| At present,extensive uniform spraying is still common in China’s field plant protection operations.Problems such as inaccurate application parameters and unreasonable operation planning are common,resulting in a high level of pesticide application in China.Especially for cotton and other typical crops with high canopy density,in order to improve the effect of plant protection and make the pesticide fully deposited in the canopy,there is a widespread phenomenon of high-dose pesticide application,which leads to more serious problems such as excessive pesticide residues,environmental pollution and ecological damage.In this paper,aiming at the optimization of spray parameters and the planning of pesticide application in the field spraying process,taking cotton crops as the object and using the boom sprayer as the carrier,the decision method of spray parameters based on cotton growth period identification is studied,and the path planning of machine tools in the process of pesticide application is studied,so as to reduce the phenomenon of re injection and leakage in the process of pesticide spraying.By means of system integration,a precision spraying decision-making system for boom sprayers based on growth period identification and spraying path planning has been developed.The test results show that the system can accurately identify the cotton growth period and achieve the automatic configuration of spray parameters,and can give suggestions for optimization of working path,improve the auxiliary navigation operation ability of the boom sprayer,greatly reduce the area of re injection and leakage in the operation process,which is of great significance to achieve the "reduction and efficiency increase " of field protection in our country.The specific contents are as follows:(1)Recognition of cotton growing period based on convolutional neural networkBased on convolution neural network,the recognition model of cotton growth period was constructed,and the structure of recognition model was optimized by orthogonal experiment.The accuracy,precision,recall,F1-score and recognition speed of the recognition model reached 93.27%,95.39%,94.31%,94.76% and 71.46ms/frame respectively.Compared with the commonly used image recognition models,the results show that the designed recognition model has better performance and is more suitable for cotton growth period identification.(2)Experiments on spray deposition of cotton plants in different growth periods and prediction of optimum spray parametersFor cotton plants in different growth periods,a variety of spray parameters were used for spray deposition experiments to verify the feasibility of optimizing spray parameter decisions based on crop growth periods and improving spray deposition effects.In addition,according to the spray deposition test results,the method of multiple nonlinear regression analysis was used to construct spray deposition regression models of cotton plants in different growth periods,and the optimal spray parameters of cotton plants in different growth periods were predicted based on the regression models.The spray deposition test is carried out with the predicted optimal spray parameters,and the test results show that the predicted spray parameters are reliable and effective.(3)Working path planning of boom sprayerAiming at the widespread repeated spraying and leaking spraying phenomenon in spraying operations,with the goal of optimizing the effective spraying range,the operation path planning of the boom sprayer was carried out,including basic path selection,line feed width optimization,boundary segmentation control strategy,and Obstacle avoidance strategy of segment control,etc.The simulation comparison test with the traditional path planning method showed that the total area of respray and missed spray was reduced by 93.36%,the false spray area was reduced by 90%,and the empty area was reduced by 87.37% by comprehensive optimization.The highest rate can reach 87.44%.(4)Development of a precision application decision-making system for boom sprayerBased on the cotton growth period identification model,the optimal spray parameters for each growth period prediction,and the path planning method of boom sprayer,the boom sprayer precision application decision-making system was developed,and the developed decisionmaking system was applied to the 3WP-1600 GA high ground clearance boom sprayer for functionality.The results show that the system can correctly identify the growth period and generate spray parameters,and the decision information can be correctly executed on the sprayer. |
