Fish feeding decision system based on fish swimming energy consumption analysis in recirculating aquaculture

Abstract

The invention discloses an intelligent fish feeding decision system based on the analysis of fish swimming energy consumption, the system includes a circulating water treatment system, a feeding machine, a circulating aquaculture pool, a PLC controller, and an LED supplementary light , high-definition camera, computer, and frequency conversion water pump; the system mainly uses machine vision and deep learning technology to analyze and evaluate the swimming energy consumption of recirculating aquaculture fish in real time, and uses the self-constructed "satiation degree-swimming energy consumption" model Predict the real-time feeding demand of farmed fish, and then realize the intelligent decision-making of feeding amount in circulating aquaculture. The system structure of the present invention is simple, and the method is simple and effective, which can greatly reduce the uncertainty brought about by relying on manual experience for decision-making of feeding amount, meet the energy supply required for the growth of aquaculture fish, and effectively improve the production efficiency of recirculating aquaculture , in line with the development needs of aquaculture welfare.

Description

technical field [0001] The invention relates to the field of analysis of fish swimming behavior and a decision-making method for feeding amount in a recirculating aquaculture system, in particular to an intelligent feeding decision system for recirculating aquaculture fish based on the analysis of energy consumption of fish swimming. Background technique [0002] Recirculating aquaculture is a type of aquaculture model that has flourished since the 21st century. It belongs to an intensive farming model with a high utilization rate of water resources. Compared with traditional aquaculture systems, this type of system can save 90%-99% of water resources, and Controllable production and environmental factors are also considered to be an inevitable development trend of fisheries in the future. In the recirculating aquaculture mode, feeding has a direct and crucial impact on the production efficiency of aquaculture. However, the efficient decision-making of feeding amount has alw...

AI Technical Summary

Problems solved by technology

[0003] The existing technical methods for predicting fish feeding demand are mainly divided into residual bait method and optical flow method. The former is a method of identifying the amount of residual bait on the water surface. On the one hand, this method will cause fish to eat too much feed and cause corresponding Health problems, on the other hand, the residual bait produced will lead to a decline in feed utilization and water quality deterioration; the latter is a method of identifying changes in water surface reflection caused by fish activities, and the accuracy of this method is limited by the number and size of fish Size, light environment and other factors are currently only suitable for aquaculture fish with a larger group size and body size

In addition, the above-mentioned technologies can only indirectly reflect the feeding needs of farmed fish through some external information. In fact, these external information need to be combined with artificial feeding experience to further predict the feeding needs of fish, and there is still a certain gap with the real-time feeding needs of fish.

Images