Sieving Model Of Agricultural Granular Materials Using Biological Neural Network And The Verification Tests

There are many problems for currently widely used single-degree-of-freedom vibration mechanism under non-uniform feeding condition,such as uneven distribution of materials on the screen surface and low utilization rate of the screen surface,which results in a high rate of material loss.Based on the vibration,the inclination angle and attitude angle of the screen surface have been adjusted.And a multi-DOF screening mechanism that can realize single-degree-of-freedom reciprocating motion and twodegree-of-freedom rotation has been designed.In order to solve the problem of nonuniform feeding,a theoretical model of vibration sieving for non-uniform feeding has been established by adopting the method of biological neural network.The theoretical calculation results have been compared with discrete element simulation and bench tests to verify its feasibility.The main research contents of this article include:1.The design of multi-degree-of-freedom vibratory sieving mechanism.A multidegree-of-freedom vibratory sieving mechanism has been designed,which can adjust the inclination angle and horizontal attitude angle of the sieve surface based on reciprocating vibration.The degrees of freedom has been calculated and the dynamic characteristics have been simulated.What's more,its circuit control system has been developed and designed based on the decoupling control of stepper motor.2.The research on simulation distribution of material sieving.The effects of different vibration frequencies,screen size,feed rate and sieve inclination angle on the material simulation distribution model under two typical vibration excitation conditions have been analyzed,which provides a theoretical basis for the establishment of the biological neural network vibration sieving model.3.The construction method of biological neural network vibration sieving model.Based on the biological neural network,a motion distribution model of agricultural materials on the vibrating sieve has been established.And its internal connection weight function has been defined.What's more,the theoretical model has been verified by discrete element simulation under two typical vibration excitation and multiple complex input conditions.4.The bench test of vibration sieving device.Based on the designed and manufactured multi-DOF vibration screening device,the vibration sieving tests of different vibration frequency,sieve hole size,feeding rate and sieve inclination angle have been carried out.The calculation results of the theoretical model and the results of the test bench have been compared and verified.The feasibility of the biological neural network vibration sieving model has been discussed.