An artificial intelligence data management method

Abstract

The present invention discloses an artificial intelligence data management method, which relates to the field of data management technology, including deploying environmental sensors to collect data in real time and obtain unstructured data; evaluating the influence of data sources and calculating data trust for data screening; classifying the screened data; performing environmental prediction and trend analysis; storing data and backing up. The present invention comprehensively utilizes structured and unstructured data, and introduces innovative data source influence scoring and trust calculation methods combined with environmental prediction algorithms, which can collect and cover a wide range of environmental data in real time, ensure the comprehensiveness and real-time nature of the data, not only improve the accuracy of data screening, but also ensure the scientific nature of data processing, making environmental data analysis more accurate and efficient. The present invention not only improves the efficiency of environmental monitoring and management, but also provides more reliable and scientific data support for environmental decision-making, which has an important positive impact on environmental protection, public health and sustainable development.

Description

Technical Field

[0001] The present invention relates to the technical field of data management, and in particular to an artificial intelligence data management method. Background Art

[0002] In the existing environmental data management, a major technical challenge is how to effectively integrate and analyze a large amount of data from multiple sources. Environmental data includes meteorological data, air quality data, water quality data, soil data, biodiversity data and climate change data. Collecting environmental data plays an extremely important role in understanding the earth system, formulating environmental policies, protecting natural resources and public health. Traditional environmental data management methods mainly rely on traditional structured data collection methods, which cannot fully capture the dynamic changes of complex environmental systems and ignore the rich sources of unstructured data, such as news reports, social media discussions and public forums. These data sources can provide instant information about public perception and environmental events. In addition, the data processing process in existing technologies often lacks scientific assessment of the influence of data sources and data trust, which may lead to the use of inaccurate or irrelevant data in environmental analysis and prediction, resulting in low data utilization efficiency and affecting the accuracy and comprehensiveness of environmental analysis and decision-making. Data management also often relies on manual operations and simple data processing technologies, which are difficult to cope with the growing amount and complexity of data, and cannot provide sufficient data security and flexibility. Summary of the invention

[0003] In view of the problems existing in the above-mentioned existing artificial intelligence data management methods, the present invention is proposed.

[0004] Therefore, the problem to be solved by the present invention is that data management often relies on manual operations and simple data processing technologies, which makes it difficult to cope with the growing amount and complexity of data and also fails to provide sufficient data security and flexibility.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: an artificial intelligence data management method, which includes deploying environmental sensors to collect data in real time and obtain unstructured data; evaluating the influence of data sources and calculating data trust for data screening; classifying the screened data; performing environmental prediction and trend analysis; storing data and backing up.

[0006] As a preferred solution of the artificial intelligence data management method described in the present invention, the deployment of environmental sensors to collect data in real time and obtain unstructured data is to select different types of sensors according to the objectives of environmental monitoring, including using particulate matter sensors in air quality sensors to monitor PM2.5 and PM10, pH sensors and dissolved oxygen sensors in water quality monitoring sensors to monitor pH value and dissolved oxygen, and temperature, humidity, wind speed and wind direction sensors in meteorological sensors to monitor temperature, humidity and wind speed. Sensors are deployed at key locations including urban traffic arteries, industrial areas and water sources to collect environmental data in real time and pre-process the data. Obtaining unstructured data is to use natural language processing technology to obtain relevant content on air quality, water pollution and climate change from news websites and environmental forums. The acquisition formula is as follows:

[0007]

[0008] Among them, D n Refers to the sum of unstructured data obtained, NLP k refers to the natural language processing function applied to the kth data source, T k refers to the text content of the kth data source, P k Pointer to the specific parameters of the k-th data source, K is the total number of data sources;

[0009] After preprocessing, the captured data is integrated with the structured data collected by the sensor to form a comprehensive environmental data set. The integration formula is as follows:

[0010]

[0011] Among them, I is the final comprehensive environmental dataset, g(D n ) refers to unstructured data D n The preprocessing function, h(S i ) refers to the i-th sensor data S i The preprocessing function, refers to the normalization function, a i is the adjustment parameter, a and b are the upper and lower limits of the integral, representing the time range of data processing, dn represents the continuous and gradual change of unstructured data during the entire data processing process, and n is the index.

[0012] As a preferred solution of the artificial intelligence data management method of the present invention, the evaluation of the influence of the data source is to use network analysis, analyze the correlation between various data sources, including analyzing the correlation between air pollutants and health problem reports or the relationship between water quality changes and public reactions in environmental forums, and score each data source for influence based on the contribution of the data source in the past prediction model and the correlation between the data source and key environmental events. The influence scoring formula is as follows:

[0013] S k =w1·C k +w2·R k +w3·A k

[0014]

[0015]

[0016]

[0017] Among them, S k is the influence score of the kth data source, C k is the weight of the kth data source in the past prediction model, R k is the correlation between the kth data source and the key environmental event, A k is the correlation score between the kth data source and other data sources, w1, w2, w3 are weight coefficients, M is the number of historical models, T is the total number of trees in each random forest model, I refers to the amount of impurity reduction, D ki Refers to the specific data of the k-th data source in the i-th model, t is the index, D kn Is the data source D k The value at the nth observation point, Is the data source D k The average value, E n is the value of the key environmental event E at the nth observation point, is the average value of the key environmental event E, N is the total number of observation points, n is the index, K is the total number of data sources, N(D k ∩D j ) is the data source D k and D j The number of simultaneous occurrences, N is the total number of observation points, N(D k ) is the data source D k The number of occurrences, j represents the index;

[0018] Score the influence of each data source kPerform analysis and comparison, prioritize data sources according to the scoring results, give priority to key data sources with high influence scores when using data, and set the judgment threshold Θ:

[0019] If S k ≥Θ, it is high-impact data, indicating that these data play a key role in understanding and predicting the current environment, and the data is used as the main input of environmental analysis and prediction models;

[0020] If S k <Θ, it is low-impact data, indicating that these data have little impact on the current environment. In this case, these data are used as auxiliary information to support the main analysis and regularly evaluate the impact of the data.

[0021] As a preferred solution of the artificial intelligence data management method of the present invention, wherein: the calculation of data trustworthiness is to construct an algorithm to calculate data trustworthiness:

[0022] T d =w r ·R d +w c ·C d +w h ·H d +w t ·Tm d

[0023] Among them, T d Refers to the trustworthiness of the data, R d Refers to the reliability score of the data source, C d Refers to the consistency score of the data, H d Refers to the historical accuracy score of the data, Tm d Refers to the timeliness score of the data, w r , w c , w h , w t is the weight coefficient;

[0024] Set the judgment threshold Ω:

[0025] If T d ≥Ω, it is high-trust data, indicating that the data is accurate and reliable. The data is used as a reliable information source for key environmental decision-making and strategy formulation, and is used preferentially in modeling and analysis, with regular quality checks;

[0026] If T d <Ω, it is low-trust data, indicating that there are doubts about the accuracy and reliability of the data. Further verification or confirmation of the data should be carried out to reduce the weight of the data.

[0027] As a preferred solution of the artificial intelligence data management method described in the present invention, the data screening is to determine the data to be used for further analysis and decision-making based on the assessed influence and trust. For data points that meet both high trust and high impact standards, they are used as the highest quality data for key environmental analysis and decision support. High trust and low impact data are selected for auxiliary analysis or as supplementary information. The screening process and results are recorded, and the screening criteria are reviewed regularly.

[0028] As a preferred solution of the artificial intelligence data management method described in the present invention, the classification of the screened data is to conduct a detailed feature analysis of the screened data including the source, type and time series characteristics of the data, determine the key features of the classification including the spatiotemporal characteristics, physical properties and source labels of the data, define the data classification standards based on the results of the data feature analysis, classify the data according to the physical properties of the data and the frequency of data collection, use the hierarchical algorithm in the clustering algorithm to group the data into different categories, for complex data sets containing multiple features, use the decision tree in the supervised learning algorithm for classification, and use the cross-validation method to verify the classification results.

[0029] As a preferred solution of the artificial intelligence data management method of the present invention, the environmental prediction and trend analysis is to construct a formula to perform environmental prediction and trend analysis:

[0030]

[0031] Where P(t) is the environmental prediction value at time t, α, γ, δ, ∈ are adjustment parameters, n represents the total number of categories of classification results, C y (t) is the value of the y-th data at time t, β y is the coefficient, which represents the impact of the y-th type of data on the predicted value, S(t) is the seasonal adjustment factor, P(t-1) is the predicted value at time t-1, which represents the impact of historical data on the current prediction, and G(t) is a nonlinear function;

[0032] Input the classified data into the formula to obtain the environmental prediction value and then verify it. Compare the consistency of the prediction results with the actual environmental data. Adjust the parameters in the formula according to the verification results. Use newly collected data to continuously optimize and calibrate the prediction model. Analyze the impact of the prediction results on the environment, society, economy and public health. Identify potential risk areas including increased pollution, resource shortages or ecosystem damage. Based on the prediction results, formulate or adjust environmental management strategies and action plans. After the decision is implemented, continuously monitor environmental changes and the effectiveness of the strategy, regularly evaluate the effectiveness of the strategy, and adjust it based on new data and information.

[0033] As a preferred solution of the artificial intelligence data management method described in the present invention, wherein: the data storage and backup is to store the environmental data in a cloud storage service, encrypt all stored environmental data using an AES encryption algorithm, set environmental data access permissions, and only allow authorized users to access sensitive environmental data, formulate backup and disaster recovery strategies, regularly back up environmental data to servers in different geographical locations, implement real-time environmental data monitoring, regularly perform integrity and consistency checks on stored environmental data, and verify the integrity and availability of backup environmental data.

[0034] A computer device comprises: a memory and a processor; the memory stores a computer program, and is characterized in that: when the processor executes the computer program, the steps of the artificial intelligence data management method are implemented.

[0035] A computer-readable storage medium having a computer program stored thereon, characterized in that: when the computer program is executed by a processor, the steps of an artificial intelligence data management method are implemented.

[0036] The beneficial effects of the present invention are as follows: the present invention comprehensively utilizes structured and unstructured data, and introduces innovative data source influence scoring and trust calculation methods combined with environmental prediction algorithms, which can collect and cover a wide range of environmental data in real time, ensuring the comprehensiveness and real-time nature of the data, which not only improves the accuracy of data screening, but also ensures the scientific nature of data processing, making environmental data analysis more accurate and efficient. The present invention not only improves the efficiency of environmental monitoring and management, but also provides more reliable and scientific data support for environmental decision-making, which has an important positive impact on environmental protection, public health and sustainable development. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0038] Figure 1 Schematic diagram of the process of artificial intelligence data management.

[0039] Figure 2 This is a comparison table of evaluation indicators between the method of the present invention and the traditional method. DETAILED DESCRIPTION

[0040] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0041] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0042] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0043] Example 1

[0044] Reference Figure 1 , which is the first embodiment of the present invention, and this embodiment provides an artificial intelligence data management method, and the artificial intelligence data management method includes the following steps:

[0045] S1, deploy environmental sensors to collect data in real time and obtain unstructured data;

[0046] Specifically, deploying environmental sensors to collect data in real time and obtain unstructured data is to select different types of sensors according to the objectives of environmental monitoring, including using particulate matter sensors in air quality sensors to monitor PM2.5 and PM10, pH sensors and dissolved oxygen sensors in water quality monitoring sensors to monitor pH value and dissolved oxygen, and temperature, humidity, wind speed and wind direction sensors in meteorological sensors to monitor temperature, humidity and wind speed. Sensors are deployed at key locations including urban traffic arteries, industrial areas and water sources to collect environmental data in real time and pre-process the data. Obtaining unstructured data is to use natural language processing technology to obtain relevant content on air quality, water pollution and climate change from news websites and environmental forums. The acquisition formula is as follows:

[0047]

[0048] Among them, D n Refers to the sum of unstructured data obtained, NLP k Refers to the natural language processing function applied to the kth data source. The specific form depends on the NLP technology and algorithm used, including text analysis, keyword extraction, sentiment analysis, etc. k refers to the text content of the kth data source, which is the original text data obtained from news websites and environmental forums. k The pointer is the specific parameter for the kth data source, which is set according to the specific data collection requirements. K is the total number of data sources, which is determined according to the actual number of monitored news websites and forums.

[0049] After preprocessing, the captured data is integrated with the structured data collected by the sensor to form a comprehensive environmental data set. The integration formula is as follows:

[0050]

[0051] Among them, I is the final comprehensive environmental dataset, g(D n ) refers to unstructured data D n The specific form of the preprocessing function depends on the characteristics of the data and may include data cleaning, standardization, feature extraction and other steps. i ) refers to the i-th sensor data S i The specific form of the preprocessing function depends on the characteristics of the data and may include data normalization, noise filtering, feature extraction and other steps. Refers to the normalization function, which is used to ensure the consistency of data from different data sources when integrating. i is an adjustment parameter used to control the degree of influence of different sensor data. a and b are the upper and lower limits of the integral, representing the time range of data processing, which are set according to the actual time period of data collection. dn represents the continuous and gradual change of unstructured data during the entire data processing process. n is an index used to represent a specific aspect of unstructured data.

[0052] The deployment of sensors is to select the sensor deployment location according to the population density, traffic flow and industrial distribution in the city, deploy water quality monitoring sensors near and upstream of the water body, and install the sensors in the location with the least impact on the environment. The data collected by the sensors is structured data. Preprocessing the data means cleaning the data, removing errors in the data set, converting the data into a format more suitable for analysis, extracting useful features from the raw data for analysis, and integrating the data into a unified data set. The prior art often only focuses on a single data source, while the present invention combines multiple types of sensors and unstructured data sources. This step can capture multiple dimensions of environmental data, such as air quality, water quality and Climate change provides a more comprehensive environmental monitoring view. Deploying sensors at key locations such as urban traffic arteries, industrial areas and water sources can monitor environmental changes in real time, capture instant environmental data, and improve the timeliness and relevance of data. Using natural language processing technology to extract relevant information from unstructured data sources, such as news reports and forum discussions, can obtain deeper environmental insights and public perceptions, supplementing the perspectives that sensor data cannot provide. The present invention can more accurately identify and respond to environmental changes, provide a scientific basis for environmental management and policy making, and the comprehensive environmental data set formed through this step provides strong data support for subsequent environmental forecasts and trend analysis.

[0053] S2. Evaluate the influence of data sources and calculate data trust for data screening;

[0054] Specifically, the impact of data sources is assessed using network analysis to analyze the correlation between various data sources, including the correlation between air pollutants and health reports or the relationship between water quality changes and public responses in environmental forums. Each data source is scored based on its contribution to past prediction models and the correlation between the data source and key environmental events. The impact scoring formula is as follows:

[0055] S k =w1·C k +w2·R k +w3·A k

[0056]

[0057]

[0058]

[0059] Among them, S k is the influence score of the kth data source, C k is the contribution of the kth data source in the past prediction model, which is obtained by analyzing the performance of the historical prediction model. k is the correlation between the kth data source and the key environmental event, obtained through statistical analysis methods, A k is the correlation score between the kth data source and other data sources, obtained through association rule learning or network analysis, w1, w2, w3 are weight coefficients, M is the number of historical models, T is the total number of trees in each random forest model, I refers to the amount of impurity reduction, D ki refers to the specific data of the k-th data source in the i-th model, t is the index used to traverse each tree in the random forest model, ΔI(D ki , t) represents the feature D ki The amount of impurity reduction caused by splitting in the t-th decision tree of the ith model, where the model refers to one of the historical models, including decision tree, random forest, and linear regression. In this embodiment, the random forest model is preferred. kn Is the data source D k The value at the nth observation point, Is the data source D k The average value, E n is the value of the key environmental event E at the nth observation point, is the average value of the key environmental event E, N is the total number of observation points, n is the index used to traverse each observation point, K is the total number of data sources, N(D k∩D j ) is the data source D k and D j The number of simultaneous occurrences, N is the total number of observation points, N(D k ) is the data source D k The number of occurrences, j represents the index, which is used to traverse all other data sources and calculate the correlation between the data source and the k data source;

[0060] Score the influence of each data source k Perform analysis and comparison, prioritize data sources according to the scoring results, give priority to key data sources with high influence scores when using data, and set the judgment threshold Θ:

[0061] If S k ≥Θ, it is high-impact data, indicating that these data play a key role in understanding and predicting the current environment, and the data is used as the main input of environmental analysis and prediction models;

[0062] If S k <Θ, it is low-impact data, indicating that these data have little impact on the current environment. In this case, these data are used as auxiliary information to support the main analysis and regularly evaluate the impact of the data.

[0063] Network analysis is to construct a network graph. Different data sources are regarded as nodes in the network, and the associations between data sources are regarded as edges. The relationships in the data (such as co-occurrence, citation, etc.) are used to construct the network graph. Various metrics of nodes in the network, such as degree centrality, closeness centrality, betweenness centrality, etc., are calculated to evaluate the importance and influence of nodes. The Girvan-Newman algorithm in the community detection algorithm is used to identify the community structure in the network, that is, a group of closely related nodes. The network structure is analyzed to understand which data sources are in a core position in the network and how they are related to each other. The data source usage strategy is adjusted according to the results of the network analysis, and those data sources that occupy a core position in the network are given priority. The setting of the threshold needs to be set according to the actual application scenario. The calculation of the environmental data influence score can It is able to accurately identify and evaluate which data sources are most critical to environmental analysis and prediction. This is particularly important for processing large amounts of environmental data. It can effectively avoid information overload and ensure that the analysis focuses on the most influential data. By analyzing the correlation between data sources and key environmental events, the relevance and accuracy of environmental prediction models can be significantly improved, ensuring that the data input to the model is not only sufficient in quantity but also highly relevant and reliable in quality, so that environmental prediction and trend analysis are based on the most valuable data, and ensuring that environmental policies and management strategies are based on the most influential and credible data, which helps to improve the scientific nature and effectiveness of decision-making. Regularly evaluating the impact of data enables environmental monitoring and management to continuously adapt to environmental changes and emerging data trends, and maintain its long-term effectiveness and adaptability.

[0064] Furthermore, the data trust is calculated by constructing an algorithm to calculate the data trust:

[0065] T d =w r ·R d +w c ·C d +w h ·H d +w t ·Tm d

[0066] Among them, T d Refers to the trustworthiness of data, which is obtained by weighted summing the scores of data source reliability, consistency, historical accuracy, and timeliness. d Refers to the reliability score of the data source, based on the statistical analysis results of the accuracy of historical data, C d Refers to the consistency score of the data, based on the proportion of consistency in the data records, H d Refers to the historical accuracy score of the data, based on the accuracy of past data predictions, that is, the number of historically accurate predictions divided by the total number of predictions, Tm d Refers to the timeliness score of the data, using the exponential decay function e-λΔt Calculate, where Δt is the time difference between data generation and the current time, λ is the decay constant, indicating the influence of time on trust, and w r , w c , w h , w t is the weight coefficient, the interval range is [0, 1]. The interval range needs to be adjusted according to the needs of the actual application scenario;

[0067] Set the judgment threshold Ω:

[0068] If T d ≥Ω, it is high-trust data, indicating that the data is accurate and reliable. The data is used as a reliable information source for key environmental decision-making and strategy formulation, and is used preferentially in modeling and analysis, with regular quality checks;

[0069] If T d <Ω, it is low-trust data, indicating that there are doubts about the accuracy and reliability of the data. Further verification or confirmation of the data should be carried out to reduce the weight of the data.

[0070] The threshold is set according to the actual application scenario. The existing technology often lacks a comprehensive assessment of data trust, which may lead to misjudgment or reliance on low-quality data for decision-making. The present invention provides a comprehensive and detailed method to evaluate data trust by accurately evaluating data trust and comprehensively considering the reliability, consistency, historical accuracy and timeliness of data sources. This comprehensive evaluation can more accurately reflect the overall quality of the data. By distinguishing between high-trust data and low-trust data, it can ensure that the most reliable data is used first in important environmental decisions and strategy formulation, thereby improving the accuracy and reliability of decisions. The set trust threshold makes data management more dynamic and flexible, allowing the trust of data to be regularly reviewed and updated to reflect the latest data quality.

[0071] Furthermore, data screening is performed to determine the data that will be used for further analysis and decision-making based on the assessed impact and trust. For data points that meet both high-trust and high-impact standards, they are used as the highest-quality data for key environmental analysis and decision support. High-trust and low-impact data are selected for auxiliary analysis or as supplementary information. The screening process and results are recorded, and the screening criteria are reviewed regularly.

[0072] By combining data trust and influence scores for screening, we ensure the high quality and high relevance of the data used, effectively filter out low-quality data or data irrelevant to the current analysis goals, and conduct refined hierarchical management of data. We select high-trust and high-impact data as the basis for key analysis, ensuring that key decisions and analysis are based on the most influential and credible data, which helps to improve the accuracy of environmental decision-making and analysis and reduce the possibility of misjudgment. At the same time, lower-impact but high-trust data is used for auxiliary analysis, which improves the overall efficiency of data utilization.

[0073] S3, classifying the filtered data;

[0074] Specifically, classifying the filtered data involves conducting a detailed feature analysis of the filtered data, including the source, type and time series characteristics of the data; determining the key features for classification, including the spatiotemporal characteristics, physical properties and source labels of the data; defining the standard for data classification based on the results of data feature analysis; classifying the data according to the physical properties of the data and the frequency of data collection; using the hierarchical algorithm in the clustering algorithm to group the data into different categories; and for complex data sets containing multiple features, using the decision tree in the supervised learning algorithm for classification; and using the cross-validation method to verify the classification results.

[0075] Detailed feature analysis is to use descriptive statistical analysis (mean, median, standard deviation, etc.) to understand the basic characteristics of the data, identify and extract key features, which may include statistical properties of the data, time series characteristics (such as trends, seasonality), spatial characteristics, etc. Hierarchical clustering algorithm is used to classify the data to determine the distance metric used to measure the similarity between data points, such as Euclidean distance or Manhattan distance. The hierarchical clustering tree of the data is constructed by gradually merging the most similar data points or data clusters. The optimal number of categories is determined by analyzing the hierarchical tree. The hierarchical tree is divided according to the determined number of categories to form the final Data classification, use the decision tree algorithm to classify complex data sets, select the features that are most helpful for classification from the data set, use algorithms (such as ID3, C4.5 or CART) to build a decision tree, use the decision tree model to classify data, and use the cross-validation method to verify the classification results, split the data set into training set and validation set, divide the data set into K subsets, set aside one subset each time as the validation set, and the rest as the training set, train the model on the training set, and evaluate the model performance on the validation set, evaluate the model's classification accuracy, recall rate, F1 score and other indicators to determine the effectiveness of the model. By comprehensively analyzing the source, type and time series characteristics of the data, as well as the spatiotemporal characteristics, physical properties and source labels, the comprehensiveness and depth of data classification are ensured. When processing complex data sets, the use of advanced algorithms such as decision trees for data classification can more accurately group data into different categories. By classifying data, data can be processed and utilized more effectively, ensuring that each type of data is appropriately analyzed and utilized, providing a solid foundation for advanced analytical activities such as environmental forecasting and trend analysis, ensuring the efficiency of data processing and the effectiveness of data analysis, thereby improving the performance and value of the entire environmental data management method.

[0076] S4. Conduct environmental forecasting and trend analysis;

[0077] Specifically, environmental forecasting and trend analysis are performed by constructing formulas:

[0078]

[0079] Where P(t) is the environmental prediction value at time t, is a polynomial function, which indicates the direct impact of classification results on prediction. α, y, δ, ∈ are adjustment parameters, and the interval range is [0, 1]. It should be noted that the interval range is adjusted according to the actual application scenario. α is used to adjust the direct impact of classification results on prediction, y is used to adjust the impact of seasonal factors on prediction, δ is used to balance the impact of historical prediction values ​​on current predictions, ∈ is used to adjust the impact of nonlinear trends and outliers on predictions, n represents the total number of categories of classification results, and C y(t) is the value of the y-th data at time t, collected from environmental sensors or other data sources, β y is the coefficient in the polynomial function, which represents the impact of the y-th type of data on the predicted value. It is calculated by statistical methods and can make the prediction model best fit the historical data. S(t) is the seasonal adjustment factor, which is used to capture the periodic changes in the data. It is calculated by analyzing the seasonal patterns in the historical data, such as using Fourier transform to identify and simulate periodic patterns. P(t-1) is the predicted value at time t-1, which represents the impact of historical data on the current prediction. It is derived from the previous prediction results. G(t) is a nonlinear function used to capture nonlinear trends and outliers in the data. The specific form depends on the characteristics of the data and can be logarithmic and exponential functions;

[0080] Input the classified data into the formula to obtain the environmental prediction value and then verify it. Compare the consistency of the prediction results with the actual environmental data. Adjust the parameters in the formula according to the verification results. Use newly collected data to continuously optimize and calibrate the prediction model. Analyze the impact of the prediction results on the environment, society, economy and public health. Identify potential risk areas including increased pollution, resource shortages or ecosystem damage. Based on the prediction results, formulate or adjust environmental management strategies and action plans. After the decision is implemented, continuously monitor environmental changes and the effectiveness of the strategy, regularly evaluate the effectiveness of the strategy, and adjust it based on new data and information.

[0081] Traditional methods usually lack real-time updating and calibration functions. The continuous optimization and calibration mechanism provided by the present invention ensures that the prediction model maintains its relevance and accuracy over time. By continuously monitoring environmental changes and policy effects, response measures can be adjusted in a timely manner to effectively respond to environmental risks. Based on the comparison between the prediction results and the actual data, the prediction model is continuously adjusted and optimized. This dynamic optimization mechanism ensures that the model continues to improve in accuracy over time and with the acquisition of new data. By including seasonal adjustment factors and the influence of historical data, the prediction model can take into account the impact of time changes on the environment, making the prediction more accurate. Accurate predictions enable environmental management strategies to be based on reliable data and in-depth analysis, thereby more effectively responding to environmental problems and contributing to the long-term sustainable management and utilization of environmental resources.

[0082] S5, store data and back it up;

[0083] Specifically, storing and backing up data means storing environmental data in cloud storage services, encrypting all stored environmental data using the AES encryption algorithm, setting environmental data access permissions to allow only authorized users to access sensitive environmental data, formulating backup and disaster recovery strategies, regularly backing up environmental data to servers in different geographical locations, implementing real-time environmental data monitoring, regularly performing integrity and consistency checks on stored environmental data, and verifying the integrity and availability of backup environmental data.

[0084] When using AES encryption, select the AES key length according to security requirements, generate a random key for encrypting and decrypting data, store the generated key securely to prevent unauthorized access, standardize the data length to a multiple of 16 bytes, select the CBC (cipher feedback mode) encryption mode to encrypt the data, store the encrypted data in the database, and use the AES encryption algorithm to encrypt all stored environmental data. It is especially important for environmental data containing sensitive or confidential information. By regularly backing up data to geographically dispersed servers, the risk of data loss due to single point failures is greatly reduced, and a more comprehensive and systematic data recovery solution is provided than traditional methods. The implementation of real-time environmental data monitoring and regular integrity checks can promptly detect and fix data problems to ensure the continued accuracy and consistency of data.

[0085] Example 2

[0086] This is the second embodiment of the present invention, and the difference between this embodiment and the previous embodiment is that:

[0087] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random AcCess Memory), disk or optical disk, etc., which can store program codes.

[0088] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0089] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0090] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

Claims

1. An artificial intelligence data management method, characterized in that: include, Deploy environmental sensors to collect data in real time and acquire unstructured data; Evaluate the influence of data sources and calculate data trust for data screening; Categorize the filtered data; Conduct environmental forecasting and trend analysis; Store data and back it up; The deployment of environmental sensors to collect data in real time and obtain unstructured data is to select different types of sensors according to the objectives of environmental monitoring, including using particulate matter sensors in air quality sensors to monitor PM2.5 and PM10, pH sensors and dissolved oxygen sensors in water quality monitoring sensors to monitor pH value and dissolved oxygen, and temperature, humidity, wind speed and wind direction sensors in meteorological sensors to monitor temperature, humidity and wind speed. Sensors are deployed at key locations including urban traffic arteries, industrial areas and water sources to collect environmental data in real time and pre-process the data. The acquisition of unstructured data is to use natural language processing technology to obtain relevant content on air quality, water pollution and climate change from news websites and environmental forums. The acquisition formula is as follows: Among them, D n Refers to the sum of unstructured data obtained, NLP k refers to the natural language processing function applied to the kth data source, T k refers to the text content of the kth data source, P k Pointer to the specific parameters of the k-th data source, K is the total number of data sources; After preprocessing, the captured data is integrated with the structured data collected by the sensor to form a comprehensive environmental data set. The integration formula is as follows: Among them, I is the final comprehensive environmental dataset, g(D n ) refers to unstructured data D n The preprocessing function, h(S i ) refers to the i-th sensor data S i The preprocessing function, refers to the normalization function, a i is the adjustment parameter, a and b are the upper and lower limits of the integral, representing the time range of data processing, dn represents the continuous and gradual change of unstructured data during the entire data processing process, and n is the index; The evaluation of the influence of data sources is to use network analysis to analyze the correlation between various data sources, including the correlation between air pollutants and health problem reports or the relationship between water quality changes and public reactions in environmental forums. The influence of each data source is scored based on the contribution of the data source in the past prediction model and the correlation between the data source and key environmental events. The influence scoring formula is as follows: S k =w1·C k +w2·R k +w3·A k Among them, S k is the influence score of the kth data source, C k is the contribution of the kth data source in the past prediction model, R k is the correlation between the kth data source and the key environmental event, A k is the correlation score between the kth data source and other data sources, w1, w2, w3 are weight coefficients, M is the number of historical models, T is the total number of trees in each random forest model, I refers to the amount of impurity reduction, D ki Refers to the specific data of the k-th data source in the i-th model, t is the index, D kn Is the data source D k The value at the nth observation point, Is the data source D k The average value, E n is the value of the key environmental event E at the nth observation point, is the average value of the key environmental event E, N is the total number of observation points, n is the index, K is the total number of data sources, N(D k ∩D j ) is the data source D k and D j The number of simultaneous occurrences, N is the total number of observation points, N(D k ) is the data source D k The number of occurrences, j represents the index; Score the influence of each data source k Perform analysis and comparison, prioritize data sources according to the scoring results, give priority to key data sources with high influence scores when using data, and set the judgment threshold Θ: If S k ≥Θ, it is high-impact data, indicating that these data play a key role in understanding and predicting the current environment, and the data is used as the main input of environmental analysis and prediction models; If S k <Θ, then it is low-impact data, indicating that the data has little impact on the current environment. In this case, the data is used as auxiliary information to support the main analysis and regularly evaluate the impact of the data; The environmental prediction and trend analysis is to construct a formula to perform environmental prediction and trend analysis: Where P(t) is the environmental prediction value at time t, α, γ, δ, ∈ are adjustment parameters, n represents the total number of categories of classification results, C y (t) is the value of the y-th data at time t, β y is the coefficient, which represents the impact of the y-th type of data on the predicted value, S(t) is the seasonal adjustment factor, P(t-1) is the predicted value at time t-1, which represents the impact of historical data on the current prediction, and G(t) is a nonlinear function; Input the classified data into the formula to obtain the environmental prediction value and then verify it. Compare the consistency of the prediction results with the actual environmental data. Adjust the parameters in the formula according to the verification results. Use newly collected data to continuously optimize and calibrate the prediction model. Analyze the impact of the prediction results on the environment, society, economy and public health. Identify potential risk areas including increased pollution, resource shortages or ecosystem damage. Based on the prediction results, formulate or adjust environmental management strategies and action plans. After the decision is implemented, continuously monitor environmental changes and the effectiveness of the strategy, regularly evaluate the effectiveness of the strategy, and adjust it based on new data and information.

2. The artificial intelligence data management method according to claim 1, characterized in that: The calculation of data trust is to construct an algorithm to calculate data trust: T d =w r ·R d +w c ·C d +w h ·H d +w t ·Tm d Among them, T d Refers to the trustworthiness of the data, R d Refers to the reliability score of the data source, C d Refers to the consistency score of the data, H d Refers to the historical accuracy score of the data, Tm d Refers to the timeliness score of the data, w r , w c , w h , w t is the weight coefficient; Set the judgment threshold Ω: If T d ≥Ω, it is high-trust data, indicating that the data is accurate and reliable. The data is used as a reliable information source for key environmental decision-making and strategy formulation, and is used preferentially in modeling and analysis, with regular quality checks; If T d <Ω, it is low-trust data, indicating that there are doubts about the accuracy and reliability of the data. Further verification or confirmation of the data should be carried out to reduce the weight of the data.

3. The artificial intelligence data management method according to claim 2, characterized in that: The data screening is to determine the data to be used for further analysis and decision-making based on the impact and trust of the assessment. For data points that meet both high-trust and high-impact standards, the data will be used as the highest quality data for key environmental analysis and decision support. High-trust and low-impact data will be selected for auxiliary analysis or as supplementary information. The screening process and results will be recorded, and the screening criteria will be reviewed regularly.

4. The artificial intelligence data management method according to claim 3, characterized in that: The classification of the filtered data is to conduct a detailed feature analysis on the filtered data including the source, type and time series characteristics of the data, determine the key features of the classification including the spatiotemporal characteristics, physical properties and source labels of the data, define the standard of data classification based on the results of the data feature analysis, classify the data according to the physical properties of the data and the frequency of data collection, use the hierarchical algorithm in the clustering algorithm to group the data into different categories, and for complex data sets containing multiple features, use the decision tree in the supervised learning algorithm for classification, and use the cross-validation method to verify the classification results.

5. The artificial intelligence data management method according to claim 4, characterized in that: The data storage and backup described herein is to store the environmental data in a cloud storage service, encrypt all stored environmental data using an AES encryption algorithm, set environmental data access permissions to allow only authorized users to access sensitive environmental data, formulate backup and disaster recovery strategies, regularly back up environmental data to servers in different geographical locations, implement real-time environmental data monitoring, regularly perform integrity and consistency checks on stored environmental data, and verify the integrity and availability of backup environmental data.

6. A computer device comprising: Memory and processor; The memory stores a computer program, wherein the processor implements the steps of any one of claims 1 to 5 when executing the computer program.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images