Research on Vegetable Greenhouse Strategy Based on Multi-objective Distributed Constraint Optimization
Meifeng Shi, Xin Liao, Hai Yang, Yuan Chen, Jun W. Wu, Shichuan Xiao · 2021
Intelligent controlled planting environment can maximize the benefits of vegetable planting. However, along with the increasingly demand for diversified vegetable cultivation and its intelligent management, it is difficult for the current planting environment for a single vegetable to achieve the maximize revenues of growing multiple vegetables simultaneously. To find the optimum environment factors for the growth of two kinds of vegetables, in this paper, a MO-DCOP model is constructed to describe the effects of environmental factors on vegetable growth in the diversified intelligent cultivation environment, in which the soil relative humidity, soil pH, temperature, carbon dioxide concentration and light intensity are modeled as the Agent respectively. Then, the excellent complete algorithm DPOP (Distributed Pseudotree Optimization Procedure) is used to solve the proposed MO-DCOP. Based on the extensive experimental results and empirical analysis, we can reasonably infer that the constructed MO-DCOP model is rigorous in theory and can realize the vegetable revenue maximization no matter what kind of mixed planting ratio.