In order to avoid deviating from actual conditions, low load rate and inefficiency in the design of traditional distributed CCHP system, this paper takes a business district as an object, transforms the optimization problem of equipment capacity of distributed energy system into a multi-objective mathematical model with the lowest comprehensive annual total cost and annual emissions. After comparing several common intelligent algorithms, it chooses a powerful and complete model. Differential Evolutionary Algorithms (DEA) with local patrol capability were used to solve the problem and obtain the optimal allocation scheme. There are many types of equipment in distributed energy system, and the influencing factors are complex. The capacity allocation of various equipment is the key to the operation efficiency of the whole system. The calculation results show that the distributed energy system optimized by differential evolution algorithm has remarkable advantages and feasibility compared with the cold, heat and power supply system. The system structure design and energy price will have an impact on the system optimization results.