摘要
神经网络的建模方法有助于实时优化调度吸收式制冷系统的运行过程。文章提出了一种简单粒子群算法(Simple Particle Swarm Optimization,SPSO)优化小波神经网络(Wavelet Neural Network,WNN)的建模方法,通过分析发生器的工作原理,建立了发生器模型的输入、输出结构;基于其内部复杂的传热、传质过程,选用单隐含层的小波神经网络作为模型的内部结构,并利用均方根误差和决定系数的试错方法确定网络模型隐含层节点数;利用Morlet小波基函数和简单粒子群算法,改进了传统神经网络的隐含层激活函数和神经网络的权值、伸缩因子和平移因子等参数的确定方法。结果表明:所建立的发生器SPSO-WNN模型比传统的WNN神经网络模型均方根误差分别降低了1.36%、1.06%;所研模型简单、有效,可作为吸收式制冷系统优化运行控制策略中的等式约束。
The modeling approach of neural network helps to optimize the operation process of dispatching absorption refrigeration system in real time.In this study,a modeling method based on simple particle swarm optimization and wavelet neural network(SPSO-WNN)is proposed.The input and output structure of the model is determined through the analysis of the working principle of the generator,while the single hidden layer WNN neural network is selected as the internal structure of the model according to the complex heat transfer and mass transfer process within the generator.The number of nodes in the hidden layer of the network model is determined by the trial-and-error method of root mean square error and the determination coefficient.The Morlet wavelet function and the simple particle swarm optimization algorithm are used to improve the activation function of the hidden layer of the traditional neural network and the method of determining the parameters such as the weights,stretch factors,translation factors of the neural network.The results indicate that the proposed generator model,SPSO-WNN,compared with the traditional WNN model,the root mean square error of the two outputs is reduced by 1.36%and 1.06%respectively.The proposed model is simple and efficient,and can be used as the equality constraint in the optimal operation control strategy of the absorption refrigeration system.
作者
丁绪东
孙梅
孙昊
马浩翔
Ding Xudong;SUN Mei;SUN Hao;MA Haoxiang(School of Information and Electrical Engineering,Shandong Jianzhu University,Jinan 250101,China;Shandong Key Laboratory of Intelligent Buildings Technology,Jinan 250101,China)
出处
《山东建筑大学学报》
2024年第3期65-75,共11页
Journal of Shandong Jianzhu University
基金
山东省重大科技创新工程项目(2019JZZY020812)
山东省自然科学基金面上项目(ZR2020MF070)。
关键词
吸收式制冷系统
发生器
WNN神经网络
简单粒子群算法
absorption refrigeration system
generator
WNN neural network
simple particle swarm algorithm
