基于v-支持向量机的事故工况下反应堆功率预测

针对事故工况下堆芯功率变化的特点和神经网络（ANNs）易陷极小值、收敛速度慢等问题，提出一种基于ν-SVR）的事故工况下堆芯功率预测方法。该方法运用k重交叉验证（k-CV）完成对ν-SVR预测器并将其用于弹棒事故（REA）和落棒事故（RDA）工况下的堆芯功率预测。研究表明，与ANNs相比，该方法具有更高的预测精度和更短的响应时间。 

Prediction of Reactor Power under Accident Conditions of Nuclear Power Plant Using ν-Support Vector Machine

Aiming at the characteristics of core power change under accident conditions and the problems of artificial neural network (ANNs) such as easy to trap minimum and slow convergence speed, a prediction method of core power under accident conditions based on? support vector regression (ν-SVR) was proposed. This method used a k-CV to optimize the parameters of ν-SVR, and then two different?ν-SVR predictors were designed. These two predictors were applied to the prediction of core power of rod ejection accident (REA) and rod drop accident (RDA). The results have shown that this method has higher prediction accuracy and shorter response time than the ANNs.