Encrypting data based on the data attributes of robots is one of the effective methods to control access users in the data outsourcing environment. Therefore, a mathematical modeling method of robot data attribute encryption based on data redundancy elimination technology is proposed. The encryption algorithm structure is analyzed based on the Bloom filter. The Hamming distance is used to calculate the similarity of big data by the Bloom filter. Finally, a big data ellipse encryption algorithm is designed according to the calculation results. The results show that during the whole experiment, the approximate fluctuation range is 0.08%~0.14%, which is not only high but also has a large fluctuation range. In contrast, the probability of occurrence of redundant data is less than 0.05% under different byte rates of redundant data, which is far lower than the two traditional methods, indicating that the application performance of the proposed method is good.