Study On The Status Identification Method Of Sugar Cane Transfer Vehicls In Hilly Areas

Guangxi is China’s sugarcane production province,according to statistics:2007-2016,Guangxi sugar cane annual average planting area and total yield ratio of 105.8,76.705 million t,accounting for 60.5 percent of the country,64.7 percent,is the country’s main sugar cane cultivation area.Because of the special geographical environment,more than 60% of the sugar cane cultivation areas in Guangxi are located in hills,slopes,small and scattered planting area,large-scale mechanical operation difficulties,so that the overall mechanization of sugar cane cultivation is low,labor intensity,high planting costs,in the international competition at a disadvantage,seriously restricting the development of Guangxi sugar cane industry.Sugarcane transfer vehicle is an important field sugar cane handling,transportation means,can be harvested in the field sugar cane harvester sugar cane and field operations required for short-distance transport of sugar cane seeds,fertilizer and other substances,such as the field harvest of sugar cane by transfer truck to the roadside large tonnage of transport vehicles,sugar cane transfer truck work process stability is very important for the mechanized production of sugar cane.In this thesis,the self-driven double-shear fork sugarcane transfer vehicle in the hilly area is used as the research object,analyzes the structure composition and working principle of the transfer vehicle,adopts the signal statistical analysis,pretreatment and feature extraction by collecting the stress strain and vibration signal in the transfer work of the transfer vehicle,uses the construction state classifier to identify the status of the transfer vehicle,checks the unstable working state in real time,and ensures the efficient and reliable operation of the sugarcane transfer vehicle.The work of this article mainly includes the following three aspects:(1)According to the simulation and the test test test of the test platform,the relationship between the stress of the main node of the transfer vehicle lifting mechanism is discussed with the load load and tilt angle,and the transfer vehicle is divided into normal,tilted and overloaded states by using the analysis results of its node stress.(2)Using the collection experience modal decomposition method(EEMD)in the time and frequency domain processing method to modalize the vibration signal of the transfer vehicle,the IMF component screening of the vibration signal of the sugarcane transfer vehicle by the correlation coefficient method,and then extracting the characteristics of the time domain and frequency domain for the relevant IMF components.(3)Using the main component analysis(PCA)low-dimensional method to reduce the dimension of the extracted high-dimensional features,according to the status identification method of the transfer vehicle,the study based on the sugarcane transporter vibration signal support vector machine and BP neural network state classification method,in view of the shortcomings of the two state classifiers,proposed the use of group intelligence algorithm for parameter optimization and model optimization.(4)In view of the shortcomings of the traditional machine learning method of splitting the feature extraction and state classification,the paper puts forward the status recognition method of the transfer vehicle based on the deep confidence network,and the state recognition method based on the deep confidence network shows that the recognition accuracy of the transfer vehicle work process is higher by the experimental research and test.