论文部分内容阅读
根据电力市场的相关历史数据准确地预测出未来的市场出清电价,对于市场中的各个参与者都具有十分重要的意义.在建立了一种粒子群优化(PSO)下的BP神经网络电价短期预测模型的基础上,采用PSO进化算法,反复抽取训练子集样本,通过对应的验证样本预测误差寻找近似最有代表性的训练子集,解决了模型的训练样本参数难以设置的问题.实验验证了该预测模型的有效性,结果表明处理好预测模型样本参数的选择问题,能够提高模型的稳定性及预测精度.
It is very important for all participants in the market to predict the future market clearing price accurately based on the historical data of the power market.With the establishment of a BP neural network based on Particle Swarm Optimization (PSO), the short-term Based on the predictive model, PSO evolutionary algorithm is used to repeatedly extract the training subset samples and find the approximate representative subset of the training samples by the corresponding prediction error of the verification sample, which solves the problem that the training sample parameters of the model are difficult to set up. The effectiveness of the prediction model shows that the selection of the sample parameters of the prediction model can be handled well and the stability and prediction accuracy of the model can be improved.