This paper proposed an enhanced sliding mode observer strategy in the area of the SOC estimation of lithium-ion batteries, which solves the problem of high-frequency chattering in traditional observation methods effectively. In the presence of bounded uncertainties and external disturbances, the state equation matching the second-order RC equivalent circuit model was established, which is the basis of the algorithm. The design parameters are obtained by analyzing and fitting the curve of the hybrid pulse power characterization test(HPPC) with the least square method(LSM). The enhanced SMO algorithm is tested using the response of the simulation model under the complex charge-discharge conditions, which demonstrate the algorithm's effectiveness and robustness.