CN-122020828-A - Correction method and device for ship lock structure simulation parameters and electronic equipment

CN122020828ACN 122020828 ACN122020828 ACN 122020828ACN-122020828-A

Abstract

The application provides a correction method, a correction device and electronic equipment for ship lock structure simulation parameters, which comprise the steps of obtaining historical displacement actual measurement data of a ship lock structure to form standardized actual measurement data, determining parameters to be corrected according to a ship lock structure simulation model, carrying out global search on the parameters to be corrected by utilizing a covariance matrix self-adaptive evolution algorithm, obtaining distribution characteristic parameters which enable a displacement response predicted value output by the model to be matched with the standardized actual measurement data to be optimal through self-adaptive learning parameter correlation and iterative optimization so as to determine initial parameter distribution of the parameters to be corrected, combining the parameters to be corrected and the displacement response predicted value to construct an augmented state space model, carrying out recursive assimilation on augmented state variables in a space model by taking real-time ship lock displacement data as observation values and adopting a Kalman filtering algorithm to realize updating correction of the parameters to be corrected. The application can assimilate real-time lock displacement data and dynamically update and correct parameters to be corrected in the lock structure model.

Inventors

ZHAO HAORAN
LONG TINGYUAN
WU JIANMING

Assignees

浙江远算科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260415

Claims (10)

1. A method for correcting simulation parameters of a ship lock structure, the method comprising: acquiring historical displacement measured data of a ship lock structure, and preprocessing to form standardized measured data; determining parameters to be corrected according to the ship lock structure simulation model; Performing global search on the parameters to be corrected by using a covariance matrix self-adaptive evolution algorithm, and obtaining a distribution characteristic parameter which enables a displacement response predicted value output by the ship lock structure simulation model to be matched with the standardized measured data optimally through correlation among self-adaptive learning parameters and iterative optimization so as to determine initial parameter distribution of the parameters to be corrected; constructing an augmentation state space model based on the parameter to be corrected, the displacement response predicted value and the initial parameter distribution; and taking the real-time ship lock displacement data as an observation value, and adopting a Kalman filtering algorithm to recursively assimilate the augmentation state variables in the augmentation state space model so as to update and correct the parameters to be corrected.
2. The method of claim 1, wherein the step of obtaining historical displacement measured data of the lock structure and forming standardized measured data by preprocessing comprises: generating standardized measured data meeting the subsequent real-time assimilation requirements through the following processing modes: performing preliminary screening on the history displacement actual measurement data, and eliminating repeated records and invalid records; correcting the missing value in the history displacement actual measurement data by adopting an adjacent mean filling method according to the distribution characteristics; Aiming at repeated observation values of the same monitoring point in the same day in the history displacement actual measurement data, calculating an arithmetic average value as a unique characterization value of the day; and resampling and time-aligning the horizontal displacement data and the vertical displacement data in the history displacement measured data based on a preset assimilation step length to ensure that different types of monitoring variables are consistent on a time scale.
3. The method according to claim 1, wherein the step of performing global search on the parameter to be corrected by using a covariance matrix adaptive evolution algorithm, obtaining a distribution characteristic parameter that optimizes a matching degree between a displacement response predicted value output by the ship lock structure simulation model and the standardized measured data by means of correlation between adaptive learning parameters and iterative optimization, so as to determine an initial parameter distribution of the parameter to be corrected, includes: setting an initial state of a covariance matrix self-adaptive evolution algorithm to form multidimensional normal distribution, wherein the initial state comprises an initial mean vector, an initial step length and an initial covariance matrix of parameters to be corrected, which are set according to a ship lock design experience value; for each iterative process, the following sampling, evolving, screening, updating steps are performed: Sampling from the multidimensional normal distribution to generate a sample set consisting of parameter vectors of a specified number of individuals, wherein each parameter vector simultaneously contains candidate value combinations of all parameters to be corrected; Substituting each sample into the ship lock structure simulation model in sequence for evolution to obtain a corresponding physical field response predicted value; calculating based on the physical field response predicted value and the corresponding value in the standardized measured data to obtain a target function value corresponding to each sample; sorting according to the objective function value corresponding to each sample, and selecting a specified number of high-quality samples; updating the distribution characteristic parameters of the multidimensional normal distribution by using the high-quality samples, wherein the distribution characteristic parameters comprise a mean value vector, a step length and a covariance matrix; And constructing and obtaining initial parameter distribution of the parameter to be corrected based on the distribution characteristic parameters of the final multidimensional normal distribution until convergence conditions are met or the maximum iteration times are reached.
4. The method of claim 1, wherein the step of constructing an augmented state space model based on the parameters to be corrected, the displacement response predictors, and the initial parameter distribution comprises: constructing an augmented state vector by taking the parameter to be corrected and the displacement response predicted value as internal state variables; constructing an initial state set comprising a plurality of augmented state vectors by adopting a Monte Carlo sampling method according to the initial parameter distribution; An augmented state space model is constructed based on the augmented state vector, the nonlinear state evolution function, and the process noise for the initial state set.
5. The method of claim 4, wherein the step of recursively assimilating the augmented state variables in the augmented state space model using the real-time lock displacement data as the observations and using a kalman filter algorithm to effect updated corrections to the parameters to be corrected comprises: Taking each vector in the current initial state set as a posterior augmentation state vector at the last moment, and executing the following steps of prediction, statistics, gain calculation and assimilation: based on the augmentation state space model, vector prediction is carried out on the posterior augmentation state vector at the previous moment, statistics is carried out on a prediction result, and a state prior deviation matrix and a displacement prediction deviation matrix at the current moment are determined; Calculating the state-displacement cross covariance of the current moment according to the state priori deviation matrix and the displacement prediction deviation matrix of the current moment; calculating a Kalman gain matrix at the current moment according to the state-displacement cross covariance, the displacement prediction auto-covariance and a preset measurement error covariance matrix at the current moment; Extracting a real-time value of the current moment corresponding to the vector from the real-time lock displacement data; Based on the Kalman gain matrix at the current moment, correcting the displacement prediction deviation of the real-time value at the current moment in real time to realize the updating correction of the parameter to be corrected; And re-using the updated and corrected posterior augmentation state vector as the posterior augmentation state vector at the previous moment, and continuously executing the steps of prediction, statistics, gain calculation and assimilation.
6. The method of claim 5, wherein the step of vector predicting the posterior augmented state vector at the previous time and counting the prediction results based on the augmented state space model, and determining the state prior bias matrix and the displacement prediction bias matrix at the current time comprises: Predicting the posterior augmentation state vector at the previous moment according to the augmentation state space model aiming at each vector in the initial state set to determine the prior augmentation state vector at the current moment; Calculating an average value based on prior augmented state vectors at the current time corresponding to the plurality of vectors in the initial state set respectively, and determining an augmented state prediction average value; Calculating a state prior deviation matrix at the current moment according to prior augmented state vectors and the augmented state prediction average value at the current moment corresponding to the vectors in the initial state set respectively, and calculating a displacement prediction deviation matrix at the current moment according to displacement prediction values and the displacement prediction average value at the current moment corresponding to the vectors in the initial state set respectively.
7. The method according to claim 5, wherein the step of correcting the displacement prediction bias of the real-time value at the current time in real time based on the kalman gain matrix at the current time to realize the update correction of the parameter to be corrected comprises: calculating a difference value of the real-time value and a displacement predicted value corresponding to the vector; Solving the product of the Kalman gain matrix at the current moment and the difference value; and obtaining the sum of the product and the prior augmented state vector at the current moment to obtain the posterior augmented state vector at the current moment.
8. A device for correcting a simulation parameter of a ship lock structure, the device comprising: The data preprocessing module is used for acquiring the history displacement actual measurement data of the ship lock structure and forming standardized actual measurement data through preprocessing; the parameter determining module is used for determining parameters to be corrected according to the ship lock structure simulation model; The evolution iteration module is used for carrying out global search on the parameter to be corrected by utilizing a covariance matrix self-adaptive evolution algorithm, and obtaining a distribution characteristic parameter which enables a displacement response predicted value output by the ship lock structure simulation model to be matched with the standardized measured data optimally through the correlation among self-adaptive learning parameters and iterative optimization so as to determine the initial parameter distribution of the parameter to be corrected; the model construction module is used for constructing an augmentation state space model based on the parameter to be corrected, the displacement response predicted value and the initial parameter distribution; And the recursion assimilation module is used for recursively assimilating the augmentation state variables in the augmentation state space model by taking the real-time lock displacement data as an observation value and adopting a Kalman filtering algorithm to realize the updating correction of the parameter to be corrected.
9. An electronic device comprising a processor and a memory, the memory storing computer-executable instructions executable by the processor, the processor executing the computer-executable instructions to implement the method of any one of claims 1 to 7.
10. A computer readable storage medium storing computer executable instructions which, when invoked and executed by a processor, cause the processor to implement the method of any one of claims 1 to 7.

Description

Correction method and device for ship lock structure simulation parameters and electronic equipment Technical Field The application relates to the technical field of analysis of safety and operation stability of a ship lock structure, in particular to a method and a device for correcting simulation parameters of the ship lock structure and electronic equipment. Background In the ship lock engineering, a system formed by a lock chamber, a lock bottom plate and a foundation is in complex environments such as water pressure, seepage pressure, temperature change and the like for a long time, and stress distribution and mechanical parameters (such as elastic modulus) in the structure can be changed continuously along with the increase of the running time. In order to ensure the anti-skid, anti-floating and overall structural safety of the ship lock under the complex working condition, it is important to acquire and correct the dynamic parameters in real time. However, the current parameter correction method still has some limitations in practical application, firstly, the existing parameter correction process is often discrete or manually triggered, and is often adjusted once in a specific time period, so that synchronous update of parameters and monitoring data cannot be realized. Secondly, most of these methods focus on local parameter correction, and lack a real-time feedback mechanism capable of effectively converting measured data into structural overall parameter update. The update of the parameters is often delayed from the real evolution process of the structure, the rule of continuous change of the parameters along with the running time is difficult to reflect, and the timeliness and the accuracy of safety monitoring are limited. Therefore, the research and development of the method capable of assimilating the monitoring data in real time and dynamically updating the structural parameters is of great significance for realizing accurate assessment of the safety state of the ship lock. Disclosure of Invention The application aims to provide a correction method and device for a ship lock structure simulation parameter and electronic equipment, which can assimilate real-time ship lock displacement data and dynamically update and correct parameters to be corrected in a ship lock structure model. The application provides a correction method of ship lock structure simulation parameters, which comprises the steps of obtaining historical displacement actual measurement data of a ship lock structure, forming standardized actual measurement data through preprocessing, determining parameters to be corrected according to a ship lock structure simulation model, carrying out global search on the parameters to be corrected by utilizing a covariance matrix self-adaptive evolution algorithm, obtaining distribution characteristic parameters which enable a displacement response predicted value output by the ship lock structure simulation model to be matched with the standardized actual measurement data optimally through self-adaptive learning of the correlation among the parameters so as to determine initial parameter distribution of the parameters to be corrected, constructing an augmented state space model based on the parameters to be corrected, the displacement response predicted value and the initial parameter distribution, recursively assimilating the augmented state variables in the augmented state space model by adopting a Kalman filtering algorithm, and achieving update correction of the parameters to be corrected. The method comprises the steps of obtaining historical displacement measured data of a ship lock structure, preprocessing the historical displacement measured data to form standardized measured data, and generating the standardized measured data meeting the subsequent real-time assimilation requirements through the following processing modes of carrying out primary screening on the historical displacement measured data, eliminating repeated records and invalid records, correcting missing values in the historical displacement measured data according to a distribution characteristic by adopting an adjacent mean filling method, calculating an arithmetic average value as a daily unique characteristic value according to repeated observed values of the same monitoring point in the same day in the historical displacement measured data, and carrying out resampling and time alignment on horizontal displacement data and vertical displacement data in the historical displacement measured data based on a preset assimilation step length to ensure that different types of monitoring variables are consistent in time scale. The method comprises the steps of setting an initial state of a covariance matrix adaptive evolution algorithm to form multi-dimensional normal distribution, wherein the initial state comprises an initial mean value vector, an initial step length and an initial covariance matrix of the parameters to b