This study develops a novel time-varying multivariable nonlinear grey model, namely TVNGM(1,N), which can capture the nonlinear and potential features of dynamic development trends. The novel multivariable nonlinear grey model has introduced a linear time-varying driving coefficient to replace the proposed model's constant parameter and added adjustment coefficient. The new model can be completely compatible with a single variable and multivariable grey models by adjusting different parameter values. For furtherly improving forecasting accuracy, the particle swarm optimization (PSO) algorithm is used to efficiently optimize the model’s parameters. Then, estimated parameters and the connotative prediction formula of the TVNGM(1,N) model are deduced by using the difference equation. To this end, two case studies are selected to prove the practicality of the method and compare it with other models. The results demonstrate that the proposed model has superior performance.