基于神经网络的鱼类营养学专家系统的研制

介绍了采用人工神经网络、经典人工智能、多目标规划、数据库技术等多种先进技术相结合，进行鱼类营养学专家系统的研制的途径与方法。该系统是一个研究鱼类饲料优化配方及科学饲养的系统。其针对鱼类不同生长阶段，生长状态及周围环境，通过建立一个鱼类营养的动力学模型，给出一个准最优饲养方案，然后进行智能化饲料配方设计及营养疾病诊断；并通过研制专家系统的知识获取、推理机制、解释程序、启发式搜索及它们的协同工作，来模拟人类专家的思维过程进行优化配方和科学饲养方案的选择和制定。以便减少饲料投资、改善水质环境、提高收益及水产品质量。

STUDIES OF FISHERIES NUTRIOLOGY EXPERT SYSTEMS BASED ON NEURAL NETWORK

This paper introduces the method and procedure of studies on the FINES (Fisheries Nutriology Ex-pert Systems). Which use a classical artificial intelligence, artificial neural network, multiple objec-tives optimal, database technology etc. It study the fine farming way by building a fish farming dynamicmodel. It can provide a quasioptimal farming way as the factors of grow phrase, physiology condition andenvironment changed. It can also automatic design the fish compound feed and diagnose nutritive dis-ease. All this keep us invest little and receive more, improve the fish quality. By studying the ES'sknowledge acquire, inference engine, explain program, heuristic search and how they work together,we can simulate the expert's mind to make an optimal feed compound and select a scientific farming way.