Coordination Control Based on Machine Learning Modified Pressure Observer for Skid-Steering Vehicles Coupled With Hydrostatic Propulsion Systems

Skid-steering vehicles (SSVs) have a wide range of applications, and coordinated control (CC) of the propulsion system is a key to improving steering capability. SSVs equipped with hydrostatic propulsion systems (HPSs) drive and steer by building up pressure, so pressure determines their steering capability. However, the strong coupling between HPS and vehicle dynamics poses a challenge for CC. To address this issue, this article proposes a CC strategy based on a machine learning (ML)-enhanced pressure observer. First, an ML approach is employed to refine the HPS’s efficiency model, thereby improving the accuracy of pressure estimation. A CC strategy is then developed to adjust the engine operating point, ensuring that the pressure demands of both the driving and steering systems are simultaneously satisfied. Finally, the effectiveness is verified by vehicle tests. Results show that the strategy enhances steering capability and reduces fuel consumption by approximately 10.98%, with a moderate increase in temperature of 1.25 °C (from 88.5 °C to 89.75 °C). This work provides both theoretical and practical support for the design and control of future SSVs.