| Recently,the intelligent detection and control of the combine harvester has become an important content in the intelligent research of agricultural machinery.The control effect of the harvester control system affects the harvest quality and harvesting efficiency of the harvester.The research of combine harvester and automatic control of harvester started late in China.Most agricultural machinery achieved mechanization but did not achieve automation,and did not achieve intelligence.The cleaning system is the last system that can affect the harvest quality of the harvesting part of the combine harvester.The detection device and control system are not perfect enough.Due to the complicated working environment of the combine harvester,the detection effect of cleaning detection device often can not meet the control requirements and can not provide suggestions for the production process.The combined harvester is complex in modeling,and the coupling and hysteresis between the systems are serious.Effective control methods are needed to overcome the complexity of the cleaning system.Aiming at the problems existing in the existing harvester control system,and aiming at the detection and control of the cleaning system,this paper proposes a cleaning vector loss detection scheme based on support vector machine,which is based on the existing cleaning system adjustment scheme.The screening sieve opening and fan speed adjustment were selected,and statistical experiments were designed.The collected data were analyzed,and the factors affecting the cleaning loss rate and grain impurity rate were selected.According to the collected test data,the cleaning system was selected.Based on the modeling,a fuzzy control scheme was proposed and the actual experiment was carried out on the combine harvester.The specific research work is as follows:1.In order to realize the detection of cleaning loss rate,provide basic conditions for system modeling and control,design and develop a cleaning detection sensor based on piezoelectric effect,design signal processing circuit based on charge signal generated by sensor,and make grain and miscellaneous The remaining impact signals are recorded,and the collected data is used to train the support vector machine classification model.The support vector machine method is embedded into the detection device,and the impact signal is classified according to the obtained support vector in real time,according to the location of the sensor.The impact of other working parameters of the harvester is used to calculate the loss rate of the cleaning.2.In order to achieve automatic adjustment of the combine harvester cleaning system.The design and development of the clearing screen opening adjustment device and the fan speed adjusting device are carried out according to the existing adjustment method.The crank slider mechanism is adopted to design different adjusting devices for the front,rear and tail screens,and the adjusting mechanism is designed.The drive circuit realizes the installation of the adjustment device without affecting the normal harvest.The adjustment of the fan speed is designed with the proportional relief valve and cycloidal motor as the main structure adjustment device,which directly drives the fan to adjust the speed.The realization of the two adjustment devices is the basis for the automatic control of the cleaning system.The least change was made.3.In order to realize the experiment of the cleaning system and the better control of the cleaning system in the non-harvesting season,the modeling system of the cleaning system was designed and analyzed,and the Plackett-Burman test and orthogonal experiment were designed.The harvesting machine working parameters of the loss rate and grain impurity rate were selected for detection and recording.The main working parameters affecting the harvesting effect were analyzed by statistical methods.The parameters were subjected to multiple regression analysis,and finally the model of the control system was obtained.The screening system adjusts the sieve opening adjustment and the fan speed adjustment mediation scheme to dynamically reflect the relationship between the various quantities of the system.4.Based on the obtained model,the control method of the cleaning system is studied.The simulation model is controlled by the paste control method based on the expert system.The fuzzy control rules are designed to overcome the inaccuracy and nonlinearity of the model to some extent.The control method is well-established on MATLAB and field trial is carried out.The comparison control was carried out,and the final control effect was obtained.The cleaning loss rate and the grain impurity content were all maintained within a certain range,which improved the harvesting speed. |