The heat source identification problem that aims to identify the spatial locations and strengths of point heat sources and know well about the heat source properties of actual engineering structure in time belongs to the field of Inverse Heat Conduction Problem (IHCP). The heat source identification problem is generally illposed

that is

the identified results are very sensitive to the measurement noise when the measured data is insufficient. In order to overcome the ill-posedness

additional constraints need to be introduced. In this paper

a novel point heat source identification approach based on sparse regularization is proposed where the sparsity of point heat sources in space is mainly considered. Due to the existence of measurement noise

a weak enforcement of measured data through a penalty term is introduced into the objective function. Moreover

to well corporate with the sparse regularization

the alternating minimization is used to iteratively solve the separated variables of temperature and heat source

and the threshold setting method is proposed to quickly and accurately find an appropriate regularization parameter. At last

a numerical example on a twodimensional steadystate case shows that the proposed approach can quickly and accurately identified both the locations and the strengths of heat source and is insensitive to measurement noise.