MPC Vector Reduction for AC Motor Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The high process load of Model Predictive Control (MPC) methods in motor drive applications, particularly with two-level voltage source inverters, leads to increased costs and restrictions on control frequency and horizon, limiting their industrial adoption.
Innovation Solution
The method reduces the number of estimation vectors in MPC by dividing the α-β plane into sectors and selecting zero vectors based on current phase switches, reducing the calculation to 4 vectors from 7, thereby simplifying the control logic and process load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the standard MPC method is used to control two-level voltage source inverters, then control performance is maintained, but the process load and computational cost become excessively high
Solution Approach 1:
The patent divides the α-β plane into six sectors and further segments the estimation vectors based on sector location. By segmenting the calculation space and applying different estimation strategies for different sectors, the method reduces the overall computational burden while maintaining control performance across all operating conditions.
Solution Approach 2:
The patent applies partial action by calculating estimation vectors for only 4 out of 7 possible vectors in certain sectors. Instead of evaluating all 7 vectors uniformly across all sectors, the method selectively reduces calculations in specific sectors where certain vectors cannot be optimal, thereby reducing process load while preserving necessary control accuracy.
2Productivity
If the number of estimation vectors is reduced to lower process load, then computational cost decreases, but control accuracy may be compromised
Solution Approach 1:
The patent applies local quality by using different numbers of estimation vectors depending on the sector location. In sectors where certain vectors cannot be optimal, only 4 vectors are evaluated. In other sectors, all 7 vectors are evaluated to ensure accuracy. This localized adaptation of calculation depth maintains control accuracy where needed while improving efficiency where possible.
Solution Approach 2:
The patent changes the parameter of estimation vector quantity based on sector location. The number of vectors to be calculated is dynamically adjusted according to which sector the current is in, transforming a static calculation approach into a dynamic one that adapts to operating conditions, thereby balancing accuracy and efficiency.
3Reliability
If all 7 estimation vectors are calculated in MPC, then complete control coverage is achieved, but switching losses increase due to frequent switching
Solution Approach 1:
The patent applies partial action by calculating only 4 out of 7 estimation vectors in specific sectors. Since certain vectors cannot be optimal in certain sectors, evaluating them would result in unnecessary switching actions that increase switching losses. The method selectively omits these redundant calculations, reducing switching frequency and energy loss while maintaining complete control coverage through the remaining vectors.
Data Source
AI summary
A method for reducing a total operational load of a method of a model predictive control-by conducting simplifications based on specific observations, in order to drive alternating current motors by using the method of the MPC with a two-level voltage source inverter. The method includes the steps of determining at which one of the predefined sectors a resultant of stator currents is present, determining a motor mode, reducing seven estimation vectors to four estimation vectors and calculating a cost function or reducing seven estimation vectors to five estimation vectors and calculating the cost function.