To achieve energy-saving, optimal dispatching of heat load in heating station with peak-shaving boiler is necessary. First, a dispatch model of integrated optimal energy-saving was built on the combination with the minimum heat energy consumption model and minimum operation cost model, this model can applicable to different heating mode. Then, a immune particle swarm algorithm(IPSO) was used to solve optimal dispatching problems, compared with PSO, this algorithm is not easy to fall into Premature and local extremum, the convergence speed is faster than PSO. An experiment results show the advantages of IPSO, optimal dispatching of heating station with peak-shaving boiler can achieve energy-saving.