A landfill leakage risk prediction method based on data assimilation algorithm

Abstract

The present invention provides a method for predicting the leakage risk of a landfill based on a data assimilation algorithm, belonging to the field of landfill leakage risk prediction, including: collecting landfill environmental data, establishing a leakage model, determining the parameters to be optimized and the optimization objective function, clarifying the prior distribution, using the Bayesian-adaptive sampling algorithm to iterate and calibrate the leakage parameters to be optimized, updating the leakage parameters of the leakage model according to the calibration results to obtain an updated landfill leachate leakage model, and using the updated landfill leachate leakage model to predict the future landfill leachate leakage to obtain leakage risk data. The present invention effectively integrates the observed data and the model prediction data, and by continuously iterating and updating the parameters of the leakage model, makes the prediction results more reliable and can timely reflect the actual situation of landfill leakage.

Description

Technical Field

[0001] The present invention relates to the field of landfill leakage risk prediction, and particularly to a method for predicting landfill leakage risk based on a data assimilation algorithm. Background Art

[0002] Hazardous waste landfills are the main places for centralized disposal of hazardous waste and also the places with high environmental risks. The main form of its environmental risk is the leakage of leachate. Most landfills face the problem of leachate leakage, which leads to a gradual increase in environmental risks. The generation and leakage of leachate directly affect the environmental risk prediction of hazardous waste landfills because leachate contains a large amount of toxic and harmful substances, such as heavy metals, organic substances, etc. If not properly treated, it will cause serious pollution to groundwater and soil and even pose a potential threat to human health.

[0003] Traditional methods for predicting landfill leachate leakage risk mainly rely on empirical models and statistical models. Although these methods can meet the prediction requirements to a certain extent, due to their many assumptions, such as factors like the composition of garbage, the structure of landfills, and climate conditions, changes in these factors may affect the generation and leakage of leachate. However, in the actual operation of landfills, these factors may vary greatly, so these assumptions are often difficult to be satisfied, and they cannot effectively integrate observed data and model prediction data and dynamically adjust the parameters of numerical models, resulting in low accuracy and stability. Summary of the Invention

[0004] In order to overcome the deficiencies of the prior art, the purpose of the present invention is to provide a method for predicting landfill leakage risk based on a data assimilation algorithm, which can effectively integrate observed data and model prediction data, and realize dynamic adjustment of model parameters to obtain a more accurate and stable prediction model.

[0005] To achieve the above purpose, the present invention provides the following solutions:

[0006] A method for predicting landfill leakage risk based on a data assimilation algorithm, comprising:

[0007] Collect geological data of the landfill and observed data of the landfill leachate, and establish a basic landfill leachate leakage model based on a single liner structure according to the geological data of the landfill;

[0008] Determine the leakage parameters to be optimized and the optimization objective function of the basic landfill leachate leakage model, and set a prior distribution for the leakage parameters to be optimized; the prior distribution includes: an initial value and a value range;

[0009] Based on the optimization objective function, use the Bayesian-adaptive sampling algorithm to iterate and calibrate the leakage parameters to be optimized to obtain a calibration result;

[0010] Update the leakage parameters of the landfill leachate leakage model according to the calibration results to obtain the updated landfill leachate leakage model;

[0011] Use the updated landfill leachate leakage model to predict the future landfill leachate leakage to obtain landfill leachate leakage risk data.

[0012] Preferably, the steps of iterating and calibrating the leakage parameters to be optimized using the Bayesian-adaptive sampling algorithm are as follows:

[0013] Determine the initial value of the leakage parameters to be optimized to obtain an initial parameter vector;

[0014] Based on the initial parameter vector, use the differential evolution algorithm to generate a set of candidate parameter vectors;

[0015] Calculate the optimization objective function according to the initial parameter vector and the candidate parameter vectors to obtain the calculation result of the optimization objective function;

[0016] Update the initial parameter vector and the candidate parameter vectors according to the update strategy in the Metropolis-Hastings algorithm and the calculation result of the optimization objective function, and adjust the node parameters of the Metropolis iteration;

[0017] Judge whether the calculation result of the optimization objective function converges. If not, return to the step "Calculate the optimization objective function according to the initial parameter vector and the candidate parameter vectors". If so, output the calibration result.

[0018] Preferably, it further includes:

[0019] Calculate the posterior distribution according to the observed data of the landfill leachate to obtain posterior distribution data;

[0020] Evaluate the fitting situation of the leakage volume and the drainage volume of the updated landfill leachate leakage model according to the posterior distribution data.

[0021] Preferably, the leakage parameters to be optimized include: pinhole leak, installation leak, and permeability coefficient of the drainage layer.

[0022] Preferably, the iteration process of the landfill leakage risk prediction model is carried out on the HELP software.

[0023] Preferably, the optimization objective function is the residual of the initial parameter vector and the candidate parameter vectors.

[0024] Preferably, the node parameters of the Metropolis iteration are: the number of parallel chains, the number of single-chain samples, the number of iterations, the scaling factor, and the crossover probability.

[0025] Preferably, the posterior distribution includes: mean square error, root mean square error, mean absolute error, coefficient of determination, and mean percentage error.

[0026] Preferably, the geomembrane is a high-density polyethylene membrane.

[0027] According to the specific embodiments provided by the present invention, the following technical effects are disclosed:

[0028] The present invention provides a landfill leakage risk prediction method based on a data assimilation algorithm, belonging to the field of landfill leakage risk prediction, including: collecting landfill environmental data and establishing a leakage model, determining the parameters to be optimized and the optimization objective function, using the Bayesian-adaptive sampling algorithm to iterate and calibrate the leakage parameters to be optimized, updating the leakage parameters of the leakage model according to the calibration results to obtain an updated landfill leachate leakage model, and using the updated landfill leachate leakage model to predict the future landfill leachate leakage to obtain leakage risk data. The present invention effectively integrates the observed data and the model prediction data, and realizes the dynamic adjustment of the model parameters, and the obtained model has high accuracy and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention, and for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0030] Figure 1 It is a flowchart of the landfill leakage risk prediction provided by the embodiment of the present invention;

[0031] Figure 2 It is a diagram of the basic landfill leachate leakage model provided by the embodiment of the present invention;

[0032] Figure 3 It is a parameter value distribution diagram provided by the embodiment of the present invention;

[0033] Figure 4 It is the correlation between parameters provided by the embodiment of the present invention;

[0034] Figure 5 It is a diagram of the leakage rate calibration provided by the embodiment of the present invention;

[0035] Figure 6 It is a diagram of the drainage rate calibration provided by the embodiment of the present invention;

[0036] Figure 7 Leakage volume verification diagram provided by the embodiment of the present invention;

[0037] Figure 8 Drainage volume verification diagram provided by the embodiment of the present invention. Detailed implementation manners

[0038] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0039] The purpose of the present invention is to provide a landfill leakage risk prediction method based on a data assimilation algorithm, which effectively integrates observed data and model prediction data, and realizes dynamic adjustment of model parameters to obtain a more accurate and stable prediction model.

[0040] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0041] Figure 1 For the landfill leakage risk prediction flow chart, as Figure 1 shown, the present invention provides a landfill leakage risk prediction method based on a data assimilation algorithm, including:

[0042] Collect geological data of the landfill and observed data of the landfill leachate, and establish a basic landfill leachate leakage model based on a single liner structure according to the geological data of the landfill;

[0043] Determine the leakage parameters to be optimized and the optimization objective function of the basic landfill leachate leakage model, and set a prior distribution for the leakage parameters to be optimized; the prior distribution includes: an initial value and a value range;

[0044] Based on the optimization objective function, use the Bayesian-adaptive sampling algorithm to iterate and calibrate the leakage parameters to be optimized to obtain a calibration result;

[0045] Update the leakage parameters of the basic landfill leachate leakage model according to the calibration result to obtain an updated landfill leachate leakage model;

[0046] Use the updated landfill leachate leakage model to predict the future landfill leachate leakage to obtain landfill leachate leakage risk data.

[0047] Further, the steps of iterating and calibrating the leakage parameters to be optimized using the Bayesian-adaptive sampling algorithm are as follows:

[0048] Determine the initial value of the leakage parameters to be optimized to obtain an initial parameter vector;

[0049] Based on the initial parameter vector, use the differential evolution algorithm to generate a set of candidate parameter vectors;

[0050] Calculate the optimization objective function according to the initial parameter vector and the candidate parameter vectors to obtain the calculation result of the optimization objective function;

[0051] Update the initial parameter vector and the candidate parameter vectors according to the update strategy in the Metropolis-Hastings algorithm and the calculation result of the optimization objective function, and adjust the node parameters of the Metropolis iteration;

[0052] Judge whether the calculation result of the optimization objective function converges. If not, return to the step of "calculating the optimization objective function according to the initial parameter vector and the candidate parameter vectors". If so, output the calibration result.

[0053] Further, it also includes:

[0054] According to the observed data of the landfill leachate, calculate the posterior distribution of the landfill leachate leakage risk data to obtain the posterior distribution data;

[0055] Evaluate the fitting situation of the leakage volume and the drainage volume of the updated landfill leachate leakage model according to the posterior distribution data.

[0056] Specifically, the leakage parameters to be optimized include: pinhole leaks, installation leaks, and the permeability coefficient of the drainage layer.

[0057] Specifically, the iteration process of the landfill leakage risk prediction model is carried out on the HELP software.

[0058] Optionally, the optimization objective function is the residual between the initial parameter vector and the candidate parameter vectors.

[0059] Specifically, the node parameters of the Metropolis iteration are: the number of parallel chains, the number of samples per chain, the number of iterations, the scaling factor, and the crossover probability.

[0060] Optionally, the posterior distribution includes: mean square error, root mean square error, mean absolute error, coefficient of determination, and mean percentage error.

[0061] Specifically, the geomembrane is a high-density polyethylene membrane.

[0062] Furthermore, the Bayesian - adaptive sampling algorithm integrates the concepts of Bayesian statistics into the adaptive sampling algorithm to improve the efficiency and accuracy of parameter calibration. The core of this integration lies in using Bayesian inference to update the probability distribution in the parameter space, while combining the concepts of prior and posterior probabilities in Bayesian statistics. The Bayesian - adaptive sampling algorithm is applied to the parameter calibration process of the landfill leachate leakage model to improve the accuracy and reliability of the model.

[0063] Specifically, first, prepare the data, that is, collect the observed data of the landfill leachate water level changes, and establish the corresponding landfill leachate leakage model according to the actual engineering situation. In the model, determine the leakage model and parameters to be optimized, such as the hole density on the landfill geomembrane and the permeability coefficient of the drainage layer, and set initial values for these parameters. Determine the value range of each parameter in the model, that is, set the prior distribution of the parameters based on previous observations or domain expert knowledge. Then, select the optimization objective function: for example, the mean square error (MSE) between the observed data and the model - predicted data can be selected as the optimization objective. Secondly, use the Bayesian - adaptive sampling algorithm to calibrate the parameters. By setting parameters such as the number of parallel chains, the number of samples on each chain, and the number of iterations, sample and update the model parameters. This algorithm uses the minimization of the mean square error between the observed data and the model - predicted data as the optimization objective. Finally, as the algorithm iterates, calculate the posterior distribution of the parameters based on the observed data to best fit the measured data. When the objective function converges, output the optimal parameter estimates. These results can be used to evaluate the accuracy and reliability of the landfill leachate leakage model.

[0064] Specifically, the Bayesian - adaptive sampling algorithm is a method of Bayesian inference for Monte Carlo Markov Chain Monte Carlo (MCMC). Its core lies in the efficient sampling of the parameter space and parameter estimation. In MCMC, setting appropriate sampling parameters has a great impact on the performance and results of the algorithm. Specifically, the key sampling parameter settings are as follows: the number of parallel chains is 4, and the number of parallel chains determines the number of chains for simultaneous MCMC sampling. By increasing the number of parallel chains, the efficiency and convergence speed of the algorithm can be improved, but more computing resources are also required; the number of samples on each chain is 1000, and the number of samples on each chain affects the exploration degree of the parameter space and the sampling accuracy; the number of iterations is set to 1000, and the number of iterations represents the number of times the entire algorithm runs. A larger number of iterations usually means a more thorough exploration of the parameter space, but also comes with a higher computational cost; the scaling factor is set to 0.5, and the scaling factor controls the degree of variation during the parameter update process. The choice of the scaling factor affects the coverage and convergence of the parameter space; the crossover probability is set to 0.9, and the crossover probability determines the probability of the crossover operation in differential evolution.

[0065] Furthermore, in the process of parameter calibration of the landfill leachate leakage model, the Bayesian-adaptive sampling algorithm can provide more accurate parameter estimation for the model, thereby improving the prediction ability and reliability of the leakage situation.

[0066] Optionally, the model fitting accuracy is the error between the model simulation value and the measured data. Common evaluation indicators include the mean squared error MSE, root mean squared error RMSE, mean absolute error MAE, coefficient of determination R 2 , mean percentage error MPE:

[0067] The mean squared error (MSE) is one of the most common fitting accuracy indicators. It calculates the average of the squared errors between the simulation values and the measured values. The smaller the MSE value, the higher the fitting accuracy of the model. The calculation formula of MSE is as follows:

[0068]

[0069] In the formula, m represents the number of samples, y i represents the measured value, represents the simulation value.

[0070] The root mean squared error (RMSE) is the square root of MSE. It measures the error in the same unit as the measured value. A smaller value indicates a higher fitting accuracy of the model. The calculation formula of RMSE is as follows:

[0071]

[0072] The mean absolute error (MAE) calculates the average of the absolute errors between the simulation values and the measured values. Different from MSE, MAE does not consider the square of the error, so it is less sensitive to outliers. The calculation formula of MAE is as follows:

[0073]

[0074] The value range of MAE is [0, +∞). When the simulation value and the measured value are exactly the same, it is equal to 0, that is, a perfect model; the larger the error, the larger this value.

[0075] The coefficient of determination (R-Square) is used to measure the ability of the model to explain the measured data. It represents the proportion of the total variance that the model can explain. R 2 The value range is between 0 and 1. If the result is 0, it means that the model fitting effect is very poor; if the result is 1, it means that the model has no errors; the closer to 1, the higher the fitting accuracy of the model. R 2 The calculation formula of is as follows:

[0076]

[0077] In the formula, the numerator part represents the sum of the squared differences between the true values and the predicted values, similar to the mean squared error MSE; the denominator part represents the sum of the squared differences between the true values and the mean, similar to the variance Var. Generally speaking, the larger the R-Squared, the better the model fitting effect.

[0078] The mean absolute percentage error measures the average of the percentage errors between the simulated values and the measured values. It can be used to solve the percentage error of the model, and the calculation formula is as follows:

[0079]

[0080] The range of MAPE is (0, +∞). A value of 0 indicates a perfect model, and a MAPE greater than 1 indicates a poor model. The smaller the value of MAPE, the better the accuracy of the model.

[0081] Reference Figure 2 , the landfill leachate leakage model established based on the landfill structure and hydrological process model is a comprehensive and widely covered model. This model is divided into three sub-models: the vertical infiltration layer, the lateral drainage layer, and the liner layer. Each sub-model has carefully considered and simulated the leakage process of different parts of the landfill. First, the vertical infiltration layer sub-model considers multiple important factors of the soil, such as porosity, field capacity, wilting point, and permeability coefficient. These factors play a key role in the vertical infiltration process of leachate in the soil. By simulating these parameters, the model can more accurately understand and predict the rate and manner of leachate infiltration downward in the landfill; second, the lateral drainage layer sub-model focuses on the drainage system on the side walls of the landfill, which includes facilities such as drainage ditches and drainage pipes. In addition, factors such as the geometric shape, size, material properties, and water flow velocity of these drainage systems are crucial for simulating the flow process of leachate in the lateral drainage system; the liner layer includes a geomembrane liner layer and a soil liner layer, each with different characteristics and functions. The geomembrane liner layer sub-model considers factors such as the type, thickness, permeability coefficient of the geomembrane, and the contact conditions with the soil. These factors are crucial for simulating the flow process of leachate in the geomembrane liner layer. On the other hand, the soil liner layer is paved with clay, and its permeability coefficient is extremely small, which can effectively prevent the leakage of leachate, thus playing a role in protecting the landfill. This landfill leachate leakage model covers the entire process of leachate generation, drainage, and leakage, providing a basis for the automatic calibration of model parameters.

[0082] Furthermore, by optimizing and adjusting the model parameters, the leakage situation of landfill leachate can be predicted more accurately, providing a scientific basis for the design and management of landfills. The sub-model parameter settings of the landfill leakage model are shown in Table 1. The precise setting of these parameters is crucial for the accuracy and practicality of the model because they directly affect the reliability and predictability of simulating the landfill leachate leakage process.

[0083] Table 1

[0084]

[0085]

[0086] Furthermore, the geomembrane liner mainly uses high-density polyethylene (HDPE) membranes. The HDPE membrane is a core component of the landfill anti-seepage system. The leakage of leachate caused by its damage will seriously pollute the soil and groundwater environment. Therefore, it is crucial to repair the holes in the HDPE membrane of landfills. The generation of holes may be related to factors such as the production process, non-standard construction, and mechanical rolling. By detecting the hole density and calibrating parameters, the safety of the anti-seepage layer can be evaluated and a basis for hole repair can be provided. In addition, as a key facility to prevent leachate from infiltrating downward, the permeability coefficient of the drainage layer directly affects the safe operation of landfills. By calibrating the permeability coefficient of the drainage layer, the flow of leachate in landfills can be described more accurately, providing a scientific basis for the design and management of landfills. In summary, parameters such as the hole density (pinhole holes, installation holes) on the landfill geomembrane and the permeability coefficient of the drainage layer are selected for parameter calibration.

[0087] Optionally, to effectively perform parameter calibration, five months of landfill leakage data are collected and divided into two parts: a calibration set and a validation set. The calibration set contains the data of the first four months and is used to optimize the model parameters in the Bayesian-adaptive sampling algorithm. The data of the last month is used as the validation set to verify the accuracy and prediction ability of the established model. Such a segmentation method can ensure the prediction ability of the model for unknown data and improve its generalization ability and reliability.

[0088] Furthermore, in the Bayesian-adaptive sampling algorithm, the parameter ranges to be optimized are set. That is, the density (# / ha) of pinholes and installation holes and the permeability coefficient (m / d) of the drainage layer are respectively identified by n2, n3, and kd, and their value ranges are determined to be [0, 5], [0, 10], and [2, 6]. The setting of these ranges is based on the actual situation of landfills and prior empirical knowledge, aiming to find the best parameter combination to make the model better fit the real data and have good interpretability.

[0089] Reference Figure 3 andFigure 4 , the value distributions of parameters n2, n3, and kd and their correlations are obtained. These charts show the interactions and influences among the parameters, which helps to more comprehensively analyze the selection and optimization process of model parameters. By deeply studying the distributions and correlations of these parameters, the behavior and performance of the model under different parameter combinations can be better understood, providing important guidance for the further improvement and optimization of the model.

[0090] Reference Figures 5 to 8 , through the parameter calibration of the Bayesian - adaptive sampling algorithm, the estimated values of the optimal parameter combination are obtained as [1.99, 4.99, 5.03]. These parameters are used to verify the model, and the actual monitoring data of the fifth month is used to verify this parameter combination. During the verification period, the drainage volume and leakage volume of the simulated leachate are simulated, and the simulated values are compared with the actual monitoring data. Figures 5 to 8 It reflects the fitting situation between the model prediction value and the actual observation value; the fitting situation is an important indicator to evaluate the prediction ability and accuracy of the model. Figures 5 to 8 The simulation results shown indicate that the fitting situation between the leakage volume and drainage volume obtained by simulation with the calibrated parameter values and the actual monitoring values is good. However, although this figure shows the comparison between the simulated values and the measured values, it is difficult to make an accurate judgment only by visual observation. Therefore, a series of model evaluation indicators, such as the mean square error (MSE), root mean square error (RMSE), coefficient of determination (R 2 ), mean absolute error (MAE), and mean percentage error (MPE), etc., are needed to comprehensively evaluate the fitting effect and prediction ability of the model, and their comparison is shown in Table 2.

[0091] Table 2

[0092]

[0093] Specifically, in order to more detailedly evaluate the effect of the model, this embodiment further compares the error analysis between the simulated values and the measured values. The mean square error (MSE) is a commonly used indicator to measure the prediction error of the model, and the smaller its value, the better the fitting degree of the model to the data. According to the data in Table 2, the mean square error of the model is determined to be 0.003, indicating that the overall fitting effect of the model is relatively good. The root mean square error (RMSE) more intuitively reflects the deviation between the model prediction error and the true value. The root mean square error of this model is 0.054, and the relatively small value indicates that the error in the prediction process of the model is relatively low.

[0094] Furthermore, in addition to the mean square error and the root mean square error, the coefficient of determination (R 2) is also an important indicator for evaluating the goodness of fit of the model. The coefficient of determination ranges from 0 to 1, and the closer the value is to 1, the higher the degree of fit of the model to the observed data. According to the data in Table 2, the coefficient of determination of the model is 0.98, indicating that the model has a strong ability to explain the actual observed data. In practical applications, the mean absolute error (MAE) and the mean percentage error (MPE) are also important evaluation indicators. MAE measures the average absolute deviation between the predicted values and the actual values of the model, while MPE measures the average percentage of the prediction error relative to the actual values. According to the data in Table 2, the MAE of this model is 0.038 and the MPE is 1.2%, indicating that the model's predictions are relatively accurate on average and the deviation is small.

[0095] Furthermore, overall, the errors of the simulated leakage volume and drainage volume during the calibration period are both controlled below 0.1, which indicates that the Bayesian-adaptive sampling algorithm shows relatively high accuracy and reliability in simulating the landfill leachate leakage model; and the fitting accuracy of the simulation results during the calibration period and the verification period to the measured values is relatively high, which means that the model has good fitting ability in predicting the landfill leakage situation and provides a reliable basis for risk assessment.

[0096] Specifically, through the comprehensive analysis of multiple evaluation indicators, it is determined that the parameter combination of the model calibrated by the Bayesian-adaptive sampling algorithm performs well in simulating the leakage volume and drainage volume, has a high degree of fit with the actual monitoring data, and relatively high prediction accuracy.

[0097] Specifically, Figures 5 to 8 In it, True value is the true value; Predictedvalue is the predicted value; the x-axis is time; the y-axis is the flow rate.

[0098] The beneficial effects of the present invention are as follows:

[0099] The present invention effectively integrates the observed data and the model prediction data, and realizes the dynamic adjustment of the model parameters, and the obtained model has high accuracy and stability.

[0100] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The same and similar parts among the various embodiments can be referred to each other.

[0101] Specific examples are used in this article to elaborate on the principles and implementation methods of the present invention. The descriptions of the above embodiments are only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A landfill leakage risk prediction method based on data assimilation algorithm, comprising: Collecting geological data of the landfill and observation data of the landfill leachate and establishing a basic landfill leachate leakage model based on a single liner structure according to the geological data of the landfill, characterized in that it also includes: Determine the leakage parameters to be optimized and the optimization objective function of the basic landfill leachate leakage model, and set a prior distribution for the leakage parameters to be optimized; the prior distribution includes: an initial value and a value range; Based on the optimization objective function, the leakage parameter to be optimized is iterated and calibrated using a Bayesian-adaptive sampling algorithm to obtain a calibration result; The leakage parameters of the basic landfill leachate leakage model are updated according to the calibration results to obtain an updated landfill leachate leakage model; Using the updated landfill leachate leakage model to predict future landfill leachate leakage, and obtaining landfill leachate leakage risk data; The steps of iterating and calibrating the leakage parameter to be optimized using the Bayesian-adaptive sampling algorithm are as follows: Determine the initial value of the leakage parameter setting to be optimized to obtain an initial parameter vector; Based on the initial parameter vector, a set of candidate parameter vectors is generated using a differential evolution algorithm; Calculate the optimization objective function according to the initial parameter vector and the candidate parameter vector to obtain a calculation result of the optimization objective function; The initial parameter vector and the candidate parameter vector are updated according to the update strategy in the Metropolis-Hastings algorithm and the calculation result of the optimization objective function, and the node parameters of the Metropolis iteration are adjusted; Determine whether the calculation result of the optimization objective function converges, if not, return to the step of "calculating the optimization objective function according to the initial parameter vector and the candidate parameter vector", if yes, output the calibration result; The basic landfill leachate leakage model includes three sub-models: vertical permeability layer, lateral drainage layer and liner layer. Each sub-model simulates the leakage process of different parts of the landfill.

2. A landfill leakage risk prediction method based on data assimilation algorithm according to claim 1, characterized in that: Also includes: According to the observation data of the landfill leachate, a posterior distribution calculation is performed on the landfill leachate leakage risk data to obtain posterior distribution data; The fitting of the leakage and drainage volume of the updated landfill leachate leakage model is evaluated according to the posterior distribution data.

3. The landfill leakage risk prediction method based on data assimilation algorithm according to claim 1 is characterized in that: The leakage parameters to be optimized include: the hole density on the landfill geomembrane and the permeability coefficient of the drainage layer, and the hole density includes: pinhole holes and installation holes.

4. The landfill leakage risk prediction method based on data assimilation algorithm according to claim 1 is characterized in that: The iteration process of the landfill leakage risk prediction model is carried out on the HELP software.

5. The landfill leakage risk prediction method based on data assimilation algorithm according to claim 1 is characterized in that: The optimization objective function is the residual of the initial parameter vector and the candidate parameter vector.

6. The landfill leakage risk prediction method based on data assimilation algorithm according to claim 1 is characterized in that: The node parameters of the Metropolis iteration are: the number of parallel chains, the number of single chain samples, the number of iterations, the scaling factor and the crossover probability.

7. The landfill leakage risk prediction method based on data assimilation algorithm according to claim 2 is characterized in that: The posterior distribution includes: mean square error, root mean square error, mean absolute error, coefficient of determination and mean percentage error.

8. The landfill leakage risk prediction method based on data assimilation algorithm according to claim 3 is characterized in that: The geomembrane is a high-density polyethylene film.

Images