| Tidal irrigation is one of the most ideal technologies in water saving irrigation in the world.Considering the introduction of this technology is short and soil humid sensor’s detecting accuracy required to be improved, this article is to foresee the water requirement via BP neutral networks per weather conditions based on the study of pakchoi grow seedlings, in order to optimize the control rule and better guide the seedling irrigation.This paper mainly focuses on the following several aspects.Through the study of pakchoi, I got the mathematical model between plug weight and substrate relative humidity via SPSS and raised the mathematical model between substrate relative humidity variation and evaporation and transpiration amount, while the substrate relative humidity is the important factor to tidal irrigation.After the correlation analysis via SPSS, I use the highly relative atmosphere temperature and humidity to represent the soil temperature, therefore the cost of water saving engineering can be reduced and it will be propitious to the promotion of tidal irrigation in seedling area.Due to the time lag and transmission distortion of common soil humidity sensors and the price for the high performance sensor is extremely expensive, I apply the BP neural network to establish the neural network model between the illumination, atmosphere temperature, humidity and substrate relative humidity. After the network model verification using the actual data, it is proven it can be used to anticipate the substrate relative humidity change rate.This model provides data for decision making of tidal irrigation of warm house grow seedling, it also provides theoretical support to promote the tidal irrigation industrialization.Research and study the automatic control solution of vegetable seedling tidal irrigation applying the forecasted soil relative humidity variation data from the BP neutral network to the PC and Mitsubishi PLC automatic control system. |
PDF Full Text Request