Method and system for monitoring operation risk of high dam of hydropower station

Abstract

The invention belongs to the technical field of dam monitoring, and provides an operation risk monitoring method and system for a high dam of a hydropower station, and the method comprises the steps: constructing a finite element simulation model of the dam; based on the finite element simulation model, obtaining a mapping relation between the low-order modal data of the dam under the simulated flood discharge working condition and dam mechanical parameters; continuously collecting vibration data at the dam gate A by using a vibration sensor, and judging whether the dam gate A has a flood discharge event in combination with upstream and downstream water level data of the dam gate A; performing modal analysis during operation on a signal acquired by the vibrating wire sensor corresponding to the dam gate A in which the flood discharge event occurs to obtain a low-order modal parameter of the dam; on the basis of the low-order modal parameters of the dam and the mapping relation between the low-order modal parameters and dam mechanical parameters, dam mechanical parameter inversion is carried out through an optimization algorithm; and judging the operation risk of the high dam of the hydropower station according to a dam mechanical parameter inversion result. According to the invention, risk monitoring can be carried out on the dam during flood discharge.

Description

Technical Field

[0001] The present invention belongs to the technical field of tight oil reservoir exploitation technology, and in particular relates to an operation risk monitoring method and system for a high dam of a hydropower station. Background Art

[0002] As the design velocity and flow rate of high dam discharge structures under flood discharge conditions continue to increase, the problem of water flow-induced dam body vibration has become increasingly prominent. At present, in the practice of high dam safety monitoring projects, the traditional dam safety monitoring system is a low-speed static monitoring system that monitors according to a set period. Due to the limitations of the performance of measuring instruments and data acquisition devices, the sampling rate can only reach a maximum of a few seconds to more than ten seconds per time. The time for the data acquisition device to complete a full-channel acquisition is generally about 1 minute. It is impossible to capture the changes in various physical quantities during the vibration process and cannot meet the needs of structural state dynamics analysis and safety assessment under flood discharge conditions.

[0003] Prior art CN 116839716 A discloses a dam safety dynamic monitoring system that uses multi-vibration source coupling, dynamic and static combination, etc., which can quickly respond, real-time dynamic acquisition and analysis under working conditions such as flood discharge, sand flushing, generator set operation and earthquake, and meet the needs of dam safety dynamic analysis and evaluation. Specifically, it includes: obtaining multiple dam vibration sources through various channels; using a distributed monitoring method for monitoring local vibration events of the dam; using a multi-vibration source coupling analysis method to make real-time judgments, and timely issuing a start or stop dynamic monitoring instruction to the entire system or a part of the system to trigger the dynamic monitoring working state of the corresponding equipment; dynamic measurement adopts the main and auxiliary principle for vibration events, with system signals as the main and local signals as the auxiliary to start or stop the dynamic measurement working state; using a data analysis sample difference method to achieve seamless compatibility between dynamic measurement and static measurement states.

[0004] The following problems exist: (1) A multi-vibration source coupling analysis method is proposed mainly for multiple vibration sources. Some of the event information required by this method depends on on-site conditions such as the power station control center, such as "indication information related to flood discharge, sand flushing or generator start-up and dam vibration caused by the hydropower station control center". However, different power stations were built at different times and have different environments, and it is very likely that there are no on-site conditions for automatically sending event information. (2) The multi-vibration source acquisition method includes the dam body vibration event information sent by the dam body vibration monitoring module of each dynamic measurement unit of the system, and the recursive STA / LTA vibration event analysis method and voting judgment method are used to analyze and judge the vibration state of the dam body. However, the recursive STA / LTA vibration event analysis method is mostly used to automatically detect effective microseismic signals of earthquakes. Flood discharge vibration is a key monitoring project for dams, especially high arch dams, in long-term operation, and a more targeted event judgment method should be adopted.

[0005] Prior art CN 106324660 B discloses a dynamic monitoring device and method for dam safety under earthquake conditions. While realizing conventional static monitoring, dynamic monitoring of dam safety under earthquake conditions is also realized. Specifically, it includes: a dynamic monitoring sensor, which is used to sense parameter information reflecting the dynamic safety state of the dam, and the parameter information includes deformation, strain and seepage; a data collector, which is used to sample the dynamic monitoring sensor signal online, and transmit and exchange data to the upper computer monitoring system through the communication network; an earthquake detection device, which is used to capture the earthquake state, and if an earthquake is detected, broadcast the earthquake event to the communication network; a host computer monitoring system, which is used to process the data sent by the data collector, and manage and control the monitoring device; a database, which is used to store detection data and parameter information.

[0006] However, its dynamic monitoring range is limited to earthquake conditions, and the dynamic monitoring scenario of flood discharge conditions is not considered.

[0007] CN 117191147 A A flood discharge dam water level monitoring and early warning method and system" relates to the technical field of water level monitoring, in particular to a flood discharge dam water level monitoring and early warning method and system, which improves the prediction accuracy and reliability and can better ensure the stability and safety of the dam.

[0008] Patent CN 117191147 A discloses pressure early warning analysis for dam flood discharge, but does not specifically disclose the comprehensive use of multi-sensor data.

[0009] Patent CN 116993163 A "A system and method for safety supervision of water conservancy and hydropower project construction" discloses a method and system for safety supervision of water conservancy and hydropower project construction, which discloses the setting of multiple sensors and a specific early warning model. However, there are the following problems: (1) The multiple sensors used in the technical solution are only used for static monitoring and cannot be used for dynamic monitoring; (2) Since it is only static monitoring, the monitoring content does not involve structural dynamic indicators, and it is even more impossible to evaluate the dynamic working state; (3) The early warning model does not disclose the specific technology, and the early warning model cannot realize the function of automatically triggering monitoring during flood discharge; (4) The specific early warning model review method, dynamic monitoring technology and parameter inversion data usage concept are not disclosed.

[0010] In summary, existing technologies are unable to perform automated dynamic monitoring of dams through multiple sensors under flood discharge conditions. Summary of the invention

[0011] To solve the above problems, the present disclosure provides a method and system for monitoring the operation risks of a high dam of a hydropower station, which adopts multiple sensors and can realize automatic dynamic monitoring of the dam under flood discharge conditions.

[0012] The following is the technology of the present invention:

[0013] A method for monitoring the operation risk of a high dam of a hydropower station, characterized by comprising:

[0014] Construct a finite element simulation model of the dam; obtain the mapping relationship between the low-order modal data of the dam under simulated flood discharge conditions and the mechanical parameters of the dam based on the finite element simulation model;

[0015] The vibration sensor is used to continuously collect vibration data at the dam gate A, and combined with the upstream and downstream water level data of the dam gate A, it is determined that a flood discharge event has occurred at the dam gate A;

[0016] Perform runtime modal analysis on the signal obtained by the vibrating wire sensor corresponding to the dam gate A where the flood discharge event occurred, and obtain the low-order modal parameters of the dam;

[0017] Based on the low-order modal parameters of the dam and the mapping relationship between the low-order modal parameters and the mechanical parameters of the dam, the mechanical parameters of the dam are inverted through the optimization algorithm;

[0018] The operation risk of the high dam of the hydropower station is judged based on the inversion results of the dam mechanical parameters.

[0019] Further,

[0020] The method of constructing a finite element simulation model of the dam and obtaining a mapping relationship between low-order modal data of the dam under a simulated flood discharge condition and mechanical parameters of the dam based on the finite element simulation model includes:

[0021] Establish multiple finite element simulation models based on engineering survey data and design data, and the multiple finite element simulation models together constitute the dam finite element simulation model;

[0022] White noise is loaded on the boundaries around the dam foundation of the finite element simulation model of the dam, and random vibration is used to simulate the flow-induced vibration load on the dam during the discharge process;

[0023] Finite element transient dynamic analysis is performed on the finite element simulation model of the dam to obtain the dynamic response of the dam under white noise vibration, and then the acceleration time history signal of the monitoring point is obtained. Based on the modal analysis method, the n-order low-order modal parameters of the dam are obtained.

[0024] The dam mechanical parameters and low-cost modal parameters of the dam finite element simulation model are used as sample data and stored in the database;

[0025] Based on the sample data in the database, the mapping relationship between the low-order modal parameters and the dam mechanical parameters is constructed using the proxy model.

[0026] Further,

[0027] The method of continuously collecting vibration data at the dam gate A by using a vibration sensor and combining the upstream and downstream water level data of the dam gate A to determine whether a flood discharge event has occurred at the dam gate A includes:

[0028] A vibration sensor is installed at the dam gate A to continuously collect the vibration amplitude, main vibration frequency, and other dominant frequency data at various locations of the dam gate A;

[0029] Use the GNSS positioning engine to obtain the upstream and downstream water level data of the dam gate A;

[0030] Combine historical water levels with flood discharge event information to determine whether a flood discharge event is likely to occur under current upstream and downstream water level data;

[0031] Compare the vibration amplitude, main vibration frequency, and other dominant frequency data at each location of the dam gate A with historical data; if a certain value exceeds the threshold, it is preliminarily determined that a flood discharge event has occurred at the orifice of the dam gate A;

[0032] A vibrating string sensor is used to dynamically measure the dam gate A, which is preliminarily judged to have a flood discharge event, at a certain frequency to obtain frequency data B. At regular intervals, a characteristic value cv representing the smoothness of the amplitude change of the frequency data B is calculated. If the cv corresponding to the dam gate A exceeds the threshold, it is determined that a flood discharge event has occurred at the dam gate A.

[0033] Further,

[0034] The same frequency and same phase excitation technology is used as the resonant excitation of the vibrating string sensor, and the output of the excitation signal is realized by DAC;

[0035] The same-frequency and same-phase excitation technology uses a single pulse wave with the same resonance frequency and phase as the steel string of the vibrating-string sensor to maintain the resonance state of the steel string; the frequency and phase of the single pulse wave are obtained by acquiring the resonance frequency and real-time motion phase of the steel string.

[0036] Further,

[0037] The method of obtaining the real-time motion phase of the steel string comprises:

[0038] The steel string motion signal is obtained by ADC sampling and a set of discrete data is obtained;

[0039] The initial sampling time T0 is known, and the phase of the first data is obtained by using FFT to calculate discrete data, and the running phase of the steel string at time T0 is obtained; assuming that the time of excitation is T1, the steel string movement phase at time T1 is obtained based on the steel string movement frequency calculated by FFT and the time difference between T1 and T2, as well as the phase at time T0.

[0040] Further,

[0041] The Cov-SSI algorithm is used to perform runtime modal analysis on the signals obtained by the vibrating wire sensor corresponding to gate A to obtain the low-order modal parameters of the dam; these include:

[0042] Construct a discrete state space equation of the vibration system; the discrete state space equation is:

[0043]

[0044] In the formula, M, C, and K are the mass matrix, damping matrix, and stiffness matrix of the system respectively; z(t) is the acceleration vector, velocity vector and displacement vector of the discrete system; F(t) is the excitation vector;

[0045] Introducing the state vector The vibration system is transformed into the corresponding state space; the state space is:

[0046] x k+1 =Ax k +w k

[0047] y k =Cx k +v k

[0048] Where A and C are the discrete time system matrix and discrete time output matrix respectively;

[0049] After finding A and C, perform eigenvalue decomposition on A:

[0050] A=ΦΛΦ -1

[0051] Where Λ is the eigenvalue matrix, Φ is the eigenvector matrix;

[0052] Arrange the diagonal elements of Λ in order of magnitude, and obtain the preliminary modal parameters of the dam:

[0053]

[0054] φ j =Cγ j

[0055] Among them, μ j is the jth eigenvalue of A, γ j is the corresponding eigenvector, Δt is the sampling time interval, f j , j and φ j are the modal frequency, damping ratio and vibration shape vector respectively;

[0056] The preliminary modal parameters of the dam are processed using the stability diagram method to obtain the low-order modal parameters of the dam.

[0057] Further,

[0058] In the stability graph method, the stability graph is composed of stability points, and the stability points satisfy the following stability criteria:

[0059]

[0060] 1-MAC(φ i ,φ i+1 )<0.02

[0061] In the formula, superscripts i and i+1 represent the system order, MAC(·,·) is the modal assurance criterion, which is used to measure the similarity between two modal vibration shapes. The expression of MAC(·,·) is:

[0062]

[0063] in,(·) H represents the conjugate transpose.

[0064] Further,

[0065] The method of inverting the dam mechanical parameters by using an optimization algorithm based on the low-order modal parameters of the dam and the mapping relationship between the low-order modal parameters and the mechanical parameters of the dam comprises:

[0066] The objective function of the optimization algorithm is constructed. The calculation goal of the objective function is to make the response characteristics of the dam structure under the dam mechanical parameters closest to the observed characteristics; the objective function is:

[0067]

[0068] Where, θ is the structural mechanics parameter to be inverted; and They represent the i-th order frequency and vibration mode of the modal parameters identified by the vibration signal; f i (θ) and φ i (θ) are the i-th order frequency and vibration mode of the structural modal parameters obtained by finite element calculation; α i and β i are the weight coefficients of different items;

[0069] The mapping relationship between low-order modal parameters and dam mechanical parameters is used as a training set and a test set, and the proxy model is trained in combination with the objective function of the optimization algorithm, so that the proxy model outputs corresponding dam mechanical parameters based on the low-order modal parameters.

[0070] Further,

[0071] It also includes: static monitoring of the operational risks of the high dam of the hydropower station; including:

[0072] When the high dam of the hydropower station is not in a state of flood discharge, dynamic data measurement of the high dam of the hydropower station is performed for a duration of x seconds, and the measured data are arithmetic averaged as static measurement data;

[0073] When the high dam of the hydropower station is in a state of flood discharge, the high dam data of the hydropower station is continuously measured, and the dynamic measurement data of x seconds are extracted and arithmetic averaged as static measurement data;

[0074] Based on static measurement data, jointly judge whether there is an operational risk of the high dam of the hydropower station.

[0075] An operation risk monitoring system for a high dam of a hydropower station is characterized by comprising:

[0076] Model building module, used to build a finite element simulation model of the dam; based on the finite element simulation model, the mapping relationship between the low-order modal data of the dam under the simulated flood discharge condition and the mechanical parameters of the dam is obtained;

[0077] The flood discharge event judgment module is used to continuously collect vibration data at the dam gate A using a vibration sensor, and combine the upstream and downstream water level data of the dam gate A to judge whether a flood discharge event has occurred at the dam gate A;

[0078] The modal parameter inversion module is used to perform runtime modal analysis on the signal obtained by the vibrating wire sensor corresponding to the dam gate A where the flood discharge event occurred, and obtain the low-order modal parameters of the dam;

[0079] Mechanical parameter inversion module, which is used to invert the mechanical parameters of the dam through optimization algorithm based on the low-order modal parameters of the dam and the mapping relationship between the low-order modal parameters and the mechanical parameters of the dam;

[0080] The judgment module is used to judge the operation risk of the high dam of the hydropower station based on the inversion results of the dam mechanical parameters.

[0081] Compared with the prior art, the present invention has the following advantages:

[0082] The finite element simulation model of the dam constructed by the present invention can simulate the flood discharge condition to obtain low-order modal data and establish a mapping relationship with mechanical parameters, providing data reference for subsequent optimization;

[0083] The vibration sensor continuously collects the vibration data of the dam gate based on the principle of physical conversion. It combines the upstream and downstream water level data and uses the hydraulic principle and historical data fusion analysis to determine whether a flood discharge event has occurred at a certain gate opening. This multi-source data fusion method improves the accuracy of judgment and is automatic and continuous.

[0084] For the dam gates where flood discharge occurs, the vibrating string sensor signals are subjected to runtime modal analysis, and the modal parameters of the dam are calculated using structural dynamics and signal processing theory. The mechanical parameters are then inverted using an optimization algorithm based on the mapping relationship with the mechanical parameters. The dam operation risk can be determined by comparing the mechanical parameters with the threshold, thereby achieving full-process automated dynamic monitoring to ensure dam safety.

[0085] Other features and advantages of the present disclosure will be described in the following description, and partly become apparent from the description, or be understood by implementing the present disclosure. The purpose and other advantages of the present disclosure can be realized and obtained by the structures pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] In order to more clearly illustrate the embodiments of the present disclosure 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 some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0087] Figure 1 A schematic diagram of the method of the present invention is shown;

[0088] Figure 2 A structural diagram of a dynamic monitoring system according to an embodiment of the present invention is shown;

[0089] Figure 3 A schematic diagram of a dam portion according to an embodiment of the present invention is shown;

[0090] Figure 4 shows an original stabilization diagram of an embodiment of the present invention;

[0091] Figure 5 The stable diagram after eliminating the false polar modes according to the embodiment of the present invention is shown. DETAILED DESCRIPTION

[0092] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0093] Figure 1 A schematic diagram of the method according to the present invention is shown, and the specific implementation details of the present invention include:

[0094] 1. Step 1: Construct an online simulation model for dam dynamic monitoring

[0095] Several groups of finite element simulation models are established based on engineering survey data and design data.

[0096] The dam is divided into three different parts according to the different concrete grades during the construction period, and the material mechanical parameters are shown in Table 1. Among them, the elastic modulus [E1, E2, E3] of the three partitions of the dam body is assumed to be the parameters to be inverted, and its value range is roughly determined based on engineering experience. The specific parameters of the several groups of finite element simulation models can be determined by experimental design methods, such as Latin hypercube sampling.

[0097] Table 1 Dam mechanical parameters

[0098]

[0099] White noise random vibration is loaded on the boundaries around the dam foundation to simulate the water flow-induced vibration load on the dam during the discharge process. Finite element transient dynamic analysis is performed on the above several groups of finite element simulation models to obtain the dynamic response of the arch dam under white noise vibration, and then the acceleration time history signal of the monitoring point is obtained. Based on the modal analysis method, the n-order low-order modes in front of the dam (n is 3 to 5) are obtained. The dam mechanical parameters and low-order modal results of the above several groups of finite element simulation models are used as sample data and stored in the dynamic monitoring online database. Based on the sample data in the database, the proxy model (Kriging model, RBF model, RSA model, etc.) is used to construct the mapping relationship between the dam mechanical parameters and low-order modal results.

[0100] 2. Step 2: Preliminary determination of whether a flood discharge event has occurred at the orifice

[0101] The vibration sensors and GNSS positioning engine of the dam gate and its vicinity are used to continuously collect the operation data of the dam gate and its vicinity. The collected operation data of the dam gate and its vicinity include the vibration amplitude at each part of the gate, the main vibration frequency and other dominant frequencies, and the upstream and downstream water levels.

[0102] Combined with historical water levels and flood discharge event information, determine whether the current upstream and downstream water levels (obtained by the GNSS positioning engine) may result in a flood discharge event. If a flood discharge event is likely to occur, then compare historical data with the vibration sensors of the dam gates and their vicinity to determine the opening status of each gate.

[0103] If the vibration intensity (such as acceleration amplitude) is used as the monitoring characteristic quantity, when the real-time monitoring value exceeds the reference threshold, it is determined that a flood discharge event has occurred at the orifice. For example, if the GNSS positioning engine obtains the current water level of 610.00m, it is judged by comparing historical data that a deep hole or medium hole flood discharge event is likely to occur at this time. If the vibration intensity monitored by the 1#, 2#, 8#, and 9# medium hole gates exceeds the preset threshold, it is preliminarily determined that the 1#, 2#, 8#, and 9# medium holes are open to discharge flood water at this time.

[0104] 3. Step 3: Trigger the vibrating string sensor for dynamic monitoring and verify the flood discharge event

[0105] A1. Dynamic monitoring technology

[0106] When a flood discharge event is detected at a certain orifice, the dynamic data acquisition device triggers the vibrating string sensor in the flow channel area of ​​the orifice to perform dynamic monitoring at a certain frequency (50 Hz or higher).

[0107] When performing dynamic measurements, in order to obtain a higher sampling rate, the dynamic data acquisition device cancels the frequency sweep excitation commonly used in static measurements of vibrating string sensors and instead uses the same frequency and same phase excitation technology to compress the measurement cycle. The output of the excitation signal is implemented using a DAC, which can output a sine wave signal at any starting position and any frequency, meeting the requirements of the same frequency and same phase excitation technology.

[0108] The same frequency and same phase excitation technology, that is, using a single pulse wave close to the resonance frequency (measured by the spectrum method) to maintain the resonance state of the steel string, the key lies in the phase control of the excitation wave and the timing of the excitation. If the frequency or phase of the excitation is inconsistent with the movement of the steel string, the resonance will be weakened.

[0109] The concept of co-frequency means that the excitation frequency is near the resonant frequency and close enough. Since the resonant frequency of the sensor changes with the change of the measured physical quantity, the frequency used for co-frequency excitation is the most recent resonant frequency measurement value during the dynamic measurement process. Between adjacent excitation windows, the movement of the steel string can be sampled and the resonant frequency can be determined. The excitation is adjusted according to the newly calculated resonant frequency to track the changes in the resonant frequency of the steel string. The excitation system needs to have good frequency resolution to accurately track subtle changes in the resonant frequency. Only a few oscillation waveforms are needed within the sampling window to determine the resonant frequency of the steel string. Compared with static measurement technology, the spectrum analysis algorithm can use fewer oscillation waveforms to accurately measure the frequency of the steel string.

[0110] The concept of in-phase means that the phase of the excitation signal is consistent with the phase of the steel string movement, that is, the excitation is applied in the direction of the steel string movement. The core of in-phase excitation is to obtain the real-time motion phase of the steel string. There are many ways to achieve the acquisition of the real-time motion phase of the steel string. The implementation methods are as follows:

[0111] (1) FFT fast Fourier transform calculation phase

[0112] FFT can calculate both frequency and phase. First, ADC samples the steel string motion signal and obtains a set of discrete data. The initial sampling time T0 is known, and then FFT calculates the discrete data to obtain the phase of the first data, and the running phase of the steel string at time T0 can be known. Assuming that the time of excitation is T1, the steel string motion phase at time T1 can be obtained based on the steel string motion frequency calculated by FFT and the time difference between T1 and T2, as well as the phase at time T0.

[0113] (2) Analyze ADC sampling data to obtain phase

[0114] First, ADC samples the steel string motion signal and obtains a set of discrete data. The initial sampling time T0 is known, and then the steel string motion frequency is calculated by FFT. According to the ADC sampling rate and the steel string motion frequency, it can be obtained that there are N sampling points in a steel string motion cycle, and the phase difference between two adjacent sampling points is 360° / N. Then the discrete data is analyzed to obtain whether the Mth point after T0 is a maximum or minimum value, and the running phase of the steel string at T0 can be inferred.

[0115] (3) Hardware comparator trigger method to obtain phase

[0116] The steel string motion signal is a sine wave. By comparing this signal with the zero-point voltage, trigger signals with two phases of 0° and 180° can be obtained.

[0117] A2. Dynamic and static combination technology

[0118] In addition, to ensure higher applicability, the present invention can flexibly switch between dynamic measurement and static measurement, using static measurement for routine monitoring and dynamic measurement when a flood discharge event occurs.

[0119] The dynamic and static combination is divided into two working conditions:

[0120] (1) Static measurement when dynamic measurement is not performed

[0121] Under this condition, the device will start a 1-second dynamic measurement and perform arithmetic averaging on the measurement data to use as static measurement data.

[0122] (2) Static measurement during dynamic measurement

[0123] Under this working condition, the device will extract 1 second of dynamic measurement data and perform arithmetic averaging to use as static measurement data.

[0124] The internal storage area is divided into a static data storage area and a dynamic data storage area to ensure that the partitioned storage does not interfere with each other.

[0125] A3. Review of flood discharge events

[0126] When the vibrating wire sensor continuously collects data, a short time window (such as 5 seconds) is set to calculate the characteristic value cv that represents the stable amplitude change. When the cv value of the signal exceeds 30%, it means that the dynamic mutation degree of the signal is large and the stability is poor. Among them:

[0127]

[0128] In the formula, σ represents the sample mean square error, and μ represents the sample mean.

[0129] When the amplitude changes stably and the characteristic value cv≤30%, it is determined that a flood discharge event has occurred. Otherwise, it is determined that the flood discharge has ended or is not a flood discharge condition, and the dynamic acquisition of the vibrating string sensor is stopped.

[0130] The dynamic data acquisition device can also end the dynamic data acquisition of the vibrating wire sensor according to a preset dynamic measurement duration.

[0131] 4. Step 4: Identification of modal parameters during dam operation

[0132] The signals obtained by dynamic monitoring of the vibrating-wire sensor are subjected to runtime modal analysis to identify the first four modal parameters of the dam. The identified modal parameters can be used for subsequent identification of mechanical parameters.

[0133] The runtime modal analysis method refers to the identification of structural modal parameters based only on the structural vibration response. It does not require the acquisition of excitation signals and is therefore widely used in health monitoring of structures such as bridges, high-rise buildings and dams. After decades of development, runtime modal analysis has developed a variety of methods including frequency domain and time domain. Taking the Cov-SSI algorithm as an example, as a time domain identification method, the algorithm has good noise resistance and high identification accuracy, and is one of the most widely used runtime modal analysis methods.

[0134] The discrete state space equation of the vibration system can be expressed as:

[0135]

[0136] In the formula, M, C, and K are the mass matrix, damping matrix, and stiffness matrix of the system respectively; z(t) are the acceleration vector, velocity vector and displacement vector of the discrete system respectively; F(t) is the excitation vector.

[0137] By introducing the state vector The vibration system (2) can be transformed into the corresponding state space:

[0138] x k+1 =Ax k +w k

[0139] y k =Cx k +v k (3)

[0140] Where A and C are the discrete-time system matrix and discrete-time output matrix, respectively.

[0141] When A and C of the state equation are obtained, perform eigenvalue decomposition on A:

[0142] A=ΦΛΦ -1 (4)

[0143] In the formula, Λ is the eigenvalue matrix and Φ is the eigenvector matrix. Arrange the diagonal elements of Λ in order of magnitude, and the modal parameters of the structure are:

[0144]

[0145] φ j =Cγ j (8)

[0146] Among them, μ j is the jth eigenvalue of A, γ j is the corresponding eigenvector, Δt is the sampling time interval, f j , j and φ j are the modal frequency, damping ratio and vibration shape vector respectively.

[0147] The modes calculated by the Cov-SSI algorithm contain false modes and need to be further processed by the stability diagram method. The stability diagram consists of stable points, which means that the modal parameters identified by a certain order system must meet the following stability criteria compared with the modal parameters identified by the previous order system:

[0148]

[0149] 1-MAC(φ i ,φ i+1 )<0.02 (11)

[0150] Where superscripts i and i+1 represent the system order, and MAC(·,·) is the modal assurance criterion, which is used to measure the similarity between two modal vibration shapes. Its expression is:

[0151]

[0152] in(·) H represents the conjugate transpose.

[0153] The Cov-SSI algorithm is used to identify the first four modal parameters of the dam. The original stability diagram and the stability diagram after removing the false modes are shown as follows: Figure 4 and Figure 5 After removing the false modes, the frequencies of the first four modes are obtained by averaging the stable points of different clusters of the stable diagram: [3.16, 3.28, 4.71, 6.08] (Hz).

[0154] 5. Step 5: Inverse analysis of dam mechanical parameters based on runtime modal analysis and proxy model

[0155] The inversion of dam mechanical parameters is completed by combining intelligent optimization algorithms (adaptive grid search algorithm, GA genetic algorithm, etc.) with the modal parameters identified by Cov-SSI.

[0156] The inversion of the mechanical parameters of the dam can be transformed into an optimization problem, that is, to find a set of solutions within the allowable parameter domain so that the response characteristics of the structure under the parameters are closest to the observed characteristics. The observed characteristics are the modal parameters obtained through vibration signal identification, and the distance between the measured characteristics, that is, the objective function in the optimization is:

[0157]

[0158] Where, θ is the structural mechanics parameter to be inverted; and represent the i-th order frequency and vibration mode of the modal parameters identified by the vibration signal; and f i (θ) and φ i (θ) are the i-th order frequency and vibration mode of the structural modal parameters obtained by finite element calculation; α i and β i are the weight coefficients of different items. Parameter inversion is equivalent to solving the optimization problem:

[0159]

[0160] Among them, D θ Represents the allowable value space of mechanical parameters.

[0161] Specifically, the mapping relationship between low-order modal parameters and dam mechanical parameters is used as a training set and a test set, and the proxy model is trained in combination with the objective function of the optimization algorithm, so that the proxy model outputs corresponding dam mechanical parameters based on the low-order modal parameters.

[0162] The examples are as follows:

[0163] Five different groups of modal parameters are set by random sampling, and the real parameter E real and the inversion parameter E inverse The relative error δ between them is calculated as follows:

[0164]

[0165] In the formula, |·| represents the bi-norm.

[0166] The inversion results are shown in Table 2. The inverted elastic moduli of the five test groups are very close to the true elastic moduli, and the maximum relative error is no more than 6%, which shows the effectiveness of the inversion method.

[0167] Table 2 Elastic modulus inversion results

[0168]

[0169] 6. Step 6: Risk identification based on inverse analysis of dam mechanical parameters

[0170] The above-mentioned inversion analysis method is used to obtain the mechanical parameters of the dam under the current operating state. According to the engineering design data or historical data, the changes in the mechanical parameters of the dam can be compared, and thresholds such as relative change rate can be set to identify the risks of the dam operation form. When the threshold is exceeded, a monitoring warning can be issued in time.

[0171] In summary, the present invention has the following innovations:

[0172] (1) Using vibration sensors + vibrating string sensors, each gate can be independently triggered for automated monitoring when a flood discharge event occurs.

[0173] (2) Dynamic monitoring of the vibrating string sensors buried during the construction period (frequency higher than 50 Hz) is achieved by using the same frequency and same phase excitation of the dynamic data acquisition device and the combined dynamic and static monitoring technology.

[0174] (3) Combine the runtime modal analysis and proxy model to realize the real-time inversion analysis of the dam mechanical parameters, and use the modal parameters to timely discover and warn of potential dangers in dam operation.

[0175] Based on the method of the present invention, the embodiment of the present disclosure also provides a dynamic monitoring system corresponding to the above method, such as Figure 2 As shown, it adopts a distributed network structure, which is divided into a monitoring management layer, a network communication layer, a data acquisition layer and a sensor layer.

[0176] The monitoring management level is equipped with data servers, workstations, switches and other equipment, as well as dam structure safety dynamic monitoring data management and analysis software.

[0177] A communication network with a bus or star topology is used between the monitoring management layer and the data acquisition layer. The monitoring data transmission generally adopts a combination of "wired communication + wireless communication", in which the wired method is used as the main channel, optical fiber Ethernet communication is used, and the wireless method is used as the backup channel, giving full play to the advantages of each network structure.

[0178] The dynamic data acquisition device of the data acquisition layer is connected to various sensors to collect sensor operation data of the dam gate, flood discharge orifice, dam foundation, dam crest, dam abutment and other parts. It can be determined whether a flood discharge event has occurred based on the vibration information of the dam gate. When a flood discharge event is confirmed, the vibrating string sensor can be immediately started for dynamic data sampling and data processing, and then communication is established with the computer equipment of the monitoring management layer through the network layer, and the data is sent to the data management and analysis software, thereby realizing online monitoring.

[0179] The sensor layer is used to collect sensor operation data at the dam gates, spillway openings, dam foundation, dam crest, dam abutments, etc. The sensors include vibration sensors installed at the gates and nearby areas; vibrating string sensors installed at the spillway openings, dam foundation, dam crest, dam abutments, etc.; and GNSS positioning engines installed at the dam crest. Figure 3 shown.

[0180] The vibration sensor includes a displacement sensor, a velocity sensor, an acceleration sensor, etc., which is responsible for vibration sensing and is used to trigger the dynamic monitoring of the vibrating string sensor.

[0181] The vibrating string sensors include crack gauges, strain gauges, steel bar stress gauges, etc., which are used to dynamically monitor the dynamic mechanical response characteristics of the dam concrete. During the construction of the dam, a large number of traditional safety monitoring sensors were buried inside the dam to statically monitor its structural form. After completion, it is generally impossible to bury new special monitoring sensors in the dam body. Vibrating string sensors are one of the most widely used monitoring sensors in the field of dam safety monitoring. If you want to continue to dynamically monitor the structural form of the dam in real time during flood discharge conditions, the ideal way is to continue to make full use of these traditional buried vibrating string sensors.

[0182] Considering that the dam structure will produce certain vibrations under flood discharge conditions, resulting in rapid changes in the opening and closing degree of the transverse joints between the dam sections, and due to the whip effect, the deformation of the middle and high elevation joints of the arch crown beam dam section may be relatively large, 10 transverse joint gauges at different elevations of 600 to 650 m near the 15# dam section were selected for dynamic monitoring.

[0183] In order to capture the dynamic mechanical response characteristics of concrete in the strong constraint area of ​​the dam under flood discharge conditions, 8 groups of 5-way strain gauges at elevations of 350m to 410m in sections 8#, 9#, 20# and 21# were selected and connected to the dynamic monitoring system.

[0184] Since the stress is generally concentrated at the orifice of the dam and the load changes are complex during gate operation and flood discharge, in order to understand the stress changes of concrete steel bars at the orifice of the dam under flood discharge conditions, 220 steel bar stress gauges near each orifice were selected and connected to the dynamic monitoring system.

[0185] The GNSS positioning engine is used to collect water level information upstream and downstream of the gate.

[0186] Although the present disclosure has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. A method for monitoring the operation risk of a high dam of a hydropower station, characterized in that: include: Construct a finite element simulation model of the dam; Based on the finite element simulation model, the mapping relationship between the low-order modal data of the dam under the simulated flood discharge condition and the dam mechanical parameters is obtained; The vibration sensor is used to continuously collect vibration data at the dam gate A, and combined with the upstream and downstream water level data of the dam gate A, it is determined that a flood discharge event has occurred at the dam gate A; Perform runtime modal analysis on the signal obtained by the vibrating wire sensor corresponding to the dam gate A where the flood discharge event occurred, and obtain the low-order modal parameters of the dam; Based on the low-order modal parameters of the dam and the mapping relationship between the low-order modal parameters and the mechanical parameters of the dam, the mechanical parameters of the dam are inverted through the optimization algorithm; The operation risk of the high dam of the hydropower station is judged based on the inversion results of the dam mechanical parameters.

2. The method for monitoring the operation risk of a high dam of a hydropower station according to claim 1, characterized in that: The finite element simulation model of the dam is constructed; Based on the finite element simulation model, the mapping relationship between the low-order modal data of the dam under the simulated flood discharge condition and the dam mechanical parameters is obtained; including: Establish multiple finite element simulation models based on engineering survey data and design data, and the multiple finite element simulation models together constitute the dam finite element simulation model; White noise is loaded on the boundaries around the dam foundation of the finite element simulation model of the dam, and random vibration is used to simulate the flow-induced vibration load on the dam during the discharge process; Finite element transient dynamic analysis is performed on the finite element simulation model of the dam to obtain the dynamic response of the dam under white noise vibration, and then the acceleration time history signal of the monitoring point is obtained. Based on the modal analysis method, the n-order low-order modal parameters of the dam are obtained. The dam mechanical parameters and low-cost modal parameters of the dam finite element simulation model are used as sample data and stored in the database; Based on the sample data in the database, the mapping relationship between the low-order modal parameters and the dam mechanical parameters is constructed using the proxy model.

3. The method for monitoring the operation risk of a high dam of a hydropower station according to claim 1, characterized in that: The method of continuously collecting vibration data at the dam gate A by using a vibration sensor and combining the upstream and downstream water level data of the dam gate A to determine whether a flood discharge event has occurred at the dam gate A includes: A vibration sensor is installed at the dam gate A to continuously collect the vibration amplitude, main vibration frequency, and other dominant frequency data at various locations of the dam gate A; Use the GNSS positioning engine to obtain the upstream and downstream water level data of the dam gate A; Combine historical water levels with flood discharge event information to determine whether a flood discharge event is likely to occur under current upstream and downstream water level data; Compare the vibration amplitude, main vibration frequency, and other dominant frequency data at each location of the dam gate A with historical data; if a certain value exceeds the threshold, it is preliminarily determined that a flood discharge event has occurred at the orifice of the dam gate A; A vibrating string sensor is used to dynamically measure the dam gate A, which is preliminarily judged to have a flood discharge event, at a certain frequency to obtain frequency data B. At regular intervals, a characteristic value cv representing the smoothness of the amplitude change of the frequency data B is calculated. If the cv corresponding to the dam gate A exceeds the threshold, it is determined that a flood discharge event has occurred at the dam gate A.

4. The method for monitoring the operation risk of a high dam of a hydropower station according to claim 1, characterized in that: The same frequency and same phase excitation technology is used as the resonance excitation of the vibrating string sensor, and the output of the excitation signal is realized by DAC; The same-frequency and same-phase excitation technology uses a single pulse wave with the same resonance frequency and phase as the steel string of the vibrating-string sensor to maintain the resonance state of the steel string; the frequency and phase of the single pulse wave are obtained by acquiring the resonance frequency and real-time motion phase of the steel string.

5. The method for monitoring the operation risk of a high dam of a hydropower station according to claim 4, characterized in that: The method of obtaining the real-time motion phase of the steel string comprises: The steel string motion signal is obtained by ADC sampling and a set of discrete data is obtained; The initial sampling time T0 is known, and the phase of the first data is obtained by using FFT to calculate discrete data, and the running phase of the steel string at time T0 is obtained; assuming that the time of excitation is T1, the steel string movement phase at time T1 is obtained based on the steel string movement frequency calculated by FFT and the time difference between T1 and T2, as well as the phase at time T0.

6. The method for monitoring the operation risk of a high dam of a hydropower station according to claim 1, characterized in that: The Cov-SSI algorithm is used to perform runtime modal analysis on the signals obtained by the vibrating wire sensor corresponding to gate A to obtain the low-order modal parameters of the dam. It includes: Construct a discrete state space equation of the vibration system; the discrete state space equation is: In the formula, M, C, and K are the mass matrix, damping matrix, and stiffness matrix of the system respectively; z(t) is the acceleration vector, velocity vector and displacement vector of the discrete system; F(t) is the excitation vector; Introducing the state vector The vibration system is transformed into the corresponding state space; the state space is: x k+1 =Ax k +w k y k =Cx k +v k Where A and C are the discrete time system matrix and discrete time output matrix respectively; After finding A and C, perform eigenvalue decomposition on A: A=FLF -1 Where Λ is the eigenvalue matrix, Φ is the eigenvector matrix; Arrange the diagonal elements of Λ in order of magnitude, and obtain the preliminary modal parameters of the dam: f j =Cγ j Among them, μ j is the jth eigenvalue of A, γ j is the corresponding eigenvector, Δt is the sampling time interval, f j , j and φ j are the modal frequency, damping ratio and vibration shape vector respectively; The preliminary modal parameters of the dam are processed using the stability diagram method to obtain the low-order modal parameters of the dam.

7. The method for monitoring the operation risk of a high dam of a hydropower station according to claim 6, characterized in that: In the stability graph method, the stability graph is composed of stability points, and the stability points satisfy the following stability criteria: 1-MAC(φ i ,f i+1 )<0.02 In the formula, superscripts i and i+1 represent the system order, MAC(·,·) is the modal assurance criterion, which is used to measure the similarity between two modal vibration shapes. The expression of MAC(·,·) is: in,(·) H represents the conjugate transpose.

8. The method for monitoring the operation risk of a high dam of a hydropower station according to claim 1, characterized in that: The method of inverting the dam mechanical parameters by using an optimization algorithm based on the low-order modal parameters of the dam and the mapping relationship between the low-order modal parameters and the mechanical parameters of the dam comprises: The objective function of the optimization algorithm is constructed. The calculation goal of the objective function is to make the response characteristics of the dam structure under the dam mechanical parameters closest to the observed characteristics; the objective function is: Where, θ is the structural mechanics parameter to be inverted; and They represent the i-th order frequency and vibration mode of the modal parameters identified by the vibration signal; f i (θ) and φ i (θ) are the i-th order frequency and vibration mode of the structural modal parameters obtained by finite element calculation; α i and β i are the weight coefficients of different items; The mapping relationship between low-order modal parameters and dam mechanical parameters is used as a training set and a test set, and the proxy model is trained in combination with the objective function of the optimization algorithm, so that the proxy model outputs corresponding dam mechanical parameters based on the low-order modal parameters.

9. The method for monitoring the operation risk of a high dam of a hydropower station according to claim 1, characterized in that: Also includes: Conduct static monitoring of the operational risks of high dams of hydropower stations, including: When the high dam of the hydropower station is not in a state of flood discharge, dynamic data measurement of the high dam of the hydropower station is performed for a duration of x seconds, and the measured data are arithmetic averaged as static measurement data; When the high dam of the hydropower station is in a state of flood discharge, the high dam data of the hydropower station is continuously measured, and the dynamic measurement data of x seconds are extracted and arithmetic averaged as static measurement data; Based on static measurement data and inverted dam mechanical parameters, it is jointly judged whether there is an operational risk of the high dam of the hydropower station.

10. An operation risk monitoring system for a high dam of a hydropower station, characterized in that: include: Model building module, used to build a finite element simulation model of the dam; Based on the finite element simulation model, the mapping relationship between the low-order modal data of the dam under the simulated flood discharge condition and the dam mechanical parameters is obtained; The flood discharge event judgment module is used to continuously collect vibration data at the dam gate A using a vibration sensor, and combine the upstream and downstream water level data of the dam gate A to judge whether a flood discharge event has occurred at the dam gate A; The modal parameter inversion module is used to perform runtime modal analysis on the signal obtained by the vibrating wire sensor corresponding to the dam gate A where the flood discharge event occurred, and obtain the low-order modal parameters of the dam; Mechanical parameter inversion module, which is used to invert the mechanical parameters of the dam through optimization algorithm based on the low-order modal parameters of the dam and the mapping relationship between the low-order modal parameters and the mechanical parameters of the dam; The judgment module is used to judge the operation risk of the high dam of the hydropower station based on the inversion results of the dam mechanical parameters.

Images