基于LSTM和CatBoost组合模型的短期负荷预测

党存禄, 杨海兰, 武文成

Short-term Load Forecasting Based on LSTM and CatBoost Combined Model

DANG Cunlu, YANG Hailan, WU Wencheng

表1 Ordered boosting 伪代码

Ordered boosting 输入：$\left\{ {\left( {{x_k},{y_k}} \right)} \right\}_{k = 1}^n,I$ $\sigma \leftarrow randompermutation{\text{ }}of{\text{ }}\left[ {1{\text{,}}\;n} \right];$ ${M_i} \leftarrow 0{\text{ for i = }}1, \cdots,{\text{n}};$ $for{\text{ t}} \leftarrow 1{\text{ }}to{\text{ I }}do$ $for{\text{ i}} \leftarrow 1{\text{ }}to{\text{ n }}do$ ${{\text{r}}_{\text{i}}} \leftarrow {{\text{y}}_{\text{i}}} - {M_{\sigma (i) - 1}}({x_i});$ $end for $ $for{\text{ i}} \leftarrow 1{\text{ }}to{\text{ n }}do$ $\eqalign{ & \Delta M \leftarrow LearnModel{\text{ }}(({x_j},{r_j}):\sigma (j) \leqslant i); \cr & {M_i} \leftarrow {M_i} + \Delta M; \cr} $ $end for $ $end for $ 输出：${{\text{M}}_{\text{n}}}$