Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Paul Takyi-Aninakwa has a total of 24 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Dynamic Multi-Adaptive Model with Temperature Interoperability Ensures Accurate Lithium-Ion Battery State of Charge Estimation

    by Paul Takyi-Aninakwa, Jude Dontoh

    Accurate state of charge (SOC) estimation is crucial for optimizing the performance and lifespan of lithium-ion batteries, particularly under dynamic operating conditions. This work proposes a novel multi-fusion approach combining a dynamic long short-term memory (DLSTM) model with an adaptive square root unscented Kalman filter (ASRUKF) to enhance SOC estimation accuracy. The DLSTM is equipped with an attention mechanism to capture complex temporal dependencies in battery data more efficiently, while the ASRUKF method provides robust state estimation by handling nonlinearities and noisy measurements, thereby improving the statistical value of the state estimate, error covariance, and measurement noise covariance matrices. The synergistic model integrates the strengths of both methods, offering improved robustness and adaptability across various operating conditions. Experimental results demonstrate the effectiveness of this multi-fusion approach in achieving precise SOC estimation under dynamic stress tests and different environmental conditions, showing minimum mean absolute error and root mean square error values of 0.00694% and 0.01614%, respectively, embracing the dynamics of different battery chemistries and capacities. The proposed model outperforms traditional estimation methods by mitigating cumulative errors and improving real-time performance, making it suitable for advanced battery management systems in electric vehicles and other applications.


  • Deep learning framework designed for high-performance lithium-ion batteries state monitoring

    by Paul Takyi-Aninakwa, Shunli Wang, Guangchen Liu, Carlos Fernandez, Wenbin Kang, Yingze Song


  • Open Access

    Enhanced Performance with Nano-Porous Silicon in TiFeSi2/C Composite Anode for Lithium-Ion Batteries

    by Alhamdu Nuhu Bage, Olusola Bamisile, Humphrey Adun, Paul Takyi-Aninakwa, Destina Godwin Ekekeh, Qingsong Howard Tu

    The innovative design of the microstructure of silicon-based composite anodes in Li-ion batteries holds great potential for overcoming inherent limitations, such as the significant volume change experienced by silicon particles. In this study, TiFeSi2/C composites prepared using micro, nano, and porous silicon showed reversible capacities of 990.45 mAh.g−1, 1137.69 mAh.g−1, and 1045.43 mAh.g−1 at C/10. The results obtained from the electrochemical characterization show that the porous structure of the composite anode material created via acid etching reduced silicon expansion during the lithiation/delithiation processes. The void spaces formed in the inner structure of the porous silicon and the presence of carbon increased the electronic conductivity between the silicon particles and, on the other hand, lowered the overall diffusion distance of Li+. This study confirms that TiFeSi2/C prepared with porous silicon dispersed in a transition metal matrix delivers better electrochemical performance compared to micro and nano silicon with a retention of 80.16%.


  • Open Access

    A novel long short-term memory-adaptive feedback-correction gain extended Kalman filter for the high-precision state-of-charge estimation of lithium-ion batteries

    by Emmanuel Appiah, Shunli Wang, Paul Takyi-Aninakwa, Etse Dablu Bobobee, Ebenezer Koukoyi

    Assessing the state of charge (SOC) is essential in guaranteeing the precise and effective use of lithium-ion batteries in electric vehicles and smart devices. For these batteries to continue to be dependable, safe to use, and have an appropriate service life in a variety of applications, such as electric vehicles and portable electronics, accurate SOC estimation by the battery management system (BMS) is essential. To examine the effects of training and testing variables on SOC estimate accuracy, this study makes use of transfer learning in a long short-term memory (LSTM) network. It also focuses on applying an adaptive feedback correction-gain extended Kalman filter (AFGEKF) and an EKF, using independently provided operational data and LSTM-estimated SOCs for performance optimization. Through iteration, this method improves denoising and SOC accuracy in a range of working conditions. The comprehensive results show that the ideal mean absolute error, mean squared error, and mean absolute percentage error are 0.4544%, 0.7326%, and 0.9371% for the LSTM model; 0.3069%, 0.4093%, and 0.3577% for the LSTM-EKF model; and 0.14687%, 0.3169%, and 0.2492% for the proposed LSTM-AFGEKF model at 0°C, 25°C, and 45°C using a ternary battery. The study shows that the training and testing hyperparameters of LSTM have a substantial impact on the accuracy of SOC estimations. Furthermore, the proposed LSTM-AFGEKF model’s capacity to provide precise SOC estimations serves as a good model with high computational efficiency.

  • Open Access

    A Critical Review of Improved Deep Convolutional Neural Network for Multi-Timescale State Prediction of Lithium-Ion Batteries

    by Shunli Wang, Pu Ren, Paul Takyi-Aninakwa, Siyu Jin, Carlos Fernandez

    Lithium-ion batteries are widely used as effective energy storage and have become the main component of power supply systems. Accurate battery state prediction is key to ensuring reliability and has significant guidance for optimizing the performance of battery power systems and replacement. Due to the complex and dynamic operations of lithium-ion batteries, the state parameters change with either the working condition or the aging process. The accuracy of online state prediction is difficult to improve, which is an urgent issue that needs to be solved to ensure a reliable and safe power supply. Currently, with the emergence of artificial intelligence (AI), battery state prediction methods based on data-driven methods have high precision and robustness to improve state prediction accuracy. The demanding characteristics of test time are reduced, and this has become the research focus in the related fields. Therefore, the convolutional neural network (CNN) was improved in the data modeling process to establish a deep convolutional neural network ensemble transfer learning (DCNN-ETL) method, which plays a significant role in battery state prediction. This paper reviews and compares several mathematical DCNN models. The key features are identified on the basis of the modeling capability for the state prediction. Then, the prediction methods are classified on the basis of the identified features. In the process of deep learning (DL) calculation, specific criteria for evaluating different modeling accuracy levels are defined. The identified features of the state prediction model are taken advantage of to give relevant conclusions and suggestions. The DCNN-ETL method is selected to realize the reliable state prediction of lithium-ion batteries.


  • A strong tracking adaptive fading‐extended Kalman filter for the state of charge estimation of lithium‐ion batteries

    by Paul Takyi‐Aninakwa, Shunli Wang, Hongying Zhang, Emmanuel Appiah, Etse Dablu Bobobee, Carlos Fernandez

    Summary: Lithium‐ion batteries are widely used as rechargeable energy and power storage system in smart devices and electric vehicles because of their high specific energy, high power densities, etc. The state of charge (SOC) serves as a vital feature that is monitored by the battery management system to optimize the performance, safety, and lifespan of lithium‐ion batteries. In this paper, a strong tracking adaptive fading‐extended Kalman filter (STAF‐EKF) based on the second‐order resistor–capacitor equivalent circuit model (2RC‐ECM) is proposed for accurate SOC estimation of lithium‐ion batteries under different working conditions and ambient temperatures. The characteristic parameters of the established 2RC‐ECM for the lithium‐ion battery are identified offline using the least‐squares curve fitting method with an average R‐squared value of 0.99881. Experimental data from the hybrid pulse power characterization (HPPC) is used for the estimation and verification of the proposed STAF‐EKF method under the complex Beijing bus dynamic stress test (BBDST) and the dynamic stress test (DST) working conditions at varying ambient temperatures. The results show that the established 2RC‐ECM tracks the actual voltage of the battery with a maximum error of 28.44 mV under the BBDST working condition. For the SOC estimation, the results show that the proposed STAF‐EKF has a maximum mean absolute error (MAE) and root mean square error (RMSE) values of 1.7159% and 1.8507%, while the EKF has 6.7358% and 7.2564%, respectively, at an ambient temperature of −10°C under the BBDST working condition. The proposed STAF‐EKF delivers optimal performance improvement compared to the EKF under different working conditions and ambient temperatures, serving as a basis for an accurate and robust SOC estimation method with quick convergence for the real‐time applications of lithium‐ion batteries.