Search

CN-122022076-A - Multi-factor coupling corrosion prediction method for municipal water supply pipeline in sea area

CN122022076ACN 122022076 ACN122022076 ACN 122022076ACN-122022076-A

Abstract

The invention provides a multi-factor coupling corrosion prediction method for municipal water supply pipelines in a sea area, which belongs to the technical field of pipeline fault prediction and health management, and comprises the steps of uniformly mapping multi-source heterogeneous information such as pipeline bodies, operation conditions, environments, loads and the like to a space-time grid unit formed by a space grid unit and a daily statistics period, using the space-time grid as a whole-flow unique minimum calculation unit, finishing differential pretreatment and space-time alignment of multi-source data, quantifying coupling weight parameters of a four-core corrosion system in the sea area, further constructing a mechanism constraint and data-driven fused double-layer corrosion prediction model, combining a double-period layering correction mechanism to realize iterative optimization of model parameters, finally finishing pipeline corrosion grade judgment, residual life quantification calculation and graded operation and maintenance result output, and converting the corrosion prediction of the municipal water supply pipelines in the sea area from scattered and single-dimensional pipeline section grade rough calculation to grid grade full-flow standardized calculation under a unified space-time reference.

Inventors

  • CHEN XIAOHUA
  • MAN CHENG
  • CUI HONGZHI

Assignees

  • 中国海洋大学

Dates

Publication Date
20260512
Application Date
20260413

Claims (10)

  1. 1. A multi-factor coupling corrosion prediction method for municipal water supply pipelines in a sea area is characterized by comprising the following steps: S1, taking a space-time grid as a minimum calculation unit, collecting attribute parameters of a pipeline body, running parameters in a pipe, parameters of soil outside the pipe and atmospheric environment and traffic load parameters based on sensors laid on the grid, and preprocessing; S2, based on the S1 data and the metal corrosion electrochemical mechanism and municipal engineering design parameters, carrying out weight calculation of salt fog-chloride ion chemical coupling, temperature and humidity-pH value coupling, water pressure fluctuation-material aging coupling, traffic load-buried load transfer-rainfall coupling to obtain four types of dimensionless coupling weight parameters; s3, inputting the coupling weight parameters obtained in the S2 and pipeline materials into a constructed universal coupling corrosion mechanism model for covering chemical corrosion and mechanical damage, and outputting a dimensionless basic corrosion rate; Meanwhile, the coupling weight parameter obtained in the step S2 is input into a machine learning prediction model after training is completed by combining the accumulated operation days of the pipeline, and the single grid daily average prediction corrosion rate is output; And S4, weighting and fusing the dimensionless basic corrosion rate and the daily average predicted corrosion rate of the single grid, calculating to obtain the daily single grid final corrosion rate, and mapping the final corrosion rate into a corresponding corrosion grade by combining a preset corrosion rate threshold.
  2. 2. The method for predicting multi-factor coupling corrosion of municipal water supply pipelines in a sea-free area according to claim 1, further comprising a model iteration correction process based on measured data, wherein the model iteration correction process comprises the steps of collecting pipeline field corrosion true value data to construct a check set, calculating errors of a model predicted value and a field measured value based on a double-period layered correction mechanism of daily small correction and monthly large correction, adopting a mean square error MSE as an index, adopting an error back propagation algorithm to adaptively adjust corresponding coupling weight parameters, model coefficients and super parameters, only adjusting global coupling corrosion mechanism model related parameters daily, adapting single-day working condition fluctuation, optimizing a machine learning prediction model driving global parameters, model fusion weights and cross-system interactive coupling coefficients in a global coupling corrosion mechanism model by means of monthly large correction, and correcting long-term prediction trends.
  3. 3. The method for predicting multi-factor coupling corrosion of municipal water supply pipelines in a sea-free area according to claim 1, further comprising quantification of residual life assessment of the pipelines, specifically: Determining three types of thresholds according to industry standards, including a corrosion rate safety threshold for judging whether the pipeline is in a high-risk corrosion state or not, a critical wall thickness for calculating the residual life, an anti-corrosion layer failure judgment standard for judging whether the anti-corrosion layer loses the protection capability or not; converting the solar corrosion rate into the annual corrosion rate according to the formula Calculating the leap year according to 366 days, calculating the safe operation period again according to the formula Calculating, wherein L is the safe operation age; An initial wall thickness for the pipe; Is critical wall thickness; Is the daily corrosion rate; Is the annual corrosion rate.
  4. 4. The multi-factor coupling corrosion prediction method for municipal water supply pipelines in a sea-free area according to claim 1, wherein the pipeline body attribute parameters comprise pipeline materials, pipeline initial wall thickness and anticorrosive layer types, the pipeline internal operation parameters comprise pipeline burying depth, operation water pressure fluctuation amplitude and pipeline accumulated operation days, the pipeline external soil and atmospheric environment parameters comprise sensor point position coordinates, soil chloride ion concentration, soil pH value, atmospheric salt spray sedimentation flux and environmental temperature, and the traffic load parameters comprise road network-pipeline topology data and road section daily traffic flow.
  5. 5. The multi-factor coupling corrosion prediction method for municipal water supply pipelines in a sea-free area of claim 4, wherein the space grid adopts a CGCS2000 national geodetic coordinate system, continuous and uniform rectangular grid subdivision is carried out on a research area by 5-20m side length, grid side length is adjusted according to the density of pipeline sections in the research area, 5-10m grids are adopted in a pipeline section dense area, 10-20m grids are adopted in a pipeline section sparse area, and unique space numbers and plane coordinate ranges are assigned to each grid; the static attribute data is directly assigned according to space attribution, the time sequence dynamic data is projected to a unit grid according to an influence range, wherein the number of discrete measurement points is used for realizing the global coverage of a research area by adopting Kriging interpolation, and an arithmetic average value is taken as a final assigned value aiming at the same type of data of the same grid.
  6. 6. The method for predicting multi-factor coupling corrosion of municipal water supply pipelines in a sea-free area according to claim 2, wherein in S2, a salt spray-chloride ion coupling weight calculation formula is as follows: ; In the middle of The salt fog-chloride ion coupling weight is adopted; The sedimentation flux of the atmospheric salt mist; Taking alpha and beta as regional working condition coupling coefficients, and max as the maximum value of a research region; The temperature and humidity-pH value coupling weight calculation formula is as follows: ; In the middle of The temperature and humidity-pH value coupling weight is adopted, T is the temperature, pH is the pH value, E a is the activation energy of metal corrosion reaction, R is the universal gas constant, gamma is the pH value influence index, all the indexes are fixed, and max is the maximum value of a research area; The water pressure fluctuation-material aging coupling weight calculation formula is as follows: ; In the middle of The weight is the water pressure fluctuation-material aging coupling weight; For operating the water pressure fluctuation amplitude, t d is the accumulated operating days of the pipeline, and the stress coefficient Max is the maximum value of the research area; the traffic load-buried load transfer-rainfall coupling weight calculation formula is as follows ; In the middle of N is the daily traffic flow of the road section; the buried depth influence coefficient is determined by the collected pipeline buried depth; and the rainfall correction coefficient on the same day is determined by grading the accumulated rainfall R on the same day, and max is the maximum value of the research area.
  7. 7. The method for predicting multi-factor coupling corrosion of municipal water supply pipelines in a sea-free area according to claim 6, wherein the model of the global coupling corrosion mechanism is as follows: ; In the middle of Is the dimensionless base corrosion rate; correcting the coefficient for global cross coupling; The cross-system interaction coupling coefficient is used for representing the synergistic acceleration corrosion amplification effect of chemical coupling and mechanical coupling; The basic corrosion resistance coefficient of the pipeline material is determined by the collected pipeline material, and is a fixed value, and the spheroidal graphite cast iron =1.0, Carbon steel =1.2, PE tube =0.1, Stainless steel =0.3; ~ And (5) calculating the coupling weight parameters.
  8. 8. The method for predicting multi-factor coupling corrosion of municipal water supply pipelines in a sea-free area according to claim 1, wherein the machine learning prediction model adopts a random forest regression model, takes a constructed space-time calculation unit as a sample basis, and inputs parameters of accumulated operation days and coupling weights of the pipelines ~ Outputting single grid daily average predicted corrosion rate The initial super parameters of the model are set to 100 decision trees, the maximum depth is 10 layers, and the minimum leaf node samples are 5.
  9. 9. The method for predicting multi-factor coupling corrosion of municipal water supply pipelines in a sea-free area according to claim 2, wherein the final corrosion rate in S4 is calculated by the following formula: = ; In the formula, The final dimensionless corrosion rate is a single day single grid; For the model weights of the global coupling mechanism, The initial values of the weights are 0.6 and 0.4 for the weight of the machine learning model, and the sum of the weights is 1; according to GB/T19285-2026 standard and combining historical experience data Mapping to I-V grade, wherein I grade is weak corrosion, II grade is weak corrosion, III grade is medium corrosion, IV grade is strong corrosion, V grade is strong corrosion, marking the grid which is more than or equal to IV grade as high risk corrosion point position, and mapping synchronously according to corrosion grade to output four states of complete corrosion layer, slight damage, moderate damage and failure.
  10. 10. The method for predicting multi-factor coupling corrosion of municipal water supply pipelines in a sea-free area according to claim 7, wherein the daily small correction takes a single day as a correction period, and based on the current day error and the actually measured constraint data, only basic parameters and coupling weights of a global coupling mechanism model are adjusted, and the following operations are specifically executed: The basic parameter reverse correction, namely, according to MSE error on the same day, adopting a gradient-descending error reverse propagation algorithm to adjust a cross coupling correction coefficient eta, regional working condition coupling coefficients alpha and beta, a stress coefficient sigma 0 and an aging coefficient mu of a mechanism model; Corrosion depth constraint correction according to actual measurement of corrosion depth on the same day Calculating the theoretical corrosion depth In the following Accumulating running days for pipeline, if the relative deviation Then, performing deviation penalty correction on eta and mu; The method comprises the steps of switching a state mode of an anticorrosive layer, namely, comparing a judging state of an insulating resistance R of the anticorrosive layer according to actual measurement on the same day with a special failure threshold value determined based on the type of the anticorrosive layer of the current pipe section, calculating by adopting conventional coupling weight, switching to an accelerated corrosion mode when the actual measurement value is equal to or greater than the corresponding threshold value, and coupling the salt fog and chloride ions to the weight Coupling weight of temperature, humidity and pH value Multiplying by a factor of 1.2; the mechanical coupling weight range constraint is that according to the deformation and vibration actual measurement value of the pipeline on the same day, if the pipeline operation safety control limit value is exceeded, the water pressure fluctuation is coupled with the weight Coupling weight for traffic vibration The upper normalization limit of (2) is relaxed from 1.0 to 1.2, and if the upper normalization limit is within the allowable range, the upper normalization limit is kept unchanged from 1.0.

Description

Multi-factor coupling corrosion prediction method for municipal water supply pipeline in sea area Technical Field The invention belongs to the technical field of pipeline fault prediction and health management, and particularly relates to a multi-factor coupling corrosion prediction method for municipal water supply pipelines in a sea-free area. Background The municipal water supply pipeline in the sea area is in a special geological environment with high salt fog, high soil chloride ions and unbalanced acid and alkali for a long time, is simultaneously influenced by the superposition of multiple factors of overhead road traffic load, fluctuation of internal water pressure of the pipeline and aging of self materials, has a corrosion failure rate higher than that in an inland common area, is easy to cause the problems of pipeline wall corrosion perforation, corrosion-resistant layer stripping cracking, pipe network leakage damage and the like, and threatens municipal water supply safety and pipe network stable operation. In order to realize accurate prejudgment of corrosion state of the municipal water supply pipeline in the sea and guarantee long-acting stable operation and maintenance of a pipe network, a targeted corrosion prediction technology is needed urgently, but the existing related technology for pipeline corrosion prediction cannot meet the actual engineering requirements of pipeline refinement and accurate corrosion prevention early warning under the complicated working condition in the sea, and the specific defects are as follows: (1) The existing corrosion prediction technology adopts single environmental factor or single structural parameter to carry out one-sided analysis, and does not construct a global coupling model aiming at the special corrosion working condition of the sea-facing area, while partial technology is incorporated into environmental or mechanical parameters, the cooperative coupling effect among multiple factors is not quantized, and four core influencing factors of salt fog-chloride ion coupling, temperature and humidity-pH value coupling, water pressure fluctuation-material aging coupling, traffic load-buried load transmission-rainfall coupling cannot be synchronously incorporated, integrated modeling calculation is carried out, and the coupling mechanism analysis is incomplete; (2) In the prior art, accurate matching of road network traffic flow data and pipeline space data is generally lost, the damage effects of vehicle running vibration and pipeline internal running load cannot be quantified, and a collaborative acceleration corrosion mechanism of chemical corrosion factors and mechanical damage factors is not considered, so that the model prediction precision is low, generalization performance is poor, and the prediction result has larger deviation from the on-site actual corrosion state; (3) Most of the existing corrosion prediction models are static unidirectional modeling, an iterative correction mechanism is not built by combining on-site real-time monitoring data, the model parameters are difficult to adapt to dynamic working conditions of seasonal climate change in a sea area and gradual change of road section traffic flow adjustment after being solidified, precision attenuation can be generated after long-term use, or the model parameters are dynamically adjusted according to actual conditions and then retrained to meet the latest environmental requirements, and the model training time is long, the cost is high, and the efficiency is low. The prior art does not develop synchronous solving calculation aiming at double mechanical coupling and multiple chemical coupling inside and outside a pipeline, and has insufficient quantification of interactive influence and synergistic acceleration corrosion effect on multisource corrosion factors. Disclosure of Invention Aiming at the problems, the invention provides a multi-factor coupling corrosion prediction method for municipal water supply pipelines in a sea area, which comprises the following steps: S1, taking a space-time grid as a minimum calculation unit, collecting attribute parameters of a pipeline body, running parameters in a pipe, parameters of soil outside the pipe and atmospheric environment and traffic load parameters based on sensors laid on the grid, and preprocessing; S2, based on the S1 data and the metal corrosion electrochemical mechanism and municipal engineering design parameters, carrying out weight calculation of salt fog-chloride ion chemical coupling, temperature and humidity-pH value coupling, water pressure fluctuation-material aging coupling, traffic load-buried load transfer-rainfall coupling to obtain four types of dimensionless coupling weight parameters; s3, inputting the coupling weight parameters obtained in the S2 and pipeline materials into a constructed universal coupling corrosion mechanism model for covering chemical corrosion and mechanical damage, and outputting a dimensionless