Test Analysis And Prediction Research On Air-and-Screen Cleaning Device

The Air and Screen Cleaning Device is an important part of combine harvester. It＇sperformance can directly influence the cleaning quality.Firstly, we introduce the structure and working principle of cleaning parts and the Airand Screen Cleaning Device. Feeding quantity is determined on2kg/s. Then, we accord tothe comprehensive analysis by the Orthogonal Test to determine a sequence of the cleaningperformance. The sequence in turn is the speed of centrifugal fan, the speed of crank, thespeed of cross flow fan, the angle of centrifugal fan and the angle of vibration sieve.Secondly, in order to reduce the experimental factors, we determine the angle ofcentrifugal fan is25°and the angle of vibrating screen is8°by doing test. The test indexincludes impurity rate, loss rate and power waste. The experimental includes speed ofcentrifugal fan, speed of crank, speed of cross flow fan. In order to get the experimentalfactors value from the test index, this article constructs a model by using the BP NeuralNetwork. We put the impurity rate, loss rate, power waste as an input and put the speed ofcentrifugal fan, speed of crank, speed of cross flow fan as an output.Then, we verify the above model. This model realizes a mapping from input to output.The impurity rate, loss rate, power waste of verification test are obtained by adjustingspeed of centrifugal fan, speed of crank, speed of cross flow fan of this test bench. The twosets of data have little difference and it proves the feasibility of this model.Last, we build a multi factor combined mathematical model, applied to Air and ScreenCleaning Device and analyze the performance of test bench. It solves the problem thatother method did not resolve and provides a theoretical basis on designing and operatingadjustment of Air and Screen Cleaning Device.