Electric Machine Predictive Torque Control for Fast Working Point Shifts

The patent applies preliminary action by pre-calculating and storing optimal control sequences (pilot control actions) for various desired working points in a lookup table during system initialization. When a working point change is required, the controller retrieves the pre-computed control sequence corresponding to the target working point and executes it directly, eliminating the need for real-time iterative optimization. This significantly reduces the response time for setting desired working points while maintaining controlled complexity through the use of pre-computed solutions.

The patent performs the computationally intensive predictive control calculations during system initialization or offline phases, storing the results in lookup tables. During runtime, when a working point change is needed, the controller simply retrieves the pre-computed control sequence from memory based on the desired working point, which requires minimal computational resources. This shifts the computational burden from runtime to initialization, achieving high productivity during operation with low energy consumption.

The patent uses the current actual working point as input to retrieve the appropriate pre-computed control sequence from lookup tables that were generated based on predictive control algorithms. This approach maintains high accuracy because the control sequences were pre-calculated using accurate predictive models, while avoiding the computational complexity of real-time predictive calculations. The system achieves precision by matching the current state to pre-analyzed scenarios stored in memory.