A multi-parameter environmental quality index calculation method

Abstract

The application discloses a multi-parameter environmental quality index calculation method, obtains multi-parameters collected by an environmental monitoring system to obtain an original parameter set, carries out anti-extreme value normalization processing on the original parameter set to obtain a normalized parameter set, inputs the normalized parameter set into a dynamic risk function model, determines a corresponding type of risk function and carries out risk mapping to obtain a risk score set, calculates weights by using an entropy weight method, combines a nonlinear fusion function containing a parameter interaction item to carry out fusion calculation, and obtains an environmental quality index EQI; and then based on historical data of the environmental quality index EQI, a quantile method is used to determine an environmental grade division threshold set, grade division is carried out, and an environmental grade result is obtained. The method provided by the application can improve comprehensive evaluation accuracy, enhance scene adaptability, and reduce the influence of data fluctuation on evaluation results.

Description

Technical Field

[0001] This application relates to the field of environmental quality monitoring and assessment technology, and in particular to a method for calculating a multi-parameter environmental quality index. Background Technology

[0002] With the increasing demands for ecological environment governance, industrial production environmental control, and refined management of facility agriculture, environmental monitoring technology is gradually evolving from single-indicator detection to multi-parameter comprehensive monitoring. In practical applications, it is typically necessary to continuously collect and analyze multiple environmental parameters such as particulate matter concentration, gas concentration, temperature, and humidity to reflect the environmental status of the target area. To facilitate unified evaluation, hierarchical management, and early warning of environmental conditions, related technologies often employ environmental quality indices to comprehensively characterize multiple environmental parameters, thereby providing quantitative basis for environmental monitoring and management.

[0003] In related technologies, the calculation of environmental quality indices is typically based on multiple parameters acquired by environmental monitoring systems. These parameters undergo data processing, followed by comprehensive calculations according to preset rules to obtain the corresponding environmental quality assessment results. While this method can achieve a comprehensive expression of multi-parameter environmental information to some extent, existing environmental quality assessment methods generally suffer from problems in practical applications, such as insufficient stability of comprehensive assessment results, weak adaptability to different scenarios, and low consistency between assessment results and actual environmental conditions. This is especially true when monitoring data is affected by sensor fluctuations, sudden environmental changes, or local anomalies, leading to significant deviations in environmental quality assessment results and impacting the accuracy and reliability of environmental quality assessments.

[0004] Therefore, in multi-parameter environmental quality assessment, the lack of accuracy in comprehensive evaluation, poor adaptability to different scenarios, and the significant impact of fluctuations in monitoring data on the evaluation results have become urgent problems that need to be solved. Summary of the Invention

[0005] This application provides a multi-parameter environmental quality index calculation method, aiming to solve the problems of insufficient comprehensive evaluation accuracy, poor adaptability to different scenarios, and significant impact of monitoring data fluctuations on evaluation results in existing multi-parameter environmental quality assessments.

[0006] This invention provides a method for calculating a multi-parameter environmental quality index, the method comprising: Obtain multiple parameters collected by the environmental monitoring system to obtain the raw parameter set; The original parameter set is subjected to anti-extreme value normalization processing to obtain a normalized parameter set, wherein the anti-extreme value normalization processing adopts at least one of reference value normalization, logarithmic compression or Box-Cox transformation; when reference value normalization is adopted, the reference value is set based on the environmental baseline. The normalized parameter set is input into the dynamic risk function model. Based on the type of each parameter in the normalized parameter set and the mathematical relationship between each parameter and risk, the corresponding type of risk function is determined. Risk mapping is performed on each parameter in the normalized parameter set to obtain a risk score set. Among them, the pollutant concentration parameter adopts a mapping method in which the risk score increases with the increase of concentration, and the temperature and humidity parameter adopts a reverse mapping method based on the suitable range, so that the risk score increases with the increase of the deviation from the suitable range. Based on the risk score set, the weights corresponding to each parameter are calculated using the entropy weight method to obtain a weight set. Based on the risk score set and the weight set, a nonlinear fusion function with parameter interaction terms is used for fusion calculation to obtain the Environmental Quality Index (EQI), wherein the nonlinear fusion function includes an interaction adjustment factor and a correlation coefficient between indicators. Based on historical data of the Environmental Quality Index (EQI), a set of thresholds for environmental classification was determined using the quantile method. Based on the set of environmental level classification thresholds, the Environmental Quality Index (EQI) is classified into levels to obtain the environmental level results.

[0007] Optionally, in the above scheme, the step of performing anti-extreme value normalization processing on the original parameter set to obtain a normalized parameter set includes: Acquire normal monitoring data from the environmental monitoring system within a preset historical time period, wherein the normal monitoring data is valid data that has not triggered a pollution warning. The reference value is determined by averaging the normal monitoring data. The original parameter set is normalized based on the reference value to obtain the normalized parameter set.

[0008] Optionally, in the above scheme, when performing risk mapping on pollutant concentration parameters in the normalized parameter set that are linearly positively correlated with risk, the method includes: Determine pollutant concentration parameters that are linearly positively correlated with risk from the set of normalized parameters; Based on the pollutant thresholds in environmental quality standards, determine the piecewise points of the piecewise linear function; Based on the segmentation points, construct a piecewise linear function corresponding to the pollutant concentration parameters; The pollutant concentration parameter is input into the piecewise linear function for piecewise mapping to obtain the risk score corresponding to the pollutant concentration parameter.

[0009] Optionally, in the above scheme, when performing risk mapping on pollutant concentration parameters in the normalized parameter set that are non-linearly positively correlated with risk, the method includes: Determine pollutant concentration parameters that are non-linearly positively correlated with risk from the set of normalized parameters; Determine the exponential function corresponding to the pollutant concentration parameter; Based on the pollutant characteristics of the pollutant concentration parameters, determine the function parameters of the exponential function; The pollutant concentration parameter is input into the exponential function for mapping to obtain the risk score corresponding to the pollutant concentration parameter.

[0010] Optionally, in the above scheme, when performing risk mapping on the temperature and humidity parameters in the normalized parameter set, the method includes: Determine the temperature and humidity parameters from the normalized parameter set; Obtain normal monitoring data within a preset historical time period; Based on the aforementioned normal monitoring data, a suitable range is determined; Based on the degree of deviation of the temperature and humidity parameters from the suitable range, the temperature and humidity parameters are reverse-mapped to obtain the risk score corresponding to the temperature and humidity parameters.

[0011] Optionally, in the above scheme, the step of calculating the weights corresponding to each parameter using the entropy weight method based on the risk score set to obtain a weight set includes: Obtain parameter monitoring samples within a preset sampling time period to obtain a sample set; Based on the sample set, the normalized proportion of each parameter in each sample is calculated to obtain the proportion set. The information entropy of each parameter is calculated based on the aforementioned proportion set to obtain the information entropy set; The weights corresponding to each parameter are calculated based on the information entropy set to obtain the weight set.

[0012] Optionally, in the above scheme, the step of performing fusion calculation based on the risk score set and the weight set using a nonlinear fusion function with parameter interaction terms to obtain the Environmental Quality Index (EQI) includes: Obtain historical monitoring data within a preset historical time period; Based on the historical monitoring data, the correlation coefficients between the parameters are calculated to obtain the correlation coefficients between the indicators. Based on the complexity of environmental pollution, determine the interaction adjustment factors; The risk score set, the weight set, the correlation coefficient between the indicators, and the interaction adjustment factor are input into the nonlinear fusion function for fusion calculation to obtain the environmental quality index EQI.

[0013] Optionally, in the above scheme, determining the set of environmental level classification thresholds based on historical Environmental Quality Index (EQI) data and using the quantile method includes: Set the threshold for basic environmental level classification; Obtain historical data for the Environmental Quality Index (EQI); Statistical analysis was performed on the historical data of the Environmental Quality Index (EQI) based on the quantile method to obtain the statistical analysis results. The basic environmental level classification thresholds are updated based on the statistical analysis results to obtain the set of environmental level classification thresholds.

[0014] Optionally, in the above scheme, the method further includes: Obtain the Environmental Quality Index (EQI) corresponding to multiple monitoring nodes to obtain the node index set; Obtain node attribute information for each monitoring node; Based on the node attribute information, determine the node weight corresponding to each monitoring node to obtain a set of node weights; The regional environmental quality index is obtained by weighting the set of node indices and the set of node weights.

[0015] Optionally, in the above scheme, the method further includes: The sliding window length is determined based on parameter fluctuations; Based on the sliding window length, the Environmental Quality Index (EQI) is smoothed by sliding window mean to obtain a smoothed EQI sequence. The rate of change is calculated based on the smoothed Environmental Quality Index (EQI) sequence to obtain the EQI change rate. The warning is triggered based on the rate of change of the Environmental Quality Index (EQI) and its duration, and the warning result is obtained. Based on the warning results, corresponding control recommendations are output.

[0016] Compared with the prior art, this application has at least the following beneficial effects: Based on further analysis and research of existing technical problems, this application recognizes the issues in multi-parameter environmental quality assessment, such as insufficient accuracy of comprehensive evaluation, poor adaptability to different scenarios, and significant impact of monitoring data fluctuations on evaluation results. This application obtains multiple parameters collected by the environmental monitoring system and forms a raw parameter set. First, the raw parameter set undergoes anti-extreme value normalization processing, thereby converting data with different dimensions, distribution characteristics, and potential abnormal fluctuations into a comparable normalized parameter set, thus reducing the interference of monitoring data fluctuations and extreme values ​​on subsequent evaluation results. Then, the normalized parameter set is input into a dynamic risk function model. Based on the type of each parameter and the mathematical relationship between each parameter and risk, the corresponding risk function is determined, and risk mapping is performed on each parameter to obtain a risk score set. Since different types of parameters are no longer processed using a single unified rule, but rather risk-characterized according to their respective risk change patterns, this approach can... This approach makes risk representation more consistent with actual environmental conditions. Subsequently, based on the risk score set, the entropy weight method is used to calculate the weights corresponding to each parameter, resulting in a weight set. This ensures that the weights are determined based on the actual data distribution of the parameters, rather than relying on fixed human experience, thereby improving the adaptability of the evaluation model under different application scenarios. Building upon this, a nonlinear fusion function containing parameter interaction terms is used to perform fusion calculation based on the risk score set and the weight set, yielding the Environmental Quality Index (EQI). Interaction adjustment factors and correlation coefficients between indicators are introduced during the fusion process, enabling the EQI to reflect not only the independent influence of each parameter but also the interaction relationships between them. Finally, based on historical EQI data, a set of environmental level classification thresholds is determined using the quantile method, and the EQI is classified accordingly to obtain environmental level results, ensuring that the evaluation results are consistent with historical environmental distribution characteristics. This application enables a unified quantitative evaluation of multi-parameter environmental conditions, improves the accuracy of comprehensive evaluation, enhances adaptability to different scenarios, and reduces the impact of monitoring data fluctuations on evaluation results. Attached Figure Description

[0017] Figure 1 A flowchart illustrating a multi-parameter environmental quality index calculation method provided in one embodiment of this application; Figure 2 A flowchart illustrating a multi-parameter environmental quality index calculation algorithm provided in one embodiment of this application; Figure 3 This is a schematic diagram of the regional EQI calculation process provided in one embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0019] In one embodiment, such as Figure 1 As shown, a multi-parameter environmental quality index calculation method is provided, including the following steps: Obtain multiple parameters collected by the environmental monitoring system to obtain the raw parameter set; The original parameter set is subjected to anti-extreme value normalization processing to obtain a normalized parameter set, wherein the anti-extreme value normalization processing adopts at least one of reference value normalization, logarithmic compression or Box-Cox transformation; when reference value normalization is adopted, the reference value is set based on the environmental baseline. The normalized parameter set is input into the dynamic risk function model. Based on the type of each parameter in the normalized parameter set and the mathematical relationship between each parameter and risk, the corresponding type of risk function is determined. Risk mapping is performed on each parameter in the normalized parameter set to obtain a risk score set. Among them, the pollutant concentration parameter adopts a mapping method in which the risk score increases with the increase of concentration, and the temperature and humidity parameter adopts a reverse mapping method based on the suitable range, so that the risk score increases with the increase of the deviation from the suitable range. Based on the risk score set, the weights corresponding to each parameter are calculated using the entropy weight method to obtain a weight set. Based on the risk score set and the weight set, a nonlinear fusion function with parameter interaction terms is used for fusion calculation to obtain the Environmental Quality Index (EQI), wherein the nonlinear fusion function includes an interaction adjustment factor and a correlation coefficient between indicators. Based on historical data of the Environmental Quality Index (EQI), a set of thresholds for environmental classification was determined using the quantile method. Based on the set of environmental level classification thresholds, the Environmental Quality Index (EQI) is classified into levels to obtain the environmental level results.

[0020] In some implementations, multiple parameters collected by the environmental monitoring system are first acquired to obtain a raw parameter set. These multiple parameters characterize the overall state of the environment to be assessed and may include pollutant concentration parameters and temperature and humidity parameters. The environmental monitoring system can be deployed in scenarios such as urban air monitoring, industrial plant environmental monitoring, or agricultural greenhouse environmental monitoring. It acquires the raw values ​​of each parameter using periodic sampling, continuous sampling, or event-triggered sampling methods, and combines the parameter values ​​at the same sampling time or within the same sampling period to form the raw parameter set.

[0021] After obtaining the original parameter set, an anti-extreme value normalization process is performed on the original parameter set to obtain a normalized parameter set. This anti-extreme value normalization process can employ at least one of the following: reference value normalization, logarithmic compression, or Box-Cox transformation. When reference value normalization is used, the reference value is set based on an environmental baseline. The environmental baseline can be understood as the historical statistical characteristics of the target monitoring scenario under normal conditions, rather than a pre-defined uniform constant. Reference value normalization allows the current sampled value to be relativized to the corresponding environmental baseline; logarithmic compression reduces the impact of sudden spikes on subsequent risk modeling; and Box-Cox transformation adjusts the parameter distribution to make it more suitable for subsequent calculations. Thus, parameters with different dimensions and fluctuation characteristics can be transformed into a normalized parameter set suitable for unified calculation.

[0022] In some implementations, the normalized parameter set is input into a dynamic risk function model. Based on the type of each parameter in the normalized parameter set and the mathematical relationship between each parameter and risk, a corresponding risk function is determined. This "corresponding risk function" is not fixed as a single function but is selected based on the risk variation pattern of the parameters. For pollutant concentration parameters, the risk score increases with increasing concentration; for temperature and humidity parameters, the risk score does not simply increase unidirectionally with an absolute increase in parameter value, but rather undergoes a reverse mapping with a suitable range as a reference. That is, the closer to the suitable range, the lower the risk score; the further away from the suitable range, the higher the risk score.

[0023] After determining the function type, risk mapping is performed on each parameter in the normalized parameter set to obtain a risk score set. Preferably, different parameters can be mapped to a unified risk scale, for example, outputting a risk score of 0-100, so as to facilitate subsequent unified fusion calculation. Thus, parameters with different physical meanings and dimensions can be converted into a comparable risk score set.

[0024] After obtaining the risk score set, the weights corresponding to each parameter are calculated using the entropy weight method, resulting in a weight set. The entropy weight method uses the information entropy of a parameter in the samples to reflect its discriminative ability. The smaller the information entropy, the more significantly the parameter represents differences in environmental quality, and the higher its corresponding weight; conversely, the larger the information entropy, the smaller the difference between samples, and the lower its corresponding weight. In this way, there is no need to repeatedly adjust the weights manually for different scenarios; a weight set adapted to the current scenario can be automatically generated based on monitoring data.

[0025] In some implementations, the Environmental Quality Index (EQI) is obtained by fusing the risk score set and the weight set using a nonlinear fusion function containing parameter interaction terms. The nonlinear fusion function includes an interaction moderating factor and an inter-indicator correlation coefficient. The inter-indicator correlation coefficient quantifies the synergistic or inhibitory effects between different parameters, while the interaction moderating factor characterizes the moderating effect of the scene's pollution complexity on the contribution of the interaction terms. By simultaneously incorporating single-parameter risk scores, weight information, and inter-parameter interaction relationships, an EQI that more closely reflects actual environmental conditions can be obtained.

[0026] After obtaining the Environmental Quality Index (EQI), a set of environmental level classification thresholds is determined based on historical EQI data and the quantile method. Then, the current EQI is classified according to this set of thresholds to obtain the environmental level results. The "historical EQI data" refers to a continuously accumulated set of EQI results over a historical period, and the "quantile method" refers to statistically analyzing threshold positions based on the distribution characteristics of the historical data to generate a set of environmental level classification thresholds suitable for the target scenario. This step converts the continuous EQI into discrete environmental level results for environmental status identification, alarm level determination, and management decision-making.

[0027] This embodiment utilizes a chained processing flow—including "anti-extreme value normalization, dynamic risk mapping, entropy weighting, nonlinear fusion with interaction terms, and quantile-based classification"—to convert multi-parameter environmental states into Environmental Quality Index (EQI) and its environmental classification results. Because this process simultaneously considers extreme value suppression, differences in parameter types, differences in parameter importance, and interactions between parameters, it improves the consistency between environmental quality assessment results and actual environmental states, and provides a unified computational framework for quantitative environmental quality analysis in various scenarios such as urban air, industrial plants, and agricultural greenhouses.

[0028] In this embodiment, the step of performing anti-extreme value normalization on the original parameter set to obtain a normalized parameter set includes: Acquire normal monitoring data from the environmental monitoring system within a preset historical time period, wherein the normal monitoring data is valid data that has not triggered a pollution warning. The reference value is determined by averaging the normal monitoring data. The original parameter set is normalized based on the reference value to obtain the normalized parameter set.

[0029] Specifically, normal monitoring data from the environmental monitoring system within a preset historical time period can be obtained first. This normal monitoring data refers to valid data that has not triggered a pollution warning. The "preset historical time period" is used to extract an environmental baseline. In a specific embodiment of this application, the preset historical time period can be, for example, the past three months. That is, historical monitoring records from the environmental monitoring system within the past three months can be retrieved, and data that triggers pollution warnings and invalid data can be removed, retaining only the monitoring data representing normal operating conditions as normal monitoring data.

[0030] After obtaining normal monitoring data, the reference value is determined by averaging the normal monitoring data. Preferably, the normal monitoring data for each monitoring parameter can be averaged separately to form a reference value set corresponding to each parameter. Subsequently, the original parameter set is normalized based on the reference values ​​to obtain the normalized parameter set. Thus, the current values ​​of each parameter in the original parameter set no longer directly participate in subsequent calculations, but are first normalized relative to the normal level of the scenario itself, thereby improving the adaptability of subsequent analysis to scenario differences.

[0031] This embodiment uses real historical data of the target scenario under normal operating conditions as the reference value, rather than a fixed general constant, thereby improving the adaptability of normalization processing to scenario differences and enhancing the stability of subsequent environmental quality index calculation.

[0032] In this embodiment, when performing risk mapping on pollutant concentration parameters in the normalized parameter set that are linearly positively correlated with risk, the method includes: Determine pollutant concentration parameters that are linearly positively correlated with risk from the set of normalized parameters; Based on the pollutant thresholds in environmental quality standards, determine the piecewise points of the piecewise linear function; Based on the segmentation points, construct a piecewise linear function corresponding to the pollutant concentration parameters; The pollutant concentration parameter is input into the piecewise linear function for piecewise mapping to obtain the risk score corresponding to the pollutant concentration parameter.

[0033] This embodiment provides a specific risk mapping method for pollutant concentration parameters that are linearly positively correlated with risk. First, pollutant concentration parameters that are linearly positively correlated with risk are determined from the normalized parameter set. Here, "linearly positively correlated" means that within the target concentration range, as the pollutant concentration increases, the risk score also increases, and its growth relationship can be fitted using a piecewise linear approach.

[0034] After determining the target pollutant concentration parameter, the segmentation points of the piecewise linear function are determined based on the pollutant thresholds in the environmental quality standards. The "pollutant thresholds in the environmental quality standards" are used to construct the boundaries of the risk change intervals. In a specific embodiment of this application, for the PM2.5 parameter, 35 μg / m³ and 75 μg / m³ can be used as segmentation nodes; the slope of each segment can be set according to the risk increase rate, for example, it can be 10-30. Then, a piecewise linear function corresponding to the pollutant concentration parameter is constructed based on the segmentation points. Finally, the pollutant concentration parameter is input into the piecewise linear function for piecewise mapping to obtain the risk score corresponding to the pollutant concentration parameter. Thus, within different concentration intervals, the pollutant concentration parameter can correspond to different linear growth rates, thereby more finely characterizing risk changes.

[0035] It should be understood that the phrase "determining segment points based on pollutant thresholds in environmental quality standards" in this embodiment is a general statement. In specific embodiments of this application, for example, 35 μg / m³ and 75 μg / m³ can be selected as segment points for PM2.5; for other pollutants, the corresponding segment points can also be determined based on the thresholds in the corresponding environmental quality standards.

[0036] This embodiment ensures that the risk calculation of linearly positively correlated pollutant concentration parameters is consistent with the standard classification range, thereby enhancing the interpretability and standard consistency of the risk score.

[0037] In this embodiment, when performing risk mapping on pollutant concentration parameters in the normalized parameter set that are non-linearly positively correlated with risk, the method includes: Determine pollutant concentration parameters that are non-linearly positively correlated with risk from the set of normalized parameters; Determine the exponential function corresponding to the pollutant concentration parameter; Based on the pollutant characteristics of the pollutant concentration parameters, determine the function parameters of the exponential function; The pollutant concentration parameter is input into the exponential function for mapping to obtain the risk score corresponding to the pollutant concentration parameter.

[0038] A specific risk mapping method is proposed for pollutant concentration parameters that exhibit a non-linear positive correlation with risk. First, pollutant concentration parameters exhibiting a non-linear positive correlation with risk are determined from the normalized parameter set. The term "non-linear positive correlation" can be understood as follows: as pollutant concentration increases, the risk score also increases, but the rate of risk increase is not constant; rather, it accelerates with increasing concentration.

[0039] After determining the concentration parameter of the pollutant, an exponential function corresponding to the pollutant concentration parameter is determined; then, based on the pollutant characteristics of the pollutant concentration parameter, the function parameters of the exponential function are determined. The "pollutant characteristics" may include attributes such as the pollutant's toxicity intensity that can affect the shape of the risk growth curve. In a specific embodiment of this application, the coefficient k in the exponential function can, for example, be between 1.2 and 2.0, and adjusted according to the pollutant's toxicity intensity. Then, the pollutant concentration parameter is input into the exponential function for mapping, obtaining the risk score corresponding to the pollutant concentration parameter. Thus, in the high concentration range, the exponential function can more sensitively reflect the risk amplification trend of highly hazardous pollutants than a linear function.

[0040] It should be understood that the terms "pollutant characteristics" and "function parameters" in this embodiment are general descriptions; in specific embodiments of this application, the function parameters may be, for example, the coefficient k of an exponential function, and its value may be selected in the range of 1.2-2.0.

[0041] This embodiment uses an exponential function to map the concentration parameters of nonlinear positively correlated pollutants to risk, enabling a more accurate representation of the upward trend of risk within the high concentration and high hazard range, thereby improving the sensitivity of nonlinear pollution risk characterization.

[0042] In this embodiment, when performing risk mapping on the temperature and humidity parameters in the normalized parameter set, the method includes: Determine the temperature and humidity parameters from the normalized parameter set; Obtain normal monitoring data within a preset historical time period; Based on the aforementioned normal monitoring data, a suitable range is determined; Based on the degree of deviation of the temperature and humidity parameters from the suitable range, the temperature and humidity parameters are reverse-mapped to obtain the risk score corresponding to the temperature and humidity parameters.

[0043] This embodiment presents a risk mapping method for temperature and humidity parameters that differs from that for pollutant concentration parameters. First, the temperature and humidity parameters are determined from the normalized parameter set. Since the risk associated with temperature and humidity parameters does not exhibit a one-way relationship of "the larger the parameter value, the more dangerous" or "the smaller the parameter value, the more dangerous," but rather depends on the deviation of the parameter value from the suitable range, a specialized reverse mapping method is required.

[0044] After determining the temperature and humidity parameters, normal monitoring data within a preset historical time period is obtained; then, based on the normal monitoring data, a suitable range is determined. Here, the "preset historical time period" is used to extract the normal distribution range of temperature and humidity. In a specific embodiment of this application, for example, normal monitoring data from the past three months can be used; the "suitable range" can be determined based on the quantiles of the normal monitoring data.

[0045] In some preferred embodiments, the temperature and humidity parameters are inversely mapped based on their deviation from the suitable range to obtain a risk score corresponding to the temperature and humidity parameters. The inverse mapping can preferably be implemented using a sigmoid function (Logistic function), which uses the normalized center value of the suitable range as the function center, and the coefficient α can be, for example, 3.0-5.0. When the temperature and humidity parameters are near the suitable range, a lower risk score is output; when the temperature and humidity parameters deviate from the suitable range, a higher risk score is output.

[0046] It should be understood that in this embodiment, "preset historical time period", "suitable interval" and "reverse mapping" are all general descriptions. In the specific embodiments of this application, the preset historical time period can be, for example, the past 3 months, the suitable interval can be determined by the quantiles of the normal monitoring data of the past 3 months, and the reverse mapping can be implemented by an S-shaped function.

[0047] This embodiment defines temperature and humidity risk by using the method of "suitable range + degree of deviation", and preferably uses an S-shaped function for inverse mapping, so that temperature and humidity parameters can be incorporated into the environmental quality index calculation system in a way that conforms to the actual comfort or suitability rules.

[0048] In this embodiment, the step of calculating the weights corresponding to each parameter using the entropy weight method based on the risk score set to obtain a weight set includes: Obtain parameter monitoring samples within a preset sampling time period to obtain a sample set; Based on the sample set, the normalized proportion of each parameter in each sample is calculated to obtain the proportion set. The information entropy of each parameter is calculated based on the aforementioned proportion set to obtain the information entropy set; The weights corresponding to each parameter are calculated based on the information entropy set to obtain the weight set.

[0049] This embodiment further defines the specific calculation process of the entropy weight method. First, parameter monitoring samples within a preset sampling time period are obtained to form a sample set. The "preset sampling time period" is used to ensure that the sample size and sample distribution are statistically representative; in a specific embodiment of this application, the preset sampling time period can be, for example, at least 30 days.

[0050] After obtaining the sample set, the normalized proportion of each parameter in each sample is calculated based on the sample set, resulting in a proportion set. Then, the information entropy of each parameter is calculated based on the proportion set, resulting in an information entropy set. Finally, the weights corresponding to each parameter are calculated based on the information entropy set, resulting in a weight set. In a preferred embodiment, the sum of all parameter weights is 1. Thus, weights can be automatically generated based on the degree of difference of parameters in the sample set, without the need for manual pre-fixing of parameter weights.

[0051] It should be understood that the "preset sampling time period" in this embodiment is a general description, and in specific embodiments of this application, it may be at least 30 days.

[0052] This embodiment automatically determines the weight set based on the information entropy of the sample distribution, enabling the parameter weights to adaptively adjust as the actual monitoring data changes, thus reducing the subjectivity brought about by manual weighting.

[0053] In this embodiment, the step of performing a fusion calculation based on the risk score set and the weight set using a nonlinear fusion function with parameter interaction terms to obtain the Environmental Quality Index (EQI) includes: Obtain historical monitoring data within a preset historical time period; Based on the historical monitoring data, the correlation coefficients between the parameters are calculated to obtain the correlation coefficients between the indicators. Based on the complexity of environmental pollution, determine the interaction adjustment factors; The risk score set, the weight set, the correlation coefficient between the indicators, and the interaction adjustment factor are input into the nonlinear fusion function for fusion calculation to obtain the environmental quality index EQI.

[0054] This embodiment further defines the nonlinear fusion calculation process including parameter interaction terms. First, historical monitoring data within a preset historical time period is acquired. This historical monitoring data is used to calculate the correlation between different parameters. In a specific embodiment of this application, the preset historical time period can be, for example, the past year.

[0055] After obtaining historical monitoring data, the correlation coefficients between various parameters are calculated based on the historical monitoring data to obtain the correlation coefficients between indicators. Preferably, the correlation coefficients can be calculated using the Pearson correlation coefficient. Then, an interaction adjustment factor is determined based on the complexity of environmental pollution. The interaction adjustment factor is used to adjust the contribution of the parameter interaction terms in the nonlinear fusion function; in a specific embodiment of this application, the interaction adjustment factor can, for example, be between 0.1 and 0.5. Finally, the risk score set, the weight set, the correlation coefficients between indicators, and the interaction adjustment factor are input into the nonlinear fusion function for fusion calculation to obtain the Environmental Quality Index (EQI).

[0056] It should be understood that the "preset historical time period", "correlation coefficient", and "interaction adjustment factor" in this embodiment are all general terms. In the specific embodiments of this application, the preset historical time period can be, for example, the past 1 year, the correlation coefficient can be, for example, the Pearson correlation coefficient, and the interaction adjustment factor can be, for example, taken from 0.1 - 0.5.

[0057] As a whole, this embodiment introduces the correlation relationship and interaction adjustment ability between parameters into the fusion calculation process of the environmental quality index EQI, so that the finally obtained EQI can not only reflect the contribution of single parameters, but also reflect the comprehensive influence brought by the synergistic or inhibitory effects of parameters.

[0058] In this embodiment, determining the environmental grade division threshold set based on the quantile method according to the historical data of the environmental quality index EQI includes: Setting the basic environmental grade division threshold; Obtaining the historical data of the environmental quality index EQI; Performing statistical analysis on the historical data of the environmental quality index EQI based on the quantile method to obtain the statistical analysis result; Updating the basic environmental grade division threshold according to the statistical analysis result to obtain the environmental grade division threshold set.

[0059] This embodiment further defines the generation method of the environmental grade division threshold set. First, set the basic environmental grade division threshold. The basic environmental grade division threshold can be used as the initial boundary of the environmental grade. In the specific embodiments of this application, the basic threshold can be set, for example, as follows: excellent corresponds to EQI ≤ 25, good corresponds to 25 < EQI ≤ 75, and polluted corresponds to EQI > 75.

[0060] After setting the basic environmental grade division threshold, obtain the historical data of the environmental quality index EQI. The historical data is used to reflect the EQI distribution of the target scenario in the historical period. In the specific embodiments of this application, the historical data of the environmental quality index EQI can be, for example, the historical EQI data of the past 1 year. Then, perform statistical analysis on the historical data of the environmental quality index EQI based on the quantile method to obtain the statistical analysis result; and then update the basic environmental grade division threshold according to the statistical analysis result to obtain the environmental grade division threshold set. In some preferred embodiments, the update can be performed quarterly, that is, the quantile threshold is recalculated by updating the historical data every quarter.

[0061] It should be understood that the terms "historical EQI data", "basic environmental level classification threshold" and "update" in this embodiment are all general descriptions. In the specific embodiments of this application, historical data may be EQI historical data of the past year, basic thresholds may be 25 and 75 as the dividing points of excellent, good and polluted, and the update cycle may be quarterly.

[0062] In this embodiment, the method further includes: Obtain the Environmental Quality Index (EQI) corresponding to multiple monitoring nodes to obtain the node index set; Obtain node attribute information for each monitoring node; Based on the node attribute information, determine the node weight corresponding to each monitoring node to obtain a set of node weights; The regional environmental quality index is obtained by weighting the set of node indices and the set of node weights.

[0063] The Environmental Quality Index (EQI) corresponding to multiple monitoring nodes is obtained to form a node index set. These multiple monitoring nodes can be multiple environmental monitoring terminals deployed in different locations within the same area, or they can be sub-nodes in multiple areas.

[0064] After obtaining the node index set, the node attribute information of each monitoring node is acquired. Then, based on the node attribute information, the node weight corresponding to each monitoring node is determined to obtain a node weight set. The "node attribute information" is a general description; in specific embodiments of this application, it may include, for example, coverage area and / or service population. Finally, the node index set and the node weight set are weighted and calculated to obtain the regional environmental quality index. Thus, the environmental quality results of multiple local monitoring nodes can be summarized into a comprehensive result at the regional scale.

[0065] This embodiment expands from single-point environmental assessment to regional environmental assessment by "obtaining node EQI, determining weights based on node attribute information, and weighted summarization", providing a foundation for locating regional pollution hotspots and carrying out regional environmental management.

[0066] In this embodiment, the method further includes: The sliding window length is determined based on parameter fluctuations; Based on the sliding window length, the Environmental Quality Index (EQI) is smoothed by sliding window mean to obtain a smoothed EQI sequence. The rate of change is calculated based on the smoothed Environmental Quality Index (EQI) sequence to obtain the EQI change rate. The warning is triggered based on the rate of change of the Environmental Quality Index (EQI) and its duration, and the warning result is obtained. Based on the warning results, corresponding control recommendations are output.

[0067] In some implementations, this embodiment adds an environmental quality trend early warning process. First, the sliding window length is determined based on parameter fluctuations. The term "sliding window length" is a general term; in specific embodiments of this application, for highly fluctuating parameters, the sliding window length may be, for example, 5-10 minutes; for low-fluctuating parameters, the sliding window length may be, for example, 30-60 minutes.

[0068] After determining the sliding window length, the Environmental Quality Index (EQI) is smoothed using the sliding window mean based on that length, resulting in a smoothed EQI sequence. This time smoothing reduces the interference of short-term random fluctuations, sensor sampling noise, or instantaneous spikes on trend judgment. Then, the rate of change is calculated based on the smoothed EQI sequence, yielding the EQI change rate. Next, a corresponding warning is triggered based on the EQI change rate and its duration, resulting in a warning result. Finally, based on the warning result, corresponding control recommendations are output. This trend warning process can trigger corresponding warnings and push control recommendations based on the change rate and its duration.

[0069] In this embodiment, "sliding window length," "early warning result," and "control suggestion information" are all general descriptions. In specific embodiments of this application, the sliding window length can be, for example, 5-10 minutes for high fluctuation parameters and 30-60 minutes for low fluctuation parameters.

[0070] This embodiment enables the Environmental Quality Index (EQI) to not only characterize the current environmental state but also the trend of environmental state changes, thereby supporting more timely risk warnings and environmental management.

[0071] In one embodiment, such as Figure 2 As shown, a multi-parameter environmental quality index calculation method is provided. This method performs anti-extreme value normalization processing on multiple parameters collected by the environmental monitoring system, generates a risk score by combining a risk function model, automatically adapts the weights for multiple scenarios using the entropy weight method, and then calculates the environmental quality index (EQI) through a nonlinear fusion function containing parameter interaction terms. Finally, environmental levels are classified based on dynamic thresholds, achieving accurate environmental quality assessment for multiple scenarios such as urban air, industrial plants, and agricultural greenhouses. The steps are as follows: Anti-extreme value normalization processing: Reference value normalization, logarithmic compression or Box-Cox transformation are adopted. The reference value is set based on the environmental baseline of the corresponding scene to suppress extreme value interference. Dynamic risk function modeling: Input normalized parameter concentration, output risk score from 0 to 100. (Higher pollutant concentrations result in higher scores; temperature and humidity are inversely mapped to the "suitable range," with deviations increasing the score.) Based on the environmental parameter types and the mathematical relationship between data and risk, a corresponding risk function is selected. The function type is one of the following: Piecewise linear functions: suitable for scenarios where pollutant concentration and risk are linearly positively correlated (goodness of fit). The segmentation points are set based on the pollutant thresholds specified in the national environmental quality standards, such as the primary threshold of PM2.5 in the atmosphere (35 μg / m³) and the secondary threshold (75 μg / m³). The slope of each segment is set to 10-30 according to the risk increment rate. The exponential function is suitable for scenarios where pollutant concentration and risk have a non-linear positive correlation. And the second derivative is positive), the function expression is: Where the base e is a natural constant, and the coefficient k takes values ​​from 1.2 to 2.0. The result is a parameter normalization result, specifically adjusted according to the toxicity intensity of the pollutant using the S-shaped function (Logistic function): applicable to temperature and humidity parameters, the function expression is as follows. ,in The normalized center value is the appropriate interval, and the coefficient a takes a value of 3.0-5.0; Entropy weight method adaptive weight allocation: weights are allocated by calculating the information entropy of parameters (the smaller the information entropy, the greater the impact of the parameter on environmental quality, and the higher the weight), which can adapt to multiple scenarios without repeated modeling; Nonlinear fusion with interaction terms: EQI calculation formula is as follows ,in This is an interactive adjustment factor (adjusted according to the complexity of scene pollution). The correlation coefficient between indicators (quantifying the synergistic or inhibitory effect of parameters); Dynamic threshold classification: Based on the historical EQI data of the corresponding region over the past year, a basic threshold is set according to the quantile, and it is updated every quarter to adapt to seasonal environmental baseline changes.

[0072] The reference value setting rule for the anti-extreme value normalization in this embodiment is as follows: the reference value is taken from the average of the normal monitoring data of the corresponding scenario in the past 3 months. Normal monitoring data refers to valid data that has not triggered a pollution warning.

[0073] In this embodiment, the "suitable range" for temperature and humidity is set based on the quantile of normal monitoring data from the corresponding scenario over the past three months. When the temperature deviates from this range, the risk score increases. It increases linearly with the degree of deviation.

[0074] In this embodiment, the specific process of automatically generating the weight matrix using the entropy weight method is as follows: Obtain parameter monitoring samples for at least 30 days in the corresponding scenario, and calculate the first... Normalized proportion of the parameters for each sample , where is the normalized value and m is the total number of samples); Calculate the information entropy according to the formula ; Calculate the weights according to the formula to ensure that the sum of all parameter weights is 1.

[0075] In this embodiment, the correlation coefficient between indicators is calculated through the Pearson correlation coefficient of the historical monitoring data of the corresponding scenario in the past year. The interaction adjustment factor ranges from 0.1 to 0.5.

[0076] As Figure 3 shown, the basic thresholds for environmental level classification in this embodiment are: excellent (EQI ≤ 25), good (25 < EQI ≤ 75), polluted (EQI > 75). The quantile thresholds are recalculated quarterly by updating historical data.

[0077] This embodiment also includes the calculation steps of the regional environmental quality index: in the multi-node deployment scenario, set the node weights according to the coverage area or service population of the monitoring nodes , and the calculation formula of the regional environmental quality index is ( is the EQI of the k-th node), which is used to locate the regional pollution hotspots.

[0078] This embodiment also includes the environmental quality trend warning steps:<A Perform time smoothing on the EQI using the moving window average. The window length is adjusted according to the parameter volatility (5 - 10 minutes for high-volatility parameters, 30 - 60 minutes for low-volatility parameters); Calculate the change rate of the smoothed EQI .

[0079] Based on the value and duration of, trigger the corresponding warning and push the control suggestions.

[0080] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combinations of these technical features do not conflict, they should all be considered as the scope described in this specification.

Claims

1. A method for calculating a multi-parameter environmental quality index, characterized in that, The method includes: Obtain multiple parameters collected by the environmental monitoring system to obtain the raw parameter set; The original parameter set is subjected to anti-extreme value normalization processing to obtain a normalized parameter set, wherein the anti-extreme value normalization processing adopts at least one of reference value normalization, logarithmic compression or Box-Cox transformation; when reference value normalization is adopted, the reference value is set based on the environmental baseline. The normalized parameter set is input into the dynamic risk function model. Based on the type of each parameter in the normalized parameter set and the mathematical relationship between each parameter and risk, the corresponding type of risk function is determined. Risk mapping is performed on each parameter in the normalized parameter set to obtain a risk score set. Among them, the pollutant concentration parameter adopts a mapping method in which the risk score increases with the increase of concentration, and the temperature and humidity parameter adopts a reverse mapping method based on the suitable range, so that the risk score increases with the increase of the deviation from the suitable range. Based on the risk score set, the weights corresponding to each parameter are calculated using the entropy weight method to obtain a weight set. Based on the risk score set and the weight set, a nonlinear fusion function with parameter interaction terms is used for fusion calculation to obtain the Environmental Quality Index (EQI), wherein the nonlinear fusion function includes an interaction adjustment factor and a correlation coefficient between indicators. Based on historical data of the Environmental Quality Index (EQI), a set of thresholds for environmental classification was determined using the quantile method. Based on the set of environmental level classification thresholds, the Environmental Quality Index (EQI) is classified into levels to obtain the environmental level results.

2. The method according to claim 1, characterized in that, The step of performing anti-extreme value normalization on the original parameter set to obtain a normalized parameter set includes: Acquire normal monitoring data from the environmental monitoring system within a preset historical time period, wherein the normal monitoring data is valid data that has not triggered a pollution warning. The reference value is determined by averaging the normal monitoring data. The original parameter set is normalized based on the reference value to obtain the normalized parameter set.

3. The method according to claim 1, characterized in that, When performing risk mapping on pollutant concentration parameters in the normalized parameter set that are linearly positively correlated with risk, the method includes: Determine pollutant concentration parameters that are linearly positively correlated with risk from the set of normalized parameters; Based on the pollutant thresholds in environmental quality standards, determine the piecewise points of the piecewise linear function; Based on the segmentation points, construct a piecewise linear function corresponding to the pollutant concentration parameters; The pollutant concentration parameter is input into the piecewise linear function for piecewise mapping to obtain the risk score corresponding to the pollutant concentration parameter.

4. The method according to claim 1, characterized in that, When performing risk mapping on pollutant concentration parameters in the normalized parameter set that exhibit a non-linear positive correlation with risk, the method includes: Determine pollutant concentration parameters that are non-linearly positively correlated with risk from the set of normalized parameters; Determine the exponential function corresponding to the pollutant concentration parameter; Based on the pollutant characteristics of the pollutant concentration parameters, determine the function parameters of the exponential function; The pollutant concentration parameter is input into the exponential function for mapping to obtain the risk score corresponding to the pollutant concentration parameter.

5. The method according to claim 1, characterized in that, When performing risk mapping on the temperature and humidity parameters in the normalized parameter set, the method includes: Determine the temperature and humidity parameters from the normalized parameter set; Obtain normal monitoring data within a preset historical time period; Based on the aforementioned normal monitoring data, a suitable range is determined; Based on the degree of deviation of the temperature and humidity parameters from the suitable range, the temperature and humidity parameters are reverse-mapped to obtain the risk score corresponding to the temperature and humidity parameters.

6. The method according to claim 1, characterized in that, Based on the risk score set, the weights corresponding to each parameter are calculated using the entropy weight method to obtain a weight set, including: Obtain parameter monitoring samples within a preset sampling time period to obtain a sample set; Based on the sample set, the normalized proportion of each parameter in each sample is calculated to obtain the proportion set. The information entropy of each parameter is calculated based on the aforementioned proportion set to obtain the information entropy set; The weights corresponding to each parameter are calculated based on the information entropy set to obtain the weight set.

7. The method according to claim 1, characterized in that, The environmental quality index (EQI) is obtained by performing a fusion calculation based on the risk score set and the weight set using a nonlinear fusion function with parameter interaction terms, including: Obtain historical monitoring data within a preset historical time period; Based on the historical monitoring data, the correlation coefficients between the parameters are calculated to obtain the correlation coefficients between the indicators. Based on the complexity of environmental pollution, determine the interaction adjustment factors; The risk score set, the weight set, the correlation coefficient between the indicators, and the interaction adjustment factor are input into the nonlinear fusion function for fusion calculation to obtain the environmental quality index EQI.

8. The method according to claim 1, characterized in that, The step of determining the set of environmental level classification thresholds based on historical Environmental Quality Index (EQI) data and using the quantile method includes: Set the threshold for basic environmental level classification; Obtain historical data for the Environmental Quality Index (EQI); Statistical analysis was performed on the historical data of the Environmental Quality Index (EQI) based on the quantile method to obtain the statistical analysis results. The basic environmental level classification thresholds are updated based on the statistical analysis results to obtain the set of environmental level classification thresholds.

9. The method according to claim 1, characterized in that, The method further includes: Obtain the Environmental Quality Index (EQI) corresponding to multiple monitoring nodes to obtain the node index set; Obtain node attribute information for each monitoring node; Based on the node attribute information, determine the node weight corresponding to each monitoring node to obtain a set of node weights; The regional environmental quality index is obtained by weighting the set of node indices and the set of node weights.

10. The method according to claim 1, characterized in that, The method further includes: The sliding window length is determined based on parameter fluctuations; Based on the sliding window length, the Environmental Quality Index (EQI) is smoothed by sliding window mean to obtain a smoothed EQI sequence. The rate of change is calculated based on the smoothed Environmental Quality Index (EQI) sequence to obtain the EQI change rate. The warning is triggered based on the rate of change of the Environmental Quality Index (EQI) and its duration, and the warning result is obtained. Based on the warning results, corresponding control recommendations are output.

Images