The authors analyse the importance of different weld control parameters on the weld pool geometry of gas tungsten arc welding using an online feature selection technique that suggests weld voltage and vertex-angle pair as more important than the weld voltage and torch speed pair. Using the selected features multilayer perceptron and radial basis function networks are developed for prediction of bead width, penetration depth, and bead area. With cross-validation the authors have extensively studied the performance of composite models (one model for all outputs) and individual models (one model for each output). The individual models are found to work better than composite models. Usually, radial basis function networks are found to work better than the multilayer perceptron networks. To assess the influence of weld control parameters the authors have studied the performance of both networks using different combination of inputs. Overall, the performance of the proposed models is found to be quite satisfactory.
In this paper, the uncertainty property is represented by Z-number as the coefficients and variables of the fuzzy equation. This modification for the fuzzy equation is suitable for nonlinear system modeling with uncertain parameters. Here, we use fuzzy equations as the models for the uncertain nonlinear systems. The modeling of the uncertain nonlinear systems is to find the coefficients of the fuzzy equation. However, it is very difficult to obtain Z-number coefficients of the fuzzy equations. Taking into consideration the modeling case at par with uncertain nonlinear systems, the implementation of neural network technique is contributed in the complex way of dealing the appropriate coefficients of the fuzzy equations. We use the neural network method to approximate Z-number coefficients of the fuzzy equations.
The use of fuzzy rule based systems to model the relationship between weld control parameters and the weld bead geometry features is explored in this paper. The Takagi-Sugeno model with linear functions of the inputs is used as the rule consequents. Given some training data, the authors use exploratory data analysis to find an initial rule base. The system parameters, e.g. consequent parameters, are estimated using a mixture of least square error (LSE) method and gradient search. The system is tested on three datasets and the performance is found to be satisfactory compared to the multilayer perceptron (MLP) and radial basis function (RBF) neural networks based systems.
This paper provides an overview of building structure modeling and control under bidirectional seismic waves. It focuses on different types of bidirectional control devices, control strategies, and bidirectional sensors used in structural control systems. This paper also highlights the various issues like system identification techniques, the time-delay in the system, estimation of velocity and position from acceleration signals, and optimal placement of the sensors and control devices. The importance of control devices and its applications to minimize bidirectional vibrations has been illustrated. Finally, the applications of structural control systems in real buildings and their performance have been reviewed.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.