Abstract: The current farming platform has been widely equipped with unmanned four-wheel independent driving and four-wheel independent steering (4WID-4WIS). However, the control system of path tracking is required for high accuracy and sufficient stability under complex working conditions. There were also complex working conditions under crop ridge cultivation, such as Π type target path, curves, initial pose deviation, various soil moisture, and bumpy ground landscapes. In this study, a control strategy of path tracking was proposed using a nonlinear disturbance observer (NOB). A mathematical in-situ steering model was introduced for the relatively low tracking errors in the turn area of the Π type path, compared with the traditional Ackerman steering model. Two steering methods were then used to realize the turn path tracking. Meanwhile, a switch control strategy was designed between yaw angle proportional integral control and pure pursuit control using in-situ steering. Furthermore, the curve and initial pose deviation shared a relatively significant impact on the accuracy of working path tracking. Moreover, the distance traveled by the 4WID-4WIS farming platform was reduced to reach the working path and the maximum lateral deviation. The tracking accuracy of the work paths was improved to design a pure pursuit control using a lookahead distance function, and a fuzzy proportional compensator using the lateral deviation, as well as the curvature of the foresight area in the work path. Besides, the feedforward compensator with NOB was designed to avoid the relatively large yaw speed disturbances from the complex soil moisture, bumpy ground landscapes, kinematic models, and measurement errors. The NOB was also constructed to achieve the precise observation of disturbance for the expected path of farming platforms. The steering compensation angle was then calculated for the feedforward compensator to counteract the disturbance. Finally, the simulation was carried out in the Ubuntu/ROS environment. The NOB strategy of path tracking effectively reduced the distance traveled by farming platforms to reach the working paths, the maximum lateral deviation, and curve tracking errors. The accuracy and stability of path tracking were achieved in the anti-interference performance, where the disturbance momentum was observed accurately. And, the outdoor experiments show that the switch control strategy performed a smaller error of turn tracking on Π type target path, compared with the traditional pure pursuit control. The tracking performance was also effectively improved. The pure pursuit with the look-ahead distance function and fuzzy proportional compensator under different initial pose deviation states reduced the distance traveled by farming platforms to reach the working paths by 32.2%-43.4%. The maximum lateral deviation, mean absolute errors of the whole line and curved area were reduced by 0-42.4%, 27.7%-49.5%, and 33.7%-39.5%, respectively, indicating the high accuracy of working paths tracking. The NOB-based feedforward compensator was reduced by 6.25% mean absolute error in the steady area under hard slate condition, 33.3% under grassland condition, and 41.7% under farmland condition. This control strategy of path-tracking effectively improved the system's robustness and path-tracking accuracy. The finding can also provide innovative ideas and technical references for the navigation system of unmanned four-wheel drive and four-rotation agricultural machinery in ridge tillage.