Study On Precision Spray Technology For Barrier And Atomization Model For Nozzles

China is rich in forest resources,but the current forest region has a large area of disease and pests,and the prevention and control rate is low.Among all methods of prevention at present,chemical methods are most effective.In the case of leaf-feeding pests with a two-way migration of ground and canopy,"Barrier treatment" application at a certain position on the trunk is the most common method for chemical control.However,there are problems such as waste of liquid medicine and environmental pollution when using this method.In response to the problem,this study has developed a tracked barrier precision spray vehicle,which applying barrier band accurately at a certain distance of the trunk from ground.This method could save more than 90%of the liquid.At the same time,a multi-threaded software recognition control system based on MFC was developed.The above studies had solved the problems that the target could be precisely obtained and the liquid spray was accurate,achieving the goal of precision spray.As for the effect of spray for the trunk,that is how large the deposition rate of the droplet on the surface of the trunk is,this research fundamentally established the atomizing model of the flat fan nozzle for precision spray of barrier.Relevant study was carried out from two aspects:the spray droplet size and the spray-axis speed of arbitrary section.The actual and simulation test methods were used to verify the model,and relevant research results are as follows:1.According to the processing flow of point cloud data returned by the two-dimensional laser scanner,the filtering,clustering and fitting algorithms were studied respectively.Of which a window filtering algorithm was used as filtering algorithm,whose main function was to filter out mixed pixels,and the filtering rate was greater than 99.5%.For the dense growth of the tree trunks,an optimized K-means clustering method was used.The number of clusters was determined using the slope changing method and the initial cluster center was determined using the Huffman tree method,meanwhile,the Huffman tree and random sampling methods were compared,results show that the average accuracy of clustering based on Huffman tree method was 95.5%,and that of random sampling method was only 76.4%,Three fitting methods based on the least square were choiced as fitting algorithms,and it was testified that the best algorithms were KASA algorithm and MLS algorithm,KASA was better,and the best angle resolution was 0.333°.Because the geometric method has the advantage of less time consuming,this study proposed an optimization scheme for commonly used geometric algorithms.The results showed that the optimized arc length algorithm and tangent algorithm could increase the accuracy of radius measurement by 10.6%and 10.7%,respectively.The absolute error of distance measurement was controlled within 66.0mm and 15.9mm,and the tangent method had better stability and measurement accuracy.2.A target location and identification system was constructed,which based on the LMS511 laser scanner as a server and an industrial computer as a client(C/S),a control system with Zhongtai USB-7660DN data acquisition card treated as the core control component was also constructed,and the related target recognition algorithm was developed,based on this,the velocity and heading angle of the spray platform were obtained,the target position interpolation was predicted too,and the spray system based on tracked chassis was also constructed.At the same time,a multi-threaded software system based on MFC was developed,through which the target positioning and identification,the speed of the spray vehicle and the total time-consuming of spray were all achieved.3.A suitable theoretical result were extracted from classical theoretical research for this study and were then improved concretely in connection with this research,and a specific atomizing model for the two types of nozzle(Lechler and Feizhuo)was obtained.This model mainly included the two most important parameters:droplet size and spray-axis velocity of arbitrary section.The obtained droplet size calculation model was validated by the actual test method,which showed that at 6 different pressures,the average absolute error of theoretical data obtained from the selected nozzles of Lechler was 6.92 ?m compared to the actual measured data,while the average relative error was 3.07%.The data of the Feizhuo nozzle was 11.41 ?m and 4.43%,respectively.The test also verified that the droplet sizes at different distances from the same nozzle did not vary dramatically at a given pressure in the close-up range(30-50 cm).In addition,the nozzle sizes at different angles also did not significantly vary at a certain pressure and flow model.However,the nozzle sizes with the same angle increased as the flow increased.The model of spray axis velocity was verified by the CFD simulation method using existing equipment and calculation conditions.Compared with the simulation data,the average absolute error of the improved theoretical data for the Lechler nozzle was 0.64 m/s for the selected nozzles,and the average relative error was 6.42%.The data of the Feizhuo nozzle was 0.67 m/s and 6.00%,respectively.The two models all have a high degree of confidence level.The research results of this paper provide new ideas and technological means for precision forestry plant protection.It has a certain reference value for actual production and scientific research and it is also conducive to the development of forestry in China.