A real-time identification method and health detection system for ship-impact bridge pier load

Abstract

The present invention relates to a real-time identification method and health detection system for ship-impacted bridge pier loads. Sensors are rationally arranged on the piers, and a health detection system is configured; collected signals are transmitted to a remote terminal via an industrial router; a data storage system and a data preprocessing system of the remote terminal are established; a finite element model of the pier is established on the terminal computer device, and the finite element model is corrected according to the collected data; a suitable load identification algorithm is selected according to the finite element model, and the finite element model is integrated into the load identification algorithm; based on the above steps, a load identification preliminary experiment is performed. Once the pier is impacted by a ship collision, the impact load time history is inverted based on the collected data and the load identification algorithm. The impact load time history identified by the present invention can be used as a reference for simulation of ship-impacted bridge piers, and can also be used for performance evaluation of the piers after being hit, providing a theoretical evaluation index for the overall safety performance of the bridge.

Description

Technical Field

[0001] The invention relates to a real-time identification method for a ship-collision load on a bridge pier and a health detection system, belonging to the field of bridge health detection. Background Art

[0002] With the continuous construction of various large-scale cross-river and cross-sea bridges, ship collisions with bridge piers have become an increasingly real problem that cannot be ignored [1]. In recent years, the number and scale of ship collisions with bridge piers have increased dramatically [2], and accidents involving ship collisions with bridge piers are not uncommon. Ship collisions with bridge piers are increasingly becoming a huge threat to the normal service of bridges and the safety of people's lives and property. Ship collisions with bridge piers can cause direct damage to the piers, which can easily lead to direct collapse of the piers; they can also indirectly affect the service life of the piers, posing a safety hazard to the normal service of the bridge in the future.

[0003] Existing research mainly focuses on how to avoid ship collisions with bridge piers[3][4]. This can reduce the probability of ship collisions with bridge piers to a certain extent. However, in many cases, ship collisions with bridge piers are caused by the loss of control of the ship. The phenomenon of ship collisions with bridge piers is still difficult to avoid. Once a bridge pier is hit, it is particularly important to evaluate the safety performance of the bridge pier after the collision. Previously, some research work on the safety performance of bridge piers after collision has been carried out[5][6][7]. These works mainly simulate the process of ship collision with bridge piers based on finite element software[8][9], and then evaluate the magnitude of ship collision force and safety performance after collision from a mechanical perspective

[10] . Among them, the ship collision force, as a load excitation of the bridge pier system, plays a vital role in the safety evaluation of the bridge pier after collision. However, the academic community has not yet fully understood the ship collision force. In the simulation process, a simplified numerical model is used to replace the ship collision force

[11] . This is obviously different from the actual ship collision force on the bridge pier. Therefore, it is particularly important to accurately obtain the actual ship collision load on the bridge pier for the state assessment and life prediction of the hit bridge pier.

[0004] However, it is not feasible to directly measure the ship impact force. On the one hand, the load value generated by the ship hitting the bridge pier is large, and it is difficult to install a load sensor with a sufficient range; on the other hand, the position of the ship hitting the bridge pier is random, which makes it difficult to determine the installation position of the load sensor. Therefore, a method for identifying the load of the ship hitting the bridge pier is urgently needed, which can identify the ship impact force in real time online and reveal the form of the ship impact force, providing important reference value for the engineering and academic circles to study the ship hitting the bridge pier.

[0005] References used above:

[0006] References:

[0007] [1] Yin Xijun. Research on ship-bridge collision and bridge pier anti-collision facilities[D]. Dalian: Dalian Maritime University, 2009.

[0008] [2] Chen Guoyu, Wang Lili. Ship collision with bridge and its prevention[M]. China Railway Publishing House, 2006.

[0009] [3] Liu Bo. Research on the application of thermal imaging technology in bridge anti-ship collision warning system[D]. Guangzhou: Guangzhou University, 2016.

[0010] [4] Zhang Jingfeng, Li Xiaozhen. Simulation analysis of anti-collision performance of steel casing-pile group composite anti-collision structure[J]. Bridge Construction, 2018, 48(6): 47-52.

[0011] [5]Fan W, Zhang Y, Liu B.Modal Combination Rule for Shock SpectrumAnalysis ofBridgeStructures Subjected to Barge Collisions[J].Journal ofEngineering Mechanics,2016,142(2):04045083.

[0012] [6] Wang Beike. Hydrodynamic simulation study on anti-ship collision arrest facilities of non-navigable openings of large bridges[D]. Ningbo: Ningbo University, 2017.

[0013] [7] Sun Yang. Study on the vulnerability of bridge to ship collision considering multiple factors[D]. Changsha: Hunan University, 2020.

[0014] [8] Wang Junjie, Chen Cheng, Wang Hong, et al. Equivalent static ship impact force on bridge based on collision numerical simulation - basic formula [J]. Technology of Highway and Transportation, 2009(2):66-70.

[0015] [9] Hu Zhiqiang, Gu Yongning, Gao Zhen, et al. Rapid estimation of ship-bridge collision force based on nonlinear numerical simulation [J]. Engineering Mechanics, 2005, 22(3): 234-240.

[0016]

[10] Chen Cheng. Simulation study on ship impact force and damage state of bridge design[D]. Shanghai: Tongji University.

[0017]

[11] SHAO Jun-hu, SONG Shuai, WU Qi-hong, et al. Research progress on bridge-ship collision force calculation model[J]. Journal of Chengdu University (Natural Science Edition), 2020, 39(1): 15-21. Summary of the invention

[0018] The present invention provides a real-time identification method and health detection system for ship impact loads on bridge piers, which overcomes the defect that the existing ship impact simulation research on bridge piers is difficult to accurately simulate the impact loads generated by the ship impact on the bridge piers, and can accurately obtain the ship impact force on the bridge piers in real time.

[0019] The technical solution adopted by the present invention to solve its technical problem is:

[0020] A real-time identification method for a ship-collision load on a bridge pier comprises the following steps:

[0021] Step S1: Arrange a number of acceleration sensors on the bridge piers and configure a matching health detection system;

[0022] Step S2: a plurality of acceleration sensors collect bridge pier information data, and transmit the information data to a remote terminal via an industrial router;

[0023] Step S3: The remote terminal integrates the collected bridge pier information data and pre-processes it through a data pre-processing system, and establishes a data storage system for the pre-processed data information;

[0024] Step S4: establishing a finite element model of the bridge pier at the remote terminal, and modifying the finite element model according to the established data storage system;

[0025] Step S5: Based on the finite element model corrected in step S4, the mass, stiffness and damping matrix of the finite element model are derived to select a matching load identification algorithm, and the finite element model is integrated into the load identification algorithm; wherein the selected load identification algorithm is a load identification algorithm based on an increased rank Kalman filter;

[0026] Step S6: When the bridge pier is impacted by a ship collision, the impact load time history is inverted based on the bridge pier information data collected by the sensors in step S2 and combined with the load identification algorithm in step S5 to verify that the identification algorithm in step S5 can effectively identify the ship collision force;

[0027] As a further preferred embodiment of the present invention, in step S1, six acceleration sensors are set at different positions of the bridge pier, wherein the selected acceleration sensor has a range of 50g and a resolution of 0.0005m / s 2 ;

[0028] In step S2, the remote terminal is a data collector, and the data collector is an 8-channel 24-bit data collector;

[0029] As a further preferred embodiment of the present invention, in step S3, the capacity of the data storage system is selected to be 1000TB;

[0030] As a further preferred embodiment of the present invention, in step S3, the data preprocessing system preprocesses the collected bridge pier information data including resampling, detrending, filtering and denoising;

[0031] As a further preferred embodiment of the present invention, the finite element model established in step S4 has a natural frequency obtained after modal analysis, which is the same as the vibration frequency of the bridge pier information data acquired by the acquisition instrument;

[0032] As a further preferred embodiment of the present invention, in step S5, the selected load identification algorithm includes the following steps:

[0033] Step S51: Construct the dynamic equation of the bridge pier, specifically:

[0034]

[0035] In formula (1), M b is the mass of the pier, C b is the damping of the bridge pier, K b is the stiffness matrix of the pier, y b (t) is the displacement vector of the pier, H(t) is the position matrix of the node force, and the position matrix is

[0036] H=[0 … 1 … 0] T (2)

[0037] In formula (2), the element 1 corresponds to the node where the ship impact force acts, and the element 0 indicates that there is no load at that location; Step S52: Establish the state equation of the pier structure based on formula (1):

[0038]

[0039] In formula (3), M b is the mass of the pier, C b is the damping of the bridge pier, K b is the stiffness matrix of the pier, y b (t) is the displacement vector of the pier;

[0040] Step S53: Discretize formula (3) to obtain the state equation of the bridge pier:

[0041] X k+1 =A d X k +B d f k +ω k (4)

[0042] In formula (4), f k is the load;

[0043] At the same time, the measurement equation of the pier state is

[0044] d k =C d Xk +D d f k +ξ k (5)

[0045] In formula (5), C d is the observation matrix, D d is the feedforward matrix, ξ k is the observation noise, f k is the load, X k is the state vector, d k is the observed value;

[0046] The load f in formula (4) and formula (5) is k The recursive relation is expressed as:

[0047] f k+1 =f k +η k (6)

[0048] In formula (6), η k is the load f k The increment at the kth time step;

[0049] Step S54: Substitute the state equation of formula (4) and the measurement equation of formula (5) into k Increase the rank to the state vector X k In this paper, we obtain the following system of simultaneous equations:

[0050]

[0051] In formula (7), C d is the observation matrix, D d is the feedforward matrix, η k is the load f k The increment at the kth time step, X k is the state vector, ω k is the process noise, d k is the observed value;

[0052] Step S55: When using the Kalman filter for load identification, prediction is first performed, wherein the rank-increasing state prediction formula is

[0053]

[0054] And the prediction error covariance matrix estimate

[0055]

[0056] In formula (8), is the estimated state vector at the kth moment, is the optimal estimate of the state vector at the k-1th moment. In formula (9), is the prediction error of the process, T is ζ k The covariance matrix of represents the error of the system model;

[0057] Step S56: Update the equations (8) and (9) predicted in step S55 to obtain the equation system:

[0058]

[0059] In formula (10), is the estimated state vector at the kth moment, is the prediction error of the process, is the optimal estimate of the state vector at the k-1th moment, P k is the estimation error of the optimal estimate, K k is the Kalman gain, T is ζ k The covariance matrix of represents the error of the system model, S is ξ k The covariance of , which represents the error of the observations;

[0060] Step S57: Select the value of covariance S in formula (10) in advance, iterate the load identification process of step S55-step S56, calculate the error of each iteration, select the value of covariance S when the error is minimum, and substitute the value of covariance S when the error is minimum into formula (10), where the error calculation method of covariance S is:

[0061]

[0062] In formula (11), x k is the observed value estimated by the state vector, d k is the observed value, κ is the error about each value of the covariance S;

[0063] Step S58: Combine T and S to calculate the optimal estimate of the state vector through the AKF algorithm, that is, estimate the optimal rank-increasing state vector as

[0064]

[0065] In formula (12), is the estimated ship impact force history;

[0066] As a further preferred embodiment of the present invention, in step S6, a force hammer is installed at the bridge pier, the force hammer is connected to the data acquisition instrument, the force hammer is started to strike the bridge pier, the data acquisition instrument obtains the load time history of the force hammer and the acceleration response of the bridge pier, and the acceleration response of the bridge pier is compared with the load time history of the force hammer to confirm the accuracy of the load identification method of steps S1 to S5;

[0067] As a further preferred embodiment of the present invention, the force measuring range of the selected hammer is 12.5T and the sensitivity is 4pc / N;

[0068] A health detection system based on the real-time identification method of the ship-impact-bridge-pier load comprises a plurality of acceleration sensors arranged at different positions of the bridge pier, a data storage system, a collector and an industrial router. The plurality of acceleration sensors are connected to the collector via the industrial router, and a data storage system is established at the collector.

[0069] Through the above technical solution, compared with the prior art, the present invention has the following beneficial effects:

[0070] 1. The real-time identification method of the ship-impact-bridge-pier load provided by the present invention can identify in real time the impact load of the bridge pier being hit by the bridge;

[0071] 2. The real-time identification method of the ship-impact-bridge-pier load provided by the present invention uses the identified impact load time history as a reference for the simulation of the ship-impact-bridge-pier, and can also be used for the performance evaluation of the bridge pier after being hit, forming a relatively important theoretical evaluation index for the overall safety performance of the bridge;

[0072] 3. The real-time identification method of ship impact load on bridge piers provided by the present invention solves the problem that the noise level in the field measurement data is difficult to accurately quantify, and can obtain a more accurate ship impact force time history. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0074] Figure 1 It is a flow chart of the real-time identification method of the ship collision load on the bridge pier provided by the present invention;

[0075] Figure 2 It is a flow chart of a load identification algorithm based on an increased rank Kalman filter provided by the present invention;

[0076] Figure 3 This is the observation value noise covariance iteration process in the real-time identification method of the ship collision bridge pier load provided by the present invention. DETAILED DESCRIPTION

[0077] As described in the background technology, current simulation studies on ship impacts with bridge piers usually use simplified numerical models to replace the ship impact force. Therefore, this is quite different from the ship impact force actually received by the bridge pier, resulting in an inability to obtain more accurate results. Therefore, the present application aims to provide a real-time identification method for the load of a ship impacting a bridge pier. The method uses load identification technology as a basis, measures the response of key points on the bridge pier to invert the time course of the bridge impact force, and can identify the ship impact force in real time online. At the same time, the present application also has a more unique innovation point. Based on the load identification technology, it reduces the degree of influence of noise on the signal during the identification process.

[0078] First of all, before explaining the identification method, this application needs to explain the health detection system configured by this application to achieve the purpose of accurate identification, such as Figure 1 As shown, it includes multiple acceleration sensors, data storage systems, collectors and industrial routers arranged at different positions of the piers. Multiple acceleration sensors are connected to the collector through the industrial router, and a data storage system is established at the collector. Multiple acceleration sensors transmit the collected signals to the remote terminal (here, the collector) through the industrial router. The data processed by the data preprocessing system is stored in the data storage system. The finite element model of the pier established by the remote terminal selects a suitable load identification algorithm, and the finite element model is integrated into the load identification algorithm to conduct a load identification pre-experiment. In actual operation, it is necessary to ensure that the health detection system is powered for a long time. Therefore, the entire health detection system needs to be waterproof, high temperature resistant, lightning weather resistant, electromagnetic interference resistant and durable to ensure that the system can perform long-term detection normally.

[0079] The real-time identification method of the ship-collision load on the bridge pier provided in the present application specifically comprises the following steps:

[0080] Step S1: Arrange several acceleration sensors on the bridge piers and configure a matching health detection system; in the preferred embodiment provided in this application, six acceleration sensors are arranged at different positions of the bridge piers. The positions of the acceleration sensors are arranged according to actual needs, with the aim of covering as many parts as possible that need to be detected; the selected acceleration sensor has a range of 50g and a resolution of 0.0005m / s 2 .

[0081] Step S2: Several acceleration sensors collect bridge pier information data and transmit the information data to a remote terminal through an industrial router; the above description states that the remote terminal is a collector, and preferably, an 8-channel 24-bit collector is used. The installation of the industrial router should also avoid electromagnetic interference. At the same time, the industrial router needs to be reasonably configured to ensure sufficient data transmission to meet the sampling frequency requirements of the health detection system.

[0082] Step S3: The remote terminal integrates the collected pier information data and pre-processes it through the data pre-processing system. The pre-processed data information establishes a data storage system; the capacity of the data storage system must meet the long-term storage requirements of the health detection system, and the recommended capacity is 1000TB. At the same time, the data pre-processing system needs to be able to pre-process the collected data, such as resampling, detrending, filtering, denoising, and other time-frequency domain processing methods.

[0083] Step S4: Establish a finite element model of the bridge pier at the remote terminal, and modify the finite element model according to the established data storage system; the establishment of the finite element model here needs to be based on the bridge pier design drawings, and in order to ensure the correctness and reliability of the finite element model, the natural frequency obtained after modal analysis of the established finite element model needs to be consistent with the vibration frequency of the data collected by the monitoring system.

[0084] Step S5: Based on the finite element model corrected in step S4, the mass, stiffness and damping matrix of the finite element model are derived to select a matching load identification algorithm, and the finite element model is integrated into the load identification algorithm; wherein the selected load identification algorithm is a load identification algorithm based on an increased rank Kalman filter, which has good robustness and can realize real-time identification of ship collision force;

[0085] Here we specifically explain the steps of the load identification algorithm based on the increased rank Kalman filter, such as Figure 2 As shown,

[0086] Step S51: Construct the dynamic equation of the bridge pier, specifically:

[0087]

[0088] In formula (1), M b is the mass of the pier, C b is the damping of the bridge pier, K b is the stiffness matrix of the pier, y b (t) is the displacement vector of the pier, H(t) is the position matrix of the node force, and the position matrix is

[0089] H=[0 … 1 … 0] T (2)

[0090] In formula (2), the element 1 corresponds to the node where the ship impact force acts, and the element 0 indicates that there is no load at that location; Step S52: Establish the state equation of the pier structure based on formula (1):

[0091]

[0092] In formula (3), M b is the mass of the pier, Cb is the damping of the bridge pier, K b is the stiffness matrix of the pier, y b (t) is the displacement vector of the pier;

[0093] Step S53: Discretize formula (3) to obtain the state equation of the bridge pier:

[0094] X k+1 =A d X k +B d f k +ω k (4)

[0095] In formula (4), f k is the load;

[0096] At the same time, the measurement equation of the pier state is

[0097] d k =C d X k +D d f k +ξ k (5)

[0098] In formula (5), C d is the observation matrix, D d is the feedforward matrix, ξ k is the observation noise, f k is the load, X k is the state vector, d k is the observed value; here the observation matrix C d With the feedforward matrix D d The response forms are different when different positions or different types of responses are used, both of which are customized according to the actual measured values;

[0099] The load f in formula (4) and formula (5) is k The recursive relation is expressed as:

[0100] f k+1 =f k +η k (6)

[0101] In formula (6), η k is the load f k The increment at the kth time step;

[0102] Step S54: Substitute the state equation of formula (4) and the measurement equation of formula (5) into k Increase the rank to the state vector X kIn this paper, we obtain the following system of simultaneous equations:

[0103]

[0104] In formula (7), C d is the observation matrix, D d is the feedforward matrix, η k is the load f k The increment at the kth time step, X k is the state vector, ω k is the process noise, d k is the observed value; Step S55: The Kalman filter can comprehensively consider the prior estimation and the observed value of the state, that is, the state equation starts from the finite element model, and the measurement equation starts from the observed value. The two finally realize the load estimation together and give the optimal a posteriori estimate of the state vector. The Kalman filter mainly has the following two steps when performing load identification. The first is to make a prediction, where the rank-increasing state prediction formula is

[0105]

[0106] And the prediction error covariance matrix estimate

[0107]

[0108] In formula (8), is the estimated state vector at the kth moment, is the optimal estimate of the state vector at the k-1th moment. In formula (9), is the prediction error of the process, T is ζ k The covariance matrix of represents the error of the system model;

[0109] Step S56: Update the equations (8) and (9) predicted in step S55 to obtain the equation system:

[0110]

[0111] In formula (10), is the estimated state vector at the kth moment, is the prediction error of the process, is the optimal estimate of the state vector at the k-1th moment, P k is the estimation error of the optimal estimate, K k is the Kalman gain, T is ζ k The covariance matrix of represents the error of the system model, S is ξ k The covariance of , which represents the error of the observations;

[0112] Step S57: The noise level in the field measurement data is difficult to quantify accurately, and the degree of influence of noise on the signal may also be different, which makes it difficult to use the formula To accurately calculate the covariance of the noise, a major innovation of the present application is to select the value of the covariance S in formula (10) in advance, iterate the load identification process of step S55-step S56, calculate the error of each iteration, select the value of the covariance S when the error is minimum (here the minimum error value is the optimal value of the covariance S), and substitute the value of the covariance S when the error is minimum into formula (10), where the error calculation method of the covariance S is:

[0113]

[0114] In formula (11), x k is the observed value estimated by the state vector, d k is the observed value, κ is the error about each value of covariance S, when x k and d k When the difference between is the smallest, the estimated state vector is the most accurate, which is given here. Figure 3 To intuitively display the iterative process diagram of the observation value noise covariance;

[0115] Step S58: Combine T and S to calculate the optimal estimate of the state vector through the AKF algorithm. Here, the smaller T is, the smaller the model error is, and the more the Kalman gain trusts the pre-estimate; conversely, the more the observed value is trusted. Finally, the optimal rank-increased state vector is estimated to be

[0116]

[0117] In formula (12), is the estimated ship impact force history.

[0118] Step S6: When the bridge pier is impacted by a ship collision, the impact load time history is inverted based on the bridge pier information data collected by the sensors in step S2 and combined with the load identification algorithm in step S5 to verify that the identification algorithm in step S5 can effectively identify the ship collision force;

[0119] Specifically, a force hammer is installed at the bridge pier, the force hammer is connected to the data collector, the force hammer is started to hit the bridge pier, the data collector obtains the load time history of the force hammer and the acceleration response of the bridge pier, and the acceleration response of the bridge pier is compared with the load time history of the force hammer. If the two time history are consistent, it means that the remote load identification system can effectively complete the identification of ship impact force.

[0120] Here, the recommended force measurement range of the hammer is 12.5T and the sensitivity is 4pc / N.

[0121] The present application can identify the impact load on the bridge pier caused by the impact of the bridge. The identified impact load time series can be used as a reference for the simulation of a ship hitting a bridge pier, and can also be used for the performance evaluation of the bridge pier after being hit, providing a theoretical evaluation index for the overall safety performance of the bridge.

[0122] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art to which this application belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with the meanings in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless defined as herein.

[0123] The meaning of "and / or" described in this application means that the situations where each exists alone or both exist at the same time are included.

[0124] The term “connection” as used in this application may mean a direct connection between components or an indirect connection between components via other components.

[0125] Based on the above ideal embodiments of the present invention, the relevant staff can make various changes and modifications without departing from the technical concept of the present invention through the above description. The technical scope of the present invention is not limited to the contents of the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. A real-time identification method for ship impact load on bridge pier, characterized in that: The specific steps include: Step S1: Arrange a number of acceleration sensors on the bridge piers and configure a matching health detection system; Step S2: a plurality of acceleration sensors collect bridge pier information data, and transmit the information data to a remote terminal via an industrial router; Step S3: The remote terminal integrates the collected bridge pier information data and pre-processes it through a data pre-processing system, and establishes a data storage system for the pre-processed data information; Step S4: establishing a finite element model of the bridge pier at the remote terminal, and modifying the finite element model according to the established data storage system; Step S5: Based on the finite element model corrected in step S4, the mass, stiffness and damping matrix of the finite element model are derived to select a matching load identification algorithm, and the finite element model is integrated into the load identification algorithm; wherein the selected load identification algorithm is a load identification algorithm based on an increased rank Kalman filter; The selected load identification algorithm includes the following steps: Step S51: Construct the dynamic equation of the bridge pier, specifically: In formula (1), M b is the mass of the pier, C b is the damping of the bridge pier, K b is the stiffness matrix of the pier, y b (t) is the displacement vector of the pier, H(t) is the position matrix of the node force, and the position matrix is H＝[0…1…0] T (2) In formula (2), the element 1 corresponds to the node where the ship impact force acts, and the element 0 means that there is no load at that location; Step S52: Establish the state equation of the pier structure based on formula (1): In formula (3), M b is the mass of the pier, C b is the damping of the bridge pier, K b is the stiffness matrix of the pier, y b (t) is the displacement vector of the pier; Step S53: Discretize formula (3) to obtain the state equation: The state equation of the bridge pier is: X k+1 =A d X k +B d f k +ω k (4) In formula (4), f k is the load; At the same time, the measurement equation of the pier state is d k =C d X k +D d f k +ξ k (5) In formula (5), C d is the observation matrix, D d is the feedforward matrix, ξ k is the observation noise, f k is the load, is the state vector, d k is the observed value; The load f in formula (4) and formula (5) is k The recursive relation is expressed as: f k+1 =f k +n k (6) In formula (6), η k is the load f k The increment at the kth time step; Step S54: Substitute the state equation of formula (4) and the measurement equation of formula (5) into k Adding rank to the state vector, we get the following system of simultaneous equations: In formula (7), C d is the observation matrix, D d is the feedforward matrix, X k is the state vector, d k is the observed value; Step S55: When using the Kalman filter for load identification, prediction is first performed, wherein the rank-increasing state prediction formula is And the prediction error covariance matrix estimate In formula (8), is the estimated state vector at the kth moment, is the optimal estimate of the state vector at the k-1th moment. In formula (9), is the prediction error of the process, T is ζ k The covariance matrix of represents the error of the system model; Step S56: Update the equations (8) and (9) predicted in step S55 to obtain the equation system: In formula (10), is the estimated state vector at the kth moment, is the prediction error of the process, is the optimal estimate of the state vector at the k-1th moment, P k is the estimation error of the optimal estimate, K k is the Kalman gain, T is ζ k The covariance matrix of represents the error of the system model, S is ξ k The covariance of , which represents the error of the observations; Step S57: Select the value of covariance S in formula (10) in advance, iterate the load identification process of step S55-step S56, calculate the error of each iteration, select the value of covariance S when the error is minimum, and substitute the value of covariance S when the error is minimum into formula (10), where the error calculation method of covariance S is: In formula (11), x k is the observed value estimated by the state vector, d k is the observed value, κ is the error about each value of the covariance S; Step S58: Combine T and S to calculate the optimal estimate of the state vector through the AKF algorithm, that is, estimate the optimal rank-increasing state vector as In formula (12), is the estimated ship impact force history; Step S6: When the bridge pier is impacted by a ship collision, the impact load time history is inverted based on the bridge pier information data collected by several sensors in step S2 and combined with the load identification algorithm in step S5 to verify that the identification algorithm in step S5 can effectively complete the identification of the ship collision force.

2. The real-time identification method of ship impact load on bridge pier according to claim 1 is characterized by: In step S1, six acceleration sensors are set at different positions of the bridge pier, wherein the selected acceleration sensor has a range of 50g and a resolution of 0.0005m / s 2 ; In step S2, the remote terminal is a data collector, and the data collector is an 8-channel 24-bit data collector.

3. The real-time identification method of ship impact load according to claim 2 is characterized by: In step S3, the capacity of the data storage system is selected to be 1000TB.

4. The real-time identification method of ship impact load on bridge pier according to claim 3 is characterized by: In step S3, the data preprocessing system preprocesses the collected bridge pier information data including resampling, detrending, filtering and denoising.

5. The real-time identification method of ship impact load on bridge pier according to claim 4 is characterized in that: The natural frequency obtained by performing modal analysis on the finite element model established in step S4 is the same as the vibration frequency of the bridge pier information data obtained by the acquisition instrument.

6. The real-time identification method of ship impact load according to claim 5 is characterized by: In step S6, a force hammer is installed at the bridge pier, the force hammer is connected to the data collector, and the force hammer is started to hit the bridge pier. The data collector obtains the load time history of the force hammer and the acceleration response of the bridge pier, and compares the acceleration response of the bridge pier with the load time history of the force hammer to confirm the accuracy of the load identification method of steps S1 to S5.

7. The real-time identification method of ship impact load on bridge pier according to claim 6 is characterized by: The force measuring range of the selected hammer is 12.5T and the sensitivity is 4pc / N.

8. A health detection system based on the real-time identification method of ship-to-bridge pier load according to claim 7, characterized in that: It includes multiple acceleration sensors, a data storage system, a collector and an industrial router arranged at different positions of the bridge pier. The multiple acceleration sensors are connected to the collector through the industrial router, and a data storage system is established at the collector.

Images