Data Envelopment Method (DEA) appeared in the 1970s. It is simple and easy to handle multi-unit complex systems, attracting a large number of scholars to innovate on the basis of this method. 1978, Data Envelopment Analysis (DEA) was first put forward by three famous logistics scientists Charnes, Cooper and Rhodes, and it was called "one stage" or traditional DEA[ 16]. Later, Fried( 1999) and others pointed out that the traditional DEA model did not consider environmental factors, and discussed and proposed to eliminate the influence of environmental factors on evaluation, which was called "two stages". Finally, Fried(2002) and others considered the influence of environmental factors and random noise on the evaluation object, and eliminated the environmental factors and random noise that were not considered by traditional DEA, thus more accurately analyzing the internal management inefficiency of decision-making units. This process is called "three stages". This is the mainstream method, but now there are other types of DEA models, such as network DEA. The practical significance of this method is that it can be used to compare the efficiency of multiple service units providing similar services by explicitly considering the use of multiple inputs (resources) and the generation of multiple outputs (services).

DEA models mainly include CCR model with constant scale return and BCC model with variable scale return. This paper uses BCC model to analyze the efficiency of logistics industry in central China, and can calculate the efficiency value, input target value and relaxation variable value of logistics industry. The BCC model is shown in formula (2- 1).