Surge suppression method and device for centrifugal air compressor

Abstract

The invention relates to the technical field of air compressors, and discloses a centrifugal air compressor surge suppression method and device.The method comprises the steps that pressure pulsation feature extraction is conducted on a pressure signal of a centrifugal air compressor, and pressure pulsation features are obtained; extracting a flow statistical feature and a frequency band energy feature of the flow signal to obtain a surge early warning feature space; carrying out characteristic decomposition on the bearing vibration signal, and extracting a characteristic frequency to obtain a surge early warning characteristic set; inputting the surge early warning feature set into a three-channel neural network for surge early warning prediction to obtain a surge early warning level; according to the surge early warning grade, control parameters of a guide vane angle controller, an anti-surge valve opening controller and a return valve opening controller are generated, the guide vane angle, the anti-surge valve opening and the return valve opening are synchronously adjusted, and a surge margin value is calculated. According to the method, accurate control over the guide vane angle, the anti-surge valve opening and the return valve opening is achieved, the surge margin value is calculated, and the control accuracy of the guide vane angle, the anti-surge valve opening and the return valve opening is improved. And real-time optimization and adjustment of the control strategy are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of air compressors, and in particular to a method and device for suppressing surge of a centrifugal air compressor. Background Art

[0002] As a key equipment in industrial production, the operation stability of centrifugal air compressors directly affects production efficiency and safety. In actual operation, due to the frequent changes in operating conditions, the internal flow field structure of the air compressor is complex and changeable, and surge is prone to occur. At present, most surge warning methods mainly rely on a single pressure or flow signal for analysis, which makes it difficult to accurately capture the weak signs before surge occurs, resulting in insufficient warning accuracy and affecting the surge suppression effect.

[0003] Traditional surge suppression methods usually use a fixed threshold control strategy, which cannot be dynamically adjusted according to the real-time operating status of the air compressor. This control method is prone to response lag or over-adjustment when facing rapid changes in parameters such as speed and pressure ratio. At the same time, due to the lack of collaborative analysis capabilities for multi-source signals, existing methods are difficult to accurately determine the critical state of surge occurrence, resulting in insufficient optimization of control parameters and affecting the operating efficiency of the air compressor. Summary of the invention

[0004] The present invention provides a method and device for suppressing surge of a centrifugal air compressor, which realizes accurate control of the guide vane angle, the opening of an anti-surge valve and the opening of a return valve, calculates the surge margin value, and ensures real-time optimization and adjustment of the control strategy.

[0005] In a first aspect, the present invention provides a method for suppressing surge of a centrifugal air compressor, the method comprising: Extract the pressure pulsation characteristics of the pressure signal of the centrifugal air compressor to obtain the pressure pulsation characteristics; Extracting flow statistics characteristics and frequency band energy characteristics of the flow signal in the centrifugal air compressor, and combining them with the pressure pulsation characteristics to obtain a surge warning feature space; Performing feature decomposition on the bearing vibration signal of the centrifugal air compressor, extracting the characteristic frequency, and combining it with the surge warning feature space to obtain a surge warning feature set; Inputting the surge warning feature set into a three-channel neural network to perform surge warning prediction to obtain a surge warning level; The control parameters of the guide vane angle controller, the anti-surge valve opening controller and the return valve opening controller are generated according to the surge warning level, and the guide vane angle, the anti-surge valve opening and the return valve opening are adjusted synchronously to calculate the surge margin value.

[0006] In a second aspect, the present invention provides a centrifugal air compressor surge suppression device, the centrifugal air compressor surge suppression device comprising: A feature extraction module is used to extract pressure pulsation features from the pressure signal of the centrifugal air compressor to obtain pressure pulsation features; A feature combination module, used to extract the flow statistical features and frequency band energy features of the flow signal in the centrifugal air compressor, and combine them with the pressure pulsation features to obtain a surge warning feature space; A feature decomposition module, used for performing feature decomposition on the bearing vibration signal of the centrifugal air compressor, extracting the feature frequency, and combining it with the surge warning feature space to obtain a surge warning feature set; An early warning prediction module is used to input the surge early warning feature set into a three-channel neural network to perform surge early warning prediction and obtain a surge early warning level; The calculation module is used to generate control parameters of the guide vane angle controller, the anti-surge valve opening controller and the return valve opening controller according to the surge warning level, and synchronously adjust the guide vane angle, the anti-surge valve opening and the return valve opening to calculate the surge margin value.

[0007] In the technical solution provided by the present invention, by performing multi-scale analysis of wavelet transform and Fourier transform on the pressure signal, combined with the principal component analysis method for dimensionality reduction processing, the time-frequency characteristics of pressure pulsation are effectively extracted, redundant information is reduced, and the efficiency of feature extraction is improved. The flow signal is processed by combining Butterworth filtering and wavelet packet transform. By extracting statistical features and frequency band energy features, the dynamic characteristics of flow fluctuations are accurately captured, and the reliability of surge warning is enhanced. Empirical mode decomposition and Hilbert envelope spectrum analysis are introduced to process bearing vibration signals. Combined with envelope demodulation technology, the accurate extraction of vibration characteristics of various bearing components is achieved, and the information dimension of surge warning is expanded. A three-channel parallel deep neural network structure is designed. Different types of features are processed respectively by convolutional layers, fully connected layers and LSTM networks, and feature fusion is performed by attention mechanism to construct an accurate surge warning model. A multi-parameter collaborative optimization method based on fuzzy rules and genetic algorithms is proposed. Through dynamic weight update and objective function optimization, the precise control of guide vane angle, anti-surge valve opening and return valve opening is achieved. A complete surge suppression effect evaluation system has been established. By monitoring the pressure ratio change rate, flow fluctuation amplitude and bearing vibration intensity, the surge margin value is calculated to ensure real-time optimization and adjustment of the control strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0009] Figure 1 A schematic diagram of the steps of a method for suppressing surge of a centrifugal air compressor according to an embodiment of the present invention; Figure 2 Schematic diagram of the structure of a centrifugal air compressor surge suppression device in an embodiment of the present invention. DETAILED DESCRIPTION

[0010] An embodiment of the present invention provides a method and device for suppressing surge of a centrifugal air compressor. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or apparatus that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or apparatus.

[0011] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 , an embodiment of a method for suppressing surge of a centrifugal air compressor in an embodiment of the present invention includes: Step S1, extracting pressure pulsation characteristics of the pressure signal of the centrifugal air compressor to obtain pressure pulsation characteristics; It is understandable that the execution subject of the present invention may be a centrifugal air compressor surge suppression device, or a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.

[0012] Specifically, the inlet pressure signal, outlet pressure signal and intermediate pressure signal of the centrifugal air compressor are collected to construct the original pressure signal sequence. These pressure signals are derived from the real-time measurement of the sensor and are acquired at a high sampling rate. Due to the complexity of fluid dynamics, these signals contain rich transient and periodic components and are affected by turbulent pulsation, pipeline resonance and changes in operating conditions. The original pressure signal sequence is input into the db4 wavelet basis function for 4-layer wavelet decomposition, and the local characteristics of the signal are analyzed at multiple scales. Wavelet transform has good time-frequency localization characteristics and can simultaneously analyze the time domain and frequency domain information of the signal. The high-frequency detail components of the signal are extracted in the first layer, which contains high-frequency pulse noise or rapidly changing information, while the medium and low-frequency components of the signal are gradually extracted from the second to fourth layers to capture the fluid instability characteristics before surge occurs. The multi-scale characteristics of the pressure signal are obtained by decomposition. The multi-scale characteristics are discrete Fourier transformed to obtain the frequency domain feature sequence of the signal. The discrete Fourier transform converts the time series signal into the expression of frequency components and identifies the characteristic frequencies related to surge. During the calculation process, Fourier transform is performed on wavelet components of different scales. Statistical analysis is performed on the original pressure signal sequence to extract important parameters for evaluating the dynamic characteristics of the signal, including indicators such as root mean square value, peak factor, waveform factor and pulse factor. The root mean square value measures the overall change of signal energy, the peak factor is used to evaluate the extreme value characteristics of the signal, the waveform factor reflects the complexity of the signal waveform, and the pulse factor characterizes the mutation phenomenon in the signal. These statistical features are combined with the frequency domain features to form the pressure pulsation feature vector. Principal component analysis is performed on the pressure pulsation feature vector to reduce the data dimension and remove redundant information. The data variance is maximized by constructing a new orthogonal coordinate system, thereby reducing the computational complexity while maintaining key information. Specifically, the covariance matrix of the eigenvector is calculated, and its eigenvalues ​​and eigenvectors are solved, and then the principal component with the largest variance information is selected to finally obtain the pressure pulsation feature after dimensionality reduction.

[0013] Step S2, extracting flow statistical characteristics and frequency band energy characteristics of the flow signal in the centrifugal air compressor, and combining them with the pressure pulsation characteristics to obtain a surge warning feature space; Specifically, the flow signals collected by the inlet flow sensor and the outlet flow sensor are sampled to obtain the original flow signal. The original flow signal is subjected to signal denoising, and the original flow signal is processed by wavelet denoising, adaptive filtering or empirical mode decomposition to remove high-frequency noise while retaining key surge characteristic information. After filtering, a relatively stable flow signal is obtained. The filtered flow signal is statistically analyzed to extract key statistical features that characterize the flow change characteristics. The statistical analysis mainly involves mean, standard deviation, skewness and kurtosis, where the mean reflects the overall level of flow, the standard deviation measures the amplitude of flow fluctuation, the skewness describes the symmetry of flow distribution, and the kurtosis is used to identify the sharpness of the flow signal. Usually, when surge is about to occur, the standard deviation of the flow signal will increase abnormally, and the changes in skewness and kurtosis will also show a specific trend, so these statistical features play an important role in surge warning. The filtered flow signal is decomposed by wavelet packet to obtain its frequency band coefficient matrix. Wavelet packet decomposition is an extended version of wavelet transform, which can decompose the full frequency band of the signal and provide finer frequency resolution than traditional wavelet transform. Wavelet packet decomposition decomposes the flow signal layer by layer into sub-signals of different frequency bands. Each sub-signal corresponds to the information of a different frequency band, forming a frequency band coefficient matrix. The energy eigenvalue is calculated for the frequency band coefficient matrix to extract the energy distribution of each frequency band. The calculation method of the energy feature is based on the square integral of the signal, that is, the signal of each frequency band component is squared and summed to obtain the energy size of the frequency band. By calculating the energy eigenvalues ​​of all sub-bands, the frequency band energy eigenvector is obtained, which reflects the energy distribution of the flow signal in different frequency ranges. The flow statistical features, frequency band energy features and pressure pulsation features are feature spliced ​​to construct the surge warning feature space.

[0014] Step S3, performing feature decomposition on the bearing vibration signal of the centrifugal air compressor, extracting the characteristic frequency, and combining it with the surge warning feature space to obtain a surge warning feature set; Specifically, a high-precision acceleration sensor is installed at the inlet and outlet bearings of the centrifugal air compressor, and the bearing vibration signal is collected at a sufficiently high sampling rate to obtain the original vibration signal sequence. The original vibration signal is subjected to empirical mode decomposition, and the non-stationary signal is decomposed into a series of intrinsic mode functions. These modal components reflect the vibration characteristics at different scales, and the component matrix of the vibration signal is obtained. The Hilbert transform is applied to each intrinsic mode function component of the vibration signal component matrix, and its instantaneous frequency and envelope spectrum are calculated to obtain the spectrum feature sequence. The characteristic frequency is extracted from the spectrum feature sequence, and the key frequency parameters of the bearing are calculated, including the inner ring passing frequency, the outer ring passing frequency, the rolling element passing frequency, and the cage rotation frequency. These characteristic frequencies are determined by the structural parameters of the bearing and are related to the damage state inside the bearing. When surge occurs, the stress state of the bearing will change, resulting in abnormal energy distribution of these characteristic frequencies. By calculating these bearing characteristic frequencies, abnormal bearing vibration caused by surge can be effectively identified. The bearing characteristic frequencies are subjected to envelope demodulation analysis to extract the vibration characteristics of various bearing components. Envelope demodulation uses bandpass filtering and Hilbert transform to extract low-frequency modulation information from high-frequency signals. This method is suitable for detecting bearing fault characteristics. By demodulating the envelope signal of the bearing characteristic frequency, the vibration characteristics of each component, such as the vibration mode of the inner ring, outer ring, rolling element and cage, are obtained, and their time domain and frequency domain characteristics are calculated to form a bearing vibration feature vector, which reflects the health status of the bearing. When surge occurs, the instability of the airflow will cause the bearing load to change, thereby causing the enhancement or attenuation of a specific vibration frequency. The bearing vibration feature vector is combined with the surge warning feature space to fuse the multimodal information of the pressure signal, flow signal and bearing vibration signal to form a high-dimensional feature set. The feature combination method uses feature splicing or data fusion technology to arrange the features of different signal sources in a certain order, and optimize the data structure through dimensionality reduction or feature selection methods to ensure the effectiveness and computational efficiency of the features, and obtain the surge warning feature set.

[0015] Step S4, inputting the surge warning feature set into the three-channel neural network to perform surge warning prediction to obtain a surge warning level; Specifically, the surge warning feature set is input into a three-channel neural network, and the corresponding features are extracted in three different channels respectively, so as to make full use of the characteristics of different signals and improve the accuracy and robustness of surge prediction. In the entire network architecture, the pressure pulsation features in the surge warning feature set are input into the first channel of the three-channel neural network, and the features are extracted layer by layer through four convolution layers in this channel to obtain the deep pattern information of the pressure signal. During the convolution operation, each layer applies multiple different convolution kernels to extract local features, and introduces nonlinear mapping capabilities through the ReLU activation function, so that the model can learn more complex feature representations. The maximum pooling operation is added after each convolution layer to reduce the size of the feature map, improve computational efficiency, and enhance the translation invariance of the model. After these four layers of convolution processing, the pressure pulsation features are converted into high-dimensional pressure feature maps, reflecting the pattern changes of pressure pulsation at different time scales, and effectively capturing the abnormal trends that lead to surge. At the same time, the flow statistics and frequency band energy features in the surge warning feature set are input into the second channel of the three-channel neural network, and the features are transformed through the three fully connected layers in the channel in turn to obtain a more compact and recognizable flow feature representation. In the first fully connected layer, the number of neurons is large to ensure that the distribution information of the original flow data is fully captured, and after ReLU activation, nonlinear transformation is introduced to enhance the modeling ability of complex flow patterns. The second fully connected layer performs dimensionality reduction processing on the features, so that the model maintains the key information of the flow signal in a lower dimensional space and eliminates redundant features. In the third fully connected layer, the network further transforms the reduced-dimensional flow features to ensure that the final flow feature representation is obtained, which effectively describes the time series changes of the flow and the abnormal pattern of the flow energy distribution when surge occurs. At the same time, the bearing vibration feature vector in the surge warning feature set is input into the third channel of the three-channel neural network, and the time series feature extraction is performed through two layers of LSTM (long short-term memory network) in this channel to obtain the time series characteristics of the bearing vibration signal. In the first layer of LSTM network, the model learns the short-term dependency features of the vibration signal and captures the local change trend, while in the second layer of LSTM network, the model integrates long-term dependency information to extract the time series evolution law of the bearing vibration features before surge occurs. After two layers of LSTM processing, the obtained frequency feature sequence fully characterizes the changes in the bearing vibration mode during surge. The attention weights are calculated for the pressure feature mapping, flow feature representation, and frequency feature sequence respectively, and a feature weight matrix is ​​generated. When calculating the attention weight, each feature is adaptively weighted, that is, the contribution of each feature is dynamically adjusted through a learnable parameter matrix, so that features with greater predictive value obtain higher weights, while features with more noise or higher redundancy are weakened.According to the feature weight matrix, the pressure feature map, flow feature representation and frequency feature sequence are weighted fused to form a fused feature vector. During the fusion process, the attention mechanism assigns different weights to different features to ensure that key features occupy a larger proportion in the final prediction while avoiding the degradation of model performance due to information redundancy. The fused feature vector is input into the output layer of the three-channel neural network, and the probability distribution of different surge warning levels is calculated by the softmax classifier. The softmax classifier maps the fused feature vector to three different surge levels and calculates the corresponding probability value for each level. Finally, the level corresponding to the maximum probability value is selected as the final surge warning level.

[0016] Step S5: generating control parameters of the guide vane angle controller, the anti-surge valve opening controller and the return valve opening controller according to the surge warning level, and synchronously adjusting the guide vane angle, the anti-surge valve opening and the return valve opening to calculate the surge margin value.

[0017] Specifically, the control range of the guide vane angle controller, the control range of the anti-surge valve opening controller, and the control range of the return valve opening controller are discretized so that they can be optimized and calculated on a limited parameter set. In the discretization process, the controller range is divided into multiple discrete levels to form a control parameter space. A fuzzy rule base is constructed based on the surge warning level so that fuzzy reasoning operations can be performed according to different surge warning states. Fuzzy control uses language rules to describe the relationship between input and output. Therefore, a series of fuzzy rules are set according to empirical data and surge characteristics, such as "if the surge level is low, the guide vane angle is adjusted appropriately, while the anti-surge valve and return valve maintain the minimum adjustment amount", "if the surge level is high, the anti-surge valve opening is increased rapidly, and the guide vane angle and return valve opening are adjusted appropriately". The discretized controller parameter space is input into the fuzzy rule base, and the fuzzy reasoning method is used for calculation to generate a fuzzy control decision sequence, which adjusts the control parameters according to different surge levels to achieve dynamic regulation of surge. The optimization objective function is set according to the fuzzy control decision sequence so that the control strategy achieves a reasonable balance between surge suppression, energy loss and adjustment time. In the optimization objective function, the surge suppression effect is set as the first weight to ensure that the primary goal of the control strategy is to effectively prevent surge. At the same time, the energy loss is set as the second weight to avoid excessive adjustment of control parameters leading to increased system energy consumption, and the adjustment time is set as the third weight to ensure that the response speed of the control strategy is fast enough to complete effective adjustment before surge occurs. In the process of constructing the objective function, different weights are adjusted according to the operating conditions. For example, when the surge risk is high, the weight of the surge suppression effect is increased, while under stable conditions, the weight of energy loss optimization is appropriately increased to improve system efficiency. Based on the optimization objective function, a genetic algorithm population is constructed to intelligently optimize the control parameters. Chromosome encoding is generated for the control parameters, in which the discretized parameters of the guide vane angle, anti-surge valve opening and return valve opening are converted into binary sequences, and the initial population is formed in the process of binary encoding mapping. Each individual represents a possible combination of control parameters, and the genetic algorithm continuously evolves these individuals to find the optimal control strategy. In the genetic calculation process, a selection operation is performed to select the better individuals to enter the next generation according to the fitness value of the objective function; then a crossover operation is performed to generate new candidate solutions by randomly crossing the gene fragments of different individuals; then a mutation operation is performed to introduce more search possibilities by randomly changing the gene sites of some individuals, thereby avoiding the algorithm from falling into a local optimum. After several generations of evolutionary calculations, the optimal chromosome is obtained, which corresponds to a set of optimal control parameters for the guide vane angle, anti-surge valve opening, and return valve opening. The optimal chromosome is decoded to convert the binary code into the corresponding physical control parameters to obtain the guide vane angle, anti-surge valve opening, and return valve opening control parameters.The adjustment of the guide vane angle is used to control the air flow entering the air compressor, thereby affecting the fluid dynamic characteristics of the system, while the adjustment of the anti-surge valve opening is used to quickly release excess pressure to prevent the occurrence of surge. The adjustment of the return valve opening realizes the return flow control of the flow, thereby optimizing the surge suppression effect. Based on the control parameters, the guide vane angle, anti-surge valve opening and return valve opening are synchronously adjusted to monitor the changes in pressure pulsation characteristics, flow statistics and characteristic frequencies, and calculate the surge margin value. The surge margin is a key indicator for measuring the surge risk of the system. It is defined as the relative distance between the surge limit and the current operating state. By calculating the surge margin value, it is determined whether the current control strategy is effective and whether the adjustment parameters need to be further optimized. If the surge margin is high, it means that the current operating condition is in a stable state, and the control parameters are appropriately adjusted to reduce unnecessary energy loss; if the surge margin is low, it means that the system is still in the surge risk area, and the control strategy needs to be further optimized to ensure that the surge is completely suppressed.

[0018] The control parameters are input into the servo motor drive system and the electric hydraulic servo system to synchronously adjust the guide vane angle, anti-surge valve opening and return valve opening. The control system receives the optimized control parameters and drives each actuator so that the guide vanes, valves and other components perform corresponding operations according to the set adjustment range to form an actuator action sequence. The response time of the actuator action sequence is detected to evaluate the adjustment speed and dynamic response characteristics of the guide vane angle, anti-surge valve opening and return valve opening. In this process, the guide vane angle adjustment response time, anti-surge valve opening adjustment response time and return valve opening adjustment response time are recorded respectively, and these data are stored in the database of the control system, and dynamic feedback adjustment is performed to optimize the control strategy of the actuator and improve the response capability of the overall system. The pressure ratio change rate, flow fluctuation amplitude and bearing vibration intensity are calculated based on the response time data to quantitatively evaluate the surge suppression effect and form a performance parameter matrix. The pressure ratio change rate is used to characterize the change trend of the compression ratio during the adjustment process of the system. If the pressure ratio change rate is too large, it means that the system is still in an unstable state and there is a risk of surge. The flow fluctuation amplitude is an important parameter to measure the flow adjustment effect. If the flow fluctuation is too large, it means that the surge has not been completely suppressed, or the opening adjustment of the anti-surge valve and the return valve has not reached the optimal state. The bearing vibration intensity reflects the impact of surge suppression on mechanical vibration. If the bearing vibration is still large after adjustment, the control strategy needs to be further optimized. By calculating these key parameters and integrating them into a performance parameter matrix, the dynamic behavior of the system during surge suppression is evaluated. The performance parameter matrix is ​​subjected to stability analysis to calculate the overshoot, stabilization time, and steady-state error. Among them, the overshoot is used to measure whether there is an overshoot phenomenon in the control strategy during the adjustment process. Excessive overshoot causes the surge suppression strategy to have a reverse effect, causing the system to enter the unstable zone. Therefore, it is necessary to optimize the control strategy as much as possible to reduce the overshoot; the stabilization time indicates the time required for the system to enter the stable state from the start of adjustment. A short stabilization time means that the system adjusts faster, which helps to improve the real-time performance of surge suppression; and the steady-state error measures whether the final surge suppression effect meets the accuracy requirements. A large steady-state error will lead to insufficient surge margin, thereby affecting the long-term stable operation of the air compressor. In the process of stability analysis, the effectiveness of the surge suppression strategy is ensured by calculating and optimizing these operating status indicators. Based on the operating status indicators obtained by the stability analysis, the surge margin is calculated to quantify the current surge safety margin of the system. The calculation method of the surge margin adopts a weighted summation model, in which the pressure ratio change rate is multiplied by the first target value, the flow fluctuation amplitude is multiplied by the second target value, and the bearing vibration intensity is multiplied by the third target value, and then these weighted results are summed to obtain the final surge margin value.

[0019] In the embodiment of the present invention, by performing multi-scale analysis of wavelet transform and Fourier transform on the pressure signal, combined with the principal component analysis method for dimensionality reduction processing, the time-frequency characteristics of pressure pulsation are effectively extracted, redundant information is reduced, and the efficiency of feature extraction is improved. The flow signal is processed by combining Butterworth filtering and wavelet packet transform. By extracting statistical features and band energy features, the dynamic characteristics of flow fluctuations are accurately captured, and the reliability of surge warning is enhanced. Empirical mode decomposition and Hilbert envelope spectrum analysis are introduced to process bearing vibration signals. Combined with envelope demodulation technology, the accurate extraction of vibration characteristics of various bearing components is achieved, and the information dimension of surge warning is expanded. A three-channel parallel deep neural network structure is designed. Different types of features are processed respectively by convolutional layers, fully connected layers and LSTM networks, and feature fusion is performed by attention mechanism to construct an accurate surge warning model. A multi-parameter collaborative optimization method based on fuzzy rules and genetic algorithms is proposed. Through dynamic weight update and objective function optimization, the precise control of guide vane angle, anti-surge valve opening and return valve opening is achieved. A complete surge suppression effect evaluation system has been established. By monitoring the pressure ratio change rate, flow fluctuation amplitude and bearing vibration intensity, the surge margin value is calculated to ensure real-time optimization and adjustment of the control strategy.

[0020] In a specific embodiment, the process of executing step S1 may specifically include the following steps: The inlet pressure signal, outlet pressure signal and intermediate pressure signal of the centrifugal air compressor are collected to obtain an original pressure signal sequence; The original pressure signal sequence is input into the db4 wavelet basis function for 4-layer wavelet decomposition to obtain the multi-scale characteristics of the pressure signal; Perform discrete Fourier transform on the multi-scale features of the pressure signal to obtain a frequency domain feature sequence; Perform statistical analysis on the original pressure signal sequence, calculate the root mean square value, peak factor, waveform factor and pulse factor, and combine them with the frequency domain feature sequence to obtain the pressure pulsation feature vector; The principal component analysis is performed on the pressure pulsation feature vector to obtain the pressure pulsation characteristics.

[0021] Specifically, during the operation of a centrifugal air compressor, the pressure state of the internal air flow will be affected by factors such as impeller rotation, air flow disturbance, and surge. Pressure sensors are arranged at different locations to collect pressure signals, among which the inlet pressure signal , outlet pressure signal and intermediate pressure signal The complete time series is obtained by high-frequency sampling, and these signals are combined into the original pressure signal sequence: in, represents the pressure signal sequence that varies with time, is the time variable, , and Represent the pressure signals of the inlet, outlet and intermediate level respectively. The original pressure signal sequence is input into the db4 wavelet basis function for 4-layer wavelet decomposition. Using the four-layer wavelet decomposition, the signal is analyzed at multiple scales to obtain information of different frequency bands. The mathematical expression of wavelet transform is as follows: in, is the wavelet coefficient, which indicates that the signal is at different scales and time shift The amount of is the wavelet basis function. The four-layer decomposition process of the db4 wavelet basis is expressed as: in, represents the low-frequency component of layer 4, Represent the high-frequency detail components of each layer respectively. Multi-scale features can effectively separate low-frequency trends and high-frequency disturbances in the signal. The multi-scale features obtained by wavelet decomposition are subjected to discrete Fourier transform to obtain the frequency domain feature sequence. The basic expression of Fourier transform is: in, In the frequency domain, frequency components, is the signal length, is an imaginary unit, is the original signal Sampling points. Through Fourier transform, the amplitude spectrum of the signal is obtained: in, and are the real and imaginary parts of the Fourier transform results respectively. It reflects the energy distribution of the signal at different frequencies, thereby revealing the main oscillation mode of the pressure signal. While obtaining the frequency domain characteristics, the original pressure signal is statistically analyzed to calculate the root mean square value (RMS), peak factor, waveform factor and pulse factor, which are used to characterize the overall characteristics of the pressure signal. The root mean square value is defined as: in, Reflects the energy level of the signal. The signal is Sample values, is the total number of samples. The crest factor is used to measure the ratio of the maximum amplitude of the signal to the root mean square value, which is defined as: in, is the maximum peak value of the signal. The waveform factor measures the waveform complexity of the signal, and the calculation formula is as follows: The impulse factor reflects the degree of spikes in the signal and is defined as: By calculating the above statistical features and combining them with the frequency domain features obtained by Fourier transform, the pressure pulsation feature vector is constructed: in, As the final high-dimensional feature vector, it contains key information in the time domain and frequency domain. Perform principal component analysis on the pressure pulsation feature vector to reduce the dimension and extract the main feature components. Principal component analysis calculates eigenvalues ​​and eigenvectors through the covariance matrix and selects the principal component with the largest variance contribution. Calculate the covariance matrix: in, is the mean of the eigenvectors. Solve for the eigenvalues ​​of the covariance matrix and the eigenvector , and select the previous The principal component corresponding to the largest eigenvalue is: in, It is the pressure pulsation characteristic after dimensionality reduction.

[0022] In a specific embodiment, the process of executing step S2 may specifically include the following steps: Sampling the flow signals collected by the inlet flow sensor and the outlet flow sensor to obtain the original flow signal, and performing signal noise reduction on the original flow signal to obtain the filtered flow signal; Perform statistical analysis on the filtered flow signal, calculate the mean, standard deviation, skewness and kurtosis, obtain the flow statistical characteristics, and perform wavelet packet decomposition on the filtered flow signal to obtain the frequency band coefficient matrix; Calculate the energy eigenvalue of the frequency band coefficient matrix, extract the energy distribution of each frequency band, and obtain the frequency band energy characteristics; The flow statistical features, frequency band energy features and pressure pulsation features are spliced ​​to obtain the surge warning feature space.

[0023] Specifically, the inlet flow signal is obtained through a high-precision flow sensor and outlet flow signal , and these traffic signals are combined into the original traffic signal sequence: in, A signal sequence representing the flow rate variation over time, is the time variable, and and Represent the instantaneous flow values ​​of the inlet and outlet respectively. Perform signal denoising on the original flow signal to obtain a smoother and more effective flow signal. Signal denoising methods include wavelet threshold denoising, empirical mode decomposition denoising and adaptive filtering. Among them, wavelet threshold denoising can better retain the mutation characteristics of the flow signal and remove high-frequency noise. Perform wavelet transform on the signal to decompose it into detail components of different scales and low frequency components : in, represents the low-frequency component of the fourth layer of wavelet, and They are the wavelet detail components of each layer. Soft threshold processing is used on the detail components to remove high-frequency noise and reconstruct the signal to obtain the filtered flow signal. . Perform statistical analysis on the filtered flow signal to calculate the mean, standard deviation, skewness and kurtosis of the flow and form the flow statistical characteristics. The mean of the flow signal Used to measure the average flow of the system, the calculation formula is: in, is the total number of flow data points, The first data points. Standard deviation It reflects the fluctuation degree of the flow signal. The calculation formula is as follows: The skewness It reflects the symmetry of traffic distribution and is defined as follows: like , indicating that the flow signal is biased to the right, that is, there are more high flow values, and if , it means that the flow is biased to the left, that is, there are more low flow values. Measures the sharpness of the flow signal, and the calculation formula is: like If it is much larger than 3, it means that the traffic signal has a peak feature and there is a mutation. After obtaining the traffic statistical characteristics, the traffic signal is decomposed by wavelet packet to extract the multi-band information of the traffic signal. Wavelet packet decomposition not only decomposes the low-frequency part, but also decomposes the high-frequency part to obtain a more refined time-frequency representation. For the filtered traffic signal , its wavelet packet decomposition process is expressed as: in, is the wavelet packet coefficient, indicating the signal at scale and time shift The energy distribution at is the wavelet packet basis function. After decomposing the 4-layer wavelet packet, the frequency band coefficient matrix is ​​obtained, where the coefficient of each frequency band reflects the characteristics of the traffic signal at different scales. The frequency band energy characteristics are calculated, that is, the energy of the wavelet packet coefficient matrix is ​​calculated, and the formula is as follows: in, Representative The energy of the layer-by-layer decomposition signal, is the corresponding decomposition coefficient. The energy values ​​of all frequency bands constitute the frequency band energy eigenvector: in, is the total number of decomposition layers. The flow statistics feature, frequency band energy feature and pressure pulsation feature are combined to obtain the surge warning feature space: in, Represents the pressure pulsation characteristics, It is the final surge warning feature vector.

[0024] In a specific embodiment, the process of executing step S3 may specifically include the following steps: The bearing vibration signal collected by the acceleration sensor at the inlet and outlet bearings of the centrifugal air compressor is sampled to obtain the original vibration signal; Performing empirical mode decomposition on the original vibration signal to obtain a vibration signal component matrix, and performing Hilbert transform on the vibration signal component matrix to obtain a frequency spectrum feature sequence; Extract characteristic frequencies from the frequency spectrum feature sequence, calculate the inner ring passing frequency, outer ring passing frequency, rolling element passing frequency and cage rotation frequency, and obtain the bearing characteristic frequency; The bearing characteristic frequency is subjected to envelope demodulation analysis to extract the vibration characteristics of each bearing component to obtain the bearing vibration characteristic vector. The bearing vibration characteristic vector is then combined with the surge warning feature space to obtain the surge warning feature set.

[0025] Specifically, high-precision acceleration sensors are arranged at the inlet and outlet bearings to collect the original vibration signals of the bearings: in, represents the acceleration signal sequence that changes with time, and represent the vibration signals of the inlet and outlet bearings respectively, and is a time variable. Empirical mode decomposition is used to decompose the signal and extract vibration components of different modes. Empirical mode decomposition decomposes the original signal by iterative , decompose it into multiple intrinsic mode functions (IMFs), each IMF represents a different frequency component of the signal. The decomposition process is expressed as: in, Indicates The intrinsic mode components, is the total number of decomposition layers, is the residual signal. In order to ensure that the decomposed IMF components have physical meaning, the first few IMF components are selected as the main vibration characteristic components, and the residual components are ignored. In this way, the original vibration signal is decomposed into multiple frequency components. The IMF component is Hilbert transformed to obtain the instantaneous frequency characteristics of the vibration signal. The definition of Hilbert transform is as follows: The time-frequency representation of the signal is obtained by calculating the instantaneous amplitude and instantaneous frequency, where the instantaneous amplitude is calculated by the following formula: The instantaneous frequency is calculated as follows: in, represents the instantaneous envelope of the vibration signal, Represents the instantaneous frequency. These features help identify the vibration mode of the bearing and analyze its dynamic changes when surge occurs. The characteristic frequency is extracted from the spectrum feature sequence to calculate the inner ring passing frequency (BPFI), outer ring passing frequency (BPFO), rolling element passing frequency (BSF) and cage rotation frequency (FTF) of the bearing. The calculation formulas for these frequencies are as follows: in, is the number of rolling elements, is the bearing inner ring diameter, is the rolling element diameter, is the contact angle, is the rotation frequency. By calculating these characteristic frequencies, the health status of the bearing is identified and the abnormal vibration mode of the bearing caused by surge is detected. After obtaining the characteristic frequencies, envelope demodulation analysis is performed to extract the vibration characteristics of each bearing component. Envelope demodulation uses bandpass filtering and Hilbert transform to extract the low-frequency modulation information of the bearing vibration signal. Bandpass filtering is performed on the vibration signal, and the frequency band range is selected to be , and then perform Hilbert transform on the filtered signal to obtain the envelope signal: in, is the filtered signal, is the envelope signal. By calculating the spectral characteristics of the envelope signal, the main vibration mode of the bearing is extracted and the bearing vibration feature vector is formed: in, Indicates the vibration energy of the envelope signal at different frequencies. In order to construct the surge warning feature set, the bearing vibration feature vector and the surge warning feature space are concatenated to form a warning feature vector: in, Indicates the pressure pulsation characteristics, Represents traffic statistics and frequency band energy characteristics, is the bearing vibration eigenvector, It is the final surge warning feature vector.

[0026] In a specific embodiment, the process of executing step S4 may specifically include the following steps: The pressure pulsation feature in the surge warning feature set is input into the first channel of the three-channel neural network, and the feature is extracted through four convolutional layers in the first channel to obtain a pressure feature map; The flow statistics features and frequency band energy features in the surge warning feature set are input into the second channel of the three-channel neural network, and the features are transformed through the three fully connected layers in the second channel in sequence to obtain the flow feature representation; The bearing vibration feature vector in the surge warning feature set is input into the third channel of the three-channel neural network, and the time series feature is extracted through the two-layer LSTM network in the third channel to obtain the frequency feature sequence; The attention weights are calculated for the pressure feature map, flow feature representation and frequency feature sequence respectively to generate a feature weight matrix; According to the feature weight matrix, the pressure feature map, the flow feature representation and the frequency feature sequence are weightedly fused to obtain a fused feature vector; The fused feature vector is input into the output layer of the three-channel neural network, and the probability distribution of the three levels is calculated through the softmax classifier. The level corresponding to the maximum probability value is taken as the surge warning level.

[0027] Specifically, the surge warning feature set is divided into three parts, in which the pressure pulsation feature is used as the first channel input, the flow statistics feature and the frequency band energy feature are used as the second channel input, and the bearing vibration feature vector is used as the third channel input. Assume that the input feature set is , then it is expressed as: in, Represents the pressure pulsation characteristics, Represents traffic statistics and frequency band energy characteristics, represents the bearing vibration characteristic vector. In order to fully extract the pressure pulsation characteristics, Input to the first channel, which uses a four-layer convolutional neural network to extract spatial features and enhance the local pattern representation of the pressure signal. Assume that the convolution kernel is , the bias term is , No. The output of the layer convolution is expressed as: Among them, * represents the convolution operation, represents the activation function (ReLU), is the initial input feature. After four layers of convolution operations, the pressure feature map is obtained. : At the same time, in order to extract traffic statistics and frequency band energy characteristics, Input to the second channel, which uses a three-layer fully connected neural network for feature transformation and dimensionality reduction. Let the weight matrix be , the bias term is , then the calculation formula of the fully connected layer is as follows: in, As the initial input, after being transformed by a three-layer fully connected network, the traffic feature representation is obtained : In order to extract the time series characteristics of bearing vibration, The input is fed into the third channel, which uses a two-layer long short-term memory network (LSTM) to learn the time dependency of bearing vibration. Let the input of the LSTM unit be , which is calculated as: in, and is the weight matrix of the LSTM network, is the bias term, Indicates After two layers of LSTM calculation, the frequency feature sequence is obtained. : In order to improve the predictive ability of the model, the pressure feature mapping , Traffic characteristics representation and frequency characteristic sequence Perform attention calculation to generate a feature weight matrix. Assume that the weight parameter of attention calculation is , then the feature attention score Calculated by the following formula: in, Represents different feature channels. By calculating the attention weight matrix , and obtain the final weighted feature vector : The fused feature vector The data is input to the output layer of the three-channel neural network, and the probability distribution of the surge warning level is calculated by the softmax classifier. The calculation formula of the softmax classifier is as follows: ; in, and are the weights and biases of the output layer, Representatives belong to The probability of level surge warning. Select the level corresponding to the maximum probability value as the surge warning level: .

[0028] In a specific embodiment, the process of executing step S5 may specifically include the following steps: Discretize the control range of the guide vane angle controller, the control range of the anti-surge valve opening controller, and the control range of the return valve opening controller to obtain the controller parameter space; A fuzzy rule base is constructed based on the surge warning level, and the controller parameter space is input into the fuzzy rule base for reasoning operation to obtain the fuzzy control decision sequence. According to the fuzzy control decision sequence, the optimization objective function is set, the surge suppression effect is set as the first weight, the energy loss is set as the second weight, and the adjustment time is set as the third weight, and the parameter optimization objective is obtained; According to the parameter optimization target, a genetic algorithm population is constructed, chromosome coding is generated, and binary coding mapping is performed on the controller parameter space to obtain the initial population; Genetic evolution calculation is performed on the initial population to obtain the optimal chromosome, and decoding operation is performed on the optimal chromosome to convert the binary code into control parameters of the guide vane angle, anti-surge valve opening and reflux valve opening; Based on the control parameters, the guide vane angle, anti-surge valve opening and return valve opening are synchronously adjusted, the changes in pressure pulsation characteristics, flow statistical characteristics and characteristic frequency are monitored, and the surge margin value is calculated.

[0029] Specifically, the control parameters are discretized to construct the controller parameter space. In the surge control system of the centrifugal air compressor, the guide vane angle Control the angle range of flow entering the compressor and the opening degree of the anti-surge valve Control the exhaust gas recirculation volume and the opening of the return valve The redistribution of the control airflow in the system is controlled. The three control variables are discretized to form a searchable control parameter space. The discretization levels are , and , then the controller parameter space is expressed as: in, represents a discrete set of guide vane angles, represents the discrete set of anti-surge valve openings, Represents a discrete set of return valve openings. Controller parameter space By all possible Combination composition. A fuzzy rule base is constructed based on the surge warning level to select a reasonable control strategy through the fuzzy control method. Surge warning level Three states are set, namely low risk (L1), medium risk (L2), and high risk (L3), and fuzzy logic is used to define the control rules. After the controller parameter space is input into the fuzzy rule base for reasoning calculation, the fuzzy control decision sequence is obtained. , that is, multiple possible Combination. Set the optimization objective function according to the fuzzy control decision sequence to ensure that the control strategy takes into account surge suppression, energy loss and regulation response time. Set the optimization objective function It consists of three parts: Surge suppression effect , measure the improvement of surge: in, is the pressure variation amplitude, The larger the value is, the better the surge suppression effect is. , measure the energy loss of anti-surge valve and backflow valve: in, is the weight coefficient, indicating that a larger opening will result in a greater energy loss. Adjustment time , which measures the responsiveness of the adjustment strategy: in, Represents the sum of the response times of the guide vane, anti-surge valve and return valve. The optimization objective function is: in, is the weight coefficient, which is used to balance the influence of surge suppression, energy loss and speed regulation. In order to optimize the control parameters, a genetic algorithm population is constructed based on the objective function. The control parameters are represented by binary coding method. for: in, is the binary code of the guide vane angle, anti-surge valve opening and return valve opening. Assume the code length is , then the mapping relationship of the controller parameters is: Perform genetic evolution calculations on the initial population, optimize the population through selection, crossover and mutation operations, and finally obtain the optimal chromosome , and then perform decoding operation to obtain the optimal control parameters Based on the optimal control parameters, the guide vane angle, anti-surge valve opening and return valve opening are adjusted synchronously, and the pressure pulsation characteristics are monitored by sensors. Traffic statistics characteristics and bearing vibration characteristics , calculate surge margin : in, is the feature weight, which is used to comprehensively evaluate the surge margin.

[0030] In a specific embodiment, the process of executing the step of synchronously adjusting the guide vane angle, the anti-surge valve opening and the return valve opening based on the control parameters, monitoring the changes of the pressure pulsation characteristics, the flow statistical characteristics and the characteristic frequency, and calculating the surge margin value may specifically include the following steps: The control parameters are input into the servo motor drive system and the electric hydraulic servo system to synchronously adjust the guide vane angle, the anti-surge valve opening and the return valve opening to obtain the actuator action sequence; Perform response time detection on the actuator action sequence, record the response time of guide vane angle adjustment, anti-surge valve opening adjustment and return valve opening adjustment, and obtain response time data; Calculate the pressure ratio change rate, flow fluctuation amplitude and bearing vibration intensity based on the response time data, evaluate the surge suppression effect, and obtain the performance parameter matrix; Perform stability analysis on the performance parameter matrix, calculate overshoot, stabilization time and steady-state error, and obtain the operating status index; The surge margin is calculated based on the operating status index, and the pressure ratio change rate is multiplied by the first target value, the flow fluctuation amplitude is multiplied by the second target value, and the bearing vibration intensity is multiplied by the third target value for weighted summation to obtain the surge margin value.

[0031] Specifically, the optimal control parameters are input into the servo motor drive system and the electric hydraulic servo system to drive the synchronous adjustment of the guide vane angle, the anti-surge valve opening and the return valve opening. Assume that the optimized control parameters are the guide vane angle , Anti-surge valve opening and reflux valve opening , then the control instruction is expressed as: in, As control input, it is transmitted to the servo motor drive system and the electric hydraulic servo system respectively to ensure the coordinated adjustment of the guide vane angle, anti-surge valve and return valve. The servo motor drive system is used to adjust the guide vane angle, and its dynamic response model is expressed as: in, is the gain of the guide vane angle system, is the guide vane angle control input, is the time constant of the servo motor. Similarly, the dynamic responses of the electrical hydraulic servo system of the anti-surge valve and the reflux valve are: in, and are the gains of the anti-surge valve and the backflow valve, and is the corresponding control input, and are their respective time constants. Through these dynamic models, the action response of the actuator is obtained to form the actuator action sequence: ; After the actuator action is completed, the response time of the actuator action sequence is detected and the guide vane angle adjustment response time is recorded. , Anti-surge valve opening adjustment response time and reflux valve opening adjustment response time The response time of the valve and guide vane is defined as the time required to reach 95% of the set value, that is: Calculate the pressure ratio change rate based on the response time data , flow fluctuation amplitude and bearing vibration intensity , in order to evaluate the surge suppression effect. The pressure ratio change rate is defined as the change in pressure ratio per unit time, that is: in, and are the outlet and inlet pressures, is the maximum response time of the control actuator. Similarly, the flow fluctuation amplitude Calculated as: in, is the flow signal, is the mean value of the flow rate. The bearing vibration intensity Calculated as the RMS value of vibration acceleration: in, The bearing vibration signal is The value of the sampling point, is the total number of samples. Perform stability analysis on the surge suppression effect and calculate the overshoot , Stabilization time and steady-state error The overshoot is defined as the maximum deviation of the system response and is calculated as: ; in, is the actuator value at the final stable state. Stabilization time Defined as the time required for the system response to reach and remain within a 5% error range: ; Steady-state error Calculated as: ; in, is the final stable state of the system, To set the target value. Calculate the surge margin value based on the operating status index , calculated by weighted summing the pressure ratio change rate, flow fluctuation amplitude and bearing vibration intensity: ; in, , and is the target weight, which is used to adjust the impact of different parameters on the surge margin.

[0032] The above describes the centrifugal air compressor surge suppression method in the embodiment of the present invention. The following describes the centrifugal air compressor surge suppression device in the embodiment of the present invention. Figure 2 In one embodiment of the present invention, a centrifugal air compressor surge suppression device comprises: A feature extraction module is used to extract pressure pulsation features from the pressure signal of the centrifugal air compressor to obtain pressure pulsation features; A feature combination module is used to extract the flow statistical features and frequency band energy features of the flow signal in the centrifugal air compressor, and combine them with the pressure pulsation features to obtain the surge warning feature space; A feature decomposition module is used to perform feature decomposition on the bearing vibration signal of the centrifugal air compressor, extract the characteristic frequency, and combine it with the surge warning feature space to obtain a surge warning feature set; An early warning prediction module is used to input the surge early warning feature set into a three-channel neural network to perform surge early warning prediction and obtain a surge early warning level; The calculation module is used to generate control parameters of the guide vane angle controller, the anti-surge valve opening controller and the return valve opening controller according to the surge warning level, and synchronously adjust the guide vane angle, the anti-surge valve opening and the return valve opening to calculate the surge margin value.

[0033] Through the cooperation of the above components, the multi-scale analysis of the pressure signal by wavelet transform and Fourier transform is combined with the principal component analysis method to reduce the dimension, effectively extract the time-frequency characteristics of the pressure pulsation, reduce redundant information, and improve the efficiency of feature extraction. The flow signal is processed by combining Butterworth filtering and wavelet packet transform. By extracting statistical features and frequency band energy features, the dynamic characteristics of flow fluctuations are accurately captured, and the reliability of surge warning is enhanced. The empirical mode decomposition and Hilbert envelope spectrum analysis are introduced to process the bearing vibration signal. Combined with the envelope demodulation technology, the vibration characteristics of each bearing component are accurately extracted, and the information dimension of surge warning is expanded. A three-channel parallel deep neural network structure is designed. Different types of features are processed respectively by convolutional layers, fully connected layers and LSTM networks, and the attention mechanism is used for feature fusion to construct an accurate surge warning model. A multi-parameter collaborative optimization method based on fuzzy rules and genetic algorithms is proposed. Through dynamic weight update and objective function optimization, the precise control of guide vane angle, anti-surge valve opening and return valve opening is achieved. A complete surge suppression effect evaluation system has been established. By monitoring the pressure ratio change rate, flow fluctuation amplitude and bearing vibration intensity, the surge margin value is calculated to ensure real-time optimization and adjustment of the control strategy.

[0034] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0035] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0036] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for suppressing surge of a centrifugal air compressor, characterized in that: The method comprises: Extract the pressure pulsation characteristics of the pressure signal of the centrifugal air compressor to obtain the pressure pulsation characteristics; Extracting flow statistics characteristics and frequency band energy characteristics of the flow signal in the centrifugal air compressor, and combining them with the pressure pulsation characteristics to obtain a surge warning feature space; Performing feature decomposition on the bearing vibration signal of the centrifugal air compressor, extracting the characteristic frequency, and combining it with the surge warning feature space to obtain a surge warning feature set; Inputting the surge warning feature set into a three-channel neural network to perform surge warning prediction to obtain a surge warning level; The control parameters of the guide vane angle controller, the anti-surge valve opening controller and the return valve opening controller are generated according to the surge warning level, and the guide vane angle, the anti-surge valve opening and the return valve opening are adjusted synchronously to calculate the surge margin value.

2. The method for suppressing surge of a centrifugal air compressor according to claim 1, characterized in that: The pressure pulsation feature extraction of the pressure signal of the centrifugal air compressor to obtain the pressure pulsation feature includes: The inlet pressure signal, outlet pressure signal and intermediate pressure signal of the centrifugal air compressor are collected to obtain an original pressure signal sequence; Input the original pressure signal sequence into the db4 wavelet basis function to perform 4-layer wavelet decomposition to obtain multi-scale features of the pressure signal; Performing discrete Fourier transform on the multi-scale features of the pressure signal to obtain a frequency domain feature sequence; Performing statistical analysis on the original pressure signal sequence, calculating the root mean square value, peak factor, waveform factor and pulse factor, and combining them with the frequency domain feature sequence to obtain a pressure pulsation feature vector; A principal component analysis is performed on the pressure pulsation feature vector to obtain the pressure pulsation feature.

3. The method for suppressing surge of a centrifugal air compressor according to claim 1, characterized in that: The method extracts the flow statistical characteristics and the frequency band energy characteristics of the flow signal in the centrifugal air compressor and combines them with the pressure pulsation characteristics to obtain the surge warning feature space, including: Sampling the flow signals collected by the inlet flow sensor and the outlet flow sensor to obtain an original flow signal, and performing signal noise reduction on the original flow signal to obtain a filtered flow signal; Performing statistical analysis on the filtered flow signal, calculating the mean, standard deviation, skewness and kurtosis to obtain flow statistical characteristics, and performing wavelet packet decomposition on the filtered flow signal to obtain a frequency band coefficient matrix; Calculating energy eigenvalues ​​for the frequency band coefficient matrix, extracting energy distribution of each frequency band, and obtaining frequency band energy characteristics; The flow statistical feature, the frequency band energy feature and the pressure pulsation feature are feature spliced ​​to obtain a surge warning feature space.

4. The method for suppressing surge of a centrifugal air compressor according to claim 1, characterized in that: The feature decomposition of the bearing vibration signal of the centrifugal air compressor is performed to extract the feature frequency, and the feature frequency is combined with the surge warning feature space to obtain a surge warning feature set, including: Sampling the bearing vibration signal collected by the acceleration sensor at the inlet and outlet bearings of the centrifugal air compressor to obtain an original vibration signal; Performing empirical mode decomposition on the original vibration signal to obtain a vibration signal component matrix, and performing Hilbert transform on the vibration signal component matrix to obtain a frequency spectrum feature sequence; Extracting characteristic frequencies from the frequency spectrum characteristic sequence, calculating inner ring passing frequency, outer ring passing frequency, rolling element passing frequency and cage rotation frequency, and obtaining bearing characteristic frequencies; The bearing characteristic frequency is subjected to envelope demodulation analysis to extract vibration characteristics of various bearing components to obtain a bearing vibration characteristic vector, and the bearing vibration characteristic vector is combined with the surge warning feature space to obtain a surge warning feature set.

5. The method for suppressing surge of a centrifugal air compressor according to claim 1, characterized in that: The step of inputting the surge warning feature set into a three-channel neural network to perform surge warning prediction and obtain a surge warning level includes: Inputting the pressure pulsation feature of the surge warning feature set into the first channel of the three-channel neural network, performing feature extraction through four convolutional layers in the first channel to obtain a pressure feature map; Inputting the flow statistics features and the frequency band energy features in the surge warning feature set into the second channel of the three-channel neural network, and sequentially performing feature transformation through the three fully connected layers in the second channel to obtain a flow feature representation; Inputting the bearing vibration feature vector in the surge warning feature set into the third channel of the three-channel neural network, performing time series feature extraction through the two-layer LSTM network in the third channel, and obtaining a frequency feature sequence; Calculating attention weights for the pressure feature map, the flow feature representation, and the frequency feature sequence respectively to generate a feature weight matrix; Performing weighted fusion on the pressure feature map, the flow feature representation and the frequency feature sequence according to the feature weight matrix to obtain a fused feature vector; The fused feature vector is input into the output layer of the three-channel neural network, the probability distribution of three levels is calculated by the softmax classifier, and the level corresponding to the maximum probability value is taken as the surge warning level.

6. The method for suppressing surge of a centrifugal air compressor according to claim 1, characterized in that: The control parameters of the guide vane angle controller, the anti-surge valve opening controller and the return valve opening controller are generated according to the surge warning level, and the guide vane angle, the anti-surge valve opening and the return valve opening are adjusted synchronously to calculate the surge margin value, including: Discretize the control range of the guide vane angle controller, the control range of the anti-surge valve opening controller, and the control range of the return valve opening controller to obtain the controller parameter space; Building a fuzzy rule base based on the surge warning level, inputting the controller parameter space into the fuzzy rule base for reasoning operation, and obtaining a fuzzy control decision sequence; According to the fuzzy control decision sequence, an optimization objective function is set, surge suppression effect is set as a first weight, energy loss is set as a second weight, and adjustment time is set as a third weight, to obtain a parameter optimization objective; Constructing a genetic algorithm population according to the parameter optimization target, generating chromosome coding, and performing binary coding mapping on the controller parameter space to obtain an initial population; Performing genetic evolution calculation on the initial population to obtain an optimal chromosome, and performing decoding operation on the optimal chromosome to convert the binary code into control parameters of the guide vane angle, the anti-surge valve opening and the return valve opening; The guide vane angle, the anti-surge valve opening and the return valve opening are synchronously adjusted based on the control parameters, the changes of the pressure pulsation characteristics, the flow statistical characteristics and the characteristic frequency are monitored, and the surge margin value is calculated.

7. The method for suppressing surge of a centrifugal air compressor according to claim 6, characterized in that: The synchronously adjusting the guide vane angle, the anti-surge valve opening and the return valve opening based on the control parameters, monitoring the changes of the pressure pulsation characteristics, the flow statistical characteristics and the characteristic frequency, and calculating the surge margin value includes: The control parameters are input into the servo motor drive system and the electric hydraulic servo system to synchronously adjust the guide vane angle, the anti-surge valve opening and the return valve opening to obtain the actuator action sequence; Perform response time detection on the actuator action sequence, record the guide vane angle adjustment response time, the anti-surge valve opening adjustment response time and the return valve opening adjustment response time, and obtain response time data; Calculate the pressure ratio change rate, flow fluctuation amplitude and bearing vibration intensity according to the response time data, evaluate the surge suppression effect, and obtain a performance parameter matrix; Perform stability analysis on the performance parameter matrix, calculate overshoot, stabilization time and steady-state error, and obtain operating status indicators; The surge margin is calculated based on the operating state index, and the pressure ratio change rate is multiplied by the first target value, the flow fluctuation amplitude is multiplied by the second target value, and the bearing vibration intensity is multiplied by the third target value for weighted summation to obtain a surge margin value.

8. A centrifugal air compressor surge suppression device, characterized in that: Used to perform the centrifugal air compressor surge suppression method according to any one of claims 1 to 7, the centrifugal air compressor surge suppression device comprises: A feature extraction module is used to extract pressure pulsation features from the pressure signal of the centrifugal air compressor to obtain pressure pulsation features; A feature combination module, used to extract the flow statistical features and frequency band energy features of the flow signal in the centrifugal air compressor, and combine them with the pressure pulsation features to obtain a surge warning feature space; A feature decomposition module, used for performing feature decomposition on the bearing vibration signal of the centrifugal air compressor, extracting the feature frequency, and combining it with the surge warning feature space to obtain a surge warning feature set; An early warning prediction module is used to input the surge early warning feature set into a three-channel neural network to perform surge early warning prediction and obtain a surge early warning level; The calculation module is used to generate control parameters of the guide vane angle controller, the anti-surge valve opening controller and the return valve opening controller according to the surge warning level, and synchronously adjust the guide vane angle, the anti-surge valve opening and the return valve opening to calculate the surge margin value.

Images