Nuclear reactor dynamics are inherently nonlinear, complex, time-varying, and some parameters depend on the output power level. External disturbances and uncertainties caused by variations in neutronic and thermal–hydraulic parameters contribute to challenges in nuclear reactor control. Load-following operation is one of the crucial operations in nuclear reactors. Therefore, the improved load-following capability is one of the most important technical issues of a VVER-1000 PWR-type nuclear reactor due to safe operations. In this work, Takagi–Sugeno (T-S) fuzzy algorithm-based PID sliding mode self-tuned adaptive control (FPID-ASMC Self-Tuned) is designed to achieve a smooth reactor core power control during load-following operation problem for Russian VVER-1000 nuclear reactor. To resolve the purpose, a validated nonlinear two-point kinetic model with three groups of delayed neutrons and constant axial offset (AO) power-distribution strategy is considered. A nonlinear two-point kinetics model comprises the normalized neutronic model, the thermal–hydraulic model, the reactivity feedback model, and the poison effect based on the top and bottom zone of the reactor core. The impact of axial xenon oscillation is made in the oscillation of power distributions in long-time operations conditions. Hence, AO, normalized axial offset (∆I), and axial xenon oscillation (AXe) index have to be considered. The chattering phenomenon, commonly observed in conventional sliding mode control, is minimized by an adaptive self-tuned fuzzy PID controller with variable gains that depend on the reactor power level and error power relative to a defined reference value, achieving favorable reference power tracking performance and adaptive disturbance rejection capability. The proposed fuzzy PID sliding mode self-tuned adaptive controller is designed to ensure the stability of the overall dynamics during both the reaching and sliding phases. The analysis of stability is guaranteed in the sense of the Lyapunov stability theorem approach. To validate the proposed controller, simulations are performed under multiple transient conditions and found preferable in load-tracking applications and adaptive disturbance rejection capability compared to conventional sliding mode controller (CSMC), fuzzy control, and PID controller with fixed parameters with standalone configurations. Furthermore, the system outputs, ∆I, and AXe are in the acceptable range based on constant AO power-distribution strategy, and these indexes are robust to parametric uncertainties of the dynamics model of nuclear reactors.