| Non-silicon MEMS has been increasingly used in military and civilian production, theassembly technology is the bottleneck of non-silicon MEMS development. In order toachieve high-precision and high-efficiency assembly of non-silicon MEMS parts, theconceptual design method and structural optimization technology of reconfigurablemicro-assembly systems are investigated in this study, in which the non-silicon MEMS ofsome small-caliber ammunition are selected as the research objects. Based on this study, weaim at providing micro-assembly system design methods and offering technical support forfurther efficient mass assembly of non-silicon MEMS, which will boost the corecompetitiveness of domestic micro/nano manufacturing.In this thesis, we first describe the research background of reconfigurablemicro-assembly systems and then review its development, design methods, optimizationtechnologies and applications at home and abroad. However, there are some problems ofreconfigurable micro-assembly systems, such as isolated various technical studies, lowersystem integration, poor reconfiguration, low assembly efficiency of the system and poordesign method. Based on these questions, we present theoretical and experimental studieson the conceptual design and structural optimization technology of reconfigurable assemblysystems. The main contents of this thesis include the following five aspects:(1) The overall technology of non-silicon MEMS reconfigurable micro-assemblysystems. First, we analyze the assembly characteristics and assembly needs of non-siliconMEMS. According to its bottlenecks of inefficient and lower precision, we propose theconcept of non-silicon MEMS reconfigurable micro-assembly systems, during which theconnotations, characteristics and its application objects are elaborated. Based on that,overall system structure is constructed and key technologies involved are analyzed, such asassembly system organizational structure configurations, transmission lines, automaticfeeding, high-precision position and orientation detection, posture adjustment,nondestructive flexible clamping, gas path, control technologies and solutions. Accordingto its functional requirements, each unit of the system including assembly subsystems,auxiliary subsystems and control subsystems can be quickly reconfigured to accommodatethe diversity of non-silicon MEMS construction. Hence, in order to realize thehuman-machine collaborative and/or fully automatic assembly of non-silicon MEMS, the system assembly precision should be less than5μm and the assembling beats must be lessthan30s.(2) Reconfigurable modular design approach for non-silicon MEMS micro-assemblysystems based on the fuzzy clustering method. First, the reconfigurable modular designprinciple of the micro assembly system is described. Then, the basic functionalrequirements are analyzed and improvements quality housing and assemblyneeds-sub-function matrix are built. According to that, product series spectrum is identified;multi-level function is decomposed; and the lowest level sub-functions are mapped to thecorresponding structural components. By considering the relevance of its functions,assembly objects, spatial configuration and system accuracy, the overall correlation matrixof system components is constructed and the module clustering is achieved by a fuzzyclustering method. Dividing module is appraised by the fuzzy evaluation method and theoptimal division results are obtained. Finally, the core of reconfigurable micro-assemblysystem-autonomous assembly unit is successfully taken as an example to do moduledivision.(3) Multi objective optimization method applied to non-silicon MEMS reconfigurablemicro-assembly systems based on modified genetic algorithm. For the reconfigurablemicro-assembly system, there are many forms of combinatorial optimization, topologyoptimization, size optimization and others. In addition, the optimization parameters includerational type variables, continuous variables and nonlinear function of multiple variables.Considering that, we propose a piecewise improved genetic algorithm based on floatingpoint and binary coding to realize the optimization of the multi objective function. Firstly,we describe the principles of the improved genetic algorithm including the hierarchicalhybrid coding method, multilayer segmentation single point crossover principle, doublemutation operation principle combining piecewise variation and non-uniform mutation, andthe procedure of the improved genetic algorithm. Then, the improved genetic algorithm isapplied to optimize the topological structure of the reconfigurable micro-assembly system.Considering the complex correlation of various modules realization scheme, we establishmulti objective optimization function model with four indicators including cost reliability,accuracy and efficiency and obtain the optimal topology based on the genetic algorithm.Compared with the results of the enumeration method, the improved genetic algorithm hasfast convergence and accuracy. Based on the determined basic topological structure of thesystem, the genetic algorithm is applied to optimize the overall size of the key components, gapless turntable based on the four bar linkage mechanism and autonomous assembly unit,in the system. The reliability of the optimization results for the turntable is verifiedtheoretical by MATLAB simulation and experimental by the measurements. The results ofexperimental test show the resolution of the turntable can reach3.5/10000°and therepeatability of positioning accuracy is less than3/1000°. For the size optimization of theautonomous assembly unit, we propose two order response surface method based on theorthogonal test to describe the relationship between design variables and response function.Then, the genetic algorithm is applied to obtain the optimal solution. At last, the reliabilityof genetic optimization is verified by probing the small range of the optimal solution.(4) Modal test research for dynamics characteristics of autonomous assembly unit.After module division, topologic structure optimization and key design parametersoptimizations for reconfiguration micro-assembly system, the author leads the researchteam to design the first national prototype of non-silicon MEMS reconfiguration assemblysystem. By using the prototype mentioned above, the dynamics characteristics on core unitof the reconfiguration micro-assembly system-autonomous assembly unit are studied;autonomous assembly unit dynamics finite element analysis (FEA) model is established;and the system natural frequency and its modes are calculated. Then, the modal test basedon MIMO method is carried out to verify the prototype’s vibration frequency and modes.Compared the FEA results with experimental results, the rationality of FEA model isvalidated, the designed micro-assembly system-autonomous assembly unit is with highstiffness and meets the design requirements. Finally, according to the modal test results,we’ve chosen optimal assembly data and made improvement recommendations on otherunits of the system.(5) Accuracy analysis and assembly experiment of autonomous assembly unit.According to characteristics of autonomous assembly unit with both static error anddynamic error, error propagation model is built based on multi-body system kinematics;and on this basis relative position, pose error among the part, detection module andsubstrate are deduced. Then, the errors to which the assembly accuracy is subtle aredetected and compensated. For straightness error of linear displacement station belongingto the assembly actuator module, offline compensation method of straightness error basedon reverse error compensation in vertical direction is proposed, the value of which is lessthan1μm after the compensation. For the depth of parallelism between the pallet andterminal of assembly actuator module, relative position and pose error detection and compensation method based on the six-dimension force sensor is proposed, the value ofwhich is less than0.03°. According to the errors mentioned above, assembly accuracy iscalculated and the value is±2.38μm. Finally, the assembly schedule of micro-assemblysystem is programmed and the assembly experiments are carried out. The experimentalresults indicate that the assembly success rate is90%and the micro-assembly system hashigh reliability and accuracy. |
PDF Full Text Request