To address the practical challenges of poor data quality, strong hyperparameter coupling, and significant peak positioning  errors  in  power  system  carbon  emission  peak  forecasting,  a  framework  integrating  robust  data  preprocessing with an improved grey-convolution hybrid model is proposed. Firstly, a dynamic quantile boundary-based outlier detection and multi-window weighted robust repair procedure is established, together with a random forest feature importance-based chained  multiple  imputation  method,  to  suppress  outlier  disturbances  and  high-dimensional  missingness  in  non-Gaussian data. Subsequently, an improved convolutional network incorporating variational mode decomposition, dilated convolution,and attention gating is constructed, with an embedded improved grey model for long-term trend extraction; hyperparameter optimization is achieved through a grey relational grade-guided whale optimization algorithm. Experimental results show that  compared  with  genetic  algorithm,  particle  swarm  optimization,  and  grey  wolf  optimization,  the  proposed  algorithm reduces mean absolute percentage error by 39.7%, 32.5%, and 25.4%, and peak prediction time deviation by 77.1%, 71.4%,and 60.0%, respectively. Compared with autoregressive integrated moving average (ARIMA), long short-term memory -prophet  (LSTM-Prophet),  time-series  transformer  (TST),  empirical  mode  decomposition-LSTM  (EMDE-LSTM),  and variational mode decomposition - gated recurrent unit (VMD-GRU), the proposed model reduces mean absolute percentage error to 2.89% and peak prediction time deviation to 0.7 h, while improving the inflection point capture rate to 93.8%. This study provides a new technical approach for accurate carbon emission peak prediction and offers data support for power system emission reduction strategy formulation.