A method for extracting and evaluating data features of a hydropower station equipment state monitoring

Abstract

The application relates to the technical field of hydropower station equipment state monitoring, and discloses a hydropower station equipment state monitoring data feature extraction and evaluation method, which comprises the following steps: firstly, an energy transmission topological network model is constructed based on the physical structure of equipment; secondly, multi-source heterogeneous monitoring signals of each node are acquired, and the signals are uniformly converted into energy dissipation rates in watt units; then, based on the topological network model and the energy dissipation rates, system-level state features including total system entropy production rates, entropy production topological centralities, main path information fidelity degrees and the like, and component-level state features including node dissipation energy spectrum entropy and the like are extracted; finally, the features are combined to generate a comprehensive state evaluation vector, fault positioning is carried out based on the system-level features, fault mode judgment is carried out based on the component-level features, and collaborative fault diagnosis is realized. Through the fusion of multi-source information under a unified energy framework, multi-dimensional and hierarchical evaluation and accurate diagnosis of the equipment state are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hydropower station equipment state monitoring, in particular to a hydropower station equipment state monitoring data feature extraction and evaluation method. BACKGROUND

[0002] Hydropower stations are a key component of modern energy systems, and their stability and safety in operation are of great importance. In a hydropower station, the hydroelectric generator set is the core power equipment that converts water energy into electrical energy. It has a complex structure, high value, and is in a high-load operation state for a long time. In order to ensure the safety of the unit, improve its operating efficiency and implement predictive maintenance, a large number of sensors are usually deployed on the key physical components of the unit to form a complex state monitoring system. The system continuously collects a large amount of monitoring data including equipment vibration, operating temperature, acoustic pressure, electromagnetic parameters, etc. These data have the characteristics of multi-source heterogeneity, with diverse sources and different physical properties, providing basic data support for accurately evaluating the health status of the equipment and warning potential failures in advance.

[0003] Existing equipment state monitoring data analysis methods face significant challenges in dealing with these multi-source heterogeneous information. Since vibration, temperature, acoustic and other signals belong to different physical categories and have completely different dimensions and physical connotations, traditional analysis methods usually treat them as independent and fragmented information streams. This processing method leads to a serious information island phenomenon, that is, it is difficult to effectively integrate and correlate the monitoring data from different sources in a unified framework. For example, vibration analysis experts and thermal analysis experts may draw conclusions based on their own data, but lack an objective and quantitative means to comprehensively evaluate the combined impact of the two phenomena on the overall state of the unit, thereby limiting the ability to conduct a comprehensive and overall assessment of the equipment state.

[0004] Current mainstream state evaluation and fault diagnosis techniques have limitations in the depth of analysis models and diagnostic conclusions. Many methods rely too much on data-driven statistical models or artificial intelligence algorithms. These models can discover some correlations from massive data, but often ignore the physical structure of the equipment, the mutual constraints between components, and the transfer and conversion rules of energy within the system, resulting in models lacking sufficient physical interpretability and making it difficult to guarantee the reliability of diagnostic conclusions. This analysis paradigm often causes the monitoring system to remain at the threshold alarm level of a single measurement point, providing fragmented state evaluations that are difficult to reveal the overall stability degradation trend or topological structure instability risk of the equipment from a system macro perspective. The final output of the fault diagnosis conclusion is also relatively single, usually only pointing out the approximate location of the fault, but unable to effectively distinguish the root mode of the fault (such as deterministic frequency fault or wideband wear fault), thereby affecting the accuracy of subsequent maintenance decisions. SUMMARY

[0005] In view of the deficiencies of the prior art, the application provides a hydropower station equipment state monitoring data feature extraction and evaluation method, which solves the problem that the existing hydropower station equipment state monitoring method is difficult to perform fusion analysis in a unified physical dimension when processing multi-source heterogeneous data, lacks dynamic evaluation of the overall state evolution trend of the system and deep diagnosis capability of the root cause of the fault, and results in insufficient comprehensiveness of state evaluation and accuracy of diagnosis conclusion.

[0006] To solve the above technical problems, the application provides a hydropower station equipment state monitoring data feature extraction and evaluation method.

[0007] The application provides the following technical scheme: a hydropower station equipment state monitoring data feature extraction and evaluation method, comprising the following steps:

[0008] Step S1: based on the physical structure of the hydropower station equipment, an energy transmission topological network model is constructed. The energy transmission topological network model is composed of a plurality of nodes representing physical components of the equipment and a plurality of directed edges representing energy transmission paths between the physical components, and the model provides a physical constraint and a topological basis for subsequent calculation.

[0009] Step S2: a plurality of multi-source heterogeneous monitoring signals corresponding to the nodes in the energy transmission topological network model are obtained, and the monitoring signals include vibration signals, temperature signals and acoustic signals.

[0010] Step S3: the multi-source heterogeneous monitoring signals are uniformly represented by energy rates to obtain energy dissipation rates representing the energy dissipation levels of each physical component. This step uniformly projects the monitoring signals of different physical dimensions into the energy rate dimension (unit: watt) to realize fusion analysis of multi-physical information. The specific conversion process is as follows:

[0011] For the vibration signal, the corresponding kinetic energy dissipation power is which can be calculated by the following formula:

[0012] ;

[0013] In the formula, is the kinetic energy dissipation power, is the equivalent mass of the physical component, is the instantaneous acceleration signal of the component, and is the instantaneous speed obtained by integrating the instantaneous acceleration signal. For the temperature signal, the corresponding heat energy dissipation power is which can be calculated by the following formula:

[0014] ;

[0015] In the formula, the power of thermal energy dissipation; the specific heat capacity of the physical component; the mass of the physical component; the instantaneous temperature signal of the component the rate of change over time.

[0016] for acoustic signals, the corresponding acoustic energy radiation power which can be calculated by the following formula:

[0017] ;

[0018] wherein, is the acoustic energy radiation power, is the acoustic pressure monitoring signal, is the density of the propagation medium, is the acoustic velocity, is the equivalent acoustic radiation area of the component. The total energy dissipation rate of a single node is the sum of all types of dissipation power.

[0019] Step S4: extracting system-level state features and component-level state features based on the energy transfer topology network model and the energy dissipation rate.

[0020] The system-level state features include:

[0021] Total system entropy production rate: summing up the energy dissipation rates of all nodes in the energy transfer topology network model, and taking the result as a macroscopic indicator for quantifying the overall energy invalid dissipation level of the hydropower station equipment.

[0022] Entropy production topology centrality: among all nodes constituting the energy transfer topology network model, identify the node with the maximum energy dissipation rate, and take the identification of the node as the entropy production topology centrality, to locate the physical component with the most significant energy dissipation.

[0023] Main path information fidelity: first, pre-set the main energy transfer path in the energy transfer topology network model; then, for adjacent nodes on the main energy transfer path, based on the time series of their energy dissipation rates, calculate the transfer entropy between them to quantify the causal relationship strength of the upstream node to the energy information flow of the downstream node; finally, weight and sum the transfer entropies calculated between all adjacent nodes on the main energy transfer path to obtain the main path information fidelity.

[0024] The topology instability tendency index is obtained by: firstly, constructing a state feature vector sequence in time dimension based on the extracted system-level state features; then, identifying topology steady states corresponding to different typical working conditions by clustering analysis on the state feature vector sequence during normal operation of the hydropower station equipment; then, establishing a baseline transition model representing the normal state transition rule of the equipment based on historical data of normal switching between topology steady states; finally, monitoring the actual transition behavior of the current state feature vector sequence between topology steady states, and quantifying the deviation of the actual transition behavior from the baseline transition model, to obtain the topology instability tendency index. The deviation is determined by comprehensively quantifying the topology steady state transition frequency per unit time, the proportion of transient time when the state feature vector sequence stays outside any identified topology steady state, and the difference between the probability distribution of the current actual transition behavior and the probability distribution of the baseline transition model.

[0025] The component-level state feature includes node dissipation energy spectrum entropy. This feature is dedicated to fault mode diagnosis of the node with the most significant energy dissipation determined by entropy production topology centrality, and the extraction steps are: firstly, for the node determined by the entropy production topology centrality, the corresponding original monitoring signal is obtained; then, the power spectrum of the original monitoring signal is calculated to obtain the dissipation energy spectrum describing the distribution of the energy dissipation of the node in the frequency domain; finally, the dissipation energy spectrum is normalized to obtain a normalized energy spectrum , and the information entropy thereof is calculated to obtain the node dissipation energy spectrum entropy .

[0026] ;

[0027] wherein, is the power spectrum of the original monitoring signal of the node, is the normalized energy spectrum obtained by normalizing the power spectrum, is the frequency, is the frequency infinitesimal.

[0028] Step S5: combining the system-level state features and the component-level state features to generate a multi-dimensional comprehensive state evaluation vector. This step provides quantitative and multi-dimensional analysis basis for the operation state evaluation and maintenance decision of the hydropower station equipment.

[0029] Step S6: based on the system-level state features and the component-level state features, cooperative fault diagnosis is performed to generate a diagnosis conclusion containing fault location information and fault mode information. The specific implementation of this step is: using the physical components determined by the entropy production topological centrality to determine the fault location information in the diagnosis conclusion; using the numerical value of the node dissipation energy spectrum entropy to determine whether the fault mode in the diagnosis conclusion belongs to the fault mode with energy concentrated in a deterministic frequency or the fault mode with energy dispersed in a wide frequency band; and associating the fault location information with the fault mode information to generate the diagnosis conclusion.

[0030] The application provides a hydropower station equipment state monitoring data feature extraction and evaluation method.

[0031] 1. The application solves the technical problem that different physical dimension data are difficult to fuse and analyze by constructing an energy transmission topological network model based on the physical structure of equipment and uniformly representing multi-source heterogeneous monitoring signals such as vibration, temperature and acoustics as energy dissipation rates. The method enables the evaluation of the equipment state to be performed in a unified energy dimension with clear physical meaning, thereby constructing a systematic analysis framework and improving the comprehensiveness and consistency of the state evaluation.

[0032] 2. The application realizes deep insight into the equipment state by extracting state features in multiple dimensions such as macro static, macro dynamic and microscopic mode, and the total system entropy production rate and entropy production topological centrality quantize the energy efficiency of the system as a whole and locate the key dissipation source; the main path information fidelity captures the causal relationship anomaly of energy flow by using transfer entropy; in particular, the topological instability tendency index realizes dynamic evaluation of the operation stability by analyzing the transfer behavior of the system between different steady states, and can provide early warning for slowly developing potential faults.

[0033] 3. The application realizes accurate judgment of the fault location and fault mode through cooperative diagnosis, improves the accuracy and guidance of the diagnosis conclusion, first determines the physical component with the most significant energy dissipation by using the entropy production topological centrality, thereby locking the fault location information; then extracts the node dissipation energy spectrum entropy of the component, and determines whether it is a deterministic fault with energy concentration or a random fault with energy dispersion by quantifying the distribution disorder of energy in the frequency domain, and finally generates a complete diagnosis conclusion containing the location and mode, which provides a direct basis for subsequent maintenance decision. BRIEF DESCRIPTION OF DRAWINGS

[0034] Figure 1 The method flowchart of the application;

[0035] Figure 2 The generator set energy transmission topological network model schematic diagram of the application;

[0036] Figure 3 The monitoring signal layout and energy rate conversion schematic diagram of the present application;

[0037] Figure 4 The multi-dimension state feature extraction flowchart of the present application;

[0038] Figure 5 The comprehensive evaluation and collaborative fault diagnosis flowchart of the present application. DETAILED DESCRIPTION

[0039] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the specification of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the present application.

[0040] Referring to the drawings Figure 1 , Figure 1 The flowchart of a hydropower station equipment state monitoring data feature extraction and evaluation method according to an embodiment of the present application. The technical solutions provided by the present application take a Francis turbine generator set as a monitoring object in a specific embodiment, and the method can include the following steps:

[0041] Step S1: based on the physical structure of the hydropower station equipment, an energy transmission topological network model is constructed, the energy transmission topological network model including nodes representing physical components of the equipment and edges representing energy transmission paths between the physical components.

[0042] Step S2: a plurality of source heterogeneous monitoring signals corresponding to the nodes are obtained.

[0043] Step S3: the plurality of source heterogeneous monitoring signals are uniformly represented by energy rates to obtain energy dissipation rates representing energy dissipation levels of each of the physical components.

[0044] Step S4: based on the energy transmission topological network model and the energy dissipation rates, system-level state features and component-level state features are extracted.

[0045] Step S5: the system-level state features and the component-level state features are combined to generate a multi-dimension comprehensive state evaluation vector.

[0046] Step S6: based on the system-level state features and the component-level state features, collaborative fault diagnosis is performed to generate a diagnosis conclusion containing fault location information and fault mode information.

[0047] In the present embodiment, first, step S1 is performed, and based on the mechanical structure and energy flow path of the Francis hydroelectric generating unit, an energy transmission topological network model thereof is constructed. Key physical components in the unit, such as the water guide mechanism, runner, main shaft, upper guide bearing, lower guide bearing, thrust bearing, generator rotor, generator stator, etc., are defined as nodes in the model. The directed edges connecting these nodes represent the paths of energy transmission, conversion and dissipation between the components.

[0048] Then, step S2 is performed, and for each defined node, multi-source heterogeneous monitoring signals in the running process thereof are acquired through the sensor network deployed on the unit. These signals include but are not limited to vibration acceleration signals installed on the bearing seat, temperature signals installed in the bearing and stator winding, and acoustic pressure signals arranged near the spiral case or draft tube of the unit.

[0049] Subsequently, step S3 is performed, and the acquired multi-source heterogeneous monitoring signals are uniformly characterized in terms of energy rate, and signals of different physical dimensions are converted into energy dissipation rates in watts W. This step is the basis for subsequent uniform analysis. The specific conversion calculation is as follows:

[0050] For vibration signals, the corresponding kinetic energy dissipation power can be calculated by the following formula:

[0051] ;

[0052] In the formula, is the kinetic energy dissipation power, is the equivalent mass of the physical component, is the instantaneous acceleration signal of the component, is the instantaneous speed obtained by integrating the instantaneous acceleration signal. For temperature signals, the corresponding thermal energy dissipation power can be calculated by the following formula:

[0053] ;

[0054] In the formula, is the thermal energy dissipation power; is the specific heat capacity of the physical component; is the mass of the physical component; is the instantaneous temperature signal of the component which is the rate of change of time.

[0055] For acoustic signals, the corresponding acoustic energy radiation power can be calculated by the following formula:

[0056] ;

[0057] In the formula, is the sound energy radiation power, is the sound pressure monitoring signal, is the density of the propagation medium, is the speed of sound, is the equivalent acoustic radiation area of ​​the component. The total energy dissipation rate of a single node is the sum of all types of dissipated power.

[0058] After obtaining the energy dissipation rate of each node, step S4 is executed to extract state features that can characterize the operating status of the unit at different levels. The system-level state features include: total system entropy production rate, entropy production topological centrality, main path information fidelity, and topological instability tendency index. The component-level state features include node dissipation energy spectrum entropy. Node dissipation energy spectrum entropy The calculation method is:

[0059] ;

[0060] in, is the power spectrum of the original monitoring signal corresponding to the node, is the normalized energy spectrum obtained by normalizing the power spectrum, is the frequency, is the frequency element.

[0061] Refer to the attached Figure 2 , Figure 2 This is a schematic diagram of a Francis turbine generator set energy transfer topology network model constructed according to an embodiment of the present invention. Step S1, constructing the energy transfer topology network model, is the foundation of the present invention's method. This step ensures that subsequent state analysis is based on clear physical structural constraints.

[0062] This step may specifically include: first, physical component identification and node abstraction; then, energy path analysis and edge definition.

[0063] During object component identification and node abstraction, the physical entities responsible for the core energy transmission, conversion, support, and dissipation functions of the Francis turbine generator set are precisely identified and abstracted as nodes in the model. Each node uniquely corresponds to a specific, monitorable physical component. For the Francis turbine generator set in this embodiment, the physical components defined as nodes include: the guide mechanism, runner, main shaft, upper guide bearing, lower guide bearing, thrust bearing, generator rotor, and generator stator.

[0064] In the energy path analysis and edge definition, the energy flow relationship between the aforementioned nodes with a clear physical direction is defined as a directed edge Edge in the model. These directed edges Edge objectively describe the process of energy flowing from one component to another, and the direction is determined by the causal relationship of energy transmission. The edges are divided into two categories according to their physical properties:

[0065] The first category is the main energy transmission edge. This type of edge represents the main path of converting hydraulic energy into mechanical energy and then into electrical energy. In this embodiment, the main energy transmission edge includes the edge from the water guide mechanism to the runner, representing the impact of high-pressure water flow on the runner, converting hydraulic energy into mechanical energy for the rotation of the runner; the edge from the runner to the main shaft, representing the mechanical energy transmission of the runner driving the main shaft to rotate; and the edge from the main shaft to the generator rotor, representing the mechanical energy transmission of the main shaft driving the generator rotor to rotate.

[0066] The second category is the energy dissipation edge. This type of edge represents the energy path that is inevitably dissipated in the form of heat, vibration, sound, etc. during energy transmission and conversion. In this embodiment, the energy dissipation edge includes the edge from the main shaft to the upper guide bearing, lower guide bearing, and thrust bearing it contacts, representing the heat and vibration energy dissipation generated in each bearing due to mechanical friction when the main shaft rotates; and the edge from the generator stator to the external environment or cooling system, representing the heat energy dissipation generated by the stator winding due to current heating effect and core loss.

[0067] After the definition of the nodes and edges described above, the energy transmission topology network model can be stored as a computer-readable data structure, such as an adjacency matrix or an adjacency list, when calculated and implemented. In this data structure, the identifier of each node and its corresponding physical component, as well as the direction and type of the edge connecting the nodes, main energy transmission or energy dissipation, are clearly recorded, providing structured input for subsequent program calls.

[0068] Referring to the accompanying drawings Figure 3 , Figure 3 is a schematic diagram of the signal layout and energy rate conversion of the hydroelectric generator set according to an embodiment of the present application. After the construction of the energy transmission topology network model, the method of the present application then performs steps S2 and S3, i.e., signal acquisition and energy rate unified representation.

[0069] The flow can specifically include: first, monitoring point layout and signal acquisition; and then, multi-source data energy conversion.

[0070] In the monitoring point arrangement and signal acquisition step S2, for each node defined in step S1, a corresponding sensor is deployed at its physical location or adjacent area to obtain real-time monitoring signals of its operating state. The type and installation location of the sensor are determined according to the physical component it corresponds to and the physical quantity to be monitored. In the embodiment, the specific arrangement of signal acquisition is as follows:

[0071] On the bearing seats of the upper guide bearing, the lower guide bearing, and the thrust bearing, three-axis acceleration sensors are respectively installed to collect vibration signals of these nodes in the radial direction, the axial direction, and the tangential direction. Meanwhile, platinum resistance temperature sensors are installed inside the pads or oil grooves of each bearing to collect operating temperature signals.

[0072] Temperature sensors are installed at the stator core or winding end of the generator to collect temperature signals when the stator is operating.

[0073] Acoustic pressure sensors or microphones are installed on the outside of the unit spiral case or the tailrace pipe wall to collect acoustic signals generated by water flow and unit operation.

[0074] All sensors convert the collected analog signals into digital time series signals through the data acquisition system, providing data input for the next calculation.

[0075] In the multi-source data energy rate conversion step S3, the monitoring signals of different physical dimensions collected in step S2 are uniformly converted into energy dissipation rates in watts.

[0076] When performing energy rate conversion calculation, the required physical component parameters such as equivalent mass, specific heat capacity, mass, and equivalent acoustic radiation area can be obtained by consulting the factory design drawings, manufacturer-provided data manual, or through modeling simulation methods such as finite element analysis (FEA). Once these parameters are determined, they are stored as fixed values in the system for calculation.

[0077] For the vibration acceleration signals collected from each bearing node, the kinetic energy dissipation power formula is used to convert them into kinetic energy dissipation power according to the equivalent mass of the component.

[0078] For the temperature signals collected from each bearing node and the generator stator node, the thermal energy dissipation power formula is used to convert them into thermal energy dissipation power according to the specific heat capacity and mass of the component.

[0079] For the collected acoustic pressure signals, the acoustic energy radiation power formula is used to convert them into acoustic energy radiation power according to the physical properties of the propagation medium and the equivalent acoustic radiation area of the component.

[0080] Through the above conversion, the original physical units of vibration, temperature, sound pressure and other signals are mapped to the unified dimension of energy dissipation rate. For a node that has multiple energy dissipation forms (for example, a bearing node generates vibration and heat at the same time), the total energy dissipation rate of the node is the arithmetic sum of all types of energy dissipation power (kinetic energy dissipation power, thermal energy dissipation power, etc.) on the node. The completion of this step lays the foundation for subsequent extraction of system-level and component-level features in a unified framework.

[0081] Referring to the accompanying drawings Figure 4 , Figure 4 is a multi-dimensional state feature extraction process diagram according to an embodiment of the present application. After completing the calculation of the energy dissipation rate of each node in step S3, the method of the present application then performs step S4, that is, based on the energy transfer topological network model and the energy dissipation rate of each node, system-level state features and component-level state features are extracted.

[0082] This step can specifically include: first, system-level state feature extraction; then, component-level state feature extraction.

[0083] When performing system-level state feature extraction, the present embodiment extracts the following four features:

[0084] Total system entropy production rate. The calculation method of this feature is to perform arithmetic summation of the energy dissipation rates of all nodes in the energy transfer topological network model defined in step S1. The calculation result is a scalar value, which quantifies the total energy loss of the hydro-generator set as a whole in unit time due to irreversible processes, providing a macroscopic physical indicator for evaluating the overall operation efficiency of the unit.

[0085] Entropy production topological centrality. The extraction method of this feature is to compare the energy dissipation rate values of all nodes in the energy transfer topological network model at the same time, identify and output the unique identifier of the node with the largest energy dissipation rate. The output result of this feature directly indicates the physical component with the most concentrated energy dissipation in the entire unit at the current time, which is used for subsequent fault source positioning.

[0086] Main path information fidelity. The extraction method of this feature is: first, determine a main energy transfer path according to the energy transfer topological network model defined in step S1, for example, from the runner node to the main shaft node and then to the generator rotor node. Then, for any two adjacent nodes on the path, obtain the time series of their respective energy dissipation rates and calculate the transfer entropy of the upstream node to the downstream node. Finally, the transfer entropies calculated between all adjacent node pairs on the main path are weighted and summed to obtain a scalar value. This value quantifies the causal fidelity of energy flow on the core transfer link, and a change in its value indicates an abnormality in the energy transfer law on the main path.

[0087] Topological instability tendency index. The extraction of this feature is a dynamic evaluation process: first, the aforementioned system-level state features are combined into a state feature vector, and a vector sequence is formed in chronological order. Then, using clustering algorithms such as K-Means, the historical state feature vector sequence of the unit during the determined normal operation is analyzed, and several stable cluster centers are identified, each center representing a typical topological steady state. Next, based on the transition frequency between each topological steady state in the normal historical data, a baseline state transition model is established, such as a Markov model. Finally, by monitoring the actual transition behavior of the current state feature vector sequence between each topological steady state, and quantifying the transition frequency, the proportion of transient time outside any steady state, and the difference between the actual transition probability distribution and the baseline transition model (e.g., using K-L divergence calculation), the topological instability tendency index is obtained.

[0088] In the extraction of component-level state features, the embodiment extracts node dissipation energy spectrum entropy. The extraction of this feature is targeted, and the extraction object is the node with the most significant energy dissipation determined by the aforementioned entropy production topological centrality. The extraction steps are as follows: first, obtain the original time-domain monitoring signal corresponding to the specific node, such as a vibration signal. Then, calculate the power spectrum of the signal by Fourier transform or other methods to obtain the dissipation energy spectrum describing the distribution of energy dissipation of the node in the frequency domain. Finally, normalize the dissipation energy spectrum and calculate it according to the formula of information entropy to obtain the node dissipation energy spectrum entropy. The value of this entropy reflects the orderliness of the distribution of energy in the frequency domain, which is used to judge the fault mode of the component in step S6.

[0089] Referring to the accompanying drawings Figure 5 , Figure 5 is a schematic diagram of the comprehensive evaluation and collaborative fault diagnosis process according to an embodiment of the present application. After the extraction of multi-dimensional state features is completed in step S4, the method of the present application then executes steps S5 and S6.

[0090] The process can specifically include: first, generating a comprehensive state evaluation vector; then, performing collaborative fault diagnosis based on the features.

[0091] In the generation of the comprehensive state evaluation vector (step S5), all the state features calculated at the same time point in step S4 are combined. Specifically, the system-level state features such as the total system entropy production rate, the entropy production topological centrality, the main path information fidelity, and the topological instability tendency index, and the component-level state feature of the node's dissipation energy spectrum entropy determined by the entropy production topological centrality are arranged in a predefined order to form a high-dimensional comprehensive state evaluation vector. This vector provides a quantitative, multi-dimensional digital snapshot of the operation state of the hydro-generator unit at the time, which can be recorded and used for subsequent device state trend analysis and maintenance decision-making.

[0092] In the weighted summation of the transfer entropy, the weight coefficients between each pair of nodes can be set according to their importance or energy transfer efficiency in the main energy transfer path. In one specific implementation, all weight coefficients can be set to 1, i.e., arithmetic mean or direct summation. In another specific implementation, nodes with higher energy transfer efficiency can be given higher weights.

[0093] In the cooperative fault diagnosis step S6, the execution of this step is usually started by a trigger condition, such as monitoring that the total system entropy production rate or the topological instability tendency index exceeds the preset normal operation threshold. Once triggered, the specific implementation of the diagnosis process is as follows:

[0094] First, the entropy production topological centrality feature is used to determine the fault location information in the diagnosis conclusion. The output result of this feature, i.e., the identifier of the node with the maximum energy dissipation rate at the current time, is directly used as the physical location positioning of the fault, for example, determining the thrust bearing node.

[0095] Then, the node dissipation energy spectrum entropy feature is used to determine the fault mode information in the diagnosis conclusion. The method extracts the value of the node dissipation energy spectrum entropy corresponding to the thrust bearing node located in the previous step. The value is compared with the pre-established energy spectrum entropy reference value of the component under normal working conditions.

[0096] If the node's dissipation energy spectrum entropy value is low, indicating that its energy dissipation is mainly concentrated in a few deterministic frequencies, the fault mode is judged as a deterministic frequency fault mode. Such faults include rotor imbalance, shaft misalignment, etc.

[0097] If the node's dissipation energy spectrum entropy value is high, indicating that its energy dissipation is widely distributed in a wide frequency band, the fault mode is judged as a wide frequency band fault mode. Such faults include random wear of bearings, deterioration of lubrication state, etc.

[0098] Finally, the aforementioned determined fault location information, such as the thrust bearing, is associated and combined with the determined fault mode information, such as the wideband fault mode, to generate a final, structured diagnostic conclusion.

[0099] The method for determining the preset normal operation threshold is as follows: collecting state characteristic data of the equipment under a plurality of typical working conditions (such as different loads and different water heads) in which no fault is confirmed, and forming a normal operation database. Subsequently, a statistical method is used to calculate the mean value and standard deviation of the total system entropy production rate or the topological instability tendency index in the database, and the result of adding three times the standard deviation to the mean value is taken as the threshold for triggering diagnosis.

[0100] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for extracting and evaluating features from hydropower station equipment status monitoring data, characterized in that: The following steps are involved: Step S1: constructing an energy transfer topology network model based on the physical structure of the hydropower station equipment, wherein the energy transfer topology network model includes nodes representing physical components of the equipment and edges representing energy transfer paths between the physical components; Step S2: Acquire multi-source heterogeneous monitoring signals corresponding to the node; Step S3: performing a unified energy rate characterization on the multi-source heterogeneous monitoring signals to obtain an energy dissipation rate characterizing the energy dissipation level of each physical component; Step S4: extracting system-level state features and component-level state features based on the energy transfer topology network model and the energy dissipation rate; Step S5: combining the system-level state features with the component-level state features to generate a multi-dimensional comprehensive state assessment vector; Step S6: performing collaborative fault diagnosis based on the system-level state characteristics and the component-level state characteristics to generate a diagnosis conclusion including fault location information and fault mode information; The step of uniformly characterizing the energy rates of the multi-source heterogeneous monitoring signals in step S3 is specifically implemented as follows: for the monitoring signals associated with a single node, uniformly converting them from their respective physical dimensions to an energy rate dimension to obtain the total energy dissipation rate of the node; wherein the conversion includes: Converting the vibration signals in the multi-source heterogeneous monitoring signals into kinetic energy dissipation power according to the equivalent mass of the corresponding physical components; Converting the temperature signal in the multi-source heterogeneous monitoring signal into thermal energy dissipation power according to the specific heat capacity and mass of the corresponding physical component; Converting the acoustic signal in the multi-source heterogeneous monitoring signal into acoustic energy radiation power based on the physical properties of its propagation medium and the equivalent acoustic radiation area; The system-level state features extracted in step S4 include entropy production topological centrality; the entropy production topological centrality is determined by: identifying a node with a maximum energy dissipation rate among all nodes constituting the energy transfer topological network model, and using the identifier of the node as the entropy production topological centrality to locate the physical component with the most significant energy dissipation; The component-level state features extracted in step S4 include node dissipation energy spectrum entropy, which is specifically used to perform fault mode diagnosis on the node with the most significant energy dissipation determined by the entropy-generating topological centrality. The extraction steps are specifically implemented as follows: First, for the node determined by the entropy production topological centrality, the corresponding original monitoring signal is obtained; Then, the power spectrum of the original monitoring signal is calculated to obtain a dissipated energy spectrum describing the distribution of energy dissipation of the node in the frequency domain; Finally, the dissipated energy spectrum is normalized and its information entropy is calculated to obtain the node dissipated energy spectrum entropy; The step of performing collaborative fault diagnosis in step S6 is specifically implemented as follows: Determining fault location information in the diagnosis conclusion using the physical component determined by the entropy-generated topological centrality; Using the value of the node dissipated energy spectrum entropy, it is determined whether the fault mode in the diagnosis conclusion is a fault mode in which energy is concentrated in a deterministic frequency or a fault mode in which energy is dispersed in a wide frequency band; The fault location information is associated with the fault mode information to generate the diagnosis conclusion.

2. A method for extracting and evaluating features from equipment status monitoring data of a hydropower station according to claim 1, characterized in that: The system-level state characteristics extracted in step S4 include the total system entropy production rate; the method for determining the total system entropy production rate is: summing the energy dissipation rates of all nodes in the energy transfer topology network model, and using the result as a macro indicator to quantify the overall energy ineffective dissipation level of the hydropower station equipment.

3. A method for extracting and evaluating features from equipment status monitoring data of a hydropower station according to claim 1, characterized in that: The system-level state features extracted in step S4 include the main path information fidelity; the main path information fidelity extraction step is specifically implemented as follows: First, a main energy transfer path is pre-set in the energy transfer topology network model; Then, for any two adjacent nodes on the main energy transfer path, the time series of their respective energy dissipation rates are obtained, and the transfer entropy between the two time series is calculated to quantify the causal relationship strength of the upstream node on the energy information flow of the downstream node; Finally, the transfer entropies calculated between all adjacent nodes on the main energy transfer path are weightedly summed to obtain the main path information fidelity.

4. A method for extracting and evaluating features from equipment status monitoring data of a hydropower station according to claim 1, characterized in that: The system-level state features extracted in step S4 also include a topological instability tendency index, and the extraction step of the index is specifically implemented as follows: First, the system-level state characteristics are transformed into a state feature vector sequence in the time dimension; Then, by clustering the state feature vector sequences of the hydropower station equipment during normal operation, the topological steady states corresponding to different typical working conditions are identified. Next, based on historical data of the device switching normally between the topological stable states, a benchmark transition model is established to characterize its normal state transition law; Finally, the topological instability tendency index is obtained by monitoring the actual transition behavior of the current state eigenvector sequence between the topological stable states and quantifying the degree of deviation between the actual transition behavior and the benchmark transition model.

5. A method for extracting and evaluating features from hydropower station equipment status monitoring data according to claim 4, characterized in that: The degree of deviation is determined by comprehensively quantifying the following factors: the frequency of topological steady-state transitions per unit time, the proportion of transient time that the state eigenvector sequence stays outside any identified topological steady state, and the difference between the probability distribution of the current actual transition behavior and the probability distribution of the benchmark transition model.

6. A method for extracting and evaluating features from hydropower station equipment status monitoring data according to claim 1, characterized in that: The step of generating a multi-dimensional comprehensive state assessment vector in step S5 is to provide a quantitative, multi-dimensional analysis basis for the operation state assessment and maintenance decision of the hydropower station equipment. This step is specifically achieved by combining the system-level state features and component-level state features extracted in step S4.

Images