Aiming at the problems of low efficiency and poor expansibility in the traditional logistics business data processing method with multi-source and heterogeneous structure, this paper studies a set of practical solutions. The acquisition system of the scheme uses the multi port adaptive method to collect, which solves the problem of low scalability of the traditional acquisition system. The acquisition and processing dual thread mode is adopted to improve the collection efficiency. This paper also studies a combination algorithm of multi-source heterogeneous data fusion. Through fuzzy reasoning algorithm, the two data fusion algorithms are combined skillfully to optimize and complement each other to improve the multi-source heterogeneous logistics business data fusion. The experimental results show that the acquisition system has high accuracy and efficiency, and the loss value of logistics business data fusion is less than 0.07.