AC Rotary Machine Controller Using Adaptive Observer for Sensorless Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing controllers for AC rotary machines face challenges in maintaining high responsiveness and error convergence in low-speed regions, especially when initial value errors occur, and struggle with smooth control at zero speed and high rotational speeds due to limitations in phase difference sensitivity and current deviation calculations.
Innovation Solution
A controller that includes current vector detection, adaptive observation, and deviation vector calculation means to calculate and amplify current and magnetic flux deviations, allowing for accurate estimation of magnetic flux phases and voltage instructions to align detected and estimated current and magnetic flux vectors, even at zero speed and high rotational speeds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a position sensor with low responsiveness is used to reduce cost, then manufacturing cost is reduced, but control responsiveness deteriorates in high-speed regions
Solution Approach 1:
The patent introduces an adaptive observer as an intermediary system that estimates magnetic flux phase and current vectors without directly using the position sensor output. This mediator processes voltage and current signals to generate accurate position information, allowing the use of low-cost sensors while maintaining high-speed control responsiveness through mathematical estimation rather than direct sensor measurement.
2Device complexity
If sensor-less control is used to eliminate position sensors, then device complexity is reduced, but measurement precision deteriorates at low speeds
Solution Approach 1:
The patent implements a dynamic adaptive observer that adjusts its estimation behavior based on operating conditions. The system uses differential equations to continuously update current and magnetic flux estimates, with the estimation dynamics adapting to maintain precision across the full speed range from zero to high speeds, eliminating the need for position sensors while preserving measurement accuracy.
Solution Approach 2:
The adaptive observer incorporates feedback mechanisms where estimated current vectors and magnetic flux phases are continuously compared and adjusted based on voltage inputs and system models. This feedback loop ensures that even without direct position sensor input, the system maintains high measurement precision by constantly refining its estimates based on electrical measurements and mathematical relationships.
3Device complexity
If high-frequency voltage superposition is used for sensor-less control, then position sensor is eliminated, but energy loss increases due to non-fundamental frequency components
Solution Approach 1:
The patent extracts only the necessary fundamental frequency information for control by using an adaptive observer that processes voltage and current signals to estimate position and speed. Unlike methods that superimpose high-frequency voltages, this approach extracts position information from the natural fundamental frequency components of the motor operation, eliminating the need for additional high-frequency energy injection and thus avoiding the associated energy losses.
Data Source
Figure 1
Figure 2(A)~2(C)
Figure 3
AI summary
To obtain a controller for an AC rotary machine which performs a smooth drive from a low-speed region including zero speed to a high-speed region, and which can attain, even if an initial value error is present in an estimated phase, desired characteristics by promptly converging the error. To this end, provided are adaptive observation means 7 for outputting an estimated magnetic flux phase θ0, an estimated current vector Isest, an estimated magnetic flux vector Phest, and an estimated speed wr0, based on a voltage instruction vector Vsref, a current deviation vector ΔIs, and an amplified deviation vector E0; deviation vector calculation means 8 for calculating the current deviation vector ΔIs between the estimated current vector Isest and a detected current vector Is from current vector detection means 3, and a magnetic flux deviation vector ΔPhi between the estimated magnetic flux vector Phest and a detected magnetic flux vector Phi from magnetic flux vector detection means 6; and deviation amplifying means 9 for amplifying the current deviation vector ΔIs and the magnetic flux deviation vector ΔPhi, and outputting the result as the amplified deviation vector E0.