摘要： 工厂化南美白对虾养殖系统中投饲模型和管控系统脱节，尚未形成一个闭环的投喂决策控制体系，导致养殖生产一体化管控系统智能化水平不足等问题。开发了一套基于多源信息融合的对虾投饲动态决策系统。该系统通过集成水下相机与专用料台，构建对虾体长与残饵量同步监测系统；在此基础上，提出一种融合生长状态与实时残饲反馈的双因子投饲决策模型，赋予系统自适应优化能力；进一步设计以PLC和上位机为核心的控制执行单元，驱动高精度投饲机实现策略的可靠实施与数据管理。对比试验表明，该系统显著提高了投饲控制精度和饲料利用效率，虾的体质量增长率提高0.26%，饲料系数降低0.14%，存活率提高0.9%，为对虾养殖的精细化、智能化和可持续发展提供了可靠的技术方案。

关键词: 南美白对虾, 智能投饲, 水下识别测量, 动态决策模型

Abstract: The feeding model and control system of the industrialized South American white shrimp farming system are disconnected, and a closed-loop feeding decision control system has not yet been formed, resulting in insufficient intelligence level of the integrated management and control system for aquaculture production. Therefore, a shrimp feeding dynamic decision-making system based on multi-source information fusion has been developed. The system integrates underwater cameras and dedicated feeding platforms to construct a synchronized monitoring system for shrimp body length and residual bait amount; On this basis, a dual factor feeding decision model that integrates growth status and real-time residual feeding feedback is proposed, endowing the system with adaptive optimization capability; Further design a control execution unit with PLC and upper computer as the core to drive the high-precision feeding machine to achieve reliable implementation of strategies and data management. Comparative experiments have shown that the system significantly improves feeding control accuracy and feed utilization efficiency, increasing shrimp body mass growth rate by 0.26%, reducing feed coefficient by 0.14%, and increasing survival rate by 0.9%. This provides a reliable technical solution for the refinement, intelligence, and sustainable development of shrimp farming.

Key words: south American white shrimp, intelligent feeding, underwater identification and measurement, dynamic decision-making model