| The two-ways mixed flow channel is a new type of flow channel of drip irrigation emitter.It can form mixed with the forward flow and backward flow to increase the local head loss and eliminate extra energy,which has good hydraulic performance.The laboratory experiment,FLUENT numerical simulation,particle image velocimetry(PIV),fluid dynamics and statistical learning theory,support vector machine(SVM),response surface analysis,genetic algorithm and other theories and techniques were used to carry out research of the hydraulic performance,energy dissipation mechanism,flow rate prediction and parameter combination optimization of emitter.These researches are of great importance to the research and development of the two-ways mixed flow emitter.The main research results obtained are as follows:(1)The orthogonal experimental design method was used and arranged 25 groups of emitter experimental schemes,laboratory experiment results showed that the flow index was from 0.432 to 0.464,hydraulic properties has greatly improved.The Range analysis and variance analysis of flow channel geometric parameter and flow index showed that the geometric parameter Z(maximal flow channel width of dividing water device and blocking water device)had th greatest influence on the flow index.(2)The comprehensive evaluation indicator system of simulation accuracy of the two-ways mixed flow emitter was established based on macroscopic flow rate and microscopic flow field,in order to evaluate the coincidence between numerical simulation and experimental data.According to the experimental data,the comprehensive indicator of 15 kinds of simulation methods of physical model and wall function was evaluated respectively.The results showed that the SST k-co model with enhanced wall function was more suitable for the simulation of the two-ways mixed flow emitter,and its comprehensive evaluation indicator value was 1.973%.(3)The study the relationship between different pressure range,two-ways flow ratio and hydraulic performance of emitter using SST k-? model with enhanced wall function.The results showed that flow ratio of the forward flow to backward flow tended to 1.indicating that the two-ways flow more fully mixed,and the hydraulic performance was more prominent.The distribution law of pressure drop along the flow channel unit was gradually reduced.And the maximum change of pressure drop mainly was appeared in mixed region of the two-ways flow,accounting for about 97%of total pressure drop.It is revealed that the mixed of the forward flow and backward flow could eliminate most of the excess energy inside flow channel.(4)In order to reveal the relationship between flow rate and pressure and geometric parameter,the flow rate prediction response surface of emitter was constructed based on support vector machine(SVM)prediction with strong generalization ability.The flow rate of emitter was predicted through training and learning of 300 training samples.The genetic algorithm was used to optimize the SVM parameter gam(C)and kernel parameter sig2(?2).The relative error between the predicted value and true value of 30 test samples was 1.810%and prediction accuracy was 0.982 when gam(C)=21898.9905 and sig2((52)=2201.0512.Compared with the traditional regression method,the accuracy of prediction was improved by 4.560%.(5)The steady flow indicator as a new kind of evaluation indicator of hydraulic performance suitable for the research and development of emitter was presented.The steady flow indicator could directly reflect the relationship between the rated flow rate of emitter and each influencing factor,and well overcome the uncertainty and blindness of flow control.The results showed that the steady flow indicator of scheme 21 was the smallest,which was 0.176,0.084,and 0.206,respectively,in the different pressure ranges,and the hydraulic performance was better.Combining with the flow rate prediction response surface and definition of steady flow indicator,the steady flow indicator response surface based on SVM was constructed.(6)The steady flow indicator was minimized as the objective function,an optimization model for the flow channel geometric parameter was established.And the hierarchical hybrid genetic algorithm based on continuous variables and discrete variables was used.The optimization method of flow channel geometric parameter was constructed based on SVM response surface and genetic algorithm.The data exchange and loop call were realized.The optimization schemes of flow channel geometric parameter with the rated flow rate of 2 L/h,2.5 L/h,3 L/h,3.5L/h,4L/h,5L/h were optimized and designed in different pressure ranges.This method can provide scientific basis for the development of emitter series products that to meet the needs of different pressure range and rated flow rate. |
PDF Full Text Request