Multidisciplinary Design Co-optimization For Mechatronics Product With6-DOF Manipulator As An Example

Product design is an intricated process, which involves multiple disciplines, and multiplesubsystems. Multidisciplinary design optimization(MDO) is a system optimization method，itexplorts the synergistic effects generated by the mutual coupling between the variousdisciplines or subsystems, to obtain the overall optimal solution, and to shorten the designcycle, which is an effective way to solve complex mechatronic system design. It has beendeveloped rapidly and widely over the past two decades in the field of aerospace, but has juststarted in the research field of complex engineering mechanical systems. This paper studiesthe issues of improved genetic algorithm with elitist strategy, design co-optimization withAnalytical Target Cascading (ATC), and product reliability analysis and optimal design underuncertainty. Co-optimization model is set up based on the Six Sigma design method,improved genetic algorithm and ATC, and is verified with a6-DOF manipulator as anexample. And main work in this paper is as following:1. Sum up and put forward a technology system of optimal design for complexmechanical product based on the indepth study achievements made at home and abroad, andon the basis of integration of the technical elements of the MDO, which gives an applicationframework of significance.2. Summarize and analyze the mathematical model of finite element modal analysis,modal extraction method, using the sensitivity analysis method combined with reducedeigenvalues and eigenvectors from Bayesian model based on monte carlo, and introducesGrid technology into the field of collaborative design, so as to improve the efficiency of theimplementation of the tasks effectively, and shorten the product development cycle. Itprovides a new solution for large-scale computation encountered in the field of collaborativedesign, and has a very good expansibility and portability.3. Reseach on the impact of the rigid-body motion of the flexible robot mechanism onelastic motion and elastic motion on a rigid body at the same time with an example of a6-DOF robot, to suppress vibration during the process of the machanism running a wide rangeof rigid-body motion, which is based on the completely inverse dynamics analysis. And itprovides clear and powerful evidence for further dynamic analysis and optimization.4. Solves the initial value compatibility issues from kinematics initial conditions whenanalysing the knowledge of computing multi-body dynamics and kinematics for the model of 6-DOF manipulator; solve the problem of large-scale sparse matrix multiplication for thecomplex multi-body system using generalized coordinates directly, which improvscomputational efficiency; solve the problem of low efficiency from large-scale equation byglobal modeling method for kinematics analysis, and by usual topological method using speedconversion matrix with virtual displacement and acceleration substituted into the equations ofmotion of the system set up by the global modeling method directly; consider elasticdeformation from the impact of the mechanism components and joint of the robot as arigid-flexible coupling body; sovle the intricacy and huge computaion problems fromderivating of the matrix and vector by Newmark and Newton-Raphson method usingmodified Newton-Raphson iteration correction formula for solving nonlinear dynamicsequations.5. Deduce a mathematical model of the two-phase hybrid stepping motor in the dqcoordinate system, on the theory basis of servo control system simulation and hybrid steppingmotor working principle, and build the6-DOF manipulator open loop simulation controlmodel, which provides a theoretical support for the dynamics/kinematics analysis, and lays afoundation for subsequent collaborative optimization.6. Research on decomposition&coordination technology, MDO mathematicaltheoretical model, Optimization strategies, and simplify the sensitivity information ofcoupling factors, construct processing system approximation model using global sensitivityequation, and then use optimization algorithm to find the optimal solution for the systemapproximation model till convergence; investigate on searching strategy for MDO space, aswell as its implementation ideas and theories, global sensitivity equation of coupled systembased on the principle of implicit function, satisfaction evaluation analysis for multi-objectivedecision and multi-objective optimization results; propose a combination optimization methodof genetic algorithm with elitist strategy, ATC and Six Sigma, which can improve thereliability and robustness.