Virtual power plant（VPP）has the problem of generation and load uncertainty

which will lead to VPP facing risk loss while gaining profit. To solve this problem

VPP can be set up by selecting units with suitable adjustable capacity from the existing multisource units in dayahead

so as to effectively avoid the risk loss and maximize the revenue of VPP. In this study

conditionalvalueatrisk（CVaR）is used as the index of risk measurement

and the optimization goal is to maximize the operating revenue and minimize the risk loss. The dynamic aggregation optimization model of multi-source units is established based on CVaR risk control. Firstly

scenario technology is used to simulate the uncertainty of generation and load in VPP. Then

the influence of risk preference of VPP manager

environmental penalty cost and purchasing electricity price on the units selection of VPP is studied. Finally

a simulation example is given to verify the correctness of the proposed model. The result shows that the dynamic aggregation model can effectively reduce the risk loss of VPP and improve the power supply stability of VPP when the VPP manager chooses the appropriate risk preference.