Development of the precise feeding system for pond culture

Feeding systems can dominate the primary efficiency and cost in aquaculture. The high management efficiency of feeding can also alleviate ecological environmental pollution. However, artificial identification has been widely used for fish feeding in the pond so far, leading to the time-consuming and laborious. Empirical feeding management cannot fully meet the large-scale pond aquaculture, in terms of production efficiency, environmental pressure, and aquaculture risk. In this study, a set of precision feeding systems was developed using fish feeding welfare. Four modules included water quality monitoring, decision control, drive execution, and remote monitoring. A systematic investigation was also implemented to clarify the effects of environmental factors on the physiology and energetics of fish. The affecting factors were then determined by the feeding demand of fish. The dissolved oxygen, temperature, and body growth were taken as the input parameters, whereas, the target demand was as the output parameters of the feeding system. The fuzzy PID control was combined with the particle swarm optimization (PSO) to realize the precision feeding operation. A comparative test of pond culture was carried out with the intensive grass carp as the research object, in order to verify the practicability and effectiveness of the system. A comprehensive evaluation was made from the aspects of regulation, growth performance, and economic and environmental benefits. The results showed that a stable and reliable control performance was achieved in the constructed precision feeding system, where the control error was less than 7%. The nash-sutcliffe (NS) index increased to 0.913 in the decision-making performance of the feeding system. whereas, the root mean squared error (RMSE) was reduced by 16.10, compared with the traditional. The feed conversion rate was significantly reduced by 11.73 % (P<0.05). Importantly, the aquaculture income increased by about 14 600 yuan/hm², whereas, the pollution was reduced by 241.40 kg per ton of fish produced. When the feed coefficient was 1.0, per ton of aquatic products brought about 500 kg of organic waste, 26 kg of nitrogen, and 13 kg of phosphorus. When the feed coefficient was 2.5, per ton of aquatic products produced 1 625 kg of organic waste, 117 kg of nitrogen, and 38 kg of phosphorus. By contrast, the grass carp produced per ton was equivalent to about 1 250 kg of organic waste, 86.70 kg of nitrogen, and 29.70 kg of phosphorus under the traditional feeding model. The total production of national grass carp in 2021 was 5.76×10⁶ t. If all grass carp farming was adopted as the precise feeding system, it was equivalent to reducing the pollution emissions of 1.17×10⁶ t organic waste, 9.43×10⁴ t nitrogen, and 2.59×10⁴ t phosphorus. The developed system can be expected to alleviate the environmental and ecological pollution caused by aquaculture. Strong comprehensive application performance can provide the theoretical reference and technical support for the research and development of other aquaculture modes and farmed fish. The feeding system should be further improved from three aspects in the future: monitoring quality and efficiency, decision-making model, and implementation cost. The sensor detection accuracy can be improved to optimize the processing efficiency of models. Systematic research of multiple feeding can be integrated using the logic line of “when to eat (feeding rhythm)-how much to eat (feeding demand characteristics)-how to eat”, in order to achieve an efficient and unified mode of breeding objects, feed, and machine.