2021 International Conference on Sensing, Measurement and Data Analytics in the era of Artificial Intelligence (ICSMD2021)

113 / 2021-07-31 23:39:10

A State Estimation Method for Nonlinear Systems Based on Uncertainty Fusion

Special Session > SS7 - Data-driven Engineering Practices in The Fields of Energy, Environment, and Biomedicine

Estimating the state of nonlinear systems is of great importance. But in practice, the state of the system cannot be measured and estimated directly, and there are many uncertain factors to improve the difficulty of state estimation. To solve the above problems, this paper proposes a fusion method of relevance vector machine and unscented particle filter to realize the state estimation of nonlinear systems on the basis of uncertainty fusion. The state estimation experiment was verified by the lithium-ion battery benchmark data set. Experimental results show that the proposed method can be used to estimate the state of nonlinear systems with small errors and to express the uncertainties, which proves the effectiveness of the proposed method. The proposed method can not only estimate the state of nonlinear systems, but also realize the adaptive expression of uncertainties due to the characteristics of uncertainty fusion.