Predictive Control Of Spray Bar Balance Of Plant Protection Machine Based On Neural Network

The plant protection machine is affected by bad roads and machine moving conditions.In the spray work,there are many problems such as low utilization rate of pesticide and uneven spraying,which cause pollution of natural ecology and endless loss of liquid medicine.For large-scale plant protection machine,small body shaking will cause strong vibration of spray bar,which makes it difficult for spray bar to follow the set value,resulting in excessive or insufficient spraying,and even causing the spray bar to scratch crops,causing crop damage or breakage at the end of spray bar,which has a great impact on the spraying progress and effect of plant protection machine.On the other hand,the spray bar system of plant protection machine is hydraulically driven,which is nonlinear and time-varying.Therefore,it is of great significance to control the spray bar tracking angle of plant protection machine with uncertainty,nonlinearity and time-varying.The main research work of this thesis is as follows.According to the structure and principle of the spray bar system of pendulum type plant protection machine suspension,the model of spray bar system of plant protection machine suspension is established by using Lagrange second kind dynamic equation,and the nonlinear kinematics model of plant protection machine system is obtained.According to the nonlinear characteristics of the spray bar system of the plant protection machine suspension,the BP neural network is used to establish the prediction model,the feedback correction is established according to the difference between the actual output and the prediction output of the BP neural network model,the performance index function of the plant protection machine system is established by using the difference between the feedback output and the set value,and the optimal control rate is obtained online through the system performance index function to determine the control signal.The simulation results show that the predictive controller based on BP neural network model has good control performance for spray bar of plant protection machine.Aiming at the problem that the modeling accuracy and network weight of BP network can not be explained,the prediction model of spray bar system of plant protection machine is established by using stochastic collocation network(SCN).The feedback correction is established according to the difference between the actual output and the predicted output of SCN model.The performance index function of plant protection machine system is established by using the difference between the feedback output and the set value The design of the controller is verified by simulation under different initial angles and set angles.The simulation results show that the predictive controller based on SCN model has good control performance and faster control speed than BP neural network model predictive controller.