MACHINE LEARNING AND ITS APPLICATION TO CIVIL ENGINEERING SYSTEMS

Today's state-of-the-art expert systems are plagued by four major problems: brittleness, lack of metaknowledge, knowledge acquisition, and validation. Knowledge acquisition by itself is a very time consuming and tedious process. The uncertainty of information and erroneous data have also caused knowledge engineers anxious moments. In order to address these problems, several machine learning techniques supported by well-formulated theories and algorithms, have been developed. In this article some of these techniques are reviewed along with examples of their application to civil engineering problems. The techniques presented either fall under the category “learning from examples” (commonly referred to as inductive learning) including the ID3 algorithm, the rough sets theory, and the PROTOS algorithm, or “learning from observations” (also known as conceptual clustering) including the COBWEB algorithm, or a combination of both.     

开发智能型计算机辅助教学课件的探讨

本文依据人工智能理论中专家系统的模型,遵循人对知识的认知过程,对开发智能型计算机辅助教学（ＩＣＡＩ）课件的特点、作用、结构及方式,并结合计算机文化基础课件实现智能化的原理进行了探讨,提出了ＩＣＡＩ课件的原理结构和学习结构模型,     

土木工程专家系统的应用和发展趋势

从专家系统的定义、组成、特点以及专家系统和一般计算机程序的区别、专家系统的开发与应用等方面介绍了专家系统的基本原理。阐述了专家系统在项目评估、诊断、决策和预测、建筑设计及优化、项目管理、施工技术以及一些特殊领域的应用。对土木工程领域专家系统的研究成果和应用情况进行总结和归纳,分析了现有土木工程专家系统存在的几个问题。在此基础上展望了专家系统在土木工程领域的应用和发展趋势,可以预见随着计算机应用的日益普及,专家系统在土木工程领域将有更为广阔的发展前景。     

Neural Networks in Structural Engineering

In the past few years literature on computational civil engineering has concentrated primarily on artificial intelligence (Al) applications involving expert system technology. This article discusses a different Al approach involving neural networks. Unlike their expert system counterparts, neural networks can be trained based on observed information. These systems exhibit a learning and memory capability similar to that of the human brain, a fact due to their simplified modeling of the brain's biological function. This article presents an introduction to neural network technology as it applies to structural engineering applications. Differing network types are discussed. A back-propagation learning algorithm is presented. The article concludes with a demonstration of the potential of the neural network approach. The demonstration involves three structural engineering problems. The first problem involves pattern recognition; the second, a simple concrete beam design; and the third, a rectangular plate analysis. The pattern recognition problem demonstrates a solution which would otherwise be difficult to code in a conventional program. The concrete beam problem indicates that typical design decisions can be made by neural networks. The last problem demonstrates that numerically complex solutions can be estimated almost instantaneously with a neural network.     

Log-Linear Models in Expert Systems

Abstract: The paper describes how log-linear models can be used to deal with uncertainty in expert systems, avoiding the common problems of many probability-based expert systems. After a general introduction to log-linear models, including hierarchical models, maximum likelihood estimation for poissonian and multinomial sampling is described and parametric and structural learning methods are illustrated by simple examples. Finally, a traffic engineering example is given.     

Incorporating Membership Functions and Certainty Factors in Fuzzy Expert Systems

Abstract: This paper demonstrates a mechanism for incorporating and independently evaluating membership functions and certainty factors during the preprocessing phrase in fuzzy expert systems. By providing this flexibility within the framework of a fuzzy expert system tool, the expert system becomes better equipped to model real-world control applications.     

Debugging Finite Element Programs Input Data with Machine Learning

Abstract: Debugging of the input data of a structural analysis program is a troublesome task which is heavily dependent on empirical knowledge. The paper describes an effort that applies machine learning to build an expert system for debugging faults in structural analysis program input data. ID3 decision tree induction algorithm is employed to build the automatic learning mechanism. This method Oflkrs an attractive potential for knowledge acquisition in the civil engineering domain in which expertise plays a dominant role. This paper is novel in at least three aspects: (1) machine learning is employed to build expert systems automatically; (2) a novel impurity function is proposed for splitting the decision tree; (3) a formulated comparison is proposed for evaluating learning results.     

The Use of Machine Learning for the Prediction of fire Resistance of Composite Shallow Floor Systems

Fire Technology - This paper is proposing an machine learning based expert system for preliminary prediction of the insulation fire resisting performance of shallow floor systems when subject to...     

CONFAULT-AN EXPERT SYSTEM FOR FAULT DIAGNOSIS IN REINFORCED CONCRETE STRUCTURES

Diagnostic applications are especially suitable for expert systems. The expert system CONFAULT diagnoses faults in reinforced concrete structures by identifying fault sub types. The knowledge base in CONFAULT is divided into modules corresponding to six major fault types, while meta rules are used to control and limit searching. A modified confidence factor approach is used to deal with uncertainty.     

Systems Theory for the civil engineer Experience with knowledge acquisition for an engineering expert system

Issues involved in the knowledge acquisition process for engineering applications of expert systems are discussed, with reference to a project underway in the reservoir management domain. Since there are no well established guidelines for the crucial task of acquiring human problem solving expertise, the individual responsible forthis task must rely on his own judgement and suggestions found in the literature. Based on the experience from this project, success in the area of knowledge acquisition appears to require the incremental development of very problem specific engineering expert systems, which facilitates the knowledge acquisition process. Other important factors described include the domain knowledge of the knowledge engineer (the person responsible for knowledge acquisition), the interaction between the expert and the knowledge engineer, and the expectations for the expert systems.     

Expert systems within the construction industry in the UK

The aim of this article is to establish a number of issues relevant to the applications of expert systems related to the UK construction industry. The following areas are discussed:

1. (1) Expert systems currently used in practical circumstances within the construction industry in the UK.

2. (2) The benefits of using expert systems in construction firms.

3. (3) The reasons affecting the acceptance of expert systems within the construction industry.

4. (4) The areas in which construction firms require expert systems development.

A questionnaire survey has been conducted involving a number of leading building companies in the UK industry to investigate to above points. A statistical analysis has been carried out and the results are presented and analyzed.     

Rough-Set-Based Knowledge Acquisition from Cases for Integrity Assessment of Bridge Structures

Many existing structures present an ever-increasing demand for maintenance and repair. It is necessary in the first place to diagnose such structures correctly. The diagnoses have been dependent to date on experts of great experience. To meet the increasing demand for maintenance and repair and perform diagnoses efficiently and correctly, the development of expert systems is hoped for. This article discusses application of rough-set theory to the acquirement of experiential knowledge contained in diagnostic cases. Rough-set theory was applied to cases in which experts diagnosed the damage of bridges, and a minimal-decision algorithm was derived that could make diagnoses equivalent to those of the experts. In addition, a method of deriving rules from the minimal-decision algorithm for the construction of expert systems is discussed.     

A Relational Database Interface for Design Expert Systems

Most of the software so far developed that is based on expert systems uses a rule-based representation of the domain knowledge. Expert System applications in the engineering design process, however, frequently require access to a large static knowledge base that is easily represented in a database. This article presents the details of a relational database module to represent such static knowledge that is callable as a function by a rule-based expert system. The software implemented on a microcomputer using a C compiler, can be accessed by any inference engine with external C function calling capability. It has been used in developing an expert system for the design of steel structural elements.     

Adaptive neuro-fuzzy based inferential sensor model for estimating the average air temperature in space heating systems

The heating systems are conventionally controlled by open-loop control systems because of the absence of practical methods for estimating average air temperature in the built environment. An inferential sensor model, based on adaptive neuro-fuzzy inference system modeling, for estimating the average air temperature in multi-zone space heating systems is developed. This modeling technique has the advantage of expert knowledge of fuzzy inference systems (FISs) and learning capability of artificial neural networks (ANNs). A hybrid learning algorithm, which combines the least-square method and the back-propagation algorithm, is used to identify the parameters of the network. This paper describes an adaptive network based inferential sensor that can be used to design closed-loop control for space heating systems. The research aims to improve the overall performance of heating systems, in terms of energy efficiency and thermal comfort. The average air temperature results estimated by using the developed model are strongly in agreement with the experimental results.     

建筑系本科五年级毕业设计课的教学方法及其理论基础

建筑系本科五年级毕业设计课为学生提供了自主导向综合实践的学习机会,进而为毕业生走上职业道路或者继续在研究生阶段的研究做准备。此类教学活动单依靠学生自身并不能完成,项目导师在提供框架限定研究问题、利用不同教学方法教学、规划组织教学活动等方面都起着重要作用。本文旨在为如何组织此类教学行为提供理论基础。首先,本文介绍了毕业设计课及其在职业建筑设计教育中的位置,而后论述了毕业设计课是如何通过强调体验式、情景式教学促使学生习得专业技能的。笔者提倡引用成人教育学理论中的教学策略作为完成教学目标的途径。最后,通过提供课程介绍和项目描述,本文建议了如何组织学生的学习过程并且描述了这一过程。结论是当学生:1)明确其学习内容、期限和目标,进而对自身的学习行为负责:2)给予其一套成熟的解决设计问题的框架;3)和充满热情的专业指导者共同进行实践时,即阿产出高水平的课程成果。     

Interpretable deep learning for roof fall hazard detection in underground mines

Roof falls due to geological conditions are major hazards in the mining industry, causing work time loss, injuries, and fatalities. There are roof fall problems caused by high horizontal stress in several large-opening limestone mines in the eastern and midwestern United States. The typical hazard management approach for this type of roof fall hazards relies heavily on visual inspections and expert knowledge. In this context, we proposed a deep learning system for detection of the roof fall hazards caused by high horizontal stress. We used images depicting hazardous and non-hazardous roof conditions to develop a convolutional neural network (CNN) for autonomous detection of hazardous roof conditions. To compensate for limited input data, we utilized a transfer learning approach. In the transfer learning approach, an already-trained network is used as a starting point for classification in a similar domain. Results show that this approach works well for classifying roof conditions as hazardous or safe, achieving a statistical accuracy of 86.4%. This result is also compared with a random forest classifier, and the deep learning approach is more successful at classification of roof conditions. However, accuracy alone is not enough to ensure a reliable hazard management system. System constraints and reliability are improved when the features used by the network are understood. Therefore, we used a deep learning interpretation technique called integrated gradients to identify the important geological features in each image for prediction. The analysis of integrated gradients shows that the system uses the same roof features as the experts do on roof fall hazards detection. The system developed in this paper demonstrates the potential of deep learning in geotechnical hazard management to complement human experts, and likely to become an essential part of autonomous operations in cases where hazard identification heavily depends on expert knowledge. Moreover, deep learning-based systems reduce expert exposure to hazardous conditions.     

黄河悬河稳定性评价专家系统的研制

从区域稳定性研究入手，以地壳、堤坝与地基三系统控制论为理论基础，系统总结了悬河研究领域中的最新成果和结论。提出了工程治理与环境治理，环境治理与经济发展，地质工程治理与水利工程治理相结合的治理策略，并用先进的人工智能技术，研制出黄河悬河稳定性评价专家系统，为黄河下游悬河分析、评价与处理以及进一步研究提供有效手段     

电梯群控技术的历史、现状和未来展望

介绍了电梯群控技术的发展历史和现状,着重分析了专家系统、模糊控制、神经网络、遗传算法、启发式搜索算法、强化学习、机器视觉等技术;列举了主要电梯企业电梯产品应用群控技术的状况;指出研究强化学习技术和机器视觉技术对电梯群控未来发展有重要意义。     

Learning heuristics for determining slurry wall panel lengths

Determining panel lengths for slurry walls is an engineering issue that involves complex geotechnical, design, and site considerations. In practice, the decision is made through a trial-and-error process. Relevant principles extracted from experts are not sufficiently detailed to generate a solution. This research proposes an inductive learning model for solving this problem. Given a new project whose panel lengths need to be determined, the model chooses similar cases from existing cases, based on case-based reasoning, performs an inductive learning, and uses correctness and coverage rates, and then static rules to verify the induced results.