Cloud-based acoustic monitoring, analysis, and diagnostics for power generation systems.

Abstract

To closely monitor noises generated from a turbine generator system.SOLUTION: A system includes an acoustic monitoring, analysis, and diagnostic system having a processor. The processor is configured to receive NF noise signals from a near field (NF) microphone array, where the NF microphone array measures noises from a power generation system in an NF. The processor is configured to receive FF noise signals from a far field (FF) microphone array, where the FF microphone array measures noises from a power generation system in an FF. The processor is configured to derive NF and FF noise measurements based on the signals and to synchronize the NF and FF noise measurements to create synchronized NF and FF noise data. The processor is configured to analyze the synchronized NF and FF noise data to create NF and FF noise signatures. The processor is configured to diagnose root causes of noises generated from the power generation system and to report the root causes.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The subject matter disclosed in this specification relates generally to a cloud-based monitoring, analysis, and diagnostic system for power generation systems using acoustic sensing technology.

Background Art

[0002] A power generation system is a system that converts a primary energy source into a secondary energy source, i.e., electricity. Examples of primary energy sources include fossil fuels (such as coal, crude oil, natural gas), hydropower (such as flowing water from a dam), nuclear reactions, wind power, solar power, and geothermal energy. In many regions, most electricity is generated from power plants that use turbines or similar machinery to drive generators. A turbine generator system uses a moving fluid (such as water, steam, combustion gas, or air) to push a series of blades attached to a shaft, thereby rotating the shaft connected to the generator. Then, the generator converts kinetic energy into electrical energy based on the relationship between magnetism and electricity. Various types of turbines include steam turbines, combustion (gas) turbines, water (hydro) turbines, and wind turbines.

[0003] During the operation of a turbine generator system, excessive noise may be generated from various moving objects (such as mechanical parts and fluids) or other physical events (such as resonance). Some noise may cause environmental problems, such as complaints from neighbors. Some noise may indicate potential problems that could lead to system / component failures if not addressed in a timely or appropriate manner. Therefore, it is necessary to closely monitor the noise generated from the turbine generator system.

Summary of the Invention

[0004] A summary of certain embodiments disclosed herein is provided below. It should be understood that these embodiments are presented merely to provide the reader with a brief overview of these specific embodiments and are not intended to limit the scope of this disclosure. In fact, this disclosure may encompass a variety of embodiments not necessarily described below.

[0005] In a first embodiment, a system is provided. The system includes an acoustic monitoring, analysis, and diagnostic system having a processor. The processor receives NF noise signals from an NF microphone array that measures noise generated from a power generation system in the near field (NF), and further receives FF noise signals from an FF microphone array that measures noise generated from the power generation system in the far field (FF). Based on the received signals, the processor derives NF noise measurements and FF noise measurements. Furthermore, the processor synchronizes the NF noise measurements and FF noise measurements to create synchronized NF noise data and synchronized FF noise data, which are analyzed by the processor to generate NF noise signatures and FF noise signatures. Based on the NF noise signatures and FF noise signatures, the processor diagnoses one or more root causes of noise generated from the power generation system and reports one or more root causes of noise generated from the power generation system.

[0006] In a second embodiment, a method is provided. According to this method, an acoustic monitoring, analysis, and diagnostic system measures noise generated from a power generation system and traveling in the near field (NF) via an NF microphone array and receives an NF noise signal from the NF microphone array. Furthermore, the acoustic monitoring, analysis, and diagnostic system measures noise generated from the power generation system and traveling in the far field (FF) via an FF microphone array and receives an FF noise signal from the FF microphone array. Based on the NF and FF signals, the acoustic monitoring, analysis, and diagnostic system derives NF noise measurements and FF noise measurements. The acoustic monitoring, analysis, and diagnostic system synchronizes the NF noise measurements and FF noise measurements to synchronized NF noise data and synchronized FF noise data. Based on the NF noise measurements and FF noise measurements, the acoustic monitoring, analysis, and diagnostic system monitors the noise performance of the power generation system. Furthermore, the acoustic monitoring, analysis, and diagnostic system analyzes synchronous NF noise data and synchronous FF noise data to generate NF noise signatures and FF noise signatures, and diagnoses the root cause of noise originating from and measured by the power generation system based on the NF noise signatures and FF noise signatures. In addition, the acoustic monitoring, analysis, and diagnostic system controls the NF microphone array and FF microphone array to continuously measure noise and generate a continuous recording acoustic signal that leads to continuous monitoring of collected data to recognize changes for early fault detection based on analysis of historical data over the lifespan of the monitored power generation system.

[0007] In a third embodiment, a non-temporary computer-readable medium for storing instructions is provided. When executed by one or more processors, the instructions cause one or more processors to control a near-field (NF) microphone array to measure noise generated from the power generation system and to receive NF noise measurements from the NF microphone array. Furthermore, the instructions cause one or more processors to control a far-field (FF) microphone array to measure noise generated from the power generation system and to receive FF noise measurements from the FF microphone array. Furthermore, the instructions cause one or more processors to synchronize the NF noise measurements and FF noise measurements to generate synchronized NF noise data and synchronized FF noise data. Furthermore, the instructions cause one or more processors to monitor the noise performance of the power generation system based on the noise measured from the NF microphone array and FF microphone array. Furthermore, the instructions cause one or more processors to analyze the synchronized NF noise data and synchronized FF noise data to generate NF noise signatures and FF noise signatures. Furthermore, the instructions cause one or more processors to diagnose the root cause of noise originating from and measured in the power generation system based on NF noise signatures and FF noise signatures. In addition, the instructions cause one or more processors to control near-field (NF) microphone arrays and far-field (FF) microphone arrays to continuously measure noise and generate continuous recording acoustic signals that enable continuous monitoring of collected data to recognize changes for early fault detection based on analysis of historical data over the lifespan of the monitored power generation system.

[0008] These features, embodiments, and advantages of the present invention, as well as other features, embodiments, and advantages, will be better understood by examining the following detailed description with reference to the accompanying drawings. Throughout the drawings, similar reference numerals represent similar parts. [Brief explanation of the drawing]

[0009] [Figure 1]This is a block diagram showing embodiments of a control system operably coupled to a machine according to some aspects of the present disclosure. [Figure 2] This is a block diagram showing an embodiment of a system according to one embodiment, including a gas turbine engine, sensors, and the control system shown in Figure 1. [Figure 3] Figure 2 is a schematic diagram of one embodiment of a cloud-based acoustic monitoring, analysis, and diagnostic (CAMAD) system that can be used with power generation systems such as gas turbine engines. [Figure 4] This is a schematic diagram showing the arrangement of near-field and far-field microphone arrays that can be used with the CAMAD system shown in Figure 3 according to one embodiment. [Figure 5] This is a flowchart of a process that allows for the analysis of acoustic measurements using the near-field and far-field microphone arrays shown in Figure 4, according to one embodiment. [Modes for carrying out the invention]

[0010] One or more specific embodiments are described below. In an effort to provide a concise description of these embodiments, some features of the actual embodiments may not be described herein. In developing such actual embodiments, it should be understood that, as with any engineering or design project, a number of decisions specific to each embodiment must be made to achieve the developer's particular goals, including compliance with system-related and business-related constraints, which may differ from embodiment to embodiment. Furthermore, it should be understood that while such development efforts may be complex and time-consuming, they will still be considered by those skilled in the art to benefit from this disclosure as merely routine design, fabrication, and manufacturing activities.

[0011] When describing elements of the various embodiments of this disclosure, the articles “a, an,” “the,” and “said” are intended to mean that there is one or more of those elements. The terms “comprising,” “including,” and “having” are intended to be inclusive and mean that there may be further elements other than those listed. Furthermore, any numerical examples in the following description are not intended to be limiting, and further numbers, ranges, and percentages are included in the technical scope of the disclosed embodiments.

[0012] Power generation systems can convert various energy sources into electricity. Energy sources may include hydrocarbons, coal, natural gas, nuclear power, solar energy, wind energy, etc. In some power generation systems, a gas turbine system can burn hydrocarbon fuels for power generation. A gas turbine system may include a compressor, combustor, gas turbine, and generator. The gas turbine is an engine used to generate rotational motion to rotate a generator. A gas turbine can burn natural gas or other hydrocarbon fuels to generate mechanical motion, which can then be used to drive a generator and produce electrical energy. More specifically, a gas turbine burns a mixture of air and fuel in a combustor, and the combustion of this air-fuel mixture produces a high-temperature pressurized gas. This high-temperature pressurized gas causes the turbine blades to rotate a shaft that connects the gas turbine engine to one or more generators, thus driving the generators, which convert the rotational motion into electricity. Some gas turbine systems may generate undesirable noise. For example, high noise levels may indicate a problem that could lead to unwanted maintenance. However, identifying the source of the noise may not be easy.

[0013] During operation, a system for rapid response and fault prediction based on noise detection can be beneficial to the operator of the power generation system. Furthermore, it can be beneficial for automated systems and operators to monitor the performance of the power generation system locally (e.g., from a field workshop) and / or remotely (e.g., from a network or cloud) under various operating conditions. In certain operations, automated systems and / or operators can check noise performance from both near-field and far-field perspectives by using sensors deployed near the power generation system and sensors deployed far away from the power generation system. For example, far-field observation that minimizes environmental noise contamination can be beneficial for automated systems and / or operators.

[0014] The technology disclosed herein includes a cloud-based acoustic monitoring, analysis, and diagnostic (CAMAD) system, which may further include hardware and / or software used to remotely monitor the noise performance of a power generation system in both near-field and far-field conditions. The CAMAD system can provide root cause analysis using noise signatures of the system and its components in both near-field and far-field conditions. In addition to its convenient use in various locations (local or remote), the CAMAD system can be used to detect problems early to prevent costly repairs, predict component failures, and avoid service interruptions.

[0015] Referring here to the drawings, Figure 1 is a block diagram showing one embodiment of a machine that can be controlled by a control system. In the illustrated embodiment, the machine 102 is operably coupled to the control system 110 so that the machine 102 can be controlled by the control system 110 to perform a commanded operation. One or more sensors 106 can be deployed to monitor the operational performance of the machine 102, ambient conditions, associated systems, etc., to bring information to the control system 110, and the control system 110 can further analyze and process the received information and generate control signals to the machine 102 to perform a commanded operation via actuators such as valves, fuel throttles, pumps, positioners, etc.

[0016] The machine 102 may be a mechanical system that can use and / or generate power to apply force and control motion based on instructed actions. During operation, the machine 102 may produce certain physical events or changes 104 such as sound, motion (e.g., vibration or displacement), heat, moisture, pressure, electromagnetic fields, light, or chemicals. These physical events or changes 104 can be detected by sensors 106. The outputs of the sensors 106 may be one or more signals 108 that can be used by a control system 110 and / or converted to be displayed on a human-readable display. Sensors 106 may be devices, modules, or sensor systems that can detect and / or respond to various changes in the physical environment and / or in the machine 102. For example, sensors 106 may be acoustic sensors, motion sensors, thermal sensors, pressure sensors, high-frequency sensors, optical sensors, chemical sensors (e.g., ozone sensors), etc.

[0017] The control system 110 may include one or more processors 114 capable of receiving a signal 108 from the sensor 106. If the signal 108 includes an analog signal, an analog-to-digital converter can be used to convert the analog signal to a digital signal, which can then be further used by the processor 114. The processor 114 can analyze and process the received signal and output a control signal 112 based on the results from the data analysis and processing. Furthermore, the control system 110 may include a memory device 116 for storing data containing computer code or instructions that can perform various processes related to signal analysis and processing. The memory device 116 may include random access memory (RAM), read-only memory (ROM), storage devices (e.g., hard drives, USB sticks), and / or storage systems (e.g., relational databases, non-relational databases). The control system 110 may further include a monitoring and alarm / warning system, including a human-machine interface (HMI) system, display, audio system, etc., which allows the user to monitor inputs to the control system 110 and the operational performance of the machine 102. The control system 110 may be a local control system (for example, located in a control room on site), a network-based control system, a cloud-based system, or a combination of these.

[0018] Figure 2 is a block diagram showing one embodiment of the power generation system 200 (e.g., machine 102) as including the turbine system, sensor (e.g., sensor 106), and turbine control system (e.g., control system 110) described in relation to Figure 1. The turbine system 210 includes two gas turbine engines 216 and 218 that can operate the turbine system 210 using liquid or gaseous fuels such as synthesis gas rich in natural gas and / or hydrogen. For example, fuel nozzles can spray a fuel supply, mix the fuel with an oxidizer (e.g., air), and distribute the oxidizer-fuel mixture to the combustor 212. Combustion of the oxidizer-fuel mixture can generate a high-temperature pressurized gas in the combustor 212, which can be directed to the exhaust section 220 through the turbine section 214, which includes a high-pressure (HP) turbine engine 216 and a low-pressure (LP) turbine engine 218. In the illustrated embodiment, the HP turbine engine 216 may be part of the HP rotor portion of the turbine section 214, and the LP turbine engine 218 may be part of the LP rotor portion. As exhaust gas passes through the HP turbine engine 216 and the LP turbine engine 218, the gas can rotate a drive shaft 222 that extends along the axis of rotation by the turbine blades. As shown in the illustration, the drive shaft 222 is connected to various components of the turbine system 210, including an HP compressor 226 and an LP compressor 228.

[0019] The drive shaft 222 of the turbine system 210 may include, for example, one or more shafts that may be aligned concentrically. The drive shaft 222 may include a shaft that connects the HP turbine engine 216 to the high-pressure compressor 226 of the compressor section 224 of the turbine system 210 to form an HP rotor. For example, the HP compressor 226 may include compressor blades coupled to the drive shaft 222. Thus, the rotation of the turbine blades of the HP turbine engine 216 can cause the compressor blades in the HP compressor 226 to rotate via the shaft connecting the HP turbine engine 216 to the HP compressor 226, thereby compressing the air in the HP compressor 226. Similarly, the drive shaft 222 may include a shaft that connects the LP turbine engine 218 to the low-pressure compressor 228 of the compressor section 224 to form an LP rotor. Thus, in the illustrated embodiment, the drive shaft 222 may include both HP and LP rotors for driving the components of the HP compressor / turbine and the components of the LP compressor / turbine, respectively. The LP compressor 228 can contain compressor blades coupled to the drive shaft 222. Therefore, the rotation of the turbine blades of the LP turbine engine 218 causes the compressor blades in the LP compressor 228 to rotate via the shaft connecting the LP turbine 218 to the LP compressor 228.

[0020] The rotation of the compressor blades of the HP compressor 226 and LP compressor 228 can act to compress the air received through the intake section 232. As shown in Figure 2, the compressed air is supplied to the combustor 212 and mixed with fuel to enable more efficient combustion. Thus, the turbine system 210 may include a double concentric shaft configuration, in which the LP turbine engine 218 is drivably connected to the LP compressor 228 by a first shaft of the drive shaft 222, and the HP turbine engine 216 is similarly drivably connected to the HP compressor 226 by a second shaft of the drive shaft 222, which may be positioned inward in a concentric arrangement with respect to the first shaft. In the illustrated embodiment, the shaft 222 may also be connected to a load 234 which can include any suitable device that operates by the rotational output of the turbine system 210. For example, the load 234 may include a vehicle or a stationary load, such as a generator in a power plant or a propeller in an aircraft. In some embodiments, the turbine system 210 may be an aero-converted gas turbine used for ship propulsion, industrial power generation, and / or ship power generation applications. Furthermore, it should be noted that while the turbine system shown in Figure 2 is a diagram of a cold-end system (e.g., the load 234 is located upstream of the intake with respect to the direction of airflow), other embodiments may include a hot-end system (e.g., the load 234 is located downstream of the exhaust 220 with respect to the direction of airflow).

[0021] To provide turbine performance information to the turbine control system 244, the gas turbine system 210 may include a set of sensors 240 configured to monitor various turbine engine parameters related to the operation and performance of the turbine system 210. The sensors 240 may include, for example, one or more inlet and outlet sensors positioned adjacent to the inlet and outlet sections of the HP turbine engine 216, LP turbine 218, HP compressor 226, LP compressor 228, and / or combustor 212, as well as the intake section 232, exhaust section 220, and / or load 234. Furthermore, the sensors 240 may include measuring sensors and / or virtual sensors. As can be understood, measuring sensors may refer to physical sensors (e.g., hardware) configured to acquire measured values ​​of specific parameters, while virtual sensors may be used to acquire estimates of parameters of interest and can be implemented using software. In some embodiments, virtual sensors may be configured to provide estimates for parameters that are difficult to measure directly using physical sensors.

[0022] For example, these various inlet and outlet sensors 240, which may include measuring sensors and virtual sensors, can sense parameters related to environmental conditions such as ambient temperature and pressure and relative humidity, as well as various engine parameters related to the operation and performance of the turbine system 10, such as compressor speed ratio, inlet differential pressure, exhaust differential pressure, inlet guide vane position, fuel temperature, generator power factor, water injection rate, compressor extraction flow rate, exhaust gas temperature and pressure, compressor discharge temperature and pressure, generator output, rotor speed, turbine engine temperature and pressure, fuel flow rate, and core speed. Furthermore, the sensors 240 can also be configured to monitor engine parameters related to various operating stages of the turbine system 210.

[0023] The measured value 242 of the turbine system parameters obtained by the sensor 240 can be provided to a turbine control system 244 configured to perform monitoring, analysis, diagnosis, and adjustment tasks of the turbine system 210. The turbine control system 244 can use local and / or cloud-based processors and / or memories (e.g., processor 114, memory 116), as well as predetermined routines (stored in a computer-readable medium), to process and analyze the received measured value 242, execute a diagnosis, and generate a control signal 246 based on the analysis and diagnosis results. The control signal 246 is transmitted to the corresponding components and loads 234 of the turbine engine 210 to execute new tasks.

[0024] The turbine control system 244 can include a local control system (e.g., arranged inside a control room close to the turbine system 210). For example, the local control system can adjust the actuators within the turbine system 210 to adjust the function of the turbine system 210 by changing parameters such as fuel flow rate, vane angle, and nozzle area, either at the instruction of an operator or in automatic mode of operation. The actuators can include mechanical, hydraulic, pneumatic, or electromagnetic actuators that manage the operation of valves that control the flow of air and fuel within the air and fuel flow paths of the turbine system 210.

[0025] Furthermore, the turbine control system 244 can include a cloud-based (or network-based) monitoring, analysis, and diagnostic system, which is described in detail below with respect to FIG. 3. The cloud-based monitoring, analysis, and diagnostic system uses the measurements 242 obtained by the sensors 240 to remotely monitor the performance of the turbine system 210, perform data analysis and simulations, predict component failures, avoid service interruptions, and perform intelligent diagnostics for troubleshooting. The cloud-based monitoring, analysis, and diagnostic system enables convenient remote access to the turbine system 210. For example, the turbine system 210 may be a gas turbine generator system (GTG) installed on a remote offshore oil rig. By using the cloud-based monitoring, analysis, and diagnostic system, an operator can remotely manage, instruct, and adjust the operation of the GTG without being physically proximate to the GTG.

[0026] Figure 3 shows a schematic diagram of one embodiment of an acoustic monitoring, analysis, and diagnostic (CAMAD) system, which may be a cloud-based acoustic monitoring, analysis, and diagnostic (CAMAD) system 310 used by a power generation system 200, which in some embodiments may include the gas turbine system 210 of Figure 2. The power generation system 200 may include a turbine system 210 and a generator (e.g., a load 234). During operation, the power generation system 200 may generate certain noise 304 (e.g., irregular sound waves) that can be detected by microphones 306 (acoustic sensors for converting sound waves into electrical signals) deployed in place. The recorded acoustic signals 308 can then be transmitted to a turbine control system 244, which may include a CAMAD system 310 and a local control system 320. In other embodiments, the CAMAD system 310 may be separate from the control system 244 but be a system that is communicatively and / or operationally coupled to the control system 244. The recorded acoustic signals 308 can be computed, processed, and analyzed by the CAMAD system 310. The processed / analyzed data information can be used to predict and detect undesirable events that may require planning for maintenance, repair, and / or other services in order to improve operational uptime, minimize downtime, and thus improve productivity. Based on the processed / analyzed data information, the turbine control system 244 can generate control signals 246 and transmit the control signals 246 to the power generation system 200 to perform new tasks. For example, one of the control signals 246 may instruct the intake unit 232 to adjust the air supply rate, fuel, flame characteristics, etc., of the air-fuel mixture that the combustor 212 burns.

[0027] Noise 304 may originate from the turbine system 210 (e.g., fan or pump), the generator (e.g., motor), or auxiliary components / equipment (e.g., connecting pipes or turbine enclosure), or any subsystem of the power generation system 200. In some situations, the source of noise 304 can be easily identified. For example, noise 304 may be caused by instability observed when mechanical parts vibrate or shake in an undesirable manner. In some other situations, the source of noise 304 may be more complex. For example, noise 304 may be caused by structural or acoustic resonant conditions such as acoustic resonance, vibrational resonance, and / or turbulence.

[0028] As shown in the figure, noise 304 can be captured by microphone 306. The microphone can be used as an acoustic sensor that detects audible frequencies by converting sound waves into electrical signals. The illustrated microphone 306 may be used to capture acoustic waves (i.e., noise 304) from components of the power generation system 200 at a specific location, and the captured acoustic waves can be converted into a recorded acoustic signal 308. The location of microphone 306 can be determined, for example, by engineers and / or operators who install, operate, and maintain the power generation system 200. Placing microphone 306 may include determining the more important and / or "noisy" components of the power generation system 200 that require monitoring over time, and determining the appropriate locations where microphone 306 should be deployed to monitor the noise performance of each component.

[0029] After detecting noise 304 and converting it into a recorded acoustic signal 308, the microphone 306 and / or associated recording equipment can transmit the recorded acoustic signal 308 to the CAMAD 310. The CAMAD 310 may include hardware and software systems used to remotely monitor the noise performance of the power generation system 200. The CAMAD 310 may include data (signal) processing-related components such as an acoustic monitoring and analysis module 312, an intelligent diagnostic module 314, and one or more processors 315. Furthermore, the CAMAD 310 may include storage-related components such as one or more databases 317 and memory 319. In addition, the CAMAD 310 may include a user interface such as a human-machine interface (HMI) 316 to facilitate operation control.

[0030] The acoustic monitoring and analysis module 312 can collect, process, and analyze recorded acoustic signals 308 via one or more processors 315. The processed and analyzed data can be used to monitor the noise performance of the power generation system 200, provide early fault detection and preparation for service or replacement, eliminate operational downtime, and increase productivity. Data processing and analysis can be performed using predetermined routines (e.g., computer programs) stored in memory 319. The collected and processed data can be classified, tagged, and stored in one or more databases 317. For example, data tags can be used by one or more processors 315 as identifiers to store the data in the appropriate location in one or more databases 317.

[0031] The data processing and analysis performed by the acoustic monitoring and analysis module 312 may include removing unwanted noise (such as background noise or ambient noise) using various filtering techniques to increase the likelihood of detecting noise of interest. Filtering can be performed in the time domain (e.g., using random noise filters, finite impulse response filters, or adaptive filters) and the frequency domain (e.g., using band-pass filters or harmonic filters). Furthermore, the data processing and analysis may include using the Fast Fourier Transform (FFT), which can transform a signal from its original domain (e.g., time) to its representation in the frequency domain and vice versa. Fourier analysis can yield noise signatures of the power generation system 200 and its components, which can be used for immediate root cause analysis. The data processing and analysis may further include other audio signal processing techniques. For example, active noise suppression can be used to reduce unwanted sounds. Two signals can be canceled out by canceling interference by generating a signal that represents the unwanted noise and, in some cases, is identical to the unwanted noise but with opposite polarity.

[0032] Based on the recorded acoustic signal 308 and processed data output from the acoustic monitoring and analysis module 312, the intelligent diagnostic module 314 can predict future system / component failures and avoid service interruptions caused by such failures. System / component failure prediction can be achieved using computer simulations that closely mimic the operation of the power generation system 200 or its components (e.g., combustor 212, HP turbine engine 216, or exhaust section 220) over time. For example, thermodynamic models, finite element analysis models (FEA), computational fluid dynamics (CFD), chemical models, combustion models, etc., can be used to model the behavior of the power generation system 200 and the resulting noise. The computer simulation can provide various noise levels and / or sound patterns by using the recorded acoustic signal 308 and / or synthesized signals under various operating conditions (e.g., turbine startup, turbine base load, turbine shutdown). For example, various noise levels can be generated in the computer simulation, such as noise levels in low-load and high-load operating modes.

[0033] To provide early fault detection and root cause analysis, the source of noise should be identified by CAMAD310. Identifying the source of noise 304 may include performing noise measurements during controlled test operation of the power generation system 200 (e.g., during initial installation testing). The noise measurement procedure may include quantifying the noise level, clarifying where and under what operating conditions the noise 304 occurs, and clarifying the characteristics of the noise (e.g., one or more dominant frequencies or signature frequencies using Fourier analysis, neural network training based on the identification of specific noise patterns, etc.).

[0034] Next, the procedure may include analyzing the noise 304 based on noise characteristics such as signature frequencies. Acoustic signature-based analysis can use frequency spectra (e.g., within the range of 50 Hz to 12 kHz generated by FFT) to distinguish different noise patterns. Signature frequencies, or combinations of signature frequencies, of a given noise (e.g., noise from computer simulations and / or noise from controlled test runs of the power generation system 200) can be used to analyze the noise 304.

[0035] For example, analysis based on acoustic signatures may reveal that noise 304 has a low-frequency tone (below approximately 60 Hz) that foreshadows an instability problem in the combustor 212. The direction of propagation of noise 304 can be determined using multi-channel measurements (e.g., inflow acoustic measurements using microphones 306 which may be configured to detect acoustic waves coming from different directions and divided into different groups). The direction of propagation of noise 304 may be the same as the direction of exhaust flow, the opposite direction, or it may be stationary (standing wave). If the measurements indicate that the direction of propagation of noise 304 is the same as the direction of exhaust flow, the combustor 212 may be the source. The other two possibilities (opposite direction and stationary) may indicate that the exhaust section 220 is causing the problem.

[0036] In the second example, the noise spectrum (e.g., frequency content) may indicate that the noise 304 is an intermediate or high-frequency tone (above 60 Hz) associated with aerodynamic phenomena in the exhaust section 220, such as vortex detachment or turbulent buffeting resulting from acoustic or structural vibration resonance. Identifying acoustic resonance may require inflow acoustic measurements using microphone 306 at strategic locations within the exhaust section 220. Identifying possible vibration resonances may include the use of impulse response tests, calculations, and / or computer simulations, which may help identify structural elements involved in vibration resonance problems.

[0037] In the third example, the noise spectrum of noise 304 may include broadband noise profiles caused by turbulence that may be caused by the turbine (such as the HP turbine engine 216 and LP turbine engine 218), the exhaust section 220, or a combination thereof. Model-based simulations such as thermodynamic models, finite element analysis models (FEA), and computational fluid dynamics (CFD) simulations may help identify the root cause of the turbulence (e.g., increased airflow velocity).

[0038] In addition to responding to the recording and analysis of real-time data, the CAMAD310 can also continuously record real-time data in one or more databases 317. Using historical data from continuous recording, the CAMAD310 can detect and recognize deviations from a typical noise profile and grow or otherwise increase the profile data for future reference. Continuous monitoring and / or recording of collected data can enable operators to recognize changes for early fault detection based on the analysis of historical data over the lifetime of the monitored system / component (for example, using baseline signatures recorded at the start of the system, the system / component can be tracked over time regarding deviations of the system / component's signature from an initial baseline). Maintaining continuous real-time monitoring of noise levels can help operators of the power generation system 200 comply with various safety / environmental requirements, such as environmental, health, and safety (EHS), Occupational Safety and Health Act (OSHA), European Union (e.g., German Technical Directive on Noise Reduction (TA LARM)), New Zealand, and Australian requirements.

[0039] The HMI 316 can be used to visually display data output from the acoustic monitoring and analysis module 312 and the intelligent diagnostic module 314. By reviewing the displayed data from the HMI 316, the operator can monitor the performance of the power generation system 200 and track potential problems indicated by the intelligent diagnostic module 314. Through the HMI 316, the operator can interact with the acoustic monitoring and analysis module 312 or the intelligent diagnostic module 314 to perform further monitoring (e.g., over longer periods), perform advanced data processing (e.g., special filtering), and / or run diagnostic routines based on historical events. Furthermore, through the HMI 316, the operator can send commands to the local control system 320 to cause the power generation system 200 to perform the intended operation based on the commands. In one or more embodiments, the turbine control system 244 may include a monitor to facilitate remote access to the CAMAD system 310. In addition to or instead of this, CAMAD may be monitored via a virtual machine in the cloud.

[0040] During operation, the operator of the power generation system 200 can continuously monitor noise performance from both near-field and far-field perspectives. For example, the operator may desire a lower or quieter near-field noise limit due to contaminated, high far-field noise, or the operator may desire far-field assurance regarding specific noise and / or noise levels free from environmental noise contamination. The CAMAD system 310 can provide a more effective way to enable far-field noise verification by continuously monitoring data to determine whether the noise is actually related to the power generation system 200 or is influenced by other noise from the environment.

[0041] Figure 4 schematically shows the arrangement of near-field and far-field microphone arrays that may be used by the CAMAD system 310. In this illustrated embodiment, the near-field (NF) microphone array 402 and the far-field (FF) microphone array 412 can be used to detect noise 304 generated from the power generation system 200 in the near-field and far-field, respectively. The near-field may be limited to a certain distance from the sound source, for example, within a range of 0 to 5 meters from the sound source (such as the power generation system 200). The far-field may begin where the near-field ends and extend (theoretically) to infinity, for example, 5 to 20 meters, 5 to 100 meters, 5 to 1000 meters, 5 to 2000 meters, or beyond.

[0042] As previously mentioned, the microphone positions of the NF microphone array 402 and FF microphone array 412 may be determined by the operator of the power generation system 200 based on operational, safety, or environmental requirements. For example, to obtain more reliable noise measurements from specific components of the power generation system 200, at least a portion of the NF microphone array 402 may be placed inside the enclosure (e.g., building) of the turbine system 210. As another example, a living area in the far field may be selected as one of the locations for the FF microphone array 412, for example, when environmental noise levels may be a concern.

[0043] Different and / or similar types of microphones can be used for the NF microphone array 402 and the FF microphone array 412. The NF microphone array 402 can be communicatively connected to the NF multichannel data acquisition module 404, which can control data acquisition for each microphone in the NF microphone array 402, data preprocessing (e.g., analog-to-digital conversion if the signal sent from the NF microphone array 402 is not a digital signal), and data communication between the NF multichannel data acquisition module 404 and the CAMAD system 310. Similarly, the FF microphone array 412 can be communicatively connected to the FF multichannel data acquisition module 414, which can control data acquisition for each microphone in the FF microphone array 412, data preprocessing (e.g., analog-to-digital conversion if the signal sent from the FF microphone array is not a digital signal), and data communication between the FF multichannel data acquisition module 414 and the CAMAD system 310.

[0044] The power and connectivity supporting the microphones (including NF and FF microphone arrays 402 and 412) and data acquisition modules (including NF and FF multi-channel data acquisition modules 404 and 414) may vary depending on the location and / or surrounding environment. For example, the NF multi-channel data acquisition module 404 can be connected to the data acquisition module 408 via a connection cable 406 (e.g., a coaxial cable). Alternatively, the FF multi-channel data acquisition module 414 can be connected to the Power over Ethernet (PoE) module 418 first via a power supply Ethernet cable 416. The PoE module 418 can then be further connected to the data acquisition module 408. In one or more embodiments, the output signals from the multi-channel data acquisition module 414 can be transmitted wirelessly (e.g., via a wireless network) to the data acquisition module 408. In such embodiments, the data acquisition module 408 and the multi-channel data acquisition module 414 may include components related to wireless communication to support wireless data transmission.

[0045] The data acquisition module 408 collects output signals via NF and FF multichannel data acquisition modules 404 and 414, converts the collected signals into recorded acoustic signals 308, and can transmit the recorded acoustic signals 308 to the CAMAD system 310 for further processing and analysis. Based on the processed / analyzed data, the local control system 320 can generate control signals 246 and transmit the control signals 246 to the power generation system 200 to perform a specific task or control operation. After completing the intended task, the new noise performance can be evaluated by the operator of the power generation system 200 via the CAMAD system 310. The evaluation can compare the real-time noise performance with previous noise performance (i.e., before completion of the intended task), which is part of the historical data from continuous recording. As mentioned above, continuous recording can provide historical data for root cause analysis (RCA) using the noise signatures of the system and components in the near field and far field.

[0046] The CAMAD system 310 can provide a synchronization mechanism between the near-field (NF) and far-field (FF) noise. For example, the operator of the power generation system 200 may desire a lower NF noise range due to contaminated high far-field noise. For example, the FF noise level (e.g., at a distance of 1 kilometer from the power generation system 200) may exceed a predetermined limit. Analysis from the CAMAD system 310 may show that the NF noise level is still within the limit. Furthermore, the diagnosis from the CAMAD system 310 may indicate that the high FF noise level is due to background noise, and there may be no indication that the power generation system 200 caused the high FF noise level problem. Therefore, no further action may be required regarding the power generation system 200. Such observed events (e.g., high FF noise levels exceeding the limit) can be recorded in one or more databases 317 for documentation or further investigation.

[0047] In another example, the "rumble" noise detected by the FF microphone array 412 could be considered a precursor to a potential unexpected maintenance event in the power generation system 200. Analysis from the CAMAD system 310 may indicate that the FF "rumble" noise has a signature frequency of approximately 100 Hz. However, NF noise measurement records taken at the time the FF "rumble" noise was detected may indicate that a similar noise pattern was not captured by the NF microphone array 412. Further investigation may reveal that the "rumble" noise originated from a nearby vibration source in the far field.

[0048] Figure 5 shows a flowchart of a process 500 suitable for processing acoustic signals transmitted from the near-field and far-field microphone arrays 402 and 412 of Figure 4. Process 500 is executable by the processor 114 and can be implemented as computer instructions or code stored in memory 116. The NF microphone array 402 can measure or otherwise transmit signals representing noise 304 generated from the power generation system 200 (block 501). An NF multi-channel data acquisition module 404 can be used to control the NF noise measurement process and transmit the NF noise measurement values ​​502 to the data acquisition module 408. Similarly, the FF microphone array 412 can measure or otherwise transmit signals representing noise 304 generated from the power generation system 200 (block 551). An FF multi-channel data acquisition module 414 can be used to control the FF noise measurement process and transmit the FF noise measurement values ​​552 to the data acquisition module 408. The data acquisition module 408 can collect NF noise measurement values ​​502 (block 503) and FF noise measurement values ​​552 (block 553), respectively, and transmit the collected measurements to the CAMAD system 310.

[0049] The CAMAD system 310 can synchronize NF noise measurements 502 and FF noise measurements 552 before performing processing, analysis, and further diagnostics (block 521). Using data synchronization, data can be synchronized between NF noise measurements and FF noise measurements, allowing for automatic updates of changes between them, for example, to maintain data consistency within the CAMAD system 310. The synchronized NF noise data 504 and synchronized FF noise data 554 can be recorded in one or more databases 317. Synchronization can link NF noise measurements 502 to FF noise measurements 552 using data identifiers (e.g., tags), so that subsequent data processing and analysis can find the appropriate data blocks in one or more databases 317 using the data identifiers embedded in the synchronized NF noise data 504 and synchronized FF noise data 554. For example, the data identifier may include tags containing the time the NF / FF noise measurements were recorded, the location of the recording, and ambient data of the recording (e.g., pressure, temperature, humidity). The tags can be used by the CAMAD system 310 to find a corresponding NF noise measurement 502 that has a time tag matching the detected noise pattern shown in the FF noise measurement 552.

[0050] The CAMAD system 310 can perform data synchronization with the data acquisition module 408. Depending on the arrangement of the NF and FF microphone arrays 402 and 412, various synchronization options can be implemented. For example, synchronization options may include the use of data processor time, GPS time, IEEE 1588 protocol, other suitable synchronization protocols, or a combination thereof, which are incorporated into the local data processor (e.g., the data processor in the data acquisition module 408). As mentioned above, environmental parameters such as temperature, pressure, and humidity can also be recorded over time.

[0051] After data synchronization, the CAMAD system 310 can process and analyze the synchronized NF noise data 504 and synchronized FF noise data 554 (block 523) to generate NF noise signatures 505 and FF noise signatures 555. As previously mentioned, data processing may include removing unwanted noise in the time and / or frequency domain using various filtering techniques, such as the use of random noise filters, finite impulse response filters, adaptive filters, band-pass filters, harmonic filters, other appropriate signal processing techniques, or a combination thereof. Based on the processed data, noise signatures can be generated using Fourier-based analysis in the frequency domain. Noise signatures can also be generated via deep learning, for example, by training one or more neural networks on baseline data. Similarly, noise signatures can be generated via other techniques such as data mining (e.g., generating baseline noise clusters as signatures), state vector machine training, or expert systems (e.g., human experts providing rules including fuzzy rules that define the baseline).

[0052] During the operation of the power generation system 200, the NF noise signature 505 and FF noise signature 555 can be used by the CAMAD system 310 in combination with other data (such as data identifiers) to detect certain events that may signal potential problems that could lead to undesirable maintenance (Decision 525). For example, the detected event may relate to vibration and noise caused by misalignment of a particular component of the power generation system 200. The CAMAD system 310 can compare the detected noise pattern identified by the FF noise signature 555 in the data analysis of the FF noise measurement 552 with a similar noise pattern identified by the NF noise signature 505 in the data analysis of the NF noise measurement 502 that has a tag (e.g., a time tag) that matches this FF noise measurement 552 (Block 525). If the detected noise pattern identified by the FF noise signature 555 matches the noise pattern identified by the NF noise signature 505, the CAMAD310 can perform a root cause analysis (RCA) using the FF noise signature 555, the NF noise signature 505, and the noise signatures of other systems and / or components in the near and far fields (block 529). Furthermore, the CAMAD310 can generate a warning message or alarm for notification to the operator of the power generation system 200 and display the results from the RCA (if available). The warning message, alarm, and RCA results can be displayed via the HMI316 and / or other suitable devices.

[0053] If the FF noise signature 555 does not match the NF noise signature 505 for the detected noise pattern shown in the FF noise measurement 552 (decision 525), the CAMAD 310 may record the detected event (e.g., the noise pattern shown in the FF noise measurement 552) in one or more databases 317 for future reference and / or continuously measure the noise 304 and send continuous recording signals to the NF and FF microphone arrays 402 and 412 to update the database 317. In one or more embodiments of process 500, the CAMAD 310 can be used for further analysis. Further analysis may use measurements from other types of sensing devices / systems based on other physical aspects of the power generation system 200 (e.g., non-acoustic changes). In addition to microphones, various types of sensors can be used, such as pressure sensors, temperature / thermal sensors, vibration sensors, position sensors, and / or optical sensors. Additional sensing devices can be incorporated into the monitoring system (e.g., the CAMAD system 310) to provide additional information about the power generation system 200 being monitored. For example, an abnormal noise pattern may be detected by the NF microphone array 402 surrounding the power generation system 200. Furthermore, the CAMAD system 310 can check vibration sensors positioned around the power generation system 200 to identify any abnormal vibrations being detected. In addition, the CAMAD system 310 can check optical sensors to verify a specific alignment (e.g., alignment between a fixed component and a rotating component) that may indicate that a particular component of the power generation system 200 is misaligned and generating vibrations and noise detected by the vibration sensors, optical sensors, and microphones. In another example, the fuel pump used to operate the power generation system 200 may have a leak. A pressure sensor can detect a pressure drop while the pump is running and transmit the pressure drop to the CAMAD system 310.The CAMAD system 310 can check the noise signatures from NF noise signature 505 and / or FF noise signature 555 for further verification in order to minimize or eliminate false alarms caused by malfunctions of the pressure sensor.

[0054] While a power system including a turbine engine (e.g., a power generation system 200 including a gas turbine engine) is used as an exemplary embodiment of the present disclosure, it should be understood that the technology presented herein is not intended to be limited to turbine generator systems / packages. The technology of the present disclosure is applicable to other types of power generation systems or turbomachinery having noise-generating components (e.g., fans, pumps, compressors, motors, turboexpanders, etc.) that can be monitored by using acoustic sensors, other suitable sensing devices, or a combination thereof.

[0055] While various modifications and alternative forms may be possible with respect to this disclosure, specific embodiments are shown as examples in the drawings and described in detail herein. However, it should be understood that this disclosure is not intended to be limited to any specific form disclosed. Rather, this disclosure is intended to encompass all modifications, equivalents, and alternatives that fall within the spirit and scope of this disclosure as defined by the appended claims below.

[0056] The technologies presented and claimed herein relate to tangible objects and specific examples of a practical nature that clearly improve the art, and apply to such tangible objects and specific examples; they are not abstract, intangible, or merely theoretical. Furthermore, if any of the claims appended to the end of this specification contain one or more elements designated as “means for performing [a function]” or “steps for performing [a function],” such elements are intended to be interpreted under § 112(f) of the United States Patent Act. However, with respect to claims containing elements designated in any other way, such elements are not intended to be interpreted under § 112(f) of the United States Patent Act. [Explanation of symbols]

[0057] 102 Machinery 104 Changes 106 Sensors 108 Signal 110 Control System 112 Control signals 114 processors 116 memory devices 200 Power generation systems, output generation systems 210 Gas turbine systems, turbine engines 212 Combustor 214 Turbine section 216 High-pressure (HP) turbine engines, gas turbine engines 218 Low-Pressure (LP) Turbine Engine 220 Exhaust section 222 Drive shaft 224 Compressor section 226 HP compressor, high-pressure compressor 228 LP compressor, low-pressure compressor 232 Intake section 234 load 240 Exit Sensor 242 measurements 244 Turbine Control System 246 Control signals 304 Noise 306 Microphone 308 Recorded acoustic signals 310 Acoustic Monitoring, Analysis, and Diagnostic (CAMAD) System 312 Acoustic Monitoring and Analysis Module 314 Intelligent Diagnostic Module 315 Processor 316 Human-Machine Interface (HMI) 317 One or more databases 319 memory 320 Local Control System 402 NF Microphone Array, Near-Field Microphone Array 404 NF Multi-Channel Data Acquisition Module 406 Connection Cable 408 Data Collection Module 412 FF microphone array, far-field microphone array 414 FF Multi-Channel Data Acquisition Module 416 Powered Ethernet Cable 418 PoE modules 500 processes 502 NF noise measurement 504 NF noise data 505 NF Noise Signature 552 FF noise measurement 554 FF noise data 555 FF Noise Signature

Claims

1. An NF noise signal is received from an NF microphone array (402) that measures noise generated from a power generation system (200) in the near field (NF). An FF noise signal is received from an FF microphone array (412) that measures noise generated from the power generation system (200) in a far-field (FF). Based on the aforementioned signal, the NF noise measurement value (502) and the FF noise measurement value (552) are derived. The NF noise measurement value (502) and the FF noise measurement value (552) are synchronized, and synchronized NF noise data (504) and synchronized FF noise data (554) are generated. The synchronized NF noise data (504) and synchronized FF noise data (554) are analyzed to generate an NF noise signature (505) and an FF noise signature (555). Based on the NF noise signature (505) and the FF noise signature (555), one or more root causes of the noise generated from the power generation system (200) are diagnosed. The report identifies one or more root causes of the noise generated from the power generation system (200). An acoustic monitoring, analysis and diagnostic system (310) comprising a processor (315) configured as follows: A system equipped with these features.

2. The system according to claim 1, wherein the diagnosis of one or more root causes includes continuously receiving a noise signal, synchronizing the NF noise measurement value (502) and the FF noise measurement value (552), generating the synchronized NF noise data (504) and the synchronized FF noise data (554), then diagnosing the one or more root causes of the noise, and then reporting the one or more root causes.

3. The system according to claim 1, wherein the processor (315) is configured to synchronize the NF noise measurement value (502) and the FF noise measurement value (552) by synchronizing data between the NF noise measurement value (502) and the FF noise measurement value (552) and by automatically updating changes between the NF noise measurement value (502) and the FF noise measurement value (552).

4. The system according to claim 1, wherein the processor (315) is configured to generate the NF noise signature (505) and the FF noise signature (555) by using Fourier-based analysis in the frequency domain.

5. The system according to claim 1, wherein the near-field includes a region between 0 and 10 meters from the gas turbine system (210) included in the power generation system (200).

6. The system according to claim 1, wherein the long-range field includes an area between 10 and 10,000 meters from the gas turbine system (210) included in the power generation system (200).

7. The system according to claim 1, further comprising a control system (244) configured to control the power generation of the power generation system (200), wherein the acoustic monitoring, analysis and diagnostic system (310) is included in the control system (244) or is communicably coupled to the control system (244).

8. The system according to claim 7, wherein the control system (244) is configured to adjust the control of the power generation system (200) to improve noise by applying one or more of the root causes.

9. The system according to claim 1, wherein the acoustic monitoring, analysis, and diagnostic system (310) comprises a cloud-based acoustic monitoring, analysis, and diagnostic system.

10. The acoustic monitoring, analysis and diagnostic system (310) comprises an acoustic monitoring and analysis module (312) configured to collect, process and analyze the NF noise signal and the FF noise signal via the processor (315), and an intelligent diagnostic module (314) configured to predict failures of the power generation system (200) and its components and to help avoid service interruptions caused by failures of the power generation system (200) and its components, according to claim 1.

11. The steps include measuring noise generated from the power generation system (200) and moving in the near field using a near-field (NF) microphone array (402), The steps include receiving an NF noise signal from the NF microphone array (402) using an acoustic monitoring, analysis, and diagnostic system (310), The steps include measuring noise generated from the power generation system (200) and traveling in the far field using a far-field (FF) microphone array (412), The acoustic monitoring, analysis, and diagnostic system (310) receives an FF noise signal from the FF microphone array (412), The acoustic monitoring, analysis, and diagnostic system (310) derives an NF noise measurement value (502) and an FF noise measurement value (552) based on the NF signal and the FF signal, The acoustic monitoring, analysis, and diagnostic system (310) synchronizes the NF noise measurement value (502) and the FF noise measurement value (552) to synchronized NF noise data (504) and synchronized FF noise data (554), The steps include: monitoring the noise performance of the power generation system (200) based on the NF noise measurement value (502) and FF noise measurement value (552) using the acoustic monitoring, analysis and diagnostic system (310); The acoustic monitoring, analysis, and diagnostic system (310) analyzes the synchronized NF noise data (504) and the synchronized FF noise data (554) to generate an NF noise signature (505) and an FF noise signature (555), The steps include: using the acoustic monitoring, analysis, and diagnostic system (310) to diagnose the root cause of the measured noise originating from the power generation system (200) based on the NF noise signature (505) and FF noise signature (555); The acoustic monitoring, analysis, and diagnostic system (310) controls the near-field (NF) microphone array (402) and the far-field (FF) microphone array (412) to continuously measure the noise and generate a continuous recording acoustic signal that results in continuous monitoring of collected data to recognize changes for early fault detection based on analysis of historical data over the lifespan of the monitored power generation system (200). A method that includes this.

12. The method according to claim 11, wherein the step of synchronizing the NF noise measurement value (502) and the FF noise measurement value (552) is performed by synchronizing data between the NF noise measurement value (502) and the FF noise measurement value (552) and automatically updating the changes between the NF noise measurement value (502) and the FF noise measurement value (552).

13. The method according to claim 11, wherein the NF noise signature (505) and the FF noise signature (555) are generated using Fourier-based analysis in the frequency domain.

14. The method according to claim 11, comprising the step of controlling the power generation of the power generation system (200) by a control system (244), wherein the acoustic monitoring, analysis and diagnostic system (310) is included in the control system (244), is communicably coupled to the control system (244), is cloud-based, or a combination thereof.

15. The method according to claim 11, further comprising the step of adjusting the control of the power generation system (200) to improve the noise emanating from the power generation system (200) by applying the root cause using a control system (244).

Images