Research On The Design Method Of Agricultural Harvester Driven By Multi-modal Data Fusion

Under the category of household production and operation organization,small farmers with the prominent feature of "small scale" occupy the dominant position in terms of quantity among the main bodies of agricultural production and operation in our country.However,due to the complex topography,fragmentation of farmland and narrow roads of machine ploughing,the cost of agricultural production organization is high in hilly and mountainous areas.Large and medium-sized agricultural harvesters cannot enter the field for operation,and the existing small agricultural harvesters cannot effectively improve labor efficiency and reduce labor intensity of farmers.This paper proposes a design method of agricultural harvester driven by multi-modal data fusion,constructs a decision-making inference model suitable for the design of agricultural machinery equipment,improves the efficiency and quality of the design and development of agricultural harvester,and designs a new type of agricultural harvester that meets the national conditions and the needs of farmers,so as to improve the working efficiency of agricultural machinery,reduce the labor intensity of farmers,and promote the healthy and sustainable development of Chinese agriculture.Firstly,this paper explores the theory and related applications of multimodal data fusion,and clarifies the applicability,significance and advantages of multimodal data fusion in the field of agricultural machinery equipment design.According to the conditional constraints of multimodal data fusion,the farmers’ demand information and agricultural machinery equipment patent information are determined as the multimodal information sources in the design of agricultural machinery equipment.The multimodal data fusion method is analyzed in depth,and the model-independent data fusion method is adopted to propose a multimodal data fusion method for the efficacy analysis of patent information technology based on farmers’ demand constraints,so as to establish the connection between the farmers’ demand information and patent data information.The feasibility of the design method of agricultural equipment driven by multimodal data fusion is verified through the scheme design of agricultural harvester.The online review information of agricultural harvester users and the patent design information of agricultural harvester are used as the data information sources in the efficacy analysis method of patent information technology based on farmers’ demand constraints.The needs of farmers were determined by text information mining technology,and the patent information of agricultural harvesters was analyzed by dimensional clustering analysis based on the constraints of farmers’ needs to obtain effective patent information,which provided knowledge support for design decisions,determined the design and development direction of agricultural harvesters and presented the design scheme.The feasibility of the method was verified,and new ideas and directions were provided for the design of related agricultural machinery equipment products.