Energy resource development area ecological environment comprehensive evaluation method

Abstract

The application discloses a kind of energy resource development regional ecological environment comprehensive evaluation method, and is applicable to ecological environment field.Its method includes: in view of the characteristics of regional energy resource development, from energy resource development, pressure, state and response four aspects extraction ecological environment evaluation index, build DPSR model, to establish energy resource development regional ecological environment evaluation index system;Data are collected and processed, and regional characteristic index system data set is established;Design particle swarm optimization projection pursuit regional ecological environment evaluation model, quantitative analysis regional ecological environment quality;Using spatial autocorrelation analysis regional ecological environment quality spatial aggregation and dispersion characteristics;Based on geographic detector, the influence mechanism of each evaluation index on the spatial distribution characteristics of ecological environment quality is explored, and the above-mentioned comprehensive evaluation of target regional ecological environment is realized.The ecological environment condition of energy resource development region can be scientifically and accurately evaluated and analyzed, and effective technical support is provided for ecological civilization construction.

Description

TECHNICAL FIELD

[0001] The present application relates to a kind of energy resource development area ecological environment comprehensive evaluation method, applicable to ecological environment field. BACKGROUND

[0002] With the rapid advancement of industrialization and urbanization, the overall situation of resources and environment is becoming increasingly severe, however, energy resource development region is mainly distributed in desert, gobi and grassland, and the ecological environment is fragile, the self-repairing ability of the region is poor, the ecological function and energy resource development and construction are in conflict, and the long-term cumulative effect may endanger the regional ecological safety and affect the quality of regional ecological environment. Therefore, it is urgent to carry out ecological environment comprehensive evaluation under the background of energy resource development, and explore the social economy, energy resource development and other factors that cause the deep change of ecological environment quality.

[0003] However, there are few studies on ecological environment evaluation in energy resource development region, and there is no standardized and standardized ecological environment evaluation technical system. Including lack of systematic ecological environment evaluation index system of energy resource development region, and no objective, scientific and reasonable ecological environment comprehensive evaluation and analysis model is constructed for energy resource development region. In view of the above problems, the present application is based on the background of energy resource development, and a regional ecological environment impact comprehensive evaluation model is constructed, which can scientifically evaluate the quality of regional ecological environment and reveal its time sequence change rule and spatial differentiation characteristics, so as to provide scientific basis for establishing regional resource environment carrying capacity early warning mechanism and improving regional ecological environment governance level.

[0004] Prior art includes:

[0005] Patent No. CN202110842725.5 discloses a kind of mine geological environment comprehensive evaluation method, its method is as follows: step one, the data acquisition module is used to collect the quality and quantity characteristics of mine environment information;Step two, the collected data is digitally processed by using data processing module;Step three, through data cleaning, supplement, based on principal component analysis algorithm, realize the dimension reduction of data, extract evaluation factor;Step four, based on the factor data modeling output by analytic hierarchy process and data analysis module, the relationship between influence factor and ecological environment quality evaluation value is established;Step five, using the determined weight value, calculate the comprehensive evaluation value of different grade mine environment quality. The method only considers the ecological environment evaluation of mine, without considering the ecological environment of other energy resource development region such as new energy development, and the analytic hierarchy process and principal component analysis method used in the method are easily disturbed by subjective conditions in evaluation, and sometimes not sensitive enough to ecological environment, so that objective and accurate ecological environment evaluation results cannot be obtained.

[0006] The patent No. CN202111263504.9 discloses a method for realizing a comprehensive evaluation index of a mine area ecological environment remote sensing, which comprises the following steps: A, collecting time series data of the research area, and constructing a ground cover type classifier; B, respectively performing remote sensing inversion on four categories of research areas according to selected characteristic indexes; C, performing parameter standardization processing through a maximum-minimum normalization model; D, extracting principal components according to the standardized data of step C, and obtaining a first principal component as an evaluation index CMEI result through pixel-by-pixel traversal. The method performs mine area ecological environment evaluation based on remote sensing data, which can perform objective long-time series analysis through satellite observation data, but cannot take into account the influence of social and economic background on the ground, and is affected by the resolution and accuracy of remote sensing data, and the available research scale is also affected. Moreover, the ecological environment status of other energy resource development areas such as new energy development is not considered. SUMMARY

[0007] Technical problem: In view of the deficiencies of the prior art, a comprehensive evaluation method for the ecological environment of an energy resource development area is provided, which comprehensively considers the influence of natural environment and social and economic factors on the quality of regional ecological environment under the background of energy resource development, improves the existing PSR model, introduces energy resource development as an influence factor of the PSR model, and proposes a DPSR model considering energy resource development. Based on the constructed DPSR model, a comprehensive evaluation index system capable of accurately representing the ecological environment status of the energy resource development area is established, and based on this, a particle swarm algorithm optimized projection pursuit regional ecological environment evaluation model is designed, which can scientifically evaluate the quality of regional ecological environment and reveal the time series variation law and spatial differentiation characteristics, which is helpful to coordinate the relationship between economic development and ecological environment, and provides a scientific basis for improving the level of regional ecological environment governance.

[0008] In order to achieve the above technical purpose, the present application discloses a comprehensive evaluation method for the ecological environment of an energy resource development area, characterized in that: the fossil energy and new energy development area and its adjacent areas are regarded as energy resource development areas, and the ecological environment of the area is comprehensively evaluated by considering the specific ecological environment factors of the energy resource development area, which comprises the following steps:

[0009] Step 1, comprehensively evaluating the ecological environment status of the energy resource development area, extracting ecological environment evaluation indexes from four aspects of energy resource development D, pressure P, state S and response R, and constructing a DPSR model to establish an ecological environment evaluation index system of the energy resource development area;

[0010] Step 2, collecting data of the evaluation index system, and after data cleaning, format conversion and boundary clipping processing, establishing a grid layer data set of each evaluation index;

[0011] Step 3, design particle swarm algorithm to optimize the projection pursuit model, build regional ecological environment evaluation model, take the attribute value of the grid unit in the obtained each evaluation index grid layer dataset as the input, quantitatively analyze the ecological environment quality, meanwhile, the best natural breaking method is used to qualitatively grade the ecological environment quality, so as to obtain the grading evaluation result of the ecological environment quality;

[0012] Step 4, the spatial autocorrelation technology is used to analyze the obtained qualitative ecological environment quality, and the spatial aggregation characteristics and dispersion degree of the regional ecological environment quality are analyzed based on Moran's I index;

[0013] Step 5, based on the geographic detector, the influence mechanism of each evaluation index on the spatial distribution characteristics of the regional ecological environment quality is analyzed, and the two-factor interaction of each evaluation index is detected, and the correlation between each evaluation index and the spatial distribution characteristics of the ecological environment quality is obtained;

[0014] Step 6, the comprehensive spatial distribution characteristic analysis result and the correlation analysis result between each evaluation index and the ecological environment quality are combined, and the comprehensive evaluation of the target regional ecological environment is completed.

[0015] Further, the DPSR model is designed for the improvement of the existing PSR model, including energy resource development D, pressure P, state S and response R four parts, each part is connected through the causal relationship: energy resource development D drives pressure P, pressure P influences state S, state S promotes response R, response R feedback to energy resource development D, pressure P and state S; four parts of indexes can completely reflect the comprehensive situation of energy resource development regional ecological environment, reveal the pressure caused by energy resource development to regional ecological environment, so as to cause the state change of regional ecological environment and natural resources, finally promote a series of corresponding response measures to ecological environment state, ecological environment pressure and energy resource development state; DPSR model covers human activities, social economy, natural resources, environment and other elements, can reflect the mutual influence relationship among them, to show the threat of human activities and social economy to regional natural resources and ecological environment under the driving of energy resource development, also can show the feedback of human activities and its environmental improvement input to reality through response index.

[0016] Further, the collected index system data first needs to be rasterized by the GIS method, and then the raster unit size is unified, each index is corresponded to a raster layer, and the attribute value of each raster unit in each raster layer corresponds to the index data; data cleaning includes removing duplicate data, filling missing data, and processing abnormal data, wherein when filling the missing data in the data set, interpolation and average value method are used according to the data characteristics, and when processing the abnormal data in the data set, the abnormal data is removed or replaced with interpolation and average value data according to the data characteristics; when boundary clipping, the raster unit range of the energy resource development region is clipped based on the administrative boundary data.

[0017] Further, the projection pursuit model obtains the best index characteristic vector and the comprehensive evaluation value of the multi-dimensional data by projecting the high-dimensional data on a one-dimensional space, and the specific steps of constructing the projection pursuit model are as follows:

[0018] 4.1 Data standardization:

[0019] For the index positively correlated with the ecological environment:

[0020] X i = [x i -min(x i )] / [max(x i )-min(x i )] Formula (1)

[0021] For the index negatively correlated with the ecological environment:

[0022] Y i = [max(y i )-y i ] / [max(y i )-min(y i )] Formula (2)

[0023] Wherein, X i represents the standardized value of the positive correlation index, x i represents the initial value of the positive correlation index, Y i represents the standardized value of the negative correlation index, y i represents the initial value of the negative correlation index, and i represents the sample data sequence.

[0024] 4.2 Projection objective function construction:

[0025] Let the sample set be {x(i,j)|i=1,2,..,n;j=1,2,..,m}, where m is the number of evaluation indexes, n is the number of samples, and the one-dimensional projection value Vi of m-dimensional data along the direction c={c(1),c(2),c(3),…,c(m)} is represented as:

[0026]

[0027] where c j represents the projection direction vector of the jth dimension data;

[0028] To meet the two requirements of maximum local projection point aggregation and overall projection dispersion as much as possible, the projection objective function Q(c) is established:

[0029] Q(c) = S(c) x D(c) Equation (4)

[0030]

[0031]

[0032] where S(c) is the inter-class distance, D(c) is the intra-class density, E(V i ) is the average value of {V i |i = 1, 2, …, n}, r ij is the inter-sample distance, r ij = (V i -V j ), R is the window radius of local density, f(R-r ij ) is a step function, where R is greater than r ij , f(R-r ij ) is equal to 1, otherwise it is equal to 0. D(c) represents the aggregation level of the projection point; the greater the value of D(c), the more the points are aggregated;

[0033] 4.3 Projection objective function optimization:

[0034] Further optimization of the projection objective function: the change of the projection objective function Q(c) is determined by the projection direction c, and different projection directions reflect different data structure characteristics, so it is necessary to find the best projection direction, which maximizes the projection objective function and its constraint condition s.t. is expressed as:

[0035] Max: Q(c) = S(c) x D(c) Equation (7)

[0036]

[0037] When the constraint conditions specified in Equation (7) and Equation (8) are met, the maximum projection objective function is the best projection direction vector.

[0038] Further, the particle swarm optimization algorithm is used to determine the best projection direction, and the specific steps are as follows:

[0039] 5.1 initialization, set the particle swarm size, particle dimension, maximum iteration number, inertia weight, particle position and velocity, learning factor, and randomly generate the position sequence and velocity sequence of the particles;

[0040] 5.2 calculate the fitness value of the particles, set the individual extreme value P best and the global extreme value g best ;

[0041] 5.3 judge whether the iteration number reaches the maximum iteration number or meets the error requirement, if yes, go to step 5.4, otherwise, update the particles, and then jump to step 5.2;

[0042] 5.4 obtain the final value, which is the global extreme value.

[0043] Put the global extreme value g best into formula (4) as the best projection direction c* to obtain the maximum projection objective function Q*(c), which satisfies the constraints specified in formula (7) and formula (8), and the direction vector corresponding to the value of the maximum projection objective function at this time is the best projection direction vector.

[0044] Further, use the components of the best projection direction vector as the weights of each evaluation index and multiply them by the corresponding evaluation index standardized value, and then sum the products to construct the regional ecological environment evaluation model EEQ, which is expressed as follows:

[0045]

[0046]

[0047] Wherein, u i represents the weight of each evaluation index; w i is the standardized value of each evaluation index; n is the number of evaluation indexes; c j represents the projection direction vector of the jth dimension data, which is the component of the best projection direction vector in the jth dimension, and is squared to make it positive.

[0048] Further, use GIS technology to perform spatial autocorrelation analysis on the obtained regional qualitative ecological environment quality:

[0049] Based on the qualitative classification evaluation results of ecological environment quality, the global Moran's I index is used to analyze the spatial clustering characteristics and dispersion degree I of the ecological environment quality space, I is between -1 and 1:

[0050] When I>0, it indicates that the ecological environment space shows a clustering trend;

[0051] When I<0, it indicates that the ecological environment space shows a discrete trend;

[0052] I=0 indicates that the ecological environment space has randomness;

[0053] Based on the analysis result of the above analysis step, the local Moran's I index is used to analyze the spatial distribution characteristics of the ecological environment quality aggregation and dispersion.

[0054] Further, based on the geographical detector, the specific method for analyzing the influence mechanism of each evaluation index on the spatial distribution characteristics of the regional ecological environment quality is as follows: each ecological environment evaluation index is taken as an explanatory variable, and the grid cell corresponding to the qualitative classification result of the regional ecological environment quality is taken as a dependent variable and introduced into the Geodetector, an analysis report is obtained by running, and the differentiation and factor detection results and the interaction detection results are extracted.

[0055] Compared with the prior art, the present application has the following beneficial effects:

[0056] The present application comprehensively considers the influence of natural environment and social economic factors on the regional ecological environment quality under the background of energy resource development, improves the existing PSR model, introduces the energy resource development as the PSR model influencing factor, and proposes a DPSR model considering the energy resource development. Based on the constructed DPSR model, an ecological environment comprehensive evaluation system is established, compared with the single biased perspective of selecting index factors in the traditional method, the natural environment and social economy and other multi-dimensional evaluation index factors are comprehensively considered, which can more accurately and intuitively represent the ecological environment condition under the background of energy resource development; at the same time, the existing technology is mainly used for ecological environment evaluation of the narrow energy resource development region, such as a certain type of mine, and the present application also includes the new energy development region in the evaluation category, so that the evaluation is more universal and applicable.

[0057] The present application uses the projection pursuit model for the ecological environment evaluation of the energy resource development region, effectively reduces the influence of human subjective factors, and the obtained evaluation result has objectivity and accuracy. Moreover, the model has strong generalization ability and can be used for ecological environment evaluation of energy resource development regions of different spatial scales. At the same time, the present application uses the particle swarm algorithm to optimize the projection pursuit model, the performance of the optimized model is excellent, the calculation efficiency is high, the evaluation precision is better than that of the traditional projection pursuit model, and the evaluation result is more consistent with the actual ecological environment condition.

[0058] Based on the ecological environment classification evaluation result, the present application analyzes the spatial aggregation characteristics of the ecological environment quality and the influence and action law of each evaluation index on the spatial distribution characteristics of the ecological environment quality based on the spatial autocorrelation and the geographical detector, which is helpful to coordinate the relationship between the economic development and the ecological environment of the energy resource development region, and provides a scientific basis for improving the regional ecological environment management level. BRIEF DESCRIPTION OF DRAWINGS

[0059] Figure 1 This is a schematic diagram of the process for the comprehensive evaluation of the ecological environment in energy resource development areas according to the present invention.

[0060] Figure 2 This is a schematic diagram of the DPSR model concept for regional ecological environment assessment in this invention.

[0061] Figure 3 This is a schematic diagram of the construction route of the particle swarm optimization projection tracking regional ecological environment assessment model in this invention.

[0062] Figure 4 This is a schematic diagram of the ecological environment spatial distribution characteristic analysis route in this invention.

[0063] Figure 5 This is a schematic diagram of the analysis route for ecological and environmental driving factors in this invention. Detailed Implementation

[0064] To more clearly illustrate the technical solutions of the embodiments of this application, the method of the present invention will be further described below with reference to the accompanying drawings and examples:

[0065] like Figure 1 The diagram illustrates a comprehensive ecological environment evaluation method for energy resource development areas according to the present invention. The steps are as follows: First, based on the characteristics of regional energy resource development, ecological environment evaluation indicators are extracted from four aspects: energy resource development (D), pressure (P), state (S), and response (R), and a DPSR model is constructed to establish an ecological environment evaluation indicator system for energy resource development areas. Data is collected and processed to establish a dataset of regional characteristic ecological environment indicator systems. Second, a particle swarm optimization algorithm is designed to optimize the projection pursuit regional ecological environment evaluation model, quantitatively analyzing the regional ecological environment quality. Next, spatial autocorrelation analysis is used to analyze the spatial clustering and dispersion characteristics of regional ecological environment quality, analyzing the commonalities and differences among different ecological environment quality levels. Finally, a geographic detector is used to explore the spatial heterogeneity of regional ecological environment quality, analyzing the influence mechanism of each evaluation indicator on the spatial distribution characteristics of ecological environment quality, thus achieving a comprehensive ecological environment evaluation of the target area.

[0066] This invention discloses a comprehensive ecological environment evaluation method for energy resource development areas, which mainly includes the following steps:

[0067] Step 1: Based on the current ecological and environmental status of the target area, and taking into account the four aspects of energy resource development (D), pressure (P), state (S), and response (R), construct a DPSR conceptual model for regional ecological and environmental assessment. Based on this, establish an ecological and environmental assessment index system for energy resource development areas. The indicators are extracted from geology, meteorology, socio-economic factors, geographic remote sensing, and energy resource development.

[0068] like Figure 2As shown, the DPSR conceptual model of the regional ecological environment evaluation of the present application: in the figure, divided into energy resource development, pressure, state and response 4 parts, each part is connected through the causal relationship arrow: energy resource development D drives pressure P, pressure P affects state S, state S promotes response R, response R feedback to energy resource development D, pressure P and state S, and is embodied through a number of indicators; 4 part indexes can completely embody the comprehensive condition of energy resource development regional ecological environment, reveal the pressure caused by energy resource development to regional ecological environment, so as to cause the state change of regional ecological environment, natural resources, and finally promote a series of corresponding response measures to ecological environment state, ecological environment pressure and energy resource development state; DPSR model covers human activities, social economy, natural resources, environment and other elements, and can reflect the mutual influence relationship among them, to show that under the driving of energy resource development, human activities and social economy bring threat to regional natural resources and ecological environment, and also can show the feedback of human activities and its environmental improvement input to reality through response index.

[0069] In the embodiment of the present application, based on the constructed DPSR model, 27 indexes in 4 levels are extracted as the basis for comprehensive evaluation and analysis, and the evaluation index system as shown in table 1 is preferably established.

[0070] Table 1 comprehensive evaluation index system of energy resource development regional ecological environment

[0071]

[0072]

[0073] Step 2, the index data required in the ecological environment comprehensive evaluation index system described in step 1 is collected, and after data cleaning, format conversion, boundary clipping and other treatments, an index system data set is established; wherein the index system data is obtained from the national Qinghai-Tibet Plateau scientific data center, the national glacier permafrost desert scientific data center, the atmospheric composition analysis website of the University of Washington in St. Louis, the resource and environment scientific data center of the Chinese Academy of Sciences, the NASA Earth Science Data website, the National Natural Resources Department, the National Meteorological Information Center, the Statistical Yearbook platform and the like; wherein the data cleaning includes removing duplicate data, filling missing data and processing abnormal data; wherein when filling the missing data in the data set, interpolation and average value methods are used according to the data characteristics; when processing the abnormal data in the data set, the abnormal data is removed or replaced with interpolation and average value data according to the data characteristics; wherein when format conversion, the index data is imported into ArcGIS software, and the raster conversion tool in the

ArcToolbox

ArcToolbox

[0074] When the data of the ecological environment comprehensive evaluation index system of the energy resource development area shown in Table 1 is rasterized, it can be divided into qualitative index data and quantitative index data. For the quantitative index data, the actual value is assigned to the raster cell; and for the qualitative data, the evaluation index is judged and assigned according to the grade division standard of the influence degree of the ecological environment, and then assigned to the corresponding raster cell.

[0075] Step 3, a particle swarm algorithm is designed to optimize the projection pursuit model, and a regional ecological environment evaluation model is constructed, the attribute value of the raster cell in the obtained raster layer data set of each evaluation index is taken as input, the ecological environment quality is quantitatively analyzed, and the best natural breaking method is used to qualitatively classify the ecological environment quality.

[0076] As shown in Figure 3 , the particle swarm optimization projection pursuit regional ecological environment evaluation model construction process of the application is: the projection pursuit model obtains the best index feature vector and comprehensive evaluation value of multi-dimensional data by projecting high-dimensional data on a one-dimensional space, and the construction of the model needs three steps, namely data standardization, projection objective function construction and projection objective function optimization; the specific steps are as follows:

[0077] 1) Data standardization

[0078] For the index positively correlated with the ecological environment:

[0079] X i = [x i -min(x i )] / [max(x i )-min(x i )] Formula (1)

[0080] For indicators with negative correlation:

[0081] X i = [max(x i )-x i ] / [max(x i )-min(x i )] Formula (2)

[0082] where X i represents the standardized value of the positive correlation indicator, x i represents the initial value of the positive correlation indicator, Y i represents the standardized value of the negative correlation indicator, y i represents the initial value of the negative correlation indicator, and i represents the sample data sequence.

[0083] 2) Projection target function construction

[0084] The sample set is {x(i,j)|i=1,2,..,n;j=1,2,..,m}, where m is the number of evaluation indicators, n is the number of samples, and the one-dimensional projection (V i ) of m-dimensional data along direction c={c(1),c(2),c(3),…,c(m)} is represented as:

[0085]

[0086] where c j represents the projection direction vector of the jth-dimensional data.

[0087] To meet the two requirements of maximum local projection point aggregation and overall projection dispersion as much as possible, the projection target function is established:

[0088] Q(c)=S(c)xD(c) Formula (4)

[0089]

[0090]

[0091] where S(c) is the inter-class distance, D(c) is the intra-class density, E(V i ) is the average value of {V i |i=1,2,…,n}, and r ij is the distance between samples.ij = (V i - V j ), R is the window radius of local density, f(R-r ij ) is a step function, where R is greater than r ij , f(R-r ij ) is equal to 1, otherwise equal to 0. D(c) represents the aggregation level of the projection points; the greater the value of D(c), the more concentrated the points.

[0092] 3) Projection objective function optimization

[0093] The initially constructed projection objective function can be further optimized. The change of the projection objective function Q(c) is determined by the projection direction c. Different projection directions can reflect different data structure characteristics, so it is necessary to find the optimal projection direction. The maximum projection objective function and its constraint conditions are expressed as:

[0094] Max: Q(c) = S(c) x D(c) Equation (7)

[0095]

[0096] When the constraint conditions specified in equation (7) and equation (8) are satisfied, the maximum projection objective function is the optimal projection direction vector.

[0097] Considering that the calculation of the optimal projection direction is a complex nonlinear optimization problem, a particle swarm optimization algorithm is used to determine the optimal projection direction, and the specific steps are as follows:

[0098] 1) Initialization, set the particle swarm size n, particle dimension D, maximum iteration number maxgen, inertia weight ω, particle position and velocity, learning factor, and randomly generate the position sequence Present and velocity sequence V of the particles; including:

[0099] a. Determine the particle swarm size n:

[0100] The particle swarm size refers to the number of all individuals in the established particle population, and its value size affects the algorithm operation complexity. The selection of the population size needs to be balanced between accuracy, stability and running time. Here, according to the existing research experience and combined with the characteristics of variable data, the population size is set to n = 50, in order to balance the model running time and accuracy requirements;

[0101] b. Determine the particle dimension D:

[0102] The particle dimension is determined by the number of independent variables in the projection objective function, which corresponds to the number of evaluation indexes.

[0103] c. Initialize the particle position and velocity:

[0104] The initial velocity of the particle is determined by the maximum velocity V max , which is the maximum limit of the velocity value taken between the current position and the optimal position. The initial velocity is taken in the range [-V max , V max ], and the maximum velocity V max is taken in the range of 10%-20% of the corresponding dimension change interval of the particle. The initial position of the particle is set in the domain of the independent variable.

[0105] d. Define the inertia weight:

[0106] Based on the principle of ensuring optimal global search performance at the beginning of the algorithm and optimal local search performance at the later stage of the algorithm, a linear decreasing scheme LDIW is used to determine the value of the inertia weight ω:

[0107]

[0108] where ω s is the initial inertia weight; ω e is the final inertia weight; t represents the current evolution generation; t max is the maximum evolution generation.

[0109] e. Define the learning factor:

[0110] The learning factor is the acceleration factor when the particle moves, and the learning factor is set to 2.

[0111] f. Define the maximum number of iterations:

[0112] The maximum evolution generation is set in the range [50, 100], and the larger the value, the slower the convergence speed. This needs to be adjusted according to the actual situation during optimization. Here, the maximum number of iterations maxgen is set to 50.

[0113] 2) Calculate the fitness value of the particle, set P best and g best ; including:

[0114] a. Determine the fitness function:

[0115] The fitness function is used to calculate the fitness value of the particle, which is as follows:

[0116]

[0117] where F is the current particle fitness value; α is a random coefficient for adjusting the fitness value to be within a reasonable range; y i is the actual output data; is the expected output data; n is the number of training samples.

[0118] b.The optimal position searched by the ith particle in the entire D-dimensional solution space is called the individual extremum, denoted as:

[0119] P best i1 i2 iD , i = 1, 2, …, N Formula (11)

[0120] c.The optimal position searched by the entire particle swarm in the entire D-dimensional solution space is called the global extremum, denoted as:

[0121] g best g1 g2 gD Formula (12)

[0122] 3) Determine whether the iteration number reaches maxgen times or meets the error requirement. If yes, go to step 4), otherwise, update the particle according to formula (13) and (14), and then jump to step 2); including:

[0123] As long as the particle finds the individual extremum and the global extremum, its state can be updated by formula (13) and (14), that is, change its current speed and position:

[0124] v id (t+1) = ω * v id (t) + c1r1(p id -x id (t)) + c2r2(p gd -x id (t)) Formula (13)

[0125] x id (t+1) = x id (t) + v id (t+1) Formula (14)

[0126] Where ω is the inertia weight; c1, c2 are learning factors, whose value range is between (0, 2); v im is the speed of the particle; t represents the tth generation; r1, r2 are arbitrary numbers between (0, 1).

[0127] 4) Obtain the final value, which is g best .

[0128] The global extremum g best ​​​​​​The maximum projection target function Q*(c) is obtained by bringing the optimal projection direction c* into formula (4), that is, the constraint conditions defined by formula (7) and formula (8) are satisfied, and the direction vector corresponding to the maximum projection target function value at this time is the optimal projection direction vector.

[0129] The construction of the regional ecological environment model EEQ is to multiply the weight of each evaluation index by the corresponding evaluation index standardized value by using the optimal projection direction vector, and the sum of the products is obtained, which can be represented by the following formula:

[0130]

[0131]

[0132] Wherein, u i represents the weight of each evaluation index; w i is the standardized value of each evaluation index; n is the number of evaluation indexes; c j represents the projection direction vector of the jth dimension data, that is, the component of the optimal projection direction vector in the jth dimension, and the square processing is performed to make it a positive value.

[0133] Finally, the ecological environment quality is qualitatively graded by using the optimal natural breaking method, and is divided into five levels of excellent, good, medium, low and poor.

[0134] The construction of the above particle swarm optimization projection tracking regional ecological environment evaluation model is carried out in the Python environment.

[0135] Step 4, using ArcGIS software to perform spatial autocorrelation analysis on the above obtained qualitative ecological environment quality, and based on Moran's I index, the spatial aggregation characteristics and dispersion degree of the regional ecological environment quality are analyzed.

[0136] As Figure 4 shown, the ecological environment spatial distribution characteristic analysis route of the present application is: using ArcGIS software to perform spatial autocorrelation analysis on the obtained regional qualitative ecological environment quality, and the specific steps include:

[0137] 1) Global Moran's I index analysis

[0138] Running the

Spatial Autocorrelation (Morans I)

ArcToolbox

[0139]

[0140] Where I is the global autocorrelation index, n is the total number of elements, x i and x j are the ecological environment quality levels of spatial units i and j, respectively, is the average value of ecological environment quality, w ij is the spatial weight coefficient matrix; I is between -1 and 1, when I > 0, it indicates that the ecological environment space shows a clustering trend; when I < 0, it indicates that the ecological environment space shows a discrete trend; I = 0 indicates that the ecological environment space has randomness.

[0141] 2) Local Moran's I index analysis

[0142] Run the

Cluster and Outlier Analysis (Anselin Local Moran's I)

ArcToolbox

[0143]

[0144] Where I p is the local autocorrelation index, n is the total number of elements, x i and x j are the ecological environment quality levels of spatial units i and j, respectively, is the average value of ecological environment quality, w ij is the spatial weight coefficient matrix.

[0145] In the obtained analysis results, Not Significant indicates no significance; High-High Cluster indicates high-high aggregation of ecological environment; High-Low Outlier indicates high-low aggregation of ecological environment; Low-High Outlier indicates low-high aggregation of ecological environment; Low-Low Cluster indicates low-low aggregation of ecological environment. Finally, the obtained evaluation results are visualized.

[0146] Step 5, based on the geographical detector, analyze the driving mechanism of each evaluation index on the spatial distribution characteristics of regional ecological environment quality, and conduct factor interaction detection, analyze the interaction results of each evaluation index on ecological environment, and obtain the correlation between each evaluation index and ecological environment spatial distribution.

[0147] For example Figure 5As shown, the ecological environment driving factor analysis step is: using the geographic detector to analyze the influence law of each evaluation index on the spatial distribution characteristics of the ecological environment quality of the energy resource development region, and the specific method is: taking each ecological environment evaluation index as an explanatory variable, and taking the grid cell corresponding to the qualitative classification result of the regional ecological environment quality as a dependent variable to import Geodetector, running to obtain an analysis report, and extracting the differentiation and factor detection results and interaction detection results.

[0148] Step 6: combining the spatial distribution characteristic analysis result and the correlation analysis result between each evaluation index and the ecological environment quality, and completing the comprehensive evaluation of the ecological environment of the target region.

[0149] Although the embodiments of the present application have been shown and described, it can be understood by those of ordinary skill in the art that various changes, modifications, replacements and variations can be made to the embodiments without departing from the principles and spirits of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for comprehensive evaluation of ecological environment in an energy resource development area, characterized in that: The fossil energy and new energy development region and its adjacent areas are taken as the energy resource development region, and the ecological environment of the region is comprehensively evaluated by considering the specific ecological environment factors of the energy resource development region, which includes the following steps: Step 1, comprehensively evaluate the ecological environment status of the energy resource development region, extract the ecological environment evaluation indexes from the four aspects of energy resource development D, pressure P, state S and response R, and build a DPSR model to establish the ecological environment evaluation index system of the energy resource development region; Step 2, collect the evaluation index system data, and after data cleaning, format conversion and boundary clipping processing, establish the grid layer data set of each evaluation index; Step 3, design particle swarm optimization algorithm to optimize the projection pursuit model, construct the regional ecological environment evaluation model, take the attribute values of the grid cells in the obtained grid layer data set of each evaluation index as input, quantitatively analyze the ecological environment quality, and at the same time, use the best natural breaking method to qualitatively classify the ecological environment quality, so as to obtain the classification evaluation result of the ecological environment quality; Step 4, analyze the above obtained qualitative ecological environment quality by using spatial autocorrelation technology, and based on Moran's I index, analyze the spatial aggregation characteristics and dispersion degree of the regional ecological environment quality; Step 5, based on the geographic detector, analyze the influence mechanism of each evaluation index on the spatial distribution characteristics of the regional ecological environment quality, and carry out two-factor interaction detection for each evaluation index, and obtain the correlation between each evaluation index and the spatial distribution characteristics of the ecological environment quality; Step 6, comprehensively analyze the spatial distribution characteristics and the correlation between each evaluation index and the ecological environment quality, and complete the comprehensive evaluation of the ecological environment of the target region.

2. The method according to claim 1, characterized in that: The existing PSR model is improved to design the DPSR model, which includes four parts of energy resource development D, pressure P, state S and response R. Each part is connected through causal relationship: energy resource development D drives pressure P, pressure P affects state S, state S promotes response R, and response R feedbacks to energy resource development D, pressure P and state S; The four parts of indexes can completely reflect the comprehensive situation of the ecological environment of the energy resource development region, reveal the pressure on the regional ecological environment caused by the energy resource development, so as to cause the change of the state of the regional ecological environment and natural resources, and finally promote a series of corresponding response measures to the ecological environment state, ecological environment pressure and energy resource development state; The DPSR model covers human activities, social economy, natural resources and environmental factors, and can reflect their mutual influence relationship, so as to show the threat of human activities and social economy to the regional natural resources and ecological environment under the driving of energy resource development, and also show the feedback of human activities and environmental improvement input to reality through response indexes.

3. The method according to claim 1, characterized in that: The collected index system data first needs to be rasterized by GIS method, and then the size of the raster unit is unified, each index corresponds to a raster layer, and the attribute value of each raster unit in each raster layer corresponds to the index data; Data cleaning includes removing duplicate data, filling missing data, and processing abnormal data. When filling the missing data in the data set, the interpolation and average value method is used according to the data characteristics. When processing abnormal data in the data set, it is removed or replaced with interpolated or average value data according to the data characteristics. When the boundary is cropped, the raster unit range of the energy resource development area is cropped based on the administrative boundary data.

4. The method according to claim 1, wherein, The projection pursuit model obtains the best index characteristic vector and comprehensive evaluation value of multi-dimensional data by projecting high-dimensional data on a one-dimensional space. The specific steps of constructing the projection pursuit model are as follows: 4.1 Data standardization: For the index positively correlated with the ecological environment: X i = [x i - min(x i )] / [max(x i ) - min(x i )] Formula (1), For the index negatively correlated with the ecological environment: Y i = [max(y i ) - y i ] / [max(y i ) - min(y i )] Equation (2), wherein X i represents a standardized value of a positive correlation index, x i represents an initial value of a positive correlation index, Y i represents a standardized value of a negative correlation index, y i represents an initial value of a negative correlation index, i represents a sample data sequence; 4.2 Projection objective function construction: Let the sample set be {x(i,j) | i = 1, 2,.., n; j = 1, 2,.., m}, where m is the number of evaluation indexes, n is the number of samples, and the one-dimensional projection value V of the m-dimensional data along the direction c = {c(1), c(2), c(3),.., c(m)} i is represented as: wherein c j represents the projection direction vector of the jth dimension data; To meet the two requirements of maximum local projection point aggregation and overall projection dispersion as much as possible, the projection objective function Q(c) is established: Q(c) = S(c) x D(c) Formula (4), where S(c) is the inter-class distance, D(c) is the intra-class density, E(V i ) is the average value of {V i |i = 1, 2, …, n}, r ij is the inter-sample distance, r ij = (V i - V j ), R is the window radius of local density, f(R-r ij ) is a step function, where R is greater than r ij , f(R-r ij ) is equal to 1, otherwise equal to 0; D(c) represents the aggregation level of the projected point; the greater the value of D(c), the more the points are aggregated; 4.3 Projection objective function optimization: Further optimize the projection objective function: The change of the projection objective function Q(c) is determined by the projection direction c. Different projection directions reflect different data structure characteristics, so the best projection direction needs to be found. The maximum projection objective function and its constraint condition s.t. are expressed as: Max: Q(c) = S(c) x D(c) Formula (7), When the constraint conditions specified in formula (7) and formula (8) are met, the maximum projection objective function is the best projection direction vector.

5. The method according to claim 4, wherein, The particle swarm optimization algorithm is used to determine the best projection direction. The specific steps are as follows: 5.1 Initialization, set particle swarm size, particle dimension, maximum iteration number, inertia weight, particle position and speed, learning factor, and randomly generate particle position sequence and speed sequence; 5.2 Calculate the fitness value of the particle, set the individual extreme value P best and the global extreme value g best ; 5.3 Determine whether the iteration number reaches the maximum iteration number or meets the error requirement. If yes, go to step 5.4, otherwise, update the particle and then jump to step 5.2; 5.4 Get the final value, which is the global extreme value; The global extreme value g best The maximum projection objective function Q*(c) is obtained by bringing the optimal projection direction c* into formula (4), that is, the constraints defined by formula (7) and formula (8) are satisfied, and the direction vector corresponding to the value of the maximum projection objective function at this time is the optimal projection direction vector.

6. The method according to claim 5, wherein, Use the components of the best projection direction vector as the weights of each evaluation index and multiply them by the corresponding evaluation index standardization value. Then sum the products to construct the regional ecological environment evaluation model EEQ, which is expressed as follows: wherein, u i w represents the weight of each evaluation index; w i is the normalized value of each evaluation index; n is the number of evaluation indexes; c j represents the projection direction vector of the jth dimension data, that is, the component of the best projection direction vector in the jth dimension, and is squared to make it a positive value.

7. The method according to claim 1, wherein, Use GIS technology to perform spatial autocorrelation analysis on the obtained regional qualitative ecological environment quality: Based on the qualitative classification evaluation results of ecological environment quality, the global Moran's I index is used to analyze the spatial clustering characteristics and dispersion degree I of ecological environment quality space, I is between -1 and 1: When I > 0, the ecological environment space shows a clustering trend; When I < 0, the ecological environment space shows a dispersion trend; I = 0, the ecological environment space has randomness; Based on the analysis results of the above analysis steps, the local Moran's I index is used to analyze the spatial distribution characteristics of ecological environment quality aggregation and dispersion.

8. The method according to claim 1, wherein, The specific method for analyzing the influence mechanism of each evaluation index on the spatial distribution characteristics of regional ecological environment quality based on the geographic detector is as follows: taking each ecological environment evaluation index as an explanatory variable and taking the grid cell corresponding to the qualitative classification result of the regional ecological environment quality as a dependent variable, importing the geographic detector, running to obtain an analysis report, and extracting the differentiation and factor detection results and the interaction detection results.

Images