Multilevel Converter Control via Predictive Switching Cell Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current multilevel converter modulation strategies face challenges in balancing capacitor voltages and switching losses, leading to harmonic distortions and unbalanced loss distributions, especially when dealing with a large number of switching levels, which requires significant computational resources, making them infeasible for many applications.
Innovation Solution
A method that analyzes each switching cell's first and second switching leg independently to determine their contribution to the transition voltage and internal conditions such as switching and conduction losses, allowing for the selection of capacitors to balance voltage levels and minimize losses, while reducing computational load by analyzing fewer states.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If carrier-based modulation with capacitor balancing control is used, then capacitor voltage balance is improved, but harmonic distortion increases
Solution Approach 1:
The method predicts the state of the control period from the present state and determines optimal switching actions in advance, rather than reactively adjusting capacitor voltages. This predictive approach allows for smoother transitions and reduced harmonic distortions while maintaining capacitor balance.
Solution Approach 2:
The method changes the control parameters by using predicted future states instead of present states for decision-making. By basing switching decisions on predicted capacitor voltages and current states, the system achieves better harmonic performance while maintaining voltage balance.
2Loss of energy
If reduced switching frequency modulation is used, then switching losses are minimized, but capacitor voltage balance deteriorates
Solution Approach 1:
The method performs preliminary prediction of capacitor voltages and switching states to determine optimal switching actions that maintain voltage balance. By predicting future states and planning switching actions in advance, the system achieves capacitor balance with fewer switching operations, thus reducing switching losses.
3Reliability
If MPC modulation strategy is used, then capacitor voltage balance is improved, but computational load increases exponentially
Solution Approach 1:
The method segments the complex MPC problem into simpler components by focusing prediction on only those switching cells that will change state. Instead of evaluating all possible switching combinations, the system identifies and predicts only the relevant switching actions, dramatically reducing computational load while maintaining capacitor voltage balance.
Solution Approach 2:
The method changes the approach from evaluating all possible switching states to predicting only the necessary switching actions based on present and future capacitor voltage states. This parameter change in the prediction methodology reduces computational complexity from exponential to manageable levels.
4Ease of manufacture
If conventional modulation strategies are used, then implementation is simple, but loss distribution among switches becomes unbalanced
Solution Approach 1:
The method performs preliminary prediction of switching states and capacitor voltages to identify optimal switching actions that balance losses among switches. By planning switching actions in advance based on predicted future states, the system achieves balanced loss distribution while maintaining implementation feasibility.
Data Source
AI summary
A method of controlling a multilevel converter, a computer program and a controller for a converter is provided. The method includes determining a transition voltage from a control period to a following control period, analyzing the switching cells of each phase leg, selecting capacitors to provide the transition voltage and synthesize the output voltage for the following control period, and connecting the selected capacitors during the following control period. The analyzing of each switching cell includes analyzing a first switching leg and a second switching leg of the switching cell. The method includes determining whether a change of state of the first switching leg would contribute to the direction of the transition voltage, and determining whether a change of state of the second switching leg would contribute to the direction of the transition voltage, and determining the internal conditions of each of the first and the second switching leg that are determined as contributing to the transition. Thus, the two legs of the H-bridge are analyzed separately. The selecting of capacitors is performed by selecting the capacitors of the switching legs of the phase leg for the transition voltage on the basis of the determined internal conditions of the switching legs, including comparing the internal conditions of all the switching legs of the phase leg. A controller is configured to control the multilevel converter by performing the method.