Digital equipment fault prediction method and system

Abstract

The invention discloses a digital equipment fault prediction method and system, and belongs to the technical field of data processing, and the method specifically comprises the steps: collecting the input power, output flow and pressure of digital equipment in real time, calculating the energy loss, and building an energy consumption database; vibration and acoustic sensors are used for monitoring signals, a noise source and the relation between the noise source and energy consumption are recognized through frequency domain analysis, and noise reduction measures are taken; historical data is collected to train a neural network model, a fault pre-judgment model is constructed, potential faults are predicted in real time and early warning is carried out, and meanwhile operation parameters are automatically adjusted according to an energy consumption analysis result; automatically starting a noise reduction device based on a noise reduction analysis result, and providing a noise treatment scheme; energy efficiency optimization and noise control of the digital equipment are realized, and the operation efficiency and reliability of the equipment are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing, and specifically relates to a digital device fault prediction method and system. Background Art

[0002] Digital devices are widely used in many fields such as industrial production. Traditional data analysis methods for digital devices are often limited to single-dimensional data monitoring and analysis, lacking systematic and effective analysis and comprehensive solutions for problems such as energy consumption, noise, and faults during the operation of digital devices, resulting in problems such as resource waste, shortened equipment lifespan, and reduced production efficiency. Therefore, it is of great significance to develop a comprehensive digital device data analysis method.

[0003] For example, the patent application with the publication number CN115859023A discloses a standardized format processing method and system for key data acquisition of a pump-motor system, including: obtaining key data signals of the pump-motor system in real time through sensors, amplifying them by a signal amplifier; then converting them into digital signals by a data collector and connecting them to an embedded system for standardized format processing; transmitting the key data signals after standardized format from the embedded system to a host computer for subsequent data analysis, completing the status test or fault diagnosis of the pump-motor, and uploading the data to the network. This technical solution is applicable to multiple sensors simultaneously perceiving parameters of multiple key data to ensure that various types of key data are collected under the same working condition of the pump-motor system; it is beneficial to more accurately judge the operating status of the pump-motor system or the existing equipment faults.

[0004] The above existing technologies all have the following problems: lacking fault diagnosis measures and not involving noise reduction measures. Summary of the Invention

[0005] In view of the deficiencies of the prior art, the present invention proposes a digital device fault prediction method and system, which can collect the input power, output flow, and pressure of digital devices in real time, calculate the energy loss, and establish an energy consumption database; monitor signals using vibration and acoustic sensors, identify noise sources and their relationship with energy consumption through frequency domain analysis, and take noise reduction measures; collect historical data to train a neural network model, construct a fault prediction model, predict potential faults in real time and give early warnings. At the same time, automatically adjust the operating parameters according to the energy consumption analysis results; automatically start the noise reduction device based on the noise reduction analysis results and provide a noise control plan; achieve energy efficiency optimization and noise control of digital devices, and improve the operating efficiency and reliability of the devices.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A digital device fault prediction method, including:

[0008] Obtain the input power, output flow rate, and pressure of the digital device through sensors, calculate the energy loss under different working conditions, and establish an energy consumption database based on the calculation results;

[0009] Real-time monitor the vibration signal and sound pressure signal of the digital device, analyze the frequency components of the noise, and take corresponding noise reduction measures in combination with the energy consumption database;

[0010] Collect the historical data during the operation of the digital device and construct a fault prediction model. The historical data includes energy consumption data, noise data, temperature data, and pressure data;

[0011] According to the fault prediction model, predict whether there are potential faults in the digital device and the types of faults, and generate an adjustment plan for the digital device based on the prediction results.

[0012] Specifically, the calculation of the energy loss under different working conditions includes:

[0013] Use a power sensor to collect the input power of the digital device in real time, and measure the output flow rate and pressure through a flow sensor and a pressure sensor;

[0014] Calculate the output power P according to the output flow rate and pressure out , and the formula is: where ρ represents the liquid density, Q represents the output flow rate of the digital device, g represents the acceleration due to gravity, H represents the head, η f represents the fluid efficiency coefficient, η l represents the energy loss coefficient, T represents the current temperature, T0 represents the reference temperature, α represents the temperature influence coefficient, C represents the current concentration, C0 represents the reference concentration, β represents the concentration influence coefficient, δ represents the time decay coefficient, t represents the operating time, and e represents the exponential;

[0015] According to the output power P out and the input power P in , obtain the energy loss loss = P in -P out .

[0016] Specifically, the real-time monitoring of the vibration signal and sound pressure signal of the digital device, analyzing the frequency components of the noise, and taking corresponding noise reduction measures in combination with the energy consumption database include:

[0017] Use vibration sensors and acoustic sensors to monitor the vibration signal and sound pressure signal in real time, and convert the time-domain signal into a frequency-domain signal through Fourier transform;

[0018] Based on the frequency-domain signal, analyze the frequency components of the noise, and combine the data in the energy consumption database to identify the noise sources of different frequencies and their relationships with energy consumption; the data in the energy consumption database are the associated data of energy loss and operating conditions.

[0019] According to the identification results, determine the noise sources and take corresponding noise reduction measures, which include optimizing the component structure to reduce mechanical vibration and improving the flow channel design to reduce fluid pulsation.

[0020] Specifically, the construction process of the fault prediction model includes:

[0021] Collect historical data during the operation of digital devices and perform tagging processing on the historical data. The historical data includes energy consumption data, noise data, temperature data, and pressure data.

[0022] Load the pre-constructed neural network model, and use the tagged historical data to train the pre-constructed neural network model to generate a trained fault prediction model.

[0023] Specifically, according to the fault prediction model, predict whether there are potential faults in the digital device and the types of faults. According to the prediction results, generate a digital device adjustment plan, including:

[0024] Real-time collect the operation data of digital devices and perform preprocessing; the operation data of digital devices includes flow rate, sound, vibration, temperature, current, electrical protection component tripping, and controller alarm.

[0025] Use signal processing methods to extract features from the preprocessed operation data to generate digital device operation feature data.

[0026] Input the digital device operation feature data into the trained fault prediction model. The fault prediction model predicts whether there are potential faults and the types of faults in the digital device according to the matching degree between the digital device operation feature data and the historical fault data features.

[0027] After the fault prediction model predicts that a fault may occur, it issues a warning message. The operator immediately responds and generates a digital device adjustment plan through threshold judgment.

[0028] Specifically, the generation of the digital device adjustment plan through threshold judgment includes:

[0029] Judge whether the energy loss exceeds the preset threshold. If it exceeds the preset threshold, automatically adjust the operation parameters of the digital device.

[0030] Determine whether the noise exceeds the preset allowable noise range. If it exceeds the preset allowable noise range, automatically activate the noise reduction device. The noise reduction device includes a shock pad and a muffler. At the same time, according to the analysis result of the noise frequency, provide a noise control solution;

[0031] Judge the type of fault according to the output result of the fault prediction model. According to the judgment result of the fault type, automatically adjust the operation mode of the digital device, including reducing the operation speed and switching to standby components.

[0032] Specifically, based on the frequency domain signal, analyze the frequency components of the noise, and combine the data in the energy consumption database to identify the noise sources of different frequencies and their relationship with energy consumption, including:

[0033] Real-time monitor the sound pressure signal of the digital device through an acoustic sensor, and convert the sound pressure signal into a frequency domain signal through Fourier transform;

[0034] Conduct spectrum analysis on the frequency domain signal to identify the main frequency components of the noise and evaluate the noise levels of each frequency component;

[0035] Extract the energy consumption data related to the current working condition from the energy consumption database, and analyze the potential correlation between the energy consumption data and the noise frequency components;

[0036] According to the known noise frequency components and energy consumption data, combined with the working principle and structural characteristics of the digital device, identify the noise sources; the noise sources include mechanical vibration, fluid pulsation, and electromagnetic interference;

[0037] For each identified noise source, analyze the relationship between the noise sources of different frequencies and energy consumption.

[0038] Specifically, the extraction of the energy consumption data related to the current working condition from the energy consumption database and the analysis of the potential correlation between the energy consumption data and the noise frequency components include:

[0039] Obtain the current operating condition of the digital device;

[0040] According to the operating condition, query and extract the corresponding energy consumption data from the energy consumption database, and preprocess the extracted energy consumption data; the energy consumption data includes power consumption, mechanical energy loss, and heat energy loss;

[0041] Obtain the noise frequency components and their corresponding noise levels of the digital device, and use the correlation analysis method to analyze the potential correlation between the energy consumption data and the noise frequency components; the potential correlation refers to whether there is a correlation between the energy consumption parameters and the noise of the frequency, and the correlation includes positive correlation and negative correlation;

[0042] Interpret the physical meaning of the correlation between energy consumption data and noise frequency components based on the analysis results, and provide optimization suggestions according to the analysis results.

[0043] A digital device fault prediction system, comprising: an energy consumption analysis module, a noise analysis module, a fault prediction module, a noise reduction control module, and a fault handling module;

[0044] The energy consumption analysis module is used to collect the input power, output flow rate, and pressure of the digital device in real time, calculate the energy loss under different working conditions, and establish an energy consumption database;

[0045] The noise analysis module is used to monitor the vibration signal and sound pressure signal in real time, analyze the frequency components of the noise, identify the noise source and its relationship with energy consumption;

[0046] The fault prediction module trains a neural network model based on the collected historical data, and predicts the potential faults and fault types of the digital device in real time. According to the prediction results, it automatically adjusts the operating parameters;

[0047] The noise reduction control module automatically starts the noise reduction device based on the noise reduction analysis results and provides a noise control solution;

[0048] The fault handling module is used to automatically adjust the operating mode of the digital device according to the fault type after the fault prediction model issues a warning, and record the fault handling process and results.

[0049] Compared with the prior art, the beneficial effects of the present invention are:

[0050] The present invention provides a digital device fault prediction method and system. Through systematic analysis of energy consumption and noise reduction, as well as fault prediction and control, it can effectively improve the energy utilization efficiency of digital devices, reduce noise pollution, detect potential faults in advance and handle them effectively, extend the service life of the devices, and improve production efficiency; at the same time, by collecting the input power, output flow rate, and pressure of digital devices in real time and calculating the energy loss, it can accurately grasp the energy consumption of digital devices under different working conditions, and then establish an energy consumption database to provide data support for subsequent energy efficiency optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a schematic diagram of a digital device fault prediction method of the present invention;

[0052] Figure 2 It is a principle flow chart of a digital device fault prediction method of the present invention;

[0053] Figure 3 It is a structural diagram of a digital device of the present invention;

[0054] Figure 4This is the architecture diagram of a digital device fault prediction system according to the present invention.

[0055] Reference numerals:

[0056] 1. Digital motor; 1a. Hybrid magnetic motor; 1c. Heat dissipation device; 2. LCD touch screen; 3. Bracket; 4. Centrifugal pump head; 4a. Pump body; 4b. Shaft; 4c. Pump cover; 5. Inlet; 6. Lifting device; 7. Multifunctional sensor; 8. Outlet; 9. Aviation plug. Detailed implementation manners

[0057] Embodiment 1

[0058] Please refer to Figures 1 - 3 , an embodiment provided by the present invention: A digital device fault prediction method. In the present invention, the digital device refers to a digital pump. In Figure 3 , the digital device includes a digital motor 1, an LCD touch screen 2, a bracket 3, a centrifugal pump head 4, a multifunctional sensor 7, and a lifting device 6; the digital motor 1 includes a hybrid magnetic motor 1a and a heat dissipation device 1c. The hybrid magnetic motor 1a forms the lower part of the digital motor 1, and the heat dissipation device 1c forms the upper part of the digital motor 1. An aviation plug 9 is provided on the digital motor 1, and the power supply is connected through the aviation plug 9. The LCD touch screen 2 is embedded in the front side of the digital motor 1. The bracket 3 is arranged below the digital motor 1 through bolt connection. The centrifugal pump head 4 is arranged below the bracket 3 through bolt connection. The centrifugal pump head 4 includes a pump body 4a, a pump cover 4c, and a shaft 4b. The output shaft of the digital motor 1 extends into the centrifugal pump head 4 and is used as the shaft of the centrifugal pump head 4. The digital motor 1 drives the impeller in the centrifugal pump head 4 to rotate through the shared shaft 4b. There is an inlet 5 on one side of the centrifugal pump head 4 and an outlet 8 on the other side. A multifunctional sensor 7 is arranged at the connection part between the bracket 3 and the pump cover 4c near the outlet 8. The probe of the multifunctional sensor 7 extends into the pump cavity formed by the pump body 4a and the pump cover 4c and is connected to the digital integrated controller in the digital device through the aviation plug 9. The lifting device 6 is connected to the digital motor 1 by bolt connection. The method includes S101 to S104, specifically:

[0059] S101: Obtain the input power, output flow rate, and pressure of the digital device through the sensor, calculate the energy loss under different working conditions, and establish an energy consumption database according to the calculation results;

[0060] S102: Real-time monitor the vibration signal and sound pressure signal of the digital device, analyze the frequency components of the noise, and take corresponding noise reduction measures in combination with the energy consumption database;

[0061] S103: Collect the historical data during the operation of the digital device, and construct a fault prediction model. The historical data includes energy consumption data, noise data, temperature data, and pressure data;

[0062] S104: Predict whether there are potential faults and the types of faults to which the digital device belongs according to the fault prediction model, and generate a digital device adjustment plan according to the prediction results.

[0063] The calculation of the energy loss under different working conditions includes:

[0064] A1: Use a power sensor to collect the input power of the digital device in real time, and measure the output flow rate and pressure through a flow sensor and a pressure sensor;

[0065] It should be noted that pressure is one of the important parameters for evaluating the working conditions of digital devices. By measuring the pressure, the operating state of the digital device under specific working conditions can be understood, so as to ensure that it works within a safe and stable range. On the one hand, although pressure is not directly involved in calculating the output power, in some cases, pressure may affect the density or flow rate of the fluid, thereby indirectly affecting the output power. For example, when the pressure increases, the density of the fluid may change, which in turn affects the calculation of the output power; on the other hand, measuring the pressure can also help with fault prediction and diagnosis. If the pressure of the digital device is abnormal, it may mean that there are certain faults or problems, such as blockage, leakage, or component wear. By monitoring the pressure change, these problems can be discovered and processed in time to avoid the further expansion of the faults.

[0066] Among them, the specific working condition refers to the working state of the digital device under specific operating conditions, and the specific conditions include the following aspects:

[0067] (1) Rotation speed or working frequency: When the digital device is at different rotation speeds or working frequencies, its energy consumption and the frequency components of the generated noise may be different. For example, when running at high speed, higher-frequency mechanical vibrations and fluid noises may be generated;

[0068] (2) Load condition: When the digital device bears different loads, its energy consumption and noise characteristics will also change. The increase in load may lead to an increase in energy consumption and an increase in the noise level;

[0069] (3) Fluid characteristics: For pump devices, the properties of the fluid will also affect its energy consumption and noise generation. For example, high-viscosity fluids may require more energy to transport and may generate greater fluid noise;

[0070] (4) Operating environment: The operating environment of the digital device, such as temperature, humidity, pressure, etc., may also affect its energy consumption and noise characteristics. For example, a high-temperature environment may cause the device efficiency to decrease, thereby increasing energy consumption; and pressure fluctuations may cause changes in fluid noise;

[0071] (5) Equipment status: The degree of wear, lubrication condition, and looseness of fasteners of the equipment can also affect its energy consumption and noise level. For example, worn bearings may generate abnormal mechanical noise and increase energy consumption;

[0072] In summary, in the energy consumption database, specific operating conditions are usually associated with energy loss data, which may come from laboratory tests, on-site monitoring, or historical operation records and are used to describe the energy consumption characteristics of equipment under different operating conditions.

[0073] A2: Calculate the output power P according to the output flow rate and pressure out , and the formula is:

[0074]

[0075] where ρ represents the liquid density, Q represents the output flow rate of the digital device, g represents the acceleration due to gravity, H represents the head, and η f represents the fluid efficiency coefficient, which refers to the efficiency improvement or loss of the fluid during transmission, and 0 ≤ η f ≤ 1, η l represents the energy loss coefficient, which is the energy loss caused by reasons such as friction and leakage, and 0 ≤ η l ≤ 1, T represents the current temperature, T0 represents the reference temperature, α represents the temperature influence coefficient, C represents the current concentration, C0 represents the reference concentration, β represents the concentration influence coefficient, δ represents the time decay coefficient, t represents the operating time, and e represents the exponential; the fluid efficiency coefficient is the efficiency of the liquid flowing in the equipment; the energy loss coefficient is the energy loss of the digital device during operation due to reasons such as friction and leakage; the temperature influence coefficient is the relative change in output power when the temperature changes; the concentration influence coefficient is the relative change in output power when the concentration changes; the time decay coefficient is the speed at which the performance of the digital device deteriorates during the change of time;

[0076] In the present invention, the values of the fluid efficiency coefficient, energy loss coefficient, temperature influence coefficient, concentration influence coefficient, and time decay coefficient are respectively: η f = 0.5, η l = 0.1, α = 0.5, β = 0.3, δ = 0.4.

[0077] It should be noted that the formula in this application can more accurately predict and evaluate the actual performance of the system by comprehensively considering multiple factors such as fluid density, flow rate, height, efficiency coefficient, energy loss coefficient, temperature, concentration, and time. The liquid density ρ, output flow rate Q, head H, and acceleration due to gravity g jointly determine the basic power output, while the fluid efficiency coefficient η f and the energy loss coefficient η lThe introduction reflects the energy conversion efficiency and loss situation of the system during the transmission process. When the fluid efficiency coefficient η f increases, the overall performance improves. When the energy loss coefficient η l increases, the overall performance decreases. At the same time, and respectively represent the influence degrees of temperature and concentration on the system performance. When the temperature and concentration deviate from the reference values, the system performance will change according to the corresponding coefficient relationship. This change may be positive or negative. Specifically: if α > 0, the performance improves with the increase of temperature; if β > 0, the performance improves with the increase of concentration. On the other hand, considering the performance change of the system over time, the time parameter t is introduced. Through the time decay coefficient δ, it reflects the performance degradation of the system due to factors such as aging and wear. To sum up, the modified formula provides a comprehensive evaluation of the performance of the fluid transmission system by introducing multiple parameters.

[0078] A3: According to the output power P out and the input power P in , the energy loss loss = P in - P out is obtained under different working conditions.

[0079] Vibrational signals and sound pressure signals of the real-time monitoring digital device are monitored, the frequency components of the noise are analyzed, and corresponding noise reduction measures are taken in combination with the energy consumption database, including:

[0080] B1: Use vibration sensors and acoustic sensors to monitor vibrational signals and sound pressure signals in real time, and convert the time-domain signals into frequency-domain signals through Fourier transform;

[0081] Furthermore, the specific steps of B1 include:

[0082] (1) Select vibration sensors and acoustic sensors according to the monitoring requirements, and install the sensors on the device or environment to be monitored, ensuring that the sensors are firmly fixed and the installation directions are correct;

[0083] (2) Use a data acquisition system to connect to the sensors wirelessly and collect vibrational signals and sound pressure signals in real time, and ensure that the sampling frequency of the acquisition system is high enough to accurately capture the high-frequency components in the signals;

[0084] (3) Perform preprocessing operations such as filtering and denoising on the collected vibrational signals and sound pressure signals to generate preprocessed time-domain signals to improve the signal quality;

[0085] (4) Use Perform a Fourier transform on the preprocessed time-domain signal to convert the time-domain signal into a frequency-domain signal. Here, X(k) represents the k-th sample of the frequency-domain signal, x(n) represents the n-th sample of the time-domain signal, k represents the sample index value in the frequency-domain signal, N represents the total number of samples of the time-domain signal, n represents the sample index value in the time-domain signal, and j represents the imaginary unit.

[0086] B2: Based on the frequency-domain signal, analyze the frequency components of the noise, and combine the data in the energy consumption database to identify the noise sources of different frequencies and their relationships with energy consumption; the data in the energy consumption database are the associated data of energy loss and operating conditions;

[0087] B3: According to the identification results, determine the noise sources and take corresponding noise reduction measures. The noise reduction measures include optimizing the component structure to reduce mechanical vibration and improving the flow channel design to reduce fluid pulsation.

[0088] Further, the specific process of determining the noise sources in B3 includes:

[0089] (1) Combine the results of frequency-domain signal analysis and energy consumption database matching to identify possible noise sources; the noise sources are generally divided into mechanical, equipment, traffic, and human voice types;

[0090] (2) Use the near-field acoustic intensity method to determine the specific positions of the noise sources;

[0091] 1) Prepare the equipment required for near-field acoustic intensity measurement, such as acoustic intensity probes and data acquisition systems;

[0092] 2) Identify the noise sources to be tested and their surrounding environments, and set the test area to ensure that there is no external interference during the test;

[0093] 3) Arrange a test grid in the test area. Here, the density and size of the grid are determined according to the characteristics of the test object and the requirements of test accuracy to ensure that the test grid covers the possible position range of the noise source;

[0094] 4) Place the acoustic intensity probe at each measurement point of the test grid, and use I = G×v to measure the acoustic intensity, and record the acoustic intensity value and the corresponding coordinates of each measurement point. Here, I represents the acoustic intensity, G represents the sound pressure, and v represents the particle vibration velocity;

[0095] It should be noted that in actual measurement, the acoustic intensity probe contains two microphones inside, which respectively measure the phase difference between the sound pressure and the particle vibration velocity, so as to directly calculate the acoustic intensity.

[0096] 5) Organize and analyze the measurement results, and draw an acoustic intensity distribution map;

[0097] 6) According to the acoustic intensity distribution map, identify the specific positions of the noise sources.

[0098] It should be understood that sound intensity is the time average of the rate of sound energy flow per unit area, representing the flow of sound energy; and the sound intensity distribution map is a graph obtained by measuring and plotting the sound intensity values at various points in a certain area. In the sound intensity distribution map, different colors or grayscales usually represent different sound intensity levels; by observing the sound intensity distribution map, find the areas with higher sound intensity values in the sound intensity distribution map. These areas indicate that the sound energy is relatively concentrated and are the locations of the noise sources.

[0099] (3) Verify and confirm the identified noise sources through data analysis methods to ensure that the identified noise sources are consistent with the actual situation and prepare to take corresponding noise reduction measures.

[0100] Based on the frequency domain signal, analyze the frequency components of the noise, and combine the data in the energy consumption database to identify the noise sources of different frequencies and their relationships with energy consumption, including:

[0101] B2.1: Real-time monitor the sound pressure signal of the digital device through an acoustic sensor, and convert the sound pressure signal into a frequency domain signal through Fourier transform;

[0102] B2.2: Perform spectrum analysis on the frequency domain signal to identify the main frequency components of the noise and evaluate the noise levels of each frequency component. Among them, spectrum analysis is the prior art content in this field and is not the creative solution of this application, so it will not be elaborated here;

[0103] B2.3: Extract the energy consumption data related to the current working condition from the energy consumption database and analyze the potential correlation between the energy consumption data and the noise frequency components;

[0104] B2.4: According to the known noise frequency components and energy consumption data, combined with the working principle and structural characteristics of the digital device, identify possible noise sources; the noise sources include mechanical vibration, fluid pulsation, and electromagnetic interference;

[0105] Furthermore, identifying possible noise sources is to identify preliminary noise sources. The specific process includes:

[0106] (1) Obtain the noise frequency component data during the operation of the digital device and the energy consumption data of the digital device;

[0107] (2) Analyze the structural characteristics of the digital device and pay attention to possible vibration sources and noise generation points, such as motor bearings, gear transmissions, and fluid channels;

[0108] (3) Match the collected noise frequency components with the structural characteristics of the digital device. Exemplarily, high-frequency noise may come from motor bearings or high-speed rotating components, while low-frequency noise may be related to fluid flow or mechanical vibration;

[0109] (4) Consider the relationship between noise at different frequencies and energy consumption. Among them, high energy consumption may be accompanied by high noise because, under high energy consumption conditions, the load on the motor or mechanical components increases, resulting in increased vibration and noise.

[0110] (5) Based on the analysis results in (3) and (4), identify the possible noise sources.

[0111] B2.5: For each identified noise source, analyze the relationship between noise sources at different frequencies and energy consumption.

[0112] Furthermore, the specific steps of B2.5 include:

[0113] (1) Mark and classify each identified noise source.

[0114] (2) Use acoustic measurement instruments to measure the frequency characteristics of each noise source, including low-frequency, medium-frequency, and high-frequency components, and record the sound pressure levels at different frequencies where G1 represents the actual sound pressure and G0 represents the reference sound pressure.

[0115] (3) For equipment or systems directly related to the noise source, such as mechanical equipment, fluid systems, and transportation vehicles, measure their energy consumption and record the energy consumption data.

[0116] (4) Compare and analyze the frequency characteristics of the noise source with the energy consumption data to find the correlation between noise sources at different frequencies and energy consumption. Among them, the correlation calculation is the prior art content in this field and is not the creative solution of this application, so it will not be elaborated here.

[0117] (5) Based on the analysis results, draw a conclusion to illustrate the relationship between noise sources at different frequencies and energy consumption.

[0118] The extraction of energy consumption data related to the current working condition from the energy consumption database and the analysis of the potential association between the energy consumption data and the noise frequency components include:

[0119] B2.31: Obtain the current operating condition of the digital device.

[0120] B2.32: According to the operating condition, query and extract the corresponding energy consumption data from the energy consumption database and preprocess the extracted energy consumption data; the energy consumption data includes power consumption, mechanical energy loss, and heat energy loss.

[0121] B2.33: Obtain the noise frequency components of the digital device and their corresponding noise levels, and use the correlation analysis method to analyze the potential correlation between the energy consumption data and the noise frequency components; the potential correlation refers to whether there is a correlation between the energy consumption parameters and the frequency noise, and the correlation includes positive correlation and negative correlation. Among them, the correlation analysis method is the prior art content in this field and is not the creative solution of this application, so it will not be elaborated here;

[0122] B2.34: According to the analysis results, explain the physical meaning of the correlation between the energy consumption data and the noise frequency components, and provide optimization suggestions based on the analysis results.

[0123] The construction process of the fault prediction model includes:

[0124] C1: Collect the historical data during the operation of the digital device and perform labeling processing on the historical data. The historical data includes energy consumption data, noise data, temperature data, and pressure data;

[0125] C2: Load the pre-constructed neural network model, and use the labeled historical data to train the pre-constructed neural network model to generate a trained fault prediction model. Among them, the neural network model is the prior art content in this field and is not the creative solution of this application, so it will not be elaborated here.

[0126] According to the fault prediction model, predict whether there are potential faults in the digital device and the types of faults, and generate a digital device adjustment plan according to the prediction results, including:

[0127] D1: Real-time collect the operation data of the digital device and perform preprocessing; the operation data of the digital device includes flow, sound, vibration, temperature, current, tripping of electrical protection components, and controller alarms;

[0128] D2: Use the signal processing method to extract features from the preprocessed operation data to generate the operation feature data of the digital device;

[0129] Furthermore, the specific steps of D2 include:

[0130] (1) Obtain the preprocessed operation data;

[0131] (2) Extract features from the preprocessed operation data in the time domain, specifically referring to calculating the mean value of the preprocessed operation data and analyzing the waveform features of the signal, such as waveform symmetry and periodicity;

[0132] (3) Perform Fourier transform on the preprocessed operation data to convert the time-domain signal into a frequency-domain signal, extract features in the frequency domain, and analyze the spectral features at the same time;

[0133] (4) Integrate the features extracted in (2) and (3) to form operating feature data;

[0134] (5) Use the principal component analysis method to optimize the operating feature data and remove redundant or irrelevant features. The principal component analysis method is the prior art content in this field and not the creative solution of this application, so it will not be elaborated here;

[0135] (6) Use the optimized operating feature data as the digital device operating feature data and label the digital device operating feature data.

[0136] D3: Input the digital device operating feature data into the trained fault prediction model. The fault prediction model predicts whether there are potential faults and the fault types of the digital device according to the matching degree between the digital device operating feature data and the historical fault data features;

[0137] D4: After the fault prediction model predicts that a fault may occur, it sends a warning message. The operator immediately responds and generates a digital device adjustment plan through threshold judgment.

[0138] The generation of the digital device adjustment plan through threshold judgment includes:

[0139] D4.1: Judge whether the energy loss exceeds the preset threshold. If it exceeds the preset threshold, automatically adjust the operating parameters of the digital device;

[0140] D4.2: Judge whether the noise exceeds the preset noise allowable range. If it exceeds the preset noise allowable range, automatically start the noise reduction device. The noise reduction device includes shock pads and mufflers. At the same time, according to the noise frequency analysis result, provide a noise control plan;

[0141] D4.3: Judge the fault type according to the output result of the fault prediction model. According to the judgment result of the fault type, automatically adjust the operating mode of the digital device, including reducing the operating speed and switching to standby components.

[0142] Embodiment 2

[0143] Please refer to Figure 4 , another embodiment provided by the present invention: A digital device fault prediction system includes:

[0144] An energy consumption analysis module, a noise analysis module, a fault prediction module, a noise reduction control module, and a fault processing module;

[0145] The energy consumption analysis module is used to collect the input power, output flow, and pressure of the digital device in real time, calculate the energy loss under different working conditions, and establish an energy consumption database;

[0146] A noise analysis module, which is used to monitor vibration signals and sound pressure signals in real time, analyze the frequency components of the noise, identify the noise sources and their relationship with energy consumption;

[0147] A fault prediction module, which trains a neural network model based on the collected historical data, and predicts the potential faults and fault types of digital devices in real time. According to the prediction results, it automatically adjusts the operating parameters;

[0148] A noise reduction control module, which automatically starts a noise reduction device based on the noise reduction analysis results and provides a noise control solution;

[0149] A fault handling module, which is used to automatically adjust the operating mode of digital devices according to the fault type after the fault prediction model issues a warning, and record the fault handling process and results.

[0150] The energy consumption analysis module includes: a power acquisition unit, a measurement unit, an energy consumption calculation unit, and a database storage unit;

[0151] The power acquisition unit, which collects the input power of digital devices in real time through power sensors, and provides the basic energy consumption data during the operation of digital devices;

[0152] The measurement unit, which measures the output flow rate and pressure through flow sensors and pressure sensors;

[0153] The energy consumption calculation unit, which calculates the output power according to the input power, output flow rate and pressure, and calculates the energy loss.

[0154] The database storage unit, which is used to establish an energy consumption database and store the energy consumption data under different working conditions.

[0155] The noise analysis module includes: a signal monitoring unit, a frequency domain analysis unit, and a correlation analysis unit;

[0156] The signal monitoring unit, which uses vibration sensors and acoustic sensors to monitor vibration signals and sound pressure signals in real time, and obtains the original data of the noise signals;

[0157] The frequency domain analysis unit, which converts the time domain signal into a frequency domain signal through Fourier transform, analyzes the frequency components of the noise, and identifies the noise sources with different frequencies;

[0158] The correlation analysis unit, which combines the data in the energy consumption database to identify the noise sources and their relationship with energy consumption.

[0159] The fault prediction module includes: a historical data collection unit, a model training unit, a prediction unit, and a warning unit;

[0160] The historical data collection unit, which is used to collect the historical data during the operation of digital devices, including energy consumption data, noise data, temperature data, and pressure data, and provides data support for the training of the neural network model;

[0161] A model training unit that trains a pre - constructed neural network model using the collected historical data to generate a trained fault prediction model;

[0162] A prediction unit that is used to collect the current operating data of the digital device in real - time and input it into the trained fault prediction model to predict whether there are potential faults and the types of faults in the digital device;

[0163] An early warning unit that, when a possible fault is predicted, issues an early warning message in a timely manner and automatically adjusts the operating parameters of the digital device according to the energy consumption analysis results to improve the reliability and operating efficiency of the device.

[0164] The fault handling module includes: a fault response unit, an operating mode adjustment unit, and a fault recording unit;

[0165] A fault response unit that, after the fault prediction model issues an early warning, the operator immediately responds;

[0166] An operating mode adjustment unit that automatically adjusts the operating mode of the digital device according to the type of fault, including reducing the operating speed and switching to standby components;

[0167] A fault recording unit that is used to automatically record the fault handling process and the handling results.

[0168] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above - mentioned specific embodiments. The above - mentioned specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can also make changes, modifications, substitutions, and variations to the above - mentioned embodiments without departing from the purpose and scope of the present invention. These all fall within the protection scope of the present invention.

[0169] If the technical solution of the present disclosure involves personal information, before the product applying the technical solution of the present disclosure processes personal information, it has clearly informed the personal information processing rules and obtained the autonomous consent of the individual. If the technical solution of the present disclosure involves sensitive personal information, before the product applying the technical solution of the present disclosure processes sensitive personal information, it has obtained the individual's separate consent and at the same time meets the requirements of "express consent". For example, at a personal information collection device such as a camera, a clear and prominent sign is set to inform that the personal information collection range has been entered and personal information will be collected. If an individual voluntarily enters the collection range, it is regarded as consenting to the collection of their personal information; or on the personal information processing device, when the personal information processing rules are informed by obvious signs / information, personal authorization is obtained through pop - up messages or by asking the individual to upload their personal information by themselves, etc. Among them, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the types of personal information processed.

Claims

1. A method for predicting faults of digital equipment, characterized in that: include: The input power, output flow and pressure of digital equipment are obtained through sensors, and the energy loss under different working conditions is calculated. Based on the calculation results, an energy consumption database is established; Real-time monitoring of vibration signals and sound pressure signals of digital equipment, analysis of the frequency components of noise, and adoption of corresponding noise reduction measures in combination with the energy consumption database; Collect historical data during the operation of digital equipment and build a fault prediction model, wherein the historical data includes energy consumption data, noise data, temperature data and pressure data; According to the fault prediction model, it is predicted whether the digital device has potential faults and the type of faults, and based on the prediction results, a digital device adjustment plan is generated.

2. A digital device fault prediction method as claimed in claim 1, characterized in that: The calculation of energy loss under different working conditions includes: Use power sensors to collect input power of digital devices in real time, and use flow sensors and pressure sensors to measure output flow and pressure; According to the output flow and pressure, calculate the output power P out , the formula is: Where ρ represents the liquid density, Q represents the output flow of the digital device, g represents the acceleration of gravity, H represents the head, η f represents the fluid efficiency coefficient, η l represents the energy loss coefficient, T represents the current temperature, T0 represents the reference temperature, α represents the temperature influence coefficient, C represents the current concentration, C0 represents the reference concentration, β represents the concentration influence coefficient, δ represents the time attenuation coefficient, t represents the operating time, and e represents the exponent; According to the output power P out and the input power P in , the energy loss under different working conditions is obtained as loss = P in -P out .

3. A digital device fault prediction method as claimed in claim 2, characterized in that: The real-time monitoring of the vibration signal and sound pressure signal of the digital device, analyzing the frequency components of the noise, combining with the energy consumption database, and taking corresponding noise reduction measures include: Use vibration sensors and acoustic sensors to monitor vibration signals and sound pressure signals in real time, and convert time domain signals into frequency domain signals through Fourier transform; Based on the frequency domain signal, the frequency components of the noise are analyzed, and combined with the data in the energy consumption database, the noise sources of different frequencies and their relationship with energy consumption are identified; the data in the energy consumption database are the correlation data between energy loss and working conditions; According to the identification results, the noise source is determined and corresponding noise reduction measures are taken, which include optimizing the component structure to reduce mechanical vibration and improving the flow channel design to reduce fluid pulsation.

4. A digital device fault prediction method as claimed in claim 3, characterized in that: The construction process of the fault prediction model includes: Collect historical data during the operation of digital devices and label the historical data, the historical data including energy consumption data, noise data, temperature data, and pressure data; Load the pre-built neural network model, use the labeled historical data to train the pre-built neural network model, and generate a trained fault prediction model.

5. A digital device fault prediction method as claimed in claim 4, characterized in that: The method of predicting whether a digital device has a potential fault and the type of the fault according to the fault prediction model, and generating a digital device adjustment plan according to the prediction result, includes: Collect the operation data of digital equipment in real time and perform preprocessing; the operation data of digital equipment includes flow, sound, vibration, temperature, current, tripping of electrical protection components, and controller alarm; Use signal processing methods to extract features from the preprocessed operation data to generate digital device operation feature data; The digital device operation characteristic data is input into a trained fault prediction model, and the fault prediction model predicts whether the digital device has a potential fault and the type of fault according to the matching degree between the digital device operation characteristic data and the historical fault data characteristics; The fault prediction model predicts the occurrence of a fault and issues a warning message, and the operator responds immediately and generates a digital equipment adjustment plan through threshold judgment.

6. A digital device fault prediction method as claimed in claim 5, characterized in that: The step of generating a digital device adjustment plan by threshold determination includes: determining whether the energy loss exceeds a preset threshold, and if so, automatically adjusting the operating parameters of the digital device; Determine whether the noise exceeds the preset noise tolerance range. If it exceeds the preset noise tolerance range, the noise reduction device is automatically activated, and the noise reduction device includes a shock-absorbing pad and a muffler. At the same time, according to the noise frequency analysis results, a noise control plan is provided; The fault type is determined according to the output result of the fault prediction model, and the operation mode of the digital device is automatically adjusted according to the determination result of the fault type, including reducing the operation speed and switching spare parts.

7. A digital device fault prediction method as claimed in claim 6, characterized in that: The method of analyzing the frequency components of the noise based on the frequency domain signal and identifying the noise sources of different frequencies and their relationship with the energy consumption in combination with the data in the energy consumption database includes: The acoustic pressure signal of the digital device is monitored in real time by an acoustic sensor, and the acoustic pressure signal is converted into a frequency domain signal by Fourier transform; Performing spectrum analysis on the frequency domain signal, identifying the main frequency components of the noise, and evaluating the noise level of each frequency component; Extract energy consumption data related to the current working conditions from the energy consumption database and analyze the potential correlation between the energy consumption data and the noise frequency components; According to the known noise frequency components and energy consumption data, combined with the working principle and structural characteristics of the digital device, the noise source is identified; the noise source includes mechanical vibration, fluid pulsation, and electromagnetic interference; For each identified noise source, the relationship between the noise source at different frequencies and the energy consumption is analyzed.

8. A digital device fault prediction method as claimed in claim 7, characterized in that: The step of extracting energy consumption data related to the current working condition from the energy consumption database and analyzing the potential correlation between the energy consumption data and the noise frequency components includes: Get the current operating status of the digital device; According to the operating conditions, query and extract corresponding energy consumption data from the energy consumption database, and pre-process the extracted energy consumption data; the energy consumption data includes electrical energy consumption, mechanical energy loss, and thermal energy loss; Obtain the noise frequency component of the digital device and its corresponding noise level, and use a correlation analysis method to analyze the potential correlation between the energy consumption data and the noise frequency component; the potential correlation refers to whether there is a correlation between the energy consumption parameter and the frequency noise, and the correlation includes positive correlation and negative correlation; Based on the analysis results, the physical meaning of the correlation between energy consumption data and noise frequency components is explained, and optimization suggestions are provided based on the analysis results.

9. A digital device fault prediction system, used to implement a digital device fault prediction method according to any one of claims 1 to 8, characterized in that: include: Energy consumption analysis module, noise analysis module, fault prediction module, noise reduction control module, fault handling module; The energy consumption analysis module is used to collect the input power, output flow and pressure of the digital device in real time, calculate the energy loss under different working conditions, and establish an energy consumption database; The noise analysis module is used to monitor the vibration signal and the sound pressure signal in real time, analyze the frequency components of the noise, identify the noise source and its relationship with energy consumption; The fault prediction module trains a neural network model based on the collected historical data, and predicts the potential faults and fault types of digital devices in real time, and automatically adjusts the operating parameters based on the prediction results; The noise reduction control module automatically starts the noise reduction device based on the noise reduction analysis results and provides a noise control solution; The fault processing module is used to automatically adjust the operation mode of the digital device according to the fault type after the fault prediction model issues an early warning, and record the fault processing process and processing results.

Images