Online and offline partial discharge detection for electric drive systems

Abstract

Aspects of the present disclosure are directed to systems and methods for real-time detection of partial discharge (PD) associated with an electric drive system based on wide bandgap semiconductors. In one aspect, a sensing device is used to detect a signal, the signal including noise associated with a drive switch and other background noise. The signal is received by a real-time spectrum analyzer. The spectrum analyzer transforms the signal into the frequency domain and determines or registers the frequency domain distribution of the signal. The spectrum analyzer performs signal differentiation between at least one other signal included in the signal and the switching noise based on the frequency domain distribution of the at least one other signal and the switching noise. Based on the differentiation analysis, the presence of a partial discharge signal can be detected. Physics-based signal differentiation methods can also be used for signal differentiation, such as utilizing the pressure dependence of the characteristics of the PD signal.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of U.S. Provisional Patent Application Serial No. 63 / 072615, filed on August 31, 2020, which is incorporated herein by reference in its entirety. Technical Field

[0003] The present subject matter generally relates to systems and methods for detecting partial discharge associated with electric drive systems. Background Art

[0004] Electric drive systems, such as those found in aircraft electric and hybrid electric propulsion systems, may employ various power electronics to control and / or convert electric power. For example, power electronics may include wide bandgap electronics, such as wide bandgap semiconductors. Such power electronics-based electric drive systems may provide, among other benefits, dynamic speed control and greater energy efficiency. However, there are certain challenges associated with such electric drive systems.

[0005] For example, repetitive fast switching pulses or so-called pulse width modulation (PWM) waveforms contain high frequency content that imposes significant challenges on the electrical insulation systems of electric drive systems. For example, the PWM voltage spectrum overlaps with the frequency band used to detect partial discharge (PD), which often makes it difficult to detect PD consistently. For power electronics, the presence of PD can lead to severe degradation of insulation materials and other packaging materials that do not have PD resistance. As a result, PWM-fed motors can suffer from premature insulation failure due to PD and dielectric overheating. Components of electric drive systems for aircraft are particularly susceptible to PD degradation at high altitudes, such as greater than 38,000 ft.

[0006] Conventional PD monitoring equipment and analyzers cannot detect PD signals, especially sporadic PD signals, in a satisfactory manner. For example, conventional systems typically analyze signals in the time domain space and utilize high-frequency filters that are synchronized with voltage signal triggering. Such an approach is neither reliable nor generally sensitive enough for PD detection under PWM. The application of such filters may result in loss or strong attenuation of the low-frequency band of the PD signal. In addition, the PD signal may be uncorrelated with the synchronized voltage signal. At this point, time domain identification of the PD signal from noise may be difficult. Therefore, the detected PD signal sometimes does not fully represent the actual PD signal, and therefore, an incorrect PD starting voltage may be reported, especially for high altitude / low voltage applications.

[0007] Therefore, systems and methods that address one or more of the challenges described above would be useful. Summary of the invention

[0008] The present invention provides a set of technical solutions, as follows.

[0009] Technical solution 1. A system comprising:

[0010] an electric drive system having an electric motor and an electric drive electrically coupled to the electric motor;

[0011] a sensing device operable to sense a signal associated with the electric drive system, the signal including a switching noise signal generated by one or more switching components of the electric drive system;

[0012] A computing system having one or more memory devices and one or more processors, the one or more processors configured to:

[0013] receiving the signal from the sensing device;

[0014] transforming the signal into the frequency domain;

[0015] determining a frequency domain distribution for a signal using the signal transformed into the frequency domain; and

[0016] Whether a partial discharge signal exists in the signal is determined by distinguishing the at least one other signal included in the signal from the switching noise signal based at least in part on the frequency domain distribution of the switching noise signal and the frequency domain distribution of at least one other signal.

[0017] Technical Solution 2. A system as described in any of the foregoing technical solutions, wherein the one or more processors are configured to distinguish the switching noise signal from the at least one other signal based at least in part on one or more statistical features associated with the frequency domain distribution of the switching noise signal and the frequency domain distribution of the at least one other signal.

[0018] Technical Solution 3. A system as described in any of the foregoing technical solutions, wherein the one or more statistical features associated with the frequency domain distribution of the switching noise signal and the one or more statistical features associated with the frequency domain distribution of at least one other signal include at least one of a mean, a variance, a skewness, and a kurtosis.

[0019] Technical Solution 4. A system as described in any of the above technical solutions, wherein, in distinguishing the switching noise signal from the at least one other signal to determine whether the partial discharge signal exists in the signal, the one or more processors are configured to:

[0020] classifying the signal using the one or more statistical features associated with the frequency domain distribution of the switching noise signal and using the one or more statistical features associated with the frequency domain distribution of the at least one other signal, and

[0021] At least one of the signals is classified as a switching noise signal, and one of the signals is classified as a partial discharge signal.

[0022] Technical Solution 5. A system as described in any of the foregoing technical solutions, wherein the one or more processors are configured to distinguish the switching noise signal from the at least one other signal based at least in part on a signal occurrence rate associated with the switching noise signal and the frequency domain distribution of the at least one other signal.

[0023] Technical Solution 6. The system as described in any of the preceding technical solutions, wherein the one or more processors transform the signal into the frequency domain by applying a fast Fourier transform to the signal.

[0024] Technical Solution 7. The system as described in any of the above technical solutions, wherein the one or more processors are further configured to:

[0025] maintaining a maximum response of the signal in the frequency domain over a time period, and

[0026] Wherein the frequency domain distribution of the at least one other signal and the frequency domain distribution of the switching noise signal are determined based at least in part on the maximum response of the signal maintained within the time period.

[0027] Technical Solution 8. The system as described in any of the preceding technical solutions, wherein the time period is preselected so as to allow the maximum response of the signal in the frequency domain to stabilize in shape and amplitude within a predetermined margin.

[0028] Technical Solution 9. A system as described in any of the above technical solutions, wherein the sensing device samples the signal using the Nyquist sampling rule, in which the signal is sampled at a rate at least twice the highest signal frequency of interest.

[0029] Technical Solution 10. The system as described in any of the preceding technical solutions, wherein the motor is operatively coupled to a load machine, and wherein the load machine is a propeller of an aircraft.

[0030] Technical Solution 11. A system as described in any of the above technical solutions, wherein the signal is a first signal sensed when the electric drive system is within a first altitude range, and wherein the one or more processors are further configured to:

[0031] receiving a second signal from the sensing device, the second signal being sensed when the electric drive system is within a second altitude range different from the first altitude range, the second signal comprising a second switching noise signal generated by the one or more switching components;

[0032] transforming the second signal into the frequency domain;

[0033] determining a frequency domain distribution of the second signal using the second signal transformed into the frequency domain; and

[0034] Determining whether a second partial discharge signal exists within the second signal by distinguishing the second switching noise signal from the at least one other signal included within the second signal based at least in part on the frequency domain distribution of the second switching noise signal and the frequency domain distribution of at least one other signal included within the second signal.

[0035] Technical Solution 12. The system as described in any of the above technical solutions, wherein the one or more processors are further configured to:

[0036] Overlapping the frequency domain distribution of the first signal and the frequency domain distribution of the second signal in a frequency domain spectrum;

[0037] comparing the frequency domain distribution of the second signal with the frequency domain distribution of the first signal overlapped in the frequency domain spectrum; and

[0038] Based on the comparison, it is determined whether at least one of the partial discharge signal and the second partial discharge signal is an actual partial discharge signal.

[0039] Technical Solution 13. A system as described in any of the foregoing technical solutions, wherein the one or more processors are configured to determine whether at least one of the local discharge signal and the second local discharge signal is an actual local discharge signal based at least in part on one or more statistical features associated with the frequency domain distribution of the first signal and one or more statistical features associated with the frequency domain distribution of the second signal.

[0040] Technical Solution 14. A method comprising:

[0041] receiving, by one or more processors of the real-time spectrum analyzer, from a sensing device, a signal associated with an electric drive system having an electric motor and an electric drive electrically coupled thereto, the signal including a switching noise signal generated by one or more switching components of the electric drive system;

[0042] transforming, by the one or more processors, the signal into a frequency domain;

[0043] determining, by the one or more processors, a frequency domain distribution for the signal using the signal transformed into the frequency domain; and

[0044] The one or more processors determine whether a partial discharge signal is included in the signal by distinguishing the switching noise signal from at least one other signal based at least in part on the frequency domain distribution of the switching noise signal and the frequency domain distribution of at least one other signal included in the signal.

[0045] Technical Solution 15. A method as described in any of the foregoing technical solutions, wherein the one or more processors distinguish the switching noise signal from the at least one other signal at least in part based on one or more statistical features associated with the frequency domain distribution of the switching noise signal and the frequency domain distribution of the at least one other signal.

[0046] Technical Solution 16. A method as described in any of the foregoing technical solutions, wherein the one or more statistical features associated with the frequency domain distribution of the switching noise signal and the one or more statistical features associated with the frequency domain distribution of at least one other signal include at least one of the mean, variance, skewness and kurtosis.

[0047] Technical solution 17. The method as described in any of the above technical solutions, wherein distinguishing the switching noise signal from the at least one other signal to determine whether the partial discharge signal is included in the signal comprises:

[0048] classifying the signal using the one or more statistical features associated with the frequency domain distribution of the switching noise signal and the one or more statistical features associated with the frequency domain distribution of the at least one other signal by the one or more processors, and

[0049] At least one of the signals is classified as the switching noise signal, and one of the signals is classified as the partial discharge signal.

[0050] Technical Solution 18. A method as described in any of the foregoing technical solutions, wherein the one or more processors distinguish the switching noise signal from the at least one other signal at least in part based on a signal occurrence rate associated with the switching noise signal and the frequency domain distribution of the at least one other signal.

[0051] Technical Solution 19. The method as described in any of the above technical solutions further comprises:

[0052] maintaining, by the one or more processors, a maximum response of the signal in the frequency domain over a time period, and

[0053] wherein the frequency domain distribution of the at least one other signal and the frequency domain distribution of the switching noise signal are determined at least in part based on the maximum response of the signal maintained during the time period, and wherein the time period is preselected so as to allow the maximum response of the signal in the frequency domain to stabilize in shape and amplitude within a predetermined margin.

[0054] Technical Solution 20. A non-transitory computer-readable medium, the non-transitory computer-readable medium comprising computer-executable instructions, which, when executed by one or more processors of a spectrum analyzer, cause the one or more processors to:

[0055] receiving a first signal detected at a first altitude, the first signal indicating the presence of a partial discharge and comprising a first switching noise signal generated by one or more switching components of an electric drive system having an electric drive electrically coupled to an electric machine;

[0056] receiving a second signal detected at a second altitude different from the first altitude, the second signal indicating the presence of partial discharge and comprising a second switching noise signal generated by the one or more switching components; and

[0057] A determination as to whether a partial discharge is present is made based at least in part on a comparison of one or more statistical features associated with the frequency domain distribution of the first signal and one or more statistical features associated with the frequency domain distribution of the second signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] A complete and enabling disclosure of the present invention, including the best mode thereof, to one of ordinary skill in the art is set forth in the specification with reference to the accompanying drawings, in which:

[0059] Figure 1 A schematic top plan view of an aircraft according to an example embodiment of the present disclosure is provided;

[0060] Figure 2A schematic block diagram of an example system including an electric drive system and a data acquisition system according to an example embodiment of the present disclosure is provided;

[0061] Figure 3 providing a graph depicting signal power of various signals transformed into the frequency domain according to example aspects of the present disclosure;

[0062] Figure 4 Provided in Figure 2 A real-time spectrum displayed on a display device of a data acquisition system;

[0063] Figure 5 providing a graph depicting signal power of various signals captured at different altitudes transformed into the frequency domain according to example aspects of the present disclosure;

[0064] Figure 6 A flow chart of a method for detecting a partial discharge signal according to an exemplary aspect of the present disclosure is provided;

[0065] Figure 7 A flow chart of a method for detecting a partial discharge signal according to another exemplary aspect of the present disclosure is provided;

[0066] Figure 8 providing a flow chart of a method for detecting a partial discharge signal according to further exemplary aspects of the present disclosure; and

[0067] Fig. 9 is a block diagram of an example computing system that can be used to implement aspects of the present disclosure. DETAILED DESCRIPTION

[0068] Reference will now be made in detail to the present embodiments of the present invention, one or more examples of which are shown in the accompanying drawings. The detailed description uses numerical and letter designations to represent features in the drawings. Similar or similar designations in the drawings and description have been used to represent similar or similar parts of the present invention. As used herein, the terms "first", "second", and "third" are used interchangeably to distinguish one component from another, and are not intended to represent the location or importance of individual components.

[0069] Aspects of the present invention are directed to systems and methods for detecting partial discharges associated with power electronics-based electric drive systems, such as power electronics-based electric drive systems for aircraft propulsion systems. In one aspect, a system includes an electric drive system having an electric motor and an electric drive electrically coupled to the motor. The electric drive controls the electric power provided to the motor. The motor can be operatively coupled to a load machine, such as a propulsion system for an aircraft. The system also includes a data acquisition system. The data acquisition system includes a sensing device operable to sense a signal associated with the electric drive system. The signal can include noise associated with a drive switch and other background noise. The signal can also include a partial discharge signal.

[0070] The signal is received by a real-time spectrum analyzer of a data acquisition system, which is configured to detect partial discharge signals, or more generally, to detect whether partial discharge is present within the system. In particular, the spectrum analyzer transforms the sensed signal from the time domain into the frequency domain and determines or registers the frequency domain profile of the signal. The maximum response of the signal may be maintained for a period of time to determine the frequency domain profile of the signal. The spectrum analyzer then performs signal differentiation based on the frequency domain profile of the signal to determine whether a partial discharge signal is present. For example, statistical features associated with the frequency domain profile may be used to differentiate or discriminate the partial discharge signal from a switching noise signal and / or other background or environmental signals.

[0071] In some instances, the spectrum analyzer may also be configured to switch the mode from the 2D frequency domain to the 3D domain so that the frequency of occurrence of the signal can be taken into account. This may allow the detection of sporadic and / or partially hidden partial discharge signals in the PWM switching noise. Such detection may be achieved at least in part by real-time spectrum analysis, for example using a seamless fast Fourier transform (FFT). In other aspects, physics-based signal differentiation methods may also be used for signal differentiation, for example using the pressure dependence of the frequency content in the partial discharge signal.

[0072] The systems and methods provided herein advantageously provide effective and reliable online partial discharge detection. Such systems and methods do not have the disadvantages of conventional detection systems, such as detection systems based on filters and signal triggers. The partial discharge signals detected using the systems and methods provided herein can be used for online health monitoring and diagnosis of insulation systems associated with electrical components of electric drive systems. The accuracy and reliability of partial discharge signal detection provided by the systems and methods provided herein can improve health estimates. Online health monitoring and diagnosis of insulation systems can provide key information for replacement and maintenance decisions of machine / drive systems, and, among other benefits, allow active action without interrupting system operation.

[0073] Figure 1 A schematic top plan view of an aircraft 100 is provided in accordance with an exemplary embodiment of the present disclosure. As shown, for this embodiment, Figure 1 The aircraft 100 is a fixed-wing aircraft having a "tube and wing" configuration. However, the inventive aspects of the present disclosure may also be applied to other aircraft configurations. In addition, the inventive aspects of the present disclosure are applicable to both manned aircraft and unmanned aircraft, or UAVs. For reference, the aircraft 100 defines a longitudinal direction L1 and a lateral direction L2 extending perpendicular to the longitudinal direction L1. The aircraft 100 also defines a lateral centerline LC. The aircraft 100 extends, for example, along the longitudinal direction L1 between a first end 110 and a second end 112. For this embodiment, the first end 110 is the front end of the aircraft 100, and the second end 112 is the rear or rear end of the aircraft 100. The lateral centerline LC divides the aircraft 100 into two halves along the lateral direction L2, or in other words, into a first side 114 and a second side 116.

[0074] The aircraft 100 includes a fuselage 118 having a starboard side 120 and a port side 122. The aircraft 100 also includes a pair of wings 130, 140, each wing extending laterally outward from the fuselage 118. In particular, the first wing 130 extends outward from the starboard side 120 of the fuselage 118, for example, along the lateral direction L2. The first wing 130 spans between a wing root 132 and a wing tip 134. The spanwise length of the first wing 130 extends between the wing root 132 and the wing tip 134. The first wing 130 also has a leading edge 136 and a trailing edge 138. The chordwise length of the first wing 130 extends between the leading edge 136 and the trailing edge 138.

[0075] The second wing 140 extends outward from the port side 122 of the fuselage 118 opposite to the first wing 130. The second wing 140 spans between the wing root 142 and the wing tip 144. The spanwise length of the second wing 140 extends between the wing root 142 and the wing tip 144. The second wing 140 also has a leading edge 146 and a trailing edge 148. The chordwise length of the second wing 140 extends between the leading edge 146 and the trailing edge 148. The aircraft 100 may also include various control surfaces for controlling the aircraft during flight. Example control surfaces include elevators, rudders, ailerons, spoilers, flaps, slats, air brakes, or trim devices, etc. Various actuators, servomotors, and other devices may be used to manipulate the various control surfaces and variable geometry components of the aircraft 100.

[0076] The aircraft 100 also includes a propulsion system 150. The propulsion system 150 is operable to generate thrust for the aircraft 100. As depicted, the propulsion system 150 includes a plurality of propellers or fans. In this example embodiment, the fans are unducted. The fans may be any suitable type of fan or propeller, including but not limited to propellers, open rotor fans, etc. In other embodiments, the fans may be ducted, fully ducted, or partially ducted. For this embodiment, the propulsion system 150 includes a first set 160 of one or more fans mounted to the first wing 130, and a second set 170 of one or more fans distributed along and mounted to the second wing 140. In particular, for this embodiment, the first set 160 includes a plurality of fans 162A, 162B, 162C distributed along the span direction of the first wing 130, and the second set 170 includes a plurality of fans 172A, 172B, 172C distributed along the span direction of the second wing 140. The fans 162A, 162B, 162C of the first set 160 each include a plurality of fan blades 164A, 164B, 164C rotatable about their respective rotational axes, and the fans 172A, 172B, 172C of the second set 170 each include a plurality of fan blades 174A, 174B, 174C rotatable about their respective rotational axes.

[0077] Although the first set 160 and the second set 170 of fans are shown as each having three unducted fans, it will be appreciated that in other embodiments, the aircraft 100 may have more or less than three unducted fans mounted to and distributed along its wings 130, 140. Furthermore, the aircraft 100 may additionally or alternatively have fans or propellers located in other suitable locations, such as to the rear end of the fuselage 118.

[0078] like Figure 1As further depicted in FIG. 1 , the propulsion system 150 includes at least one motor for driving the fan blades 164A, 164B, 164C of the first set 160 of fans 162A, 162B, 162C and the fan blades 174A, 174B, 174C of the second set 170 of fans 172A, 172B, 172C. In some embodiments, the propulsion system 150 may include a single motor for driving the fan blades 164A, 164B, 164C of the first set 160 of fans 162A, 162B, 162C and the fan blades 174A, 174B, 174C of the second set 170 of fans 172A, 172B, 172C. For example, the motor may be operable to drive the fans 162A, 162B, 162C, 172A, 172B, 172C. In some other embodiments, the propulsion system 150 may include a first motor for driving the fan blades 164A, 164B, 164C of the fans 162A, 162B, 162C of the first set 160 and a second motor for driving the fan blades 174A, 174B, 174C of the fans 172A, 172B, 172C of the second set 170. Both the first motor and the second motor may be operable to drive the fans 162A, 162B, 162C, 172A, 172B, 172C.

[0079] For this embodiment, the fans 162A, 162B, 162C of the first set 160 and the fans 172A, 172B, 172C of the second set 170 each have a dedicated motor operatively coupled thereto. In particular, the fans 162A, 162B, 162C of the first set 160 each have an associated motor 166A, 166B, 166C, and the fans 172A, 172B, 172C of the second set 170 each have an associated motor 176A, 176B, 176C. Motors 166A, 166B, 166C are operable to drive fan blades 164A, 164B, 164C of their respective fans 162A, 162B, 162C, and motors 176A, 176B, 176C are operable to drive fan blades 174A, 174B, 174C of their respective fans 172A, 172B, 172C. The motors may be, for example, electric motors, generators, or combination motor / generators. In other example embodiments, propulsion system 150 may have other suitable configurations and numbers of motors.

[0080] Furthermore, for this embodiment, the propulsion system 150 of the aircraft 100 is an electric propulsion system and includes one or more electric power sources, power electronics, transmission lines, etc., for providing electric power to or from the motors 166A, 166B, 166C, 176A, 176B, 176C. For example, Figure 1 As shown in , the aircraft 100 includes an electric power source 180. For example, the electric power source 180 can be one or more energy storage devices. The one or more energy storage devices can include one or more batteries, battery packs, super capacitors, etc. The electric power source 180 can be installed in the fuselage 118 or in another suitable position. Electric power can be transmitted from the electric power source 180 to various loads that consume electric power, such as motors 166A, 166B, 166C, 176A, 176B, 176C. The power bus 182 can transport electric power from the electric power source 180 to the loads that consume power. The power bus 182 can include one or more transmission lines or cables, switches, and other electrical components.

[0081] A plurality of power electronics may be positioned along the power bus 182. For example, for this embodiment, a first power converter 184 may be positioned along the power bus 182. The first power converter 184 may be a DC-DC power converter operable to change the voltage of the electric power provided by the electric power source 180. The electric power conditioned by the first power converter 184 may be provided as direct current (DC) along the power bus 182 to a plurality of electric drives or power conversion devices, such as inverters. For this embodiment, each motor 166A, 166B, 166C, 176A, 176B, 176C has an associated power conversion device 168A, 168B, 168C, 178A, 178B, 178C. The power conversion device 168A, 168B, 168C, 178A, 178B, 178C may convert the input DC current into alternating current (AC). Thus, the power conversion devices 168A, 168B, 168C, 178A, 178B, 178C may be DC-AC inverters. The AC current may be provided to the motors 166A, 166B, 166C, 176A, 176B, 176C. It will be appreciated that Figure 1 The configuration of the electric distributed unducted fan propulsion system provided in is provided by way of example and is not intended to be limiting. Other suitable configurations for providing electrical power to the motors 166A, 166B, 166C, 176A, 176B, 176C are possible.

[0082] In some embodiments, the propulsion system 150 of the aircraft 100 can be a hybrid electric distributed non-ducted fan propulsion system. In such an embodiment, for example, the propulsion system 150 may include a mechanical power source, such as a gas turbine engine 186. The gas turbine engine 186 can be operatively coupled to an electric motor 188. The electric motor 188 is operable to generate electrical power. In this way, the electric motor 188 is capable of outputting electrical power. The electrical power output by the electric motor 188 can be provided to the electric power source 180 and / or to the electric motors 166A, 166B, 166C, 176A, 176B, 176C, for driving their respective fans 162A, 162B, 162C, 172A, 172B, 172C.

[0083] The motors 166A, 166B, 166C, 176A, 176B, 176C, their respective fans 162A, 162B, 162C, 172A, 172B, 172C, and their respective electric drives or power conversion devices 168A, 168B, 168C, 178A, 178B, 178C may each form an individual electric drive system and may collectively form an electric drive system for the propulsion system 150. As will be provided below, the systems and methods of the present invention provided herein provide for online partial discharge detection associated with components of an electric drive system, such as a motor. Figure 1 Any electric drive system in the electric drive system of the propulsion system 150 of the aircraft. However, it will be appreciated that the inventive aspects of the present disclosure are not limited to aviation or aerospace applications. Instead, aspects of the present invention may be applied to a wide range of applications and industries, including but not limited to HVAC / HVDC power transmission and power conversion, healthcare, renewable energy, and transportation applications.

[0084] Figure 2 A schematic block diagram of an example system 200 including an electric drive system 208 according to an example embodiment of the present disclosure is provided. For example, the electric drive system 208 may be, for example Figure 1 An electric drive system in the electric drive system of the aircraft propulsion system 150. Figure 2As shown in , the electric drive system 208 includes an electric motor 210 and an electric drive 212 electrically coupled to the electric motor 210. A load machine 214 or device is operatively coupled to the electric motor 210. The electric motor 210 can be an electric motor, a generator, or a combined motor / generator operable in a drive mode and in a generator mode. In the drive mode, the electric motor 210 drives the load machine 214. In the generator mode, the electric motor 210 generates electric power based on the torque provided by the load machine 214. As an example, the load machine 214 can be a propeller or a fan, which can be operated to generate thrust for a vehicle such as an aircraft. The electric drive 212 of the power electronics of the electric drive system 208 can be operated to control and / or convert electric power, which is provided to or received from the electric motor 210. The electric power can be provided to or transported away from the electric drive 212 via a power bus 216. The electric drive 212 can include one or more switch components 218, such as wide bandgap electronic devices, such as wide bandgap semiconductors. As an example, the one or more switch components 218 may include one or more SiC MOSFETs. In this regard, the electric drive system 208 may be a wide bandgap based electric drive system.

[0085] As another example, the switch assembly 218 may include one or more diodes. For example, in some embodiments, the electric drive system 208 may include a generator (not shown), such as a permanent magnet generator (PMG). The generator may include an intermediate exciter stage having a rectifier. The rectifier may include one or more switch assemblies 218, which may be, for example, diodes. The switch assembly 218 of the rectifier may convert AC power into DC power for main rotor excitation.

[0086] The controller 228 is communicatively coupled to the electric drive 212, for example, to control the switch assembly 218. The controller 228 may have one or more non-transitory memory devices and one or more processors, such as a microcontroller. The controller 228 may receive feedback from the electric drive 212 (e.g., current and / or voltage readings), feedback from the motor 210 (e.g., temperature readings), and / or feedback from the load machine 214 (e.g., speed, position, and / or torque readings). The controller 228 may control the electric drive 212 based at least in part on the received feedback.

[0087] For this embodiment, the motor 210 is electrically coupled to the electric drive 212 via a multi-phase cable 220. The multi-phase cable 220 includes a first cable 222 associated with a first phase, a second cable 224 associated with a second phase, and a third cable 226 associated with a third phase. In this regard, the multi-phase cable 220 is a three-phase cable and the motor 210 is a three-phase motor. However, it will be appreciated that Figure 2The electric drive system 208 depicted in FIG. 2 can have any multi-phase configuration, such as a six-phase configuration. In yet other embodiments, the electric drive 212 and the motor 210 are integrated.

[0088] like Figure 2 As further shown in FIG. 2 , the system 200 includes a data acquisition system 230. The data acquisition system 230 includes a sensing device 232 that is operable to sense signals transmitted between the motor 210 and the electric drive 212 or elsewhere along the electrical system. The signals may include general background noise and noise associated with the drive switches, or pulse width modulation (PWM) switching noise generated by the switching device of the electric drive 212. In addition, the signals captured by the sensing device 232 may include partial discharge signals. In some embodiments, the sensing device 232 samples or captures the signals using the Nyquist sampling rule, in which the signals are sampled at a rate at least twice the highest signal frequency of interest.

[0089] In addition, for the present embodiment, the sensing device 232 is a radio frequency current transformer (RFCT). As shown, the sensing device 232 or RFCT is clamped to the first cable 222, the second cable 224, and the third cable 226 of the multi-phase cable. In the case of a ground lead, the sensing device 232 or RFCT may also be clamped to such ground leads of the first cable 222, the second cable 224, and the third cable 226. In other embodiments, the sensing device 232 may be a high frequency current transformer (HFCT). In an alternative embodiment, the sensing device 232 may be an ultra-high frequency (UHF) electromagnetic antenna. The UHF electromagnetic antenna may be located within 1-2 inches (2.5-5 cm) of the desired test object (e.g., the multi-phase cable 220). In yet other embodiments, the sensing device 232 may be a capacitive coupler directly connected to the live portion of the system 200. It will be appreciated that the sensing device 232 may also be positioned in other suitable locations.

[0090] In addition, if Figure 2, the data acquisition system 230 includes a spectrum analyzer 240. For example, the spectrum analyzer 240 can be a real-time spectrum analyzer. The spectrum analyzer 240 is communicatively coupled to the sensing device 232, for example, via one or more wired and / or wireless communication links. The spectrum analyzer 240 includes a display device 260, which is operable to display various information, such as signals captured in real time. The spectrum analyzer 240 also includes a computing system 250, which has one or more memory devices and one or more processors. As will be further explained herein, the computing system 250 of the spectrum analyzer 240 can receive the signals captured by the sensing device 232 and can analyze the received signals to detect partial discharge signals associated with the electric drive system 208. In particular, the computing system 250 of the spectrum analyzer 240 can distinguish between partial discharge signals and PWM switching noise generated by the electric drive 212 based at least in part on the statistical distribution of the frequency spectrum of the partial discharge signals and the PWM switching noise generated by the electric drive 212. In addition, in some embodiments, physics-based signal differentiation can also be applied, for example, to confirm the presence of partial discharge signals. As one example, as will be further explained herein, physics-based signal differentiation can be based on the pressure dependence of partial discharge signals. Spectrum analyzer 240 can be used to detect signals associated with partial discharges either in-line (e.g., on an aircraft or other structure to which components of electric drive system 208 are mounted) or off-line (e.g., in a test laboratory).

[0091] like Figure 2 As further shown in FIG. 1 , capacitive coupler 234 is electrically coupled to the live portion of system 200. For this embodiment, multiphase cable 220 and, in particular, third cable 226 are electrically coupled to capacitive coupler 234. In other embodiments, capacitive coupler 234 may be electrically coupled to the live portion of system 200 at another suitable location, particularly when multiphase cable 220 is insulated. Capacitive coupler 234 is electrically coupled to partial discharge coupler 236. Partial discharge coupler 236 may be electrically coupled to grounding system 238, such as Figure 2 As an example, grounding system 238 may be an aircraft grounding system. Capacitive coupler 234 and partial discharge coupler 236 may form a sensing circuit with sensing device 232. For example, these elements may form an RLC impedance network. The output across the RLC network may be a voltage pulse having an amplitude proportional to the PD amplitude.

[0092] Still reference Figure 2 , example manners in which spectrum analyzer 240 may detect and / or distinguish partial discharge signals from switching noise provided in a PWM modulated signal will now be provided.

[0093] In order to perform an online or offline partial discharge signal test, electric power is provided to the electric drive 212 from an electric power source (e.g., a battery or line voltage) via a power bus 216. In order to control the electric power provided to the motor 210, the electric drive 212 performs various switching operations. In particular, the switch assembly 218 of the electric drive 212 can be quickly turned on and off to control the electric power provided to the motor 210. In this way, the signal provided by the electric drive 212 can include a signal with a PWM waveform. The switching operation performed by the electric drive 212 can effectively change the torque and / or speed of the motor 210. For example, the switching operation can be performed by a wide bandgap electronic device (e.g., a SiC MOSFET). The signal is transmitted from the electric drive 212 to the motor 210 via a multi-phase cable 220.

[0094] The sensing device 232 captures a signal. As described above, the signal may include various components or component signals. For example, the signal may include general background or ambient noise, noise associated with the driver switching, or noise generated by the rapid switching of the switch assembly 218 to generate a PWM signal, and in some instances, the signal may include a partial discharge signal. For this embodiment, the signal captured by the sensing device 232 is directly input into the spectrum analyzer 240. In this regard, for this embodiment, no filter is required, and no voltage sensing or signal triggering is required. Therefore, the system of the present disclosure does not suffer from the disadvantages associated with passing the sensor signal through a filter.

[0095] When a signal is received, the computing system 250 of the spectrum analyzer 240 transforms the signal into the frequency domain. As an example, the computing system 250 can transform the signal into the frequency domain by applying FFT to the signal. In particular, the computing system 250 can perform a scan to calculate the FFT bucket of the frequency range or band of interest in real time. The FFT buckets can be seamless or overlapping so that sporadic signals are not missed. Applying FFT to the signal can transform the signal from the time domain to the frequency domain. That is, one or more processors of the computing system 250 can transform the received signal from the time domain signal to the frequency domain signal. In some embodiments, the computing system 250 of the spectrum analyzer 240 can perform multiple scans. In some instances, performing one or more scans on the received signal may not capture the narrowband signal of interest or only capture a small part of the narrowband signal of interest. The transformation of the signal into the frequency domain is performed in real time.

[0096] In the case of transforming the signal into the frequency domain, the computing system 250 of the spectrum analyzer 240 may perform a "MAX HOLD" or maximum hold function to hold the maximum response of one or more signals of the signal in the frequency domain for a period of time. The period of time may be selected so as to allow the maximum response of the signal in the frequency domain to stabilize in shape and amplitude within a predetermined margin.

[0097] The computing system 250 of the spectrum analyzer 240 is further configured to determine a frequency domain distribution for the signal using the signal transformed into the frequency domain.The frequency domain distribution of the signal is determined based at least in part on a maximum response of the signal maintained over the time period.

[0098] As an example, Figure 3 A graph is provided depicting the signal power or amplitude of a received signal transformed into the frequency domain. In particular, Figure 3 The received signals are depicted, which include transient signals 270, 272 that are not subjected to the maximum hold function, and the first signal 274 and the second signal 276 to which the maximum hold function is applied. Maintaining the maximum response of the first and second signals 274, 276 over time facilitates seamless recording of the frequency characteristics of the first and second signals 274, 276. Therefore, it can be shown as Figure 3 The frequency domain distributions of the first signal 274 and the second signal 276 are constructed as shown in FIG.

[0099] In addition, reference Figure 2 and Figure 3 , the computing system 250 of the spectrum analyzer 240 can detect a partial discharge signal included in the signal by distinguishing a switching noise signal (e.g., a first signal 274) from at least one other signal (e.g., a second signal 276) included in the signal based at least in part on the frequency domain distribution of the switching noise signal and the frequency domain distribution of at least one other signal. The at least one other signal can be a signal suspected of being a partial discharge signal, for example, based on its frequency domain distribution. In this regard, the at least one other signal can be a signal of interest. As an example, the computing system 250 can distinguish the switching noise signal from the at least one other signal based at least in part on one or more statistical features associated with the frequency domain distribution of the switching noise signal and the frequency domain distribution of the at least one other signal. The one or more statistical features can be any suitable statistical features that provide a quantitative characterization of the frequency domain distribution associated with the received signal. Example statistical features can include, but are not limited to, a mean, variance, skewness, kurtosis, high-order features, combinations thereof, and the like associated with a given distribution.

[0100] In some embodiments, in order to distinguish or discern between partial discharge signals and signals associated with PWM switching noise, computing system 250 of spectrum analyzer 240 may classify the signals using one or more determined statistical features associated with the signals. In such embodiments, for example, at least one of signals 274, 276 may be classified as a switching noise signal, and one of the signals may be classified as a partial discharge signal. Figure 3 As depicted in , based on the statistical characteristics of the signals, the signals can be classified. In particular, the first signal 274 can be classified as a switching noise signal, and the second signal 276 can be classified as a partial discharge signal. As shown, the frequency characteristics of the second signal 276 are very different from the switching noise signal or the first signal 274. The curve of the second signal 276 has distinguishing characteristics. For example, the envelope of the curve of the second signal 276 exceeds the noise maximum response distribution, which indicates that the second signal 276 may be a partial discharge signal, thereby indicating the presence of partial discharge. In other words, the curve of the second signal 276 is a frequency domain maximum hold response associated with partial discharge. For Figure 3 In the frequency domain spectrum shown in , the second signal 276 or the partial discharge signal held at maximum hold grows above the first signal 274 or the switching noise signal. However, sometimes the low frequency end response of the second signal 276 (the response associated with the partial discharge) may completely or significantly overlap with the first signal 274 (the noise response). In other examples, the partial discharge response may only differ from the noise response in a narrow band. In this way, in other frequency domains, the maximum hold frequency distribution of the partial discharge response may look similar to the distribution of the noise response signal.

[0101] As mentioned, the switching noise generated by the switching assembly 218 may overlap significantly with the detection frequency band used to detect signals associated with partial discharge. Figure 3 As depicted in , it will be appreciated that transforming the signals into the frequency domain, recording their distribution while applying a maximum hold function thereto, and considering the statistical characteristics of the frequency domain distribution of the signals can cause the partial discharge signal, or the second signal 276 in this example, to stand out relative to the noise signal associated with the PWM switching. In this regard, occasional partial discharge signals can be captured and false positive detections avoided or greatly reduced. The partial discharge approach provided herein also provides the opportunity to distinguish signals in 2D, and in some instances, as will be further explained below in 3D. The detected partial discharge signals can be used for various purposes, such as for determining the health of the insulation system of the electric drive system 208 and / or diagnosing certain areas of the insulation system that need repair or replacement.

[0102] In some examples, the real-time spectrum analyzer 240 can be used to detect sporadic signals (such as partial discharges under DC voltage) and partial discharge signals with narrow signal frequency bands and low power levels compared to the switching noise signal spectrum. In such examples, the spectrum is completely "hidden" in the presence of PWM switching noise. Therefore, in some embodiments, the spectrum analyzer 240 can implement persistence spectrum analysis, in which all signals are superimposed in a single graph, such as on a display device. For example, Figure 4 As shown in FIG. 1 , all signals are shown as superimposed on the spectrum analyzer 240 ( Figure 2 ) in a single diagram on the display device 260. Figure 4 , the color coding of the signal trace shows how often the signal occurs at a particular frequency (i.e., occurrence frequency) and power level. In this regard, the continuous spectrum helps detect "hidden" signals, such as hidden signal 278. In this example, hidden signal 278 is a partial discharge signal. As depicted, hidden signal 278, or partial discharge signals in this example, are "hidden" under the PWM signal and switching noise, but they can be detected by the continuous spectrum analysis tool of spectrum analyzer 240.

[0103] Therefore, in some embodiments, the computing system 250 of the spectrum analyzer 240 can distinguish the switching noise signal from other signals based at least in part on the signal occurrence rate associated with the switching noise signal and the frequency domain distribution of other component signals included in the signal. For example, the mode of the spectrum analyzer 240 can be switched to a 3D mode so that the occurrence frequency of the signal can be presented on the display device 260. This allows the detection of hidden and / or sporadic signals.

[0104] According to another inventive aspect of the present disclosure, a physics-based signal differentiation method can be used to detect partial discharge signals and / or confirm that the detected partial discharge signals are actually actual partial discharge signals. The physics-based signal differentiation method provided herein utilizes the pressure dependence of the partial discharge signal or the effect of altitude / pressure on the partial discharge signal to detect and / or confirm the partial discharge signal. As will be appreciated, partial discharges may have different partial discharge initiation voltages (PDIV) at different altitudes / pressures. For example, at sea level, there may be no partial discharge signal at 900V rms, but at high altitudes such as 38,000 ft, and under the same voltage excitation, there may be partial discharges. Typically, the PDIV of the insulation system of an electric drive system is reduced at low voltage / high altitude.

[0105] Reference now Figure 2, an example approach in which a physics-based approach can be used to detect and / or confirm a partial discharge signal will be provided. For this example embodiment, a spectrum analyzer 240 receives a first signal from a sensing device 232. The first signal is provided to a motor 210 of an electric drive system 208 and is captured by the sensing device 232. The first signal may include various components or component signals. For example, the first signal may include a first switching noise signal generated by an electric drive 212 of the electric drive system 208 or more particularly by a switching component 218 thereof. In particular, the first signal is sensed by the sensing device 232 when the electric drive system 208 is at a first altitude (e.g., 35,000 ft). The air pressure at the first altitude may have a first pressure value. The first signal may include other component signals, for example, in some instances, a partial discharge signal.

[0106] The partial discharge signal may be detected from the first signal, for example, in the above-described exemplary manner. In particular, the computing system 250 of the spectrum analyzer 240 may detect the first partial discharge signal included in the first signal by distinguishing the first switching noise signal from one or more of the other component signals of the first signal based at least in part on the frequency domain distribution associated with the first switching noise signal and the frequency domain distribution associated with the other component signals compared thereto. For example, the spectrum analyzer 240 may distinguish the first signal using one or more statistical features of the frequency domain distribution or spectrum corresponding to the first signal.

[0107] In addition, the spectrum analyzer 240 receives a second signal from the sensing device 232. The second signal is provided to the motor 210 of the electric drive system 208 and is sensed by the sensing device 232. The second signal includes various components or component signals. For example, the second signal may include a second switching noise signal generated by the electric driver 212 of the electric drive system 208. In particular, when the electric drive system 208 is at a second altitude of, for example, 38,000 ft, the second signal is sensed. The second altitude is different from the first altitude. The air pressure at the first altitude may have a second pressure value that is less than the first pressure value. The second signal may include other component signals, for example, in some instances, a partial discharge signal.

[0108] The partial discharge signal may be detected from the second signal, for example, in the exemplary manner described herein. In particular, the computing system 250 of the spectrum analyzer 240 may detect the second partial discharge signal included in the second signal by distinguishing the second switching noise signal from one or more other component signals based at least in part on the frequency domain distribution associated with the second switching noise signal and the frequency domain distribution associated with the other component signals included in the second signal. For example, the spectrum analyzer 240 may distinguish the second signal using one or more statistical features of the frequency domain distribution or spectrum corresponding to the second signal.

[0109] The computing system 250 of the spectrum analyzer 240 can confirm the presence of a partial discharge by comparing one or more statistical features associated with the frequency domain distribution of a first signal captured at a first altitude or a first pressure condition and one or more statistical features associated with the frequency domain distribution of a second signal captured at a second altitude or a second pressure condition. In this manner, it can be determined or confirmed that one or both of the suspected partial discharge signals are actually partial discharge signals. In this regard, the spectrum analyzer 240 can analyze the suspected partial discharge signals captured at two different altitudes / pressures to make such a determination. In some example embodiments, the spectrum analyzer 240 can determine whether at least one of the first partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on one or more statistical features associated with the frequency domain distribution of the first signal and one or more statistical features associated with the frequency domain distribution of the second signal.

[0110] For example, now refer to Figure 2 and 5 , Figure 5 A graph is provided depicting the signal power of various signals transformed into the frequency domain according to an example aspect of the present disclosure. In particular, Figure 5 , for the same voltage level, a frequency domain distribution of a first signal 280 captured at a first altitude and a frequency domain distribution of a second signal 282 captured at a second altitude (e.g., at a high altitude corresponding to a low voltage condition) are shown. In this example, the first altitude is a low altitude corresponding to a high voltage condition (e.g., sea level), and the second altitude is a high altitude corresponding to a low voltage condition (e.g., 35,000 ft above sea level). Furthermore, for this example, both the first and second signals 280, 282 are suspected of having characteristics associated with partial discharge, or in other words, both the signals 280, 282 are suspected of containing partial discharge signals. Furthermore, in Figure 5 , in order to more clearly illustrate the suspected partial discharge components of the signals 280 , 282 , the signals 280 , 282 are depicted with switching noise and other noise removed.

[0111] As shown, the frequency domain distribution of the first signal 280 is overlapped with the frequency domain distribution of the second signal 282 in the frequency domain spectrum by the computing system 250 of the spectrum analyzer 240. The computing system 250 can compare the frequency domain distribution of the second signal 282 with the overlapped frequency domain distribution of the first signal 280 in the frequency domain spectrum. Based on the comparison, the computing system 250 can determine whether there is a partial discharge in the system, or more specifically, whether at least one of the first suspected partial discharge signal included in the first signal 280 and the second suspected partial discharge signal included in the second signal 282 is an actual partial discharge signal.

[0112] In some embodiments, as described above, the computing system 250 may determine whether at least one of the signals 280, 282 is an actual partial discharge signal based at least in part on a comparison between one or more statistical features associated with the frequency domain distribution of the first signal 280 and one or more statistical features associated with the frequency domain distribution of the second signal 282. Thus, the computing system 250 may determine the statistical features of the frequency domain distributions of the first signal 280 and the second signal 282, and may compare both of them to an expected trend and also to determine differences in their frequency characteristics.

[0113] In particular, when comparing the statistical features of the first and second signals 280, 282, the computing system 250 may compare the amplitude of the signal power of the first signal 280 with the amplitude of the signal power of the second signal 282 at various frequencies (e.g., across the entire frequency spectrum or within a preselected frequency band). The computing system 250 may also compare the frequency distribution of the first signal 280 with the frequency distribution of the second signal 282 within a frequency range (e.g., across the entire frequency spectrum or within a preselected frequency band). The difference in amplitude and frequency distribution may indicate whether the first and second signals 280, 282 include a partial discharge signal or feature. In this example, the first and second signals 280, 280 have significant differences in their amplitudes at certain frequencies, and their distribution trends indicate that both the first and second signals 280, 282 have partial discharge signals. Since the frequency domain distributions or features of the first and second signals 280, 282 have significant differences and trends relative to their corresponding expected frequency domain features associated with partial discharge, there is confidence that they are actually both partial discharge signals. The confidence level can be output by the system based on the statistical closeness of these frequency domain distributions relative to expected distributions and / or relative to each other. In general, if the frequency domain distribution of the first signal 280 and the frequency domain distribution of the second signal 282 are significantly different due to pressure changes, this indicates that at least one of the signals 280, 282 contains a partial discharge signal.

[0114] Typically, partial discharge signals captured under high altitude / low pressure conditions have narrower frequency distributions, and their high frequency content is attenuated, while their low frequency content becomes stronger in signal power amplitude compared to signals under low altitude / high pressure conditions. If the amplitudes of the first and second signals 280, 282 tend to be closer to each other and / or if there is a minimal difference in their respective frequency distributions, this may indicate that a partial discharge feature is not present in one or both of the signals 280, 282.

[0115] In other cases, only one of the signals 280, 282 may indicate a partial discharge signature. For example, the second signal 282 captured at a high altitude / low pressure (e.g., 35,000 ft above sea level) may include a partial discharge signature, while the first signal 280 captured at a low altitude / high pressure (e.g., at sea level) may not include a partial discharge signature. When comparing the statistical features of the first and second signals 280, 282, the computing system 250 may determine that for the same voltage level, the second signal 282 has features indicative of a partial discharge, while the first signal 280 contains only noise, or more precisely, does not include features indicative of a partial discharge. A missing first signal 280 would indicate that the first signal 280 does not include a partial discharge signature, while the second signal 282 does include a partial discharge signature.

[0116] refer to Figure 2 , another example approach will be provided in which a physics-based approach may be used to detect and / or confirm a partial discharge signal. For this example embodiment, the computing system 250 of the spectrum analyzer 240 is configured to detect a first partial discharge signal included in a first signal sensed by the sensing device 232 at a first time and at a first altitude or within a first altitude range (e.g., within 100 ft of 40,000 ft). The first signal is provided to the motor 210 and sensed by the sensing device 232. The first signal includes a first switching noise signal generated by one or more switching components 218. The first partial discharge signal may be detected using any of the example approaches described herein.

[0117] In addition, the computing system 250 of the spectrum analyzer 240 is configured to detect a second partial discharge signal included in a second signal sensed by the sensing device 232 at a second time and at a first altitude or within a first altitude range. The second time is later in time than the first time. In this regard, the first signal and the second signal are captured at the same or substantially the same altitude but at different times. For example, the electric drive system 208 may be embodied on an aircraft, and the first time may be associated with a time when the aircraft is generally ascending (e.g., during a climb phase of flight), and the second time may be associated with a time when the aircraft is generally descending (e.g., during a descend phase of flight). The second signal is provided to the motor 210 and sensed by the sensing device 232. The second signal includes a second switching noise signal generated by one or more switch components 218.

[0118] In addition, the computing system 250 of the spectrum analyzer 240 is configured to determine whether at least one of the first partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on a comparison of one or more statistical features associated with the frequency domain distribution of the first signal and the frequency domain distribution of the second signal. For example, similar analysis described above with respect to embodiments in which the signals are sensed at different altitudes can be used to determine whether a partial discharge exists. For example, a characteristic or statistical feature of the frequency domain distribution associated with the first signal can be compared with a characteristic or statistical feature of the frequency domain distribution associated with the second signal. As described above, the confirmation of the presence of a partial discharge can be used for various purposes, such as for determining the health of an insulation system of an electric drive system and / or diagnosing certain areas of the insulation system that need repair or replacement.

[0119] Figure 6 A flow chart of a method (300) of detecting partial discharge signals associated with an electric drive system according to exemplary aspects of the present disclosure is provided. In certain exemplary aspects, the method (300) may utilize the data acquisition system 230 described above to implement certain aspects thereof.

[0120] At (302), the method (300) includes receiving, by one or more processors of a spectrum analyzer, a signal provided to a motor of an electric drive system from a sensing device, the signal including a switching noise signal generated by one or more switching components. For example, the spectrum analyzer may be Figure 2 The spectrum analyzer 240 depicted in FIG. The motor and electric drive may be Figure 2 The motor 210 and the electric drive 212 of the electric drive system 208 of FIG. 20 are shown in FIG. 21. The switching noise signal may be generated by one or more switch components 218 of the electric drive 212. In some implementations, the electric drive has one or more wide bandgap switch components. As will be appreciated, the signal may include other component signals, such as other noise and environmental signals, and in some instances, include partial discharge signals.

[0121] In some implementations, the electric drive is electrically coupled to the motor via a multiphase cable. The multiphase cable may include a first cable associated with a first phase, a second cable associated with a second phase, and a third cable associated with a third phase. Therefore, the multiphase cable may be a three-phase cable. In other implementations, the multiphase may have any suitable number of phases. The sensing device may be an RFCT. The RFCT may be clamped to the first cable, the second cable, and / or the third cable of the multiphase cable. In other implementations, the RFCT may be clamped to the ground lead of the first cable, the second cable, and / or the third cable. In some further embodiments, the RFCT may be clamped to all phases of the system 200. In this way, the fundamental frequency current may be eliminated to avoid saturation of the current transformer. In addition, in some further implementations, the sensing device is a UHF electromagnetic antenna. The sensing device may sample or sense a signal using the Nyquist sampling rule, wherein the signal is sampled at a rate at least twice the highest signal frequency of interest. In some implementations, the electric machine is operatively coupled to a load machine, such as a propeller of an aircraft.

[0122] At (304), the method (300) includes transforming, by one or more processors, the signal into the frequency domain. In some implementations, transforming, by the one or more processors, the signal into the frequency domain includes applying an FFT to the signal. Thus, in some implementations, applying the FFT to the signal transforms the signal from the time domain into the frequency domain.

[0123] At (306), the method (300) includes determining, by one or more processors, a frequency domain distribution of the signal using the signal transformed into the frequency domain. For example, in some implementations, the method (300) may include maintaining, by one or more processors, a maximum response of the signal in the frequency domain over a time period. In other words, the one or more processors may apply a maximum hold function to the signal over the time period. In such implementations, the switching noise signal (e.g., Figure 3 The time period may be pre-selected or selected so as to allow the maximum response of the signal in the frequency domain to stabilize in shape and amplitude within a predetermined margin.

[0124] At (308), the method (300) includes determining, by one or more processors, whether a partial discharge signal is included within the signal by distinguishing the switching noise signal from the at least one other signal based at least in part on a frequency domain distribution of the switching noise signal and a frequency domain distribution of the at least one other signal. In some implementations, for example, the one or more processors distinguish the switching noise signal from the at least one other signal based at least in part on one or more statistical features associated with the frequency domain distribution of the switching noise signal and one or more statistical features associated with the frequency domain distribution of the at least one other signal. The one or more statistical features associated with the frequency domain distribution of the switching noise signal and the one or more statistical features associated with the frequency domain distribution of the at least one other signal may include at least one of a mean, a variance, a skewness, a kurtosis, a combination thereof, and the like.

[0125] In yet some implementations, distinguishing the switching noise signal from the at least one other signal to detect the partial discharge signal at (308) includes classifying the signals by one or more processors using one or more statistical features associated with a frequency domain distribution of the switching noise signal and one or more statistical features associated with a frequency domain distribution of the at least one other signal. In such implementations, at least one of the signals is classified as a switching noise signal and one of the signals is classified as a partial discharge signal.

[0126] In some further implementations, the one or more processors distinguish the switching noise signal from the at least one other signal based at least in part on a signal occurrence rate associated with the switching noise signal and a frequency domain distribution of the at least one other signal. For example, as described above, the mode of the spectrum analyzer can be switched to a 3D mode so that the frequency of occurrence of the signal can be presented on a display device for analysis. This allows detection of hidden and / or sporadic signals that may not be detectable in the 2D frequency domain.

[0127] In some implementations, the detection of partial discharge may be confirmed using a physics-based approach. In such implementations, the signal is a first signal sensed when the electric drive system is at a first altitude or within a first altitude range. The method (300) may further include receiving a second signal from the sensing device by one or more processors. The second signal is sensed when the electric drive system is at a second altitude or within a second altitude range different from the first altitude / first altitude range. The second signal includes a second switching noise signal generated by the electric drive (or more particularly, by rapid switching of its switching components). In addition, the method includes transforming the second signal into a frequency domain by one or more processors, and determining a frequency domain distribution of the second signal using the second signal transformed into the frequency domain by one or more processors. In addition, the method includes detecting a second partial discharge signal included in the second signal by distinguishing the second switching noise signal from at least one other signal included in the second signal based at least in part on the frequency domain distribution of the second switching noise signal and the frequency domain distribution of at least one other signal included in the second signal.

[0128] In such an implementation, the method (300) may further include overlapping, by one or more processors, the frequency domain distribution of the first signal and the frequency domain distribution of the second signal in the frequency domain spectrum. The method may also include comparing, by one or more processors, the frequency domain distribution of the second signal with the overlapped frequency domain distribution of the first signal in the frequency domain spectrum. Then, based on the comparison, the method may include determining, by the one or more processors, whether a partial discharge exists. In other words, based on the comparison, the method may include determining, by the one or more processors, whether at least one of the partial discharge signal and the second partial discharge signal is an actual partial discharge signal. The one or more processors may determine whether at least one of the partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on one or more statistical features associated with the frequency domain distribution of the first signal and the frequency domain distribution of the second signal.

[0129] Figure 7 A flow chart of a method (400) of detecting partial discharge signals associated with an electric drive system according to exemplary aspects of the present disclosure is provided. In certain exemplary aspects, the method (400) may utilize the data acquisition system 230 described above to implement certain aspects thereof.

[0130] At (402), the method (400) includes receiving, by one or more processors of a spectrum analyzer, from a sensing device, a first signal provided to a motor of an electric drive system, the first signal including a first switching noise signal generated by an electric drive of the electric drive system, and sensing the first signal when the electric drive system is at a first altitude. In some implementations, the electric drive system is embedded in a propulsion system of an aircraft. For example, the electric drive system may be Figure 1For example, the spectrum analyzer may be a real-time spectrum analyzer.

[0131] At (404), the method (400) includes detecting, by one or more processors, a first partial discharge signal included in the first signal by distinguishing the first switching noise signal from at least one other signal included in the first signal based at least in part on a frequency domain distribution associated with the first switching noise signal and a frequency domain distribution associated with at least one other signal.

[0132] At (406), the method (400) includes receiving, by one or more processors, a second signal provided to the motor, the second signal including a second switching noise signal generated by the electric drive, and sensing the second signal when the electric drive system is at a second altitude different from the first altitude. In some implementations, the first altitude differs from the second altitude by at least 1,000 ft. In yet other implementations, the first altitude is less than 30,000 ft and the second altitude is greater than 30,000 ft.

[0133] At (408), the method (400) includes detecting, by one or more processors, a second partial discharge signal included in the second signal by distinguishing the second switching noise signal from at least one other signal included in the second signal based at least in part on a frequency domain distribution associated with the second switching noise signal and a frequency domain distribution associated with at least one other signal included in the second signal.

[0134] At (410), the method (400) includes determining, by one or more processors, whether a partial discharge is present. In particular, the method (400) includes determining whether at least one of the first partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on one or more statistical features associated with a frequency domain distribution of the first signal and a frequency domain distribution of the second signal. In this regard, the presence of a partial discharge may be detected without passing the first signal or the second signal through a filter.

[0135] Figure 8 A flow chart of a method (500) of detecting partial discharge signals associated with an electric drive system according to exemplary aspects of the present disclosure is provided. In certain exemplary aspects, the method (500) may utilize the data acquisition system 230 described above to implement certain aspects thereof.

[0136] At (502), method (500) includes detecting, by one or more processors of a spectrum analyzer, a first partial discharge signal included in a first signal sensed when the electric drive system is at a first altitude, the first signal being provided to a motor of the electric drive system and including a first switching noise signal generated by an electric drive of the electric drive system.

[0137] At (504), the method (500) includes detecting, by one or more processors, a second partial discharge signal included in a second signal sensed when the electric drive system is at a second altitude different from the first altitude, the second signal provided to the electric machine and including a second switching noise signal generated by the electric drive device.

[0138] At (506), the method (500) includes determining, by one or more processors, whether at least one of the first partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on one or more statistical features associated with the frequency domain distribution of the first signal and one or more statistical features associated with the frequency domain distribution of the second signal. For example, the one or more statistical features associated with the frequency domain distribution of the first signal may be compared with the one or more statistical features associated with the frequency domain distribution of the second signal. Based on this comparison, it may be determined whether a partial discharge exists within the electrical system.

[0139] In some example embodiments, a Figure 2 An aircraft of the system 200. In such an example embodiment, PD online detection can be performed as follows. Prior to performing the PD detection analysis, the system is equipped with knowledge regarding the switching noise frequency distribution (or distributions) associated with the switching component. In some embodiments, the spectrum can span from very low frequencies (e.g., in the kHz range) to relatively high frequencies (e.g., up to 150 MHz for SiC switching devices). PD online detection analysis can be implemented when a "quiet zone" appears to search for PDs in a 2D domain (e.g., in 2D signal power vs. frequency domain). Given online measurements under certain environmental conditions, if the frequency 2D distribution is significantly different from the general switching noise related distribution, a PD may be present. Statistical analysis can be implemented to determine whether the measured spectrum contains switching noise characteristics or whether a PD exists.

[0140] Additionally, additional PD online measurements can be performed at other altitudes, especially if a suspected new frequency signature changes with altitude. These additional measurements or tests can confirm the presence of a PD. In the event that no new frequency signature is detected beyond the general switching frequency fingerprint or distribution, a 3D domain analysis (e.g., power and signal occurrence rate vs. frequency domain) can be performed to confirm if there is a suspected PD signal hidden beneath the switching frequency distribution, such as Figure 4 If such a hidden signal is detected, additional online measurements can be performed to check the behavior of the hidden signal at different altitudes.

[0141] In one example embodiment, an aircraft is provided. The aircraft includes an electric drive system having an electric motor, a propeller operatively coupled to the electric motor, and an electric drive electrically coupled to the electric motor. In addition, the aircraft includes a sensing device and a computing system having one or more memory devices and one or more processors. The one or more processors are configured to detect a first partial discharge signal included in a first signal sensed by the sensing device when the aircraft is at a first altitude, the first signal being provided to the electric motor and including a first switching noise signal generated by one or more switching components of the electric drive system. In addition, the one or more processors are configured to detect a second partial discharge signal included in a second signal sensed when the aircraft is at a second altitude different from the first altitude, the second signal being provided to the electric motor and including a second switching noise signal generated by one or more switching components. In addition, the one or more processors are configured to determine whether at least one of the first partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on a comparison of one or more statistical features associated with a frequency domain distribution of the first signal and one or more statistical features associated with a frequency domain distribution of the second signal.

[0142] In another exemplary embodiment, an aircraft is provided. The aircraft includes an electric drive system having a motor, a propeller operatively coupled to the motor, and an electric drive electrically coupled to the motor. In addition, the aircraft includes a sensing device and a computing system having one or more memory devices and one or more processors. The one or more processors are configured to detect a first partial discharge signal, the first partial discharge signal being included in a first signal sensed by the sensing device when the aircraft is at a first altitude or within a first altitude range, the first signal being provided to the motor and including a first switching noise signal generated by one or more switch components of the electric drive system. In addition, the one or more processors are configured to detect a second partial discharge signal included in a second signal, the second signal being sensed at a second time when the aircraft is at a first altitude or within a first altitude range, the second time being later in time than the first time, the second signal being provided to the motor and including a second switching noise signal generated by one or more switch components. In addition, the one or more processors are configured to determine whether at least one of the first partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on a comparison of one or more statistical features associated with the frequency domain distribution of the first partial discharge signal and one or more statistical features associated with the frequency domain distribution of the second partial discharge signal. In such an embodiment, the aircraft is ascending to a preselected altitude range at a first time and descending at a second time. In this regard, the partial discharge signal at a given pressure can be checked, for example, while ascending to a preselected altitude range, and then checked again at the same pressure, for example, while descending from the preselected altitude range. Statistical consistencies and / or inconsistencies in the first and second signals can be identified and used to determine the likelihood that a partial discharge has occurred.

[0143] In another example embodiment, a non-transitory computer-readable medium is provided. The non-transitory computer-readable medium includes computer-executable instructions that, when executed by one or more processors of a spectrum analyzer, cause the one or more processors to: detect a first partial discharge signal included in a first signal sensed by a sensing device at a first altitude, the first signal including a first switching noise signal generated by one or more switch components of an electric drive system, the electric drive system including an electric drive electrically coupled to a motor; detect a second partial discharge signal included in a second signal sensed by the sensing device at a second altitude different from the first altitude, the second signal including a second switching noise signal generated by the one or more switch components; and determine whether at least one of the first partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on a comparison of one or more statistical features associated with a frequency domain distribution of the first signal and one or more statistical features associated with a frequency domain distribution of the second signal.

[0144] In another example embodiment, a non-transitory computer-readable medium is provided. The non-transitory computer-readable medium includes computer-executable instructions that, when executed by one or more processors of a spectrum analyzer, cause the one or more processors to: receive a signal from a sensing device, the signal including a switching noise signal generated by one or more switching components of an electric drive system, the electric drive system including an electric drive and a motor; transform the signal into a frequency domain; use the signal transformed into the frequency domain to determine a frequency domain distribution for the signal; and detect a partial discharge signal included in the signal by distinguishing the switching noise signal from at least one other signal based at least in part on the frequency domain distribution of the switching noise signal and the frequency domain distribution of at least one other signal. The at least one other signal can be selected for a distinguishing analysis based on a preliminary analysis indicating that the at least one other signal exhibits an inconsistent frequency domain distribution relative to expected switching noise, environmental, and / or other signal responses for a given operating condition.

[0145] Fig. 9 A block diagram of a computing system 600 according to exemplary aspects of the present disclosure is provided. The computing system 600 is an example of a suitable computing system for implementing certain aspects of the present disclosure. Fig. 9 The computing system 600 is configured in a similar manner to the spectrum analyzer 240 ( Figure 2 ) of the computing system 250. In addition, the controller 228 may include Fig. 9 Some or all of the computing components of computing system 600.

[0146] like Fig. 9As shown in , the computing system 600 may include one or more processors 604 and one or more memory devices 606. The one or more processors 604 and the one or more memory devices 606 may be embodied in one or more computing devices 602. The one or more processors 604 may include any suitable processing device, such as a microprocessor, a microcontroller, an integrated circuit, a logic device, or other suitable processing device. The one or more memory devices 606 may include one or more computer-readable media, including but not limited to non-transitory computer-readable media or media, RAM, ROM, hard drives, flash drives, and other memory devices, such as one or more buffer devices.

[0147] The one or more memory devices 606 may store information accessible by the one or more processors 604, including computer readable instructions 608 executable by the one or more processors 604. The instructions 608 may be any set of instructions that, when executed by the one or more processors 604, cause the one or more processors 604 to perform operations. The instructions 608 may be software written in any suitable programming language or may be implemented in hardware. The instructions 608 may be any computer readable instructions described herein.

[0148] The memory device(s) 606 may further store data 610 accessible by the processor(s) 604. For example, the data 610 may include received sensor data, data sets of known signal distributions, etc. Furthermore, according to example embodiments of the present disclosure, the data 610 may include one or more tables, functions, algorithms, models, equations, etc.

[0149] The one or more computing devices 602 may also include a communication interface 612 for communicating with other components, systems or devices, for example. The communication interface 612 may include any suitable component for interfacing with one or more networks, including, for example, a transmitter, receiver, port, controller, antenna, or other suitable component.

[0150] The technology discussed herein relates to computer-based systems and the actions taken thereby and the information sent to and from computer-based systems. It should be appreciated that the inherent flexibility of computer-based systems allows for multiple possible configurations, combinations and divisions of tasks and functionality between and among components. For example, a single computing device or multiple computing devices working in combination can be used to implement the processes discussed herein. Databases, memories, instructions and applications can be implemented on a single system or distributed across multiple systems. Distributed components can operate sequentially or in parallel.

[0151] This written description uses examples, including the best mode, to disclose the invention, and also to enable any person skilled in the art to practice the invention, including making and using any device or system, and performing any combined method. The patentable scope of the invention is defined by the claims, and may include other examples that occur to a person skilled in the art. Such other examples are intended to be within the scope of the claims if they include structural elements that are identical to the literal language of the claims, or if they include equivalent structural elements that have insubstantial differences from the literal language of the claims.

[0152] Additional exemplary aspects are described below with respect to the following clauses:

[0153] 1. A system comprising: an electric drive system having an electric motor and an electric drive electrically coupled to the electric motor; a sensing device operable to sense a signal associated with the electric drive system, the signal comprising a switching noise signal generated by one or more switching components of the electric drive system; a computing system having one or more memory devices and one or more processors, the one or more processors being configured to: receive the signal from the sensing device; transform the signal into a frequency domain; determine a frequency domain distribution of the signal using the signal transformed into the frequency domain; and determine whether a partial discharge signal is present in the signal by distinguishing the switching noise signal from at least one other signal included in the signal based at least in part on the frequency domain distribution of the switching noise signal and the frequency domain distribution of at least one other signal.

[0154] 2. A system according to any preceding clause, wherein the one or more processors are configured to distinguish the switching noise signal from the at least one other signal based at least in part on one or more statistical features associated with the frequency domain distribution of the switching noise signal and the frequency domain distribution of the at least one other signal.

[0155] 3. A system according to any preceding clause, wherein the one or more statistical features associated with the frequency domain distribution of the switching noise signal and the one or more statistical features associated with the frequency domain distribution of the at least one other signal include at least one of a mean, a variance, a skewness, and a kurtosis.

[0156] 4. A system according to any preceding clause, wherein, in distinguishing the switching noise signal from the at least one other signal to determine whether the partial discharge signal is present within the signal, the one or more processors are configured to: classify the signal using the one or more statistical features associated with the frequency domain distribution of the switching noise signal and using the one or more statistical features associated with the frequency domain distribution of the at least one other signal, and wherein at least one of the signals is classified as the switching noise signal and one of the signals is classified as the partial discharge signal.

[0157] 5. A system according to any preceding clause, wherein the one or more processors are configured to distinguish the switching noise signal from the at least one other signal based at least in part on a signal occurrence rate associated with the switching noise signal and a frequency domain distribution of the at least one other signal.

[0158] 6. A system according to any preceding clause, wherein the one or more processors transform the signal into the frequency domain by applying a Fast Fourier Transform to the signal.

[0159] 7. A system according to any preceding clause, wherein the one or more processors are further configured to: maintain a maximum response of the signal in the frequency domain over a time period, and wherein the frequency domain distribution of at least one other signal and the frequency domain distribution of the switching noise signal are determined at least in part based on the maximum response of the signal maintained over the time period.

[0160] 8. A system according to any preceding clause, wherein the time period is preselected so as to allow the maximum response of the signal in the frequency domain to settle in shape and amplitude within a predetermined margin.

[0161] 9. A system according to any preceding clause, wherein the sensing device samples the signal using the Nyquist sampling rule in which the signal is sampled at a rate of at least twice the highest signal frequency of interest.

[0162] 10. The system of any preceding clause, wherein the electric machine is operably coupled to a load machine, and wherein the load machine is a propeller of an aircraft.

[0163] 11. A system according to any of the preceding clauses, wherein the signal is a first signal sensed when the electric drive system is within a first altitude range, and wherein the one or more processors are further configured to: receive a second signal from a sensing device, sense the second signal when the electric drive system is within a second altitude range different from the first altitude range, the second signal comprising a second switching noise signal generated by one or more switching components; transform the second signal into a frequency domain; determine a frequency domain distribution of the second signal using the second signal transformed into the frequency domain; and determine whether a second partial discharge signal exists in the second signal by distinguishing the second switching noise signal from at least one other signal included in the second signal based at least in part on the frequency domain distribution of the second switching noise signal and the frequency domain distribution of at least one other signal included in the second signal.

[0164] 12. A system as described in any preceding clause, wherein the one or more processors are further configured to: overlap the frequency domain distribution of the first signal and the frequency domain distribution of the second signal in a frequency domain spectrum; compare the frequency domain distribution of the second signal with the overlapped frequency domain distribution of the first signal in the frequency domain spectrum; and determine whether at least one of the local discharge signal and the second local discharge signal is an actual local discharge signal based on the comparison.

[0165] 13. A system according to any preceding clause, wherein the one or more processors are configured to determine whether at least one of the partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on one or more statistical features associated with the frequency domain distribution of the first signal and one or more statistical features associated with the frequency domain distribution of the second signal.

[0166] 14. A method, comprising: receiving, by one or more processors of a real-time spectrum analyzer, a signal associated with an electric drive system from a sensing device, the electric drive system having a motor and an electric drive electrically coupled thereto, the signal comprising a switching noise signal generated by one or more switching components of the electric drive system; transforming, by the one or more processors, the signal into a frequency domain; determining, by the one or more processors, a frequency domain distribution for the signal using the signal transformed into the frequency domain; and determining, by the one or more processors, whether a partial discharge signal is included in the signal by distinguishing the switching noise signal from at least one other signal included in the signal based at least in part on the frequency domain distribution of the switching noise signal and the frequency domain distribution of the at least one other signal.

[0167] 15. A method according to any preceding clause, wherein the one or more processors distinguish the switching noise signal from the at least one other signal based at least in part on one or more statistical features associated with the frequency domain distribution of the switching noise signal and the frequency domain distribution of the at least one other signal.

[0168] 16. A method according to any preceding clause, wherein the one or more statistical features associated with the frequency domain distribution of the switching noise signal and the one or more statistical features associated with the frequency domain distribution of the at least one other signal include at least one of mean, variance, skewness and kurtosis.

[0169] 17. A method according to any preceding clause, wherein distinguishing the switching noise signal from the at least one other signal to determine whether the partial discharge signal is included in the signal comprises: classifying the signal by the one or more processors using the one or more statistical features associated with the frequency domain distribution of the switching noise signal and the one or more statistical features associated with the frequency domain distribution of the at least one other signal, and wherein at least one of the signals is classified as a switching noise signal and one of the signals is classified as a partial discharge signal.

[0170] 18. A method according to any preceding clause, wherein the one or more processors distinguish the switching noise signal from the at least one other signal based at least in part on a signal occurrence rate associated with the switching noise signal and a frequency domain distribution of the at least one other signal.

[0171] 19. The method according to any of the preceding clauses further includes: maintaining, by the one or more processors, a maximum response of the signal in the frequency domain over a time period, and wherein the frequency domain distribution of at least one other signal and the frequency domain distribution of the switching noise signal are determined at least in part based on the maximum response of the signal maintained over the time period, and wherein the time period is preselected so as to allow the maximum response of the signal in the frequency domain to stabilize in shape and amplitude within a predetermined margin.

[0172] 20. A non-transitory computer-readable medium comprising computer-executable instructions which, when executed by one or more processors of a spectrum analyzer, cause the one or more processors to: receive a first signal detected at a first altitude, the first signal indicating the presence of partial discharge and comprising a first switching noise signal generated by one or more switching components of an electric drive system having an electric drive electrically coupled to a motor; receive a second signal detected at a second altitude different from the first altitude, the second signal indicating the presence of partial discharge and comprising a second switching noise signal generated by the one or more switching components; and determine whether partial discharge is present based at least in part on a comparison of one or more statistical features associated with a frequency domain distribution of the first signal and one or more statistical features associated with a frequency domain distribution of the second signal.

[0173] 21. A method, comprising: receiving, by one or more processors of a spectrum analyzer, a first signal provided to a motor of an electric drive system from a sensing device, the first signal comprising a first switching noise signal generated by one or more switch components of the electric drive system, and sensing the first signal when the electric drive system is at a first altitude; detecting, by the one or more processors, a first partial discharge signal included in the first signal by distinguishing the first switching noise signal from the first partial discharge signal based at least in part on a frequency domain distribution associated with the first switching noise signal and a frequency domain distribution associated with the first partial discharge signal; receiving, by the one or more processors, a second signal provided to the motor, the second signal comprising a first switching noise signal generated by the one or more switch components a second switching noise signal generated by a component, sensing the second signal when the electric drive system is at a second altitude different from the first altitude; detecting, by the one or more processors, a second partial discharge signal included in the second signal by distinguishing the second switching noise signal from the second partial discharge signal based at least in part on a frequency domain distribution associated with the second switching noise signal and a frequency domain distribution associated with the second partial discharge signal; and determining, by the one or more processors, whether at least one of the first partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on one or more statistical features associated with the frequency domain distribution of the first partial discharge signal and the frequency domain distribution of the second partial discharge signal.

[0174] 22. A method according to any preceding clause, wherein the first altitude differs from the second altitude by at least 1000 ft.

[0175] 23. A method according to any preceding clause, wherein the first altitude is less than 30,000 ft and the second altitude is greater than 30,000 ft.

[0176] 24. A method according to any preceding clause, wherein the electric drive system is embedded within a propulsion system of an aircraft.

[0177] 25. A method according to any preceding clause, wherein the first partial discharge signal and the second partial discharge signal are detected without passing the first signal or the second signal through a filter.

[0178] 26. A method comprising: detecting, by one or more processors of a spectrum analyzer, a first partial discharge signal, the first partial discharge signal included in a first signal sensed when the electric drive system is at a first altitude, the first signals being provided to a motor of the electric drive system and including a first switching noise signal generated by one or more switch components of the electric drive system, the electric drive system including an electric drive electrically coupled to the motor; detecting, by the one or more processors, a second partial discharge signal included in a second signal sensed when the electric drive system is at a second altitude different from the first altitude, the second signal being provided to the motor and including a second switching noise signal generated by the one or more switch components; and determining, by the one or more processors, whether at least one of the first partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on one or more statistical features associated with a frequency domain distribution of the first partial discharge signal and a frequency domain distribution of the second partial discharge signal.

[0179] 27. An aircraft, comprising: an electric drive system having an electric motor, a propeller operably coupled to the electric motor, and an electric drive electrically coupled to the electric motor; a sensing device; a computing system having one or more memory devices and one or more processors, the one or more processors configured to: detect a first partial discharge signal included in a first signal sensed by the sensing device when the aircraft is at a first altitude, the first signal being provided to the electric motor and comprising a first switching noise signal generated by one or more switching components of the electric drive system; detect a second partial discharge signal included in a second signal sensed when the aircraft is at a second altitude different from the first altitude, the second signal being provided to the electric motor and comprising a second switching noise signal generated by the one or more switching components; and determine whether at least one of the first partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on a comparison of one or more statistical features associated with a frequency domain distribution of the first partial discharge signal and a frequency domain distribution of the second partial discharge signal.

[0180] 28. A non-transitory computer-readable medium comprising computer-executable instructions that, when executed by one or more processors of a spectrum analyzer, cause the one or more processors to: detect a first partial discharge signal included in a first signal sensed by a sensing device at a first altitude, the first signal comprising a first switching noise signal generated by one or more switch components; detect a second partial discharge signal included in a second signal sensed by the sensing device at a second altitude different from the first altitude, the second signal comprising a second switching noise signal generated by the one or more switch components; and determine whether at least one of the first partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on a comparison of one or more statistical features associated with a frequency domain distribution of the first partial discharge signal and a frequency domain distribution of the second partial discharge signal.

[0181] 29. A non-transitory computer-readable medium comprising computer-executable instructions, which, when executed by one or more processors of a spectrum analyzer, cause the one or more processors to: receive a signal from a sensing device, the signal comprising a switching noise signal generated by one or more switching components; transform the signal into a frequency domain; use the signal transformed into the frequency domain to determine a frequency domain distribution for the signal; and detect a partial discharge signal included in the signal by distinguishing the switching noise signal from the partial discharge signal based at least in part on the frequency domain distribution of the switching noise signal and the frequency domain distribution of the partial discharge signal.

[0182] 30. An aircraft, comprising: an electric drive system having an electric motor, a propeller operably coupled to the electric motor, and an electric drive electrically coupled to the electric motor; a sensing device; a computing system having one or more memory devices and one or more processors, the one or more processors configured to: detect a first partial discharge signal included in a first signal sensed by the sensing device when the aircraft is at a first altitude or within a first altitude range, the first signal being provided to the electric motor and comprising a first switching noise signal generated by one or more switch components; detect a second partial discharge signal included in a second signal sensed at a second time when the aircraft is at the first altitude or within the first altitude range, the second time being later in time than the first time, the second signal being provided to the electric motor and comprising a second switching noise signal generated by the one or more switch components; and determine whether at least one of the first partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on a comparison of one or more statistical features associated with a frequency domain distribution of the first partial discharge signal and a frequency domain distribution of the second partial discharge signal.

[0183] 31. An aircraft according to any preceding clause, wherein the aircraft is ascending to a pre-selected altitude range at the first time and is descending at the second time.

Claims

1. A system for partial discharge detection, comprising: an electric drive system having an electric motor and an electric drive electrically coupled to the electric motor; a sensing device operable to sense a signal associated with the electric drive system, the signal including a switching noise signal generated by one or more switching components of the electric drive system; A computing system having one or more memory devices and one or more processors, the one or more processors configured to: receiving the signal from the sensing device; transforming the signal into the frequency domain; determining a frequency domain distribution for a signal using the signal transformed into the frequency domain; as well as determining whether a partial discharge signal exists in the signal by distinguishing the at least one other signal included in the signal from the switching noise signal based at least in part on a frequency domain distribution of the switching noise signal and a frequency domain distribution of at least one other signal; wherein the signal is a first signal sensed when the electric drive system is within a first altitude range, and wherein the one or more processors are further configured to: receiving a second signal from the sensing device, the second signal being sensed when the electric drive system is within a second altitude range different from the first altitude range, the second signal comprising a second switching noise signal generated by the one or more switching components; transforming the second signal into the frequency domain; determining a frequency domain distribution of the second signal using the second signal transformed into the frequency domain; and Determining whether a second partial discharge signal exists within the second signal by distinguishing the second switching noise signal from the at least one other signal included within the second signal based at least in part on the frequency domain distribution of the second switching noise signal and the frequency domain distribution of at least one other signal included within the second signal.

2. The system of claim 1, wherein: The one or more processors are configured to distinguish the switching noise signal from the at least one other signal based at least in part on one or more statistical features associated with the frequency domain distribution of the switching noise signal and the frequency domain distribution of the at least one other signal.

3. The system of claim 2, wherein: The one or more statistical features associated with the frequency domain distribution of the switching noise signal and the one or more statistical features associated with the frequency domain distribution of the at least one other signal include at least one of mean, variance, skewness, and kurtosis.

4. The system of claim 2, wherein: In distinguishing the switching noise signal from the at least one other signal to determine whether the partial discharge signal is present in the signal, the one or more processors are configured to: classify the signal using the one or more statistical features associated with the frequency domain distribution of the switching noise signal and using the one or more statistical features associated with the frequency domain distribution of the at least one other signal, and At least one of the signals is classified as a switching noise signal, and one of the signals is classified as a partial discharge signal.

5. The system of claim 1, wherein: The one or more processors are configured to distinguish the switching noise signal from the at least one other signal based at least in part on a signal occurrence rate associated with the switching noise signal and the frequency domain distribution of the at least one other signal.

6. The system of claim 1, wherein: The one or more processors transform the signal into the frequency domain by applying a Fast Fourier Transform to the signal.

7. The system of claim 1, wherein: The one or more processors are further configured to: maintaining a maximum response of the signal in the frequency domain over a time period, and Wherein the frequency domain distribution of the at least one other signal and the frequency domain distribution of the switching noise signal are determined based at least in part on the maximum response of the signal maintained within the time period.

8. The system of claim 7, wherein: The time period is preselected so as to allow the maximum response of the signal in the frequency domain to stabilize in shape and amplitude within a predetermined margin.

9. The system of claim 1, wherein: The sensing device samples the signal using the Nyquist sampling rule in which the signal is sampled at a rate at least twice the highest signal frequency of interest.

10. The system of claim 1, wherein: The electric machine is operatively coupled to a load machine, and wherein the load machine is a propeller of an aircraft.

11. The system of claim 1, wherein: The one or more processors are further configured to: Overlapping the frequency domain distribution of the first signal and the frequency domain distribution of the second signal in a frequency domain spectrum; comparing the frequency domain distribution of the second signal with the frequency domain distribution of the first signal overlapped in the frequency domain spectrum; and Based on the comparison, it is determined whether at least one of the partial discharge signal and the second partial discharge signal is an actual partial discharge signal.

12. The system of claim 1, wherein: The one or more processors are configured to determine whether at least one of the partial discharge signal and the second partial discharge signal is an actual partial discharge signal based at least in part on one or more statistical features associated with the frequency domain distribution of the first signal and one or more statistical features associated with the frequency domain distribution of the second signal.

13. A method for partial discharge detection, comprising: receiving, by one or more processors of the real-time spectrum analyzer, from a sensing device, a signal associated with an electric drive system having an electric motor and an electric drive electrically coupled thereto, the signal including a switching noise signal generated by one or more switching components of the electric drive system; transforming, by the one or more processors, the signal into a frequency domain; determining, by the one or more processors, a frequency domain distribution for the signal using the signal transformed into the frequency domain; as well as determining, by the one or more processors, whether a partial discharge signal is included in the signal by distinguishing the switching noise signal from the at least one other signal based at least in part on the frequency domain distribution of the switching noise signal and the frequency domain distribution of at least one other signal included in the signal; wherein the signal is a first signal sensed when the electric drive system is within a first altitude range, and wherein the one or more processors are further configured to: receiving a second signal from the sensing device, the second signal being sensed when the electric drive system is within a second altitude range different from the first altitude range, the second signal comprising a second switching noise signal generated by the one or more switching components; transforming the second signal into the frequency domain; determining a frequency domain distribution of the second signal using the second signal transformed into the frequency domain; and Determining whether a second partial discharge signal exists within the second signal by distinguishing the second switching noise signal from the at least one other signal included within the second signal based at least in part on the frequency domain distribution of the second switching noise signal and the frequency domain distribution of at least one other signal included within the second signal.

14. The method of claim 13, wherein: The one or more processors distinguish the switching noise signal from the at least one other signal based at least in part on one or more statistical features associated with the frequency domain distribution of the switching noise signal and the frequency domain distribution of the at least one other signal.

15. The method of claim 14, wherein: The one or more statistical features associated with the frequency domain distribution of the switching noise signal and the one or more statistical features associated with the frequency domain distribution of the at least one other signal include at least one of a mean, a variance, a skewness, and a kurtosis.

16. The method of claim 15, wherein: Distinguishing the switching noise signal from the at least one other signal to determine whether the partial discharge signal is included in the signal comprises: classifying, by the one or more processors, the signal using the one or more statistical features associated with the frequency domain distribution of the switching noise signal and the one or more statistical features associated with the frequency domain distribution of the at least one other signal, and At least one of the signals is classified as the switching noise signal, and one of the signals is classified as the partial discharge signal.

17. The method of claim 13, wherein: The one or more processors distinguish the switching noise signal from the at least one other signal based at least in part on a signal occurrence rate associated with the switching noise signal and the frequency domain distribution of the at least one other signal.

18. The method of claim 13, further comprising: A maximum response of the signal in the frequency domain is maintained within a time period by the one or more processors, and wherein the frequency domain distribution of the at least one other signal and the frequency domain distribution of the switching noise signal are determined at least in part based on the maximum response of the signal maintained within the time period, and wherein the time period is preselected so as to allow the maximum response of the signal in the frequency domain to stabilize in shape and amplitude within a predetermined margin.

19. A non-transitory computer readable medium comprising computer executable instructions that, when executed by one or more processors of a spectrum analyzer, cause the one or more processors to: receiving a first signal detected at a first altitude, the first signal indicating the presence of a partial discharge and comprising a first switching noise signal generated by one or more switching components of an electric drive system having an electric drive electrically coupled to an electric machine; receiving a second signal detected at a second altitude different from the first altitude, the second signal indicating the presence of partial discharge and comprising a second switching noise signal generated by the one or more switching components; and A determination as to whether a partial discharge is present is made based at least in part on a comparison of one or more statistical features associated with the frequency domain distribution of the first signal and one or more statistical features associated with the frequency domain distribution of the second signal.

Images