Fan operation control method and system based on internet of things

Abstract

The application relates to the field of fan operation control, in particular to a fan operation control method and system based on the Internet of Things. The application locks specific rotating fan surfaces with hidden danger tendencies by collecting the blade tip passing time, aerodynamic pressure data and infrared thermal image time sequence data, fusing frequency offset features and pressure waveform offset features and forming comprehensive drift features. Then, infrared thermal image verification is introduced. If the infrared verification standard is not met, regularity verification is further carried out to judge the regularity of the rotating fan surface for fault analysis. If the rotating fan surface that does not meet the infrared verification standard has regularity, the fault rotating fan surface is determined. If there is no regularity, the target fan operation parameters are adjusted, the hidden danger tendency is verified to be eliminated, and the infrared similarity is verified to determine whether a warning signal is sent. The application improves the reliability of early fault warning of the fan operation state through the above method.

Description

Technical Field

[0001] This invention relates to the field of fan operation control, and in particular to a fan operation control method and system based on the Internet of Things. Background Technology

[0002] As modern industrial equipment continues to develop towards larger scale, higher intelligence, and greater reliability, industrial fans, as key airflow and heat exchange components, are not only closely related to the energy efficiency and stability of the entire system, but also to production safety, equipment lifespan, and maintenance costs. Therefore, accurate sensing of fan operating health and early fault warning have become an indispensable part of the Industrial Internet of Things (IIoT) and predictive maintenance systems.

[0003] Chinese Patent Publication No. CN116085282A discloses a fan operation control method, a fan operation control device, a readable storage medium, and a fan. The fan operation control method includes: acquiring the target distance between the fan and the target user and the environmental parameters of the environment in which the fan is located; determining the target operating wind speed of the fan based on the target distance and the environmental parameters; and controlling the fan to operate at the target operating wind speed.

[0004] However, the following problems still exist in the existing technology.

[0005] In traditional fan control strategies, due to the lack of dynamic assessment of the overall thermal state of the system, isolated overheating events may cause frequent and unnecessary overheating protection, thereby masking the true trend of the continuous development of the fault and the early fault characteristics, and may lead to unnecessary shutdowns or speed reductions, thus reducing the stability and reliability of the system operation. Summary of the Invention

[0006] To address this, the present invention provides a fan operation control method and system based on the Internet of Things, which overcomes the problem in the prior art that, due to the lack of dynamic assessment of the overall thermal state of the system, isolated overheating events may cause frequent and unnecessary overheating protection, thereby masking the true trend of the continuous development of the fault and the early fault characteristics, and may lead to unnecessary shutdowns or speed reductions, thus reducing the stability and reliability of the system operation.

[0007] To achieve the above objectives, in one aspect, the present invention provides a fan operation control method based on the Internet of Things, comprising:

[0008] Step S1: Collect the passing time, aerodynamic pressure data, and infrared thermal image time series data of the target fan blade tip;

[0009] Step S2: Based on the passing frequency reference of each blade tip of the target fan recorded over time, extract the transient pressure pulse waveform sequence of the blade tip within a predetermined period based on the aerodynamic pressure data to analyze the comprehensive drift characteristics of each rotating fan surface of the target fan. The comprehensive drift characteristics include frequency offset characteristics and pressure waveform offset characteristics.

[0010] Step S3: Select rotating fan surfaces with potential hazards based on the comprehensive drift characteristics;

[0011] Step S4: Based on the infrared thermal image time series data, extract several frames of infrared freeze-frame images corresponding to the rotating fan surface with potential hidden dangers within the time domain segment, and compare them with standard infrared images to verify the infrared similarity.

[0012] Step S5: In response to non-compliance with infrared verification standards, perform regularity verification based on infrared freeze-frame images, and conduct fault analysis based on the regularity verification results, including...

[0013] Identify the faulty rotating sector;

[0014] Adjust the target fan's operating parameters, verify whether the potential hazard has been eliminated, and verify the infrared similarity to determine whether to issue a warning signal.

[0015] Furthermore, the process based on recording the passing frequency reference of each blade tip of the target fan over time includes,

[0016] Based on the passage time of the target fan blade tip, calculate the average rotation period of the target fan blade tip within a predetermined time.

[0017] The reciprocal of the average rotation period is determined as the blade tip passage frequency reference.

[0018] Furthermore, the process of analyzing the comprehensive drift characteristics of each rotating fan surface of the target fan includes,

[0019] Based on the frequency reference of each blade tip, the frequency offset value corresponding to the pressure pulse waveform is calculated and determined as the frequency offset feature;

[0020] Extract each transient pressure pulse waveform from each of the aforementioned transient pressure pulse waveform sequences;

[0021] Calculate the similarity between the transient pressure pulse waveform and the standard pressure pulse waveform point by point, and determine the mean similarity value;

[0022] The reciprocal of the mean similarity value is determined as the pressure waveform offset feature;

[0023] The frequency offset feature and the pressure waveform offset feature are weighted and summed to obtain the comprehensive drift feature.

[0024] Furthermore, the process of selecting rotating fan surfaces with potential hazards includes,

[0025] If the overall drift characteristic is greater than or equal to the preset overall drift threshold, then the rotating fan surface is selected as a rotating fan surface with potential hidden dangers.

[0026] Furthermore, the process of comparing the infrared images with standard infrared images to verify infrared similarity includes,

[0027] Determine the average image similarity between each infrared freeze-frame image and the standard infrared image;

[0028] If the average image similarity is less than the image similarity threshold, then the rotating fan-shaped surface with potential risks does not meet the infrared verification standard.

[0029] Furthermore, the process of verifying regularity based on infrared freeze-frame images includes,

[0030] Identify several frames of infrared freeze-frame images corresponding to the rotating fan that does not meet the infrared verification standard, and determine the inter-frame similarity.

[0031] Determine the mean value of the inter-frame similarity;

[0032] If the mean of inter-frame similarity is greater than or equal to the preset inter-frame similarity threshold, then the rotating fan that does not meet the infrared verification standard is verified to have regularity.

[0033] If the mean of inter-frame similarity is less than the preset inter-frame similarity threshold, then the rotating fan that does not meet the infrared verification standard is found to be irregular.

[0034] Furthermore, the fault analysis based on the regularity verification results includes,

[0035] If the rotating sector that does not meet the infrared verification standard shows a regularity, then the faulty rotating sector is identified.

[0036] If the rotating fan surface that does not meet the infrared verification standard is irregular, the operating parameters of the target fan are adjusted to verify whether the potential hazard has been eliminated and to verify the infrared similarity in order to determine whether to issue a warning signal.

[0037] Furthermore, the process of verifying whether the potential hazard has been eliminated includes,

[0038] If the overall drift characteristic of the rotating fan surface after adjusting the target fan's operating parameters is greater than or equal to the overall drift threshold, then the potential hazard has not been eliminated.

[0039] If the overall drift characteristic of the rotating fan surface after adjusting the target fan's operating parameters is less than the overall drift threshold, then the potential hazard is verified to be eliminated.

[0040] Furthermore, the process of determining whether to issue a warning signal includes,

[0041] The infrared similarity between the infrared freeze-frame image corresponding to the rotating fan surface after adjusting the target fan's operating parameters and the standard infrared image was determined to verify the infrared similarity.

[0042] If the potential hidden danger is not eliminated, or if the infrared similarity is less than the preset infrared similarity threshold, an early warning signal will be issued.

[0043] If the potential hazard is verified to be eliminated, or if the infrared similarity is greater than or equal to a preset infrared similarity threshold, then no warning signal will be issued.

[0044] Furthermore, on the other hand, a system for controlling fan operation based on the Internet of Things is also provided, including:

[0045] The acquisition module is used to acquire the passing time of the target fan blade tip, aerodynamic pressure data, and infrared thermal image time series data;

[0046] The feature analysis module, which is connected to the acquisition module, is used to extract the transient pressure pulse waveform sequence of the blade tips within a predetermined period based on the aerodynamic pressure data, based on the passing frequency reference of each blade tip of the target fan recorded over time, so as to analyze the comprehensive drift characteristics of each rotating blade of the target fan, including frequency offset characteristics and pressure waveform offset characteristics.

[0047] A selected module is connected to the analytical feature module to select rotating fan surfaces with potential risks based on the comprehensive drift features.

[0048] The verification module is connected to the acquisition module and the selection module respectively, and is used to extract several frames of infrared freeze-frame images corresponding to the rotating fan with potential hidden dangers within the time domain segment based on the infrared thermal image time series data, and compare them with standard infrared images to verify the infrared similarity.

[0049] The fault analysis module, connected to the acquisition module, feature analysis module, selection module, and verification module, is used to perform regularity verification based on infrared freeze-frame images in response to non-compliance with infrared verification standards, and to perform fault analysis based on the regularity verification results, including...

[0050] Used to identify faulty rotating fan surfaces;

[0051] It is used to adjust the operating parameters of the target fan, verify whether the potential hazards have been eliminated, and verify the infrared similarity in order to determine whether to issue a warning signal.

[0052] Compared with existing technologies, this invention collects fan blade tip passage time, aerodynamic pressure data, and infrared thermal image time-series data, fusing frequency offset features and pressure waveform offset features to form a comprehensive drift feature. This feature is used to pinpoint specific rotating fan surfaces with potential hazards. Subsequently, infrared thermal imaging verification is introduced. If the fan does not meet infrared verification standards, further regularity verification is performed to determine the regularity of the rotating fan surface for fault analysis. If the verification shows a regularity in the rotating fan surface that does not meet the infrared verification standards, the faulty rotating fan surface is identified. If there is no regularity, the target fan's operating parameters are adjusted to verify whether the potential hazard has been eliminated and to verify infrared similarity, in order to determine whether to issue a warning signal. Through the above method, the reliability of early fault warnings for fan operation is improved.

[0053] In particular, this invention extracts the transient pressure pulse waveform sequence of the blade tip within a predetermined period based on the aerodynamic pressure data to analyze the comprehensive drift characteristics of each rotating blade of the target fan. Since the effects of potential problems such as blade cracks, surface wear, or uneven dust accumulation are extremely weak in the initial stage, their physical characteristics are often neither simply increased vibration nor a significant overall temperature rise, but rather manifest as a subtle, synergistic change in the rotational dynamics and local aerodynamic properties of a specific blade. For example, each high-speed rotating blade of the fan periodically compresses the air in front of it at its tip, generating a weak pressure wave; each blade generates a pressure pulse as it passes by. Ideally, all blades are completely identical, so these pulses should be equally spaced, of the same shape, and of the same amplitude. When blades have cracks, wear, or dust accumulation, the morphology and periodicity of these pressure pulses change, leading to pressure waveform and frequency shifts. Therefore, integrating frequency and pressure waveform shift characteristics to determine comprehensive drift characteristics is based on profound physical mechanisms and monitoring logic: frequency shift characteristics capture, in the time domain, the periodic dynamic imbalance trend caused by minute changes in blade mass or aerodynamic drag torque, which is strictly synchronized with rotation, and is a macroscopic integrated reflection of the fault; while pressure waveform shift characteristics, in the spatial domain, capture, through the analysis of pulse waveform morphology, the characteristic aerodynamic disturbances caused by microscopic changes in blade geometry or surface condition. These two parameters comprehensively cover the dynamic and aerodynamic characteristics of the blade from both the time and spatial domains. Among them, the pressure waveform is extremely sensitive to the blade surface condition but is easily affected by instantaneous flow field disturbances; frequency shift is robust to changes in mass or torque but has poor localization. When the two are combined, they form a characteristic amplifier that is strong in responding to real faults and weak in responding to random noise. This is something that cannot be achieved by a single feature or simple parallel observation, thus providing a reliable basis for subsequent fault analysis and improving the reliability of early fault warning of fan operating status.

[0054] In particular, this invention extracts several frames of infrared still images corresponding to the rotating fan-shaped blades exhibiting potential hazards within a time domain segment based on the time-series data of the infrared thermal images. These images are then compared with standard infrared images to verify infrared similarity. In complex industrial environments, the "potential hazards" identified earlier based on dynamic and aerodynamic characteristics may originate from occasional interferences such as instantaneous airflow disturbances, slight load fluctuations, or sensor noise. While these interferences may cause signal drift, they do not necessarily correspond to actual physical damage. However, most progressive faults in rotating machinery, such as crack propagation, increased friction, and worsening imbalance, are essentially energy dissipation processes. Ultimately, these faults will manifest as abnormal thermodynamic characteristics at the fault location, namely, local temperature rise or changes in heat distribution patterns. For example, when blades have cracks or wear, the thermal conductivity of these areas changes, leading to local temperature increases or decreases. This results in thermal characteristics in the infrared image that differ from those of normal blades. Infrared thermal imaging technology can capture these thermo-radiative signals generated by the fault mechanism itself non-contactly and with high spatial resolution. Verifying infrared similarity provides a crucial physical verification step for the previously identified potential hazards. Specifically, the system identifies potentially problematic rotating fan surfaces and extracts multiple frames of infrared images over a period of time. By comparing these images with standard infrared images, the similarity of the thermal images is quantified. This verification method based on infrared thermal imaging not only provides fault characteristics independent of aerodynamic and mechanical properties but also allows for secondary confirmation of potential problems from a thermodynamic perspective, thereby further improving the reliability of early fault warnings for fan operation.

[0055] In particular, this invention verifies regularity based on infrared freeze-frame images and analyzes faults based on the verification results. By extracting and analyzing infrared image sequences of the suspected fan surface across multiple consecutive rotation cycles and calculating their inter-frame similarity, it can quantify whether the abnormal thermal features exhibit a recurring pattern synchronized with the rotational speed. If the verification shows regularity, it proves that the thermal anomaly source is solidified on the rotating fan surface itself, strongly coupled with the rotational motion, thus directly identifying the faulty rotating fan surface and achieving precise location. Conversely, if the verification shows no regularity, it indicates that the initially detected thermal anomaly is more likely due to transient interference or overall, non-local operational state deviations, such as sudden changes in system load or fluctuations in cooling conditions, rather than structural damage to a specific fan surface. In this case, blindly shutting down for repair is unnecessary; a more reasonable strategy is to adjust the overall operating parameters of the target fan and observe whether the previously determined comprehensive drift features and infrared similarity meet the threshold under this adjustment. This design upgrades fault diagnosis from a single feature alarm to an intervention-oriented and verifiable systematic approach. It not only avoids invalid warnings triggered by occasional interference, but also provides key decision-making basis for distinguishing between local fixed faults and overall transient imbalances through systematic parameter control experiments, thereby improving the reliability of early fault warnings for fan operation status. Attached Figure Description

[0056] Figure 1 This is a schematic diagram illustrating the steps of an IoT-based fan operation control method according to an embodiment of the invention.

[0057] Figure 2 A logic decision diagram for selecting a rotating sector with potential risks in an embodiment of the invention;

[0058] Figure 3 This is a logic block diagram for regularity verification based on infrared freeze-frame images, as an embodiment of the invention.

[0059] Figure 4 This is a logic block diagram illustrating how to determine whether to issue a warning signal according to an embodiment of the invention. Detailed Implementation

[0060] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0061] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0062] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the term "connection" should be interpreted broadly. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0063] Please see Figure 1 The diagram illustrates the steps of an IoT-based fan operation control method according to an embodiment of the invention. The IoT-based fan operation control method of this embodiment includes:

[0064] Step S1: Collect the passing time, aerodynamic pressure data, and infrared thermal image time series data of the target fan blade tip;

[0065] Step S2: Based on the passing frequency reference of each blade tip of the target fan recorded over time, extract the transient pressure pulse waveform sequence of the blade tip within a predetermined period based on the aerodynamic pressure data to analyze the comprehensive drift characteristics of each rotating fan surface of the target fan. The comprehensive drift characteristics include frequency offset characteristics and pressure waveform offset characteristics.

[0066] Step S3: Select rotating fan surfaces with potential hazards based on the comprehensive drift characteristics;

[0067] Step S4: Based on the infrared thermal image time series data, extract several frames of infrared freeze-frame images corresponding to the rotating fan surface with potential hidden dangers within the time domain segment, and compare them with standard infrared images to verify the infrared similarity.

[0068] Step S5: In response to non-compliance with infrared verification standards, perform regularity verification based on infrared freeze-frame images, and conduct fault analysis based on the regularity verification results, including...

[0069] Identify the faulty rotating sector;

[0070] Adjust the target fan's operating parameters, verify whether the potential hazard has been eliminated, and verify the infrared similarity to determine whether to issue a warning signal.

[0071] In practice, there are no restrictions on the methods for collecting the passing time of the target fan blade tip, aerodynamic pressure data, and infrared thermal image timing data. The passing time of the target fan blade tip can be collected by non-contact sensors such as laser displacement sensors or eddy current sensors installed on the stationary casing of the fan or nearby supports. Transient aerodynamic pressure pulses can be captured by installing a high-frequency response dynamic pressure sensor on a stationary component near the blade tip sweep path. Infrared thermal image timing data can be obtained by using an online infrared thermal imager installed at a fixed position on the side of the fan or by using a handheld infrared thermal imager that is inspected periodically. As long as there is a precise timing alignment relationship or a unified synchronous acquisition clock reference, it is acceptable. This will not be elaborated further.

[0072] In implementation, there are no restrictions on the method for extracting the transient pressure pulse waveform sequence of the blade tip within a predetermined period. It can be achieved through time window segmentation or peak trigger capture. The former, based on precisely synchronized blade tip passage times, extracts a fixed-length data segment forward and backward from each passage time center on the continuous pressure data stream, thus directly obtaining the pressure pulse waveform corresponding to each blade. The latter first identifies local peaks in the pressure data, then extracts waveform segments based on these peak points, and finally classifies these waveform segments into the corresponding rotating sector according to the temporal relationship of the blade tip passage times. It is only necessary to ensure that each extracted transient pressure pulse waveform uniquely and accurately corresponds to a specific blade tip passage event, i.e., a specific rotating sector; further details are omitted here.

[0073] Specifically, the process based on recording the passing frequency reference of each blade tip of the target fan over time includes,

[0074] Based on the passage time of the target fan blade tip, calculate the average rotation period of the target fan blade tip within a predetermined time.

[0075] The reciprocal of the average rotation period is determined as the blade tip passage frequency reference.

[0076] In practice, the predetermined time is usually selected within the range of [10s, 60s], and is preferably 30s.

[0077] Specifically, the process of analyzing the comprehensive drift characteristics of each rotating fan surface of the target fan includes,

[0078] Based on the frequency reference of each blade tip, the frequency offset value corresponding to the pressure pulse waveform is calculated and determined as the frequency offset feature;

[0079] Extract each transient pressure pulse waveform from each of the aforementioned transient pressure pulse waveform sequences;

[0080] Calculate the similarity between the transient pressure pulse waveform and the standard pressure pulse waveform point by point, and determine the mean similarity value;

[0081] The reciprocal of the mean similarity value is determined as the pressure waveform offset feature;

[0082] The frequency offset feature and the pressure waveform offset feature are weighted and summed to obtain the comprehensive drift feature.

[0083] In implementation, the standard pressure pulse waveform is predetermined. Those skilled in the art can collect the blade tip transient pressure pulse waveforms of multiple consecutive rotation cycles as the original sample library during the healthy operation phase after the target fan has passed acceptance. By calculating the correlation coefficient between each sample waveform and the temporary mean waveform, outlier abnormal samples with correlation coefficients lower than a preset threshold are removed. The preset threshold is preferably 0.95. The retained healthy samples are phase aligned using interpolation resampling technology to eliminate phase drift caused by speed fluctuations. Finally, the point-by-point arithmetic mean of all aligned samples is calculated, and the waveform of this average value is determined as the standard pressure pulse waveform.

[0084] In practice, there are no restrictions on the method for determining the similarity between the transient pressure pulse waveform and the standard pressure pulse waveform. Feature extraction of the pressure pulse waveform can be performed using image processing algorithms, or correlation coefficient analysis can be used to determine the similarity between the two waveforms. The only requirement is to ensure that the waveform accurately reflects the degree of similarity between the two; further details will not be elaborated upon here.

[0085] In practice, when the frequency offset feature and the pressure waveform offset feature are weighted and summed, the weight of the frequency offset feature is 0.55, and the weight of the pressure waveform offset feature is 0.45.

[0086] This invention extracts the transient pressure pulse waveform sequence of the blade tip within a predetermined period based on the aerodynamic pressure data to analyze the comprehensive drift characteristics of each rotating blade of the target fan. Since the effects of potential problems such as blade cracks, surface wear, or uneven dust accumulation are extremely weak in the initial stage, their physical characteristics are often neither simply increased vibration nor a significant overall temperature rise, but rather manifest as a subtle, coordinated change in the rotational dynamics of a specific blade and its local aerodynamic properties. For example, each high-speed rotating blade of the fan periodically compresses the air in front of it at its tip, generating a weak pressure wave; each blade generates a pressure pulse as it passes by. Ideally, all blades are completely identical, so these pulses should be equally spaced, of the same shape, and of the same amplitude. When blades have cracks, wear, or dust accumulation, the morphology and periodicity of these pressure pulses change, leading to pressure waveform and frequency shifts. Therefore, integrating frequency and pressure waveform shift characteristics to determine comprehensive drift characteristics is based on profound physical mechanisms and monitoring logic: frequency shift characteristics capture, in the time domain, the periodic dynamic imbalance trend caused by minute changes in fan mass or aerodynamic drag torque, which is strictly synchronized with rotation, representing a macroscopic integrated reflection of the fault; while pressure waveform shift characteristics, in the spatial domain, capture, through the analysis of pulse waveform morphology, the characteristic aerodynamic disturbances caused by microscopic changes in blade geometry or surface condition. These two parameters comprehensively cover the dynamic and aerodynamic characteristics of the blades from both the time and spatial domains, providing a reliable basis for subsequent fault analysis and improving the reliability of early fault warnings for fan operation.

[0087] Please see Figure 2 As shown, this is a logic diagram for selecting a rotating fan surface with a potential for hidden dangers according to an embodiment of the invention. Specifically, the process of selecting a rotating fan surface with a potential for hidden dangers includes,

[0088] If the overall drift characteristic is greater than or equal to the preset overall drift threshold, then the rotating fan surface is selected as a rotating fan surface with potential hidden dangers.

[0089] In implementation, the purpose of the comprehensive drift threshold is to characterize the difference in features between fan blades under normal operating conditions and under conditions with potential hazards, in order to distinguish between normal operation and situations with a tendency to have potential hazards. The comprehensive drift threshold is predetermined, and is typically selected within the range of [0.5, 0.9], with 0.7 being preferred in implementation.

[0090] Specifically, the process of comparing the infrared images with standard infrared images to verify infrared similarity includes,

[0091] Determine the average image similarity between each infrared freeze-frame image and the standard infrared image;

[0092] If the average image similarity is less than the image similarity threshold, then the rotating fan-shaped surface with potential risks does not meet the infrared verification standard.

[0093] In practice, the standard infrared image is predetermined. Those skilled in the art can, during the healthy operation phase after the target fan has passed acceptance testing, acquire infrared images of each rotating fan surface within multiple rotation cycles using an infrared thermal imager as a raw sample library; denoise the raw images to eliminate background thermal noise interference, and use an image registration algorithm to rotate and align each frame to eliminate displacement caused by minor fan vibrations or shooting angle deviations; finally, calculate the pixel-by-pixel grayscale average of all aligned images, and determine this average image as the standard infrared image corresponding to that rotating fan surface.

[0094] In implementation, the purpose of the image similarity threshold is to characterize the critical limit at which the infrared thermal distribution of the tested fan deviates from the standard healthy state. The image similarity threshold is predetermined. Those skilled in the art can collect a large number of standard infrared image samples of the rotating fan during stable, fault-free operation and determine their average image similarity to represent the stability of the infrared thermal distribution under normal conditions. To indicate potential problems, the image similarity threshold is set to a predetermined multiple of the average value. Typically, the predetermined multiple is selected within the range of [0.65, 0.85], and is preferably 0.75 in implementation.

[0095] This invention extracts several frames of infrared still images corresponding to rotating fan surfaces with potential hazards within a time domain segment based on the time-series data of infrared thermal images. These images are then compared with standard infrared images to verify infrared similarity. In complex industrial environments, potential hazards identified earlier based on dynamic and aerodynamic characteristics may originate from occasional interferences such as instantaneous airflow disturbances, slight load fluctuations, or sensor noise. While these interferences may cause signal drift, they do not necessarily correspond to actual physical damage. However, most progressive faults in rotating machinery, such as crack propagation, increased friction, and worsening imbalance, are essentially energy dissipation processes. Ultimately, these faults will manifest as abnormal thermodynamic characteristics at the fault location, namely, local temperature rise or changes in heat distribution patterns. For example, when blades have cracks or wear, the thermal conductivity of these areas changes, leading to local temperature increases or decreases. This results in thermal characteristics in infrared images that differ from those of normal blades. Infrared thermal imaging technology can capture these thermo-radiative signals generated by the fault mechanism itself non-contactly and with high spatial resolution. Verifying infrared similarity provides a crucial physical verification step for the previously identified potential hazards. Specifically, the system will lock onto the rotating fan surface with potential problems and extract multiple frames of infrared images of it over a period of time. By comparing them with standard infrared images, the similarity of their thermal images is quantified. The verification method based on infrared thermal imaging can not only provide fault characteristics independent of aerodynamic and mechanical properties, but also confirm potential problems from a thermodynamic perspective, thereby further improving the reliability of early fault warnings for fan operation status.

[0096] Please see Figure 3 As shown, it is a logic block diagram of a regularity verification based on infrared freeze-frame images according to an embodiment of the invention. Specifically, the process of regularity verification based on infrared freeze-frame images includes:

[0097] Identify several frames of infrared freeze-frame images corresponding to the rotating fan that does not meet the infrared verification standard, and determine the inter-frame similarity.

[0098] Determine the mean value of the inter-frame similarity;

[0099] If the mean of inter-frame similarity is greater than or equal to the preset inter-frame similarity threshold, then the rotating fan that does not meet the infrared verification standard is verified to have regularity.

[0100] If the mean of inter-frame similarity is less than the preset inter-frame similarity threshold, then the rotating fan that does not meet the infrared verification standard is found to be irregular.

[0101] In implementation, the purpose of the inter-frame similarity threshold is to quantify whether an infrared abnormal thermal feature has a stable repetitive occurrence pattern that is strictly synchronized with the rotation cycle. The inter-frame similarity threshold is predetermined. Those skilled in the art can collect infrared image sequences of the same healthy rotating fan surface over multiple consecutive rotation cycles during a period of stable, fault-free fan operation, calculate the similarity between every two adjacent frames in the sequence, and determine their average value. This represents the inherent normal random fluctuation level between adjacent frames under normal conditions, caused by a combination of factors such as transient changes in environmental radiation background, sensor background noise, and non-periodic thermal flow disturbances. To distinguish between the "normal random fluctuations" under healthy conditions and the highly repetitive patterns caused by faults, the inter-frame similarity threshold is set as the product of the average value and the inter-frame precision coefficient. Typically, the inter-frame precision coefficient is selected within the range of [0.75, 1.05], and is preferably 0.85 in implementation.

[0102] Specifically, the fault analysis based on regularity verification results includes,

[0103] If the rotating sector that does not meet the infrared verification standard shows a regularity, then the faulty rotating sector is identified.

[0104] If the rotating fan surface that does not meet the infrared verification standard is irregular, the operating parameters of the target fan are adjusted to verify whether the potential hazard has been eliminated and to verify the infrared similarity in order to determine whether to issue a warning signal.

[0105] This invention utilizes infrared freeze-frame images for regularity verification, and then performs fault analysis based on the verification results. By extracting and analyzing infrared image sequences of suspected fan surfaces across multiple consecutive rotation cycles and calculating their inter-frame similarity, it quantifies whether the abnormal thermal features exhibit a recurring pattern synchronized with the rotational speed. If the verification demonstrates regularity, it proves that the thermal anomaly source is firmly embedded in the rotating fan surface itself, strongly coupled with the rotational motion, thus directly identifying the faulty rotating fan surface and achieving precise location. Conversely, if the verification lacks regularity, it indicates that the initially detected thermal anomaly is more likely due to transient interference or a holistic, non-local operational state shift, such as sudden changes in system load or fluctuations in cooling conditions, rather than structural damage to a specific fan surface. In this case, blindly shutting down for repair is unnecessary; a more reasonable strategy is to adjust the overall operating parameters of the target fan and observe whether the previously determined comprehensive drift characteristics and infrared similarity subsequently meet the threshold under this adjustment. This design upgrades fault diagnosis from a single feature alarm to an intervention-oriented and verifiable systematic approach. It not only avoids invalid warnings triggered by occasional interference, but also provides key decision-making basis for distinguishing between local fixed faults and overall transient imbalances through systematic parameter control experiments, thereby improving the reliability of early fault warnings for fan operation status.

[0106] Specifically, the process of verifying whether the potential risks have been eliminated includes,

[0107] If the overall drift characteristic of the rotating fan surface after adjusting the target fan's operating parameters is greater than or equal to the overall drift threshold, then the potential hazard has not been eliminated.

[0108] If the overall drift characteristic of the rotating fan surface after adjusting the target fan's operating parameters is less than the overall drift threshold, then the potential hazard is verified to be eliminated.

[0109] Please see Figure 4 The diagram shown is a logic block diagram for determining whether to issue a warning signal according to an embodiment of the invention. Specifically, the process of determining whether to issue a warning signal includes:

[0110] The infrared similarity between the infrared freeze-frame image corresponding to the rotating fan surface after adjusting the target fan's operating parameters and the standard infrared image was determined to verify the infrared similarity.

[0111] If the potential hidden danger is not eliminated, or if the infrared similarity is less than the preset infrared similarity threshold, an early warning signal will be issued.

[0112] If the potential hazard is verified to be eliminated, or if the infrared similarity is greater than or equal to a preset infrared similarity threshold, then no warning signal will be issued.

[0113] In practice, there are no restrictions on the way to adjust the operating parameters of the target fan. The fan speed can be adjusted by changing the motor drive frequency or by adjusting the system load. As long as the adjustment can actively and controllably change the operating state of the target fan, it can provide a basis for observing and verifying the changing trend and response of the comprehensive drift characteristics and infrared similarity under this new state. This will not be elaborated further.

[0114] In implementation, the purpose of the infrared similarity threshold is to characterize whether the infrared thermal state of the detected rotating fan surface has recovered to an acceptable healthy range after adjusting the operating parameters. Specifically, those skilled in the art can, during the fan's stable, fault-free operation, repeatedly adjust the operating parameters and collect infrared images under corresponding conditions, calculating the average similarity between these images and a benchmark infrared image. This represents the permissible level of normal difference in infrared thermal images between different stable operating conditions under normal circumstances. When the similarity between the adjusted infrared image and the standard image exceeds the average level of normal difference, i.e., performs better than daily fluctuations, the thermal state is considered to have fully recovered. Therefore, the infrared similarity threshold is set as the product of the average similarity and the error coefficient. Typically, the error coefficient is selected within the range of [0.7, 0.95], and preferably 0.8 in implementation.

[0115] In practice, there are no restrictions on the way the warning signal is issued. It can be issued by triggering an audible and visual alarm device or by sending alarm information to a remote monitoring terminal. As long as the warning signal can promptly and accurately notify relevant personnel or systems to take appropriate measures, it is acceptable.

[0116] Specifically, a system that provides a fan operation control method based on the Internet of Things (IoT) is also provided, including:

[0117] The acquisition module is used to acquire the passing time of the target fan blade tip, aerodynamic pressure data, and infrared thermal image time series data;

[0118] The feature analysis module, which is connected to the acquisition module, is used to extract the transient pressure pulse waveform sequence of the blade tips within a predetermined period based on the aerodynamic pressure data, based on the passing frequency reference of each blade tip of the target fan recorded over time, so as to analyze the comprehensive drift characteristics of each rotating blade of the target fan, including frequency offset characteristics and pressure waveform offset characteristics.

[0119] A selected module is connected to the analytical feature module to select rotating fan surfaces with potential risks based on the comprehensive drift features.

[0120] The verification module is connected to the acquisition module and the selection module respectively, and is used to extract several frames of infrared freeze-frame images corresponding to the rotating fan with potential hidden dangers within the time domain segment based on the infrared thermal image time series data, and compare them with standard infrared images to verify the infrared similarity.

[0121] The fault analysis module, connected to the acquisition module, feature analysis module, selection module, and verification module, is used to perform regularity verification based on infrared freeze-frame images in response to non-compliance with infrared verification standards, and to perform fault analysis based on the regularity verification results, including...

[0122] Used to identify faulty rotating fan surfaces;

[0123] It is used to adjust the operating parameters of the target fan, verify whether the potential hazards have been eliminated, and verify the infrared similarity in order to determine whether to issue a warning signal.

[0124] In implementation, there are no restrictions on the method for extracting several frames of infrared freeze-frame images corresponding to the rotating fan surface with potential hazards within the time domain. It can be achieved through hardware synchronous triggering, using the pulse signal generated by the sensor through the leaf tip to trigger the infrared thermal imager to capture images when the target fan surface passes through a specific phase; or through software post-processing, based on the synchronously recorded leaf tip passage time and rotation period information, extracting the corresponding fan surface image frames from the continuous infrared video stream using an image registration algorithm. It is only necessary to ensure that the extracted frames of infrared freeze-frame images all correspond to the same physical rotating fan surface and have a precise temporal correspondence; this will not be elaborated further.

[0125] In implementation, there are no restrictions on the structure of the parsing feature module, selection module, verification module, and fault analysis module. They can be composed of logic components or combinations of logic components. Logic components include field-programmable processors, computers, or microprocessors in computers, which will not be elaborated further.

[0126] In implementation, the structure of the acquisition module is not limited and can be composed of a combination of various sensors. For example, a laser displacement sensor can be used to acquire the passage time of the target fan blade tip, a high-frequency response dynamic pressure sensor can be used to acquire aerodynamic pressure data, and an online infrared thermal imager or a handheld infrared thermal imager can be used to acquire infrared thermal image time-series data. As long as the sensors can accurately and stably acquire the required data, this will not be elaborated further.

[0127] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A fan operation control method based on the Internet of Things, characterized in that, include, Step S1: Collect the passing time, aerodynamic pressure data, and infrared thermal image time series data of the target fan blade tip; Step S2: Based on the passing frequency reference of each blade tip of the target fan recorded over time, extract the transient pressure pulse waveform sequence of the blade tip within a predetermined period based on the aerodynamic pressure data to analyze the comprehensive drift characteristics of each rotating fan surface of the target fan, wherein the comprehensive drift characteristics include frequency offset characteristics and pressure waveform offset characteristics. Step S3: Select rotating fan surfaces with potential hazards based on the comprehensive drift characteristics; Step S4: Based on the infrared thermal image time series data, extract several frames of infrared freeze-frame images corresponding to the rotating fan surface with potential hidden dangers within the time domain segment, and compare them with standard infrared images to verify the infrared similarity. Step S5: In response to non-compliance with infrared verification standards, perform regularity verification based on infrared freeze-frame images, and conduct fault analysis based on the regularity verification results, including... Identify the faulty rotating sector; Adjust the target fan's operating parameters, verify whether the potential hazard has been eliminated, and verify the infrared similarity to determine whether to issue a warning signal.

2. The fan operation control method based on the Internet of Things according to claim 1, characterized in that, The process based on the passing frequency reference of each blade tip of the target fan through time includes: Based on the passage time of the target fan blade tip, calculate the average rotation period of the target fan blade tip within a predetermined time. The reciprocal of the average rotation period is determined as the blade tip passage frequency reference.

3. The fan operation control method based on the Internet of Things according to claim 1, characterized in that, The process of analyzing the comprehensive drift characteristics of each rotating fan surface of the target fan includes, Based on the frequency reference of each blade tip, the frequency offset value corresponding to the pressure pulse waveform is calculated and determined as the frequency offset feature; Extract each transient pressure pulse waveform from each of the aforementioned transient pressure pulse waveform sequences; Calculate the similarity between the transient pressure pulse waveform and the standard pressure pulse waveform point by point, and determine the mean similarity value; The reciprocal of the mean similarity value is determined as the pressure waveform offset feature; The frequency offset feature and the pressure waveform offset feature are weighted and summed to obtain the comprehensive drift feature.

4. The fan operation control method based on the Internet of Things according to claim 1, characterized in that, The process of selecting rotating fan surfaces with potential hazards includes, If the overall drift characteristic is greater than or equal to the preset overall drift threshold, then the rotating fan surface is selected as a rotating fan surface with potential hidden dangers.

5. The fan operation control method based on the Internet of Things according to claim 1, characterized in that, The process of comparing the infrared images with standard infrared images to verify infrared similarity includes: Determine the average image similarity between each infrared freeze-frame image and the standard infrared image; If the average image similarity is less than the image similarity threshold, then the rotating fan-shaped surface with potential risks does not meet the infrared verification standard.

6. The fan operation control method based on the Internet of Things according to claim 1, characterized in that, The process of verifying regularity based on infrared freeze-frame images includes, Identify several frames of infrared freeze-frame images corresponding to the rotating fan that does not meet the infrared verification standard, and determine the inter-frame similarity. Determine the mean value of the inter-frame similarity; If the mean of inter-frame similarity is greater than or equal to the preset inter-frame similarity threshold, then the rotating fan that does not meet the infrared verification standard is verified to have regularity. If the mean of inter-frame similarity is less than the preset inter-frame similarity threshold, then the rotating fan that does not meet the infrared verification standard is found to be irregular.

7. The fan operation control method based on the Internet of Things according to claim 6, characterized in that, The fault analysis based on the regularity verification results includes, If the rotating sector that does not meet the infrared verification standard shows a regularity, then the faulty rotating sector is identified. If the rotating fan surface that does not meet the infrared verification standard is irregular, the operating parameters of the target fan are adjusted to verify whether the potential hazard has been eliminated and to verify the infrared similarity in order to determine whether to issue a warning signal.

8. The fan operation control method based on the Internet of Things according to claim 1, characterized in that, The process of verifying whether the potential risks have been eliminated includes... If the overall drift characteristics of the rotating fan surface after adjusting the target fan's operating parameters are greater than or equal to the overall drift threshold, then the potential hazard has not been eliminated. If the overall drift characteristic of the rotating fan surface after adjusting the target fan's operating parameters is less than the overall drift threshold, then the potential hazard has been verified to be eliminated.

9. The fan operation control method based on the Internet of Things according to claim 8, characterized in that, The process of determining whether to issue a warning signal includes... The infrared similarity between the infrared freeze-frame image corresponding to the rotating fan surface after adjusting the target fan's operating parameters and the standard infrared image was determined to verify the infrared similarity. If the potential hidden danger is not eliminated, or if the infrared similarity is less than the preset infrared similarity threshold, an early warning signal will be issued. If the potential hazard is verified to be eliminated, or if the infrared similarity is greater than or equal to a preset infrared similarity threshold, then no warning signal will be issued.

10. A system applying the Internet of Things-based fan operation control method according to any one of claims 1-9, characterized in that, include: The acquisition module is used to acquire the passing time of the target fan blade tip, aerodynamic pressure data, and infrared thermal image time series data; The feature analysis module, which is connected to the acquisition module, is used to extract the transient pressure pulse waveform sequence of the blade tip within a predetermined period based on the aerodynamic pressure data, based on the passing frequency reference of each blade tip of the target fan recorded over time, so as to analyze the comprehensive drift characteristics of each rotating blade of the target fan. The comprehensive drift characteristics include frequency offset characteristics and pressure waveform offset characteristics. A selected module is connected to the analytical feature module to select rotating fan surfaces with potential risks based on the comprehensive drift features. The verification module is connected to the acquisition module and the selection module respectively, and is used to extract several frames of infrared freeze-frame images corresponding to the rotating fan with potential hidden dangers within the time domain segment based on the infrared thermal image time series data, and compare them with standard infrared images to verify the infrared similarity. The fault analysis module, connected to the acquisition module, feature analysis module, selection module, and verification module, is used to perform regularity verification based on infrared freeze-frame images in response to non-compliance with infrared verification standards, and to perform fault analysis based on the regularity verification results, including... Used to identify faulty rotating fan surfaces; It is used to adjust the operating parameters of the target fan, verify whether the potential hazards have been eliminated, and verify the infrared similarity in order to determine whether to issue a warning signal.

Images