| Military equipment is a crucial force for executing combat missions,and its full potential in mobility is the foundation for ensuring the completion of these missions.Weather conditions are one of the key factors affecting the mobility of equipment,and the degree of impact on equipment mobility cannot be overlooked.To meet the urgent needs of our military’s core military capability building and intelligent application of meteorological data,it is imperative to improve our understanding of the meteorological environment and the assessment of military equipment effectiveness.Traditional methods for evaluating military equipment mobility have several limitations,mostly relying on expert experience,subjective evaluation,or statistical analysis.Thus,this can lead to significant subjectivity in the evaluation results.Furthermore,existing methods cannot comprehensively consider the complex interactions between meteorological impact indicators and other meteorological factors,and they struggle to meet the growing demand for rapid increases in meteorological data.Therefore,this dissertation focuses on the evaluation of military equipment mobility under meteorological conditions and conducts the following research content:(1)In response to the limitations of current efficiency evaluation methods in extracting spatiotemporal features from meteorological data,the Weather Trans-Model efficiency evaluation model was constructed in this dissertation based on the proposed mobility impact indicators.This model uses different modules to extract features from input data based on their characteristics and employs a Transformer variant as the sequence prediction module.Additionally,environmental factors were introduced as auxiliary variables to learn the complex interactions between different meteorological factors.Ultimately,the model output is converted into a classification probability distribution,enabling the prediction of the degree of impact on equipment mobility at a future time.To meet the demand for fast response to military decision-making under the increasing amount of meteorological data,the model was parallelized using Map Reduce.Finally,the effectiveness of the model’s component design,environmental factors,and parallelization performance was validated through ablation experiments on accuracy,precision,recall,and F1-score.(2)In response to the application requirements of the Weather Trans-Model efficiency evaluation model,the overall architecture and technical architecture scheme of the military equipment mobility evaluation system based on Hadoop in the meteorological environment were designed by combining existing technologies.Subsequently,based on the clarification of system requirements,the design of a system use case diagram and algorithmic structural class diagram based on UML was completed.Finally,in order to provide high-quality and high-efficiency data support to achieve the main functions of the system,data cleansing methods oriented towards Map Reduce and hierarchical storage methods oriented towards HBase + Hive were designed.(3)Based on the design of the military equipment mobility evaluation system in the meteorological environment,a scalable and high-performance Hadoop cluster environment was constructed by combining the Weather Trans-Model parallel prediction model.Secondly,the system’s data center was implemented based on the data processing workflow design process,and multiple data access interfaces were provided.Finally,utilizing the concept of front-end and back-end separation service architecture,the specific functions of the military equipment mobility evaluation system in the meteorological environment were implemented,including system management,data management,equipment management,model management,and 2D/3D visualization modules.This further validates the practical value of the proposed model in this dissertation. |
PDF Full Text Request