New data optimization framework for parameter estimation under uncertainties with application to lithium-ion battery

Data optimization, or optimal experiment design, is an effective way to improve and guarantee the accuracy of state and parameter estimation, as the quality of data has significant impact on the estimation accuracy. Such capability is especially critical for energy systems requiring high reliability. The common practice of data optimization is to design input excitation by maximizing the Fisher information, and hence minimizes the variance of the estimation error. However, such approach suffers from fundamental limitations, including negligence of estimation bias and system uncertainties in measurement, model, and parameter, which severely restrict the applicability and effectiveness of the method. This paper aims at establishing new criteria and a novel framework for data optimization and estimation error quantification to overcome the fundamental limitations. First, a generic formula is derived for quantifying the estimation error subject to sensor, model, and parameter uncertainties for the commonly used least-squares algorithm. Based on the formula, desirable data structures, which could minimize the errors caused by each uncertainty, are identified. These structures are then used as new criteria to formulate the novel data optimization framework. The proposed methodology is applied to the parameter estimation problem of a lithium-ion battery electrochemical model in simulation and experiments, showing up to two orders of magnitude improvement in estimation accuracy compared with the traditional Fisher-information-based approach and other baselines.