基于PSO-ELM的热力系统参数预测  

Prediction of thermal system parameters based on PSO-ELM hybrid algorithm

作　　者：赵丽娟 马良玉[1] 王晓霞[1] ZHAO Lijuan;MA Liangyu;WANG Xiaoxia(School of Control and Computer Engineering,North China Electric Power University,Baoding 071003,China)

机构地区：[1]华北电力大学控制与计算机工程学院,河北保定071003

出　　处：《电力科学与工程》2018年第3期70-74,共5页Electric Power Science and Engineering

基　　金：国家自然科学基金资助项目(61174111);河北省自然科学基金资助项目(F2016502093);中央高校基本科研业务费专项资金资助项目(2016MS119)

摘　　要：高精度的参数预测对热力系统变负荷过程中的实时监控和故障诊断有着至关重要的意义。极端学习机是一种单隐层前馈神经网络,具有泛化能力强,学习速度快等优点。但是,由于输入权值和偏差的随机性使得ELM需要较多的隐含层节点才能达到理想精度,为了提高极端学习机对热力系统参数的预测精度,利用粒子群算法优化极端学习机的输入权值矩阵和隐含层偏差,从而计算出输出权值矩阵。仿真实验结果表明,相比ELM算法,经过粒子群优化的极端学习机对热力系统参数的预测具有较高的精度。Parameter prediction with high precision is of great importance for real-time condition monitoring and fault diagnosis of the thermal system during variable load process.Extreme Learning Machine(ELM)is a feed-forward neural network algorithm with single hidden layer,which has good generalization ability and fast learning ability.However,since the input weights and deviations are randomly determined,ELM needs more hidden layer nodes to achieve the desired precision.To improve the prediction accuracy of the extreme learning machine for the thermal system,Particle Swarm Optimization(PSO)is used to optimize the input weight matrix and hidden layer deviation of extreme learning machines,and then the output weight matrix is calculated.The simulation results show that the proposed hybrid algorithm is more accurate and effective for the prediction of the thermal system parameters.

关 键 词：高压加热器 参数预测 极端学习机 粒子群算法 

分 类 号：TK264.1[动力工程及工程热物理—动力机械及工程]