Combine Harvester Low Crushing Rate Operation Strategy Research by Using Bayesian Network

As the main harvesting machinery, combine harvester is often due to improper adjustment of its operating parameters resulting in increased crushing rate and grain waste during process. To quickly obtain working range key under low rate, this study conducted field tests on relevant affecting finally selected travel speed, feed threshing drum concave clearance, as node variables for construction Bayesian network model. Based “search-and-score” algorithm, best structure can be obtained using combination Akaike Information Criterion (AIC) scoring function hill-climbing method. In network, adjusting proportion lowest level nodes 100 %, operation strategy condition broken by reasoning was: < 6 kg/s, speed 5 km/h, clearance = 10 mm, 900 rpm. Three trials were carried out optimized strategy, measured rates 0.93 0.95 1.07 respectively, average was 0.98 %. At same time, when not used, were, 1.12 1.41 1.93 1.48 The test results prove that based inference effectively reduce