Observation Method for the Operation Parameters of Greenhouse Electric Tractor Rotary Tillage Based on GABP-STUKF

Traditional greenhouse electric tractors are not equipped with high-precision sensors, which make it difficult to conduct rotary tillage parameter measurements.Therefore, problems due to imperfect rotary tillage parameter measurements are encountered, such as inconvenient measurement, low data resolution, poor sensing accuracy of operation parameters, and large noise.To improve the accuracy of data collected by electric tractor rotary tillage operations, this study presented a strong tracking untracked Kalman filter (STUKF) parameter observation method based on a genetic algorithm (GA) optimization neural network.First, a back propagation neural network was used to predict the state of the greenhouse tractor, and GA was used to avoid the local minimum in the prediction process.Second, untrace transform and strong tracking filter were applied to solve the strong nonlinearity and imprecision problems of the network model.Third, a method of operation parameter observation based on genetic algorithm back propagation (GABP)-STUKF was proposed.Finally, the actual vehicle test was carried out on the washboard pavement of the Jiangsu Agricultural Machinery Test and Identification Base.Results show that, the mean absolute error (MAE) and root mean square error (RMSE) of longitudinal vehicle speed observation were 0.017 and 0.025 m/s, respectively.The MAE and RMSE of cutter roll speed prediction were 8.941 and 18.413 rpm, respectively.The MAE and RMSE predicted by ploughing depth were 0.029 and 0.038 m, respectively.With the acquisition methods of motor driver and rope sensor taken as the control group, the MAE observed by GABP-STUKF for the above three operating parameters increased by 10%, 46.8%, and 7.54%, respectively.This study provides a new idea for the parameter observation of greenhouse electric tractors..