In the reconstructed phase space, the equal-dimension and new information prediction model of coal washery daily water consumption time series was established based on LSSVM.

• 用微熵率法求得选煤厂日用水量时间序列的最佳嵌入维数和最佳延迟参数，重构相空间，建立了基于最小二乘支持向量机的选煤厂日用水量时间序列等维信息一步预侧模型。