A method and system for intelligent control of structural system conversion in support displacement construction

Abstract

This invention proposes an intelligent control method and system for structural system transformation during bridge support relocation construction, belonging to the field of bridge technology. It acquires sensor data from key locations of the bridge support relocation construction structure, filters the sensor data using an improved Kalman filter method, performs wavelet enhancement on the filtered results, matches the enhancement results with a preset structural system transformation control rule library, and performs intelligent control of structural system transformation based on the matching results. This invention improves the Kalman filter method by dynamically adjusting the observation noise covariance, adapting to the spike interference of sudden changes in electromagnetic interference intensity on sensor signals, and achieving dynamic electromagnetic interference suppression. In wavelet enhancement processing, the optimal wavelet decomposition level is automatically selected based on the signal length, ensuring denoising effect while avoiding over-decomposition. The selection and setting of soft and hard thresholds in the wavelet threshold denoising process balances the discontinuity and amplitude deviation of the denoised results.

Description

Technical Field

[0001] This invention belongs to the field of bridge technology, and in particular relates to an intelligent control method and system for structural system transformation during support relocation construction. Background Technology

[0002] The scaffolding relocation construction technology for steel-concrete composite beam bridges refers to the technique of connecting steel box girder segments into a single unit within one or more spans without the need for temporary scaffolding at the bolt joints during the splicing process. This technology can resolve the impact of temporary scaffolding on traffic flow when steel-concrete composite beam bridges cross major traffic arteries. In the specific construction process, temporary scaffolding is installed at the bolt joints of the side span steel beams that do not obstruct traffic, as well as in the central median strip of the major traffic artery. Then, multiple steel box girder segments are hoisted and bolted together. After completing structural construction such as the pouring of the concrete bridge deck, the temporary scaffolding is removed, completely transforming the structural system from a temporary support state to a permanent load-bearing system.

[0003] To ensure the safety of temporary scaffold dismantling, it is necessary to rely on wireless monitoring of the stress and / or displacement of key sections such as the mid-span section, L / 4 section, temporary scaffold support section, and pier top (wireless monitoring avoids complex wiring and meets the monitoring flexibility requirements of the dynamically changing construction environment of temporary scaffolds). However, dynamic electromagnetic interference generated by equipment or construction machinery such as frequency converters, walkie-talkies, cranes, and concrete pump trucks at the bridge construction site can easily contaminate the stress and / or displacement data of key sections monitored wirelessly, affecting the accuracy of stress and / or displacement data monitoring, which may in turn affect the judgment of the timing of temporary scaffold dismantling, or even cause serious consequences such as structural overturning. Summary of the Invention

[0004] To address the aforementioned problems in the existing technology, this invention proposes an intelligent control method and system for structural system transformation during support relocation construction. This method can at least partially alleviate the interference of electromagnetic interference at the bridge construction site on the monitoring of stress and / or displacement data of key sections, thereby improving the accuracy of stress and / or displacement data monitoring of key sections.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent control method for structural system transformation during bridge support relocation construction, comprising the following steps: S1: acquiring sensor data at key locations of the bridge support relocation construction structure; S2: filtering the sensor data based on an improved Kalman filter method to obtain the final improved Kalman filter result of the sensor data; S3: performing wavelet enhancement on the final improved Kalman filter result of the sensor data to obtain a processed result; S4: matching the processed result with a preset structural system transformation control rule library to obtain a matching result; and performing intelligent control of structural system transformation based on the matching result.

[0006] Furthermore, the key locations of the bridge support relocation construction structure in step S1 include the mid-span section, the support fulcrum, and the top of the pier.

[0007] Furthermore, the improved Kalman filtering method in step S2 specifically includes:

[0008] S21: Acquire the sensor data and construct an initial sensor data matrix; initialize a memory space with the same size as the initial sensor data matrix;

[0009] S22: Using the number of sensors as the maximum value of the outer loop sequence, process the data stream of each sensor in sequence; for each sensor, initialize the following variables: process noise covariance, initial observation noise covariance, forgetting factor, state estimate, estimation error covariance, state transition matrix, and observation matrix;

[0010] S23: Using the total amount of data monitored by each sensor as the maximum value of the inner loop sequence, calculate the observation residual for each sampled data; specifically, the calculation of the observation residual involves first calculating the product of the state estimation variable, the observation matrix variable, and the state transition matrix variable, and then subtracting the product from the initial matrix of sensor data to obtain the observation residual;

[0011] S24: Adjust the initial observation noise covariance based on the observation residuals and sliding window; calculate the Kalman filter gain based on the adjusted initial observation noise covariance; and obtain the improved Kalman filter result of the sensor data based on the Kalman filter gain.

[0012] Further, in step S24, the initial observation noise covariance is adjusted based on the observation residuals and the sliding window, specifically as follows:

[0013] S241: Initialize a sliding window matrix that matches the initial matrix of sensor data; then extract all data from the row where the sensor is located in the corresponding inner loop sequence from the sliding window matrix as the current window; then remove the first element of the current window and add the observation residual to the end; reassign the updated current window back to the row corresponding to the sliding window matrix to obtain the updated sliding window matrix;

[0014] S242: Calculate the variance of the updated sliding window matrix. When the variance is greater than m1 times the estimated error covariance, update the estimated error covariance to k1 times the original estimated error covariance. When the variance is less than m2 times the estimated error covariance, update the estimated error covariance to k2 times the original estimated error covariance.

[0015] S243: Calculate the product of the observation residual and its transpose as the first product; calculate the product of the state transition matrix variable, the estimation error covariance variable, and the transpose of the state transition matrix variable, add it to the process noise covariance variable to obtain the prediction covariance matrix, then calculate the product of the observation matrix, the prediction covariance matrix, and the transpose of the observation matrix to obtain the second product; then subtract the second product from the first product to obtain the adjustment amount;

[0016] S244: Calculate the difference between 1 and the forgetting factor, and then multiply the difference by the estimated error covariance updated in step S242 as the first adjustment component; calculate the product of the forgetting factor and the adjustment amount in step S243 as the second adjustment component; add the first adjustment component and the second adjustment component to obtain the adjusted initial observation noise covariance.

[0017] Further, in step S24, the Kalman filter gain is calculated based on the adjusted initial observation noise covariance. Specifically, the product of the prediction covariance matrix and the transpose of the observation matrix is ​​used as the dividend, and the sum of the second product and the adjusted initial observation noise covariance is used as the divisor to calculate the division result. The division result is then used as the Kalman filter gain. In step S24, the improved Kalman filter result of the sensor data is obtained based on the Kalman filter gain. Specifically, the product of the Kalman filter gain and the observation residual is calculated, and the product of the state transition matrix variable and the state estimation variable is calculated. The two products are then added together to obtain the improved Kalman filter result of the sensor data. The estimation error covariance is updated as the iteration variable for subsequent loops. After executing the inner and outer loops, the improved Kalman filter result of the sensor data is used as the final improved Kalman filter result of the sensor data.

[0018] Furthermore, in step S3, the final improved Kalman filter result of the sensor data is subjected to wavelet enhancement to obtain the processed result, specifically as follows:

[0019] S31: Obtain the signal length of the final improved Kalman filter result of the sensor data, denoted as N; calculate the logarithm of N to the base 2 and round down, and take the smaller value between the rounded result and the first preset decomposition layer threshold as the wavelet decomposition layer number;

[0020] S32: Calculate the signal standard deviation of the final improved Kalman filter result of the sensor data as the first standard deviation; calculate the standard deviation after performing a moving average on the final improved Kalman filter result of the sensor data as the second standard deviation; calculate the ratio of the first standard deviation to the second standard deviation and then subtract the absolute value of 1 to obtain the standard deviation change rate.

[0021] The result after the moving average processing is taken, and the absolute value of the difference between adjacent points of the result after the moving average processing is calculated, and then the average value is calculated to obtain the first average value; the absolute value of the difference between adjacent points of the final improved Kalman filter result of the sensor data is taken, and then the average value is calculated to obtain the second average value; the first average value is divided by the second average value to obtain the gradient change rate;

[0022] S33: The standard deviation change rate and the gradient change rate are weighted and summed. When the summation result is greater than the preset summation threshold, the wavelet decomposition level obtained in step S31 is used as an input to the wden function, and the threshold processing mode of the wden function is set to soft thresholding to denoise the final improved Kalman filter result of the standardized sensor data. When the summation result is less than or equal to the preset summation threshold, the wavelet decomposition level obtained in step S31 is used as an input to the wden function, and the threshold processing mode of the wden function is set to hard thresholding to denoise the final improved Kalman filter result of the standardized sensor data. The denoised signal is used as the processed result.

[0023] This invention also proposes an intelligent control system for structural system transformation during support relocation construction, used to execute the aforementioned intelligent control method for structural system transformation during support relocation construction. The system includes a data acquisition module, a data processing module, a matching module, and a result output module. The data acquisition module is connected to the data processing module, the data processing module is connected to the matching module, and the matching module is connected to the result output module. The data acquisition module acquires sensor data at key locations of the bridge support relocation construction structure and inputs it to the data processing module. The data processing module filters and enhances the input signal and inputs the filtered and enhanced signal to the matching module. The matching module matches the filtered and enhanced signal with a preset structural system transformation control rule library and inputs the matching result to the result output module. The result output module generates a structural system transformation control signal based on the matching result to control the transformation of the structural system during support relocation construction.

[0024] The beneficial technical effects of this invention compared with the prior art are as follows: (1) The Kalman filtering method is improved by dynamically adjusting the observation noise covariance, adapting to the spike interference of the sensor signal caused by the sudden change in electromagnetic interference intensity, and realizing dynamic electromagnetic interference suppression; (2) The optimal wavelet decomposition layer is automatically selected according to the signal length, ensuring the denoising effect while avoiding excessive decomposition; (3) The selection and setting of soft and hard thresholds in the wavelet threshold denoising process balances the discontinuity and amplitude deviation of the denoised result. Attached Figure Description

[0025] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0026] Figure 1 This is a flowchart of an intelligent control system method for structural system conversion during support relocation construction according to the present invention;

[0027] Figure 2 This is a simplified flowchart of the steps of the improved Kalman filtering method of the present invention;

[0028] Figure 3 This is a system block diagram of an intelligent control system for structural system conversion during support relocation construction according to the present invention. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] The concepts involved in this application will first be described with reference to the accompanying drawings. It should be noted that the following descriptions of various concepts are only for the purpose of making the content of this application easier to understand and do not constitute a limitation on the scope of protection of this application; furthermore, the embodiments and features in the embodiments of this application can be combined with each other unless otherwise specified. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0031] Refer to the instruction manual. Figure 1This invention proposes an intelligent control method for structural system transformation during bridge support relocation construction, comprising the following steps: S1: Acquiring sensor data at key locations of the bridge support relocation construction structure; the key locations of the bridge support relocation construction structure include core locations that can reflect the overall stress and displacement of the bridge, such as the mid-span section, support fulcrum, and pier top. S2: Filtering the sensor data based on an improved Kalman filter method to obtain the final improved Kalman filter result of the sensor data; S3: Performing wavelet enhancement on the final improved Kalman filter result of the sensor data to obtain a processed result; S4: Matching the processed result with a preset structural system transformation control rule library to obtain a matching result; and performing intelligent control of structural system transformation based on the matching result. It is understandable that the aforementioned pre-set structural system conversion control rule library can be designed using a three-level database structure of bridge type-component-section, embedding a decoupling algorithm to decouple the geometric constraints of the above three levels; the matching process based on the matching results includes a dual verification mechanism of literal feature similarity calculation and semantic feature similarity analysis, generates a unique identity code for the matching results through a hash function algorithm, and achieves accurate matching by combining custom conversion rules. When a match is successful, the structural system conversion program is automatically triggered, generating a structural system conversion control signal to control the conversion of the structural system during the support relocation construction.

[0032] Refer to the instruction manual. Figure 2 The improved Kalman filtering method in step S2 of the intelligent control method for structural system transformation during support relocation construction of the present invention specifically includes: S21: acquiring the sensor data and constructing an initial sensor data matrix; initializing a memory space with the same size as the initial sensor data matrix to ensure that the data organization format in memory is strictly consistent with the algorithm requirements.

[0033] S22: Using the number of sensors as the maximum value of the outer loop sequence, process the data stream of each sensor sequentially; for each sensor, initialize the following variables: process noise covariance, initial observation noise covariance, forgetting factor, state estimate, estimation error covariance, state transition matrix, and observation matrix; S23: Using the total amount of data monitored by each sensor as the maximum value of the inner loop sequence, calculate the observation residual for each sampled data; specifically, the calculation of the observation residual involves first calculating the product of the state estimate variable, the observation matrix variable, and the state transition matrix variable, and then subtracting the product from the initial sensor data matrix to obtain the observation residual; S24: Adjust the initial observation noise covariance based on the observation residual and the sliding window; and calculate the Kalman filter gain based on the adjusted initial observation noise covariance; obtain the improved Kalman filter result of the sensor data based on the Kalman filter gain.

[0034] In step S24 described above, adjusting the initial observation noise covariance based on the observation residuals and the sliding window specifically includes the following sub-steps:

[0035] S241: Initialize a sliding window matrix that matches the initial matrix of sensor data; then extract all data from the row where the sensor is located in the corresponding inner loop sequence from the sliding window matrix as the current window; then remove the first element of the current window and add the observation residual to the end; reassign the updated current window back to the row corresponding to the sliding window matrix to obtain the updated sliding window matrix;

[0036] S242: Calculate the variance of the updated sliding window matrix. When the variance is greater than m1 times the estimation error covariance, update the estimation error covariance to k1 times the original estimation error covariance; when the variance is less than m2 times the estimation error covariance, update the estimation error covariance to k2 times the original estimation error covariance. It can be understood that the value of k1 should be greater than 1, and the value of k2 should be less than 1. k1 is preferably 1.08, and k2 is preferably 0.95. The setting of k1 is used to reduce the confidence in the current measurement value and avoid the state estimation being interfered with by sudden and severe measurement noise, thereby maintaining the smoothness of the estimation. The setting of k2 is used to increase the confidence in the current measurement value and increase the weight of the current measurement value in the state update, so that the state estimation can track the real changes of the system more quickly.

[0037] S243: Calculate the product of the observation residual and its transpose as the first product; calculate the product of the state transition matrix variable, the estimation error covariance variable, and the transpose of the state transition matrix variable, and add it to the process noise covariance variable to obtain the prediction covariance matrix; then calculate the product of the observation matrix, the prediction covariance matrix, and the transpose of the observation matrix to obtain the second product; then subtract the second product from the first product to obtain the adjustment amount; this step, through the calculation of the adjustment amount, can measure the initial observation noise covariance in real time, which is better adapted to time-varying noise problems.

[0038] S244: Calculate the difference between 1 and the forgetting factor, and then multiply the difference by the estimated error covariance updated in step S242 as the first adjustment component; calculate the product of the forgetting factor and the adjustment amount in step S243 as the second adjustment component; add the first adjustment component and the second adjustment component to obtain the adjusted initial observation noise covariance.

[0039] In step S24, the Kalman filter gain is calculated based on the adjusted initial observation noise covariance. Specifically, the product of the prediction covariance matrix and the transpose of the observation matrix is ​​used as the dividend, and the sum of the second product and the adjusted initial observation noise covariance is used as the divisor to calculate the division result. The division result is then used as the Kalman filter gain. In step S24, the improved Kalman filter result of the sensor data is obtained based on the Kalman filter gain. Specifically, the product of the Kalman filter gain and the observation residual is calculated, and the product of the state transition matrix variable and the state estimation variable is calculated. The two products are then added together to obtain the improved Kalman filter result of the sensor data. The estimation error covariance is updated as the iteration variable for subsequent loops. After executing the inner and outer loops, the improved Kalman filter result of the sensor data is used as the final improved Kalman filter result of the sensor data.

[0040] In step S3 of this invention, wavelet enhancement is performed on the final improved Kalman filter result of the sensor data to obtain the processed result, specifically as follows:

[0041] S31: Obtain the signal length of the final improved Kalman filter result of the sensor data, denoted as N; calculate the logarithm of N to the base 2 and round down, and take the smaller of the rounded result and the first preset decomposition layer threshold as the wavelet decomposition layer number; the present invention automatically adjusts the granularity of wavelet decomposition according to the characteristics of sensor data (signal length can reflect the characteristics of sensor data), ensuring that the main features of the signal can be captured under different noise environments and operating conditions, while ensuring the efficiency of data processing.

[0042] S32: Calculate the signal standard deviation of the final improved Kalman filter result of the sensor data as the first standard deviation; calculate the standard deviation after performing a moving average on the final improved Kalman filter result of the sensor data as the second standard deviation; calculate the ratio of the first standard deviation to the second standard deviation and then subtract the absolute value of 1 to obtain the standard deviation change rate.

[0043] The result after the moving average processing is taken, and the absolute value of the difference between adjacent points of the result after the moving average processing is calculated, and then the average value is calculated to obtain the first average value; the absolute value of the difference between adjacent points of the final improved Kalman filter result of the sensor data is taken, and then the average value is calculated to obtain the second average value; the first average value is divided by the second average value to obtain the gradient change rate;

[0044] S33: The standard deviation change rate and the gradient change rate are weighted and summed. When the summation result is greater than a preset summation threshold, the wavelet decomposition level obtained in step S31 is used as an input to the wden function, and the threshold processing mode of the wden function is set to soft thresholding to denoise the final improved Kalman filter result of the standardized sensor data. When the summation result is less than or equal to the preset summation threshold, the wavelet decomposition level obtained in step S31 is used as an input to the wden function, and the threshold processing mode of the wden function is set to hard thresholding to denoise the final improved Kalman filter result of the standardized sensor data. The denoised signal is used as the processed result. The above selection of soft and hard thresholds in the wavelet thresholding denoising process can balance the discontinuity and amplitude deviation of the denoised result.

[0045] Refer to the instruction manual. Figure 3 This invention also proposes an intelligent control system for structural system transformation during scaffold relocation construction. This control system can be constructed using a centralized control platform that manages multiple devices collaboratively. The intelligent control system for structural system transformation during scaffold relocation construction executes the aforementioned intelligent control method for structural system transformation during scaffold relocation construction. It includes a data acquisition module, a data processing module, a matching module, and a result output module. The data acquisition module is connected to the data processing module, the data processing module is connected to the matching module, and the matching module is connected to the result output module. The data acquisition module acquires sensor data at key locations of the bridge scaffold relocation construction structure and inputs it to the data processing module. The module can employ a general-purpose AI chip (e.g., CPU / GPU), a dedicated AI chip (e.g., ASIC), or an FPGA architecture to filter and enhance the input signal. The filtered and enhanced signal is then input into the matching module. The matching module matches the filtered and enhanced signal with a preset structural system conversion control rule library and inputs the matching result into the result output module. The result output module generates a structural system conversion control signal based on the matching result (the aforementioned structural system conversion control signal includes at least a set of bridge temporary support dismantling instructions (including dismantling sequence instructions, dismantling timing instructions), bridge deck paving instructions, and railing installation instructions) to control the conversion of the structural system during support relocation construction.

[0046] The embodiments and / or implementation methods described above are merely preferred embodiments and / or implementation methods for implementing the technology of the present invention, and are not intended to limit the implementation methods of the technology of the present invention in any way. Any person skilled in the art can make some modifications or alterations to other equivalent embodiments without departing from the scope of the technical means disclosed in the content of the present invention, but they should still be regarded as the technology or embodiments that are substantially the same as the present invention.

[0047] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. The above descriptions are only preferred embodiments of this application. It should be noted that due to the limitations of written expression, while there are objectively infinite specific structures, those skilled in the art can make several improvements, modifications, or changes without departing from the principles of this application, and can also combine the above technical features in an appropriate manner. These improvements, modifications, changes, or combinations, or the direct application of the inventive concept and technical solution to other situations without modification, should all be considered within the scope of protection of this application.

Claims

1. A method for intelligent control of structural system conversion in a support displacement construction, characterized in that, The method comprises the following steps: S1: obtaining sensor data at a key position of a bridge support displacement construction structure; S2: filtering the sensor data based on an improved Kalman filtering method to obtain a final improved Kalman filtering result of the sensor data; S3: performing wavelet enhancement on the final improved Kalman filtering result of the sensor data to obtain a processed result; S4: matching the processed result with a preset structure system conversion control rule library to obtain a matching result; performing intelligent control on structure system conversion according to the matching result; The S2 specifically comprises: S21: obtaining the sensor data and constructing a sensor data initial matrix; initializing a memory space with a size same as that of the sensor data initial matrix; S22: taking the number of sensors as an outer loop sequence maximum value, and sequentially processing data streams of each sensor; for each sensor, initializing the following variables: process noise covariance, initial observation noise covariance, forgetting factor, state estimation, estimation error covariance, state transition matrix and observation matrix; S23: taking the total amount of monitoring data of each sensor as an inner loop sequence maximum value, and calculating observation residuals for each sampling data; the calculation of the observation residuals specifically comprises: firstly calculating the product of the state estimation variable, the observation matrix variable and the state transition matrix variable, and then subtracting the product from the sensor data initial matrix to obtain the observation residuals; S24: adjusting the initial observation noise covariance according to the observation residuals and a sliding window, calculating a Kalman filtering gain based on the adjusted initial observation noise covariance, and obtaining an improved Kalman filtering result of the sensor data based on the Kalman filtering gain; In the S3, the final improved Kalman filtering result of the sensor data is subjected to wavelet enhancement to obtain a processed result, and the wavelet enhancement specifically comprises: S31: obtaining the signal length of the final improved Kalman filtering result of the sensor data, denoted as N; calculating the logarithm of N with 2 as the base and taking the down rounding result as the wavelet decomposition layer number; S32: calculating the signal standard deviation of the final improved Kalman filtering result of the sensor data as a first standard deviation; calculating the standard deviation of the final improved Kalman filtering result of the sensor data after sliding average processing as a second standard deviation; calculating the ratio of the first standard deviation to the second standard deviation and then subtracting the absolute value of 1 to obtain a standard deviation change rate; taking the result after the sliding average processing and calculating the average value of the absolute values of the differences between adjacent points of the result after the sliding average processing to obtain a first average value; taking the average value of the absolute values of the differences between adjacent points of the final improved Kalman filtering result of the sensor data to obtain a second average value; dividing the first average value by the second average value to obtain a gradient change rate; S33: the standard deviation change rate and the gradient change rate are weighted and summed, when the sum result is greater than a preset sum threshold value, the wavelet decomposition layer number obtained in the step S31 is taken as an input quantity of a wden function, a threshold value processing mode of the wden function is set as a soft threshold value, and a final improved Kalman filtering result of the normalized sensor data is denoised; when the sum result is less than or equal to the preset sum threshold value, the wavelet decomposition layer number obtained in the step S31 is taken as an input quantity of the wden function, and a threshold value processing mode of the wden function is set as a hard threshold value, and the final improved Kalman filtering result of the normalized sensor data is denoised; and the denoised signal is taken as a processed result.

2. The intelligent control method for structural system conversion in the construction of a support displacement according to claim 1, characterized in that, The key positions of the bridge support displacement construction structure in the S1 include a midspan section, a support fulcrum and a pier top.

3. The intelligent control method for structural system conversion in the construction of a support displacement according to claim 1, characterized in that, In the S24, the initial observation noise covariance is adjusted according to the observation residual and the sliding window, and specifically: S241: a sliding window matrix matched with an initial matrix of sensor data is initialized, and then all data corresponding to the rows of sensors in the inner loop sequence are extracted from the sliding window matrix as a current window; The first element of the current window is removed, and the observation residual is added at the end; the updated current window is re-assigned to the corresponding row of the sliding window matrix to obtain an updated sliding window matrix; S242: the variance of the updated sliding window matrix is calculated, when the variance is greater than m1 times the estimated error covariance, the estimated error covariance is updated to k1 times the original estimated error covariance; when the variance is less than m2 times the estimated error covariance, the estimated error covariance is updated to k2 times the original estimated error covariance; wherein the value of k1 is greater than 1, and the value of k2 is less than 1; S243: the product of the observation residual and its transpose is calculated as a first product quantity; The product of the state transition matrix variable, the estimated error covariance variable and the transpose of the state transition matrix variable is calculated, and the product is added to the process noise covariance variable to obtain a prediction covariance matrix, then the product of the observation matrix, the prediction covariance matrix and the transpose of the observation matrix is calculated to obtain a second product quantity; and the first product quantity is subtracted from the second product quantity to obtain an adjustment quantity; S244: the difference between 1 and the forgetting factor is calculated, and the difference is multiplied by the estimated error covariance updated in the step S242 to obtain a first adjustment component; The product of the forgetting factor and the adjustment quantity in the step S243 is calculated as a second adjustment component; and the first adjustment component and the second adjustment component are added to obtain an adjusted initial observation noise covariance.

4. The intelligent control method for structural system conversion in the construction of a support displacement according to claim 3, characterized in that, The step S24 is to calculate the Kalman filter gain based on the adjusted initial observation noise covariance, specifically, to calculate a division result by taking the product of the prediction covariance matrix and the transpose of the observation matrix as the dividend and the sum of the second product and the adjusted initial observation noise covariance as the divisor; and to take the division result as the Kalman filter gain; the step S24 is to obtain the improved Kalman filter result of the sensor data based on the Kalman filter gain, specifically, to calculate the product of the Kalman filter gain and the observation residual, to calculate the product of the state transition matrix variable and the state estimation variable, and to add the two products to obtain the improved Kalman filter result of the sensor data; The estimated error covariance is updated as an iteration variable for the subsequent loop; after the inner loop and the outer loop are executed, the improved Kalman filter result of the sensor data is taken as the final improved Kalman filter result of the sensor data.

5. An intelligent control system for structural system conversion in support displacement construction, used to perform an intelligent control method for structural system conversion in support displacement construction according to any one of claims 1-4, characterized in that, The data acquisition module, the data processing module, the matching module and the result output module are included, the data acquisition module is connected with the data processing module, the data processing module is connected with the matching module, and the matching module is connected with the result output module; The data acquisition module is used for acquiring sensor data at key positions of a bridge support displacement construction structure and inputting the sensor data to the data processing module, the data processing module is used for filtering and enhancing an input signal and inputting the filtered and enhanced signal to the matching module, the matching module is used for matching the filtered and enhanced signal with a preset structure system conversion control rule library and inputting a matching result to the result output module, and the result output module is used for generating a structure system conversion control signal according to the matching result to control conversion of a structure system in support displacement construction.

Images