论文部分内容阅读
为了使电梯群控系统更好地跟踪电梯交通流的变化以提高群控系统的性能,提出了基于支持向量回归(Support Vector Regression,SVR)的电梯交通流预测方法。针对电梯交通流时间序列小样本的特性,考虑了电梯交通流的横向和纵向变化趋势,采用SVR算法建立了电梯交通流时间序列的预测模型。给出了预测的评价指标,研究了SVR模型中的参数对预测效果的影响,利用试验寻优的方法确定了SVR预测模型的最优参数。最后,与电梯交通流RBF神经网络预测模型进行了比较研究,分析了数据样本中波动较大部分的预测效果,结果表明SVR算法比RBF神经网络方法具有更好的预测性能、泛化能力和鲁棒性,实现了电梯交通流较好的拟合和预测。
In order to make the elevator group control system better track the change of the elevator traffic flow to improve the performance of the group control system, an elevator traffic flow prediction method based on Support Vector Regression (SVR) is proposed. Considering the characteristics of the small sample of the elevator traffic flow time series, taking into account the horizontal and vertical trend of the elevator traffic flow, SVR algorithm was used to establish the forecast model of the elevator traffic flow time series. The evaluation index of the prediction is given, the influence of the parameters in the SVR model on the prediction effect is studied, and the optimal parameters of the SVR prediction model are determined by using the experimental optimization method. Finally, compared with the RBF neural network prediction model of elevator traffic flow, the prediction results of the larger part of the data samples are analyzed. The results show that the SVR algorithm has better prediction performance than the RBF neural network method, Excellent performance of elevator traffic flow fitting and forecast.