Abstract: The operating environment of corn harvesters in hilly and mountainous areas is complex. Conventional harvesters are prone to body tilt during contour operation on slopes, which affects driving comfort, operational safety, and harvesting quality. To address these problems, this study designed a tracked leveling chassis and an adaptive lateral leveling system based on a tracked chassis, a three-point leveling mechanism, and the Beetle Antennae Search (BAS) algorithm. First, according to the stability and passability requirements of hilly and mountainous operations, the parameters of the tracked traveling and driving systems were determined, and an overall scheme of a dual-hydraulic-cylinder lateral leveling mechanism was proposed. Second, a kinematic model of the leveling mechanism was established. Based on the lateral leveling range of ±15°, key parameters such as the hydraulic cylinder stroke and cylinder diameter were determined, and the hydraulic system was designed. On this basis, a BAS Fuzzy-PID leveling controller was developed using the Beetle Antennae Search algorithm. PID, Fuzzy-PID, and BAS-PID controllers were used as comparative methods, and Simulink was employed to simulate the controllers and obtain simulation curves of the body lateral angle, hydraulic cylinder displacement, and velocity. The simulation results showed that, under a 15° slope condition, the BAS Fuzzy-PID controller achieved a shorter leveling time of 1.04 s and a higher leveling speed than the comparative controllers, with no overshoot, demonstrating faster leveling response and better stability. Under sinusoidal and random signal inputs, the BAS Fuzzy-PID controller also exhibited better signal tracking capability. Finally, a hardware-in-the-loop simulation platform was built based on dSPACE to verify the leveling response capability of the controller under slope and random road conditions. The results can provide a reference for the design and control of lateral leveling chassis for corn harvesters in hilly and mountainous areas.