| Abstract:In view of the problems of reduced speed and increased fuel consumption under transient sudden load conditions due to turbo lag effect, a transient control algorithm for an agricultural diesel engine based on load torque active observation was proposed. First of all, a prediction model of agricultural machine arable land resistance and engine brake torque was established. By adopting a tracking differentiator to implement the operation mode recognition, as well as the prediction of plough depth and load torque, a boost pressure feedforward controller was designed. Secondly, under the premise that the deviation of the prediction model and the external torque disturbance were regarded as “total disturbance”, an expanded state observer was used for online estimation and compensation. Then, based on the dynamic information of plough depth and engine speed, a parameter self-learning scheme for the load torque model was designed using a recursive optimization algorithm to improve feedforward accuracy, and reduce the observation burden of total disturbance. Finally, the algorithm was verified on a calibrated high-precision SIMULINK simulation platform. Results show that compared with traditional control method, this algorithm can increase the transient air supply speed by 53.1%, and accelerate the indicated torque response by 44.7%, thus reducing the engine speed fluctuation by 98.8%, and the transient fuel consumption by 7.0%. |