<p>Control systems in many real-world applications are often affected by imprecision and subjective human judgment, making it difficult to construct accurate mathematical models or probabilistic descriptions of disturbances. Within uncertainty theory, uncertain inference controllers provide a rule-based control framework using uncertain sets and uncertain measures. Moreover, existing uncertain inference rules are largely developed for dense rule bases. However, in practice, rule bases are generally sparse rather than dense. To overcome this limitation, this paper develops a new uncertainty inference rule for sparse rule bases. Based on the proposed rule, a novel uncertain inference controller is designed to ensure well-defined control outputs even under severely limited rule coverage. The effectiveness of the proposed method was demonstrated by applying it to a trolley-inverted pendulum system, in which the controller successfully stabilized the pendulum in an upright equilibrium position. The results indicate that the proposed sparse rule-based uncertain inference mechanism is a practical and effective tool for challenging nonlinear control problems under sparse knowledge structures.</p>