Khajuria, A (1994) Quality assurance of concrete using a hybrid knowledge-based expert system. PhD thesis, Rutgers State University of New Jersey, School of Graduate Studies, USA.
Abstract
Strength of concrete at 28 days is universally chosen as the single most important criterion for its acceptance. If this strength is found to be less than the required strength, the concrete in-place has to be either removed or retrofitted. In order to avoid such a situation, construction specifications are written. These specifications typically include various quality control tests to be conducted and an allowable range of values for their results. A quality assurance program based on this methodology suffers from some serious drawbacks. For example, a batch of concrete received at the construction site is disqualified for use if it fails to meet the allowable limits given in the specifications. For a given construction project, if this problem is encountered quite often, it can result in construction delays with a subsequent significant loss of time and money. To overcome this problem, a hybrid knowledge-based expert system has been developed. The developed system named CONCEX (CONCrete EXpert) basically consists of three main modules. The first module helps in a basic review of the concrete mix. The second module helps in finding the reasons responsible for causing a given variability in a quality control test. This module was developed using a shell called RuleMaster2$\sp{\rm TM}. $ The third module helps in the estimation of 28-day concrete strength. Models were developed using the method of regression analysis and artificial neural networks. NeuroShell2$\sp{\rm TM}$ was used for building artificial neural networks. It has been observed that a hybrid knowledge-based expert system can be successfully used for assuring the quality of concrete. Knowledge representation in the form of example tables is easy and well suited to the problem of quality assurance. The accuracy of estimation of concrete strength improves as the number of quality control tests are increased in the models. Models developed using backpropagation neural networks gave better estimates of concrete strength. It is also possible to develop global models using artificial neural networks. When models have to be developed using incomplete data sets, the missing values of the predictor variables should be replaced by an average of the minimum and the maximum value specified in the range of values for that particular variable. Models with 9 predictor variables developed using this technique gave reasonable estimates of 28-day strength.
| Item Type: | Thesis (Doctoral) |
|---|---|
| Thesis advisor: | Williams, T P and Balaguru, P |
| Uncontrolled Keywords: | accuracy; construction site; expert system; quality assurance; quality control; regression analysis |
| Index terms: | quality assurance, accuracy, artificial neural network, neural network, construction site, specification, expert system, quality assurance program, methodology, construction delay, construction project, estimation, variability, regression analysis, estimate, quality control, module |
| Subjects: | project delivery, financial and cost management, artificial intelligence, work location, modelling and simulation, professional development, contractual condition, statistical analysis, project controls, data management, production management, research methods, architectural elements, quality assurance |
| Topics: | Quality Management, Project Management, Contract Administration, Time Control, Site Management, Design Practice, Digital Applications, Research Practice, Information Management, Cost Management |
| Descriptive scope: | 4 PCTA |
N.B. Descriptive scope is a count of how many of the five facets of empirical research are indicated by the words used in title, abstract and keywords. It is not intended as a judgement on the research; merely a count of the kind of word we would expect to indicate Phenomenon, Concepts, Theoretical framing, Empirical techniques, Analytical techniques. If all five are present, then a code of “5 PCTEA” will indicate this. If you feel the coding for this record is questionable, we welcome discussion around the terms we matched or the way we categorized them. The facet you would expect may not be coded, or a facet may be coded inappropriately. This can also bear on a larger question, of which facets should be treated as defining in construction management research. Please get in touch, and we will look at it. More details here