[关键词]
[摘要]
为更好地对汽轮机排汽比焓进行测量,将粒子群优化算法引入支持向量回归SVR模型中,构建与之相匹配的排汽比焓软测量预测模型。根据实例校验方法对该模型展开校验,采用汽轮机15种参数作为输入参数,排汽比焓作为输出参数。对某300 MW机组和200 MW机组数据进行仿真,并将该模型与标准SVR模型和双隐层RBF过程神经网络模型预测结果进行对比,对于300 MW机组,该模型的预测平均相对误差为0.101%,均方根误差为0.110%;对于200 MW机组,该模型的预测平均相对误差为0.057%,均方根误差为0.062%;与其他两种模型相比,PSOSVR模型的预测平均相对误差和均方根误差均最小。实例证明PSO-SVR的排汽比焓软测量预测模型在精确度以及泛化能力等方面呈现出一定的优势,具有较好的预测能力。
[Key word]
[Abstract]
In order to measure the specific enthalpy of steam turbine exhaust better,the particle swarm optimization algorithm is introduced into the support vector regression SVR model,and the prediction model of steam turbine exhaust specific enthalpy is built.According to the experimental verification method,the model is verified with 15 parameters of steam turbine as input parameters,and exhaust specific enthalpy as output parameters.The proposed model is used to simulate the 300 MW unit and 200 MW unit,and compare with standard SVR model and the dual process of hidden layer RBF neural network.For 300 MW unit,the model prediction has the average relative error of 0.101%,and the root mean square error of 0.110%.For 200 MW unit,the average relative error of the model is 0.057% and the mean square root error is 0.062%.Compared with the other two models,the PSO-SVR model has the smallest average relative error and root mean square error.These results show that the PSO-SVR model has certain advantages in accuracy and generalization ability,and good predictability.
[中图分类号]
TK262
[基金项目]
上海市“科技创新行动计划”高新技术领域项目(16111106300,17511109400)