To overcome the low precision of the basic grey model (GM) in forecasting power loads of medium and longterm, a co-evolutionary particle swarm optimization (CPSO) Grey Model (CPSO-GM) is proposed in this paper. This is done by employing the CPSO to optimize the parameters of the grey model based on the modified formula of the background value. They conduct the simulation experiments on the power load data of medium and long-term by applying the CPSO-GM. The experimental results show that the proposed algorithm is superior to the three different grey prediction models and better to forecast pow...