Mechanisms’ Comprehensive Analysis And Evaluation Oriented Function And Performance

The key to improve the mechanical properties and intelligent controlling level isto optimize the major mechanisms’ performance constantly, while the core problemsare how to evaluate the mechanisms’ performance reasonably and how to control themachine fine. Therefore, it’s important to research the analysis and evaluation methodfor mechanisms’ comprehensive performance based on single performance indexes.The principal research contents are as follows:(1) Applicability of Principal Component Analysis (PCA). The commonly usedstatistical evaluation methods are analyzed. Due to the correlation and diversity of themechanical single performance indexes, PCA is much more applicable forcomprehensive performance analysis and evaluation. Further, for the shortcomings oftraditional PCA, there are4extension methods introduced in mechanisms’comprehensive performance analysis and evaluation, including Kernel PrincipalComponent Analysis (KPCA), fuzzy principal component analysis (FPCA), globalanalysis based on PCA-BP neural network and Relative Principals ComponentsAnalysis (RPCA), and the algorithms are always proposed.(2) Comprehensive performance analysis and evaluation of typical open-chainmechanisms. Kinematic and dynamic single performance indexes of open-chainmechanisms are summarized, while their calculation method and physical meaningare explained. The planar serial robot and space serial robot of differentconfigurations and dimensions with specific tasks are researched respectively. Thecorrelations among each single performance indexes are analyzed, and thecomprehensive kinematic dexterity and comprehensive performance could beevaluated by KPCA, which provides scientific research basis for the optimum order ofconfigurations, dimensions and tasks. When the space robot finishes some definitetasks, the robot may be in singular configuration, the comprehensive kinematicdexterity should be evaluated by FPCA, and the extreme single performance indexesvalues of singularity could be absorbed to improve the accuracy and reliability ofevaluation results.(3) Comprehensive performance analysis and evaluation of typical closed-chainmechanisms. Kinematic and dynamic single performance indexes of closed-chain mechanisms are summarized, while their calculation method and physical meaningare explained. The PCA-BP neural network model for comprehensive performanceevaluation of crank-rocker mechanism is constructed, and then the comprehensiveperformance of crank-rocker mechanism within a certain range of dimensions couldbe predicted reasonably. The KPCA-BP neural network model for comprehensiveperformance evaluation of planar2-RR parallel mechanism is constructed, and thenthe comprehensive performance of planar2-RR parallel mechanism with differenttasks in a certain workspace could be predicted reasonably. Finally, the shaper’s mainmovement mechanisms of different configurations and dimensions with same tasksare researched to verify the effectiveness of comprehensive performance analysis andevaluation based on KPCA, which provides scientific research basis for the optimumorder of configurations and dimensions.(4) Simulation and comprehensive performance analysis and evaluation ofaircraft’s landing gear mechanism. Because the structure of landing gear mechanismis relatively complex, it’s difficult to analyze kinematics and dynamics performanceby analytic method. And the single performance indexes are calculated based onsimulation by CAD/CAE software. The parameterized virtual prototype modle ofaircraft’s landing gear mechanism is established by UG software, and then digitalprototype modle could be got in ADAMS software. After setting the parameters ofvirtual simulation environment, the single performance indexes values of landing gearmechanism with different dimensions could be calculated through kinematics anddynamics simulation. The comprehensive performance of landing gear mechanism isanalyzed and evaluated based on RPCA, and the dimension of landing gearmechanism with relative optimal comprehensive performance is chosen, which provesvalidity of the optimization design for aircraft’s landing gear mechanism combinedwith CAD/CAE technique and RPCA in engineering practice.The study above has developed a set of theory and method for mechanism’comprehensive performance analysis and evaluation based on PCA or its extensionmethods, analyzing the correlations among each single performance indexes, andproviding scientific research basis for the optimum order of tasks, mechanismstructure synthesis and dimensional synthesis.