Study On Thermal Environment Control Strategy And Heating Technology Of Large Scale Aquaculture Greenhouse

The temperature plays an important role in the growth of fish. Suitable temperature has important influence on feeding and immune function. PID control strategy and position control are adopted in thermal environment regulation of traditional aquaculrure greenhouse. However, these two control strategies not only has high energy consumption, but the stability of thermal environment is also poor. It is not conductive to the growth of the fish. In this study, the control effect of BP neural network based on PID control strategy and the traditional PID control strategy are compared and analyzed, and a model predictive control-MPC is proposed to control the thermal environment of aquaculture greenhouse. The main results of this paper are as follows:(1) Based on the thermal physical model of aquaculture greenhouse, the PID control strategy based on BP neural network and the traditional PID control strategy are compared. Research object is an aquaculture greenhouse located in Hangzhou. Results show that the overshoot of PID control strategy based on BP neural network is 18%, which does not cause large fluctuations in control system, and the overshoot of conventional PID is 80%, which cause the system fluctuation more severe. Transition time of PID control strategy based on BP neural network is 400s, and the traditional PID control strategy is 1000s. Former control strategy can achieve the steady state more quickly. PID control strategy based on BP neural network is significantly stronger than traditional PID control strategy.(2) In order to evaluate the effect of MPC control strategy on thermal environment of aquaculture greenhouse, the control effect of MPC on greenhouse is compared with traditional PID control and position control. The research object is a greenhouse in Hangzhou and the temperature data is from a typical year of Hangzhou. Results show that the stability of model predictive control is much higher than that of traditional PID control and position control, and the water temperature variance of each month is smaller than that of traditional PID control and position control. Model predictive control has less energy consumption and more stable. From April to November, the total energy consumption of model predictive control is 3167831 MJ, that of the traditional PID control is 3361020 MJ, and that of the position control is 3742200 MJ. Compared to PID control, the total energy consumption of MPC decreased by 5.75%, of which the highest saved energy in April is 16.69% and the lowest saved energy in January is 1.07%. Compared to position control, the total energy consumption of PID decreased by 10.19%, of which the highest saved energy in April is 23.9% and the lowest saved energy in January is 1.14%. To greenhouse air temperature, the control effect of the three control strategies is close.(3) A greenhouse in Hangzhou is treated as an object. The boiler design capacity with model predictive control is 0.72 MW, that with PID control is 1.45 MW and that with position control is 1.82 MW. In actual operation, the boiler efficiency of model predictive control is more close to rated power, and efficiency of PID control strategy is lower. In the whole heating period, the energy consumption of model predictive control is far less than traditional PID control, energy saving up to 9.91% and the energy consumption of PID is far less than position control, energy saving up to 3.28%. Using anthracite, diesel, natural gas, coal gas and electricity as raw materials, analyze and compare the economics of model predictive control, traditional PID control and position control. Results show that the energy cost of model predictive control is significantly lower than PID control and the energy cost of PID control is significantly lower than position control. To burn anthracite as an example, from November to April, compared to PID control, the total energy savings of model predictive control is 15.5 thousand RMB and reduce 30276 kg CO2,33 kg SO2, and 146 kg NOx respectively; compared to position control, the total energy savings of PID control is 5.3 thousand RMB and reduce 10355 kg CO2,11 kg SO2, and 50 kg NOx respectively.