Results in Physics (Jun 2024)
A quadratic programming optimization of field leveling for large-scale photovoltaic plant
Abstract
Field leveling is crucial for efficiency and performance optimization of concentrated photovoltaic (PV) plants due to their large areas and uneven terrain. Manual adjustment combined with the block plane method is conventionally used for PV field leveling design, which is time-consuming and laborious with non-optimized results. There’s a lack of optimization method for large-scale PV field leveling. This paper introduces a quadratic programming-based optimization approach, dividing the leveling plane into grids and utilizing a hyperbolic paraboloid to model the surface. Constraints are imposed based on earthwork balance and slope ratio, with the objective function being the minimization of total cut-and-fill volume. The globally optimal solution for large-scale PV field leveling is through MATLAB R2022a programming. Case studies demonstrate that the quadratic optimization method, compared to the traditional block plane method, can yield savings of 55 % to 78 % in cut-and-fill volumes, with faster computation and more practical outputs. Moreover, the method proves to be more effective as site area and topography complexity increase. Parametric analyses obtain a critical slope ratio of 2 %, below which the cut-and-fill volumes are significantly impacted by the slope ratio, while above which, the influence diminishes. Meanwhile, the cut-and-fill volumes are found to increase continuously as the block size increases, while remaining mostly constant when the grid size changes within a range of 5–30 m. These findings highlight the significant influence of design slope ratio and block size on the cut-and-fill volume, while the grid size has a minor impact. Recommended parameters for PV field leveling design include a design slope ratio of 3 % to 7 %, a grid size of 5 to 20 m, and a block size of 30 to 50 m. This proposed method provides an optimization method for field leveling design of large-scale PV plants. It achieves the least engineering quantity, substantially improves design efficiencies, and provides more applicable outputs.
