Parallel Gradient Optimization for Electrical Converter Switching Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electrical converter control methods face challenges in achieving short response times, low harmonic distortion, and low switching losses due to high computational demands in solving optimization problems required for efficient control.
Innovation Solution
A method and controller that utilize a gradient-based optimization approach, implemented in parallel across multiple computing units, such as an FPGA or multi-core processor, to determine the next switching state by minimizing an objective function, thereby reducing computational complexity and improving control performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If standard QP solvers are used to determine switching instants in MP3C control, then optimal performance during transients and low harmonic current distortion are achieved, but high computational demands result in long response times
Solution Approach 1:
The controller is divided into multiple computing units, each capable of independently solving the quadratic programming problem. This parallel architecture segments the computational task, allowing simultaneous evaluation of multiple candidate switching states, thereby reducing overall computation time while maintaining optimal performance
Solution Approach 2:
The controller pre-calculates and stores optimized pulse patterns offline before real-time operation. During transient conditions, these pre-computed patterns serve as initial guesses or starting points for the QP solver, significantly reducing the computational burden and acceleration the response time during critical transient events
2Manufacturing precision
If model predictive control with quadratic programming is implemented for converter control, then optimal torque and flux control with low harmonic distortion are achieved, but high computational complexity increases device requirements
Solution Approach 1:
The control algorithm is segmented into modular computing units that can be independently implemented and configured. Each unit handles specific aspects of the MPC calculation (objective function evaluation, constraint checking, optimization), allowing the system to achieve optimal torque and flux control while distributing computational complexity across multiple simpler, specialized components
Solution Approach 2:
The controller dynamically adjusts prediction horizons, switching frequencies, and optimization parameters based on operating conditions. By changing these parameters adaptively, the system maintains high precision torque and flux control while reducing computational complexity during steady-state operation where full optimization is less critical
Data Source
Figure 1~2
Figure 3~5
Figure 4
AI summary
A method for controlling an electrical converter (12) comprises the steps of: receiving an actual electrical quantity (20) relating to the electrical converter (12) and a reference quantity (22); determining a future state of the electrical converter (12) by minimizing an objective function based on the actual electrical quantity and the reference quantity as initial optimization variables; and determining the next switching state for the electrical converter (12) from the future state of the electrical converter (12). The objective function is iteratively optimized by: calculating optimized unconstrained optimization variables based on computing a gradient of the objective function with respect to optimization variables; and calculating optimization variables for a next iteration step by projecting the unconstrained optimization variables on constraints. The computation of the gradient and/or the projection is performed in parallel in more than one computing unit (44, 50, 70).