When the current method is used to identify lidar hardware failure data, the characteristics of Lidar echo data under different weather conditions are not analyzed and compared, which leads to the problems of high false positive rate, low recall rate and harmonic index. A method for identifying lidar hardware fault data based on data mining technology is proposed. Firstly, the combined method of EMD and wavelet threshold is used to remove clutter in the lidar echo data, and then the denoised data is clustered by fuzzy C-means clustering. Moreover, its data characteristics are mined, and finally the characteristic probability distribution model of normal echo signal data and fault echo signal data under different weather influences is developed to complete the identification of fault data. Experimental results show that the proposed method can effectively reduce the misjudgment rate and improve the recall rate and the coordination index.