With the development of aquaculture, the demand for aquatic products in our country is increasing year by year, and the majority of aquaculture means is follow traditional artificial experience. Artificial feeding has some problems, such as inaccurate control of feeding quantity, low degree of automation and large labor cost. Therefore, a precise and intelligent feeding method is needed to improve the utilization rate of bait. The six-axis sensor is used to analyze the surface wave of the fish pond, and the feeding state function curve is obtained through the sensor. The principal component analysis is used to select representative characteristic parameters and compare the contribution rate of characteristic parameters at different intervals. The results showed that the analysis results with an interval of 60 seconds were more close to the actual situation. Six parameters with higher contribution rate were selected as indicators to determine whether to continue feeding. If two characteristic parameters were lower than the set threshold for three consecutive times, feeding was considered to be over. The feeding method based on the study on the feeding rules of fish schools makes the feeding process more in line with the feeding needs of fish schools. The utilization rate of bait reaches more than 85%, saves the breeding cost and reduces the pollution. The results show that this method can realize precise feeding of fish swarm, and provide reference for realizing automatic feeding of fish swarm.
