Research Of Precision Fertigation Control System And Equipment Technology In Facility Agriculture

Agricultural modernization is the priority among priorities in the future development of Chinese agricultural. Modern agricultural technologies and equipments will be the important means to improve agricultural productivity and ensuring the sustainable development of agriculture. Under this background, focusing on the water-fertilizer production process of facility agriculture, arounding the intelligent management system and the intelligent equipment technology in the precise water and fertilizer controlling process, following aspects were researched in this paper.（1） A multi-variable fuzzy relation model between water-fertilizer management factors(W_q, F_q, F_feq, F_c, F_o) and crop growth indexes（PMI, PI, RGI, YQ） were established using fuzzy modeling theory. Model computing was carried out for mainly indexes of crop growth, including plant height Hp, leaf area index LAI, Leaf photosynthesis rate FG, Root nitrogen uptake RN. and yield Y. The coefficient of determination of model computing is respectively 0.9608, 0.9541, 0.9633, 0.9764, 0.9873, which means there is high relativity between the model computing and the actual measurement. Test results show that the proposed model is able to reflect relations between water-fertilizer management factors and crop growth indexes accurately.（2） Numerical simulation studies of transport characteristics of water and nutrients in soil under drip irrigation were carried out. HYDRUS-3D was applied to simulate:（1）change of soil moisture content and soil nutrient concentration with time and space.（2）influence on soil moisture content and soil nutrient concentration of different irrigation and fertilization system（fertilization frequency, irrigation intensity）. Simulation results show that:（1）Effects of irrigation and fertilization process on the water and nutrient distribution was decrease with increasing soil depth, among which 0²0cm soil layer was the most affected. Changes of soil nutrient concentration lag in that of soil moisture content.（2）After fertilization, there was maximum nutrient concentration in 10 cm soil layer, relatively uniform distribution in 30⁷0cm soil layer, but minimum nutrient concentration in less than 10 cm soil layer. Over time, the nutrient concentration distribution space moved downward overall, and the total concentration decreased, there was minimum nutrient concentration in 0²0cm soil layer.（3） When working hours are allowed, it would be benefit for the absorption of nutrients by the crop roots and improving fertilization efficiency, by using drippers with smaller flow, higher fertilization frequency, and lower level of fetiliazation concentration.（3） General theory and methodology of the hydraulic pipe network system design under fertigation was further improved based on the traditional theory of water saving irrigation, by considering the fertilization process. A cost-oriented hydraulic network optimization model with multi-constraints was established. In this model, the investment cost of construction and operation of hydraulic pipe network system was optimized object, while minimizing the total annual cost W of system was optimization goal. The system design indexs were described as constraint conditions, such as pipe diameter restriction, velocity constraint, pressure constraint, rotational irrigation group flow constraint. In order to find the optimal solution of this model, a simulated annealing genetic hybrid algorithm（SAGA） was proposed for improving the capability both of global search and local search. The method of penalty function was introduced to convert the multi constrainted optimization problem to an unconstrained optimization problem. In order to verify the effectiveness of the SAGA hybrid algorithm, a practical optimization project was introduced. The results show that the total annual cost of system was 174177 yuan by using traditional optimization method, while just 139735 yuan by using proposed SAGA hybrid algorithm. Cost savings was 19.77%.（4） A kind of platform architecture of intelligent fertigation control system was proposed by adopting hierarchical design principles. The whole platform was divided into 4 levels, including plant layer, control layer, local management layer, and remote decision layer, in which the data management and the access operation were core. The remote decision layer was a remote Web service platform based on models and knowledge bases. A distributed local control system based on CAN bus was constituted of local management layer, plant layer and control layer. The local monitoring station was a link of information interaction between remote and local end, through deploying an application with C/S and B/S hybrid architecture. A combination type automatic back-wash filter and a low cost open-tank intelligent fertigation equipment were development with modular design and embedded technology. The control task of automatic back-wash filter was realized by MCU based on Cortex-M3 and HMI. The multi-task management was realized by ARM and μC/OS –II, which improved the real time and stability of control system.（5） The structure model of venturi injector was established and then structure parameters were determined. In order to test the suction characteristics of venturi injector, experiments were designed to analyze the relations among venturi drive flow, suction flow and suction concentration under different inlet and outlet pressure. The test results show that proposed venturi injector can obtain more suction flow under smaller driving force, of which the optimal import and export pressure are 14 m and 3m respectively, the max suction flow and the max suction concentration are 430L/h and 41.2% respectively. The mechanism model of concentration control process in intelligent fertigation equipment was deduced, which has revealed the characteristics, including nonlinearity, pure hysteresis, time-varying, of the real water-fertilizer mixing process. Moreover, time varying of system was increased due to uncertainty of rotational irrigation schedule. The mechanism model of p H control process in intelligent fertigation equipment was deduced from a general model of p H neutralization process, which has also revealed the characteristics, including nonlinearity, pure hysteresis, time-varying, uncertainty, of p H control process.（6） A new kind of variable universe fuzzy control strategy（VFPI） was proposed for the concentration control process. In this algorithm, considering the influence of varied rotational irrigation schedule, a co-fuzzy controller with one-dimensional was introduced to regulate the reference value of integral coefficient of PI controller, through monitoring the system flow change on line. Several simulation tests and field experiments were carried out to verify the validity and advantage of proposed VFPI algorithm. The simulation results and experimental results demonstrate that the proposed VFPI control strategy has better adjustability than the traditional PI control method. When the recipe contents or rotational irrigation situations varied, the VFPI controller has smaller overshoot, better dynamic quality and reliability than the traditional PI controller. A control algorithm with cloud inference in the variable universe fuzzy-PI controller（CVFPI） was proposed for the p H control process. Several simulation tests and field experiments were carried out to verify the validity and advantage of proposed CVFPI algorithm. The simulation results and experimental results demonstrate that the proposed CVFPI control algorithm can meet the requirements of p H control process. It has better characteristics both in dynamic process and steady state. Compared to PI and VFPI, CVFPI has smaller overshoot, shorter stable time, and stronger ability of anti-interference.