摘要
在水质评价过程中,水环境质量等级之间存在非线性关系,这给水质评价工作带来难度。针对此,以某地地表水质作为研究对象,对单因子分析法、离散Hopfield神经网络模型和T-S模糊神经网络模型3种水质评价方法进行分析比较。仿真结果表明:模糊神经网络算法结合了神经网络系统和模糊系统的优势,集学习、联想、识别与信息处理于一体,评价结果稳定、合理,优于离散Hopfield神经网络模型,非常适合对于此类问题的研究;而单因子分析法存在一定的局限性。
In water quality assessment,nonlinear relationship between different water quality levels troubles the water quality evaluation.Taking the surface water from a place as the object of study,the single factor analysis method and discrete Hopfield neural network model and T-S fuzzy neural network algorithm for the water quality analysis were compared.The simulation results show that the fuzzy neural network algorithm which boasting of the superiority of both neural network system and fuzzy system in learning,association,recognition and information processing outperforms the discrete Hopfield neural network model in the water quality assessment,the assessment result from it is scientific and rational,but the single factor analysis method has some limitations.
出处
《化工自动化及仪表》
CAS
2012年第7期900-903,909,共5页
Control and Instruments in Chemical Industry
关键词
水质分析
水环境质量等级
非线性
模糊神经网络算法
单因子分析法
HOPFIELD
仿真
water quality analysis,water environment quality level,nonlinearity,fuzzy neural network algorithm,single factor analysis method,Hopfield,simulation
