Abstract:The optimization of multibody systems requires accurate and efficient methods for sensitivity analysis. The adjoint method is probably the most efficient way to analyze sensitivities, especially for optimization problems with numerous optimization variables. This paper discusses sensitivity analysis for dynamic systems in gradient-based optimization problems. A discrete adjoint gradient approach is presented to compute sensitivities of equality and inequality constraints in dynamic simulations. The constraints are combined with the dynamic system equations, and the sensitivities are computed straightforwardly by solving discrete adjoint algebraic equations. The computation of these discrete adjoint gradients can be easily adapted to deal with different time integrators. This paper demonstrates discrete adjoint gradients for two different time-integration schemes and highlights efficiency and easy applicability. The proposed approach is particularly suitable for problems involving large-scale models or high-dimensional optimization spaces, where the computational effort of computing gradients by finite differences can be enormous. Three examples are investigated to validate the proposed discrete adjoint gradient approach. The sensitivity analysis of an academic example discusses the role of discrete adjoint variables. The energy optimal control problem of a nonlinear spring pendulum is analyzed to discuss the efficiency of the proposed approach. In addition, a flexible multibody system is investigated in a combined optimal control and design optimization problem. The combined optimization provides the best possible mechanical structure regarding an optimal control problem within one optimization.

Hessian Chain Bracketing

A New Algorithm For Unconstrained Optimization Problem With The Form Of Sum Of Squares Minimization

Optimization of Generalized Jacobian Chain Products without Memory Constraints

A matrix algebra approach to approximate Hessians

Matrix-Free Jacobian Chaining

Second-Order Min-Max Optimization with Lazy Hessians

Second-order optimization with lazy Hessians

Useful Compact Representations for Data-Fitting

A structured diagonal Hessian approximation method with evaluation complexity analysis for nonlinear least squares

Review and Unification of Methods for Computing Derivatives of Multidisciplinary Computational Models

Potential Hessian Ascent: The Sherrington-Kirkpatrick Model

Scalable Subspace Methods for Derivative-Free Nonlinear Least-Squares Optimization

A discrete adjoint gradient approach for equality and inequality constraints in dynamics

Decentralized Implicit Differentiation

New Methods for Parametric Optimization via Differential Equations

Complexity results and active-set identification of a derivative-free method for bound-constrained problems

Model Hessian for accelerating first-principles structure optimizations

Higher-Order Corrections to Optimisers based on Newton's Method

Gradient and Hessian of functions with non-independent variables

A Hessian for Gaussian Mixture Likelihoods in Nonlinear Least Squares

A general approach to computing derivatives for Hessian-based seismic inversion