Method and apparatus for estimating electromagnetic forces active in an electric machine

The invention relates to a method and apparatus for estimat- ing electromagnetic forces active in an electric machine.

The suppression of undesirable noise and vibration caused by electric machines, in particular by rotating electric ma- chines has been an important research issue in auto industry. It is due to the increasing ubiquity of electrified compo- nents inside a passenger vehicle. The electromagnetic forces in electric machines can be responsible for high vibration and/or noise, which may result in acoustic disturbances for the passengers or to any other person around the electric ma- chine. Reducing high vibration and/or noise needs a predic- tion of the electromagnetic forces, in particular, in the ar- ea where they occur.

Measuring electromagnetic (EM) forces in rotating electric machines through common sensoring technologies is not feasi- ble. The electromagnetic forces occur between the stator and the rotor of the electric machines, in particular between the rotor teeth and stator teeth. A common sensoring technology for measuring electromagnetic forces is called "force trans- ducer". The forces determined in electric machines have mag- nitudes of several tenths of kilo-Newtons . For that, the transducers are larger than 30 mm of size, while the air gap between the stator and rotor is usually between 0,3 mm and 3 mm which is a first limitation for usage. A second limitation of these common sensoring comes from the fact that, due to their functioning principle (electrical resistor forms strain gauge) , they are sensible to the magnetic field, generated at the air gap, which is significantly high at these air gaps of the rotating electric machines and will substantially affect the sensor outputs. Fig. 2 shows an exemplary electric machine comprising a sta- tor St and a rotor Ro . In the stator, the stator teeth St_Te are formed. In the rotor, the rotor teeth Ro_Te are formed. Between the stator teeth and the rotor teeth, the air gap is formed. The air gap size is usually between 0,3 mm to 3 mm.

Due to the aforementioned reason, electromagnetic forces are never measured. Eventually, a way that is sometimes used to estimate electromagnetic forces is to measure the polyphase currents that would feed an electromagnetic finite element model for magneto static analyses. However, using this pro- cess, the physical output, e.g. vibrations of such estimated forces are not correlated with the physical input, e.g. cur- rent, making the estimation of the electromagnetic forces not completely robust nor validated.

Accordingly, it is an object of the present invention to pro- vide an improved and efficient method and an apparatus which allows prediction of electromagnetic forces active in elec- tric machines.

This object is achieved according to a first aspect of the present invention by a method for estimating electromagnetic forces active in an electric machine during operation of the electric machine comprising the features of claim 1.

The invention provides according to the first aspect a method for estimating electromagnetic forces active in an electric machine during operation of the electric machine, the method comprising the steps of:

Measuring at least one first operation parameter of the elec- tric machine while the electric machine is operated under at least one operational condition, and

Estimating electromagnetic forces active in an electric ma- chine during operation of the electric machine by multiplying the measured at least one first operation parameter and a re- spective second operation parameter provided by a stored structural/vibro-acoustic model.

In connection with the present invention, an electric machine can be used for the traction of an electric vehicle. In a further arrangement, the electric machine can include auxil- iaries of the vehicle, such as, air compressor, hybrid boost- er etc. The method and apparatus of the present invention can also be used to estimate electromagnetic forces within high speed electric machines.

In connection with the present invention, "during operation of the electric machine" includes that the electric machine is operating with a specific rotational speed and a specific torque value dependent on the operation task and load con- nected to the electric machine. The electric machine can be operated under several configurable rotational speed values and/or several configurable torque values.

The electromagnetic force is responsible for most of the in- teractions that can be seen in our environment. The electro- magnetic force holds electrons in their orbit around the nu- cleus. These electrons interact with other electrons to form electron bonds among elements and produce molecules and, eventually, visible matter. The electromagnetic force, also called the Lorentz force, explains how both moving and sta- tionary charged particles interact. The electromagnetic force includes the formerly distinct electric force and the magnet- ic force.

In connection with the present invention, a "structur- al/vibro-acoustic model" can be understood to be an abstract model that organizes elements of data and standardizes how these elements relate to one another and to properties of the electric machine. It can describe a dataset relevant to the electric machine in a specific manner as well as comprises data describing and/or specifying functionalities and/or the relations of the electric machine. In particular, the struc- tural/vibro-acoustic model comprises data that can be updat- ed. The structural/vibro-acoustic model can be implemented in a database stored on a local or centralized computer. The da- tabase can be further stored in a distributed database on a distributed server system or cloud system.

With the method and apparatus according to the present inven- tion it is possible to estimate forces occurring in the elec- tric machine without relying on a non-updated electromagnetic model. Moreover, the presented technique is quick and allows system level integration of electric machines whining noise simulation and later enabling condition monitoring purposes to be performed in real-time environment.

In addition, the present invention allows by estimating the forces in electric machines monitoring operation condition of the electric machines and trouble shooting in case of detect- ed unknown and/or undesired electromagnetic forces within the electric machine.

Further, the present invention may support applications where conformity is quite important. For instance, in the car in- dustry the electromagnetic forces and/or force shapes, which has the greatest contribution on the annoying noise component can be detected. If these are detected or identified, it is possible to apply several control strategies to monitor these electromagnetic forces as well as it is possible to reconsid- er, modify and/or redesign the structure of the component and/or electric machine causing the electromagnetic force.

Due to the detected or identified electromagnetic forces, a modification of the structure can be used to remove resonanc- es or shift the resonances to higher frequencies.

By the present invention, it is possible to detect which electromagnetic force affect which kind of frequencies. As a relations of the electric machine. In particular, the struc- tural/vibro-acoustic model comprises data that can be updat- ed. The structural/vibro-acoustic model can be implemented in a database stored on a local or centralized computer. The da- tabase can be further stored in a distributed database on a distributed server system or cloud system.

With the method and apparatus according to the present inven- tion it is possible to estimate forces occurring in the elec- tric machine without relying on a non-updated electromagnetic model. Moreover, the presented technique is quick and allows system level integration of electric machines whining noise simulation and later enabling condition monitoring purposes to be performed in real-time environment.

In addition, the present invention allows by estimating the forces in electric machines monitoring operation condition of the electric machines and trouble shooting in case of detect- ed unknown and/or undesired electromagnetic forces within the electric machine.

Further, the present invention may support applications where conformity is quite important. For instance, in the car in- dustry the electromagnetic forces and/or force shapes, which has the greatest contribution on the annoying noise component can be detected. If these are detected or identified, it is possible to apply several control strategies to monitor these electromagnetic forces as well as it is possible to reconsid- er, modify and/or redesign the structure of the component and/or electric machine causing the electromagnetic force.

Due to the detected or identified electromagnetic forces, a modification of the structure can be used to remove resonanc- es or shift the resonances to higher frequencies.

By the present invention, it is possible to detect which electromagnetic force affect which kind of frequencies. In a further possible embodiment, at least one unitary force is applied to the structural/vibro-acoustic model and the at least one unitary force corresponds to the respective at least one intrinsic behavior of the electric machine. The up- dated structural/vibro-acoustic model is used in conjunction with unitary force waves representative of the typical loads of an electric machine to obtain normalized vibration/acous- tic output. The force waves are manually generated. All the force waves represent a particular shape component that are all put together and form the total force wave. Each force amplitude of the force waves is normalized to generate a unit energy load for every frequency line of a broadband spectrum. Thus, applying these forces to the updated model, the vibra- tion/acoustic output is obtained for each force shape.

In a further possible embodiment, the operational condition comprises a plurality of certain torque parameters of the electric machine and/or a plurality of certain speed parame- ters of the electric machine. In this way, in a testing phase data are measured from a running electric machine under oper- ational condition to receive real life data. The received re- al life data provides noise and/or vibration information within certain torque parameters and/or certain speed parame- ters of the electric machine. Specific noise or vibration can be detected and analyzed as well as structural defects and disturbance resulting from the detected noise or vibration may be identified and remedied. The operational condition can be configurable or comprises a static value.

In a further possible embodiment, the operational condition is configurable by using a ramping up or slowing down func- tion. In this way, the operational condition is dynamically configurable by varying the speed by ramping up or slowing down .

In a further possible embodiment, the at least first opera- tion parameter comprises at least acoustic pressure and/or vibration values caused by acceleration of the electric ma- chine. Acoustic pressure and vibration values caused by ac- celeration of the electric machine can be measured. The ac- celeration can be configurable, such as, each acceleration value that cause acoustic pressure and vibration can be meas- ured . In a further possible embodiment, the at least first opera- tion parameter comprises at least acoustic pressure and/or vibration values caused by velocity of the electric machine. Acoustic pressure and vibration values caused by velocity of the electric machine can be measured. The velocity can be configurable, such as, each acceleration value that cause acoustic pressure and vibration can be measured.

In a further possible embodiment, the at least first opera- tion parameter comprises at least acoustic pressure and/or vibration values caused by displacement of the electric ma- chine. Acoustic pressure and vibration values caused by dis- placement of the electric machine can be measured. In this way, acoustic pressure and vibration caused by displacement of the electric machine can be measured.

In a further possible embodiment, an accelerometer is used to measure the vibration values. An accelerometer is a sensor that measures its acceleration. This is usually done by de- termining the inertial force acting on a test mass. Thus, for example, it can be determined whether there is an increase or decrease in speed. The acceleration estimation may depend on the accelerometer technology. Usually Piezoelectric (PE) ac- celerometers are used and use the fact that the instantaneous stress (generated by changes in acceleration) on the PE ele- ment produces a charge that is proportional to the accelera- tion. This charge goes to the electrical terminals of the sensor, thus giving voltage to measure digital signal. In this way the acceleration can be measured and provided as a digital signal.

In a further possible embodiment, a microphone is used to measure the acoustic pressure. A microphone is a sound trans- ducer that converts airborne sound as sound pressure oscilla- tions into corresponding electrical voltage changes as a mi- crophone signal. In this way the acoustic pressure can be measured and provided as a digital signal. In a further possible embodiment, the structural/vibro- acoustic model comprises the shape of the electric machine. A shape can be a mode shape and can refer to shapes and natural frequencies. Usually, the shapes and natural frequencies are obtained by solving the so-called equation of motion of the system (electric machine) . They form the eigen-vectors and eigen-frequencies of the mathematical system.

In a further possible embodiment, the structural/vibro- acoustic model comprises the structure of the electric ma- chine. The structure describes the electric machine compo- nents assembled together (stator, rotor, winding, housing, etc.) . They are modelled independently or together, depending on the model used. The models also require the material me- chanical properties of each component (stiffness, mass, damp- ing) .

In a further possible embodiment, the structural/vibro- acoustic model comprises the pressure of the structure from the electric machine. The vibration of the machine generates vibrations of the medium around it (air usually) , which pro- duces pressure fluctuation, eventually leading to acoustic noise. The acoustic pressure model then requires the sur- rounding medium properties, and is usually calculated away from the machine.

In a further possible embodiment, the experimental modal analysis comprises an application of an external physical force to the electric machine. In this way, the reaction of the electric machine in particular to the external physical force can be observed and measured. The external physical force can by physical stress caused by a hammer or shaker to the electric machine. The experimental modal analysis deals with the functional relationships between the technique and the environment. The experimental modal analysis is used to describe, explain, predict and control behavior. The behavior can be understood by the function of the behavior and in what context the behavior occurs.

In a further possible embodiment, the experimental modal analysis comprises an application of an external physical force to single components of the electric machine. In this way reaction of, in particular a single component of the electric machine due to the affect caused by the external physical force can be measured.

In a further possible embodiment, the experimental modal analysis comprises measuring the vibration/acoustic output of the electric machine. The experimental modal analysis is used to update the stored structural/vibro-acoustic model. In the experimental modal analysis, the shape and frequencies be- tween the two aspects (experimental and model) are correlat- ed .

Further, the stored structural/vibro-acoustic model is used in conjunction with unitary forces to obtain the normalized vibration/acoustic output. By applying the unitary forces to the stored structural/vibro-acoustic model, the vibra- tion/acoustic output is obtained for each force shape.

In a further possible embodiment, the experimental modal analysis comprises estimating the intrinsic behavior of the electric machine. Each structure does have any intrinsic be- havior that does not depend on forces. The intrinsic behavior can be used to update the stored structural/vibro-acoustic model .

In a further possible embodiment, the at least one unitary force represents certain loads of an electric machine. These unitary forces comprise forces of one Newton amplitude. These unitary forces are generated with particular shapes and ap- plied to the stored structural/vibro-acoustic model. In this way, for each applied unitary force, one vibration response is supplied corresponding to a specific force shape. The electromagnetic force can be reconstructed from the specific force shapes.

Up to now, the invention has been described with respect to the claimed method. Features, advantages or alternative em- bodiments herein can be assigned to the other claimed objects (e.g. the computer program or a device, i.e. the apparatus or a computer program product) and vice versa. In other words, the subject matter which is claimed or described with respect to the device can be improved with features described or claimed in the context of the method and vice versa. In this case, the functional features of the method are embodied by structural units of the system and vice versa, respectively. Generally, in computer science a software implementation and a corresponding hardware implementation are equivalent. Thus, for example, a method step for "storing" data may be per- formed with a storage unit and respective instructions to write data into the storage. For the sake of avoiding redun- dancy, although the apparatus may also be used in the alter- native embodiments described with reference to the method, these embodiments are not explicitly described again for the apparatus .

The invention further provides according to a second aspect an apparatus for estimating electromagnetic forces active in an electric machine during operation of the electric machine comprising the features of claim 13.

The invention further provides according to the second aspect an apparatus for estimating electromagnetic forces active in an electric machine during operation of the electric machine, comprising :

a measuring unit, adapted to measure at least one first oper- ation parameter of the electric machine while the electric machine is operated under at least one operational condition, and

an estimating unit, adapted to estimate electromagnetic forc- es active in an electric machine during operation of the electric machine by multiplying the measured at least one first operation parameter and a respective second operation parameter provided by a stored structural/vibro-acoustic mod- el .

The invention further provides according to a third aspect a system for estimating electromagnetic forces active in an electric machine during operation of the electric machine comprising the features of claim 13.

The invention further provides according to the third aspect a system for estimating electromagnetic forces active in an electric machine during operation of the electric machine comprising :

- an electric machine, wherein the electric machine comprises an electric motor powered by direct current sources or by al- ternating current; and

- an apparatus according to claim 13.

The invention further provides according to a fourth aspect a computer program product for estimating electromagnetic forc- es active in an electric machine during operation of the electric machine comprising the features of claim 15.

In the fourth aspect the invention relates to a computer pro- gram product comprising a computer program, the computer pro- gram being loadable into a memory unit of a computing unit, including program code sections to make the computing unit execute the method for estimating electromagnetic forces ac- tive in an electric machine during operation of the electric machine according to the first aspect of the invention, when the computer program is executed in said computing unit.

It is part of the invention that not all steps of the method necessarily have to be performed on the same component or computer instance, but can also be performed on different computer instances. In addition, it is possible that individual steps of the method described above can be carried out in one unit and the remaining components in another unit, as a distributed sys- tem.

The properties, features and advantages of this invention de- scribed above, as well as the manner they are achieved, be- come clearer and more understandable in the light of the fol- lowing description and embodiments, which will be described in more detail in the context of the drawings. This following description does not limit the invention on the contained em- bodiments. Same components or parts can be labeled with the same reference signs in different figures. In general, the figures are not for scale. It shall be understood that a pre- ferred embodiment of the present invention can also be any combination of the dependent claims or above embodiments with the respective independent claim.

These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments de- scribed hereinafter.

In the following possible embodiments of the different as- pects of the present invention are described in more detail with reference to the enclosed figures.

Fig. 1 shows a flowchart of a possible exemplary embodi- ment of a method for estimating electromagnetic forces active in an electric machine during opera- tion of the electric machine;

Fig. 2 shows an exemplary block diagram of an electric ma- chine ;

Fig. 3 shows a block diagram of a possible exemplary em- bodiment of an apparatus for estimating electromag- netic forces active in an electric machine during operation of the electric machine; Fig. 4 shows a block diagram of a possible exemplary em- bodiment of a system for estimating electromagnetic forces active in an electric machine during opera- tion of the electric machine;

Fig. 5 shows a schematic diagram of a specific example for illustrating the operation of a method and an appa- ratus estimating electromagnetic forces active in an electric machine during operation of the elec- tric machine.

Fig. 1 shows schematically a flowchart of a possible exempla- ry embodiment of a method for estimating electromagnetic forces active in an electric machine during operation of the electric machine. Electromagnetic forces active has to be un- derstood as forces that occurs while the electric machine 20 is under current and voltage as well as in motion, and there- fore in operation. The operation can be with or without load connected to the electric machine 20.

The method comprises in the illustrated exemplary embodiment several main steps. In a first step SI, at least one first operation parameter Y _tot of the electric machine 20 is meas- ured while the electric machine 20 is operated under at least one operational condition. The at least one first operation parameter Y _tot depending on the frequency w describes the vi- bration response comprising, e.g. acceleration, velocity, displacement, or the acoustic response, e.g. pressure, meas- ured on the outer surface of the electric machine 20 and around the electric machine 20 under operation of the elec- tric machine 20. The frequency w is the frequency (in rad/s) at which the measurement and simulation of the method accord- ing to the present invention are evaluated.

In a further step S2, the electromagnetic forces F _ot active in an electric machine 20 during operation of the electric machine 20 are estimated by multiplying the measured at least one first operation parameter Y _tot and a respective second op- eration parameter provided by a stored structural/vibro- acoustic model M. The second operation parameter provided by the stored structural/vibro-acoustic model M may comprise the vibration or acoustic response obtained from the vibro- acoustic analysis performed using the applied unitary forces F _{u, I} and the stored structural/vibro-acoustic model M. The ap- plied unitary forces F _{u, I} are as well dependent of the fre- quency w. The unitary forces F _{u, I} are unitary forces of ampli- tude 1 N applied to the stored structural/vibro-acoustic mod- el M, which corresponds to spatial shapes with 1, 2, 3,...

lobes respectively. The electromagnetic forces F _ot comprise the total estimated force that applies under operating condi- tions .

Fig. 3 shows schematically a block diagram of a possible ex- emplary embodiment of an apparatus 10 for estimating electro- magnetic forces active in an electric machine 20 during oper- ation of the electric machine 20.

As can be seen from the block diagram of Fig. 3, the appa- ratus 10 is used for estimating electromagnetic forces F _ot active in an electric machine 20 during operation of the electric machine 20.

The apparatus 10 comprises in the illustrated embodiment a measuring unit 11 and an estimating unit 12. The measuring unit 11 is adapted to measure at least one first operation parameter Y _tot of the electric machine 20 while the electric machine 20 is operated under at least one operational condi- tion. The estimating unit 12 is adapted to estimate electro- magnetic forces F _tot active in an electric machine 20 during operation of the electric machine 20 by multiplying the meas- ured at least one first operation parameter Y _ot and a respec- tive second operation parameter provided by a stored struc- tural/vibro-acoustic model M. The apparatus for estimating electromagnetic forces F _ot ac- tive in an electric machine 20 during operation of the elec- tric machine 20 may be a computer, personal computer or a workstation in a computer network and includes a central pro- cessing unit, a system memory, and a system bus that couples various system components including the system memory to the central processing unit. The system bus may be any of several types of bus structures including a memory bus or memory con- troller, a peripheral bus, and a local bus using any of a va- riety of bus architectures. The system memory may include read only memory (ROM) and/or random-access memory (RAM) . A basic input/output system (BIOS), containing basic routines that help to transfer information between elements within the personal computer, such as during start-up, may be stored in ROM. The computer may also include a hard disk drive for reading from and writing to a hard disk. The hard disk drive may be coupled with the system bus by a hard disk drive in- terface. The drive and its associated storage media provide nonvolatile storage of machine-readable instructions, data structures, program modules and other data for the computer. Although the exemplary environment described herein employs a hard disk, those skilled in the art will appreciate that oth- er types of storage media, such as flash memory cards, digi- tal video disks, random access memories (RAMs) , read only memories (ROM) , and the like, may be used instead of, or in addition to, the storage devices introduced above. A number of program modules may be stored on the hard disk, ROM or RAM, such as an operating system, one or more application programs, like the method for estimating and/or other program modules, and/or program data for example.

The measuring unit 11 and the estimating unit 12 may be im- plement as a single instance or as different instances on the central processing unit. In a further example, the measuring unit 11 and the estimating unit 12 may be implemented on dif- ferent central processing units of a computer and/or on chips adapted to incorporate the measuring unit 11 and estimating unit 12. In a further embodiment, the apparatus 10 may comprise an in- terface adapted to connect an accelerometer to measure the vibration values and/or to connect a microphone to measure the acoustic pressure.

In a further embodiment, the apparatus 10 may comprise a memory to store the structural/vibro-acoustic model M.

Fig. 4 shows a block diagram of a possible exemplary embodi- ment of a system 100 for estimating electromagnetic forces active in an electric machine 20 during operation of the electric machine 20.

As can be seen from the block diagram of Fig. 4, the system 100 is used for estimating electromagnetic forces F _ot active in an electric machine 20 during operation of the electric machine 20.

The system 100 comprises an electric machine 20. The electric machine 20 comprises in an embodiment an electric motor pow- ered by direct current sources. In an alternative embodiment, the electric machine 20 comprises an electric motor powered by alternating current. The system 100 further comprises an apparatus 10 according to the present invention.

Fig. 5 shows a schematic diagram of a specific example for illustrating the operation of a method 1 and an apparatus 10 estimating electromagnetic forces active in an electric ma- chine 20 during operation of the electric machine 20.

In Fig. 5 the operation of the method 1 according to the pre- sent invention is shown comprising four operation steps a to d. The method according to the present invention combines a testing phase and a simulation phase. The testing phase com- prises the operation steps a and b and the simulation phase comprises operation step c. From the results of the testing phase and the simulation phase, the electromagnetic forces F _tot can be estimated in the operation step d. The sequence of the operational steps a to c can be varied if necessary.

In operational step a, vibration data and/or acoustic data specific for an electric machine 20 to be evaluated are ex- perimentally gathered for at least one operational condition. The operational condition may comprise a plurality of certain torque parameters of the electric machine 20 and/or a plural- ity of certain speed parameters of the electric machine 20. The operational step a is performed with a running electric machine under current and voltage, wherein the operational condition is configurable and can be configured by using a ramping up or slowing down function. Advantageously, the op- erational step a can provide real life data from the electric machine 20. While the operational step a is performed, at least one first operation parameter Y _tot can be measured. The at least one first operation parameter Y _tot may comprise at least acoustic pressure and/or vibration values. The first operational parameter Y _tot depends on the frequency w . The first operation parameter Y _tot (w) is the vibration response (acceleration, velocity, displacement) or acoustic response (pressure) measured on the outer surface of the machine and around the electric machine under operation. The frequency w is the frequency (in rad/s) at which the measurement and sim- ulation are evaluated. The acoustic pressure and/or vibration values may be caused by at least one of acceleration of the electric machine 20, velocity of the electric machine 20, and/or displacement of the electric machine 20. The accelera- tion of the electric machine can be measured, for instance by an accelerometer and the acoustic pressure can be measured, for instance by a microphone, which is placed around the electric machine 20. In the operational step a, a testing is performed comprising changes in speed and torque of the elec- tric machine 20 to find the required operating point with the vibration and noise, which are disturbing. The operational step a may be performed in method step SI. In operational step b, an experimental modal analysis is per- formed. With respect to the experimental modal analysis a structural/vibro-acoustic model M is generated and/or updat- ed. The structural/vibro-acoustic model M is generated by us- ing at least one intrinsic behavior of the electric machine 20 determined within the experimental modal analysis. The ex- perimental modal analysis is performed while the electric ma- chine 20 is in a stationary state. In the experimental modal analysis, the intrinsic behavior of the structure (electric machine 20), which does not depend on any forces is tested. The intrinsic behavior describes a specific behavior of the electric machine 20 under a specific condition, while the ma- chine is in a stationary state. In the stationary state, the electric machine 20 is without current and voltage as well as without any motion. The specific condition (represented in Fig. 5 by the flash) may comprise a physical stress on the structure caused for instance by a hammer. Further tools and/or actions may be used to detect the intrinsic behavior of the electric machine. After hitting the electric machine or structure, for instance with a hammer, the vibration data and/or pressure data created by the electric machine can be detected and collected. A software- and analysis algorithm can be used to determine the corresponding shapes (oval cir- cle) for the detected and collected vibration data and/or pressure data. The shapes are determined at different fre- quencies co . The shapes and frequencies w are used to update. The structural/vibro-acoustic model M can be stored in a da- tabase locally or centralized.

In operational step c, a structural/vibro-acoustic model M update is performed (circle arrow) , so that the structur- al/vibro-acoustic model M fits exactly what has been measured in operational step b. The correct shapes and frequencies co, in particular the natural frequencies of the structure meas- ured in operational step b are embedded in the structur- al/vibro-acoustic model M. Further, in operational step c, unitary forces F _u,I (w) are applied to the structural/vibro- acoustic model M. The unitary forces F _u,I (w) are generated manually with the particular shape corresponding to the simi- lar shape generated in operational step b. The unitary forces F _{u, I} ((o) are unitary forces of an amplitude 1 N applied to the structural/vibro-acoustic model M, which corresponds to the spatial shapes. For instance, if shapes 1, 2, and 3 are de- termined, three unitary forces F _{u, I} (w) , F _u , ₂ (w) , and F _u , ₃ (G)) are applied to the structural/vibro-acoustic model M.

In operational step c, the updated structural/vibro-acoustic model M is used in conjunction with the unitary force waves F _{u, I} (w) representative of the typical loads of the electric machine 20 to obtain normalized vibration/acoustic output.

The manually generated force waves represent a particular shape component that, all put together, form the total force wave. That means that the force can be reconstructed from the shapes. Each force amplitude is normalized to generate a unit energy load for every frequency w line of a broadband spec- trum. Thus, applying the unitary forces F _{u, I} (w) to the updat- ed structural/vibro-acoustic model M, the vibration/acoustic output Y _{u, I} (w) is obtained for each force shape. Therefore, according to the number of shapes corresponding vibrations or acoustic responses Y _{u, I} (w) , Y _u ,2(w) , ..., Y _{u, I} (w) are provided. Y _{u, I} (w) are the vibration or acoustic response obtained from the vibro-acoustic analysis performed using the unitary forc- es F _{u, I} (w) and the stored structural/vibro-acoustic model M.

In operational step d, by multiplying the vibration/acoustic output Y _{u, I} (w) to the operational load comprising the first operation parameter Y _tot of operational step a, the electro- magnetic forces F _tot can be estimated, respectively back- calculated. F _tot (w) as the total estimated force that applies under operating condition can be estimated by performing an inverse vibration synthesis using the following equation:

Y _tot (w) = F _K (w) x Y _u , I (w) .

F _K (w) is the estimated force component of shape K that apply under operating condition. From the equation, the vibration response can be estimated from a summation of the estimated force with a vibration or acoustic response for each force shape. The inverting matrix can be used to collect F _tot (w) = å FK(w) . To ensure invertibility of the matrix to estimate the electromagnetic forces F _ot (w) , it is required to take particular care of the number of measured output denoted by the size of Y _ot (w) ·

In summary the invention relates to method and apparatus for estimating electromagnetic forces active in an electric ma- chine. The method comprises the steps of:

Measuring (SI) at least one first operation parameter (Y _tot ) of the electric machine (20) while the electric machine (20) is operated under at least one operational condition, and Estimating (S2) electromagnetic forces (F _ot ) active in an electric machine (20) during operation of the electric ma- chine (20) by multiplying the measured at least one first op- eration parameter (Y _tot ) and a respective second operation pa- rameter provided by a stored structural/vibro-acoustic model (M) .

Due to the present invention, electromagnetic forces within an electric machine, in particular in the air gaps between the stator and rotor of the electric machine can be measured and disturbing noises and/or vibration resulting from the electromagnetic forces can be identified as well as eliminat- ed .