Multi-Objective Lower Irrigation Limit Simulation and Optimization Model for Lycium Barbarum Based on NSGA-III and ANN

Lycium barbarum has rich medicinal value and is an important economic tree species in China, with annual output of 21 billion RMB. The yield the quality dry fruit are crucial issues that affect cultivation income farmers Ningxia water shortage area. According to local acquisition standard barbarum, amount per 50 g (ADF-50) key factor evaluating determining purchase price. In order optimize irrigation lower limit automatic drip system multiple objectives, ADF-50 selected be optimal objectives. limits full flowering stage, summer fruiting early autumn stage optimized by third generation non-dominated sorting genetic algorithm (NSGA-III) this paper. mathematical relationships between quantity, amount, yield, were established balance model, production function (WPF), artificial neural network model (ANN), respectively. accuracy ANN verified experiments. experiments optimization results show that: (1) quantity calculated accurate, their Nash–Sutcliffe coefficient 0.83 0.66; (2) a certain range competitive relation. By solving NSGA-III schemes, which tend different compromise scheme can obtained; (3) Compared original water, scheme’s improved 10.7% 8.8% system’s improve not only but also barbarum. This provides new idea for establishing planting