In the recent phase, product cost variance analysis has been widely used. Traditional product cost variance analysis is based on data fusion method, which is constrained by its own openness and self-organizing condition. In the process of product cost variance analysis interaction, it is very easy to cause a large amount of data, resulting in system crashes, influencing the effect of data analysis of product cost difference, which brings a great hidden danger to the final cost analysis. To this end, we proposed an analysis and calculation method of product difference data based on complex number privacy protection - CNPD (ComplexNumber-based PrivacyData Aggregation). The calculation method mainly used the complex algebraic expression of the product cost differential data to fuse by using the data hiding, the data information was sent and received by node to node in the process of product cost difference data interaction, which effectively prevented the cost difference data to be recovered. Simulation results show that the speed of product cost difference data analysis is greatly improved, and the accuracy and integrity of the data is guaranteed by using CNPD. The method can be widely used.
