# US 12,118,280 B1 · Electric drive unit simulation > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,118,280 B1 - Document ID: US12118280B1 - Status: Granted - Kind: B1 - Application number: 18/625,435 - Filed: 2024-04-03 - Granted: 2024-10-15 - Earliest claimed: 2023-08-10 - Assignees: Monumo Limited - Applicants: Monumo Limited - Inventors: Chun-Ting Lau; Daniel Bates; Kevin Bersch; William Gallafent; Jaroslaw Pawel Rzepecki; Alexey Kostin; Jonathan Rayner; Markus Kaiser; Nicolas Durrande; Rupert Tombs; Ian Murphy - Web page: https://patentgrep.com/patent/US12118280B1 ## CPC - G06F 30/27: using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model - G06F 30/15: Vehicle, aircraft or watercraft design - G06F 30/17: Mechanical parametric or variational design - G06F 30/20: Design optimisation, verification or simulation (optimisation, verification or simulation of circuit designs G06F30/30) - G06F 2119/06: Power analysis or power optimisation - G06F 2119/08: Thermal analysis or thermal optimisation - G06F 2119/18: Manufacturability analysis or optimisation for manufacturability - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract A computer-implemented method of simulating operation of an electric drive unit to predict one or more performance parameters of the electric drive unit is provided. The electric drive unit comprises at least an electric motor. The method comprises obtaining parameters defining physical properties of the electric motor, obtaining parameters defining drive currents for driving the electric motor, processing the obtained parameters using a machine learning module trained a priori to predict a spatially varying electromagnetic and/or mechanical and/or thermal profile within the electric motor during operation, and providing as output a predicted profile for the electric motor, and using the predicted profile to compute the one or more performance parameters of the electric drive unit. ## Claims ### Claim 1 A computer-implemented method of simulating operation of an electric drive unit to predict one or more performance parameters of the electric drive unit, the electric drive unit comprising at least an electric motor, the method comprising: obtaining parameters defining physical properties of the electric motor; obtaining parameters defining drive currents for driving the electric motor; processing the obtained parameters using a machine learning module trained a priori to predict a spatially varying electromagnetic and/or mechanical and/or thermal profile within the electric motor during operation, and providing as output a predicted profile for the electric motor and a parameter indicating an uncertainty associated with the predicted profile, determining that the uncertainty associated with the predicted profile is greater than a threshold; generating, in response to determining that the uncertainty associated with the predicted profile is greater than the threshold, a new predicted profile comprising a training electromagnetic and/or mechanical and/or thermal profile within the electric motor during operation generated by solving a system of predefined equations for the obtained parameters; performing additional training of the machine learning module using the new predicted profile and the obtained parameters; and using the new predicted profile to compute the one or more performance parameters of the electric drive unit. ### Claim 2 (depends on 1) The method according to claim 1, wherein the using the new predicted profile to compute the one or more performance parameters of the electric drive unit comprises one of: a) solving a further system of predefined equations for the new predicted profile; and b) processing the new predicted profile using a further machine learning module trained a priori to compute one or more performance parameters of an electric drive unit. ### Claim 3 (depends on 2) The method according to claim 2, wherein the further machine learning module is trained using a training set comprising a plurality of training items, each training item comprises one or more performance parameters generated by solving the further system of predefined equations for the respective training profile. ### Claim 4 (depends on 1) The method according to claim 1, wherein said spatially varying electromagnetic and/or mechanical and/or thermal profile within the electric motor during operation specifies, for each of a plurality of locations within the electric motor, one of: a magnetic flux density; a magnetic field strength; a magnetic vector potential; a Maxwell stress tensor; a mechanical displacement/velocity/acceleration; an energy or power loss; and a temperature. ### Claim 5 (depends on 1) The method according to claim 1, wherein the machine learning module is trained using a training set comprising a plurality of training items, each training item comprises a training electromagnetic and/or mechanical and/or thermal profile within the electric motor during operation generated by solving the system of predefined equations for the parameters. ### Claim 6 (depends on 1) The method according to claim 1 wherein the one or more performance parameters being selected from a group comprising: one or more electromagnetic performance parameters; one or more mechanical performance parameters; one or more thermal performance parameters; and one or more energy performance parameters. ### Claim 7 (depends on 1) The method according to claim 1, wherein the one or more performance parameters comprise one or more of a torque of the electric motor, a torque ripple, and an inverter loss. ### Claim 8 (depends on 1) The method according to claim 1, wherein the parameters defining physical properties of the electric motor comprise parameters specifying a geometry of the electric motor. ### Claim 9 (depends on 8) The method according to claim 8, wherein the electric motor comprises a rotor and a stator, and the parameters specifying the geometry of the electric motor comprise parameters specifying a geometry of the rotor and/or the stator. ### Claim 10 (depends on 9) The method according to claim 9, wherein the parameters specifying a geometry of the electric motor specify a stator tooth geometry and/or a rotor tooth geometry. ### Claim 11 (depends on 1) The method according to claim 1, wherein the electric drive unit whose operation is to be simulated comprises an inverter for providing said drive currents. ### Claim 12 (depends on 1) A computer-implemented method of optimising a design of an electric drive unit with respect to one or more performance parameters, the electric drive unit comprising at least an electric motor, the design comprising parameters defining physical properties of the electric motor and parameters defining drive currents for driving the electric motor, the method comprising: jointly optimising the parameters defining physical properties of the electric motor and the parameters defining the drive currents over a plurality of iterations, wherein each iteration comprises: performing the method of claim 1 to predict the one or more performance parameters of the electric drive unit; and adjusting the parameters defining physical properties of the electric motor and the parameters defining the drive currents with an objective of improving the one or more performance parameters. ### Claim 13 (depends on 12) The method according to claim 12, wherein the parameters are adjusted based on a genetic optimisation algorithm. ### Claim 14 (depends on 12) The method according to claim 12, wherein the parameters are adjusted based on a Bayesian optimisation algorithm. ### Claim 15 (depends on 12) The method according to claim 12 further comprising outputting the optimised design for manufacture of a corresponding the electric drive unit. ### Claim 16 (depends on 12) A method of manufacturing an electric drive unit comprising: implementing the method according to claim 12 to obtain optimised parameters defining physical properties of the electric motor and the parameters defining the drive currents; and manufacturing the electric drive unit according to the obtained parameters. ### Claim 17 (depends on 1) A non-transitory computer readable medium comprising executable instructions that, when executed by a processor, cause the processor to implement the method according to claim 1.