An IPSO-WNN method for short term load prediction of integrated energy system is proposed to solve the problem that the conventional WNN model has the disadvantages of slow convergence speed and easy to fall into local optimum, which leads to the low prediction accuracy. Firstly, the Pearson coefficient is used to analyze the influencing factors and select the appropriate factors as the input quantity of the prediction. Secondly, the traditional particle swarm optimization algorithm is improved. Chaos algorithm is introduced in PSO and different particle inertia weight selection strategies are adopted according to particle fitness, and then IPSO-WNN prediction model was established based on the improved particle swarm algorithm (IPSO) to realize load prediction of comprehensive energy system. Finally, compared with the traditional WNN prediction model, the ipso-based WNN prediction model has improved the prediction time and accuracy.