CN117291349A - Groundwater level restoration prediction method and system - Google Patents

Groundwater level restoration prediction method and system Download PDF

Info

Publication number
CN117291349A
CN117291349A CN202311576343.8A CN202311576343A CN117291349A CN 117291349 A CN117291349 A CN 117291349A CN 202311576343 A CN202311576343 A CN 202311576343A CN 117291349 A CN117291349 A CN 117291349A
Authority
CN
China
Prior art keywords
data
groundwater
flow
water level
obtaining
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202311576343.8A
Other languages
Chinese (zh)
Other versions
CN117291349B (en
Inventor
于芳
孟冲
于丽
申鹏凯
李俊伟
王峥
肖娟
郝振杰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jining Yulongyuan Water Service Co ltd
Original Assignee
Jining Yulongyuan Water Service Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Jining Yulongyuan Water Service Co ltd filed Critical Jining Yulongyuan Water Service Co ltd
Priority to CN202311576343.8A priority Critical patent/CN117291349B/en
Publication of CN117291349A publication Critical patent/CN117291349A/en
Application granted granted Critical
Publication of CN117291349B publication Critical patent/CN117291349B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/20—Ensemble learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/02—Agriculture; Fishing; Forestry; Mining
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Marketing (AREA)
  • Data Mining & Analysis (AREA)
  • Health & Medical Sciences (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Health & Medical Sciences (AREA)
  • Development Economics (AREA)
  • Operations Research (AREA)
  • Game Theory and Decision Science (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Primary Health Care (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • Quality & Reliability (AREA)
  • Evolutionary Biology (AREA)
  • Mining & Mineral Resources (AREA)
  • Animal Husbandry (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Agronomy & Crop Science (AREA)
  • Marine Sciences & Fisheries (AREA)
  • Educational Administration (AREA)
  • Public Health (AREA)
  • Medical Informatics (AREA)
  • Water Supply & Treatment (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Geophysics And Detection Of Objects (AREA)

Abstract

The invention relates to the technical field of groundwater level prediction, in particular to a groundwater level recovery prediction method and a groundwater level recovery prediction system, wherein the method comprises the following steps: the method comprises the steps of forming a time sequence matrix by collecting related data of each period affecting the change of the underground water level; obtaining a periodic hysteresis relation of the data according to the frequency spectrum curve relation of each row of data and the water level height data; obtaining an environmental change factor according to the periodic hysteresis relationship and the bias correlation coefficient of the groundwater flow, the flow speed and the flow direction data; obtaining a hydrologic environment influence factor according to the environment change factor, the period hysteresis of each row of data and the data distribution condition; and (3) improving the initial weight of each height data in the groundwater in the Adaboost algorithm according to the hydrologic environmental impact factor, and training the prediction result of the output water level height sample of the weak classifier. The method improves the robustness and accuracy of the Adaboost algorithm.

Description

Groundwater level restoration prediction method and system
Technical Field
The application relates to the field of groundwater level prediction, in particular to a groundwater level recovery prediction method and a groundwater level recovery prediction system.
Background
Groundwater is an extremely important fresh water resource and is critical to various fields such as drinking water supply, agricultural irrigation, industrial production and the like, however, continuous drop of groundwater level not only causes shortage of water resource, but also has negative effects on ecosystems including wetland degradation, exhaustion of river and loss of biodiversity, and in addition, groundwater level drop may also cause risks such as land subsidence, geological disasters and damage to infrastructure.
It is worth noting that agricultural irrigation is highly dependent on underground water resources, the reduction of underground water level possibly causes serious harm to agricultural production, the yield of agricultural products is reduced, the possibility of frequent natural disasters such as drought and fire is increased, and the prediction of the recovery of the underground water level has extremely important significance in order to ensure sustainable utilization of the underground water resources.
The traditional groundwater level recovery prediction system is relatively mature, however, the traditional water level prediction method is usually only aimed at the angles of groundwater flow, groundwater quality change and the like, the research on the groundwater level is not well combined with the complicated groundwater environment of the groundwater, the groundwater environment comprises the influence of geological structures and human factors, the complicated groundwater environment can influence the flowing and storage of the groundwater, and therefore, the groundwater level needs to be predicted by comprehensively considering the factors such as the groundwater hydrologic environment and the like so as to better understand and manage the groundwater system.
Disclosure of Invention
In order to solve the technical problems, the invention provides a groundwater level restoration prediction method and a groundwater level restoration prediction system, and the adopted technical scheme is as follows:
in a first aspect, an embodiment of the present invention provides a groundwater level restoration prediction method, including the steps of:
the method comprises the steps of forming a time sequence matrix by collecting related data of each period affecting the change of the underground water level;
acquiring a spectrum curve of each row of data in a time sequence matrix; for each row of data of the time sequence matrix, obtaining a periodic hysteresis relation of the data according to a distribution relation between the data and a frequency spectrum curve of the water level height data; calculating partial correlation coefficients among three data of groundwater flow, flow speed and flow direction in the time sequence matrix; obtaining an environmental change factor according to the periodic hysteresis relationship of groundwater flow, flow speed and flow direction data in the time sequence matrix and the partial correlation coefficient;
obtaining a hydrologic environment influence factor according to the environment change factor, the periodic hysteresis of each row of data of the time sequence matrix and the distribution condition of each row of data; according to the hydrologic environmental impact factors, improving the initial weight of each height data of the underground water in the Adaboost algorithm; and training the weak classifier according to the initial weight of each height data in the underground water to output a prediction result of the water level height sample.
Preferably, the forming a time series matrix by collecting the related data of each period affecting the groundwater level change includes:
collecting relevant data affecting groundwater level changes includes: rainfall, groundwater temperature, groundwater level height, groundwater flow speed, groundwater flow direction, groundwater ion concentration data, and groundwater pH data;
for each period, each data acquired in the period is formed into each row vector of the time series matrix.
Preferably, the acquiring the spectral curve of each row of data in the time sequence matrix includes:
and carrying out frequency domain transformation on each data of the time sequence matrix by adopting a fast Fourier transformation algorithm, and obtaining a frequency spectrum curve corresponding to each data.
Preferably, the obtaining the periodic hysteresis relation of the data according to the distribution relation between the data and the frequency spectrum curve of the water level height data includes:
obtaining the maximum amplitude value of a frequency spectrum curve of the data and the water level height data, and calculating the frequency corresponding to the maximum amplitude value of the data and the water level height data;
obtaining the maximum value of the frequency of the data and the water level height data;
calculating the absolute value of the difference value of the frequency between the data and the water level height data, and taking the ratio of the maximum amplitude value of the data and the water level height data as the true number of a logarithmic function taking a natural constant as a base number;
dividing the product of the absolute value of the calculation result of the logarithmic function and the absolute value of the difference by the maximum value to obtain a periodic hysteresis relation of data; wherein the cycle hysteresis relationship of the data is set to 0 when the data and the water level height data are unchanged in the cycle.
Preferably, the calculating the partial correlation coefficient between the groundwater flow, the flow speed and the flow direction three data in the time sequence matrix includes:
for each row of data of groundwater flow, flow speed and flow direction in the time sequence matrix, controlling the groundwater flow data into a constant, and analyzing the net correlation between the groundwater flow speed and the groundwater flow direction data to obtain a partial correlation coefficient of the groundwater flow data;
controlling the groundwater flow speed data to be constant, and analyzing the net correlation between the groundwater flow speed data and the groundwater flow speed data to obtain the partial correlation coefficient of the groundwater flow speed data;
and controlling the groundwater flow data to be constant, and analyzing the net correlation between the groundwater flow and the groundwater flow data to obtain the partial correlation coefficient of the groundwater flow data.
Preferably, the obtaining the environmental change factor according to the periodic hysteresis relationship and the partial correlation coefficient of the groundwater flow, the flow speed and the flow direction data in the time sequence matrix includes:
respectively taking the groundwater flow, the flow speed and the flow direction data in the time sequence matrix as each target line data, and respectively calculating the partial correlation coefficient mean value between each target line data and other two lines data;
taking the opposite number of the cycle hysteresis of the target line data as an index of an index function taking a natural constant as a base, calculating the product of the calculation result of the index function and the absolute value of the mean value of the partial correlation coefficient, and taking the sum of the product of the groundwater flow, the flow speed and the flow direction data as an environment change factor.
Preferably, the obtaining the hydrographic environmental impact factor according to the environmental change factor, the period hysteresis of each row of data of the time sequence matrix and the distribution condition of each row of data includes:
for each row of data of the time sequence matrix, obtaining the maximum value and the minimum value of the data;
taking the opposite number of the periodic hysteresis of the data as an index of an exponential function based on a natural constant, calculating the product of the difference result of the maximum value and the minimum value and the calculation result of the exponential function, and taking the product of the average value of the products of all the data of the time sequence matrix and an environmental change factor as a hydrographic environmental influence factor.
Preferably, the improving the initial weight of each height data of the groundwater in the Adaboost algorithm according to the hydrographic environmental impact factor includes:
for each collected height data of the underground water, obtaining a ratio of the height data to the height data adjacent to the previous height data, taking the product of the opposite number of the ratio and the hydrologic environment influence factor as an index of an exponential function based on a natural constant, and taking the calculation result of the exponential function as the initial weight of the underground water height data in an Adaboost algorithm.
Preferably, the training the prediction result of the water level height sample by the weak classifier according to the initial weight of each height data in the groundwater includes:
training the G weak classifiers to obtain strong classifiers by using initial weights of all height data in the underground water, wherein the input of the strong classifiers is the height data of the underground water, and the input of the strong classifiers is the prediction result of the height data of the underground water; wherein G is the number of preset weak classifiers.
In a second aspect, an embodiment of the present invention further provides a groundwater level recovery prediction system, including a memory, a processor, and a computer program stored in the memory and running on the processor, where the processor implements the steps of any one of the methods described above when executing the computer program.
The invention has at least the following beneficial effects:
the invention provides a method and a system for predicting the recovery of an underground water level, which are used for constructing the cycle hysteresis degree of the underground water level change caused by different data based on non-transient change generated by the influence of other factors on the underground water level change, and the cycle hysteresis degree reflects the correlation between the maximum periodicity of two data by comparing the maximum amplitude difference of different data sets, so that whether the water level is directly and rapidly influenced by certain type of data change can be well explained;
the environmental change factor is constructed by combining the period hysteresis change with the analysis of the groundwater environment, the dominant degree of various environmental influences is judged through the calculation of the partial correlation coefficient, and the influence of the environmental factors on the water level is subjected to decision analysis;
according to the method, the hydrologic environment influence factors are constructed to reflect the influence degree of the comprehensive hydrologic environment on the water level, and the initial weights of the samples in the Adaboost algorithm are changed according to the hydrologic environment influence factors, so that each sample has unique weight after initialization, the greater the influence of the hydrologic environment on the water level at the moment is, the lower the weight is, the more easily and correctly classified the samples at the moment are, and the robustness and accuracy of the Adaboost algorithm are improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a groundwater level restoration prediction method provided by the invention;
fig. 2 is a flow chart of index construction for groundwater level data prediction.
Detailed Description
In order to further describe the technical means and effects adopted by the present invention to achieve the preset purpose, the following detailed description refers to specific embodiments, structures, features and effects of a method and a system for predicting groundwater level recovery according to the present invention, with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the groundwater level restoration prediction method and system provided by the invention with reference to the accompanying drawings.
The embodiment of the invention provides a groundwater level restoration prediction method and a groundwater level restoration prediction system.
Specifically, referring to fig. 1, the following method for predicting groundwater level recovery is provided, and the method includes the following steps:
and S001, collecting related data influencing groundwater level change, preprocessing the collected data, and constructing a time sequence matrix.
In order to more accurately predict the groundwater level, a water level well is arranged at each interval W meter at the position of the groundwater, and related sensors are arranged on the water level well to acquire required data, and environmental information, groundwater level information, groundwater flow information and groundwater content information are periodically acquired. In this embodiment, the verification value 100 is taken for W, and the practitioner can set the verification value by himself.
The environment information comprises rainfall and underground temperature information, wherein the rainfall obtains total rainfall after the last data acquisition time from the data acquisition time by installing a rainfall meter near a water level well, and the underground temperature information is acquired by the water level well; the water level information of the underground water is the water level height of the underground water, and is collected through a water level well; the flow information of the groundwater comprises groundwater flow, groundwater flow speed and groundwater flow direction, wherein the groundwater flow and the groundwater flow speed are collected in a mode of installing a flow sensor and a flow speed sensor on a water level well, and the groundwater flow direction is used for determining the flow direction of the groundwater by judging the pressure change condition of the groundwater at different points through installing a pressure sensor on the water level well; the underground water content information comprises underground water ion (nitrate and heavy metal ion) concentration data and underground water pH data, is collected by installing a chemical sensor array (comprising ion selective electric shock and pH sensors) on a water level well, and performs unified normalization processing on all collected data.
The total m data to be acquired are provided, the acquisition period of each data is T days, and n times of data are uniformly acquired in one period, wherein m takes an empirical value of 8,T, n takes an empirical value of 30, and an implementer can set the data by himself. Sequencing according to the data acquisition sequence to construct a time sequence matrix acquired in the Nth period：
Wherein,time series matrix representing the nth period acquisition, < >>Represents->Seed data->Sub-acquisition of results, e.g.)>And (5) normalizing the specific value of the groundwater temperature acquired by the nth time of the water level well.
And step S002, analyzing the effect of each row of data in the time sequence matrix on the groundwater level in a targeted manner, thereby constructing a hydrologic environment influence factor.
Groundwater is taken as an important component in natural environment, the water level of the groundwater is affected by the rock stratum environment and artificial activities, the permeability and the porosity of the rock stratum directly model the movement characteristics of the groundwater, the rock stratum with high permeability and large porosity is beneficial to the flow and storage of water, the influence of the artificial activities on the groundwater level is mainly represented in aspects of pumping, drainage system, underground exploitation, underground gas storage, oil storage and the like, excessive pumping can lead to water level reduction, the urban and land utilization changes reduce the penetration of water, and the underground engineering and storage activities can change the path of groundwater flow, so that the change of the groundwater level is a complex process, and the geological factors and the artificial activities need to be comprehensively considered to ensure accurate prediction on the groundwater level.
Since the change in the groundwater level is affected by various factors, the change in the groundwater level is not instantaneous with respect to other effects, but has a time delay. Therefore, the factors influencing the water level change usually cause the periodic change of the groundwater level after a period of time, for example, after the periodic change of the rainfall is performed for a period of time, the factors firstly flow to a river or a lake in a surface runoff form, then gradually permeate soil to form a saturated zone, and finally migrate to a deeper groundwater layer, and the periodic change of the groundwater level only occurs at the moment, in this case, the periodic change of the water level change data and the periodic change of the groundwater flow have a certain time lag relationship.
The present embodiment uses a fast fourier transform to convert each row of data in the time series matrix from the time domain to the frequency domain, for the firstThe data of the rows are fourier transformed to obtain a corresponding frequency curve. The fast fourier transform is a known technique, and this embodiment is not described in detail.
For the firstFrequency spectrum curve obtained by carrying out Fourier transformation on data, and the maximum amplitude value is recorded as +.>Frequency corresponding to maximum amplitude +.>Calculating the cycle lag relation of the ith row data relative to the 3 rd row (water level height) data as the frequency corresponding to the main periodicity of the row data +.>：
Wherein,for the periodic hysteresis of the ith row relative to the third row data, +.>And->Maximum amplitude values of spectral curves corresponding to the ith row and 3 rd row data respectively,/->And->Maximum amplitude value corresponding frequency of data of ith row and 3 rd row respectively, +.>The maximum value of the data in brackets is taken as a function of the maximum value.
The represented hysteresis relationship can reveal the periodic variation hysteresis of each data relative to the water level height data, which is helpful for analyzing the time delay characteristics among different variables, < >>The larger the value, the more->The stronger the period hysteresis of the data relative to the water level data is, +.>Representing the ratio of the maximum amplitude value of each row of data to the maximum amplitude value of the water level height data, reflecting the difference condition of the amplitude of each row of data relative to the amplitude of the water level data, expressing the strength of periodical change, and when the difference of the maximum amplitude values of the two data is smaller, calculating the difference of the maximum amplitude value of the two data>Approaching 0, whereas the larger the difference is +.>The closer to infinity, ++>Along with->Is increased by an increase in (a); />By comprehensively considering the frequency difference between each row of data and the water level height data, the hysteresis state of each row of data relative to the water level height data is reflected, and the denominator acts as a difference value of quantized molecules, +.>Representing the time delay relationship of the periodic variation between the line i data and the water level data, the Fourier transform converts the signals from the time domain to the frequency domain, the components of the signals at different frequencies can be known, when comparing the spectra of the two signals, the dominant periodic variation between them can be known from their maximum frequency components, if there is a time delay between the two signals, their spectra will undergo a certain shift, which shift may reflect the delay of one signal relative to the other, i.e.)>The bigger the->The larger.
It should be noted that if the ith data is unchanged, for example, there is no rainfall in the acquisition period, the data in the ith line is identical, and the frequency domain of the data after fourier transformation at this time will show a very sharp peak because all the energy is concentrated on one frequency, which will result in amplitudeParticularly large, at this timeApproaching infinity, it is indicated that there is no hysteresis in the data for lines i and 3, or that hysteresis does not occur during this period; similarly, if the line 3 water level data is hardly changed, +.>Also approaching infinity; when the data of the ith row and the 3 rd row are not changed, the frequency domain representation of the two signals has very sharpSharp peaks, but no hysteresis relationship exists due to no change in the data they correspond to, whereEqual to 0 means that the data of row i and row 3 do not have a hysteresis reaction during this period, no significant relative delay, i.e.>Set to 0.
The groundwater environment affects the groundwater level mainly by environmental and human influences, and the groundwater level is comprehensively affected by various natural and human factors including climate change, rainfall and evaporation, geology and soil characteristics, human activities and the like. For example, slower groundwater flow and flow speed results in groundwater accumulation, raising groundwater level; and the groundwater level may be lowered due to factors such as excessive pumping of groundwater, precipitation reduction caused by climate change, etc. Thus, the change in groundwater level is the result of interactions of various factors in the groundwater environment.
Among environmental influences, different geological conditions of groundwater have different influences on groundwater flow characteristics of the groundwater, for example, water absorption, permeability and pore structure of a rock stratum where the groundwater is located can influence flow rate and flow velocity of the groundwater, and structures such as faults and cracks of the rock stratum where the groundwater is located can influence flow direction of the groundwater.
In order to eliminate the influence of the geological conditions of the groundwater on the groundwater, in the embodiment, one variable is controlled to be constant for the flow speed, the flow rate and the flow direction of the groundwater, and the partial correlation coefficients of the variable and the other two variables are calculated by taking lines 4, 5 and 6 in the acquired time sequence matrix as input、/>And->Wherein->Representing the partial correlation coefficient of flow direction and flow rate in the case of controlled flow, +.>A partial correlation coefficient representing the flow and direction in the case of a controlled flow rate, +.>Represents the partial correlation coefficient of the flow rate and the flow rate under the condition of controlling the flow direction, wherein the partial correlation coefficient is a known technology, and the description of the embodiment is omitted.
And calculate the mean value of the partial correlation coefficient between one variable and the other two variables、/>And->，/>Represents the mean value of the partial correlation coefficients of flow and flow velocity and flow direction, wherein +.>A mean value of the partial correlation coefficients representing the flow rate and the flow and direction, < >>The average value of the partial correlation coefficients representing the flow direction, the flow rate and the flow velocity, the environmental change factor Z is constructed according to the partial correlation coefficients among the fourth data, the fifth data and the sixth data, and the hydrologic environmental influence factor is constructed according to the variation amplitude of various data and the maximum periodicity analysis>：
Wherein,is the hydrologic environment influence factor, comprehensively judging the influence of the hydrologic environment on the water level change, Z is the environment change factor, m is the number of the acquired data types, and the number of the acquired data types is +.>And->Maximum value and minimum value of q-th line data, respectively,/->Is an exponential function based on a natural constant e, +.>For the periodic hysteresis of the q-th line data relative to the third line data, r takes a value of 6,/->The data is the average value of the partial correlation coefficients of the p-th data and the other two types of groundwater flow characteristics.
In the case where the environmental change affects the Z,representing the periodic hysteresis of the p-th data relative to the water level data when +.>The larger the p-th line data, the longer the time interval between the change of the p-th line data and the change of the water level data, the smaller the direct influence of the p-th line data on the water level data, the smaller Z is, and +.>Comprehensively considering the partial correlation coefficient with other flow characteristics to reflect the degree that the flow characteristic change caused by the groundwater environment is dominant, wherein the larger the value is, the more the p-th variable has strong correlation with the other variable when one variable is controlled, namely the influence of the p-th data on the other variable is not limited by the other variable, the larger the dominant position is occupied, the stronger the influence on the water level is, the more the p-th data has strong influence on the water level>The larger.
While in the context of a hydrographic environmental impact factorIn (c) by calculating the variation amplitude of the q-th line data +.>Taking the product of the exponential function of the lag relation of the q-th line data relative to the water level data into comprehensive consideration, wherein the influence of the variation range of the q-th line data and the periodic information on the water level variation is +.>The larger the data representing the q-th row is, the larger the variation amplitude of the data is, the greater the possibility of affecting the water level is, and +.>With a consequent increase in->The larger represents the longer the time interval of the water level change caused by the q-th data change, possibly limited by other factors when affecting the water level, +.>And consequently decreases.
It is worth mentioning that when q=7 and q=8, the magnitude of the change in nitrate, heavy metal ion concentration and pH can describe the influence of human factors on the water level significantly, because significant changes in nitrate, heavy metal ion and pH are caused when agricultural or industrial water is discharged into groundwater.
And step S003, the initial weight of the water level sample in the Adaboost algorithm is improved by combining the hydrologic environment influence factor, and prediction is made for groundwater level recovery influenced by the hydrologic environment.
In this embodiment, adaboost (adaptive enhancement algorithm) is used to predict the groundwater level, adaboost is an integrated learning algorithm, and is used to improve the overall classification performance of the weak classifier, only the height data of groundwater is selected as a sample, and no consideration is given to other hydrologic environmental factors, that is, each data point in the height data of the groundwater in the third row in the detected time series matrix is taken as a sample, the weight distribution of the training sample is initialized, and the original algorithm generally initializes the weight of each sample to 1/N, where N is the number of samples. The initial weight of each sample is improved, and the initial weight of the height data of the ith groundwater after improvement is set asThen:
wherein,is the initial weight of the ith acquisition data in the height data of the groundwater, +.>Data value representing the ith acquisition of groundwater elevation data,/th acquisition of groundwater elevation data>Data values representing the i-1 st acquisition of the height data of groundwater, ++>The larger the value of the hydrographic environment influence factor is, the larger the influence of the hydrographic environment on the water level in the detected moment period is represented, and if the current acquired data is relative to the phaseThe larger the data value of the next-to-last acquired data, the smaller the data point needs to be given a smaller initial weight, reducing the likelihood that the sample is misclassified.
Wherein, for the first data of the groundwater, the formula is given bySet to->。
Improved initial weightSamples with critical effects of the hydrographic environment on water level prediction can be better focused, and since the hydrographic environment has a larger effect on the groundwater level with a larger hydrographic environment influence factor, by reducing the initial weight of samples with large hydrographic environment influence factors, it is easier to classify correctly at the beginning of such samples, so that the algorithm can guide the model to learn these critical features earlier.
After the improved initial weight is allocated to each training sample, setting the iteration number D as an empirical value 20, and prescribing that the training sample can be combined with various weak classifiers of different types, such as decision trees, random forests and the like, combining G weak classifiers in total, taking the empirical value 10, wherein the types of the weak classifiers can be the same or different, the specific combination mode and the number can be flexibly adjusted according to actual conditions, and training one weak classifier according to the current weight distribution in each iteration, and calculating the classification error rate of the weak classifier.
Calculating the weight of the classifier through the error rate of the classifier, updating the weight distribution of the sample, increasing the weight of the sample subjected to error classification, reducing the weight of the sample subjected to correct classification, stopping iteration after reaching the preset iteration times, continuously adjusting the weight of the weak classifier and the weight of the sample along with the progress of iteration, so that the sample subjected to error classification is more concerned in the subsequent iteration, thereby improving the overall classification performance, finally carrying out weighted combination according to the weight of each weak classifier to obtain a strong classifier, and the strong classifier can receive the water level detection sequence of a time sequence matrix as input and output the prediction result of the water level height sample, and aims to finish accurate prediction of groundwater level recovery and provide reliable estimation for groundwater level change.
The index construction flow chart of the groundwater level data prediction is shown in fig. 2.
So far, groundwater level prediction data can be obtained through the method.
Based on the same inventive concept as the above method, the embodiment of the invention further provides a groundwater level restoration prediction system, which comprises a memory, a processor and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of any one of the above groundwater level restoration prediction methods.
The embodiment of the invention provides a method and a system for predicting groundwater level recovery, which are used for constructing the periodic hysteresis degree of groundwater level change caused by different data based on non-transient change generated by the influence of groundwater level change by other factors, and the periodic hysteresis degree reflects the association between the maximum periodicity of two data by comparing the maximum amplitude difference of different data sets, so that whether the water level is directly and rapidly influenced by certain type of data change can be well explained;
the environmental change factor is constructed by combining the period hysteresis change with the analysis of the groundwater environment, the dominant degree of various environmental influences is judged through the calculation of the partial correlation coefficient, and the influence of the environmental factors on the water level is subjected to decision analysis;
according to the embodiment of the invention, the hydrologic environment influence factors are constructed to reflect the influence degree of the comprehensive hydrologic environment on the water level, and the initial weights of the samples in the Adaboost algorithm are changed according to the hydrologic environment influence factors, so that each sample has unique weight after initialization, the larger the influence of the hydrologic environment on the water level at the moment is, the lower the weight is, the more easily and correctly classified the samples at the moment are, and the robustness and accuracy of the Adaboost algorithm are improved.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. And the foregoing description has been directed to specific embodiments of this specification. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and the same or similar parts of each embodiment are referred to each other, and each embodiment mainly describes differences from other embodiments.
The above embodiments are only for illustrating the technical solution of the present application, and are not limiting; the technical solutions described in the foregoing embodiments are modified or some of the technical features are replaced equivalently, so that the essence of the corresponding technical solutions does not deviate from the scope of the technical solutions of the embodiments of the present application, and all the technical solutions are included in the protection scope of the present application.

Claims (10)

1. The groundwater level restoration prediction method is characterized by comprising the following steps of:
the method comprises the steps of forming a time sequence matrix by collecting related data of each period affecting the change of the underground water level;
acquiring a spectrum curve of each row of data in a time sequence matrix; for each row of data of the time sequence matrix, obtaining a periodic hysteresis relation of the data according to a distribution relation between the data and a frequency spectrum curve of the water level height data; calculating partial correlation coefficients among three data of groundwater flow, flow speed and flow direction in the time sequence matrix; obtaining an environmental change factor according to the periodic hysteresis relationship of groundwater flow, flow speed and flow direction data in the time sequence matrix and the partial correlation coefficient;
obtaining a hydrologic environment influence factor according to the environment change factor, the periodic hysteresis of each row of data of the time sequence matrix and the distribution condition of each row of data; according to the hydrologic environmental impact factors, improving the initial weight of each height data of the underground water in the Adaboost algorithm; and training the weak classifier according to the initial weight of each height data in the underground water to output a prediction result of the water level height sample.
2. A groundwater level restoration prediction method according to claim 1, wherein the forming a time series matrix of the collected data of each period affecting the groundwater level change comprises:
collecting relevant data affecting groundwater level changes includes: rainfall, groundwater temperature, groundwater level height, groundwater flow speed, groundwater flow direction, groundwater ion concentration data, and groundwater pH data;
for each period, each data acquired in the period is formed into each row vector of the time series matrix.
3. The method of claim 2, wherein the obtaining the spectral curve of each row of data in the time series matrix comprises:
and carrying out frequency domain transformation on each data of the time sequence matrix by adopting a fast Fourier transformation algorithm, and obtaining a frequency spectrum curve corresponding to each data.
4. A groundwater level restoration prediction method according to claim 3, wherein the obtaining the periodic hysteresis relationship of the data according to the distribution relationship between the data and the spectral curve of the water level height data comprises:
obtaining the maximum amplitude value of a frequency spectrum curve of the data and the water level height data, and calculating the frequency corresponding to the maximum amplitude value of the data and the water level height data;
obtaining the maximum value of the frequency of the data and the water level height data;
calculating the absolute value of the difference value of the frequency between the data and the water level height data, and taking the ratio of the maximum amplitude value of the data and the water level height data as the true number of a logarithmic function taking a natural constant as a base number;
dividing the product of the absolute value of the calculation result of the logarithmic function and the absolute value of the difference by the maximum value to obtain a periodic hysteresis relation of data; wherein the cycle hysteresis relationship of the data is set to 0 when the data and the water level height data are unchanged in the cycle.
5. The groundwater level restoration prediction method according to claim 2, wherein calculating the partial correlation coefficient between three data of groundwater flow, flow speed and flow direction in the time series matrix comprises:
for each row of data of groundwater flow, flow speed and flow direction in the time sequence matrix, controlling the groundwater flow data into a constant, and analyzing the net correlation between the groundwater flow speed and the groundwater flow direction data to obtain a partial correlation coefficient of the groundwater flow data;
controlling the groundwater flow speed data to be constant, and analyzing the net correlation between the groundwater flow speed data and the groundwater flow speed data to obtain the partial correlation coefficient of the groundwater flow speed data;
and controlling the groundwater flow data to be constant, and analyzing the net correlation between the groundwater flow and the groundwater flow data to obtain the partial correlation coefficient of the groundwater flow data.
6. The method of claim 1, wherein the obtaining the environmental change factor according to the periodic hysteresis relationship and the partial correlation coefficient of the groundwater flow, the flow speed, and the flow direction data in the time series matrix comprises:
respectively taking the groundwater flow, the flow speed and the flow direction data in the time sequence matrix as each target line data, and respectively calculating the partial correlation coefficient mean value between each target line data and other two lines data;
taking the opposite number of the cycle hysteresis of the target line data as an index of an index function taking a natural constant as a base, calculating the product of the calculation result of the index function and the absolute value of the mean value of the partial correlation coefficient, and taking the sum of the product of the groundwater flow, the flow speed and the flow direction data as an environment change factor.
7. The method of claim 1, wherein the obtaining the hydrographic environmental impact factor according to the environmental change factor, the periodic hysteresis of each line of data of the time series matrix, and the distribution of each line of data comprises:
for each row of data of the time sequence matrix, obtaining the maximum value and the minimum value of the data;
taking the opposite number of the periodic hysteresis of the data as an index of an exponential function based on a natural constant, calculating the product of the difference result of the maximum value and the minimum value and the calculation result of the exponential function, and taking the product of the average value of the products of all the data of the time sequence matrix and an environmental change factor as a hydrographic environmental influence factor.
8. The groundwater level restoration prediction method according to claim 7, wherein the improving the initial weight of each height data of groundwater in the Adaboost algorithm according to the hydrographic environmental impact factor comprises:
for each collected height data of the underground water, obtaining a ratio of the height data to the height data adjacent to the previous height data, taking the product of the opposite number of the ratio and the hydrologic environment influence factor as an index of an exponential function based on a natural constant, and taking the calculation result of the exponential function as the initial weight of the underground water height data in an Adaboost algorithm.
9. The groundwater level restoration prediction method according to claim 8, wherein training the weak classifier based on the initial weight of each height data in groundwater outputs a predicted result of a water level height sample, comprising:
training the G weak classifiers to obtain strong classifiers by using initial weights of all height data in the underground water, wherein the input of the strong classifiers is the height data of the underground water, and the input of the strong classifiers is the prediction result of the height data of the underground water; wherein G is the number of preset weak classifiers.
10. A groundwater level restoration prediction system comprising a memory, a processor and a computer program stored in the memory and running on the processor, characterized in that the processor implements the steps of the method according to any one of claims 1-9 when executing the computer program.
CN202311576343.8A 2023-11-24 2023-11-24 Groundwater level restoration prediction method and system Active CN117291349B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311576343.8A CN117291349B (en) 2023-11-24 2023-11-24 Groundwater level restoration prediction method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311576343.8A CN117291349B (en) 2023-11-24 2023-11-24 Groundwater level restoration prediction method and system

Publications (2)

Publication Number Publication Date
CN117291349A true CN117291349A (en) 2023-12-26
CN117291349B CN117291349B (en) 2024-03-19

Family

ID=89244739

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311576343.8A Active CN117291349B (en) 2023-11-24 2023-11-24 Groundwater level restoration prediction method and system

Country Status (1)

Country Link
CN (1) CN117291349B (en)

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH04152920A (en) * 1990-10-16 1992-05-26 Sekisui Chem Co Ltd Device for monitoring bathtub water level
CN104406623A (en) * 2014-07-23 2015-03-11 青岛理工大学 Side slope dynamic stability coefficient determination method based on underground water level and displacement monitoring
CN107704973A (en) * 2017-10-31 2018-02-16 武汉理工大学 Water level prediction method based on neutral net Yu local Kalman filtering mixed model
WO2019224739A1 (en) * 2018-05-25 2019-11-28 University Of Johannesburg System and method for real time prediction of water level and hazard level of a dam
CN110703360A (en) * 2019-10-16 2020-01-17 西北大学 Three-dimensional effect model for landslide prediction based on rainfall intensity and threshold value
CN111310981A (en) * 2020-01-20 2020-06-19 浙江工业大学 Reservoir water level trend prediction method based on time series
CN113240274A (en) * 2021-05-13 2021-08-10 水利部水利水电规划设计总院 Reasonable ecological water level-based method for evaluating exploitable amount of underground water in plain area
US20220321394A1 (en) * 2021-03-31 2022-10-06 Equifax Inc. Techniques for prediction models using time series data
GB202213018D0 (en) * 2021-10-08 2022-10-19 Univ Hohai Basin similarity classification method and device
CN115560721A (en) * 2022-09-23 2023-01-03 福州大学 Rockfill concrete dam deformation prediction method based on environment factor hysteresis analysis
CN117005526A (en) * 2023-10-07 2023-11-07 济宁御龙源水务有限公司 Road drainage device for municipal works

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH04152920A (en) * 1990-10-16 1992-05-26 Sekisui Chem Co Ltd Device for monitoring bathtub water level
CN104406623A (en) * 2014-07-23 2015-03-11 青岛理工大学 Side slope dynamic stability coefficient determination method based on underground water level and displacement monitoring
CN107704973A (en) * 2017-10-31 2018-02-16 武汉理工大学 Water level prediction method based on neutral net Yu local Kalman filtering mixed model
WO2019224739A1 (en) * 2018-05-25 2019-11-28 University Of Johannesburg System and method for real time prediction of water level and hazard level of a dam
CN110703360A (en) * 2019-10-16 2020-01-17 西北大学 Three-dimensional effect model for landslide prediction based on rainfall intensity and threshold value
CN111310981A (en) * 2020-01-20 2020-06-19 浙江工业大学 Reservoir water level trend prediction method based on time series
US20220321394A1 (en) * 2021-03-31 2022-10-06 Equifax Inc. Techniques for prediction models using time series data
CN113240274A (en) * 2021-05-13 2021-08-10 水利部水利水电规划设计总院 Reasonable ecological water level-based method for evaluating exploitable amount of underground water in plain area
GB202213018D0 (en) * 2021-10-08 2022-10-19 Univ Hohai Basin similarity classification method and device
CN115560721A (en) * 2022-09-23 2023-01-03 福州大学 Rockfill concrete dam deformation prediction method based on environment factor hysteresis analysis
CN117005526A (en) * 2023-10-07 2023-11-07 济宁御龙源水务有限公司 Road drainage device for municipal works

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
LIU, GD等: "Hysteresis of Dam Slope Safety Factor under Water Level Fluctuations Based on the LEM Coupled with FEM Method", 《CMES-COMPUTER MODELING IN ENGINEERING & SCIENCES》, vol. 133, no. 2, pages 351 - 375 *
余世鹏;杨劲松;刘广明;: "基于离散小波变换与模糊神经算法的河口日水位预测", 水力发电学报, no. 06, pages 43 - 49 *
刘杨等: "时间序列模型在查哈阳农场降雨量预测中应用", 《黑龙江水利科技》, vol. 40, no. 02, pages 5 - 8 *
朱鹏涛等: "海原甘盐池地震观测井水位动态特征分析", 《地震工程学报》, vol. 44, no. 06, pages 1489 - 1496 *
王智磊;孙红月;刘永莉;尚岳全;: "降雨与边坡地下水位关系的时间序列分析", 浙江大学学报(工学版), no. 07, pages 166 - 172 *

Also Published As

Publication number Publication date
CN117291349B (en) 2024-03-19

Similar Documents

Publication Publication Date Title
CN106599571B (en) Watershed hydrological model stage calibration method considering both flow and evaporation
CN110852515B (en) Water quality index prediction method based on mixed long-time and short-time memory neural network
Zhang et al. A temporal LASSO regression model for the emergency forecasting of the suspended sediment concentrations in coastal oceans: Accuracy and interpretability
CN105981025A (en) Actually-measured marine environment data assimilation method based on sequence recursive filtering three-dimensional variation
CN110109080B (en) Weak signal detection method based on IA-SVM model
CN105137498A (en) Underground target detection and recognition system and method based on feature fusion
CN110309608B (en) Blast furnace molten iron silicon content forecasting method aiming at time lag uncertain information
CN101968957A (en) Voice detection method under noise condition
CN115099500B (en) Water level prediction method based on weight correction and DRSN-LSTM model
CN103336305B (en) A kind of method dividing Sandstone Gas Reservoir high water cut based on gray theory
CN104820993A (en) Underwater weak target tracking method combining particle filtering with track before detect
CN104237853B (en) A kind of for the particle filter method of trace point mark sequence before multi frame detection
Zhang et al. First arrival picking of microseismic signals based on nested U-Net and Wasserstein Generative Adversarial Network
Guo et al. Drought trigger thresholds for different levels of vegetation loss in China and their dynamics
CN115423163A (en) Method and device for predicting short-term flood events of drainage basin and terminal equipment
CN110363349A (en) A kind of LSTM neural network hydrologic(al) prognosis method and system based on ASCS
Azari et al. Evaluation of machine learning methods application in temperature prediction
CN116090625A (en) Coastal city flood rapid prediction method based on LightGBM and hydrological hydrodynamic model
CN112818512B (en) Storm mode-based maximum single-wave height Hmax calculation method
CN110907393A (en) Method and device for detecting saline-alkali stress degree of plants
CN113822201B (en) Deep learning method for underwater object shape recognition based on flow field velocity component time course
CN104268408A (en) Energy consumption data macro-forecast method based on wavelet coefficient ARMA model
CN113642484A (en) Magnetotelluric signal noise suppression method and system based on BP neural network
CN108647425A (en) K-means high flow or low flow time forecasting procedures based on particle group optimizing
CN113792685A (en) Microseism event detection method based on multi-scale convolution neural network

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant