Application of pain feedback and herding central point election algorithm in intelligent herding systems

As traditional animal husbandry advances towards digitalization, an intelligent herding system was investigated using pain feedback and herding central point election algorithm to achieve fine-grained monitoring and control．Core server and IoT terminal devices were connected to the same MQTT channel, adopted a unified command format for communication, and enabled visualization through data exchange at web frontend．Each animal was equipped with an IoT terminal device to continuously monitor MQTT channel commands relevant to its status．When an animal deviated beyond permitted radius, pain feedback algorithm would calculate shock intervals and pause intervals based on command data, yaw angle, and location information．Stimulus modules on IoT device were positioned in four directions, guiding animals to advance in predetermined directions by applying penalties at varying frequencies．Meanwhile, location data uploaded from IoT terminals to core server's cluster center point election algorithm through convolution calculations to calculate optimal herding central point, then encapsulated commands were sent to MQTT channel．Results demonstrated that this network architecture reduced coupling, pain feedback mechanism enabled fine-grained animal control, and herding central point election algorithm improved herding central point positioning accuracy．