Water turbine governor hydraulic system state early warning method, device, equipment and medium

Abstract

The invention relates to the technical field of hydraulic turbine governor fault diagnosis, and discloses a hydraulic turbine governor hydraulic system state early warning method, device and equipment and a medium. The method comprises the steps that key parameters of a hydraulic turbine governor hydraulic system are collected and processed to obtain all state quantities associated with the state of the hydraulic turbine governor hydraulic system; assigning a corresponding weight according to the influence degree of each state quantity on the operation state; normalizing each state quantity, and constructing a state vector representing the health state of the hydraulic turbine governor hydraulic system; calculating a mean vector of the state vector set to obtain a health state center vector; calculating a weighted Euclidean distance between each state vector in the state vector set and the health state center vector, and taking the weighted Euclidean distance as a real-time state early warning index; and calculating an average value and a standard deviation of continuous n weighted Euclidean distances as fault alarm indexes. According to the invention, real-time risk assessment and fault trend early warning can be carried out on the operation state of the hydraulic turbine governor hydraulic system.

Description

Technical Field

[0001] This invention relates to the field of fault diagnosis technology for hydro turbine governors, and in particular to a method, device, equipment, and medium for early warning of the status of a hydro turbine governor hydraulic system. Background Technology

[0002] As the proportion of intermittent renewable energy sources such as wind and solar power gradually increases in the current power system, hydropower will undertake more peak-shaving and frequency regulation tasks to ensure good integration of intermittent renewable energy with the existing power system, placing higher demands on the safe operation and management of hydropower. The turbine governor, as the core control equipment of the hydro-generator unit, plays a crucial role in regulating the turbine unit's frequency, rapidly synchronizing with the grid, and adjusting load. The hydraulic system of the turbine governor, as the power source for governor regulation, directly affects the governor's ability to accurately and quickly regulate the unit load to meet the grid's peak-shaving and frequency regulation needs. Therefore, to ensure the safe and stable operation of hydro-generator units and the power grid, it is essential to improve the fault diagnosis technology of the turbine governor's hydraulic system, especially its real-time early warning of fault trends, to ensure the safe, stable, and continuous operation of the turbine governor.

[0003] Currently, the main method for diagnosing faults in the hydraulic system of turbine governors is to use fault threshold judgment and post-event handling. That is, when the operating parameters of a certain device exceed the set range or a vulnerable component is damaged and unable to function, a corresponding fault signal is reported, or the device is stopped for handling, or the damaged component is replaced online. This method is limited to functions such as monitoring equipment parameters and controlling the command operation of major equipment, and lacks the ability to provide automated and intelligent fault early warning. This inevitably leads to unplanned equipment shutdowns and reduces equipment operating efficiency. Summary of the Invention

[0004] To address the shortcomings of existing fault diagnosis technologies for turbine governors, this invention proposes a method, device, equipment, and medium for early warning of the hydraulic system status of a turbine governor. It fully utilizes the massive historical data accumulated by the unit monitoring system, incorporates big data concepts, and can use historical data to perform self-learning, update the system operation status early warning model, and improve the accuracy of early warning.

[0005] The technical solution adopted in this invention is as follows: A method for early warning of the status of a hydraulic system of a turbine governor includes: Key parameters of the hydraulic system of the turbine governor are collected, processed to obtain various state variables associated with the state of the hydraulic system of the turbine governor, and assigned corresponding weights according to the degree of influence of each state variable on the operating state. Normalize each state variable to construct a state vector representing the healthy state of the hydraulic system of the turbine governor. Multiple state vectors form a state vector set. Calculate the mean vector of the state vector set under normal state of the hydraulic system of the turbine governor to obtain the center vector of the healthy state. The weighted Euclidean distance between each state vector in the state vector set and the center vector of the healthy state is calculated as a real-time state early warning indicator of the hydraulic system of the turbine governor; the average value and standard deviation of n consecutive weighted Euclidean distances are calculated as fault alarm indicators of the hydraulic system of the turbine governor.

[0006] Furthermore, the process of collecting key parameters of the turbine governor hydraulic system and processing them to obtain various state variables associated with the state of the turbine governor hydraulic system includes: collecting historical data of key parameters of the turbine governor hydraulic system; based on the complexity of the monitoring values ​​and operating conditions of the turbine governor hydraulic system, and combining the correlation between the state variables and the operating state of the turbine governor hydraulic system, setting judgment logic for each state variable, and extracting state variables that can reflect the overall state of the turbine governor hydraulic system.

[0007] Furthermore, the normalization process for each state variable constructs a state vector representing the health state of the hydraulic system of the turbine governor. Multiple state vectors form a state vector set, including: Based on the minimax method, each state variable is linearly normalized. If there are n state variables and each state variable has m data points, then the first state vector will be an n-dimensional vector.

[0008] A set of state vectors is composed of multiple state vectors:

[0009] Right now: .

[0010] Furthermore, the mean vector of the set of state vectors of the hydraulic system of the turbine governor under normal conditions is used to obtain the center vector of the healthy state, including:

[0011] in, Let the health status center vector be... It is the nth mean vector of the set of state vectors.

[0012] Furthermore, the calculation of the weighted Euclidean distance between each state vector in the set of state vectors and the center vector of the healthy state includes: calculating the squared difference between each state vector in the set of state vectors under the operating conditions of the hydraulic system of the turbine governor and the center vector of the healthy state, multiplying it by the weight of the corresponding state vector, and then summing all the results and taking the square root to obtain the weighted Euclidean distance.

[0013] Furthermore, the calculation of the average and standard deviation of n consecutive weighted Euclidean distances as fault alarm indicators of the turbine governor hydraulic system includes: calculating the average value MEAN and standard deviation STD of n consecutive weighted Euclidean distances, and using MEAN+3*STD as the fault alarm indicator of the turbine governor hydraulic system.

[0014] Furthermore, when one or more real-time status warning indicators of the turbine governor hydraulic system are abnormal, the overall system operating status is determined to be in a state of alert; when the fault alarm indicator exceeds the threshold and one or more real-time status warning indicators are abnormal, or when three or more real-time status warning indicators are abnormal, the overall system operating status is determined to be in an abnormal state.

[0015] A hydraulic system status early warning device for a water turbine governor, comprising: The first calculation module is configured to collect key parameters of the hydraulic system of the turbine governor, process them to obtain various state quantities associated with the state of the hydraulic system of the turbine governor, and assign corresponding weights according to the degree of influence of each state quantity on the operating state. The second calculation module is configured to normalize each state variable, construct a state vector representing the healthy state of the hydraulic system of the turbine governor, and form a set of state vectors from multiple state vectors; calculate the mean vector of the set of state vectors under the normal state of the hydraulic system of the turbine governor to obtain the center vector of the healthy state. The status early warning module is configured to calculate the weighted Euclidean distance between each state vector in the state vector set and the center vector of the healthy state, as a real-time status early warning indicator of the turbine governor hydraulic system; and to calculate the average and standard deviation of n consecutive weighted Euclidean distances, as a fault alarm indicator of the turbine governor hydraulic system.

[0016] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the hydraulic system status early warning method for a turbine governor.

[0017] A computer-readable storage medium storing a computer program that, when executed by a processor, implements a method for early warning of the hydraulic system status of a turbine governor.

[0018] The beneficial effects of this invention are as follows: (1) In view of the shortcomings of the current fault early warning method of the hydraulic system of the turbine governor of the hydropower unit, the present invention assigns weight coefficients to the importance of each key parameter. Based on the Euclidean space clustering model and Gaussian threshold, the weighted Euclidean distance and Gaussian threshold method are combined to conduct real-time risk assessment and fault trend early warning of the operating status of the hydraulic system of the turbine governor.

[0019] (2) This invention utilizes the historical health status monitoring signal of the hydraulic system of the turbine governor, fully considers the complexity of its key parameter monitoring values ​​and the variability of its operating conditions, designs the judgment logic of each state monitoring quantity, and extracts the state monitoring quantity that can reflect the overall state of the system.

[0020] (3) This invention makes full use of the massive historical data accumulated by the unit monitoring system, incorporates the concept of big data, and can use the historical data of the unit for self-learning to update the early warning model of the hydraulic system of the turbine governor and improve the accuracy of the early warning.

[0021] (4) By analyzing the real-time early warning status of the hydraulic system of the turbine governor, before a fault occurs, most of the operating information, health status and development trend of the hydraulic system of the turbine governor can be grasped, guiding the operation and maintenance personnel to make adjustments, extend the equipment operation cycle, ensure the effective performance of the entire hydropower station equipment, minimize the probability of accidents, prevent faults from occurring or reduce the impact and consequences of faults; provide a basis for maintenance decisions, make the purpose of maintenance work clear and the methods scientific, shorten maintenance time, and at the same time provide a basis for the formulation of a reasonable inspection and maintenance system, which is a guarantee for the safe and reliable operation of the hydropower station and the acquisition of great economic and social benefits. Attached Figure Description

[0022] Figure 1 This is a flowchart of a hydraulic system status early warning method for a turbine governor according to Embodiment 1 of the present invention.

[0023] Figure 2 This is a flowchart of a hydraulic system status early warning method for a turbine governor according to Embodiment 2 of the present invention. Detailed Implementation

[0024] To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific embodiments are now described. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention; that is, the described embodiments are only a part of the embodiments of the invention, not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0025] Example 1 like Figure 1 As shown, this embodiment provides a method for early warning of the status of a hydraulic system of a turbine governor, including: Key parameters of the hydraulic system of the turbine governor are collected, processed to obtain various state variables associated with the state of the hydraulic system of the turbine governor, and assigned corresponding weights according to the degree of influence of each state variable on the operating state. Normalize each state variable to construct a state vector representing the healthy state of the hydraulic system of the turbine governor. Multiple state vectors form a state vector set. Calculate the mean vector of the state vector set under normal state of the hydraulic system of the turbine governor to obtain the center vector of the healthy state. The weighted Euclidean distance between each state vector in the state vector set and the center vector of the healthy state is calculated as a real-time state early warning indicator of the hydraulic system of the turbine governor; the average value and standard deviation of n consecutive weighted Euclidean distances are calculated as fault alarm indicators of the hydraulic system of the turbine governor.

[0026] It should be noted that this method, by screening key parameters and assigning weights, can highlight factors that significantly affect the system state and avoid irrelevant parameters interfering with subsequent analysis; by eliminating the influence of dimensions through normalization, different types of state quantities can be analyzed uniformly, and the health state center vector provides a clear benchmark for the normal state of the system; real-time status early warning indicators can dynamically reflect the deviation between the current state and the healthy state of the system, and fault alarm indicators take into account both the overall trend and fluctuation of the data, effectively improving the timeliness and accuracy of early warning.

[0027] Preferably, key parameters of the turbine governor hydraulic system are collected and processed to obtain various state variables associated with the state of the turbine governor hydraulic system. This includes: collecting historical data of key parameters of the turbine governor hydraulic system; based on the complexity of the monitoring values ​​and operating conditions of the turbine governor hydraulic system, and combining the correlation between the state variables and the operating state of the turbine governor hydraulic system, setting judgment logic for each state variable, and extracting state variables that can reflect the overall state of the turbine governor hydraulic system.

[0028] Specifically, the process begins by comprehensively collecting historical data on key parameters of the turbine governor's hydraulic system under different operating stages and conditions. This historical data is then categorized, organized, and validated, with abnormal and invalid data removed. Next, the complexity of the system's monitored values ​​is analyzed to clarify the changing patterns and interrelationships of each value. Furthermore, considering common system operating conditions, the characteristics of parameter changes under different conditions are analyzed. Finally, based on the close correlation between state variables and system operating status, a targeted judgment logic is established. This logic is used to filter and transform the processed historical data, removing parameters with weak correlations and retaining and extracting state variables that comprehensively and accurately reflect the overall operating status of the system.

[0029] It should be noted that the above steps utilize historical data to provide sufficient basis for the extraction of state variables, ensuring that the state variables have historical relevance and practicality; the judgment logic is set by combining the complexity of monitoring values, operating conditions and relevance, making the extracted state variables more targeted and effective, and able to accurately capture the core characteristics of the overall system state; at the same time, irrelevant parameters are eliminated, reducing the workload of subsequent data processing and improving the operating efficiency of the early warning method.

[0030] Preferably, each state variable is normalized to construct a state vector representing the health state of the hydraulic system of the turbine governor. Multiple state vectors form a state vector set, including: Based on the minimax method, each state variable is linearly normalized. If there are n state variables and each state variable has m data points, then the first state vector will be an n-dimensional vector.

[0031] A set of state vectors is composed of multiple state vectors:

[0032] Right now: .

[0033] Specifically, for each extracted state variable, linear normalization is performed using the minimax method to unify the value range of each state variable to the same interval, eliminating the impact of differences in dimensions and magnitudes. Based on the total number of state variables and the number of valid data points corresponding to each state variable, the normalized values ​​of each state variable corresponding to each data point are arranged in a preset order, forming a multidimensional state vector with the same dimension as the number of state variables. Multiple multidimensional state vectors generated by the system under different time periods and operating conditions during normal operation are collected and integrated according to set rules to form a state vector set that comprehensively covers the healthy operating state of the system.

[0034] It should be noted that the maximum-minimum method is simple to operate and can effectively eliminate differences in dimensions and magnitudes, making different types of state variables comparable and operable. The multidimensional state vector can centrally reflect the comprehensive health status of the system under a single data point, while the vector set comprehensively covers various state scenarios of normal system operation, laying a unified and reliable data foundation for subsequent calculation of the health status center vector and state deviation analysis.

[0035] Preferably, the mean vector of the set of state vectors of the hydraulic system of the turbine governor under normal conditions is calculated to obtain the center vector of the healthy state, including:

[0036] in, Let the health status center vector be... It is the nth mean vector of the set of state vectors.

[0037] Specifically, firstly, a set of state vectors generated under normal system operation is selected, ensuring that all vectors in the set originate from fault-free and stable system operation. Then, statistical analysis is performed on all vectors in the state vector set, calculating the mean value of all vectors corresponding to each dimension (i.e., the dimension corresponding to each state variable). Finally, the means of each dimension are arranged in dimensional order to form a mean vector consistent with the dimensions of the state vectors. This mean vector is the health state center vector representing the baseline of normal system operation.

[0038] It should be noted that calculating the mean vector based on the vector set of normal states ensures that the health state center vector accurately reflects the typical state of normal system operation and has reliable benchmark properties. The mean calculation of each dimension can comprehensively reflect the average level of the state quantity under normal operation, making each dimension of the center vector representative. This provides a clear and unified comparison benchmark for subsequent judgment of the deviation between the real-time state vector and the normal state, improving the accuracy of deviation judgment.

[0039] Preferably, calculating the weighted Euclidean distance between each state vector in the state vector set and the center vector of the healthy state includes: calculating the squared difference between each state vector in the state vector set under the operating conditions of the hydraulic system of the turbine governor and the center vector of the healthy state, multiplying it by the weight of the corresponding state vector, summing all the results and taking the square root to obtain the weighted Euclidean distance.

[0040] Specifically, first, obtain the real-time state vector generated under the current system operating conditions, and the pre-determined health state center vector. Then, calculate the difference between the real-time state vector and the health state center vector along their corresponding dimensions, and square the difference for each dimension. Multiply the squared result of each dimension by the corresponding state variable weight to obtain the weighted squared difference for each dimension. Summate the weighted squared differences for all dimensions, and then take the square root of the sum. The final result is the weighted Euclidean distance between the real-time state vector and the health state center vector.

[0041] It should be noted that calculating the squared difference can amplify the degree of deviation from the normal state, making the deviation easier to identify. The introduction of weights can highlight the influence of key state quantities, making the distance calculation more in line with the actual operating characteristics of the system. The weighted Euclidean distance can comprehensively reflect the overall deviation between the real-time state and the normal state in all dimensions, accurately quantifying the degree of difference between the current operating state and the healthy state of the system. It provides a scientific and reasonable quantitative basis for real-time state early warning indicators, ensuring the accuracy and pertinence of real-time early warning.

[0042] Preferably, the average and standard deviation of n consecutive weighted Euclidean distances are calculated as fault alarm indicators of the hydraulic system of the turbine governor, including: calculating the average value MEAN and standard deviation STD of n consecutive weighted Euclidean distances, and using MEAN+3*STD as the fault alarm indicator of the hydraulic system of the turbine governor.

[0043] Specifically, weighted Euclidean distance data generated during system operation is continuously recorded, and multiple consecutive weighted Euclidean distance data points are selected as analysis samples in chronological order. Statistical operations are performed on the selected consecutive weighted Euclidean distance samples to calculate the average value of the sample set to reflect the overall level of system state deviation during this period, and simultaneously calculate the standard deviation of the sample set to reflect the degree of fluctuation in deviation during this period. Using a preset combination method, the calculated average value and standard deviation are integrated to form a fault alarm indicator characterizing the system's fault risk, serving as the basis for judging whether the system has fault risk.

[0044] It should be noted that the average value reflects the overall trend of system state deviation, while the standard deviation reflects the fluctuation of deviation. The combination of the two allows the fault alarm indicator to take into account both the overall trend and local fluctuations. This indicator can effectively identify abnormal situations that exceed the normal fluctuation range, reduce false alarms caused by abnormal single data points, and avoid missing potential fault risks. It provides a reliable judgment standard for system fault early warning and improves the accuracy and stability of fault early warning.

[0045] Preferably, when one or more real-time status warning indicators of the turbine governor hydraulic system are abnormal, the overall operating status of the system is determined to be in a state of alert; when the fault alarm indicator exceeds the threshold and one or more real-time status warning indicators are abnormal, or when three or more real-time status warning indicators are abnormal, the overall operating status of the system is determined to be in an abnormal state.

[0046] Specifically, the system monitors various real-time status warning indicators generated by the real-time monitoring system, determines whether each indicator is within the normal range, and counts the number of abnormal real-time status warning indicators. Simultaneously, it monitors the values ​​of fault alarm indicators and determines whether they exceed preset thresholds. According to preset judgment rules, if one or more abnormal real-time status warning indicators are detected, the system is directly determined to enter a state of alert; if a fault alarm indicator exceeds the threshold and at least one real-time status warning indicator is abnormal, or if three or more abnormal real-time status warning indicators are detected, the system is determined to enter an abnormal state. Finally, corresponding status prompts are output for different judgment results, providing a basis for subsequent processing.

[0047] It should be noted that by clarifying the judgment conditions for different operating states, the state judgment has a clear rule basis, avoiding subjectivity and ambiguity; by distinguishing between the attention state and the abnormal state, the risk level-based early warning can be realized, which makes it easier for staff to take corresponding measures according to different risk levels and improves the practicality of the early warning; the combination of multiple conditions can effectively filter out minor and occasional anomalies, accurately identify the real fault risks, and reduce misoperation and oversight.

[0048] Example 2 This embodiment is based on embodiment 1: like Figure 2 As shown, this embodiment provides a method for early warning of the status of a hydraulic system of a turbine governor, including: Step 1: Collect key parameters a1, a2, a3...a1 of the hydraulic system of the turbine governor. n The process yields various state variables associated with the hydraulic system state of the turbine governor; Step 2: Based on the varying degrees of importance of each key parameter to the operational status, assign different weighting coefficients b1, b2, ... b to the aforementioned key parameters. n The total weights sum to 1; Step 3: Normalize each state variable, and then construct a state vector representing the health state of the hydraulic system of the turbine governor. Multiple state vectors form a set of state vectors. Step 4: Calculate the mean vector of the set of state vectors of the hydraulic system of the turbine governor under normal conditions, and obtain the center vector of the system health state; Step 5: Obtain the set of state vectors under the operating conditions of the hydraulic system of the turbine governor according to Steps 1-3. Calculate the weighted Euclidean distance between each state vector in the vector set and the center vector of the healthy state. Calculate the average (MEAN) and standard deviation (STD) of n consecutive weighted Euclidean distances. Use MEAN+3*STD as the fault alarm index of the hydraulic system of the turbine governor. Step 6: Following steps 1-3, obtain the set of state vectors under the operating conditions of the turbine governor hydraulic system. Calculate the weighted Euclidean distance between each state vector in the set and the center vector of the healthy state. That is, calculate the squared difference between each vector in the set of state vectors under the operating conditions of the turbine governor hydraulic system and the center vector of the healthy state, multiply it by the weight of the corresponding state vector parameter, sum all the results and take the square root. Then calculate the average value (MEAN) of n consecutive weighted Euclidean distances, which serves as the real-time status early warning indicator of the turbine governor hydraulic system.

[0049] Preferably, in order to calculate the various state quantities associated with the hydraulic system state of the turbine governor in step 1, it is first necessary to construct a real-time early warning system for the operating status of the hydraulic system of the turbine governor. By utilizing massive historical measurement data and fully considering the complexity of the monitoring values ​​and operating conditions of the hydraulic system of the turbine governor, and based on the strong correlation between the state monitoring quantities and the system operating status, the judgment logic of each state monitoring quantity is designed to extract the state monitoring quantities that can reflect the overall system status.

[0050] Preferably, in step 2, different weighting coefficients b1, b2...b are constructed to assign different values ​​to the key parameters based on their varying degrees of importance to the operational status. n ,b1+b2+.....+b n =1.

[0051] Preferably, the construction of the state vector representing the healthy state of the hydraulic system of the turbine governor in step 3 will be achieved by constructing a multi-dimensional vector: If the hydraulic system state monitoring system of the turbine governor constructed in step 1 contains n state monitoring quantities, and after density unification, each state monitoring quantity has m data points, then the first state vector will be an n-dimensional vector:

[0052] Preferably, the set of state vectors in steps 1 and 2 is as follows:

[0053] Right now

[0054] Preferably, in steps 3 and 4, the mean vector of the set of state vectors of the turbine governor hydraulic system under normal conditions is used as the center vector of the healthy state. : .

[0055] Preferably, the state vector set of the hydraulic system of the turbine governor under operating conditions is obtained according to steps 1 and 3. The weighted Euclidean distance between each vector in the vector set and the center vector of the healthy state is calculated. The weighted Euclidean distance of the first vector is = x + +...+ The weighted Euclidean distance of the nth vector = x + +...+ Then, the average value of n consecutive weighted Euclidean distances (MEAN) is calculated as a real-time status early warning indicator for the hydraulic system of the turbine governor.

[0056] Preferably, according to step 5, the average value (MEAN) and standard deviation (STD) of n consecutive weighted Euclidean distances are calculated, and MEAN+3*STD is used as the fault alarm index of the hydraulic system of the turbine governor.

[0057] Preferably, according to step 6, when one or more status monitoring quantities of the system exceed the real-time status warning indicator, the overall operating status of the system is determined to be "attention"; when the system fault warning indicator exceeds the threshold, and at the same time one or more status monitoring quantities exceed the real-time status warning indicator, or when three or more status monitoring quantities of the system exceed the real-time status warning indicator, the overall operating status of the system is determined to be "abnormal".

[0058] Example 3 This embodiment provides a status early warning device for the hydraulic system of a water turbine governor, including: The first calculation module is configured to collect key parameters of the hydraulic system of the turbine governor, process them to obtain various state quantities associated with the state of the hydraulic system of the turbine governor, and assign corresponding weights according to the degree of influence of each state quantity on the operating state. The second calculation module is configured to normalize each state variable, construct a state vector representing the healthy state of the hydraulic system of the turbine governor, and form a set of state vectors from multiple state vectors; calculate the mean vector of the set of state vectors under the normal state of the hydraulic system of the turbine governor to obtain the center vector of the healthy state. The status early warning module is configured to calculate the weighted Euclidean distance between each state vector in the state vector set and the center vector of the healthy state, as a real-time status early warning indicator of the turbine governor hydraulic system; and to calculate the average and standard deviation of n consecutive weighted Euclidean distances, as a fault alarm indicator of the turbine governor hydraulic system.

[0059] Specifically, after the first calculation module is started, it establishes a data connection with the monitoring equipment of the hydraulic system of the turbine governor to collect key parameters during the system operation; it performs preprocessing operations such as filtering and integration on the collected parameters to convert them into state variables related to the system state; then, through the built-in weight assignment logic, it assigns weights to each state variable according to the degree of influence of each state variable on the system operation, and transmits the processed state variables and weights to the second calculation module.

[0060] After receiving the data, the second calculation module processes each state variable using a preset normalization algorithm; it constructs a multi-dimensional state vector based on the number of state variables and the data distribution, collects multiple state vectors during normal system operation to form a state vector set; it performs statistical analysis on the set through built-in mean calculation logic to obtain the health state center vector, and transmits the set and center vector to the state early warning module.

[0061] After receiving data, the status early warning module calculates the weighted Euclidean distance between the current system status vector and the health status center vector in real time as a real-time status early warning indicator; at the same time, it continuously stores the continuously generated weighted Euclidean distance data, calculates its average value and standard deviation through the built-in statistical algorithm to form a fault alarm indicator; finally, it outputs two types of indicators to provide data support for system status determination.

[0062] It should be noted that this device adopts a modular design with clear division of labor among modules, improving stability and scalability; data transmission between modules is smooth, realizing fully automated operation from parameter acquisition and data processing to index calculation, reducing manual intervention and improving early warning efficiency; the device strictly follows preset logic and algorithms to ensure the accuracy of data processing and index calculation, providing reliable hardware and software support for the status early warning of the hydraulic system of the turbine governor.

[0063] Example 3 This embodiment is based on embodiment 1: This embodiment provides a computer device, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the hydraulic system status early warning method for a turbine governor as described in Embodiment 1. The computer program can be in the form of source code, object code, executable file, or some intermediate form.

[0064] Example 4 This embodiment is based on embodiment 1: This embodiment provides a computer-readable storage medium storing a computer program. When executed by a processor, this computer program implements the hydraulic system status early warning method for a turbine governor as described in Embodiment 1. The computer program can be in the form of source code, object code, executable file, or some intermediate form. The storage medium includes any entity or device capable of carrying computer program code, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0065] The above description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the concept described herein through the above teachings or related technologies or knowledge. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

[0066] It should be noted that, for the sake of simplicity, the foregoing method embodiments are described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

Claims

1. A method for early warning of the status of a hydraulic system of a turbine governor, characterized in that, include: Key parameters of the hydraulic system of the turbine governor are collected, processed to obtain various state variables associated with the state of the hydraulic system of the turbine governor, and assigned corresponding weights according to the degree of influence of each state variable on the operating state. Normalize each state variable to construct a state vector representing the healthy state of the hydraulic system of the turbine governor. Multiple state vectors form a state vector set. Calculate the mean vector of the state vector set under normal state of the hydraulic system of the turbine governor to obtain the center vector of the healthy state. The weighted Euclidean distance between each state vector in the state vector set and the center vector of the healthy state is calculated as a real-time state early warning indicator of the hydraulic system of the turbine governor; the average value and standard deviation of n consecutive weighted Euclidean distances are calculated as fault alarm indicators of the hydraulic system of the turbine governor.

2. The method for early warning of the hydraulic system status of a turbine governor according to claim 1, characterized in that, The process involves collecting key parameters of the turbine governor's hydraulic system and processing them to obtain various state variables associated with the state of the turbine governor's hydraulic system. This includes: collecting historical data of key parameters of the turbine governor's hydraulic system; based on the complexity of the monitoring values ​​and operating conditions of the turbine governor's hydraulic system, and combining the correlation between the state variables and the operating state of the turbine governor's hydraulic system, setting judgment logic for each state variable, and extracting state variables that can reflect the overall state of the turbine governor's hydraulic system.

3. The method for early warning of the hydraulic system status of a turbine governor according to claim 1, characterized in that, The normalization process for each state variable constructs a state vector representing the health state of the hydraulic system of the turbine governor. Multiple state vectors form a state vector set, including: Based on the minimax method, each state variable is linearly normalized. If there are n state variables and each state variable has m data points, then the first state vector will be an n-dimensional vector. A set of state vectors is composed of multiple state vectors: Right now: 。 4. The method for early warning of the hydraulic system status of a turbine governor according to claim 3, characterized in that, The mean vector of the set of state vectors of the hydraulic system of the turbine governor under normal conditions is used to obtain the health state center vector, including: in, Let the health status center vector be... It is the nth mean vector of the set of state vectors.

5. The method for early warning of the hydraulic system status of a turbine governor according to claim 1, characterized in that, The calculation of the weighted Euclidean distance between each state vector in the set of state vectors and the center vector of the healthy state includes: calculating the squared difference between each state vector in the set of state vectors under the operating conditions of the hydraulic system of the turbine governor and the center vector of the healthy state, multiplying it by the weight of the corresponding state vector, summing all the results and taking the square root to obtain the weighted Euclidean distance.

6. The method for early warning of the hydraulic system status of a turbine governor according to claim 1, characterized in that, The calculation of the average and standard deviation of n consecutive weighted Euclidean distances as fault alarm indicators of the hydraulic system of the turbine governor includes: calculating the average value MEAN and standard deviation STD of n consecutive weighted Euclidean distances, and using MEAN+3*STD as the fault alarm indicator of the hydraulic system of the turbine governor.

7. The method for early warning of the hydraulic system status of a turbine governor according to claim 1, characterized in that, When one or more real-time status warning indicators of the turbine governor hydraulic system are abnormal, the overall system operating status is determined to be in a state of alert; when the fault alarm indicator exceeds the threshold and one or more real-time status warning indicators are abnormal, or when three or more real-time status warning indicators are abnormal, the overall system operating status is determined to be in an abnormal state.

8. A status early warning device for the hydraulic system of a water turbine governor, characterized in that, include: The first calculation module is configured to collect key parameters of the hydraulic system of the turbine governor, process them to obtain various state quantities associated with the state of the hydraulic system of the turbine governor, and assign corresponding weights according to the degree of influence of each state quantity on the operating state. The second calculation module is configured to normalize each state variable, construct a state vector representing the healthy state of the hydraulic system of the turbine governor, and form a set of state vectors from multiple state vectors; calculate the mean vector of the set of state vectors under the normal state of the hydraulic system of the turbine governor to obtain the center vector of the healthy state. The status early warning module is configured to calculate the weighted Euclidean distance between each state vector in the state vector set and the center vector of the healthy state, as a real-time status early warning indicator of the turbine governor hydraulic system; and to calculate the average and standard deviation of n consecutive weighted Euclidean distances, as a fault alarm indicator of the turbine governor hydraulic system.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the hydraulic system status early warning method for the turbine governor as described in any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the hydraulic system status early warning method for the turbine governor as described in any one of claims 1-7.

Images