Irrigation decision method, device, computer device and storage medium

Abstract

The application relates to an irrigation decision method, device, computer device and storage medium. The method comprises the following steps: acquiring an environment variable measured value of a target irrigation area and thus determining a target leaf blade water potential, a target leaf blade air hole conductivity and a target soil humidity; determining a reinforcement learning reward function and a target reward value according to the target leaf blade water potential, the target leaf blade air hole conductivity and the target soil humidity; obtaining an initial irrigation strategy aboutirrigation water flow and irrigation time based on an improved particle swarm optimization algorithm; acquiring an irrigation result for executing the initial irrigation strategy, evaluating the irrigation result according to the reinforcement learning reward function, and acquiring a reward value of the initial irrigation strategy; storing a reinforcement learning model meeting conditions according to relations between the reward value and the target reward value and training termination conditions of the reinforcement learning model; and determining an irrigation strategy related to the irrigation water flow and irrigation time according to the reinforcement learning model. The irrigation decision method can improve utilization ratio of irrigation information.

Description

technical field [0001] The present application relates to the field of agricultural irrigation, in particular to an irrigation decision-making method, device, computer equipment and storage medium. Background technique [0002] With the development of irrigation technology, the accuracy of irrigation decision-making is continuously improving, and the required data parameters are also increasing, such as historical accumulation data, field measured information (soil moisture information, crop physiological indicators, crop morphological indicators, etc.), weather forecast information, crop water demand information, agricultural remote sensing information from different scales and different regions, three-dimensional spatial data related to geographic location, and expert experience. Since each sensor (such as humidity, temperature, moisture, stem flow, etc.) has its own characteristics, usually a sensor can only reflect the suitability of crop moisture from a certain aspect o...

AI Technical Summary

Problems solved by technology

Since each sensor (such as humidity, temperature, water, stem flow, etc.) has its own characteristics, usually a sensor can only reflect the suitability of crop moisture from a certain aspect or within a certain range, and only obtain partial or One-sided information. At the same time, due to the influence of human operation, external interference and noise, if the information provided by a certain sensor or the experience of experts is used alone to make irrigation decisions, it may cause waste of irrigation information and reduce the utilization of irrigation information. Rate

Images