为13个,预测样本均为4个,前者的平均预测精度为1.85%～24.86%,后者为1.23%～15.28%。研究发现,粗糙集在保留关键信息的前提下可有效地对数据表进行约简,约简后的神经网络预测结果与实际情况吻合,并比无约简时总体精度有较大幅度提高。     

基于PCA-GA-BP神经网络的瓦斯含量预测分析

为提高煤层瓦斯含量预测的效率和准确率,提出了先采用主成份分析（PCA）方法来降低变量间的相关性,然后将遗传算法（GA）与BP神经网络相结合的煤层瓦斯含量预测的新方法。为了避免BP神经网络收敛速度慢、易陷入局部极小值等问题,算法采用GA对BP神经网络的权值和阈值进行优化,利用Matlab软件进行编程,建立了BP神经网络和GA-BP神经网络瓦斯含量预测模型。选取淮南某矿瓦斯含量及其影响因素作为实验数据对该模型进行了实例分析,将主成份回归和BP网络算法预测结果与该模型进行了对比分析。结果表明:PCA-GA-BP网络预测模型平均相对误差为2.759%,预测效果明显优于主成份回归和BP网络预测模型,可以准确的预测煤层瓦斯含量。     

A hybrid approach to faults detection and diagnosis in batch and semi-batch reactors by using EKF and neural network classifier

This work deals with a new hybrid approach for the detection and diagnosis of faults in different parts of fed-batch and batch reactors. In this paper, the fault detection method is based on the using of Extended Kalman Filter (EKF) and statistical test. The EKF is used to estimate on-line in added to the state of reactor the overall heat transfer coefficient (U). The diagnosis method is based on a probabilistic neural network classifier. The Inputs of the probabilistic classifier are the input–output measurements of reactor and the parameter U estimated by EKF, while the outputs of the classifier are fault types in reactor. This new approach is illustrated for simulated as well as experimental data sets using two cases of reactions: the first is the oxidation of sodium thiosulfate by hydrogen peroxide and the second is alkaline hydrolyse of ethyl benzoate in homogeneous hydro-alcoholic. Finally, the combination of the estimated parameter U using EKF and probabilistic neural network classifier provided the best results. These results show the performance of the proposed approach to monitoring the semi-batch and batch reactors.     

A unique intelligent approach for forecasting project completion time in oil refineries

This study presents a hybrid approach for accurate forecasting of project completion time with noisy and uncertain safety factors in oil refineries. The hybrid approach is based on artificial neural network (ANN), fuzzy mathematical programming (FMP) and conventional regression. Three indictors, namely, number of occupational injuries, number of employees and ratio of maximum useful hours over useful hour per month are considered as inputs. Also, project completion time is considered as the main output. To achieve the objective of this study, five sets of data with respect to oil refinery construction projects in various cities of Iran are collected and analyzed through statistical methods. It is shown that for the actual case of this study, ANN presents lowest mean absolute percentage error (MAPE). Also, analysis of variance (ANOVA) is used to verify and validate the results of this study. This is the first study that presents a hybrid approach for accurate estimation and forecasting of project completion time with complex, noisy and uncertain occupational factors.     

基于BP网络的建筑安装施工现场安全综合评价的研究

针对目前我国建筑安装施工现场安全评价技术的不成熟和欠科学性的现状 ,笔者分析和综合了目前安全评价技术 ,结合建筑业特点 ,提出了基于BP神经网络的建筑安装施工现场安全评价方法 ,并对该评价模型的原理、方法及算法进行了研究。首先 ,结合建筑安装施工现场安全生产的特点建立评价指标体系 ,随后 ,运用层次分析法确定指标及准则层的权重 ,并运用模糊综合评价法生成评价样本集 ,最后 ,利用样本集训练BP网络 ,待误差满足要求后 ,即可运用训练成功的BP神经网络进行安全评价。     

An adaptive neural network algorithm for assessment and improvement of job satisfaction with respect to HSE and ergonomics program: The case of a gas refinery

Researchers have been continuously trying to improve human performance with respect to Health, Safety and Environment (HSE) and ergonomics (hence HSEE). This study proposes an adaptive neural network (ANN) algorithm for measuring and improving job satisfaction among operators with respect to HSEE in a gas refinery. To achieve the objectives of this study, standard questionnaires with respect to HSEE are completed by operators. The average results for each category of HSEE are used as inputs and job satisfaction is used as output for the ANN algorithm. Moreover, ANN is used to rank operators performance with respect to HSEE and job satisfaction. Finally, Normal probability technique is used to identify outlier operators. Moreover, operators with inadequate job satisfaction with respect to HSEE are identified. This would help managers to see if operators are satisfied with their jobs in the context of HSEE. This is the first study that introduces an integrated ANN algorithm for assessment and improvement of human job satisfaction with respect to HSEE program in complex systems.     

岩层移动角选取的神经网络方法研究

岩层移动角是进行各类保护煤柱设计时的关键性参数 ,涉及地表建 (构 )筑物的安全。在综合分析影响岩层移动角因素的基础上 ,采用人工神经网络方法建立岩层移动角选取的模型。该模型采用改进的BP算法 ,运用我国典型的地表移动观测站资料作为学习训练样本和测试样本 ,对模型的计算结果与实测值进行了对比分析。分析结果表明 :用人工神经网络方法求算岩层移动角考虑的因素更为全面 ,结果更接近于实际。笔者为岩层移动角的理论计算探索出了一种新的方法。     

Pulling force prediction using neural networks

Purpose. In ergonomics and human factors investigations, pulling force (PF) estimation has usually been achieved using various types of biomechanical models, and independent approximation of PF was done with the help of upper extremity joints. Recently, multiple regression methods have gained importance for task-relevant inputs in predicting PF. Artificial neural networks (ANNs) also play a vital role in fitting the data; however, their use in work-related biomechanics and ergonomics is inadequate. Therefore, the current research aimed to accomplish comparative investigation of ANN and regression models by assessing their capacity to predict PF values. Methods. Multipositional PF data were acquired from 200 subjects at three different handle heights and body locations. ANN and regression models were formed using a random sample of three subsets (75% training, 15% selection, 10% validation) for proving the outcomes. Results. The comparison of ANN and regression models shows that the predictions of ANN models had a profoundly explained variance and lower root mean square difference values for the PF data at three handle heights. Conclusions. These outcomes advise that ANNs offer a precise and robust substitute for regression methods, and should be considered a useful method in biomechanics and ergonomics task assessments.     

Traffic accident reconstruction and an approach for prediction of fault rates using artificial neural networks: A case study in Turkey

Objective: Currently, in Turkey, fault rates in traffic accidents are determined according to the initiative of accident experts (no speed analyses of vehicles just considering accident type) and there are no specific quantitative instructions on fault rates related to procession of accidents which just represents the type of collision (side impact, head to head, rear end, etc.) in No. 2918 Turkish Highway Traffic Act (THTA 1983). The aim of this study is to introduce a scientific and systematic approach for determination of fault rates in most frequent property damage–only (PDO) traffic accidents in Turkey.

Methods: In this study, data (police reports, skid marks, deformation, crush depth, etc.) collected from the most frequent and controversial accident types (4 sample vehicle–vehicle scenarios) that consist of PDO were inserted into a reconstruction software called vCrash. Sample real-world scenarios were simulated on the software to generate different vehicle deformations that also correspond to energy-equivalent speed data just before the crash. These values were used to train a multilayer feedforward artificial neural network (MFANN), function fitting neural network (FITNET, a specialized version of MFANN), and generalized regression neural network (GRNN) models within 10-fold cross-validation to predict fault rates without using software. The performance of the artificial neural network (ANN) prediction models was evaluated using mean square error (MSE) and multiple correlation coefficient (R).

Results: It was shown that the MFANN model performed better for predicting fault rates (i.e., lower MSE and higher R) than FITNET and GRNN models for accident scenarios 1, 2, and 3, whereas FITNET performed the best for scenario 4. The FITNET model showed the second best results for prediction for the first 3 scenarios. Because there is no training phase in GRNN, the GRNN model produced results much faster than MFANN and FITNET models. However, the GRNN model had the worst prediction results. The R values for prediction of fault rates were close to 1 for all folds and scenarios.

Conclusions: This study focuses on exhibiting new aspects and scientific approaches for determining fault rates of involvement in most frequent PDO accidents occurring in Turkey by discussing some deficiencies in THTA and without regard to initiative and/or experience of experts. This study yields judicious decisions to be made especially on forensic investigations and events involving insurance companies. Referring to this approach, injury/fatal and/or pedestrian-related accidents may be analyzed as future work by developing new scientific models.     

Long-term prediction model of rockburst in underground openings using heuristic algorithms and support vector machines

Rockburst possibility prediction is an important activity in many underground openings design and construction as well as mining production. Due to the complex features of rockburst hazard assessment systems, such as multivariables, strong coupling and strong interference, this study employs support vector machines (SVMs) for the determination of classification of long-term rockburst for underground openings. SVMs is firmly based on the theory of statistical learning algorithms, uses classification technique by introducing radial basis function (RBF) kernel function. The inputs of models are buried depth H, rocks’ maximum tangential stress σ_θ, rocks’ uniaxial compressive strength σ_c, rocks’ uniaxial tensile strength σ_t, stress coefficient σ_θ/σ_c, rock brittleness coefficient σ_c/σ_t and elastic energy index W_et. In order to improve predictive accuracy and generalization ability, the heuristic algorithms of genetic algorithm (GA) and particle swarm optimization algorithm (PSO) are adopted to automatically determine the optimal hyper-parameters for SVMs. The performance of hybrid models (GA + SVMs = GA-SVMs) and (PSO + SVMs = PSO-SVMs) have been compared with the grid search method of support vector machines (GSM-SVMs) model and the experimental values. It also gives variance of predicted data. A rockburst dataset, which consists of 132 samples, was employed to evaluate the current method for predicting rockburst grade, and the good results of overall success rate were obtained. The results indicated that the heuristic algorithms of GA and PSO can speed up SVMs parameter optimization search, the proposed method is robust model and might hold a high potential to become a useful tool in rockburst prediction research.     

应急状态下矿井瓦斯爆炸致灾因子传播快速分类器研究*

为了在矿井瓦斯爆炸灾变发生后,快速确定瓦斯爆炸冲击波的压力、温度、有毒有害气体等致灾因子在井巷网络中的传播情况。利用CFD数值模拟或爆炸实验获得瓦斯爆炸冲击波的压力、温度、有毒有害气体等致灾因子传播大数据,将影响瓦斯爆炸传播的因素以及观测点等参数作为人工神经网络的输入节点,压力、温度等致灾因子作为输出节点,建立瓦斯爆炸致灾因子传播快速预测机器学习模型,解决CFD数值模拟的建模、计算及数据分析处理等过程耗时大、不适应灾变应急的快速响应等问题。研究结果表明:在给定爆炸位置和爆炸当量的均直巷道,获得任一点的爆炸冲击波压力、温度以及有毒有害气体所需时间是瞬时的,人工神经网络平均训练误差为6.92 %,有训练样本的验证误差为5.24 %,无训练样本的验证误差为6.88 %。     

改进BP算法在煤与瓦斯突出预测中的应用

为了正确预测煤与瓦斯突出的趋势与危险性 ,基于反向BP神经网络 ,笔者提出了一种改进的BP网络模型 :为了加快BP网络的收敛速度 ,增强其跳出局部极小点的能力 ,采用了自适应变步长法和改进模拟退火法 (SA法 )相结合的方法。实际应用表明 ,该模型收敛速度快 ,准确性高 ,具有较高的可靠性和实用性 ,是一种十分有效的煤与瓦斯突出危险性预测方法。     

基于神经网络方法的复杂超静定结构失效概率分析

将整体结构按拓扑关系划分为若干模块,根据力的传递原理对模块结构进行失效概率计算,获得各模块结构的失效概率信息;运用有限元模拟分析获得整体结构的失效概率信息。再将模块结构的失效概率作为输入,整体结构的失效概率作为输出,构造样本集。以BP神经网络进行失效概率分析,既可提高计算速度和精度,也可利用其泛化能力对相同拓扑结构的超静定结构进行失效概率计算。算例中对包含5个模块结构的整体结构单元进行基于神经网络的失效概率分析,以网络外推能力计算了包含7模块结构的整体结构单元的失效概率,获得较好的计算精度,从而验证了该方法的有效性。     

中空夹层钢管再生混凝土柱火灾后力学性能分析

基于ABAQUS建立中空夹层钢管再生混凝土短柱理论模型,进行受火后加载模拟,受火过程采用ISO-834标准升、降温曲线。进而探究柱构件在受火全过程中的温度分布规律和受火后的力学性能。并分析了受火时间、空心率、取代率、名义含钢率、钢管强度和混凝土强度等参数对柱构件剩余承载力的影响。结果表明:受火时间、含钢率、混凝土强度和钢材强度对构件剩余承载力影响较明显,空心率和取代率对构件剩余承载力影响较小;受火45 min和90 min后,空心率取33.90%能较大限度减轻构件自重,且具备良好的抗火性能。     

应用电性拓扑状态指数预测烷烃自燃点

建立了一个基于人工神经网络的定量结构-性质相关性模型,用于52种烷烃化合物自燃点的预测研究。应用原子类型电性拓扑状态指数作为表征分子结构特征的描述符。该指数既能表征分子的电子特性,又反映其拓扑特征,同时易于计算,并有较强的同分异构体区分能力。采用误差反向传播（BP）神经网络方法对烷烃自燃点与电性拓扑状态指数间可能存在的非线性关系进行拟合。将52种烷烃样本随机划分为训练集（30种）、验证集（8种）和测试集（14种）,并通过“试差法”确定网络的最优参数。运用最佳网络结构[64—1]对实验样本进行模拟,结果表明,多数样本的自燃点预测值与实验值符合良好,对于测试集,平均预测绝对误差为8．4℃,均方根误差为11．8,优于多元线性回归方法和传统基团贡献法所得结果。该方法的提出为工程上提供了一种根据分子结构预测有机物白燃点的有效方法。