ABSTRACT
In Chapter 8, a neuro-fuzzy model was presented for estimating the work zone capacity
taking into account seventeen different numeric and linguistic factors and data collection locality. A backpropagation neural network was employed to estimate the parameters
associated with the bell-shaped Gaussian membership functions used in the fuzzy
inference mechanism. An optimum generalization strategy was used in order to avoid
overgeneralization and achieve accurate results. In this chapter, the subtractive clustering approach described in Section 8.4.1 is
judiciously integrated with the radial basis function (RBF) and backpropagation (BP) neural network models to create the clustering-RBF and clustering-BF neural network
models (RBFNN and BFNN), respectively. The two clustering-neural network models
are developed for estimating the work zone capacity in a freeway work zone as a function of eighteen different factors: 1) percentage of trucks (jc j) , 2) pavement grade (vertical
slope in the longitudinal plane) (x2), 3) number of lanes (хз), 4) number of lane closures (x4), 5) lane width (x5), 6) work zone layout (x6), 7) work intensity (jc7), 8) work zone length (length of closure) (xg), 9) work zone speed (xç), 10) proximity of ramps (x10), 11) work zone location (rM), 12) work zone duration (x12), 13) work time (x13), 14) work day (хм), 15) weather conditions (xis), 16) pavement conditions (хіб), 17) driver composition (χ\η), and 18) data collection locality (xig). A detailed discussion of impact of these factors was presented in Section 8.2. The variables are quantified and normalized using the methods described in Section 8.3. Spline-based nonlinear functions are used to quantify each linguistic as well as binaiy-valued variable mathematically. Spline-based
nonlinear functions are also assigned to numeric variables in order to model the impact of their variations on the work zone capacity.
