Optimization Of Nutrient Solution Formula Based On Integrated Water And Fertilizer In Plant Factory

A plant factory is a product of modern agricultural development and represents the most advanced and efficient emerging agricultural production method.The nutrient solution system is one of the main facility components within a plant factory,as it provides the water and fertilizer required by crops.The nutrient solution formula has a decisive influence on crop yield and quality.Through literature review,it was found that there are few studies on using machine learning to predict lettuce growth.The purpose of this project is to provide technical guidance for the development of hydroponic lettuce in a plant factory.This study focuses on hydroponic lettuce in a plant factory,exploring the effects of different nutrient solutions on yield,quality,and ecological characteristics at different time periods by changing the content of nitrogen species and the ratio of nitrogen,phosphorus,and potassium in the nutrient solution.By establishing an SSA-BP neural network prediction model and combining it with hydroponic experimental data,the nutrient solution of hydroponic lettuce is researched.Conducting experimental research under specific plant factory environmental conditions is of great significance for optimizing hydroponic lettuce technology,increasing yield,and improving planting efficiency.Based on Hoagland nutrient solution,this study used hydroponic cultivation,indoor experiments,and observation methods to study the effects of different nitrogen sources and different nitrogen-phosphorus-potassium ratios on the yield and quality of hydroponic lettuce.The SSA-BP neural network prediction model was established by using the sparrow search optimization algorithm to predict lettuce leaf length,plant height,and leaf number.Finally,the best nutrient solution formula was screened through a combination of experiments and the SSA-BP model.The main conclusions drawn from the experiments and data analysis are as follows:(1)Without changing the total nitrogen content in the nutrient solution,increasing the content of nitrate nitrogen in the nutrient solution(i.e.,replacing an equal amount of ammonium nitrogen in the Hoagland nutrient solution with1.75mg/L～5.25mg/L of nitrate nitrogen)is more conducive to promoting the growth of lettuce leaves,leaf number,and plant height compared to the control group(Hoagland nutrient solution).Additionally,increasing the nitrate nitrogen content in the nutrient solution can shorten the growth cycle of lettuce.The quality of lettuce leaf length and plant height in the N3 treatment group was significantly better than that in the CK treatment group,with an increase of 19.43%and 10.21%,respectively.In terms of yield,the N3 treatment group had a greater advantage,with a fresh weight and yield of 14266.18g/m~2and 17135.25g/m~2,respectively,which increased by50.716%compared to the CK treatment group.(2)Increasing the nitrate nitrogen content from 1.75mg/L to 7mg/L can enhance the content of vitamin C and soluble protein in lettuce to varying degrees,and reduce the risk of consuming lettuce with high nitrate levels.Increasing the nitrate nitrogen content while decreasing the ammonium nitrogen content can significantly improve the quality and yield of hydroponic lettuce,and increase the efficiency of hydroponic cultivation.However,excessive nitrate nitrogen content can also increase the nitrate levels in lettuce.(3)Appropriate increase of potassium content in the nutrient solution can promote leaf length,leaf number,and plant height of lettuce.However,when the nitrogen content in the nutrient solution is too high,increasing potassium content can have a suppressive effect on lettuce yield.When the nitrogen content is moderate,an increase in phosphorus and potassium content in the nutrient solution can promote leaf length,leaf number,and plant height of lettuce.The effect of phosphorus on lettuce yield is not obvious,and the effect of potassium on lettuce yield is greater than that of phosphorus.(4)In hydroponic experiments with different nitrogen-phosphorus-potassium ratios,the lettuce fresh weight was highest in the P2 treatment(9N:1P:8.8K),and all treatments were superior to the P3 treatment.The above-ground dry and fresh weights were in the order of P2>K2>P1>N3>K1>P3.Increasing the phosphorus and potassium content in the nutrient solution on the basis of the N3 treatment can increase the yield of lettuce.The P2 treatment showed a greater growth advantage with a weight of 128.461g/plant,which was 12.58%higher than the N3 treatment.(5)The SSA-BP neural network was used to fit the dataset of hydroponic lettuce leaf number,leaf length,and plant height,and the highest R~2value obtained was0.97113.This indicates that the model can predict the dynamic changes of hydroponic lettuce growth with good accuracy.By using this model to test hydroponic experiments in a plant factory,the accuracy of the predicted values compared to the actual growth values exceeded 83.31%,demonstrating that the SSA-BP neural network can be used to predict the nutrient solution formula for hydroponic lettuce and determine the optimal nutrient solution formula as formula 2.