农业工程学报
農業工程學報
농업공정학보
2013年
17期
9-15
,共7页
廖庆喜%杨松%廖宜涛%丛锦玲%王磊
廖慶喜%楊鬆%廖宜濤%叢錦玲%王磊
료경희%양송%료의도%총금령%왕뢰
农业机械%神经网络%气力装置%系统建模%油菜%整数规划
農業機械%神經網絡%氣力裝置%繫統建模%油菜%整數規劃
농업궤계%신경망락%기력장치%계통건모%유채%정수규화
agricultural machinery%neural networks%pneumatic equipment%system modeling%rapeseed%integer programming
为了了解油菜精量联合直播机气力排种系统工作参数及排种性能对油菜直播机作业性能的直接影响,该文应用JPS-12排种器性能检测试验台及自行研制的气力式油菜精量排种检测系统,开展了气力排种系统排种器单体负压流量与压强、系统各负压支路输气管压强及系统总负压输气管流量的试验,构建了基于系统总负压输气管流量与各支路输气管流量的神经网络模型,得出了系统总负压流量与风机转速、正压泄气孔大小的关系式及系统各支路输气管压强与流量的转换关系模型;同时以排种系统各行排量一致性变异系数<5%、各排种器单转排量>36粒等为约束方程,建立了以风机转速、排种轴转速及正压泄气孔大小为变量的整数规划模型,研究表明,所建立的系统工作参数及排种性能模型,对排种系统的性能优化与结构改进具有重要的理论价值和实际指导意义。
為瞭瞭解油菜精量聯閤直播機氣力排種繫統工作參數及排種性能對油菜直播機作業性能的直接影響,該文應用JPS-12排種器性能檢測試驗檯及自行研製的氣力式油菜精量排種檢測繫統,開展瞭氣力排種繫統排種器單體負壓流量與壓彊、繫統各負壓支路輸氣管壓彊及繫統總負壓輸氣管流量的試驗,構建瞭基于繫統總負壓輸氣管流量與各支路輸氣管流量的神經網絡模型,得齣瞭繫統總負壓流量與風機轉速、正壓洩氣孔大小的關繫式及繫統各支路輸氣管壓彊與流量的轉換關繫模型;同時以排種繫統各行排量一緻性變異繫數<5%、各排種器單轉排量>36粒等為約束方程,建立瞭以風機轉速、排種軸轉速及正壓洩氣孔大小為變量的整數規劃模型,研究錶明,所建立的繫統工作參數及排種性能模型,對排種繫統的性能優化與結構改進具有重要的理論價值和實際指導意義。
위료료해유채정량연합직파궤기력배충계통공작삼수급배충성능대유채직파궤작업성능적직접영향,해문응용JPS-12배충기성능검측시험태급자행연제적기력식유채정량배충검측계통,개전료기력배충계통배충기단체부압류량여압강、계통각부압지로수기관압강급계통총부압수기관류량적시험,구건료기우계통총부압수기관류량여각지로수기관류량적신경망락모형,득출료계통총부압류량여풍궤전속、정압설기공대소적관계식급계통각지로수기관압강여류량적전환관계모형;동시이배충계통각행배량일치성변이계수<5%、각배충기단전배량>36립등위약속방정,건립료이풍궤전속、배충축전속급정압설기공대소위변량적정수규화모형,연구표명,소건립적계통공작삼수급배충성능모형,대배충계통적성능우화여결구개진구유중요적이론개치화실제지도의의。
2BFQ-6 precision planter for rapeseed can be used for sowing, fertilization, rotary tillage, stubbing and trenching. Pneumatic seed-metering system is the key component of this planter, and its working parameters and seeding performances directly affect the planter’s operating performances that include the consistency of seeding quantity per row and the stability of the total seeding quantity of the planter. The purpose of this paper was to explore the relationships among the parameters of pneumatic seed-metering system that include flow rate and pressure of each negative pressure branch tube, flow rate of the total negative pressure tube, rotation speeds of the fan and sowing disc, and the outlet size of positive pressure tube. These parameters’ action laws had direct effect on seeding performances of the system, then the models of these relationships and action laws was constructed. So that we could improve seeding performances of the planter by controlling these parameters, and predict the planter’s seeding performances under a certain system condition. On the basis of JPS-12 and testing system of pneumatic seed-metering device for rapeseed, a two-layer Artificial Neural Networks model had been constructed which was about the relationship between flow rate of the total negative pressure tube and flow rate of each branch negative pressure tube. A binary quadratic polynomial model was established to describe the relationship between the flow rate of total negative pressure tube and fan rate, outlet size of positive pressure tube. And according to the regression equation, pressure with flow rate of each negative pressure branch tube was well fitted by testing flow rate and pressure of the seed-metering device. At a certain fan rate and outlet size of positive pressure tube, flow rate of the total negative pressure tube, flow rate and pressure of each negative pressure branch tube could be calculated through these models. This fan rate and outlet size can also be modified by some equations established in this paper. An integer programming model with 3 variables which were rotational speed of fan, rotational speed of sowing disc, the outlet size of positive pressure tube, and 2 constraint equations which were the equations of seed rate of the system and its uniformity was explored in this paper. In addition, more than 80,000 integer solutions were solved by using the MATLAB software, all of which could make sure that the average seeding quantity of the system was more than 36 and the variation coefficient (which was about the consistency of seeding quantity of the system) was less than 5%. The models constructed in this paper had expounded the mutual laws among the parameters of the system and also had described the action laws of these parameters to the operating performances of the system. The results showed that these models were valuable for optimization and improvement of operating performance and structure of the system.