基于BP神经网络的水产健康养殖专家系统设计与实现

Design and Implementation on Expert System of Health Aquiculture Based on BP Neural Networks

针对传统专家系统的知识获取困难、推理能力弱、智能水平低和实用性差等缺点,阐述了BP神经网络运用于水产健康养殖专家系统的设计思想,对水产养殖中的饲养、水环境调控、疾病诊断的模糊描述进行量化,从系统模型和实现流程上说明本专家系统的特点,并以水质评价子系统为例,对平台功能和性能进行测试.实验数据表明,误差小于1%.该平台克服了完全依靠专家经验的主观性,诊断效率高,具有较高的实用性、通用性和灵活性.

In allusion to the insufficiencies such as difficulties in knowledge acquisition,weak inference capability, low intelligence level and bad practicality of traditional expert systems, this paper introduces design concepts of BP neural network applies to expert system of health aquiculture, and quantifies the fuzzy descriptions, such as aquaculture, water environment control, disease diagnosis. The characteristics of the expert system are illuminated from the system model and the realization of that process, and water quality e- valuation system as an example for testing the functions and performance of the platform. Experimental data shows that the error rate is less than 1 % . The platform has been completely overcome the subjectivity of experience to rely on experts, which has the advantages of diagnostic efficiency, high practicality, versatility and flexibility.