Asynchronous State Estimation for High Bandwidth Motor Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional digital field-oriented control (FOC) architectures for electric motors are limited by the computational time required for state estimation, which restricts the control frequency and can lead to sacrifices in accuracy or the need for complex architectures to achieve high bandwidth motor control.
Innovation Solution
Implementing a control architecture where state estimation operates asynchronously with the control loop, allowing state estimates to be calculated as a background task and stored with timestamps, enabling the use of the most recent estimates for control signal generation without limiting the control frequency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If state estimation is performed synchronously with the control loop at high frequency, then control accuracy and bandwidth are improved, but computational time requirements increase and system complexity increases
Solution Approach 1:
The patent segments the control system into two independent parts: a state estimation function that operates asynchronously and a control loop that operates at the required control frequency. The state estimation function processes feedback data samples independently and stores results with timestamps, while the control loop retrieves the most recent estimates without being blocked by computation time. This segmentation resolves the contradiction by allowing each function to operate at its own optimal frequency without interfering with the other.
Solution Approach 2:
The state estimation function performs preliminary calculations of rotor position and velocity estimates in advance, storing these results with timestamps before they are needed by the control loop. This preliminary action allows the control loop to simply retrieve pre-computed values without waiting for computation, thereby maintaining high control frequency while avoiding the computational time burden during critical control intervals.
2Speed
If state estimation operates at the control loop frequency, then control bandwidth is improved, but device complexity increases due to computational requirements
Solution Approach 1:
The patent divides the computational workload by separating state estimation from the control loop. The state estimation function operates independently at a lower frequency, performing complex calculations only when needed and storing results with timestamps. The control loop operates at high frequency by simply retrieving pre-computed estimates, avoiding the need for complex real-time computation at control bandwidth frequencies.
Solution Approach 2:
The patent introduces an intermediary mechanism consisting of timestamped data storage between the state estimation function and the control loop. The state estimation function stores its results with timestamps, and the control loop retrieves the most recent estimate based on timestamp comparison. This intermediary allows high-speed control without requiring high-speed computation, reducing device complexity while maintaining control bandwidth.
3Manufacturing precision
If synchronous state estimation is implemented, then control precision is improved, but productivity decreases due to computational overhead
Solution Approach 1:
The patent segments the processing architecture so that state estimation and control operations can proceed independently without blocking each other. The state estimation function processes feedback data and generates rotor position and velocity estimates without waiting for the control loop, while the control loop continuously generates control signals using the most recent available estimates. This segmentation eliminates computational overhead that would otherwise reduce system throughput while maintaining control precision.
Solution Approach 2:
The patent ensures continuous operation of both state estimation and control loop functions without interruption. The state estimation function continuously processes feedback data samples and updates estimates in the background, while the control loop continuously generates control signals by retrieving the most recent estimates. This continuous operation of both functions simultaneously improves productivity by eliminating idle time while maintaining control precision through constant updates.
Data Source
AI summary
A motor controller architecture and method of operating the same. The motor controller asynchronously generates multiphase control signals for a multi-phase electric motor, relative to the estimation of various state parameters used in generating those control signals. Latency between the state estimation task and the control signal generation task is addressed by storing a timestamp with each input data sample from the sensors, and maintaining that timestamp with the output data from state estimation. Knowledge of the timestamp value allows the control task to update the state estimates to compensate for the time difference between the input data sample and the current sampling period.