Research And Development And Application Of Variable Control System For Liquid Fertilizer Spraying Based On Neural Network Algorithm

"one control,two reductions and three fundamentals" put forward the goal of reducing weight and increasing efficiency under the national conditions of modern agriculture in China.It is required to achieve accurate fertilization and improve the use of machinery in reducing the amount of agricultural chemical fertilizer and improving the use of machinery.The amount of chemical fertilizer in China is obviously higher than the world level(using chemical fertilizer 225 kg per unit area).The large amount of chemical fertilizer has played the purpose of increasing production,in order to further improve the utilization rate of chemical fertilizer and increase production and efficiency.The development of variable rate fertilization technology provides an efficient platform for the accurate application of fertilizer.The research on the technology of variable rate fertilization will be conducive to the green,efficient and sustainable development of agriculture in China in the future,and it is a key issue in the field of precision agriculture and even intelligent agriculture in China.This paper takes the liquid fertilizer spraying variable control system as the research object,and develops the liquid fertilizer spraying variable control system on the basis of neural network PID variable control algorithm research and variable control system lag correction method.Through the verification and debugging of the performance parameters and field effect of the system,the performance of the system is further optimized and improved,and a set of variable control system suitable for spraying liquid fertilizer in the field is formed.The main results are as follows:(1)The control algorithm is simulated by taking the opening transfer function of the electric proportional valve as the control object.the results show that the rise time,adjustment time,overshoot and peak value of the BP neural network are 0.0298,0.0059,2.8% and 0.0576 respectively,which are the best compared with the fuzzy control algorithm and the conventional control algorithm.combined with the response of the control algorithm to the interference signal,BP neural network PID control algorithm shows the fastest response speed,good stability,the highest control accuracy and strong adaptive adjustment ability.In the indoor parameter experimental verification,the linear correlation between the system pressure and the number of opening valves and the pressure and flow value under different system pressure shows a minimum of 0.92 and a maximum of 0.98.The results show that under the BP neural network PID control algorithm strategy,the system has strong control ability to the linear controllability and stability of the fluid.In the indoor accuracy test experiment,the lowest accuracy is 4.6% and the highest is 1%,and there is no significant difference in accuracy under each pressure.It shows that under the strategy of BP neural network PID control algorithm,the system realizes high precision control and stability of fluid and strong self-adaptive ability under different pressure.(2)By analyzing the factors affecting the operation of the variable control system,and establishing the lag model according to the main influencing factors,and based on the lag signal generated by the lag model with navigation to further ensure the accuracy of fertilization.In order to verify the lag effect,the experimental results show that when spraying fertilizer at different speeds,the error is as low as 4.1% and the highest is 3%.Compared with the indoor precision data in static environment,the field operation accuracy has not decreased,and the lag correction effect shows practicability and stability.(3)Through the demand analysis of the variable control system of liquid fertilizer spraying,the research and development of the variable control system needs to follow the actual requirements of high precision,stability,reliability and intelligence in order to better adapt to the field working environment.Through the design of the overall scheme of the variable control system for liquid fertilizer spraying,the hardware composition of the variable control system and the method of flow control are determined.Through the construction of the environment,the arrangement of the design ideas,the design of the program structure of the lower computer and the design of the interface of the upper computer,the embedded development of the variable control system is completed.Field experiments were carried out to verify the effect and application of the control system.The results showed that under foliar fertilizer spraying and conventional fertilization treatment,the yield of the former treatment was 4.26% higher than that of the second treatment.Compared with the conventional treatment,the yield of foliar spraying and slow-release fertilizer increased by 13.49% and 7.9%,respectively.Under the treatment of foliar fertilizer spraying and slow-release fertilizer,the yield was 13.49% and 7.9% higher than that of conventional fertilizer and 11.5% and 5.99% more than that of foliar fertilizer and conventional fertilizer.The treatment of foliar fertilizer spraying and slow-release fertilizer could increase the yield of cotton.Through the monitoring of the amount of fertilizer sprayed in the operation,the results show that the highest accuracy is 2.3% and the lowest is 4.6%.Compared with the outdoor lag correction verification test(the highest accuracy is 2.3%,the lowest is 4.1%),the accuracy is basically the same.In the field experiment,the liquid fertilizer spraying variable control system shows high control precision and meets the operation requirements of liquid fertilizer spraying in the field.