| Under the rapid development of the agricultural mechanization,developing intelligent agricultural machinery can significantly improve the production quality and efficiency,reduce environmental pollution,and take advantage of the agricultural resources efficiently.Both economic and environmental benefits will be gained.In order to realize the automatic guidance accurately,an intelligent vehicle was designed as the experiment platform.RTK-DGPS(Real time kinematics Differential GPS)was chosen as the main navigation sensor.PID(Proportion Integration Differentiation),FOPID(Fractional order PID)and ADRC(Active Disturbances Rejection Controller)control algorithms were researched in detail.Research contents and results were summarized as below:1.An intelligent agricultural vehicle was developed as the experiment platform.The design criterion of the platform is stabilization,high precision,fast response and easy to control.The platform was composed of sensor system,actuators and CAN-Bus communication network.In the sensor system,the global positioning coordinate,the posture,the wheel angle and the speed of the vehicle were measured respectively by RTK-DGPS,inertial sensor,absolute type encoder and incremental type encoder.Useful information was extracted from the original sensor data via data processing.Actuators included the steering,the braking and the speed control mechanisms.Steering speed was controlled by the frequency of the stepping motor.Braking was realized by transposing the voltage of the linear actuator.Speed control was realized by changing the input voltage of the drive motor.Each node with the CAN controller and the CAN transceiver was communicated to each other rapidly,steadily and reliably according to the designed CAN-Bus protocol.2.A method for calculating navigation parameters with anti-interference performance was proposed.When the soil environment is uneven,the error between the RTK-DGPS measurement point and the centroid of the vehicle will be enlarged.So the lever-arm effect error was compensated via the geometric conversion by using the data from the inertial sensor.On the uneven road,test results showed that the mean errors between the calculated values and the measured values were 0.026 m and 0.054 m after and before the compensation,so more accurate lateral deviation was got via the compensation.The error of the heading deviation calculated by the inertial sensor will be enlarged when there is strong magnetic interference around the working environment.Therefore,using the Least-square method to process the RTK-DGPS dynamic positioning points was proposed to calculate the heading deviation.Test results showed that,when the vehicle was drived along a straight line,a circle and a curve on the concrete pavement,the mean errors between two methods were 0.4798°,1.8029°and 1.2277°.It was proved that the heading deviation could be calculated by RTK-DGPS effectively.So the proposed method can avoid the soil environmental interference and magnetic-field interference.3.Mathematical model of the navigation control system was established.In order to reduce the complexity of the control system design,simple and accurate mathematical model are expected to establish,and the disturbances from the model uncertainties are expected to be compensated by the robust controller.The motion model of the vehicle was established according to the kinematics rules.And the open-loop model of the steering mechanism was established according to the relationship between the stepping motor frequency and the wheel angle.A steering proportion controller was designed to enhance the dynamic performance of the steering system.Then the closed-loop model of the steering system was identified by adding the step signal artificially.Then,heading control system model and lateral control system model were established based on the motion model of the vehicle and the closed-loop model of the steering system.4.Based on the performance indexes,path following PID controllers for agricultural vehicle were designed.Considering the shortcoming of the traditional parameter tuning methods,a method for tuning the PID controller with performance indexes was proposed.Performance indexes included ISE(integral square error),IAE(integral absolute error),ITAE(integral time absolute error)and ITSE(integral time square error).The control performance of these four optimal PID controllers were discussed though simulation in detail.By comparison,the optimal PD controller CITAE?9s+ 0.1 tuned by ITAE performance index was chosen since it can minimized the system oscillation.Overshoot and settling time of the closed-loop system controlled by the controller were 21.9%and 0.69 s respectively.5.An optimal FOPID controller was designed for the path tracking control.The controller parameters were tuned by two methods.One was the previous flat phase method.The other was an improved tuning method proposed in this paper,which can compensate the missing parts of the flat phase method caused by computation.Comparison and analysis about two tuning methods were given in detail.And two FOPD controllers tuned by two methods were compared on aspects of frequency domain,time domain,and robustness.Simulation results demonstrated that the FOPD controller CFOPD1(s)= 0.0218(1 + 300s1.017)tuned by the improved tuning method was more superior.Moreover,the optimal FOPD controller was taken to compare with the optimal IOPD(integer order proportion differentiation)controller CPD(s)=9s+0.1.Simulation results demonstrated that the optimal FOPD controller enabled the closed-loop system have smaller ITAE value,and overshoot and settling time of the closed-loop system controlled by the FOPD controller were 11.48%and 0.79 s respectively.6.An ADRC heading controller was designed based on the stability analysis of the heading control system.The assumptive constant disturbance and time-varying disturbances were observed effectively by the ESO(Extended state observer).After the disturbance compensation,the system became linear.Then an error feedback law was designed for the linear system.Simulation results demonstrated that ADRC had excellent capacity of resisting the given disturbances,and overshoot and settling time of the closed-loop system controlled by the FOPD controller were 2.13%and 0.66 s respectively.7.The control performances of the heading and lateral control systems were verified via the experiments.A simple control strategy was designed,and the speed of the vehicle was set to be 0.5 m s-1.The results of the heading control tests showed that the settling time were all in 2 s and the tracking precisions were all less than 1° when the initial heading deviation were-86° and 84°.The control performance was satisfactory.On the other hand,both optimal PD controller and FOPD controller were verified in the lateral control experiments.Based on the optimal PD controller,test 1 result showed that the maximum,mean and standard deviation of the lateral deviation was 10.4 cm,6.4 cm and 2.6 cm,and test 2 result showed that the maximum,mean and standard deviation was 9.2cm,3.5 cm and 1.7 cm.Based on the optimal FOPD controller,test 1 result showed that the maximum,mean and standard deviation of thelateral deviation was 9.2 cm,3.0 cm and 1.7 cm,and test 2 result showed that the maximum,mean and standard deviation was 6.2 cm,2.4 cm and 1.6 cm.Limited by the precision of the steering mechanism and the sensor,there is no significant difference between two controllers,however,both controllers made the vehicle track the desired path steadily and the tracking precision are all in centimeter-level. |
PDF Full Text Request