CN-122020930-A - Landfill well group optimal layout method based on permeability coefficient field inversion

CN122020930ACN 122020930 ACN122020930 ACN 122020930ACN-122020930-A

Abstract

The invention discloses a landfill well group optimizing layout method based on permeability coefficient field inversion, relates to the field of well group optimizing layout, and utilizes a finite element forward model to couple and observe water level data, and performs dimension reduction treatment on grid nodes through feature decomposition. The technology core is to introduce a bimodal distribution nonlinear mapping mechanism, reconstruct a low-dimensional parameter vector obtained by inversion into a high-resolution permeability coefficient field with clear spatial information of a fracture domain and a matrix domain. According to the physical field distribution, the potential poorly wetted areas are located by identifying the locations of the preferential flow channels and the resulting hydraulic response dead zones. And then, based on the spatial relation between the geometric centers of the dead zones and the heterogeneous flow field, the initial well group coordinates are corrected and encrypted in a targeted manner. Thus, the technical problem that the prior art cannot accurately distribute wells according to the complex structure in the landfill is effectively solved.

Inventors

ZHANG CHENCHENG
HU JIE
MENG MENG
KE HAN
LAN JIWU
ZHAN LIANGTONG
CHEN YUNMIN

Assignees

浙江大学

Dates

Publication Date: 20260512
Application Date: 20260410

Claims (10)

1. A landfill well group optimizing layout method based on permeability coefficient field inversion is characterized by comprising the following steps: s1, constructing an observation water level vector and a finite element forward model based on initial well group coordinates, single well recharging flow and monitoring well water level time sequence data of a target landfill; s2, performing space correlation calculation and feature decomposition truncation processing on grid node coordinates of the finite element forward model to obtain a feature vector matrix and a parameter reconstruction operator; S3, based on the observed water level vector, carrying out parallel iterative optimization on a plurality of groups of initial low-dimensional parameter vectors generated randomly so as to obtain an optimal low-dimensional parameter vector capable of minimizing simulation errors; s4, inputting an optimal low-dimensional parameter vector into a parameter reconstruction operator, and carrying out high-resolution physical field reconstruction by combining linear combination of feature vector matrixes and double-peak distribution nonlinear mapping so as to obtain a permeability coefficient field distribution diagram containing space information of a fracture domain and a matrix domain; And S5, performing well group layout optimization based on heterogeneous characteristics on the permeability coefficient field distribution map based on the initial well group coordinates to obtain a landfill well group optimal layout scheme.
2. The landfill well group optimizing layout method based on permeability coefficient field inversion according to claim 1, wherein step S1 comprises: Extracting initial stable water level and sampling transient water level from the water level time sequence data of the monitoring well, and vector assembling the initial stable water level and the sampling transient water level to obtain an observed water level vector; Defining a core research area and a buffer zone according to the initial well group coordinates, and discretizing the core research area and the buffer zone by using a finite element method to obtain finite element grid data; and constructing a finite element forward model based on the finite element grid data and the single well recharging flow.
3. The landfill well group optimizing layout method based on the permeability coefficient field inversion according to claim 2, wherein step S2 comprises: based on a preset priori statistical parameter set, carrying out space correlation quantization on the finite element grid data to obtain a space covariance matrix; performing feature decomposition and truncation based on principal component analysis on the space covariance matrix and the priori statistical parameter set to obtain a truncated feature vector matrix and a truncated feature value matrix; And constructing a parameter reconstruction operator based on the cut-off eigenvector matrix, the cut-off eigenvalue matrix and the priori statistical parameter set.
4. The landfill well group optimizing layout method based on the permeability coefficient field inversion according to claim 3, wherein the constructing the parameter reconstruction operator based on the cut-off eigenvector matrix, the cut-off eigenvalue matrix and the priori statistical parameter set comprises constructing a nonlinear mapping chain comprising linear Gaussian mapping, probability integral transformation and bimodal distribution inverse cumulative distribution function transformation based on the cut-off eigenvector matrix, the cut-off eigenvalue matrix and the priori statistical parameter set as the parameter reconstruction operator.
5. The landfill well group optimizing layout method based on the permeability coefficient field inversion according to claim 1, wherein step S3 comprises: based on the observation water level vector, the parameter reconstruction operator and the finite element forward model, constructing a comprehensive objective function comprising a weighted least squares data fitting item and a priori distribution regularization item; Inputting a plurality of groups of randomly generated initial low-dimensional parameter vectors into a comprehensive objective function for parallel iterative optimization so as to generate a convergence parameter vector set; and screening the optimal low-dimensional parameter vector from the convergence parameter vector set.
6. The landfill well group optimizing layout method based on the permeability coefficient field inversion according to claim 1, wherein step S4 comprises: Inputting the optimal low-dimensional parameter vector into a parameter reconstruction operator to perform linear Gaussian field recovery and nonlinear inverse cumulative distribution function transformation so as to obtain an optimal logarithmic permeability coefficient field; Converting the optimal logarithmic permeability coefficient field into a physical permeability coefficient field, and carrying out region classification according to a preset hydraulic conductivity threshold to obtain a permeability coefficient field distribution diagram for distinguishing a fracture domain from a matrix domain; And carrying out connected domain searching and path tracking on the fracture domains in the permeability coefficient field distribution diagram to identify the position information of the preferential flow channel.
7. The landfill well group optimizing layout method based on the permeability coefficient field inversion according to claim 1, wherein step S5 comprises: based on the initial well group coordinates, carrying out hydraulic response blind area identification and extraction on the permeability coefficient field distribution map to obtain a potential poor wetting blind area; Calculating the geometric centroid of the potential poor wetting blind area, and carrying out coordinate correction according to the space distance threshold value of the position information of the preferential flow channel to obtain corrected encryption well coordinates; and combining the corrected encrypted well coordinates and the initial well group coordinates to obtain the well group optimal layout scheme of the landfill site.
8. The optimal placement method for well groups in a landfill based on inversion of a permeability coefficient field according to claim 7, wherein the identifying and extracting the hydraulic response blind areas of the permeability coefficient field distribution map based on the initial well group coordinates to obtain the potential poor wetting blind areas comprises: Identifying a matrix domain in the permeability field profile; Determining an effective coverage area based on the initial well group coordinates; The matrix area and effective footprint are spatially differentiated to identify potential poorly wetting dead zones.
9. The landfill well group optimizing layout method based on the osmotic coefficient field inversion of claim 8, wherein the spatial difference between the matrix area and the effective coverage area is used for identifying the potential poorly wetting blind area, and the method comprises the following steps of spatial difference between the matrix area and the effective coverage area, wherein the formula is as follows: Wherein, the In order to potentially wet out the dead zone, As a set of matrix domains, For the coordinate vector of the ith existing well, In order to design the influence radius, For the number of wells to be present, To calculate any spatial coordinate vector within the domain, The domain is calculated for the entire landfill, In order to be a sphere or a circular region operator with a certain point as a center, Is the set of effective coverage areas.
10. The landfill well group optimizing layout method based on the osmotic coefficient field inversion according to claim 5, wherein the screening out the optimal low-dimensional parameter vector from the convergence parameter vector set comprises: Performing structure fidelity divergence estimation of a posterior field on the convergence parameter vector set, the parameter reconstruction operator and the prior statistical parameter set to obtain a structure divergence set; performing normalized coupling of fitting residual errors and structure divergences on the structure divergences set and the convergence parameter vector set to obtain a physical data comprehensive scoring set; and identifying and extracting a solution with the minimum comprehensive score by global minimum value searching and index locking based on the physical data comprehensive score set and the convergence parameter vector set so as to obtain an optimal low-dimensional parameter vector.

Description

Landfill well group optimal layout method based on permeability coefficient field inversion Technical Field The invention belongs to the field of well group optimal layout, in particular to the field of well group optimal layout, and more particularly relates to a landfill well group optimal layout method based on permeability coefficient field inversion. Background At present, sanitary landfill is still a main way to treat urban solid waste worldwide, and in order to accelerate the stabilization process of landfill bodies and reduce the treatment cost of percolate, a bioreactor landfill technology is widely applied, and the core of the technology is to optimize the water and nutrient distribution inside the landfill bodies through the recharging of percolate. However, urban solid waste is used as an artificial medium with great complexity, the internal structure of which is generally composed of high-permeability fissured domains (such as macropores and uncompacted voids) and low-permeability matrix domains (such as compacted garbage and soil) in a staggered manner, and the remarkable heterogeneity and the characteristic of double-peak voids cause a strong preferential flow phenomenon. During recharging, the percolate tends to migrate rapidly along a few high permeability channels, while most of the waste matrix is bypassed, resulting in a low recharging efficiency, a very uneven distribution of dry and wet inside the landfill body, and difficulty in achieving the desired accelerated degradation effect. The accurate characterization of the spatial distribution of permeability coefficients within a landfill and the optimization of recharging well layout accordingly is key to solving the above problems. Existing landfill well group layout schemes are designed depending on empirical formulas or homogenization assumptions. In conventional engineering practice, it is often assumed that the landfill body has uniform hydraulic characteristics, and well group layout is performed by adopting a regular grid shape or a plum blossom shape according to a single equivalent permeability coefficient and a design influence radius. However, this deployment approach based on the homogeneity assumption ignores the objectively existing spatial variability within the landfill body, failing to identify and cover the huge hydraulic response dead zone created by the preferential flow channel contending for flow. Although some inversion methods based on hydraulic tomography or geostatistics are used for exploring the parameter distribution of an underground medium at present, when aiming at a special medium such as a landfill site, the prior art is difficult to accurately reconstruct a bimodal distribution permeability coefficient field which accords with the physical characteristics of the underground medium, and the fracture domain and the matrix domain cannot be effectively distinguished, and on the other hand, the existing inversion results are multi-stop in parameter characterization, and a systematic method capable of directly converting a high-resolution physical field reconstruction result into well group space coordinate optimization is lacked. Therefore, how to automatically identify the blind area with poor wetting and correct the encrypted well position based on the refined permeability coefficient field obtained by inversion of the monitoring data becomes a technical problem to be solved in the engineering design of the current landfill. Disclosure of Invention The invention provides a landfill well group optimal layout method based on permeability coefficient field inversion. The technical scheme includes that the landfill well group optimal layout method based on the osmotic coefficient field inversion comprises the following steps of S1, constructing an observation water level vector and a finite element forward model based on initial well group coordinates, single well recharging flow and monitoring well water level time sequence data of a target landfill, S2, conducting space correlation calculation and characteristic decomposition truncation processing on grid node coordinates of the finite element forward model to obtain a characteristic vector matrix and a parameter reconstruction operator, S3, conducting parallel iterative optimization on a plurality of groups of randomly generated initial low-dimensional parameter vectors based on the observation water level vector to obtain an optimal low-dimensional parameter vector capable of minimizing simulation errors, S4, inputting the optimal low-dimensional parameter vector into the parameter reconstruction operator, conducting high-resolution physical field reconstruction through combination of linear combination of the characteristic vector matrix and double-peak distribution nonlinear mapping to obtain an osmotic coefficient field distribution map comprising spatial information of a matrix domain, and S5, conducting non-homogeneous characteristic-based well group