The state of charge (SOC) is an important parameter that indicates the remaining capacity and performance of lithium-ion batteries in Electric Vehicles (EVs). Accurate and reliable SOC estimation can help optimize the Battery Management System (BMS) and extend the battery life. In this paper, a novel SOC estimation method that combines the Sigma-Point Kalman Filter (SPKF), the Enhanced Self-Correcting (ESC) model also known as the Plett model, and the bar-delta filter is presented to handle nonlinearities, adapt to changing conditions, and offer robust estimations even in the presence of uncertainties and noise, which might be limitations of some traditional optimization algorithms. The SPKF is a state estimation method that can handle nonlinear and noisy systems better than the Extended Kalman Filter (EKF), which relies on linearizing the models around the current state estimate. The Plett model is a battery model that incorporates the effects of capacity fading, temperature, and hysteresis on the battery dynamics. The bar-delta filter estimates the average SOC of the battery pack using the SPKF method. Through comprehensive simulations and comparative analyses, the effectiveness of the developed approach is demonstrated using both MATLAB programming and octave codes. The results show that the proposed method achieves higher accuracy and stability than the EKF method.