为了对匹配决策问题进行建模与预测,提出了一种具有更多神经生理学特征的稀疏回声状态网络(ESN),并基于在线监督学习方法对网络进行训练.为了评估网络的匹配决策性能,设计了三组测试数据集对网络性能进行测试,并提出了一种基于网络期望输出与实际输出序列最大相关系数的评价方法.仿真结果表明,新模型只需要较少的训练时间即可获得较好的决策性能,且对发放时间间隔、平移和网络噪声具有较好的鲁棒性. In order to model and predict matching decision problems, a sparse echo state network (ESN) with more neurophysiological features is proposed and the network is trained based on the online supervised learning method.In order to evaluate the matching performance of the network, Three sets of test data sets are used to test the performance of the network, and an evaluation method based on the maximum correlation coefficient between the network expected output and the actual output sequence is proposed. The simulation results show that the new model can get better performance with less training time Decision-making performance, and the release of time intervals, translation and network noise has good robustness.