Abstract: For modeling and solving problems of complex nonlinear systems whose input/output have uncertainty and are associated with time or process, a cloud process neural network model is built in the paper. It has uncertainty information processing ability by combining process neural networks processing ability for time-varying signal with cloud model transformation ability between qualitative and quantitative concepts. In addition, the cat swarm optimization algorithm is used to optimize the network structure and parameters simultaneously, and it helps to improve network approximation?and generalization performance. The effective extension of neural networks in time domain and uncertain information processing field is realized. Experimental results verify the effectiveness and feasibility of the model and algorithm.