<p>This paper proposes a radial basis function (RBF)-adaptive terminal sliding mode control (RBF-ATSMC) strategy with friction compensation. It is designed to address the degradation of force-tracking performance in electro-hydraulic load simulators (EHLSs). This degradation arises from lumped uncertainties, strong nonlinearities (such as nonlinear friction), and position coupling. To overcome the difficulty of accurately identifying LuGre model parameters, which limits precise friction compensation, an improved sparrow search algorithm (ISSA) is introduced. The ISSA incorporates a chaotic reverse elite-guided strategy (CREGS) to mitigate uneven initial population distribution and a dynamic layered strategy (DLS) to optimize the update mechanism. These enhancements decrease the maximum identification errors of the LuGre static and dynamic parameters from 2.23% and 2.64% using SSA to 0.24% and 0.02%, respectively. These improvements enable the reliable identification of friction parameters. Using the identified LuGre model, a friction-compensated RBF-ATSMC strategy is developed. Adaptive laws are integrated to accommodate time-varying parameters online, and an RBF neural network estimates and compensates for unknown lumped disturbances in real time. Furthermore, a nonsingular terminal sliding mode control (NTSMC) scheme is adopted to guarantee rapid response and finite-time convergence of tracking errors. The stability is rigorously proven using Lyapunov theory. Finally, the effectiveness and feasibility of both the proposed ISSA and the friction-compensated RBF-ATSMC strategy are validated through comparative simulations and experiments.</p>