Research On Dissolved Oxygen Intelligent Control System Of Recirculating Aquaculture

At present,the domestic circulating aquaculture model has been vigorously promoted and developed,and intelligent water quality control equipment also plays an important role in aquaculture production.Due to the high density of circulating aquaculture,pure oxygen is usually used for oxygenation.The traditional dissolved oxygen regulation method requires workers to control the aeration time and oxygen flow rate of the fish pond according to their own practical experience.This operation is greatly dependent on the workers’ breeding experience,which may cause waste of pure oxygen or insufficient oxygen supply,especially in high-temperature and high-density breeding.Under the working conditions,a little carelessness will lead to the disaster of the flood pool.Existing conventional dissolved oxygen control systems are oriented to blowers and waterwheels,and save electric energy by controlling the rotational speed,but the dissolved oxygen control system for pure oxygen aeration has not been reported yet.To realize the intelligent regulation under the condition of pure oxygen and oxygenation,improve the precise management ability of water quality,and enhance the risk prevention ability in the aquaculture process,the following researches are carried out in this paper:(1)To study the oxygen consumption rate of the red-finned puffer,mainly by controlling the water temperature and feeding rate,the oxygen consumption of the redfinned puffer was analyzed.The results showed that at 28°C,the oxygen consumption rate of the red-finned puffer was significantly higher than In the 20℃ group(P<0.05),under different feeding rates,there was no significant difference in the peak oxygen consumption rate of the red-finned pufferfish.The main difference was the time required for the oxygen consumption rate to recover before feeding after feeding.The recovery time at 20℃ was 2.5-7.5 hours at 28°C and 3-15 hours at 28°C.There is a tendency for the recovery time to increase significantly with the increase in water temperature and feeding rate.The anesthesia cannula feeding used in this experiment did not significantly affect the oxygen consumption rate of the fish.(2)Build a fish oxygen consumption rate prediction model.In this paper,three machine learning algorithms such as recurrent neural network(RNN),long short-term memory neural network(LSTM),and gated recurrent unit(GRU)are used to realize the prediction function of fish oxygen consumption rate.Build a network model,set the corresponding parameters in the neural network,filter and divide the data set,and standardize the content of the data set.The processed training samples are put into different neural network models for training and prediction.The prediction results of the three oxygen consumption rate prediction models were analyzed.The results showed that the root means n square error RMSE and average error MAE of the GRU prediction model were 3.66 and 2.693,respectively,and the root mean square error was reduced by22.13% compared with the LSTM and RNN prediction models.and 13.52%,the average error is reduced by 21.7% and 17.09% compared with the LSTM and RNN prediction models,and the R2 value is slightly better than the LSTM and RNN prediction models,so the GRU redfin puffer oxygen consumption rate prediction model can better predict the corresponding parameters Changes in oxygen consumption rate.(3)Design a dissolved oxygen intelligent control system with high safety and control precision.This paper designed a set of dissolved oxygen intelligent control systems,mainly designing the predictive control algorithm based on the GRU oxygen consumption rate prediction model,the closed-loop electronic flow regulator,the circuit and program design of the controller,and the design of safeguard measures.The control logic of the main program and subprograms,as well as the display and data recording methods,are described in detail.(4)Test and test the dissolved oxygen intelligent control system.Nitrogen simulated the feeding and oxygen consumption of 100kg/m3 redfin pufferfish,and compared with manual control to observe the control performance of three intelligent control algorithms,fuzzy control PID,neural network PID and model predictive control under the disturbance of fish pond feeding and the multi-objective value.tracking energy.The results show that the model predictive control has a high control accuracy,the dissolved oxygen fluctuation range is within ±0.1mg/L after the control to the target value,and it also has a good tracking ability under the multi-target tracking experiment.The dissolved oxygen intelligent control system for circulating aquaculture built in this paper realizes the active control of dissolved oxygen under the condition of pure oxygen aeration.The system has low cost,low control delay,and high control accuracy and reliability.The application of this system in recirculating aquaculture can improve the protection capability of dissolved oxygen,and can cope with the fluctuation of dissolved oxygen caused by feeding and temperature changes.