Intelligent model selection with updating parameters during staged excavation using optimization method

Various constitutive models have been proposed, and previous studies focused on identifying parameters of specified models. To develop the smart construction, this paper proposes a novel optimization-based intelligent model selection procedure in which parameter identification is also performed during staged excavation. To conduct the model selection, a database of seven constitutive models accounting for isotropic or anisotropic yield surface, isotropic or anisotropic elasticity, or small strain stiffness for clayey soils is established, with each model numbered and deemed as one additional parameter for optimization. A newly developed real-coded genetic algorithm is adopted to evaluate the performance of simulation against field measurement. As the process of optimization goes on, the soil model exhibiting good performance during simulation survives from the database and model parameters are also optimized. For each excavation stage, with the selected model and optimized parameters, wall deflection and ground surface settlement of the subsequent unexcavated stage are predicted. The proposed procedure is repeated until the entire excavation is finished. This proposed procedure is applied to a real staged excavation with field data, which demonstrates its effectiveness and efficiency in engineering practice with highlighting the importance of anisotropic elasticity and small strain stiffness in simulating excavation. All results demonstrate that the current study has both academic significance and practical significance in providing an efficient and effective approach of adaptive optimization-based model selection with parameters updating in engineering applications.

     

水文模型参数多目标率定及最优非劣解优选

针对概念性水文模型参数众多、相互制约,且多目标参数优化率定最优参数求解困难、易受决策者主观因素影响的问题,采用多目标优化算法对水文模型参数进行率定,得到模型参数最优非劣解集,在此基础上,引入最小最大后悔值决策理论,并结合Pareto支配基本理论,提出了一种多目标最优非劣解选取准则。以柘溪流域为研究对象,采用三目标MOSCDE优化率定新安江模型的参数,并与单目标SCE-UA优化结果进行对比分析。结果表明,提出的非劣解选取方法可以有效从大规模非劣解集中筛选出最优非劣解,大大缩短参数率定耗时。     

Use of evolutionary computing for modelling some complex problems in geotechnical engineering

In this paper, the feasibility of using evolutionary computing for solving some complex problems in geotechnical engineering is investigated. The paper presents a relatively new technique, i.e. evolutionary polynomial regression (EPR), for modelling three practical applications in geotechnical engineering including the settlement of shallow foundations on cohesionless soils, pullout capacity of small ground anchors and ultimate bearing capacity of pile foundations. The prediction results from the proposed EPR models are compared with those obtained from artificial neural network (ANN) models previously developed by the author, as well as some of the most commonly available methods. The results indicate that the proposed EPR models agree well with (or better than) the ANN models and significantly outperform the other existing methods. The advantage of EPR technique over ANNs is that EPR generates transparent and well-structured models in the form of simple and easy-to-use hand calculation formulae that can be readily used by practising engineers.     

Settlement of Shallow Foundations on Cohesionless Soils Based on SPT Value Using Multi-Objective Feature Selection

Accurate prediction of settlement for shallow footings on cohesionless soil is a complex geotechnical problem due to large uncertainties associated with soil. Prediction of the settlement of shallow footings on cohesionless soil is based on in situ tests as it is difficult to find out the properties of soil in the laboratory and standard penetration test (SPT) is the most often used in situ test. In data driven modelling, it is very difficult to choose the optimal input parameters, which will govern the model efficiency along with a better generalization. Feature subset selection involves minimization of both prediction error and the number of features, which are in general mutual conflicting objectives. In this study, a multi-objective optimization technique is used, where a non-dominated sorting genetic algorithm (NSGA II) is combined with a learning algorithm (neural network) to develop a prediction model based on SPT data based on the Pareto optimal front. Pareto optimal front gives the user freedom to choose a model in terms of accuracy and model complexity. It is also shown how NSGA II can be effectively applied to select the optimal parameters and besides minimizing the error rate. The developed model is compared with existing models in terms of different statistical criteria and found to be more efficient.     

Optimal design of groundwater remediation systems using a multi-objective fast harmony search algorithm

A new multi-objective optimization methodology is developed, whereby a multi-objective fast harmony search (MOFHS) is coupled with a groundwater flow and transport model to search for optimal design of groundwater remediation systems under general hydrogeological conditions. The MOFHS incorporates the niche technique into the previously improved fast harmony search and is enhanced by adding the Pareto solution set filter and an elite individual preservation strategy to guarantee uniformity and integrity of the Pareto front of multi-objective optimization problems. Also, the operation library of individual fitness is introduced to improve calculation speed. Moreover, the MOFHS is coupled with the commonly used flow and transport codes MODFLOW and MT3DMS, to search for optimal design of pump-and-treat systems, aiming at minimization of the remediation cost and minimization of the mass remaining in aquifers. Compared with three existing multi-objective optimization methods, including the improved niched Pareto genetic algorithm (INPGA), the non-dominated sorting genetic algorithm II (NSGAII), and the multi-objective harmony search (MOHS), the proposed methodology then demonstrated its applicability and efficiency through a two-dimensional hypothetical test problem and a three-dimensional field problem in Indiana (USA).     

An evolutionary approach to modelling the thermomechanical behaviour of unsaturated soils

A new data‐mining approach is presented for modelling of the stress–strain and volume change behaviour of unsaturated soils considering temperature effects. The proposed approach is based on the evolutionary polynomial regression (EPR), which unlike some other data‐mining techniques, generates a transparent and structured representation of the behaviour of systems directly from raw experimental (or field) data. The proposed methodology can operate on large quantities of data in order to capture nonlinear and complex relationships between contributing variables. The developed models allow the user to gain a clear insight into the behaviour of the system. Unsaturated triaxial test data from the literature were used for development and verification of EPR models. The developed models were also used (in a coupled manner) to produce the entire stress path of triaxial tests. Comparison of the EPR model predictions with the experimental data revealed the robustness and capability of the proposed methodology in capturing and reproducing the constitutive thermomechanical behaviour of unsaturated soils. More importantly, the capability of the developed models in accurately generalizing the predictions to unseen data cases was illustrated. The results of a sensitivity analysis showed that the models developed from data are able to capture and represent the physical aspects of the unsaturated soil behaviour accurately. The merits and advantages of the proposed methodology are also discussed. Copyright © 2014 John Wiley & Sons, Ltd.     

Identification of coupling parameters between shear strength behaviour of compacted soils and chemical’s effects with an evolutionary-based data mining technique

When subjected to a very high-pH, most of the soil minerals undergo physico-chemical transformation. This could induce strong modifications of the shear strength behaviour of the soil. This issue is of high interest in the framework of the design of deep nuclear wastes repositories, since the degradation of the concrete lining of deep galleries after thousands of years will generate an alkaline solute (pH > 12) that would circulate through the backfill, and alter its hydromechanical characteristics. A study was undertaken to assess the impact of high-pH fluid circulation on the shear strength behaviour of a backfill material. Because of the complexity of the existing constitutive theories, a new approach was used, based on evolutionary polynomial regression (EPR), for modelling of these processes. EPR is an evolutionary data mining technique that generates a transparent and structured representation of the behaviour of a system directly from data. An EPR model was developed and validated using results from a comprehensive set of triaxial tests. Through a sensitivity analysis, the EPR model permitted to identify the specific surface, and to a lesser extent the micropore void ratio, as coupling parameters between hydromechanical behaviour alteration during alkaline fluid circulation and a physical process.     

Comparison of surrogate models with different methods in groundwater remediation process

Surrogate modelling is an effective tool for reducing computational burden of simulation optimization. In this article, polynomial regression (PR), radial basis function artificial neural network (RBFANN), and kriging methods were compared for building surrogate models of a multiphase flow simulation model in a simplified nitrobenzene contaminated aquifer remediation problem. In the model accuracy analysis process, a 10-fold cross validation method was adopted to evaluate the approximation accuracy of the three surrogate models. The results demonstrated that: RBFANN surrogate model and kriging surrogate model had acceptable approximation accuracy, and further that kriging model’s approximation accuracy was slightly higher than RBFANN model. However, the PR model demonstrated unacceptably poor approximation accuracy. Therefore, the RBFANN and kriging surrogates were selected and used in the optimization process to identify the most cost-effective remediation strategy at a nitrobenzene-contaminated site. The optimal remediation costs obtained with the two surrogate-based optimization models were similar, and had similar computational burden. These two surrogate-based optimization models are efficient tools for optimal groundwater remediation strategy identification.     

Landform classification: a high-performing mapping unit partitioning tool for landslide susceptibility assessment—a test in the Imera River basin (northern Sicily,Italy)

In landslide susceptibility studies, the type of mapping unit adopted affects the obtained models and maps in terms of accuracy, robustness, spatial resolution and geomorphological adequacy. To evaluate the optimal selection of these units, a test has been carried out in an important catchment of northern Sicily (the Imera River basin), where the spatial relationships between a set of predictors and an inventory of 1608 rotational/translational landslides were analysed using the multivariate adaptive regression splines (MARS) method. In particular, landslide susceptibility models were prepared and compared by adopting four different types of mapping units: the largely adopted grid cells (PX), the typical contributing area–controlled slope units (5000_SLU), the recently optimized parameter-free multiscale slope units (PF_SLU) and a new type (LCL_SLU) of slope unit obtained by crossing classic hydrological partitioning with landform classification. At the same time, once a pixel-based model was prepared, four different SLU modelling strategies were applied to each of the obtained slope unit layers, including two different types of pixel score zoning, a pixel score re-modelling and a factor-based SLU re-modelling. According to the achieved results, LCL_SLUs produced the highest performance and reliability, offering an optimal compromise between the high-performing but scattered and the smoothed but lower-performing prediction images that were obtained from pixel-based and hydrologic SLU–based modelling, respectively. Additionally, among the four adopted SLU modelling strategies, the new proposed procedure, which uses the zoned pixel–based score deciles as the LCL_SLU predictors for a new regression, resulted in the best outstanding performance (ROC_AUC?=?0.95).

     

多目标系统模糊模型在强夯参数优选中的应用

将一种多目标系统的模糊模型用于地基强夯参数的优选,使这种定性的问题得到了定量化的处理。实例分析表明,模糊优选较其它方法科学、合理。     

Intelligent modelling of clay compressibility using hybrid meta-heuristic and machine learning algorithms

Compression index Ccis an essential parameter in geotechnical design for which the effectiveness of correlation is still a challenge.This paper suggests a novel modelling approach using machine learning(ML)technique.The performance of five commonly used machine learning(ML)algorithms,i.e.back-propagation neural network(BPNN),extreme learning machine(ELM),support vector machine(SVM),random forest(RF)and evolutionary polynomial regression(EPR)in predicting Cc is comprehensively investigated.A database with a total number of 311 datasets including three input variables,i.e.initial void ratio e0,liquid limit water content wL,plasticity index Ip,and one output variable Cc is first established.Genetic algorithm(GA)is used to optimize the hyper-parameters in five ML algorithms,and the average prediction error for the 10-fold cross-validation(CV)sets is set as thefitness function in the GA for enhancing the robustness of ML models.The results indicate that ML models outperform empirical prediction formulations with lower prediction error.RF yields the lowest error followed by BPNN,ELM,EPR and SVM.If the ranges of input variables in the database are large enough,BPNN and RF models are recommended to predict Cc.Furthermore,if the distribution of input variables is continuous,RF model is the best one.Otherwise,EPR model is recommended if the ranges of input variables are small.The predicted correlations between input and output variables using five ML models show great agreement with the physical explanation.     

Selection of sand models and identification of parameters using an enhanced genetic algorithm

Numerous constitutive models of granular soils have been developed during the last few decades. As a consequence, how to select an appropriate model with the necessary features based on conventional tests and with an easy way of identifying parameters for geotechnical applications has become a major issue. This paper aims to discuss the selection of sand models and parameters identification by using genetic algorithm. A real‐coded genetic algorithm is enhanced for the optimization with high efficiency. Models with gradually varying features (elastic‐perfectly plastic modelling, nonlinear stress–strain hardening, critical state concept and two‐surface concept) are selected from numerous sand models as examples for optimization. Conventional triaxial tests on Hostun sand are selected as the objectives in the optimization. Four key points are then discussed in turn: (i) which features are necessary to be accounted for in constitutive modelling of sand; (ii) which type of tests (drained and/or undrained) should be selected for an optimal identification of parameters; (iii) what is the minimum number of tests that should be selected for parameter identification; and (iv) what is the suitable and least strain level of objective tests to obtain reliable and reasonable parameters. Finally, a useful guide, based on all comparisons, is provided at the end of the discussion. Copyright © 2015 John Wiley & Sons, Ltd.     

城市固体废弃物卫生填埋场选址评价的模糊集方法

针对固体废弃物卫生填埋场地适宜性评价具有多目标性和模糊性的特点,引入改进的层次分析法(AHP)确定评价指标的权重,在此基础上构造多目标模糊模式识别模型,发展了模糊模式识别模型和模糊综合评判模型相结合的多层次多目标复杂系统模糊集评价方法。利用该方法对某地固体废弃物卫生填埋场进行了选址评价,论证了固体废弃物卫生填埋场的场地适宜性。     

三峡水库围堰发电期调度模拟优化研究

汛限水位是协调综合利用水库防洪和兴利矛盾的重要参数。三峡水库围堰发电期调度规则将汛期水库水位控制在134．9-135．4m运行,本文将汛期(6-9月)按旬划分为12个分期,建立模拟模型和动态规划模型;采用宜昌站120年实测日流量序列,通过所建模型计算,得到了优化的分期汛限水位方案。模型实现了模拟和优化的有机结合,将带机遇约束的多目标优化问题转化为确定性约束条件下的单目标优化问题,降低了复杂性。结果表明,模型在不增加防洪风险的前提下,发电效益增加了2．21％(2005 年)和2．32％(2006年)。     

Optimization of reinforcement strategies for dangerous dams considering time-average system failure probability and benefit�Ccost ratio using a life quality index

Lifetime-oriented multi-objective optimization for structural reinforcement based on series-system reliability and benefit?Ccost analysis has already been proposed. It is still needed, however, to incorporate the life quality index (LQI) in the lifetime reinforcement optimization process of complex massive infrastructure engineering with correlated series?Cparallel failure modes (e.g., dams). An improved technique combining overall system failure probability with benefit?Ccost analysis based on the LQI is developed. An approach to obtain time-average system failure probability with correlated series?Cparallel failure modes is proposed to measure the structural performance. Then, the concept for benefit?Ccost ratio based on LQI including failure consequence and life quality objective is introduced. As an application of the methodology, the optimal reinforcement strategy for an existing earth dam is shown. Three types of reinforcement strategies, that is, preventive reinforcement, essential reinforcement and that lies between them, are selected. The results show that the preventive reinforcement strategy is the most beneficial for a dam, whose failure loss involving human life is tremendous. The advantage of the proposed approach is its ability to harmonize overall structural safety with reinforcement cost and can be extended to optimization of reinforcement strategies for other massive infrastructure engineering projects.     

High-dimensional geostatistical history matching

Hydrocarbon reservoir modelling and characterisation is a challenging subject within the oil and gas industry due to the lack of well data and the natural heterogeneities of the Earth’s subsurface. Integrating historical production data into the geo-modelling workflow, commonly designated by history matching, allows better reservoir characterisation and the possibility of predicting the reservoir behaviour. We present herein a geostatistical-based multi-objective history matching methodology. It starts with the generation of an initial ensemble of the subsurface petrophysical property of interest through stochastic sequential simulation. Each model is then ranked according the match between its dynamic response, after fluid flow simulation, and the observed available historical production data. This enables building regionalised Pareto fronts and the definition of a large ensemble of optimal subsurface Earth models that fit all the observed production data without compromising the exploration of the uncertainty space. The proposed geostatistical multi-objective history matching technique is successfully implemented in a benchmark synthetic reservoir dataset, the PUNQ-S3, where 12 objectives are targeted.     

Evaluation of liquefaction potential based on CPT results using evolutionary polynomial regression

In this paper a new approach is presented, based on evolutionary polynomial regression (EPR), for determination of liquefaction potential of sands. EPR models are developed and validated using a database of 170 liquefaction and non-liquefaction field case histories for sandy soils based on CPT results. Three models are presented to relate liquefaction potential to soil geometric and geotechnical parameters as well as earthquake characteristics. It is shown that the EPR model is able to learn, with a very high accuracy, the complex relationship between liquefaction and its contributing factors in the form of a function. The attained function can then be used to generalize the learning to predict liquefaction potential for new cases not used in the construction of the model. The results of the developed EPR models are compared with a conventional model as well as a number of neural network-based models. It is shown that the proposed EPR model provides more accurate results than the conventional model and the accuracy of the EPR results is better than or at least comparable to that of the neural network-based models proposed in the literature. The advantages of the proposed EPR model over the conventional and neural network-based models are highlighted.     

石家庄地区水资源利用与经济发展协调模型研究

运用大系统递阶优化控制理论对石家庄地区水资源开发利用与经济发展协调管理问题进行分析和研究,提出该地区水资源开发利用与经济发展协调管理的三级递阶模型;探讨了模型系统优化识别问题和系统协调管理模式及其优化方法。通过数字仿真、灵敏度分析,说明模型运行结果准确可靠,可作为管理部门决策的依据。     

大连周水子海水入侵区地下水多目标优化管理模型

将改进后的遗传算法GA（添加了小生境、Pareto解集过滤器等模块）与变密度地下水流及溶质运移模拟程序SEAWAT-2000相耦合,新开发了变密度地下水多目标模拟优化程序MOSWTGA。将MOSWTGA应用于求解大连周水子地区以控制抽水井所在含水层不发生海水入侵为约束的地下水开采多目标优化管理模型,得到地下水最大开采量与海水入侵面积之间一系列Pareto近似最优解。研究成果不仅为实行合理的地下水资源配置提供了科学的实用模型,同时也为解决多个优化目标下的变密度地下水优化管理问题提供高效可靠的模拟优化工具,具有重要的潜在环境经济效益。     

三峡梯级枢纽多目标生态优化调度模型及其求解方法

针对三峡梯级枢纽综合效益的充分发挥及其对长江流域典型生态系统修复及持续改善的科学需求,通过分析发电效益与生态效益之间的制约竞争关系,以发电量最大和生态缺水量最小为目标建立了梯级电站多目标生态优化调度模型,对三峡梯级枢纽多目标生态优化调度进行了研究。同时,针对传统优化方法难以同时处理多个调度目标的固有缺陷,提出一种改进多目标差分进化算法对所构建模型进行高效求解。该方法针对差分进化算法在多目标协同优化和全局寻优能力等方面的不足,依据问题的特点重新设计了差分进化算法的进化算子,同时设计了一种多目标混沌搜索策略以加强算法的局部搜索能力。最后,依据多目标生态优化调度问题的特点设计了一种不需要设置惩罚因子的约束处理方法。通过三峡梯级枢纽多目标生态优化调度的实例应用,验证了本文所构建模型的合理性以及所提出算法的有效性和工程实用性。