Vectorized Sparse Second-Order Forward Automatic Differentiation for Optimal Control Direct Methods
By: Yilin Zou, Fanghua Jiang
Potential Business Impact:
Makes computers solve hard problems faster.
Direct collocation methods are widely used numerical techniques for solving optimal control problems. The discretization of continuous-time optimal control problems transforms them into large-scale nonlinear programming problems, which require efficient computation of first- and second-order derivatives. To achieve computational efficiency, these derivatives must be computed in sparse and vectorized form, exploiting the problem's inherent sparsity structure. This paper presents a vectorized sparse second-order forward automatic differentiation framework designed for direct collocation methods in optimal control. The method exploits the problem's sparse structure to efficiently compute derivatives across multiple mesh points. By incorporating both scalar and vector nodes within the expression graph, the approach enables effective parallelization and optimized memory access patterns while maintaining flexibility for complex problems. The methodology is demonstrated through application to a prototype optimal control problem. A complete implementation for multi-phase optimal control problems is available as an open-source package, supporting both theoretical research and practical applications.
Similar Papers
Implicit differentiation with second-order derivatives and benchmarks in finite-element-based differentiable physics
Computational Engineering, Finance, and Science
Makes computer simulations solve problems faster.
Scalable Analysis and Design Using Automatic Differentiation
Numerical Analysis
Makes computer simulations of complex problems faster.
A second order numerical scheme for optimal control of non-linear Fokker-Planck equations and applications in social dynamics
Numerical Analysis
Helps predict and change how groups think.