Method and system for evaluating safety degree of multi-machine system of water-turbine generator set

Abstract

The invention relates to the technical field of water-turbine generator set safety monitoring, in particular to a water-turbine generator set multi-machine system safety degree evaluation method and system, and the method comprises the steps: obtaining corresponding evaluation indexes according to the change conditions of various types of data, and training a neural network model through the evaluation indexes of each set in all transition processes; obtaining a first overall safety degree of the water-turbine generator set through the trained neural network model; constructing a fuzzy relation matrix according to the value of each evaluation index of each unit in each transition process and a fuzzy evaluation range corresponding to each safety level, obtaining a fuzzy entropy evaluation matrix through information entropy of membership degrees of all safety levels of each evaluation index, and obtaining a second overall safety degree of the water-turbine generator set; and the final safety degree of the multi-machine system of the water-turbine generator set is obtained through the first overall safety degree and the second overall safety degree. According to the method, the accuracy of safety degree evaluation of the multi-machine system of the water-turbine generator set is improved.

Description

Technical Field

[0001] The invention relates to the technical field of safety monitoring of a hydro-turbine generator set, and in particular to a safety evaluation method and system for a multi-machine system of a hydro-turbine generator set. Background Art

[0002] Under the background of green and low-carbon development, hydropower is often used as a regulating resource for the power grid to participate in the peak load regulation and frequency regulation process. The frequent transition process in this process will lead to stability problems such as unit vibration, swing and water pressure fluctuation. The number of operating risks and the degree of loss of various components of the hydropower unit are also different, which seriously affects the operating life of the hydropower generator unit. As a complex nonlinear power system, the hydropower generator unit has many types of faults and complex impacts.

[0003] In the prior art, risk assessment is usually conducted on a single important component in a hydro-turbine generator set; however, there are still certain challenges in comprehensively and holistically evaluating the safety of a hydro-turbine generator set; and hydro-turbine generator sets have obvious time-varying characteristics, while existing research methods for risk assessment of hydro-turbine generator sets are mostly based on static models, which make it difficult to handle the correlation between various data indicators of the hydro-turbine generator set that change over time. As a result, the accuracy of the predictive assessment of the safety of a multi-machine system of a hydro-turbine generator set is insufficient, reducing the accuracy of the safety assessment of the multi-machine system. Summary of the invention

[0004] The present invention provides a method and system for evaluating the safety of a multi-machine system of a hydro-generator set, so as to solve the existing problems.

[0005] A method and system for evaluating the safety of a multi-machine system of a hydro-generator set of the present invention adopts the following technical solutions: A first aspect of the present invention is to provide a method for evaluating the safety of a multi-machine system of a hydro-generator set, the method comprising the following steps: Obtain various types of data for each unit during all transitions during the simulation of the hydro-generator unit; According to the changes of various types of data in each transition process of each unit, the evaluation index in each transition process of each unit is obtained, a sample set is constructed through the evaluation indexes in all transition processes of each unit, and the neural network model is trained through the sample set and the artificially calibrated safety value to obtain the trained neural network model; according to the sample set to be analyzed at the current moment and the trained neural network model, the first overall safety degree of the hydro-generator unit is obtained; Each evaluation index is divided into several safety levels, and a fuzzy evaluation range of each safety level is obtained. A fuzzy relationship matrix is ​​constructed according to the value of each evaluation index of each unit in each transition process and the fuzzy evaluation range corresponding to each safety level, and the elements in the fuzzy relationship matrix are recorded as membership degrees; the fuzzy relationship matrix is ​​adjusted according to the information entropy of the membership degrees of all safety levels of each evaluation index in the fuzzy relationship matrix; the second overall safety degree of the hydro-generator unit is obtained through the data distribution in the adjusted matrix; The final safety degree of the multi-machine system of the hydro-generator set is obtained through the first overall safety degree and the second overall safety degree of the hydro-generator set.

[0006] Furthermore, the various types of data include: water head value of the water diversion pipeline, water head value of the tailwater pipeline, motor output power, terminal voltage, excitation voltage and frequency.

[0007] Furthermore, the step of obtaining the evaluation index of each unit in each transition process according to the change of various types of data in each transition process of each unit includes the following specific steps: The evaluation indexes include: frequency fluctuation index, power fluctuation index, voltage amplitude index, water head fluctuation index of water diversion pipeline and water head fluctuation index of tailwater pipeline; The construction process of the frequency fluctuation index is as follows: Will The frequency within the interval is recorded as the limited frequency. The frequencies outside the interval are recorded as out-of-limit frequencies, where: To preset the reference frequency, is the preset frequency error; Among them, the data in each transition process of each unit was obtained with a sampling interval of 1 second; Among them, the formula corresponding to the frequency fluctuation index is:

[0008] In the formula, It represents the mean value of the out-of-limit frequency at all time points in each transition process of each unit, It represents the mean value of the frequency within the limit at all time points in each transition process of each unit, represents the maximum frequency corresponding to all time points in each transition process of each unit, where , , They are , , The weight of Represents the frequency fluctuation index in each transition process of each unit; The formula corresponding to the power fluctuation index is:

[0009] In the formula, Indicates the maximum motor output power of each unit in each transition process, represents the initial motor output power of each unit during each transition process, represents the final motor output power of each unit during each transition process, Represents the power fluctuation index of each unit in each transition process; The formula corresponding to the voltage amplitude index is:

[0010] In the formula, Indicates the maximum value of the terminal voltage of each unit during each transition process. Indicates the initial terminal voltage of each unit in each transition process, Indicates the maximum value of the excitation voltage in each transition process of each unit. Indicates the initial excitation voltage of each unit in each transition process, Indicates the voltage amplitude index in each transition process of each unit; The reference peaks of the water head value of the water diversion pipe and the water head value of the tailwater pipe of each unit in each transition process during the simulation time are obtained, and the water head fluctuation index of the water diversion pipe and the water head fluctuation index of the tailwater pipe are quantified by the fluctuation of the reference peaks of the water head value of each unit in each transition process during the simulation time.

[0011] Further, the reference peaks of the water head value of the water diversion pipeline and the water head value of the tailwater pipeline of each unit in each transition process during the simulation time are obtained, and the water head fluctuation index of the water diversion pipeline and the water head fluctuation index of the tailwater pipeline are quantified by the fluctuation of the reference peak of the water head value of each unit in each transition process during the simulation time, and the specific steps include the following: The specific method for obtaining the reference peak is: Select the starting point of each unit in each transition process during the simulation time. The water pipe head value at the peak of the wave is The mean of the water pipe head values ​​at the peaks is recorded as the front water head mean; then the back water head of each unit at the end of each transition process during the simulation time is obtained. The water pipe head value at the peak of the wave is The mean of the water pipe head values ​​at the peaks is recorded as the mean water head after the simulation. If the mean water head before and after the simulation is greater than the mean water head after the simulation, the mean water head before and after the simulation is recorded as the mean water head after the simulation. The peaks are taken as reference peaks; if the current water head mean is less than the latter water head mean, the latter water head of each unit at the end of each transition process during the simulation time is taken as the reference peak. peaks as reference peaks; among them, is the preset quantity parameter; The reference peak value of the water head value of the water diversion pipeline and the reference peak value of the tailwater pipeline of each unit in each transition process during the simulation time are obtained by the above-mentioned reference peak acquisition method; The formula corresponding to the water head fluctuation index of the water diversion pipeline is:

[0012] In the formula, Indicates that each unit has The water head value of the water diversion pipeline of the reference wave peak, Indicates the water head value of the water supply pipe that is closest to 0 in each unit during each transition process. It indicates the water head fluctuation index of the water diversion pipeline of each unit in each transition process; The formula corresponding to the tailwater pipe head fluctuation index is:

[0013] In the formula, Indicates that each unit has the first The tailwater pipe head value of the reference wave peak, Indicates the tailwater pipe head value that is closest to 0 for each unit in each transition process. It represents the tailwater pipe head fluctuation index of each unit during each transition process.

[0014] Furthermore, the step of constructing a sample set by using the evaluation indicators of all transition processes of each unit includes the following specific steps: Normalize all evaluation indicators of each unit in each transition process to obtain normalized results; obtain the normalized evaluation indicators of all transition processes in the historical simulation process, and use the normalized evaluation indicators of all transition processes in each simulation process as a sample set, wherein the sample set is an M×H matrix, wherein M is the total number of evaluation indicators, and H is the total number of all transition processes in each simulation process.

[0015] Furthermore, the first overall safety degree of the hydro-generator set is obtained according to the sample set to be analyzed at the current moment and the trained neural network model, and the specific steps include the following: The sample set that needs to be analyzed at the current moment is obtained, and the sample set that needs to be analyzed at the current moment is input into the trained BiLSTM neural network to obtain the initial safety degree of each unit; according to the initial safety degree of all units, the fuzzy entropy weight method is used to obtain the first overall safety degree of the hydro-generator unit.

[0016] Furthermore, each evaluation index is divided into several safety levels, a fuzzy evaluation range of each safety level is obtained, and a fuzzy relationship matrix is ​​constructed according to the value of each evaluation index of each unit in each transition process and the fuzzy evaluation range corresponding to each safety level. The specific steps include the following: The fuzzy evaluation range of the six safety levels corresponding to each evaluation index of each unit in each transition process is obtained by expert experience method or historical data method, and the membership degree corresponding to each safety level is obtained by the membership function according to the value of each evaluation index; among which, the six safety levels include stable, relatively stable, critically stable, weak risk, medium risk, and strong risk; With all evaluation indicators as rows and all safety levels as columns, a fuzzy relationship matrix is ​​obtained.

[0017] Furthermore, the fuzzy relationship matrix is ​​adjusted by the information entropy of the membership of all safety levels of each evaluation index in the fuzzy relationship matrix; and the second overall safety degree of the hydro-generator set is obtained by the data distribution in the adjusted matrix, which includes the following specific steps: Calculate the entropy value of the membership degree corresponding to all safety levels of each evaluation indicator, construct the entropy values ​​of all evaluation indicators into an entropy matrix with 1 row and M columns, and multiply the entropy matrix with the fuzzy relationship matrix to obtain a matrix, which is recorded as the fuzzy entropy evaluation matrix, where the fuzzy entropy evaluation matrix has 1 row and N columns; Perform linear normalization on the values ​​in the fuzzy entropy evaluation matrix; According to the maximum membership principle, the maximum value of the N values ​​in the fuzzy entropy evaluation matrix is ​​obtained and recorded as the reference safety degree of each unit in each transition process. The average value of the reference safety degree of each unit in all transition processes is recorded as the reference safety degree of each unit. According to the reference safety degree of all units, the second overall safety degree of the hydro-turbine generator set is obtained by the fuzzy entropy weight method.

[0018] Furthermore, the final safety degree of the multi-machine system of the hydro-generator set is obtained by using the first overall safety degree and the second overall safety degree of the hydro-generator set, and the specific steps include the following: The average of the first overall safety degree and the second overall safety degree of the hydro-turbine generator set is taken as the reference overall safety degree of the hydro-turbine generator set. The reference overall safety degree of the hydro-turbine generator set is normalized by the standard normalization method to obtain the final safety degree of the multi-machine system of the hydro-turbine generator set.

[0019] The second aspect of the present invention is to provide a multi-machine system safety assessment system for a hydro-generator set, the system comprising the following modules: Data acquisition module: used to collect various types of data of each unit in all transition processes during the simulation of the hydro-generator unit; Data analysis module: used to analyze and obtain the first overall safety degree of the hydro-generator set according to the neural network model, and to analyze and obtain the second overall safety degree of the hydro-generator set through fuzzy theory; Safety evaluation module: used to obtain the final safety of the multi-machine system of the hydro-generator set through the first overall safety and the second overall safety of the hydro-generator set.

[0020] The beneficial effects of the technical solution of the present invention are: obtaining the evaluation index of each unit in each transition process according to the change of various types of data in each transition process of each unit, constructing a sample set through the evaluation index of all transition processes of each unit, training the neural network model through the sample set and the artificially calibrated safety value to obtain the trained neural network model; obtaining the first overall safety of the hydro-turbine generator set according to the sample set and the trained neural network model that need to be analyzed at the current moment; improving the accuracy of the analysis of the change of the hydro-turbine generator set over time; and according to the value of each evaluation index of each unit in each transition process. A fuzzy relationship matrix is ​​constructed according to the fuzzy evaluation range corresponding to each safety level, and the elements in the fuzzy relationship matrix are recorded as membership degrees; the fuzzy relationship matrix is ​​adjusted according to the information entropy of the membership degrees of all safety levels of each evaluation indicator in the fuzzy relationship matrix to obtain a fuzzy entropy evaluation matrix; the second overall safety degree of the hydro-generator set is obtained through the fuzzy entropy evaluation matrix; the accuracy of complex, fuzzy and uncertain analysis of the hydro-generator set is improved; the final safety degree of the multi-machine system of the hydro-generator set is obtained through the first overall safety degree and the second overall safety degree of the hydro-generator set, and the accuracy of the safety degree evaluation of the multi-machine system of the hydro-generator set is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1 A flow chart of the steps of a method for evaluating the safety of a multi-machine system of a hydro-generator set according to the present invention; Figure 2 This is a flow chart for safety assessment of multi-machine system of hydro-generator set; Figure 3 The present invention is a module flow chart of a multi-machine system safety assessment system for a hydro-generator set. DETAILED DESCRIPTION

[0023] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a method and system for assessing the safety of a multi-machine system of a hydro-generator set proposed by the present invention, its specific implementation method, structure, features and effects, in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0024] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0025] The following is a detailed description of a method and system for assessing the safety of a multi-machine system of a hydro-turbine generator set provided by the present invention in conjunction with the accompanying drawings.

[0026] like Figure 1 As shown, the first aspect of the present invention is to provide a method for evaluating the safety of a multi-machine system of a hydro-generator set, the method comprising the following steps: Step S001: Collect various types of data of each unit in all transition processes during the simulation of the hydro-generator unit.

[0027] It should be noted that during the power generation process, the hydropower generator set will experience a transition process of starting, shutting down or load change. Since the operating status of the unit in different transition processes is very different and each transition process contains more detailed information, the safety of the hydropower generator set is analyzed through the data changes in each transition process; therefore, it is necessary to collect various types of data for each transition process for analysis.

[0028] It should be further explained that the transition process refers to the process in which the unit operating state transitions from one stable operating point to another stable operating point when the hydro-turbine generator unit is started, shut down or the load changes. Therefore, a complete simulation process involves multiple transition processes, and there is a timeline between multiple transition processes.

[0029] Specifically, the present invention builds a real-time simulation model of a multi-machine system of a hydro-turbine generator set based on RTDS (real-time digital simulation system) software. By coupling the RTDS model with the hydraulic model simulated by the fourth-order Runge-Kutta method, a complete simulation framework of a multi-machine system of a hydro-turbine generator set is established. The coupled model can simulate the transition process of the two-stage guide vane opening control strategy, and then output the key operating parameters of each unit in the multi-machine system, including the water head value of the water diversion pipeline, the water head value of the tailwater pipeline, the motor output power, the terminal voltage, the excitation voltage and frequency, etc. Among them, the RTDS model, the fourth-order Runge-Kutta method, the hydraulic model and the two-stage guide vane opening control strategy are all well-known technologies and will not be described in detail here.

[0030] During the simulation, for each transition process, m equally spaced guide vane opening inflection points are selected, and a series of system parameters are calculated and output according to the time and opening value of the guide vane opening inflection points. When the guide vane opening value is known, the hydraulic machinery parameters such as the flow rate, speed and turbine torque of the unit can be obtained by the turbine full characteristic curve interpolation method. Subsequently, these parameters are substituted into the generalized basic equations of hydraulic machinery, and the frequency, pipe head value and other parameters of the unit are further calculated using the internal characteristic method. The system voltage and power are directly output through RTDS simulation based on the full electromagnetic transient model.

[0031] In this embodiment, m is a preset parameter, and in this embodiment, m=12 is taken as an example for analysis; in this embodiment, the preset parameter m is not specifically limited, and the implementer may determine it according to the specific situation.

[0032] Among them, the turbine full characteristic curve interpolation method, generalized basic equation, internal characteristic method and full electromagnetic transient model are all well-known technologies and will not be described in detail here.

[0033] In the above discrete data collection process, data is collected with a sampling interval of 1 second, wherein the data sampling interval is not specifically limited.

[0034] At this point, various types of data of the hydro-generator set in each transition process are obtained; wherein the various types of data include: water head value of the water diversion pipeline, water head value of the tailwater pipeline, motor output power, terminal voltage, excitation voltage and frequency.

[0035] Step S002: Obtain evaluation indicators for each transition process of each unit according to changes in various types of data in each transition process of each unit, construct a sample set through the evaluation indicators in all transition processes of each unit, train the neural network model through the sample set and the artificially calibrated safety value, and obtain the trained neural network model; obtain the first overall safety of the hydro-turbine generator set according to the sample set that needs to be analyzed at the current moment and the trained neural network model.

[0036] It should be noted that during the operation of the hydro-turbine generator set, when the various types of data of the hydro-turbine generator set are relatively stable, it means that the safety of the hydro-turbine generator set is higher, and vice versa, it means that the safety of the hydro-turbine generator set is lower. Therefore, the corresponding evaluation index is quantified according to the fluctuation of each type of data, and the safety of the hydro-turbine generator set is evaluated through the quantitative evaluation index.

[0037] Specifically, five evaluation indicators corresponding to each transition process of each unit are constructed based on various types of data in each transition process of each unit according to multiple units of all units in a complete simulation process, wherein the five evaluation indicators include: frequency fluctuation index, power fluctuation index, voltage amplitude index, water head fluctuation index of water diversion pipeline, and water head fluctuation index of tailwater pipeline. In this embodiment, five evaluation indicators are used as an example for description, but are not specifically limited. The implementer can determine according to specific scenarios and situations.

[0038] The construction process of the frequency fluctuation index is as follows: Preset a reference frequency , in this embodiment, the reference frequency Take 50 Hz as an example for explanation. In this embodiment, the preset reference frequency There is no specific limitation and implementers can decide based on specific circumstances.

[0039] Preset a frequency error , in this embodiment the frequency error Take 0.2 Hz as an example for explanation, where in this embodiment, the preset frequency error There is no specific limitation and implementers can decide based on specific circumstances.

[0040] Will The frequency within the interval is recorded as the limited frequency. Frequencies outside the interval are recorded as out-of-limit frequencies.

[0041] Among them, the formula corresponding to the frequency fluctuation index is:

[0042] In the formula, It represents the mean value of the out-of-limit frequency at all time points in each transition process of each unit, It represents the mean value of the frequency within the limit at all time points in each transition process of each unit, represents the maximum frequency corresponding to all time points in each transition process of each unit, where , , They are , , In this embodiment, the weight is not specifically limited. Represents the frequency fluctuation index of each unit during each transition process.

[0043] The formula corresponding to the power fluctuation index is:

[0044] In the formula, Indicates the maximum motor output power of each unit in each transition process, represents the initial motor output power of each unit during each transition process, represents the final motor output power of each unit during each transition process, Represents the power fluctuation index of each unit during each transition process.

[0045] The formula corresponding to the voltage amplitude index is:

[0046] In the formula, Indicates the maximum value of the terminal voltage of each unit during each transition process. Indicates the initial terminal voltage of each unit in each transition process, Indicates the maximum value of the excitation voltage in each transition process of each unit. Indicates the initial excitation voltage of each unit in each transition process, Represents the voltage amplitude index of each unit during each transition process.

[0047] It should be noted that when analyzing the fluctuation of water pipe head value, due to its different fluctuations at different times, generally when the turbine generator set is started or shut down, the water pipe head value changes greatly, and the data with larger changes can better reflect the fluctuation of water pipe head value. Therefore, the data with larger peak fluctuation in the water pipe head value signal curve is first selected as the data for evaluation index analysis for analysis.

[0048] It should be further explained that, since it is unknown which water pipe head value fluctuates more at the beginning and the end, the changes in the peaks of the beginning data and the end data are analyzed before selection.

[0049] Specifically, the method for obtaining the reference peak is as follows: select the front peak of each unit at the beginning of each transition process during the simulation time. The water pipe head value at the peak of the wave is The mean of the water pipe head values ​​at the peaks is recorded as the front water head mean; then the back water head of each unit at the end of each transition process during the simulation time is obtained. The water pipe head value at the peak of the wave is The mean of the water pipe head values ​​at the peaks is recorded as the mean water head after the simulation. If the mean water head before and after the simulation is greater than the mean water head after the simulation, the mean water head before and after the simulation is recorded as the mean water head after the simulation. The peaks are taken as reference peaks; if the current water head mean is less than the subsequent water head mean, the subsequent water head of each unit at the end of each transition process during the simulation time is taken as the reference peak. peaks as reference peaks; among them, is a preset quantity parameter. ,right Without further limitation, in this embodiment For the purpose of illustration, implementers may decide on a case-by-case basis.

[0050] The reference peak value of the water head value of the water diversion pipeline and the reference peak value of the water head value of the tailwater pipeline of each unit in each transition process during the simulation time are obtained by the above-mentioned reference peak acquisition method.

[0051] The water head fluctuation index of the diversion pipeline and the tailwater pipeline are quantified by the fluctuation of the reference wave peak of the water pipe head value of each unit in each transition process during the simulation time; the quantification process is as follows: The formula corresponding to the water head fluctuation index of the water diversion pipeline is:

[0052] In the formula, Indicates that each unit has the first The water head value of the water diversion pipeline of the reference wave peak, Indicates the water head value of the water supply pipe that is closest to 0 in each unit during each transition process. It represents the water head fluctuation index of the water diversion pipeline of each unit in each transition process. In this embodiment, the first 5 are taken for analysis, that is, n is 5, but n is not specifically limited, and the implementer can determine it according to the specific scene and situation.

[0053] The formula corresponding to the tailwater pipe head fluctuation index is:

[0054] In the formula, Indicates that each unit has the first The tailwater pipe head value of the reference wave peak, Indicates the tailwater pipe head value that is closest to 0 for each unit in each transition process. It represents the tailwater pipe head fluctuation index of each unit in each transition process. Among them, the value of n in the formula of tailwater pipe head fluctuation index is the same as that in the formula of diversion pipe head fluctuation index.

[0055] The frequency fluctuation index, power fluctuation index, voltage amplitude index, water head fluctuation index of the water diversion pipeline and the water head fluctuation index of the tailwater pipeline of each unit in each transition process during the simulation process are obtained; the safety degree corresponding to each unit in each transition process during the simulation process is artificially calibrated with a safety degree value, wherein the calibrated safety degree value ranges from 0 to 1.

[0056] Normalize all evaluation indicators of each unit in each transition process to obtain normalized results. Obtain the normalized evaluation indicators of all transition processes in the historical simulation process, and use the normalized evaluation indicators of all transition processes in each simulation process as a sample set, wherein the sample set is an M×H matrix, wherein M is the total number of evaluation indicators, and H is the total number of all transition processes in each simulation process; wherein different sample sets represent different simulation processes, and the transition processes in different simulation processes are the same.

[0057] The BiLSTM neural network is trained through all sample sets in history and calibrated safety values ​​to obtain a trained BiLSTM neural network.

[0058] Obtain the sample set that needs to be analyzed at the current moment, input the sample set that needs to be analyzed at the current moment into the trained BiLSTM neural network, and obtain the initial safety degree of each unit. According to the initial safety degree of all units, the fuzzy entropy weight method is used to estimate the first overall safety degree of the hydro-generator unit.

[0059] Among them, the BiLSTM neural network and the fuzzy entropy weight method are both well-known technologies and will not be described in detail here.

[0060] At this point, the first overall safety degree of the hydro-generator set is obtained.

[0061] Step S003: Divide each evaluation indicator into several safety levels, obtain the fuzzy evaluation range of each safety level, and construct a fuzzy relationship matrix according to the value of each evaluation indicator of each unit in each transition process and the fuzzy evaluation range corresponding to each safety level, wherein the elements in the fuzzy relationship matrix are recorded as membership degrees; adjust the fuzzy relationship matrix through the information entropy of the membership degrees of all safety levels of each evaluation indicator in the fuzzy relationship matrix to obtain a fuzzy entropy evaluation matrix; process the fuzzy entropy evaluation matrix, and obtain the second overall safety degree of the hydro-turbine generator set through the data distribution in the processed fuzzy entropy evaluation matrix.

[0062] It should be noted that, due to the error in obtaining the overall safety of the hydro-generator set through the neural network, the safety of the hydro-generator set analyzed with only one safety degree is inaccurate, so another safety degree evaluation method is needed to analyze the safety degree.

[0063] It should be further explained that since the safety analysis of hydro-turbine generator sets faces a complex, ambiguous and uncertain environment, many influencing factors may not be completely clear or difficult to quantify. Also, the role of fuzzy theory is mainly to help process and analyze information that is uncertain and ambiguous; therefore, processing and modeling fuzzy information through fuzzy theory can help improve the accuracy and reliability of safety assessment, enhance the scientific nature of decision-making and the robustness of the system.

[0064] Specifically, the fuzzy evaluation range of the six safety levels corresponding to each evaluation indicator of each unit in each transition process is obtained, and the membership degree corresponding to each safety level is obtained through the membership function according to the value of each evaluation indicator; wherein, the six safety levels include stable, relatively stable, critical stability, weak risk, medium risk, and strong risk; wherein, the safety level decreases in the order of stable, relatively stable, critical stability, weak risk, medium risk, and strong risk.

[0065] Among them, the fuzzy evaluation range can be obtained through expert experience method, historical data method and literature method; among them, the fuzzy evaluation range is between 0 and 1.

[0066] With all evaluation indicators as rows and all security levels as columns, a fuzzy relationship matrix is ​​obtained; wherein the elements in the fuzzy relationship matrix are the membership of each evaluation indicator to the corresponding security level. The sum of the membership of all security levels of each evaluation indicator is 1. The fuzzy relationship matrix has M rows and N columns; wherein M is the total number of evaluation indicators and N is the total number of security levels.

[0067] Calculate the entropy value of the membership degree corresponding to all security levels of each evaluation indicator, construct the entropy values ​​of all evaluation indicators into an entropy matrix of 1 row and M columns, multiply the entropy matrix by the fuzzy relationship matrix to obtain a matrix, recorded as the fuzzy entropy evaluation matrix; perform linear normalization on the values ​​in the fuzzy entropy evaluation matrix. The fuzzy entropy evaluation matrix is ​​1 row and N columns. The entropy value is calculated by the information entropy function, and the information entropy function is a well-known technology and will not be described here.

[0068] According to the maximum membership principle, the maximum value of the N values ​​in the fuzzy entropy evaluation matrix is ​​obtained and recorded as the reference safety degree of each unit in each transition process. The mean value of the reference safety degree of each unit in all transition processes is recorded as the reference safety degree of each unit. According to the reference safety degree of all units, the second overall safety degree of the hydro-turbine generator set is estimated by the fuzzy entropy weight method.

[0069] Among them, the maximum membership principle and the fuzzy entropy weight method are both well-known technologies and will not be described in detail here.

[0070] Step S004: Obtain the final safety degree of the multi-machine system of the hydro-generator set through the first overall safety degree and the second overall safety degree of the hydro-generator set.

[0071] It should be noted that, due to the error in obtaining the overall safety degree of the hydro-turbine generator set through the neural network, the safety of the hydro-turbine generator set analyzed with only one safety degree is inaccurate, so the safety degree is adjusted by the safety degree obtained through fuzzy theory.

[0072] Specifically, the average of the first overall safety degree and the second overall safety degree of the hydro-turbine generator set is used as the reference overall safety degree of the hydro-turbine generator set, and the reference overall safety degree of the hydro-turbine generator set is normalized by the standard normalization method to obtain the final safety degree of the multi-machine system of the hydro-turbine generator set. Among them, the standard normalization method is to first perform standardization processing, and then perform Min-Max normalization processing; among them, the standardization processing and the Min-Max normalization processing are both well-known technologies, and will not be described in detail here. Among them, the safety degree evaluation flow chart of the multi-machine system of the hydro-turbine generator set is as follows: Figure 2 shown.

[0073] At this point, the final safety degree of the turbine-generator multi-machine system is obtained.

[0074] like Figure 3 As shown, the second aspect of the present invention is to provide a multi-machine system safety assessment system for a hydro-generator set, the system comprising the following modules: Data acquisition module 101: used to collect various types of data of each unit in all transition processes during the simulation of the hydro-generator unit; Data analysis module 102: used to analyze and obtain a first overall safety degree of the hydro-generator set according to a neural network model, and to analyze and obtain a second overall safety degree of the hydro-generator set through fuzzy theory; Safety evaluation module 103: used to obtain the final safety degree of the multi-machine system of the hydro-generator set through the first overall safety degree and the second overall safety degree of the hydro-generator set.

[0075] The third aspect of the present invention is to provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, a method for evaluating the safety of a multi-machine system of a hydro-turbine generator set is implemented.

[0076] A fourth aspect of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements a method for evaluating the safety of a multi-machine system of a hydro-turbine generator set.

[0077] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for evaluating the safety of a multi-machine system of a hydro-generator set, characterized in that: The method comprises the following steps: Obtain various types of data for each unit during all transitions during the simulation of the hydro-generator unit; According to the changes of various types of data in each transition process of each unit, the evaluation index in each transition process of each unit is obtained, a sample set is constructed through the evaluation indexes in all transition processes of each unit, and the neural network model is trained through the sample set and the artificially calibrated safety value to obtain the trained neural network model; according to the sample set to be analyzed at the current moment and the trained neural network model, the first overall safety degree of the hydro-generator unit is obtained; Each evaluation index is divided into several safety levels, and a fuzzy evaluation range of each safety level is obtained. A fuzzy relationship matrix is ​​constructed according to the value of each evaluation index of each unit in each transition process and the fuzzy evaluation range corresponding to each safety level, and the elements in the fuzzy relationship matrix are recorded as membership degrees; the fuzzy relationship matrix is ​​adjusted according to the information entropy of the membership degrees of all safety levels of each evaluation index in the fuzzy relationship matrix; the second overall safety degree of the hydro-generator unit is obtained through the data distribution in the adjusted matrix; The final safety degree of the multi-machine system of the hydro-generator set is obtained through the first overall safety degree and the second overall safety degree of the hydro-generator set.

2. A method for evaluating the safety of a multi-machine system of a hydro-generator set according to claim 1, characterized in that: The various types of data include: water head value of the water diversion pipeline, water head value of the tailwater pipeline, motor output power, terminal voltage, excitation voltage and frequency.

3. A method for evaluating the safety of a multi-machine system of a hydro-generator set according to claim 2, characterized in that: The specific steps of obtaining the evaluation index of each transition process of each unit according to the change of various types of data in each transition process of each unit are as follows: The evaluation indexes include: frequency fluctuation index, power fluctuation index, voltage amplitude index, water head fluctuation index of water diversion pipeline and water head fluctuation index of tailwater pipeline; The construction process of the frequency fluctuation index is as follows: Will The frequency within the interval is recorded as the limited frequency. The frequencies outside the interval are recorded as out-of-limit frequencies, where: To preset the reference frequency, is the preset frequency error; Among them, the data in each transition process of each unit was obtained with a sampling interval of 1 second; Among them, the formula corresponding to the frequency fluctuation index is: In the formula, It represents the mean value of the out-of-limit frequency at all time points in each transition process of each unit, It represents the mean value of the frequency within the limit at all time points in each transition process of each unit, represents the maximum frequency corresponding to all time points in each transition process of each unit, where , , They are , , The weight of Represents the frequency fluctuation index in each transition process of each unit; The formula corresponding to the power fluctuation index is: In the formula, Indicates the maximum motor output power of each unit in each transition process, represents the initial motor output power of each unit during each transition process, represents the final motor output power of each unit during each transition process, Represents the power fluctuation index of each unit in each transition process; The formula corresponding to the voltage amplitude index is: In the formula, Indicates the maximum value of the terminal voltage of each unit during each transition process. Indicates the initial terminal voltage of each unit in each transition process, Indicates the maximum value of the excitation voltage in each transition process of each unit. Indicates the initial excitation voltage of each unit in each transition process, Indicates the voltage amplitude index in each transition process of each unit; The reference peaks of the water head value of the water diversion pipe and the water head value of the tailwater pipe of each unit in each transition process during the simulation time are obtained, and the water head fluctuation index of the water diversion pipe and the water head fluctuation index of the tailwater pipe are quantified by the fluctuation of the reference peaks of the water head value of each unit in each transition process during the simulation time.

4. A method for evaluating the safety of a multi-machine system of a hydro-generator set according to claim 3, characterized in that: The reference peaks of the water head value of the water diversion pipeline and the water head value of the tailwater pipeline of each unit in each transition process during the simulation time are obtained, and the water head fluctuation index of the water diversion pipeline and the water head fluctuation index of the tailwater pipeline are quantified according to the fluctuation of the reference peaks of the water head value of each unit in each transition process during the simulation time. The specific steps include the following: The specific method for obtaining the reference peak is: Select the starting point of each unit in each transition process during the simulation time. The water pipe head value at the peak of the wave is The mean of the water pipe head values ​​at the peaks is recorded as the front water head mean; Then obtain the final value of each unit at the end of each transition process during the simulation time. The water pipe head value at the peak of the wave is The mean of the water pipe head values ​​at the peaks is recorded as the mean water head after the simulation. If the mean water head before and after the simulation is greater than the mean water head after the simulation, the mean water head before and after the simulation is recorded as the mean water head after the simulation. The peaks are taken as reference peaks; if the current water head mean is less than the subsequent water head mean, the subsequent water head of each unit at the end of each transition process during the simulation time is taken as the reference peak. peaks as reference peaks; among them, is the preset quantity parameter; The reference peak value of the water head value of the water diversion pipeline and the reference peak value of the tailwater pipeline of each unit in each transition process during the simulation time are obtained by the above-mentioned reference peak acquisition method; The formula corresponding to the water head fluctuation index of the water diversion pipeline is: In the formula, Indicates that each unit has the first The water head value of the water diversion pipeline of the reference wave peak, Indicates the water head value of the water supply pipe that is closest to 0 in each unit during each transition process. It indicates the water head fluctuation index of the water diversion pipeline of each unit in each transition process; The formula corresponding to the tailwater pipe head fluctuation index is: In the formula, Indicates that each unit has the first The tailwater pipe head value of the reference wave peak, Indicates the tailwater pipe head value that is closest to 0 for each unit in each transition process. It represents the tailwater pipe head fluctuation index of each unit during each transition process.

5. A method for evaluating the safety of a multi-machine system of a hydro-generator set according to claim 1, characterized in that: The specific steps of constructing a sample set by using the evaluation indicators of all transition processes of each unit are as follows: Normalize all evaluation indicators of each unit in each transition process to obtain normalized results; obtain the normalized evaluation indicators of all transition processes in the historical simulation process, and use the normalized evaluation indicators of all transition processes in each simulation process as a sample set, wherein the sample set is an M×H matrix, wherein M is the total number of evaluation indicators, and H is the total number of all transition processes in each simulation process.

6. A method for evaluating the safety of a multi-machine system of a hydro-generator set according to claim 1, characterized in that: The first overall safety degree of the hydro-generator set is obtained according to the sample set to be analyzed at the current moment and the trained neural network model, and the specific steps include the following: The sample set that needs to be analyzed at the current moment is obtained, and the sample set that needs to be analyzed at the current moment is input into the trained BiLSTM neural network to obtain the initial safety degree of each unit; according to the initial safety degree of all units, the fuzzy entropy weight method is used to obtain the first overall safety degree of the hydro-generator unit.

7. A method for evaluating the safety of a multi-machine system of a hydro-generator set according to claim 1, characterized in that: The method of dividing each evaluation index into several safety levels, obtaining the fuzzy evaluation range of each safety level, and constructing a fuzzy relationship matrix according to the value of each evaluation index of each unit in each transition process and the fuzzy evaluation range corresponding to each safety level, includes the following specific steps: The fuzzy evaluation range of the six safety levels corresponding to each evaluation index of each unit in each transition process is obtained by expert experience method or historical data method, and the membership degree corresponding to each safety level is obtained by the membership function according to the value of each evaluation index; among which, the six safety levels include stable, relatively stable, critically stable, weak risk, medium risk, and strong risk; With all evaluation indicators as rows and all safety levels as columns, a fuzzy relationship matrix is ​​obtained.

8. A method for evaluating the safety of a multi-machine system of a hydro-generator set according to claim 1, characterized in that: The fuzzy relationship matrix is ​​adjusted by the information entropy of the membership of all safety levels of each evaluation index in the fuzzy relationship matrix; the second overall safety degree of the hydro-generator set is obtained by the data distribution in the adjusted matrix, and the specific steps include the following: Calculate the entropy value of the membership degree corresponding to all safety levels of each evaluation indicator, construct the entropy values ​​of all evaluation indicators into an entropy matrix with 1 row and M columns, and multiply the entropy matrix with the fuzzy relationship matrix to obtain a matrix, which is recorded as the fuzzy entropy evaluation matrix, where the fuzzy entropy evaluation matrix has 1 row and N columns; Perform linear normalization on the values ​​in the fuzzy entropy evaluation matrix; According to the maximum membership principle, the maximum value of the N values ​​in the fuzzy entropy evaluation matrix is ​​obtained and recorded as the reference safety degree of each unit in each transition process. The average value of the reference safety degree of each unit in all transition processes is recorded as the reference safety degree of each unit. According to the reference safety degree of all units, the second overall safety degree of the hydro-turbine generator set is obtained by the fuzzy entropy weight method.

9. A method for evaluating the safety of a multi-machine system of a hydro-generator set according to claim 1, characterized in that: The method of obtaining the final safety degree of the multi-machine system of the hydro-generator set by using the first overall safety degree and the second overall safety degree of the hydro-generator set comprises the following specific steps: The average of the first overall safety degree and the second overall safety degree of the hydro-turbine generator set is taken as the reference overall safety degree of the hydro-turbine generator set. The reference overall safety degree of the hydro-turbine generator set is normalized by the standard normalization method to obtain the final safety degree of the multi-machine system of the hydro-turbine generator set.

10. A multi-machine system safety assessment system for a hydro-turbine generator set, characterized in that: The system includes the following modules: Data acquisition module: used to collect various types of data of each unit in all transition processes during the simulation of the hydro-generator unit; Data analysis module: used to analyze and obtain the first overall safety degree of the hydro-generator set according to the neural network model, and to analyze and obtain the second overall safety degree of the hydro-generator set through fuzzy theory; Safety evaluation module: used to obtain the final safety of the multi-machine system of the hydro-generator set through the first overall safety and the second overall safety of the hydro-generator set.

Images