| The automatic navigation control method is a key part of restricting the precise operation of intelligent agricultural machinery,which is conducive to the intelligent development of modern agriculture in China.With the proposal of the national strategy of accelerating the construction of an agricultural power,and China is a large country in fruit planting area and the outflow of rural labor.Therefore,in order to respond to the needs of national strategic development and solve the shortage of agricultural labor,it is imperative to develop automation and intelligence of orchard operation machinery.However,at present,some orchards in China are hilly and mountainous,limited by topography,irregular distribution of fruit trees,high labor intensity,high labor costs,and low production efficiency.In view of the above problems,in order to realize the precise operation of automatic navigation of orchard crawler vehicles,the accuracy and stability of automatic navigation are improved.In this paper,the tracked chassis is used as the research platform to carry out the research of the automatic navigation control method of orchard.Carry out the research on turning model of orchard crawler,the key technology research of orchard navigation,the research on automatic navigation control method of orchard crawler,and the integration and test of automatic navigation system of orchard crawler.The main research and conclusions of this paper are as follows:（1）Research on turning model of orchard crawler vehicle.In order to improve the accuracy of turn-control of crawler vehicles in orchards,a prediction model of crawler turning parameters based on support vector machine regression algorithm is proposed.Firstly,through the design nesting experiment,600 sets of crawler slope,theoretical speed of left track,theoretical speed of right track,actual speed of left track,actual speed of right track and turning radius of 600 sets of crawler at different speeds were collected.Secondly,a variety of machine learning algorithms are used to construct,analyze and compare the crawler turning model,and select the optimal algorithm to construct the turning parameter value prediction model.Finally,80 sets of data are selected to verify the effect of the established crawler turning parameter value model.The results show that the crawler turn prediction model built based on the support vector machine algorithm has the highest prediction accuracy and the best fitting effect.The coefficients of determination of the prediction model based on the parameter values of the support vector machine algorithm were 0.979653,0.999834 and 0.999816,and the root mean square errors were 0.00717925,4.98791×10-5,4.8164×10-5,respectively.（2）Research on orchard point cloud map construction and navigation path optimization method.In order to solve the problems of large canopy crossing and irregular distribution of fruit trees in orchards,a lidar-based point cloud map construction and navigation path optimization method for orchards is proposed.Firstly,solid-state lidar is used to collect three-dimensional point cloud data of orchard environment,and point cloud processing software is used to perform point cloud pass-through filtering,voxel downsampling filtering and statistical filtering preprocessing of the collected point cloud data,and the ground point cloud plane fitting algorithm is applied to split the ground point cloud from the non-ground point cloud.Secondly,the ground point cloud data are fitted to the ground by using the overall least squares method,random sampling consensus algorithm and eigenvalue method,and the comparative fitting effect is analyzed.Finally,according to the preliminary investigation of the planting mode of oil olive,according to the environmental characteristics of the orchard,the shortcomings of the existing navigation algorithm are analyzed,the algorithm is improved and optimized,and the navigation path is compared and analyzed by simulation software.The simulation results show that the running time of the original A*algorithm is 30.600207s,and the running time of the improved A*algorithm is 0.203237s,which greatly improves its operating efficiency;moreover,the length of the planned path is also shortened,and the number of turns is reduced from the original 8 to 6,which is reduced by 25%,which improves the smoothness of automatic navigation of orchard crawler vehicles.（3）Research on automatic navigation control method of orchard crawler vehicle.In order to realize the precise control of automatic navigation of orchard crawler and improve the accuracy and stability of automatic navigation,an adaptive navigation control decision-making method based on fuzzy control is proposed,and an automatic navigation embedded control system is designed and developed.Firstly,the overall architecture of the automatic navigation control system of the orchard crawler is designed.Secondly,determine the hardware and software structure of each key subsystem;Then,based on fuzzy control and traditional PID control methods,the adaptive navigation control decision-making method is studied,the mathematical model of automatic navigation is established,and the simulation research of automatic navigation control system is carried out in MATLAB.Through simulation,it can be seen that fuzzy PID is optimized by 41.31%and 34.78%compared with the overshoot and steady-state errors of conventional PID control methods,respectively.（4）Orchard crawler automatic navigation system integration and test.In order to verify the control accuracy and robustness of the automatic navigation of the orchard crawler,the software and hardware integration of the automatic navigation control system of the orchard crawler were integrated,and the test was carried out at different speeds under the two control modes of conventional PID and fuzzy PID.The test results show that under the conventional PID control mode,the orchard crawler drove at a speed of 0.1m/s,0.2m/s,0.3m/s and 0.4m/s,with the average transverse deviation of 3.63cm,4.77cm,5.30cm and 6.82cm,and the average longitudinal deviation of 5.32cm,5.18cm,4.41cm and 6.37cm,respectively.The average lateral deviations of orchard crawler turning under the fuzzy PID control mode were2.72cm,2.83cm,3.53cm and 4.65cm,and the average longitudinal deviations were 5.00cm,4.25cm,5.52cm and 6.64cm,respectively.When the speed of the crawler is 0.1m/s,the lateral error and longitudinal error of the vehicle are the smallest,and the average transverse error and longitudinal error of fuzzy PID adaptive control are increased by 25.07%and 46.80%,respectively,compared with the traditional PID adaptive control. |
PDF Full Text Request