Predicting continuous finger kinematics from surface electromyography (sEMG) signals provides crucial input for the intuitive proportional control of extreme upper-limb robotic prostheses. However, this regression task remains a complex challenge due to the noisy and non-linear dynamics of muscle activations. In this study, a hybrid deep spatio-temporal model is proposed that combines a Convolutional Neural Network (CNN) and a Long Short-Term Memory (LSTM) network to predict simultaneous finger joint angles from raw sEMG signals. Evaluated on 27 subjects from the Ninapro DB1 dataset using a rigorous repetition-wise split, the model was tested across three input window sizes (400ms, 600ms, and 800ms). Results demonstrate that the Hybrid CNN-LSTM consistently outperforms standard LSTM and Bidirectional LSTM (BiLSTM) baselines. The optimal 800ms window achieved the best overall performance (R2 = 0.7733, PCC =0.8841, RMSE =11.65°), demonstrating high-fidelity tracking particularly in the highly active digits. These findings highlight the feasibility of deploying efficient, single-modality deep learning estimators for robust, real-time prosthetic control without requiring complex multimodal sensor fusion.
Sukrit Ghosh, Yashkrit Singh, K. K. Sah et al.· International Conference on...· 0 citations
This paper presents a nonlinear model predictive control (NMPC) framework for quadrotor trajectory tracking under nonlinear dynamics and aerodynamic disturbances. Unlike conventional proportional–derivative (PD) controllers that require trajectory-dependent gain retuning, the proposed NMPC employs a single set of weighting matrices to track multiple three-dimensional trajectories. The proposed controller is evaluated on helical and Lissajous trajectories and compared against an optimized PD controller. Simulation results show that the NMPC framework achieves improved tracking accuracy, reduced overshoot, and smoother control effort across varying trajectory complexities. Quantitatively, NMPC reduces the Integral of Squared Error (ISE) by 53–99.6% and the Integral of Time Absolute Error (ITAE) by 82.5–98.5% compared to the optimized PD controller, while maintaining robust performance without retuning. These results demonstrate the suitability of fixed-tuning NMPC for agile quadrotor trajectory tracking.
N. Ahir, Jyotindra Narayan, G. Bhandari· International Conference on...· 0 citations
Traditional tea harvesting methods often lack precision, leading to significant leaf damage and reduced operational efficiency. To address these limitations, this paper proposes an airflow-driven end-effector designed for selective tea harvesting, utilizing pneumatic transport to move harvested leaves from the shearing interface to storage. A theoretical model based on force equilibrium was developed to determine the critical transport velocity (CTV), calculated to be approximately 3.6 m/s. The theoretical prediction was further evaluated through CFD analysis using ANSYS Fluent with a k–ω Shear Stress Transport (SST) turbulence model across inlet velocities ranging from 2.5 m/s to 4.5 m/s. Simulation results indicate that inlet velocities at or above 3.6 m/s provide airflow conditions favourable for sustained pneumatic leaf transport, with 4.5 m/s achieving outlet velocities up to 6.586 m/s and mass flow rates of approximately 1.59×10−3 kg/s, while lower velocities fail to maintain sufficient transport momentum. Despite considerable velocity decay along the transport path, adequate velocity is retained in critical regions for sustained pneumatic conveyance. These findings demonstrate the feasibility of airflow-based transport for selective harvesting and provide a quantitative basis for optimizing end-effector design.
Sharamjeet Shaurya, Arjun Agarwal, Dhirendra Kumar Verma et al.· International Conference on...· 0 citations