A novel hybrid algorithm based on ga and m-nn for an optimal design of a weather radar network

Air traffic management (ATM) relies on the reports provided by the meteorologists to circumvent the impending severe weather based on the weather radars (WRs). Therefore, we present in this study a hybrid approach for optimal deployment of a weather radar network (WRN) over Algiers Flight Information Region (FIR) airspace in view to achieve maximum radar coverage. Our proposed model, coined CM-NNGA, leverages a genetic algorithm (GA) to explore significant and complex spaces of possible solutions with a modified neural network (M-NN) method to escape local optimums. The digital elevation model (DEM) that represents the spatial solutions is discretized into a grid system, where each cell represents a feasible solution. The maximization of a WRN coverage is achieved by minimizing of partial beam blockage (PBB) given several constraints, namely the ground elevation, the radar beam elevation and the distance between any two pairs of radars. To reduce the computational time, skipping schemes are employed. The results show that the proposed method performs qualitatively superior to selected state-of-the-art algorithms, with a somewhat higher computational time consumption, especially for large areas.