Method for controlling reactor vibration isolation device, and electronic apparatus

Abstract

The invention relates to the technical field of reactor detection, and particularly provides a reactor vibration isolation device control method and electronic equipment, and the method comprises the steps: collecting the working frequency of a reactor in real time, determining the existence of non-dominant frequency resonance, collecting a vibration signal under a corresponding non-preset working frequency, and extracting vibration characteristic parameters; under the condition that it is judged that the reactor body works normally, based on a multi-body dynamic model, the vibration characteristic parameters serve as input excitation, simulation calculation is conducted, and simulation response data after vibration isolation is predicted; the predicted simulation response data after vibration isolation are compared with actually measured signals after vibration isolation, and the isolation effect of the vibration isolation device is evaluated; and the vibration isolation device is controlled according to the isolation effect, so that non-dominant frequency resonance is suppressed, and the problem that an existing vibration isolation device cannot meet the requirement for effective vibration isolation due to the fact that an electric reactor of a power grid in the related technology is prone to generating non-dominant frequency resonance with wide frequency distribution and various vibration modes is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric reactor detection, and in particular to a control method of a vibration isolation device of an electric reactor and an electronic device. BACKGROUND

[0002] When an electric reactor in a power grid operates at a non-pre-set working frequency, due to the structural characteristics of the electric reactor itself and external conditions, the electric reactor or connected equipment is prone to be damaged. In actual operation, the working environment of the electric reactor is constantly changing, and sudden changes in load and fluctuations in power grid frequency can all cause the electric reactor to produce non-primary frequency resonance. These non-primary frequency resonances have extensive frequency distribution and various vibration modes, which brings great challenges to the design of the vibration isolation device.

[0003] The existing vibration isolation device has a simple structure and is ultimately an overall system as a single-degree-of-freedom spring-mass system, and its natural frequency is generally single, usually designed near the primary frequency, which results in that it can only play a good vibration isolation effect near one frequency, and the vibration isolation effect is often greatly discounted for those atypical resonance frequencies far away from the primary frequency.

[0004] In view of the problem that the electric reactor of the power grid is prone to produce non-primary frequency resonance with extensive frequency distribution and various vibration modes, resulting in that the existing vibration isolation device cannot meet the effective vibration isolation requirement, no effective solution has been proposed at present. SUMMARY

[0005] The control method of the vibration isolation device of the electric reactor and the electronic device provided by the embodiments of the present application at least solve the problem that the electric reactor of the power grid is prone to produce non-primary frequency resonance with extensive frequency distribution and various vibration modes, resulting in that the existing vibration isolation device cannot meet the effective vibration isolation requirement.

[0006] According to an aspect of an embodiment of the present application, a control method of a vibration isolation device of a reactor is provided, comprising: collecting a working frequency of the reactor in real time, determining whether a resonance frequency corresponding to the working frequency is a non-primary frequency resonance; for the non-primary frequency resonance, collecting a vibration signal at a corresponding non-pre-set working frequency, extracting a vibration characteristic parameter of the vibration signal to determine whether the reactor body is working normally; in the case that the reactor body is working normally, based on a multi-body dynamics model, taking the vibration characteristic parameter as an input excitation to perform simulation calculation and predict simulation response data after isolation; wherein the multi-body dynamics model is established according to a calibrated reactor model and current design parameters of the vibration isolation device; the calibrated reactor model is obtained by comparing and calibrating a vibration signal under different working states according to a theoretical model of the reactor; comparing the predicted simulation response data after isolation with the actually measured signal after isolation to evaluate the isolation effect of the vibration isolation device; controlling the vibration isolation device according to the isolation effect to suppress the non-primary frequency resonance.

[0007] As an optional solution, controlling the vibration isolation device according to the isolation effect comprises: determining a transfer rate-frequency curve of the vibration isolation device as a performance boundary based on the multi-body dynamics model according to the isolation effect of the vibration isolation device; comparing the performance boundary with a pre-set isolation threshold to determine whether each frequency point meets the standard; in the case that the isolation effect of the frequency point is lower than a pre-set effect threshold, modifying the isolation parameters of the multi-body dynamics model to find an optimal parameter combination through an optimization algorithm; and controlling the vibration isolation device according to the optimal parameter combination.

[0008] As an optional solution, comparing the performance boundary with the pre-set isolation threshold to determine whether each frequency point meets the standard comprises: determining a transfer rate value corresponding to each frequency point based on the transfer rate-frequency curve of the performance boundary; calculating a performance index of each frequency point according to the transfer rate value and in combination with an established evaluation index system, wherein the performance index comprises at least one of the following: isolation efficiency, vibration attenuation rate, frequency characteristic and energy transfer ratio; comparing the performance index of each frequency point with a corresponding pre-set isolation threshold, in the case that the performance index is better than the corresponding pre-set isolation threshold, the corresponding frequency point meets the standard, and in the case that the performance index is worse than the corresponding pre-set isolation threshold, the corresponding frequency point does not meet the standard.

[0009] As an optional approach, when the vibration isolation effect at a frequency point is lower than a preset threshold, the vibration isolation parameters of the multibody dynamics model are modified, and the optimal parameter combination is found through an optimization algorithm. This includes: obtaining response data at the weak performance frequency point where the vibration isolation effect is lower than the preset threshold, the response data including three parameters: amplitude, phase, and transmissivity; calculating the sensitivity coefficient of the vibration isolation effect relative to each vibration isolation design parameter based on the multibody dynamics model; determining the vibration isolation design parameter that has the greatest impact on the vibration isolation effect based on the sensitivity coefficient; and using a parameter optimization algorithm to search for the optimal parameter combination of the vibration isolation design parameters.

[0010] As an optional approach, controlling the vibration isolation device according to the optimal parameter combination includes: real-time acquisition of changes in the grid frequency and load parameters of the power grid where the reactor is located, and real-time monitoring of the frequency fluctuations of the power grid; calculating the optimal control parameters of the controllable parameters of the vibration isolation device using an adaptive control algorithm based on the optimal parameter combination, changes in grid frequency and load parameters; and adjusting the controllable parameters of the vibration isolation device to the corresponding optimal control parameters when the frequency fluctuations of the power grid where the reactor is located exceed a preset threshold, so as to control the vibration isolation device.

[0011] As an optional approach, the reactor's operating frequency is acquired in real time to determine whether the corresponding resonant frequency is a non-dominant frequency resonance. This includes: acquiring the reactor's operating current and voltage signals in real time; preprocessing the acquired operating current and voltage signals, wherein the preprocessing includes at least one of the following: cleaning, noise reduction, and outlier detection; performing frequency domain analysis on the preprocessed operating current and voltage signals using Fast Fourier Transform to obtain the amplitude and phase information of each resonant frequency, and identifying the real-time operating frequency and each resonant frequency; and determining whether each resonant frequency is a non-dominant frequency resonance based on the amplitude and phase information.

[0012] As an optional approach, for non-dominant frequency resonance, vibration signals are collected at the corresponding non-preset operating frequency, and vibration characteristic parameters of the vibration signals are extracted to determine whether the reactor body is operating normally. This includes: in the presence of non-dominant frequency resonance, collecting vibration signals of the reactor at a non-preset operating frequency, wherein the non-preset operating frequency is the operating frequency corresponding to the non-dominant frequency resonance; after denoising and filtering preprocessing of the vibration signals, analyzing the preprocessed vibration signals using time-frequency analysis methods to extract vibration characteristic parameters of the vibration signals, wherein the vibration characteristic parameters include at least one of the following: frequency, amplitude, and energy distribution; and determining whether the reactor body is operating normally by inputting the vibration characteristic parameters into a trained machine learning model, wherein the machine learning model is trained based on multiple sets of training data, each set of training data including vibration characteristic parameters and whether the corresponding reactor body is operating abnormally.

[0013] As an optional approach, for non-dominant frequency resonance, under the condition that the reactor body is working normally, before performing simulation calculations based on a multibody dynamics model and using the vibration characteristic parameters as input excitation to predict the simulation response data after vibration isolation, the method further includes: collecting vibration signals of the reactor at different operating frequencies; creating a theoretical model of the reactor and calculating the theoretical natural frequencies and theoretical mode shapes of the reactor at different operating frequencies; comparing the theoretical natural frequencies and theoretical mode shapes with the measured frequencies and mode shapes of the vibration signals, iteratively optimizing the model parameters of the theoretical model to calibrate the reactor model; and establishing a multibody dynamics model of the vibration isolation device based on the calibrated reactor model and the current design parameters of the vibration isolation device.

[0014] As an optional solution, the method further includes: real-time acquisition of the reactor's operating frequency; if the resonant frequency corresponding to the operating frequency is determined to be the main frequency resonance, then generating a fault alarm signal for the reactor and sending it to the relevant terminal; using a rule-based expert system to comprehensively analyze the reactor's operating parameters and the vibration spectrum features extracted by the time-frequency analysis module, matching them with fault rules in the knowledge base to determine the fault type of the reactor; and generating corresponding maintenance decision suggestions based on the fault type, and sending the decision suggestions to the relevant terminal.

[0015] According to another aspect of the present invention, an electronic device is provided, including: a processor, and a memory storing a program, the program including instructions that, when executed by the processor, cause the processor to perform the methods described above.

[0016] According to another aspect of the present invention, a non-transitory machine-readable medium storing computer instructions for causing the computer to perform the above-described method is provided.

[0017] The control method for the vibration isolation device of the reactor provided in this invention collects the operating frequency of the reactor in real time to determine whether the resonant frequency corresponding to the operating frequency is a non-dominant frequency resonance; for non-dominant frequency resonance, vibration signals at the corresponding non-preset operating frequency are collected, and vibration characteristic parameters of the vibration signals are extracted to determine whether the reactor body is working normally; effectively eliminating the influence of abnormal operation of the reactor body on the evaluation of the vibration isolation effect of the vibration isolation device, and improving the accuracy of vibration isolation effect detection of the vibration isolation device.

[0018] Under normal operating conditions of the reactor, based on a multibody dynamics model, vibration characteristic parameters are used as input excitation to perform simulation calculations and predict the simulated response data after vibration isolation. The predicted simulated response data after vibration isolation is then compared with the measured signal after vibration isolation to evaluate the isolation effect of the vibration isolation device. This allows for real-time and accurate evaluation of the isolation effect of the vibration isolation device.

[0019] Moreover, the multibody dynamics model is established based on the calibrated reactor model and the current design parameters of the vibration isolation device. The calibrated reactor model is obtained by comparing and calibrating the reactor's theoretical model with vibration signals under different operating conditions, which can further improve the accuracy of the predicted simulation response data, thereby improving the accuracy of the vibration isolation effect evaluation of the vibration isolation device.

[0020] The vibration isolation device is controlled based on its isolation effect to suppress non-dominant frequency resonances. By accurately assessing the real-time vibration isolation effect, the device can be controlled promptly to suppress non-dominant frequency resonances. This solves the problem in related technologies where power grid reactors easily generate non-dominant frequency resonances with wide frequency distributions and diverse vibration modes, causing existing vibration isolation devices to fail to meet effective vibration isolation requirements. This approach achieves accurate assessment of the vibration isolation effect and effectively controls and adjusts the vibration isolation device based on the effect, thus improving the technical effectiveness of the vibration isolation device in suppressing non-dominant frequency resonances. Attached Figure Description

[0021] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention, and those skilled in the art can obtain other embodiments based on these drawings without creative effort.

[0022] Figure 1This is a flowchart of a control method for a vibration isolation device of a reactor, according to an embodiment of the present invention.

[0023] Figure 2 This is a schematic diagram of the control device of a reactor vibration isolation device according to an embodiment of the present invention.

[0024] Figure 3 This is a schematic diagram of the structure of the electronic device created by this invention. Detailed Implementation

[0025] Embodiments of the present invention will now be described in more detail with reference to the accompanying drawings. While some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0026] With the increasing number of reactors in the power grid, the vibration and noise problems generated during their operation are becoming increasingly serious, negatively impacting the lives of surrounding residents and the health of workers. Traditional vibration isolation measures, such as rubber pads, have high natural frequencies and small participating masses, making it difficult to meet actual vibration reduction and isolation requirements. Moreover, reactors play a crucial role in the power system, yet the vibration problems they generate during operation have been troubling engineers.

[0027] Especially when the reactor operates at non-typical operating frequencies, the reactor or the connected equipment is easily damaged due to the structural characteristics of the reactor itself and the influence of external conditions. If a capacitor bank without series connection of the reactor is used, resonance problems may occur, high voltage distortion will be generated on the bus with the capacitor bank, and a series of problems will be caused to damage to the user equipment.

[0028] Vibration isolation devices in related technologies are generally used as single-degree-of-freedom spring-mass systems, with relatively simple natural frequencies. This makes them prone to non-dominant frequency resonances with wide frequency distributions and diverse vibration modes, resulting in existing vibration isolation devices failing to meet the requirements for effective vibration isolation.

[0029] Figure 1 This is a flowchart of a control method for a vibration isolation device of a reactor according to an embodiment of the present invention, as shown below. Figure 1 As shown, an embodiment of the present invention provides a control method for a vibration isolation device of a reactor, comprising:

[0030] Step S101: Real-time acquisition of the reactor's operating frequency to determine whether the resonant frequency corresponding to the operating frequency is a non-dominant frequency resonance.

[0031] Step S102: For non-master frequency resonance, collect vibration signals at the corresponding non-preset operating frequency, extract vibration characteristic parameters of the vibration signals, and determine whether the reactor body is working properly.

[0032] Step S103: Under normal working conditions of the reactor body, based on the multibody dynamics model, the vibration characteristic parameters are used as input excitation to perform simulation calculations and predict the simulation response data after vibration isolation. The multibody dynamics model is established based on the calibrated reactor model and the current design parameters of the vibration isolation device. The calibrated reactor model is obtained by comparing and calibrating the vibration signals under different working conditions based on the theoretical model of the reactor.

[0033] Step S104: Compare the predicted simulation response data after vibration isolation with the measured signal after vibration isolation to evaluate the isolation effect of the vibration isolation device.

[0034] Step S105: Control the vibration isolation device according to the isolation effect to suppress non-dominant frequency resonance.

[0035] The embodiments of this invention provide that by real-time acquisition of the reactor's operating frequency, the resonant frequency corresponding to the operating frequency is determined to be a non-dominant frequency resonance; for non-dominant frequency resonance, vibration signals at the corresponding non-preset operating frequency are acquired, and vibration characteristic parameters of the vibration signals are extracted to determine whether the reactor body is working normally; effectively eliminating the influence of abnormal reactor body operation on the evaluation of vibration isolation effect of vibration isolation device, and improving the accuracy of vibration isolation effect detection of vibration isolation device.

[0036] Under normal operating conditions of the reactor, based on a multibody dynamics model, vibration characteristic parameters are used as input excitation to perform simulation calculations and predict the simulated response data after vibration isolation. The predicted simulated response data after vibration isolation is then compared with the measured signal after vibration isolation to evaluate the isolation effect of the vibration isolation device. This allows for real-time and accurate evaluation of the isolation effect of the vibration isolation device.

[0037] Moreover, the multibody dynamics model is established based on the calibrated reactor model and the current design parameters of the vibration isolation device. The calibrated reactor model is obtained by comparing and calibrating the reactor's theoretical model with vibration signals under different operating conditions, which can further improve the accuracy of the predicted simulation response data, thereby improving the accuracy of the vibration isolation effect evaluation of the vibration isolation device.

[0038] The vibration isolation device is controlled based on its isolation effect to suppress non-dominant frequency resonances. By accurately assessing the real-time vibration isolation effect, the device can be controlled promptly to suppress non-dominant frequency resonances. This achieves accurate assessment of the vibration isolation effect and allows for effective control and adjustment of the device based on that effect, thus improving the technical effectiveness of the vibration isolation device in suppressing non-dominant frequency resonances.

[0039] This solves the problem in the aforementioned related technologies that power grid reactors are prone to generating non-dominant frequency resonances with wide frequency distributions and diverse vibration modes, which makes existing vibration isolation devices unable to meet the requirements for effective vibration isolation.

[0040] The entity performing the above steps can be the control device of the reactor.

[0041] In step S101 above, the operating frequency of the reactor is acquired in real time to determine whether the resonant frequency corresponding to the operating frequency is a non-dominant frequency resonance. The reactor's operating frequency can be tracked in real time, and real-time operating frequency data under different operating conditions can be captured. The collected real-time operating frequency data will be used for frequency analysis to identify the resonant frequencies present during reactor operation and determine whether the resonant frequencies are non-dominant frequency resonances.

[0042] Specifically, the process includes the following steps: Real-time acquisition of the reactor's operating current and voltage signals. Preprocessing of the acquired operating current and voltage signals, including at least one of the following: cleaning, noise reduction, and outlier detection. Frequency domain analysis of the preprocessed operating current and voltage signals using Fast Fourier Transform to obtain the amplitude and phase information of each resonant frequency, and identification of the real-time operating frequency and each resonant frequency. Based on the amplitude and phase information, determining whether each resonant frequency is a non-dominant frequency resonance.

[0043] The real-time acquired operating frequency data, including the aforementioned operating current and voltage signals, can be transmitted to the data processing module for storage and analysis. The data processing module preprocesses the real-time operating frequency data to obtain preprocessed operating current and voltage signals. This preprocessing includes data cleaning, outlier detection, and noise reduction to ensure data accuracy and reliability.

[0044] Then, the Fast Fourier Transform (FFT) algorithm is used to perform frequency domain analysis on the preprocessed operating current and voltage signals to obtain the various frequency components present in the reactor operation, as well as their amplitude and phase information. The FFT algorithm is used to convert the time-domain data into frequency-domain data, and the amplitude and phase of each frequency component are calculated using the FFT algorithm. Based on the frequency resolution and data length, an appropriate number of FFT points and a window function, such as a Hamming window or a Blackman window, are selected. The FFT results are then normalized to obtain the relative amplitude and phase distribution of each frequency component.

[0045] Based on the amplitude and phase distribution of the frequency components, a matching identification is performed with the amplitude and phase distribution of the designed main frequency resonance. If the main frequency resonance is identified, it indicates a reactor malfunction. As an optional embodiment, the method further includes: real-time acquisition of the reactor's operating frequency; if the resonant frequency corresponding to the operating frequency is determined to be the main frequency resonance, a reactor body fault alarm signal is generated and sent to relevant terminals; a rule-based expert system comprehensively analyzes the reactor's operating parameters and vibration spectrum features extracted through the time-frequency analysis module, matching them with fault rules in the knowledge base to determine the reactor body fault type; based on the fault type, corresponding maintenance decision suggestions are generated and sent to relevant terminals.

[0046] In other words, after determining whether the resonant frequency corresponding to the operating frequency is a non-primary frequency resonance, if the identified resonant frequency is a primary frequency resonance or a non-preset operating frequency resonance with an amplitude exceeding 20% ​​of the rated operating frequency amplitude, a first-level fault alarm signal for the reactor body is generated.

[0047] After generating a fault alarm signal for the reactor body, fault diagnosis is initiated. The fault diagnosis module uses a rule-based expert system to comprehensively analyze the reactor's current, voltage, and temperature operating parameters, as well as the vibration spectrum characteristics extracted by the time-frequency analysis module, and matches them with fault rules in the knowledge base to determine the fault type of the reactor body. The fault type includes at least one of the following: loose winding, failure of iron core clamping force, or fracture of internal structural components. The vibration spectrum characteristics include: frequency in the range of 65Hz to 85Hz, vibration amplitude greater than 30% of the rated value, and duration exceeding 30 seconds.

[0048] After determining the fault type of the reactor body, corresponding maintenance decision suggestions are generated based on the fault type. For example, if the fault type is a loose iron core, the generated maintenance decision suggestion is to reduce the reactor's operating current to reduce the vibration amplitude, and to recommend that the reactor be disassembled, inspected, and tightened during the next power outage maintenance. If the fault type is a loose winding, the generated maintenance decision suggestion is to plan an immediate shutdown maintenance.

[0049] After generating a fault alarm signal for the reactor body, the parameter optimization operation of the vibration isolation device is suppressed, and a prompt message is output to clearly indicate that the root cause of the fault lies in the reactor body rather than the vibration isolation device, thus preventing misoperation.

[0050] The evaluation module for the operating parameters of the vibration state adopts the support vector machine (SVM) algorithm and uses the radial basis function (RBF) kernel function. The SVM model has 1000 training samples, including 600 normal state samples and 400 abnormal state samples. The optimal model parameters are determined by 5-fold cross-validation. The classification accuracy of the model on the test set is not less than 98%.

[0051] Frequency domain analysis is performed using the Fast Fourier Transform (FFT) algorithm, including: setting the frequency resolution to 0.1Hz and identifying the resonant frequencies of the reactor outside the rated operating frequency of 50Hz by analyzing the spectrum; when a resonant frequency with an amplitude exceeding 20% ​​of the rated operating frequency amplitude is identified in the range of 65Hz to 85Hz, it is determined to be a resonance caused by a non-preset operating frequency.

[0052] The preprocessed vibration signal was analyzed using a time-frequency analysis method, including: selecting the Hanning window as the time window with a window length of 1 second and an overlap rate of 50%; performing a 512-point FFT on the data within each time window to generate a time-frequency plot; and extracting the amplitude, frequency, and phase characteristic parameters of each frequency component from the time-frequency plot to form a feature vector for evaluating the vibration state.

[0053] Based on the identified non-dominant frequency resonance, it is compared with a preset threshold to determine whether the non-dominant frequency resonance exceeds the normal range. If the non-dominant frequency resonance exceeds the normal range, an alarm message is generated and sent to the monitoring system for processing.

[0054] If it is a non-dominant frequency resonance, the load changes of the reactor are monitored in real time, and the changes in load current and voltage are recorded. The fluctuation of the power grid frequency is analyzed, and the magnitude and duration of the frequency deviation are determined by comparing it with the standard frequency.

[0055] By correlating load changes and frequency fluctuations with the occurrence time of resonant frequencies, the causal relationship between these data and the resonant frequencies can be determined. Machine learning algorithms, such as decision trees or support vector machines, can be used to establish a mapping model between load changes, frequency fluctuations, and resonant frequencies, enabling automatic identification of the causes of resonant frequency occurrences.

[0056] Based on the characteristics and causes of the resonant frequency, a corresponding control strategy is generated, and control commands are issued to relevant equipment for execution. Specifically, based on the characteristics and causes of the resonant frequency, a corresponding control strategy is automatically generated, and a knowledge base is established between the resonant frequency characteristics and the control strategy. This knowledge base includes characteristics such as the amplitude, duration, and frequency range of the resonant frequency, as well as corresponding control measures, such as adjusting the inductance value of the reactor or changing the power grid operating parameters. Once the resonant frequency is identified, a matching control strategy is searched in the knowledge base based on its characteristics.

[0057] If no perfectly matching strategy is found in the knowledge base, the closest strategy can be identified through similarity calculation, and appropriate adjustments can be made based on the actual situation. The generated control strategy is then converted into specific control commands, such as adjusting the inductance setting of reactors or regulating the reactive power compensation device of the power grid, and sent to the corresponding execution equipment.

[0058] The operating frequency and resonance information of the reactor are monitored and fed back to the monitoring center in real time to form a closed-loop control system. This system enables dynamic optimization of the reactor's operating status and fault prevention.

[0059] As an optional implementation, the real-time operating frequency data of the reactor is acquired at a sampling rate of 100 times per second, and the data is transmitted to the data processing module via a high-speed data bus. The data processing module uses a data analysis library (such as Python's Pandas library) to clean the data, uses the 3σ principle to identify and remove outliers that exceed the normal range, and uses wavelet transform to denoise the data. The accuracy of the denoised data is improved by more than 95%.

[0060] Then, the FFT function in a scientific computing library (such as Python's NumPy library) was used to perform frequency domain analysis on the preprocessed data. A 1024-point FFT was selected, and a Hamming window function was used for windowing to obtain a spectrum with a frequency resolution of 0.1 Hz. By analyzing the spectrum, it was found that there was a non-dominant frequency resonance with an amplitude of 20% of the fundamental frequency and a frequency of 150 Hz during reactor operation.

[0061] To analyze the cause of the resonant frequency, the monitoring center recorded the load current and voltage data of the reactor over the past hour, and also obtained the fluctuation of the grid frequency during the same period. It was found that the grid frequency fluctuated by ±0.2Hz for 10 minutes.

[0062] By correlating load changes and frequency fluctuations with the occurrence time of the resonant frequency, it was found that the occurrence time of the resonant frequency coincided with the frequency fluctuation time. Therefore, it was determined that the frequency fluctuation was the main cause of the resonant frequency.

[0063] To further verify this conclusion, a mapping model between frequency fluctuation and resonant frequency was established using the decision tree algorithm. The mapping model was then trained using historical data, and a prediction model with an accuracy of 98% was finally obtained.

[0064] Based on the preset resonant frequency threshold, the system determines that the current resonant frequency exceeds the normal range, thus generating an alarm message and sending it to the monitoring center. Simultaneously, based on the characteristics of the resonant frequency, a corresponding control strategy is retrieved from the knowledge base: the inductance value of the reactor is adjusted to 80% of its original value, and the reactive power compensation device of the power grid is activated to suppress the resonance.

[0065] Step S102: For non-dominant frequency resonance, collect vibration signals at the corresponding non-preset operating frequency, extract vibration characteristic parameters of the vibration signals to determine whether the reactor body is working normally. Based on the frequency analysis results, obtain the resonant frequency caused by the non-preset operating frequency, start the data acquisition system to collect the reactor's vibration data at the non-preset operating frequency to determine whether the reactor is in an abnormal vibration state. An abnormal vibration state can characterize the abnormal operation of the reactor body.

[0066] As an optional embodiment, for non-dominant frequency resonance, vibration signals are collected at the corresponding non-preset operating frequency, and vibration characteristic parameters of the vibration signals are extracted to determine whether the reactor body is working normally. This includes: in the presence of non-dominant frequency resonance, collecting vibration signals of the reactor at a non-preset operating frequency, where the non-preset operating frequency is the operating frequency corresponding to the non-dominant frequency resonance; after denoising and filtering preprocessing of the vibration signals, analyzing the preprocessed vibration signals using time-frequency analysis methods to extract vibration characteristic parameters of the vibration signals, where the vibration characteristic parameters include at least one of the following: frequency, amplitude, and energy distribution; and determining whether the reactor body is working normally by inputting the vibration characteristic parameters into a trained machine learning model, where the machine learning model is trained based on multiple sets of training data, each set of training data including vibration characteristic parameters and whether the corresponding reactor body is malfunctioning.

[0067] Specifically, through frequency analysis, the resonant frequency generated by the reactor during operation is identified, and it is determined whether the resonant frequency is caused by a non-preset operating frequency. If the resonant frequency is caused by a non-preset operating frequency, the data acquisition system is triggered to enter the deep acquisition mode to collect the vibration data of the reactor at the non-preset operating frequency.

[0068] The collected vibration data undergoes denoising and filtering in the data preprocessing module to obtain processed vibration data. This processed vibration data is then stored in the database and simultaneously transmitted to the data analysis module for analysis. The collected vibration data is processed through denoising and filtering in the data preprocessing module to improve data quality. The processed vibration data is then stored in the database and simultaneously transmitted to the data analysis module for analysis.

[0069] Data analysis employs time-frequency analysis to perform time-frequency domain analysis on the processed vibration data, obtaining energy distribution data of the vibration signal at different times and frequencies, and extracting vibration characteristic parameters of the vibration signal. The time-frequency analysis method is used to perform time-frequency domain analysis on the vibration data to determine the target frequency range and time resolution of the analysis, and to select appropriate time window lengths and overlap rates.

[0070] The vibration signal is segmented, and a short-time Fourier transform is performed within each time window to obtain its spectrum. The spectrums of each time window are arranged chronologically to form a time-frequency diagram, representing the energy distribution of the vibration signal at different times and frequencies. Vibration characteristic parameters, such as the amplitude, frequency, and phase of each frequency component, are extracted from the time-frequency diagram for subsequent condition assessment and fault diagnosis.

[0071] Based on the vibration feature parameters of the extracted vibration signals, an evaluation model for the reactor's vibration state is constructed. Machine learning algorithms are then used to classify and evaluate the reactor's vibration state, determining whether it is in an abnormal vibration state. The selection of a suitable machine learning algorithm is based on the specific requirements of the reactor vibration state evaluation; for example, Support Vector Machines are suitable for binary classification problems, while Random Forests are suitable for multi-class classification problems. Sufficient reactor vibration data, including data from normal and various abnormal states, is collected as both training and testing datasets.

[0072] The training data is preprocessed to extract vibration feature parameters from the vibration signals, followed by feature selection and normalization. Cross-validation and other methods are used to train and tune the evaluation model, optimizing its classification performance. The model's generalization ability is then evaluated on the test dataset. The trained model is applied to online vibration state assessment, performing real-time classification and anomaly detection on newly acquired vibration data.

[0073] Based on the reactor's design parameters and operational experience, normal ranges and alarm thresholds for vibration amplitude, frequency, and other indicators are preset. The vibration state classification results output by the evaluation model are compared with the preset thresholds; if they exceed the normal range, they are judged as abnormal vibration states.

[0074] If the reactor is in an abnormal vibration state, a vibration abnormality alarm message is generated and sent to the monitoring center. Simultaneously, fault diagnosis is triggered to analyze and diagnose the factors contributing to the vibration abnormality. By comprehensively analyzing the reactor's operating parameters, vibration data, and historical operating data, the factors causing the abnormal vibration are inferred and diagnosed, resulting in a fault diagnosis report.

[0075] Considering the duration and trend of vibration signals, thresholds for the duration and rate of change of abnormal vibrations are set to further confirm the abnormal vibration state. By comprehensively analyzing the reactor's operating parameters, vibration data, and historical operating data, methods such as expert systems or fault trees are used to reason about and diagnose the causes of abnormal vibrations, determine possible fault types and locations, and generate a fault diagnosis report.

[0076] Based on the fault diagnosis report, corresponding control strategies and maintenance suggestions are generated. Control commands are then sent to the field equipment, and maintenance suggestions are simultaneously sent to maintenance personnel. These strategies and suggestions, such as adjusting reactor operating parameters or replacing faulty components, guide the troubleshooting and equipment maintenance, ensuring the safe and stable operation of the reactor.

[0077] As an optional implementation, a Fast Fourier Transform (FFT) algorithm is used to perform spectral analysis on the vibration signal, with a frequency resolution set to 0.1 Hz. By analyzing the spectrum, the resonant frequencies of the reactor outside the rated operating frequency of 50 Hz can be identified. For example, if a resonant frequency with an amplitude of 20% of the rated value is found at 75 Hz, it is determined to be a resonance caused by a non-preset operating frequency. The data acquisition system immediately enters deep acquisition mode, increasing the sampling frequency to 10 kHz, and continuously acquires vibration data for 2 minutes.

[0078] Data preprocessing employed a 5th-order Butterworth low-pass filter with a cutoff frequency of 500Hz to effectively remove high-frequency noise. Wavelet thresholding was then used for noise reduction, improving the signal-to-noise ratio by 20dB. The time-frequency analysis module selected the Hanning window as the time window, with a window length of 1 second and an overlap rate of 50%. A 512-point FFT was performed on the data within each time window to generate a time-frequency plot. The amplitude, frequency, and phase of each frequency component were extracted from the time-frequency plot to form a feature vector.

[0079] Vibration state assessment employed a Support Vector Machine (SVM) algorithm with a radial basis function kernel. The training sample size was 1000, including 600 normal state samples and 400 abnormal state samples. Optimal model parameters were determined using 5-fold cross-validation, achieving a classification accuracy of 98% on the test set. Based on historical operating data of the reactor, vibration amplitude exceeding 30% of the rated value, lasting for more than 30 seconds, and with a frequency between 65Hz and 85Hz were classified as abnormal vibration. Currently, the reactor's vibration amplitude is 35% of the rated value, and the duration has reached 1 minute. The assessment model determined it to be in an abnormal vibration state, generating a vibration anomaly alarm and triggering the fault diagnosis module.

[0080] Fault diagnosis employs a rule-based expert system, comprehensively analyzing the reactor's operating parameters such as current, voltage, and temperature, as well as vibration spectrum characteristics. This data is matched against fault rules in the knowledge base, initially determining the resonance to be caused by a loose internal core of the reactor, and generating a fault diagnosis report. Based on the diagnosis results, a control strategy is generated, suggesting reducing the reactor's operating current to decrease vibration amplitude. Simultaneously, maintenance recommendations are generated, suggesting that the reactor be disassembled and inspected during the next power outage maintenance to check the core's fixation and tighten it.

[0081] In step S103 above, under the condition that the reactor body is working normally, the vibration characteristic parameters are used as input excitation based on the multibody dynamics model to perform simulation calculation and predict the simulation response data after vibration isolation. The multibody dynamics model is established based on the calibrated reactor model and the current design parameters of the vibration isolation device. The calibrated reactor model is obtained by comparing and calibrating the vibration signals under different working conditions based on the theoretical model of the reactor.

[0082] In practice, the vibration mode caused by the preset frequency is identified by analyzing the vibration characteristics. Combined with the design parameters and actual working state of the reactor, the vibration signal of the reactor is decomposed by signal processing methods. Using a multibody dynamics model, the isolation or weakening effect of the vibration isolation device is determined.

[0083] As an optional embodiment, for non-dominant frequency resonance, under the condition that the reactor body is working normally, before performing simulation calculations based on the multibody dynamics model and using vibration characteristic parameters as input excitation to predict the simulation response data after vibration isolation, the method further includes: collecting vibration signals of the reactor at different operating frequencies; creating a theoretical model of the reactor and calculating the theoretical natural frequencies and theoretical mode shapes of the reactor at different operating frequencies; comparing the theoretical natural frequencies and theoretical mode shapes with the frequencies and mode shapes of the measured vibration signals, iteratively optimizing the model parameters of the theoretical model to calibrate the reactor model; and establishing a multibody dynamics model of the vibration isolation device based on the calibrated reactor model and the current design parameters of the vibration isolation device.

[0084] Vibration sensors are used to collect vibration signals from the reactor under different operating conditions, and the collected vibration signals are transmitted to a data processing system. Time-domain and frequency-domain analyses are performed on the vibration signals, and the spectral characteristics of the vibration signals are calculated using a power spectral density estimation algorithm to identify the main frequency components and their corresponding vibration modes.

[0085] Based on the design parameters and actual operating conditions of the reactor, a theoretical model of the reactor's vibration characteristics is established. The vibration characteristics of the reactor are simulated and calculated using the finite element analysis method. The theoretical natural frequencies and theoretical mode shapes of the reactor under different operating conditions are obtained and compared with the measured vibration frequencies and mode shapes to determine the actual vibration mode of the reactor.

[0086] Wavelet transform signal processing is used to perform time-frequency analysis and feature extraction on the collected vibration signal. The vibration signal is decomposed into sub-signals of different frequency bands by wavelet packet decomposition, and the energy distribution of each sub-signal is calculated to determine the frequency band where the vibration energy is mainly concentrated and to identify the vibration mode caused by the preset frequency.

[0087] Based on the structural characteristics and installation location of the reactor, determine the layout scheme of the vibration isolation device, select the vibration isolation material and the model of the vibration isolation device, and calculate the natural frequency and transmission characteristics of the vibration isolation device through dynamic characteristic analysis of the vibration isolation device, and judge the isolation effect of the vibration isolation device on the preset frequency vibration.

[0088] Multibody dynamics simulation software was used to model and simulate the reactor-vibration isolation device system. The measured vibration signal was input, the response characteristics of the system at different vibration frequencies were calculated, the vibration transfer function before and after vibration isolation was obtained, and the effect of the vibration isolation device on reducing vibration at a preset frequency was evaluated.

[0089] If the vibration isolation effect is not ideal, the parameter settings of the vibration isolation device are optimized based on the vibration characteristic analysis results and the dynamic characteristics of the vibration isolation device. The optimal vibration isolation scheme is obtained through iterative calculation and simulation verification.

[0090] The optimized vibration isolation scheme was applied to the actual reactor installation. Through vibration testing and long-term monitoring, the actual isolation effect and service life of the vibration isolation device were evaluated. Based on the feedback from the monitoring data, the vibration isolation scheme was continuously improved and perfected.

[0091] As an optional implementation, the vibration sensor collects the vibration signal of the reactor under rated operating conditions, with a sampling frequency of 10kHz and a sampling duration of 60 seconds. The data processing system uses a 6th-order Butterworth low-pass filter to preprocess the vibration signal, with a cutoff frequency set to 1000Hz to filter out high-frequency noise.

[0092] The vibration signal was then windowed using a Hanning window with a window length of 1 second and an overlap rate of 50%. A 2048-point FFT was used to perform spectral analysis on the windowed signal, yielding a spectrum with a resolution of 0.5 Hz. Based on the peak positions in the spectrum, the main frequency components were identified as 50 Hz, 150 Hz, and 250 Hz, with corresponding peak amplitudes of 0.2 g, 0.05 g, and 0.01 g, respectively.

[0093] The power spectral density of the spectrum was estimated using the Welch's T-test (unequal variance T-test) with a frequency resolution of 1 Hz. The relative intensities of the main frequency components were determined to be 0.8, 0.15, and 0.05, respectively. Based on the theoretical model, the reactor's natural mode shape at 50 Hz is bending vibration, at 150 Hz it is torsional vibration, and at 250 Hz it is axial vibration.

[0094] Finite element simulation analysis revealed that the vibration mode of the reactor under actual operating conditions was basically consistent with the theoretical model, but some frequency shift and mode coupling phenomena existed. A 5-level wavelet packet decomposition of the vibration signal was performed using db4 wavelets, yielding the energy distribution of 32 sub-frequency bands. Comparative analysis of the energy distribution with the theoretical vibration mode determined that the reactor's main vibration energy was concentrated in the bending vibration mode around 50Hz, accounting for over 70% of the total energy.

[0095] Taking into account the structural characteristics and vibration properties of the reactor, a rubber-steel wire rope composite vibration isolation device was selected as the vibration isolation device. A multibody dynamics model of the reactor-vibration isolation device coupled system was established using simulation software. The stiffness and damping parameters of the vibration isolation device were optimized so that the natural frequency of the vibration isolation device avoids the main vibration frequency of the reactor.

[0096] Simulation results show that the optimized vibration isolation scheme can reduce the vibration transmissibility near 50Hz by more than 80%, effectively suppressing the resonance phenomenon of the reactor. Finally, the optimized vibration isolation scheme was applied in the actual reactor installation, and online monitoring was conducted for 6 months using an accelerometer. The results show that the actual vibration isolation effect of the vibration isolation device is basically consistent with the simulation prediction, the overall vibration level of the reactor is reduced by more than 70%, and the operating condition is good.

[0097] Specifically, in creating the multi-vibration dynamic model, a simulation method is used to model the vibration isolation device, creating a simulation model to simulate the vibration isolation effect at different resonant frequencies. The simulation model will be calibrated based on the actual operating conditions of the reactor and the aforementioned analysis results.

[0098] Based on the structural parameters and material properties of the reactor, a three-dimensional solid model of the reactor is established using the finite element method. Mesh generation and boundary condition settings are then performed to obtain the numerical model of the reactor, which serves as the basis for modeling the vibration isolation device. Based on the type and parameters of the vibration isolation device, the type of mechanical model is selected, and a mathematical model of the vibration isolation device is established. This model is then coupled with the reactor model to form a complete numerical model of the vibration isolation device.

[0099] When selecting a vibration isolation device model, the following factors need to be considered: the type and structural characteristics of the vibration isolation device, such as rubber vibration isolation devices, spring vibration isolation devices, and damping vibration isolation devices. Different types of vibration isolation devices have different mechanical properties and applicable ranges. The material properties and nonlinear characteristics of the vibration isolation device, such as the hyperelasticity and viscoelasticity of rubber materials, and the plasticity and fatigue characteristics of metal materials, require the selection of an appropriate mechanical model based on the constitutive relations of the materials.

[0100] The frequency range and dynamic load characteristics of vibration isolation devices, such as low-frequency vibration isolation, high-frequency vibration isolation, and impact vibration isolation, all influence the requirements for the vibration isolation device model. Therefore, it is necessary to select a model with corresponding characteristics. The purpose and accuracy requirements of simulation analysis, such as vibration isolation effect evaluation, fatigue life prediction, and failure mechanism analysis, also influence the level of detail and computational efficiency of the vibration isolation device model. A balance must be struck between accuracy and efficiency.

[0101] By employing both modal analysis and harmonic response analysis, three characteristic parameters of the vibration isolation device—natural frequency, mode shape, and frequency response function—under different excitation frequencies are calculated. This determines the inherent characteristics and dynamic response law of the vibration isolation device, providing a theoretical basis for subsequent vibration isolation effect analysis. Based on the actual operating conditions of the reactor, such as voltage, current, and temperature, the input and constraint conditions of the simulation model are determined. The results of the aforementioned vibration characteristic analysis are used as the initial conditions and verification basis for the simulation model. Parameter identification and correction are performed on the simulation model to obtain a multibody dynamics model, improving the accuracy and reliability of the simulation results.

[0102] Multiphysics coupling analysis methods, such as electromagnetic field and structural field coupling, are used to simulate the vibration response of the reactor under actual working conditions. The influence of the reactor's electromagnetic excitation characteristics, hydrodynamic characteristics and acoustic radiation characteristics on vibration is analyzed, and the vibration distribution and transmission characteristics of the reactor at different resonant frequencies are obtained.

[0103] By changing the parameter settings of the vibration isolation device, such as stiffness coefficient, damping coefficient, and arrangement position, multiple sets of simulation calculations are performed to obtain vibration isolation effect data under different vibration isolation schemes. Comparative analysis is then conducted to determine the optimal combination of vibration isolation parameters and arrangement method, thereby achieving optimized design of vibration isolation effect.

[0104] During optimization, the objective function and constraints of the optimization design are first determined, such as maximum vibration isolation efficiency, minimum vibration isolation device volume, minimum cost, and displacement limits, strength restrictions, and stability requirements of the vibration isolation device. A suitable optimization algorithm is then selected, such as parametric scanning, response surface methodology, genetic algorithm, or particle swarm optimization. The appropriate optimization method is chosen based on factors such as the type and number of design variables, the characteristics of the objective function, and computational cost.

[0105] An experimental scheme was designed, and based on the range and distribution characteristics of the design variables, orthogonal experiments and Latin hypercube experiments were used to generate a set of representative test sample points to construct a surrogate model of the vibration isolation effect. Simulation calculations were performed, and simulation analysis was conducted for each sample point according to the experimental scheme to obtain the response value of the vibration isolation effect. The simulation results were then compiled and analyzed to evaluate the distribution law and influencing factors of the vibration isolation effect.

[0106] A surrogate model is constructed. Based on simulation results, methods such as multinomial regression, radial basis functions, and Kriging models are used to establish an approximate relationship between vibration isolation effects and design variables. This model replaces time-consuming simulation calculations, improving optimization efficiency. Optimization is then performed by combining the surrogate model with optimization algorithms. Based on the objective function and constraints, the optimal combination of design parameters is solved to obtain the global or local optimum solution. The optimization results are then verified and analyzed.

[0107] The process involves making a decision on the vibration isolation scheme. Based on the optimization results and other factors such as manufacturing cost, reliability, and maintainability, different vibration isolation schemes are comprehensively evaluated and weighed to select the final vibration isolation device design scheme. Detailed design and verification are then conducted. The optimized vibration isolation scheme is applied to the actual reactor vibration isolation device. Vibration testing and long-term monitoring are used to verify the effectiveness and stability of the vibration isolation effect. Based on the measured data, the simulation model is further revised and improved, establishing a closed-loop optimization mechanism between the simulation model and the actual system. This provides reliable technical support for the design and optimization of reactor vibration isolation devices.

[0108] As an alternative implementation, the reactor is made of Q235 steel. Based on its geometric dimensions and structural characteristics, it is modeled using finite element analysis software (taking ANSYS as an example), with the SOLID186 element type selected and the mesh size set to 50mm. Fixed constraints are applied to the bottom of the reactor, and a simple harmonic excitation load with an amplitude of 1kN and a frequency range of 0-500Hz is applied to the top.

[0109] The vibration isolation device uses natural rubber material, and its nonlinear hyperelastic characteristics are described using the Mooney-Rivlin model. The material parameters are C10=1.2MPa and C01=0.8MPa. The mechanical behavior of the rubber vibration isolation device is simulated using COMBIN14 elements, with a stiffness coefficient of 2000kN / m and a damping ratio of 0.05. Coupled constraints connected to the reactor are applied at both ends of the vibration isolation device to form a finite element model of the reactor-vibration isolation device coupled system.

[0110] Modal analysis yielded the first 10 natural frequencies of the coupled system, which are 12.6 Hz, 35.8 Hz, 68.3 Hz, etc. Resonance response analysis obtained the vibration transmissibility curves at the top of the reactor under different excitation frequencies, and found that there is a significant resonance peak near 43.5 Hz, with a vibration transmissibility as high as 8.2.

[0111] Based on the reactor's rated operating voltage of 35kV, current of 600A, and ambient temperature of 40℃, the load conditions of the simulation model were determined, and the 43.5Hz resonant frequency obtained from the previous analysis was selected as the key frequency of interest. Using an electromagnetic-structural coupling analysis method, considering the electromagnetic field distribution and electromagnetic force within the reactor, its vibration response at the 43.5Hz excitation frequency was simulated. The results showed that the vibration displacement at the top of the reactor reached 2.5mm, and the stress was concentrated in the fixed area at the bottom, with a maximum stress of 56MPa.

[0112] Through parameter sensitivity analysis, it was determined that the stiffness and damping coefficient of the rubber vibration isolation device are the main factors affecting the vibration isolation effect. Using orthogonal experimental design method, stiffness coefficient and damping ratio were selected as design variables. Five levels were taken in the ranges of 1000-5000kN / m and 0.02-0.1, for a total of 25 sets of tests. The vibration isolation efficiency under each set of tests was obtained through simulation calculation, and a quadratic response surface model of vibration isolation efficiency with respect to design variables was constructed.

[0113] The surrogate model was optimized and solved using a sequential quadratic programming algorithm. The optimal vibration isolation parameters were found to be a stiffness coefficient of 3600 kN / m and a damping ratio of 0.08. Under these parameters, the vibration isolation efficiency can reach more than 85%, and the vibration displacement at the top of the reactor is reduced to less than 0.3 mm.

[0114] The optimized design results were applied to the actual reactor vibration isolation device and monitored for 6 months. The test results showed that the optimized vibration isolation device had a stable vibration isolation effect under the reactor operating conditions, which was in good agreement with the simulation prediction results, thus verifying the effectiveness and reliability of the simulation optimization design.

[0115] Step S104: Compare the predicted simulated response data after vibration isolation with the measured signal after vibration isolation to evaluate the isolation effect of the vibration isolation device. Based on the obtained simulation results (i.e., the simulated response data), if the amplitude of a non-dominant frequency resonant component exceeds a preset threshold, it is defined as a preset non-dominant frequency resonance. The isolation capability of the vibration isolation device for the preset non-dominant frequency resonance is then determined. Finally, the vibration isolation performance at the frequency point is evaluated to determine the performance boundary of the vibration isolation device.

[0116] The vibration response data of the vibration isolation device under different frequency excitations are obtained. The vibration response data includes vibration displacement, velocity, and acceleration. The vibration displacement, velocity, and acceleration response data are stored in a database to establish a vibration isolation performance database, providing data support for subsequent performance evaluation.

[0117] Based on the vibration response data, non-dominant frequency resonant components and their corresponding frequency points are identified. If the amplitude of a non-dominant frequency resonant component exceeds a preset threshold, it is defined as a preset non-dominant frequency resonant component.

[0118] During implementation, the range of the dominant frequency amplitude is pre-set based on the reactor's design parameters and operating experience, such as 5% to 10% of the rated voltage. The actual amplitude of the dominant frequency component is determined by performing spectral analysis on the reactor's measured vibration data. The amplitudes of the non-dominant frequency components are then compared with the dominant frequency amplitude to calculate their relative amplitude ratios.

[0119] Based on the vibration control requirements and relevant standards for reactors, the amplitude limit of non-dominant frequency resonant components is set, such as 20% or 30% of the dominant frequency amplitude. Quantitative analysis is performed on each indicator, and calculation formulas and measurement methods are established to ensure the measurability and comparability of the indicators.

[0120] Methods such as the analytic hierarchy process (AHP) and entropy weighting were employed to rank the importance of each indicator and assign weights. The determination of weights comprehensively considered expert experience, measured data, and theoretical analysis results. The indicator weights were normalized, and a comprehensive evaluation model for vibration isolation performance was constructed, such as a weighted summation model or a product model. The evaluation model was tested and optimized through case studies and simulation verification to ensure its scientific validity and effectiveness.

[0121] A multi-index weighted evaluation method was adopted to establish a vibration isolation performance evaluation index system, which includes four indicators: vibration isolation efficiency, vibration attenuation rate, frequency characteristics, and energy transfer ratio.

[0122] Based on the comprehensive evaluation results of vibration isolation performance at various frequency points, and according to the characteristics of vibration isolation performance indicators and the evaluation purpose, an appropriate comprehensive evaluation method is selected. Commonly used methods include weighted average method, geometric average method, ideal point method, and grey relational method.

[0123] The weighted average method is suitable for situations where the weights of indicators differ greatly and are highly complementary. It can reflect the importance of each indicator, but it is easily affected by extreme values.

[0124] The geometric mean method is suitable for situations where the differences in indicator weights are small and the balance is strong. It can balance the differences between indicators, but it is difficult to highlight key indicators. The ideal point method is suitable for judging how close the indicator value is to the ideal value. It can intuitively reflect the distance between the evaluated object and the ideal state, but it requires reasonable setting of the ideal point and negative ideal point.

[0125] Grey relational analysis is suitable for situations where the correlation between indicators is strong and the number of samples is small. It can effectively handle incomplete and uncertain information, but the calculation process is relatively complex.

[0126] In this embodiment, various comprehensive evaluation methods can be used for comparative analysis according to specific circumstances to improve the reliability and persuasiveness of the evaluation results. The simulation response data of the vibration isolation device is substituted into the evaluation model to calculate the vibration isolation performance index values ​​at each frequency point, obtaining the curve of vibration isolation performance changing with frequency. This curve is then compared with preset performance standards to determine the effectiveness level of the vibration isolation device at each frequency point.

[0127] By substituting the simulation response data of the vibration isolation device into the index system, the vibration isolation performance index values ​​at each frequency point are calculated, and the curve of vibration isolation performance changing with frequency is obtained.

[0128] Through comprehensive evaluation, the overall performance index of the vibration isolation device is calculated, its performance level and effectiveness boundary under the existing design are determined, and the weak performance range and influencing factors are identified.

[0129] For a preset non-dominant frequency resonant point, extract the vibration response data at that frequency point, calculate the vibration transmissibility before and after vibration isolation, evaluate the isolation effect of the vibration isolation device on the resonant component, and mark it as a weak point in performance if the isolation effect does not meet the requirements.

[0130] Based on the comprehensive evaluation results of vibration isolation performance at various frequency points, the overall performance index of the vibration isolation device is calculated to determine its performance level and effectiveness boundary under the existing design.

[0131] Based on the vibration isolation performance evaluation results, vibration isolation parameters and structural improvements are carried out for weak performance ranges and non-dominant frequency resonances, such as adjusting the stiffness and damping coefficient of the vibration isolation device and optimizing the arrangement of the vibration isolation device. Through repeated simulation verification and evaluation, the optimal vibration isolation design scheme that meets the performance requirements is obtained, and it is applied and verified in actual engineering projects to continuously improve the design and optimization methods of the vibration isolation device.

[0132] As an optional implementation method, for a vibration isolation device of a certain type of reactor, the vibration response data in the frequency range of 0-500Hz is obtained by finite element simulation calculation, including the time domain waveforms and frequency domain curves of vibration displacement, velocity and acceleration.

[0133] The simulation data was discretized with a frequency step of 1 Hz, and characteristic parameters at each frequency point were extracted to construct a vibration isolation performance database. Measured vibration signals from the reactor were acquired at a sampling frequency of 5 kHz and a sampling duration of 60 seconds. Spectral analysis of the measured signals was performed using Fast Fourier Transform (FFT), determining the dominant frequency component to be 50 Hz with an amplitude of 0.2 m / s². Based on this, the threshold for non-dominant frequency resonance was set at 0.06 m / s², which is 30% of the dominant frequency amplitude.

[0134] Peak search of the spectral data revealed resonance peaks at 150Hz and 305Hz, with amplitudes of 0.08m / s² and 0.12m / s², respectively. These exceeded the judgment threshold and were identified as non-dominant frequency resonances. Based on the type and installation location of the vibration isolation device, five indicators, including vibration isolation efficiency, displacement transmissibility, and acceleration amplification factor, were selected to construct a vibration isolation performance evaluation index system.

[0135] The weights of each index were determined using the Analytic Hierarchy Process (AHP). A comprehensive evaluation model for vibration isolation performance was established through index normalization and weighted summation. Simulation data was imported into the evaluation model to calculate the comprehensive vibration isolation effectiveness level of the isolation device at each frequency point, and the frequency response curve of the vibration isolation performance was plotted.

[0136] The non-dominant resonant frequencies of 150Hz and 305Hz were specifically evaluated. The vibration transmissibility before and after vibration isolation was calculated to be -12dB and -8dB, respectively, both failing to meet the design requirement of -20dB, and were marked as weak points in performance. A weighted average method was used to comprehensively evaluate the vibration isolation effectiveness at each frequency point, and the overall performance index of the vibration isolation device was obtained as 0.78, which is at a medium level.

[0137] Further analysis revealed that the horizontal and vertical stiffness of the vibration isolation device were 15% and 22% lower than the design values, respectively. This resulted in a lower natural frequency, close to the resonant frequency, which was the main factor causing the performance degradation. Based on this, the stiffness parameters of the vibration isolation device were optimized. Through repeated simulations and iterations, it was determined that increasing the horizontal stiffness by 20% and the vertical stiffness by 30% improved the overall performance index of the vibration isolation device to 0.92, meeting the design requirements.

[0138] An optimized design parameter table was generated to guide the production and installation of the vibration isolation device. The device was then tested and verified on an actual reactor. Ultimately, it was determined that the optimized vibration isolation scheme could effectively suppress non-dominant frequency resonances and had good vibration isolation performance.

[0139] Step S105 involves controlling the vibration isolation device based on its isolation effect to suppress non-dominant frequency resonance. This mainly includes: determining the transmissibility-frequency curve of the vibration isolation device as its performance boundary based on a multibody dynamics model, according to the isolation effect of the device; comparing the performance boundary with a preset isolation threshold to determine whether it meets the standard at each frequency point; modifying the isolation parameters of the multibody dynamics model when the isolation effect at a frequency point is lower than the preset threshold, and finding the optimal parameter combination through an optimization algorithm; and controlling the vibration isolation device based on the optimal parameter combination.

[0140] As an optional embodiment, the performance boundary is compared with a preset vibration isolation threshold to determine whether the standard is met at each frequency point. This includes: determining the transmissibility value corresponding to each frequency point based on the transmissibility-frequency curve of the performance boundary; calculating the performance index of each frequency point based on the transmissibility value and the established evaluation index system, wherein the performance index includes at least one of the following: vibration isolation efficiency, vibration attenuation rate, frequency characteristics, and energy transfer ratio; comparing the performance index of each frequency point with the corresponding preset vibration isolation threshold, and determining whether the corresponding frequency point meets the standard if the performance index is better than the corresponding preset vibration isolation threshold, and whether the corresponding frequency point does not meet the standard if the performance index is worse than the corresponding preset vibration isolation threshold.

[0141] In other words, the above-mentioned comprehensive evaluation methods are used to calculate the overall performance index of the vibration isolation device, determine its performance level and effectiveness boundary under the existing design, and identify the weak performance range and influencing factors.

[0142] Based on the above vibration isolation effect evaluation, the frequency points where the vibration isolation effect is lower than the vibration isolation threshold at a specific resonant frequency were identified as the weak performance points. These were then the focus of optimization design, and response data was extracted for subsequent parameter identification and optimization analysis. For the weak performance frequency points, i.e., those where the vibration isolation effect is lower than the vibration isolation threshold at the preset resonant frequency, the vibration isolation parameters were modified, and the spring stiffness and damping material properties in the vibration isolation device were adjusted.

[0143] As an optional embodiment, when the vibration isolation effect at a frequency point is lower than a preset effect threshold, the vibration isolation parameters of the multibody dynamics model are modified, and the optimal parameter combination is found through an optimization algorithm. This includes: obtaining response data at the weak performance frequency point where the vibration isolation effect is lower than the preset effect threshold, the response data including three parameters: amplitude, phase, and transmissivity; calculating the sensitivity coefficient of the vibration isolation effect relative to each vibration isolation design parameter based on the multibody dynamics model; determining the vibration isolation design parameter that has the greatest impact on the vibration isolation effect based on the sensitivity coefficient; and using a parameter optimization algorithm to search for the optimal parameter combination of the vibration isolation design parameters.

[0144] Response data of the vibration isolation device at frequencies where the vibration isolation effect is below the isolation threshold are extracted. This response data includes three parameters: amplitude, phase, and transmissibility. Based on the structural type and working principle of the vibration isolation device, a mathematical model is established. This model is related to three key parameters: spring stiffness, damping coefficient, and mass distribution. The objective function and design variables for sensitivity analysis are determined, such as using isolation efficiency as the objective function and spring stiffness and damping coefficient as design variables.

[0145] A combination of theoretical analysis and experimental identification was used to determine the initial parameters of the vibration isolation device model. These parameters were then corrected and verified by comparing them with measured response data, resulting in the dynamic model of the vibration isolation device. Sensitivity analysis was employed to investigate the influence of key parameters of the vibration isolation device on its isolation effect. Appropriate sensitivity analysis methods, such as the finite difference method, the adjoint variable method, and the automatic differentiation method, were selected based on the characteristics of the problem and the solution requirements, choosing a method that balances computational efficiency and accuracy.

[0146] Sensitivity analysis was employed to calculate the response changes of the vibration isolation device by altering the spring stiffness and damping coefficient of the mathematical model, thereby obtaining the sensitivity coefficients of each parameter. Within the range of design variables, small perturbations were applied to each variable, and the changes in the objective function were calculated to obtain the partial derivatives of the objective function with respect to the design variables, i.e., the sensitivity coefficients.

[0147] The sensitivity coefficients are normalized to obtain the relative sensitivity of each design variable. The relative sensitivity is then ranked and compared to determine its impact on the objective function, thus identifying the key design variables.

[0148] Sensitivity analysis was conducted to verify and confirm the accuracy and reliability of the sensitivity coefficients by applying positive and negative perturbations to the design variables, and to analyze the locality and globality of the sensitivity. Based on this, the key parameters that have the greatest impact on the vibration isolation effect were identified as the main control variables for the optimization design.

[0149] Based on the nonlinear characteristics of damping materials, a constitutive model is used to describe their mechanical behavior. The constitutive model combines the hyperelasticity, viscoelasticity and damping properties of the material.

[0150] To address the nonlinear characteristics of damping materials, a constitutive model is employed to describe their mechanical behavior. Considering the types and properties of damping materials, such as rubber, elastomers, and viscoelastic materials, different types exhibit different mechanical behaviors and applicable ranges. The stress-strain relationship and strain rate dependence of the materials are analyzed, including their nonlinearity, anisotropy, and viscoelasticity, to select a constitutive model that accurately describes these characteristics.

[0151] Assess the temperature and frequency dependence of materials, such as how material properties change with temperature and frequency, and select a constitutive model that can account for the influence of these factors. Balance the difficulty of parameter identification with computational efficiency, such as the physical meaning and measurability of model parameters, and the numerical solution efficiency of the model, achieving a balance between accuracy and practicality. Obtain the constitutive parameters of the material through material testing and parameter identification, and establish a nonlinear dynamic model of the damping material.

[0152] By combining the structural parameters of the vibration isolation device and the properties of the damping material, a multi-parameter optimization model of the vibration isolation device is established, and a multidisciplinary optimization design method is used to solve it.

[0153] Taking into account the structural parameters and damping material properties of the vibration isolation device, a multi-parameter optimization model of the vibration isolation device is established. The objective function is to improve the vibration isolation efficiency at a specific resonant frequency. The design variables are the size, shape, and material selection of the vibration isolation device, and the constraints are the mass, strength, and stability of the system. The parameter optimization model is constructed and solved using a multidisciplinary optimization design method.

[0154] By analyzing the sensitivity of the design parameters, the parameters that contribute the most to the vibration isolation effect are identified. The optimal parameter combination is then searched using a parameter optimization algorithm to obtain the improved design parameters of the vibration isolation device.

[0155] During the optimization design process, sensitivity analysis of design parameters is used to identify the parameters that contribute most to the vibration isolation effect, such as spring stiffness, damping coefficient, elastic modulus of damping material, and loss factor, and these parameters are then optimized in a focused manner.

[0156] Based on the characteristics and scale of the optimization problem, such as the type and number of design variables, the form and complexity of the objective function, and the types and number of constraints, a suitable optimization algorithm should be selected, such as genetic algorithms or particle swarm optimization. The global and local search capabilities of the optimization algorithm should be balanced, considering factors such as the algorithm's exploration and development mechanisms, convergence speed, and accuracy. An algorithm that balances search efficiency and optimization effect should be selected according to the needs of the problem.

[0157] Through sensitivity analysis and experimental optimization of algorithm parameters, the optimal combination of parameters such as population size, crossover probability, and mutation probability is determined. The optimized vibration isolation device design parameters are applied to the simulation model of the vibration isolation device to predict and analyze its vibration isolation performance. The performance is then compared with that before optimization to evaluate the effectiveness of the optimized design.

[0158] Furthermore, targeted strategies such as domain knowledge, problem decomposition, and multi-objective optimization can be introduced to improve and integrate the algorithm, thereby enhancing its optimization performance and adaptability. The optimal parameter combination is searched using a parameter optimization algorithm to obtain the improved design parameters for the vibration isolation device. These optimized design parameters are then applied to the simulation model of the vibration isolation device to predict and analyze its vibration isolation performance. The results are then compared with the performance before optimization to evaluate the effectiveness of the optimized design.

[0159] The effectiveness and reliability of the optimization algorithm are evaluated through comparison with other algorithms and experimental verification, and the algorithm is further improved and refined. If the optimization effect meets the requirements, an optimized design parameter table for the vibration isolation device is generated to guide engineering implementation; if the optimization effect is not ideal, the process is returned to the optimization design process, and the optimization model and algorithm are further adjusted until a satisfactory design result is obtained, forming a closed-loop optimization design process that is continuously iterated and updated to continuously improve the performance of the vibration isolation device.

[0160] As an alternative implementation method, simulation analysis and experimental testing of the vibration isolation device revealed a significant decrease in vibration isolation effectiveness near 250Hz, with the isolation efficiency dropping from 85% to 60%, far below the design requirement of 80%. Response data of the vibration isolation device were extracted for this resonant frequency, including an amplitude of 1.2mm, a phase of -90°, and a transmissibility of -8dB.

[0161] Based on the structural parameters of the vibration isolation device, a 2-degree-of-freedom dynamic model was established, taking into account factors such as spring stiffness k, damping coefficient c, and mass m. A genetic algorithm was used to identify the model parameters, and the initial parameters k=10kN / m, c=500N·s / m, and m=20kg were obtained. The fitting error with the measured data was 5%.

[0162] Sensitivity analysis of spring stiffness and damping coefficient was performed using the central difference method with a step size of 10%. The resulting sensitivity coefficients were 0.8 and -0.6, respectively, indicating that spring stiffness has a greater impact on vibration isolation. Considering the nonlinear characteristics of the damping material, a three-parameter Maxwell model was selected to describe its dynamic mechanical behavior. Through dynamic mechanical property testing, the material parameters E′=8MPa, E″=1.5MPa, and τ=2ms were obtained.

[0163] Using the structural and material properties of the vibration isolation device as design variables, and with vibration isolation efficiency and frequency ratio as objective functions, and mass, volume, and displacement as constraints, a multi-objective optimization model is constructed. The NSGA-II genetic algorithm (Nondominated Sorting Genetic Algorithm II with elitist strategy) is selected for optimization, with a population size of 50, a crossover probability of 0.8, a mutation probability of 0.1, and 100 generations of iterations to obtain the Pareto optimal solution set.

[0164] Balancing vibration isolation efficiency and cost, the optimal solution k=15kN / m, c=800N·s / m, and E′=10MPa was selected. Applying the optimization results to the simulation model, the vibration isolation efficiency near 250Hz was improved to 88%, meeting the design requirements. Comparing the optimization effects of the genetic algorithm and the particle swarm optimization algorithm, the former has a faster convergence speed and better optimization effect.

[0165] Finally, an optimized design parameter table for the vibration isolation device was generated and applied to the manufacturing of the engineering prototype. Through testing, the dynamic performance of the vibration isolation device was significantly improved, providing an effective approach for the optimized design of the vibration isolation device.

[0166] As an optional embodiment, controlling the vibration isolation device according to the optimal parameter combination includes: real-time acquisition of changes in the grid frequency and load parameters of the power grid where the reactor is located, and real-time monitoring of grid frequency fluctuations; based on the optimal parameter combination and changes in grid frequency and load parameters, using an adaptive control algorithm to calculate the optimal control parameters of the controllable parameters of the vibration isolation device; when the frequency fluctuation of the power grid where the reactor is located is detected to exceed a preset threshold, adjusting the controllable parameters of the vibration isolation device to the corresponding optimal control parameters to control the vibration isolation device.

[0167] The optimized vibration isolation parameters are input into the control unit of the vibration isolation device, and the device parameters are adjusted when the next power grid frequency fluctuation or load change event occurs. The optimized vibration isolation parameters are then input into the database as target values ​​for vibration isolation control.

[0168] Vibration signals from the vibration isolation device are collected by vibration sensors and transmitted to the control unit for processing. Power monitoring devices collect parameters such as grid frequency, voltage, and current, while load detection devices collect parameters such as load amplitude, phase, and harmonics, which are then transmitted to the control unit.

[0169] An adaptive control algorithm is designed in the control unit to calculate the optimal control parameters of the vibration isolation device in real time based on changes in grid frequency and load parameters. If grid frequency fluctuations or load changes exceed preset thresholds, the vibration isolation parameter adjustment mechanism is triggered, and the calculated control parameters are sent to the actuator.

[0170] Vibration isolation parameters include spring stiffness, damping coefficient, and damping material properties. Based on the structural characteristics and control requirements of the vibration isolation device, a suitable control unit, such as a programmable logic controller (PLC) or embedded controller, is selected and equipped with sufficient input / output interfaces and communication protocols to realize information interaction and control command issuance with hardware devices such as sensors and actuators.

[0171] The actuator, based on instructions issued by the control unit, adjusts the spring stiffness and damping coefficient in the vibration isolation device in real time by changing the displacement of the hydraulic cylinder and the current of the electromagnetic coil, thereby achieving active control of the dynamic characteristics of the vibration isolation device.

[0172] Vibration sensors, such as accelerometers and displacement sensors, are placed at key locations in the vibration isolation device to collect vibration signals from the device in real time. The collected data is then transmitted to the control unit via a bus or wireless network. After signal conditioning such as filtering and amplification, the data is converted into digital quantities for use by the control algorithm.

[0173] Transmitting these digital parameter information to the control unit in real time serves as the basis for judging power grid frequency fluctuations and load changes. Commonly used adaptive control algorithms include adaptive PID (Proportional-Integral-Derivative) control, model reference adaptive control, and adaptive sliding mode control. Adaptive PID control adjusts PID parameters online to adapt the controller to changes in the controlled object's parameters, achieving adaptive control performance, and is suitable for linear time-invariant systems.

[0174] Reference adaptive control uses a reference model to provide an ideal input-output relationship, and adjusts the controller parameters to make the input-output relationship of the actual system approximate the reference model. It is suitable for linear time-varying systems. Adaptive sliding mode control, by designing a sliding surface and adaptive law, allows the system state to move on the sliding surface. It has strong robustness to parameter perturbations and external disturbances and is suitable for nonlinear systems.

[0175] Based on the dynamic model and control requirements of the vibration isolation device, the structure and parameters of the adaptive controller are designed, including control gain, the form of the adaptive law, and the design of the disturbance observer. Simulation verification and parameter tuning are then performed. In the actual control system, issues such as the discretization implementation of the adaptive control algorithm, computational efficiency optimization, and numerical stability need to be considered. Furthermore, it needs to be combined with other control strategies to form a composite control system.

[0176] Intelligent actuators such as electro-hydraulic servo actuators or electromagnetic rheodynamic dampers are used to adjust the parameters in the vibration isolation device in real time according to the instructions issued by the control unit. Different control methods can be used for different types of vibration isolation devices.

[0177] For hydraulic vibration isolation devices, stiffness and damping can be adjusted by changing the displacement of the hydraulic cylinder; for electromagnetic rheological vibration isolation devices, the damping coefficient can be adjusted by changing the current of the electromagnetic coil; for rubber vibration isolation devices, stiffness and damping can be adjusted by changing the pre-compression of the rubber material.

[0178] The dynamic characteristics of the vibration isolation device are actively controlled by the actions of the actuators. A human-machine interface for the vibration isolation control system is established, which displays information such as the vibration status, control parameters, and energy consumption of the vibration isolation device in real time through a graphical monitoring screen, and provides functions such as remote debugging, fault diagnosis, and parameter optimization.

[0179] Establish a digital twin model of the vibration isolation device, and realize digital management of the entire life cycle of the vibration isolation device through physical modeling, data-driven methods, etc., including design, manufacturing, installation, operation, and maintenance stages. Develop fault diagnosis and predictive maintenance algorithms for the vibration isolation device, and realize functions such as health status monitoring, remaining life prediction, and fault cause diagnosis of the vibration isolation device through vibration signal analysis, machine learning, and other technologies, thereby improving the reliability and maintainability of the system.

[0180] Optimize the energy consumption and cost of vibration isolation control systems by employing multi-objective optimization algorithms and reinforcement learning techniques to minimize energy consumption and maintenance costs of vibration isolation devices while meeting performance requirements, achieving a balance between economy and performance. Establish a knowledge base and expert system for vibration isolation devices, collecting and organizing design specifications, fault cases, optimization experiences, and other knowledge in the field to form an intelligent decision support system that assists engineers in designing and optimizing vibration isolation devices and troubleshooting problems.

[0181] Through big data analytics and cloud computing platforms, remote monitoring and centralized management of vibration isolation devices can be achieved, providing online diagnostics, early warning, and optimization services to improve the intelligence level and management efficiency of vibration isolation devices and continuously enhance the stability and reliability of vibration isolation control.

[0182] As an optional implementation, for a vibration isolation device of a certain type of reactor, a set of optimal parameters were obtained through multi-objective genetic algorithm optimization, including a spring stiffness of 120 kN / m, a damping coefficient of 3 kN·s / m, and a dynamic elastic modulus of 15 MPa for the damping material. These parameters were then input into a PLC-based vibration isolation control system with a sampling period of 10 ms and a quantization precision of 16 bits.

[0183] Four triaxial acceleration sensors with a range of ±10g and a sensitivity of 500mV / g are installed between the reactor base and the foundation. They communicate with the PLC via the ModBus serial communication protocol to collect vibration signals in real time, and the signals are preprocessed by a digital low-pass filter with a cutoff frequency of 500Hz.

[0184] The system utilizes power monitoring devices to collect grid frequency, voltage, and current data at a sampling frequency of 1kHz. The switching status of capacitor banks and load changes are determined via a switching interface. The control system employs an adaptive fuzzy PID control algorithm, using membership functions to describe the degree of change in frequency, voltage, and current. Twenty-seven fuzzy control rules are established, and the centroid defuzzification method is used to calculate the correction amount for the output vibration isolation parameters.

[0185] When the frequency fluctuation exceeds ±0.5Hz or the load change exceeds 20% of the rated value, adaptive control is triggered, and the output vibration isolation parameter correction is superimposed on the initial parameters to obtain the corrected vibration isolation parameters. For the hydraulic vibration isolation device of the reactor, a servo proportional valve is used to control the displacement of the hydraulic cylinder. The relationship between the valve core displacement and the control current is 0.1mm / mA, and the relationship between the hydraulic cylinder displacement and the spring stiffness is 10kN / m per mm. By changing the control current, the vibration isolation stiffness can be adjusted online, with an adjustment range of 50~200kN / m.

[0186] A human-machine interface for the vibration isolation device is built, and an industrial monitoring and data acquisition system is adopted. The changes of parameters such as frequency, voltage, and current are displayed through dynamic curves, and the vibration isolation parameters and vibration level are displayed through digital instruments. An alarm function is set up so that when the vibration level exceeds 150% of the rated value, an audible and visual alarm is issued.

[0187] A simulation model of the vibration isolation device was established. Dynamic simulation software (such as Matlab / Simulink) was used to obtain the dynamic equation of the reactor through parameter identification. The backpropagation (BP) neural network was trained through experimental data to achieve fault diagnosis, with a diagnosis accuracy of over 95%.

[0188] A genetic algorithm is used to optimize the energy consumption of the vibration isolation device. With vibration isolation efficiency and energy consumption as dual objectives, the Pareto optimal solution is used to balance the weights of the two, thereby reducing energy consumption by 20%. Finally, 5G communication technology is used to realize remote monitoring of the vibration isolation device, and big data analysis is carried out using a cloud computing platform. Machine learning algorithms are used to predict the remaining lifespan of the vibration isolation device and provide early warning one month in advance, realizing intelligent operation and maintenance of the vibration isolation device.

[0189] Figure 2 This is a schematic diagram of a control device for a vibration isolation device of a reactor, according to an embodiment of the present invention. Figure 2 As shown, an embodiment of the present invention also provides a control device for a reactor's vibration isolation device, applied to the control of the reactor's vibration isolation device. The control device for the reactor's vibration isolation device includes:

[0190] The main frequency identification module 201 is used to collect the operating frequency of the reactor in real time and determine whether the resonant frequency corresponding to the operating frequency is a non-main frequency resonance.

[0191] The anomaly detection module 202 is connected to the main frequency identification module 201 mentioned above. It is used to collect vibration signals at the corresponding non-preset operating frequency for non-main frequency resonance, extract vibration characteristic parameters of the vibration signals, and determine whether the reactor body is working normally.

[0192] The predictive response module 203, connected to the aforementioned anomaly judgment module 202, is used to perform simulation calculations based on a multibody dynamics model, using vibration characteristic parameters as input excitation, to predict the simulation response data after vibration isolation, when the reactor body is working normally. The multibody dynamics model is established based on the calibrated reactor model and the current design parameters of the vibration isolation device. The calibrated reactor model is obtained by comparing and calibrating vibration signals under different operating conditions based on the theoretical model of the reactor.

[0193] The effect evaluation module 204, connected to the prediction response module 203, is used to compare the predicted simulation response data after vibration isolation with the measured signal after vibration isolation to evaluate the isolation effect of the vibration isolation device.

[0194] The feedback control module 205 is connected to the effect evaluation module 204 and is used to control the vibration isolation device according to the isolation effect in order to suppress non-master frequency resonance.

[0195] The embodiments of this invention provide that by real-time acquisition of the reactor's operating frequency, the resonant frequency corresponding to the operating frequency is determined to be a non-dominant frequency resonance; for non-dominant frequency resonance, vibration signals at the corresponding non-preset operating frequency are acquired, and vibration characteristic parameters of the vibration signals are extracted to determine whether the reactor body is working normally; effectively eliminating the influence of abnormal reactor body operation on the evaluation of vibration isolation effect of vibration isolation device, and improving the accuracy of vibration isolation effect detection of vibration isolation device.

[0196] Under normal operating conditions of the reactor, based on a multibody dynamics model, vibration characteristic parameters are used as input excitation to perform simulation calculations and predict the simulated response data after vibration isolation. The predicted simulated response data after vibration isolation is then compared with the measured signal after vibration isolation to evaluate the isolation effect of the vibration isolation device. This allows for real-time and accurate evaluation of the isolation effect of the vibration isolation device.

[0197] Moreover, the multibody dynamics model is established based on the calibrated reactor model and the current design parameters of the vibration isolation device. The calibrated reactor model is obtained by comparing and calibrating the reactor's theoretical model with vibration signals under different operating conditions, which can further improve the accuracy of the predicted simulation response data, thereby improving the accuracy of the vibration isolation effect evaluation of the vibration isolation device.

[0198] The vibration isolation device is controlled based on its isolation effect to suppress non-dominant frequency resonances. By accurately assessing the real-time vibration isolation effect, the device can be controlled promptly to suppress non-dominant frequency resonances. This achieves accurate assessment of the vibration isolation effect and allows for effective control and adjustment of the device based on that effect, thus improving the technical effectiveness of the vibration isolation device in suppressing non-dominant frequency resonances.

[0199] This solves the problem in the aforementioned related technologies that power grid reactors are prone to generating non-dominant frequency resonances with wide frequency distributions and diverse vibration modes, which makes existing vibration isolation devices unable to meet the requirements for effective vibration isolation.

[0200] Embodiments of the present invention also provide a non-transitory machine-readable medium storing a computer program, wherein the computer program, when executed by a computer's processor, is used to cause the computer to perform a method according to an embodiment of the present invention.

[0201] Embodiments of the present invention also provide a computer program product, including a computer program, wherein the computer program, when executed by a computer's processor, is used to cause the computer to perform the method of an embodiment of the present invention.

[0202] An embodiment of the present invention also provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores a computer program executable by the at least one processor, which, when executed by the at least one processor, causes the electronic device to perform the method of the embodiment of the present invention.

[0203] refer to Figure 3 The present invention will now describe a structural block diagram of an electronic device that can serve as an embodiment of the present invention, serving as an example of a hardware device applicable to various aspects of the present invention. The electronic device is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0204] like Figure 3As shown, the electronic device includes a computing unit 301, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) 302 or a computer program loaded from a storage unit 308 into a random access memory (RAM) 303. The RAM 303 may also store various programs and data required for the operation of the electronic device. The computing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0205] Multiple components in the electronic device are connected to I / O interface 305, including: input unit 306, output unit 307, storage unit 308, and communication unit 309. Input unit 306 can be any type of device capable of inputting information into the electronic device. Input unit 306 can receive input digital or character information and generate key signal inputs related to user settings and / or function control of the electronic device. Output unit 307 can be any type of device capable of presenting information and may include, but is not limited to, a display, speaker, video / audio output terminal, vibrator, and / or printer. Storage unit 308 may include, but is not limited to, disks and optical discs. Communication unit 309 allows the electronic device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks, and may include, but is not limited to, modems, network cards, infrared communication devices, and / or wireless communication transceivers, such as Bluetooth devices, WiFi devices, WiMax devices, cellular communication devices, and / or the like.

[0206] The computing unit 301 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 301 include, but are not limited to, CPUs, graphics processing units (GPUs), various special-purpose artificial intelligence (AI) computing units, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. The computing unit 301 performs the various methods and processes described above. For example, in some embodiments, the method embodiments of the present invention can be implemented as computer programs tangibly contained in a machine-readable medium, such as storage unit 308. In some embodiments, part or all of the computer program can be loaded and / or installed on an electronic device via ROM 302 and / or communication unit 309. In some embodiments, the computing unit 301 can be configured to perform the methods described above by any other suitable means (e.g., by means of firmware).

[0207] Computer programs for implementing the methods of embodiments of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0208] In the context of embodiments of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable signal medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, or infrared systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0209] It should be noted that the term "comprising" and its variations used in the embodiments of this invention are open-ended, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The modifications of "one" and "a plurality" mentioned in the embodiments of this invention are illustrative and not restrictive, and those skilled in the art should understand that unless explicitly indicated otherwise in the context, they should be understood as "one or more".

[0210] The steps described in the method embodiments provided by the present invention can be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of protection of the present invention is not limited in this respect.

[0211] The term "embodiment" in this specification refers to a specific feature, structure, or characteristic described in connection with an embodiment that may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily imply the same embodiment, nor does it imply independence or alternativeity from other embodiments. The various embodiments in this specification are described in a related manner, with reference to each other for similar or identical parts. In particular, for apparatus, device, and system embodiments, since they are substantially similar to method embodiments, the description is relatively simple, and relevant details are referred to in the description of the method embodiments.

[0212] The above-described embodiments are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of protection. It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A control method for a vibration isolation device of a reactor, characterized in that, include: Real-time acquisition of the reactor's operating frequency to determine whether the resonant frequency corresponding to the operating frequency is a non-dominant frequency resonance; For non-dominant frequency resonance, vibration signals at the corresponding non-preset operating frequency are collected, and vibration characteristic parameters of the vibration signals are extracted to determine whether the reactor body is working properly. When the reactor body is working normally, the vibration characteristic parameters are used as input excitation based on the multibody dynamics model to perform simulation calculations and predict the simulation response data after vibration isolation. The multibody dynamics model is established based on the calibrated reactor model and the current design parameters of the vibration isolation device. The calibrated reactor model is obtained by comparing and calibrating the reactor's theoretical model with vibration signals under different operating conditions. The predicted simulated response data after vibration isolation is compared with the measured signal after vibration isolation to evaluate the isolation effect of the vibration isolation device. The vibration isolation device is controlled according to the isolation effect to suppress the non-dominant frequency resonance.

2. The method according to claim 1, characterized in that, Controlling the vibration isolation device based on the isolation effect includes: Based on the isolation effect of the vibration isolation device, and based on the multibody dynamics model, the transmissivity-frequency curve of the vibration isolation device is determined as the performance boundary. The performance boundary is compared with the preset vibration isolation threshold to determine whether the standard is met at each frequency point; If the vibration isolation effect at a frequency point is lower than the preset effect threshold, the vibration isolation parameters of the multibody dynamics model are modified, and the optimal parameter combination is found through an optimization algorithm. The vibration isolation device is controlled according to the optimal parameter combination.

3. The method according to claim 2, characterized in that, The performance boundary is compared with the preset vibration isolation threshold to determine whether the standard is met at each frequency point, including: Based on the transfer rate-frequency curve of the performance boundary, determine the transfer rate value corresponding to each frequency point. Based on the stated transmissivity value and the established evaluation index system, the performance index at each frequency point is calculated, wherein the performance index includes at least one of the following: vibration isolation efficiency, vibration attenuation rate, frequency characteristics, and energy transfer ratio. The performance indicators of each frequency point are compared with the corresponding preset vibration isolation threshold. If the performance indicators are better than the corresponding preset vibration isolation threshold, the corresponding frequency point meets the standard. If the performance indicators are worse than the corresponding preset vibration isolation threshold, the corresponding frequency point does not meet the standard.

4. The method according to claim 2, characterized in that, When the vibration isolation effect at a given frequency point is lower than a preset threshold, the vibration isolation parameters of the multibody dynamics model are modified, and the optimal parameter combination is found through an optimization algorithm, including: Obtain response data at the weak frequency point where the vibration isolation effect is lower than a preset effect threshold. The response data includes three parameters: amplitude, phase, and transmissibility. Based on the multibody dynamics model, the sensitivity coefficient of the vibration isolation effect relative to each vibration isolation design parameter is calculated; Based on the sensitivity coefficient, determine the vibration isolation design parameters that have the greatest impact on the vibration isolation effect; The optimal combination of vibration isolation design parameters is searched using a parameter optimization algorithm.

5. The method according to claim 2, characterized in that, Controlling the vibration isolation device according to the optimal parameter combination includes: The system collects real-time data on the grid frequency and load parameter changes of the power grid where the reactor is located, and monitors the frequency fluctuations of the power grid in real time. Based on the optimal parameter combination, and considering the changes in grid frequency and load parameters, an adaptive control algorithm is used to calculate the optimal control parameters for the controllable parameters of the vibration isolation device. When the frequency fluctuation of the power grid where the reactor is located is detected to exceed the preset threshold, the controllable parameters of the vibration isolation device are adjusted to the corresponding optimal control parameters in order to control the vibration isolation device.

6. The method according to claim 1, characterized in that, Real-time acquisition of the reactor's operating frequency to determine whether the resonant frequency corresponding to the operating frequency is a non-dominant frequency resonance, including: Real-time acquisition of the reactor's operating current and operating voltage signals; The real-time acquired operating current and operating voltage signals are preprocessed, wherein the preprocessing includes at least one of the following: cleaning, noise reduction, and outlier detection; The preprocessed operating current signal and operating voltage signal are analyzed in the frequency domain using Fast Fourier Transform to obtain the amplitude and phase information of each resonant frequency, and the real-time operating frequency and each resonant frequency are identified. Based on amplitude and phase information, determine whether each resonant frequency is a non-dominant frequency resonance.

7. The method according to claim 1, characterized in that, For non-dominant frequency resonance, vibration signals are collected at the corresponding non-preset operating frequency, and vibration characteristic parameters of the vibration signals are extracted to determine whether the reactor body is working properly, including: In the presence of non-dominant frequency resonance, the vibration signal of the reactor at a non-preset operating frequency is collected, wherein the non-preset operating frequency is the operating frequency corresponding to the non-dominant frequency resonance. After denoising and filtering preprocessing, the vibration signal is analyzed using time-frequency analysis to extract vibration characteristic parameters, wherein the vibration characteristic parameters include at least one of the following: frequency, amplitude, and energy distribution. By inputting the vibration characteristic parameters into a trained machine learning model, the working state of the reactor body can be determined. The machine learning model is trained based on multiple sets of training data, each set of training data including vibration characteristic parameters and whether the corresponding reactor body is malfunctioning.

8. The method according to claim 1, characterized in that, For non-dominant frequency resonance, assuming the reactor body is operating normally, before performing simulation calculations based on a multibody dynamics model, using the vibration characteristic parameters as input excitation, and predicting the simulation response data after vibration isolation, the method further includes: Collect vibration signals of the reactor at different operating frequencies; Create a theoretical model of the reactor and calculate the theoretical natural frequencies and theoretical mode shapes of the reactor at different operating frequencies; The theoretical natural frequency and theoretical mode shape are compared with the frequency and mode shape of the measured vibration signal, and the model parameters of the theoretical model are iteratively optimized to calibrate the reactor model. Based on the calibrated reactor model and the current design parameters of the vibration isolation device, a multibody dynamics model of the vibration isolation device is established.

9. The method according to claim 1, characterized in that, The method further includes: The reactor's operating frequency is collected in real time. If the resonant frequency corresponding to the operating frequency is determined to be the main frequency resonance, a fault alarm signal for the reactor body is generated and sent to the relevant terminal. The rule-based expert system comprehensively analyzes the operating parameters of the reactor and the vibration spectrum features extracted by the time-frequency analysis module, and matches them with the fault rules in the knowledge base to determine the fault type of the reactor body. Based on the fault type, corresponding maintenance decision suggestions are generated and sent to the relevant terminal.

10. An electronic device, comprising: A processor and a memory storing a program, characterized in that the program includes instructions that, when executed by the processor, cause the processor to perform the method according to any one of claims 1 to 9.

Images