<p>This paper addresses the force and position control problem in constrained reconfigurable manipulators, explicitly incorporating both actuator and manipulator dynamics. While prior research has considered actuator dynamics in fixed or modular robots, this study presents the first unified control framework that integrates actuator-level electrical dynamics and manipulator mechanical dynamics for rigid-link electrically driven reconfigurable manipulators. This approach effectively manages challenges arising from dynamic reconfiguration, actuator uncertainties, and environmental interactions. By including actuator dynamics alongside manipulator dynamics, the proposed control strategy enhances system accuracy, stability, and disturbance rejection, accounting for practical issues such as delays and nonlinearities. Traditional model-based controllers struggle to handle the combined system uncertainties and dynamic variations inherent in reconfigurable manipulators. To overcome these challenges, a hybrid control approach combining model-based backstepping and an adaptive radial basis function neural network (RBFNN) is developed. The RBFNN approximates unknown electrical and mechanical dynamics in real time, while an adaptive compensator mitigates external disturbances, friction, and neural approximation errors. The proposed control strategy guarantees the asymptotic convergence of position tracking errors and constraint forces while regulating the direct current motors to deliver the required currents and torques. System stability is rigorously established using Lyapunov theory. Simulation results on a 2-DOF constrained reconfigurable manipulator, supported by quantitative analyses, demonstrate the superior effectiveness and robustness of the proposed hybrid neural network-based backstepping controller in handling actuator and manipulator uncertainties.</p>