Earth observation system performance evaluation method and platform based on artificial intelligence

Abstract

The invention discloses an earth observation system performance evaluation method and platform based on artificial intelligence, and relates to the technical field of performance evaluation, and the method comprises the steps: obtaining multi-source observation data according to an earth observation system; establishing a primary evaluation index system according to a plurality of primary evaluation indexes of the earth observation system; based on the primary evaluation index system, establishing a multi-dimensional performance evaluation model of the earth observation system; inputting the multi-source observation data into a multi-dimensional performance evaluation model to obtain a plurality of first-level performance evaluation results; weight distribution is carried out based on the first-level evaluation index system, and a performance evaluation weighting model is established; and inputting the plurality of primary performance evaluation results into the performance evaluation weighting model to obtain a performance evaluation result of the earth observation system. The technical problems of low evaluation precision and low response efficiency in the prior art are solved, and the technical effect of improving the evaluation precision and the system response efficiency is achieved.

Description

Technical Field

[0001] The present invention relates to the field of performance evaluation technology, and in particular to an artificial intelligence-based earth observation system performance evaluation method and platform. Background Art

[0002] With the growth of global environmental monitoring needs and the advancement of earth observation technology, earth observation systems have been widely used in ecological protection, resource management, disaster emergency response and other fields. However, with the diversification of data collection platforms and the continuous increase in data volume, the performance evaluation and management of earth observation systems face unprecedented challenges. Traditional earth observation system evaluation methods usually rely on simple statistical analysis and weighting methods, which have problems such as low evaluation accuracy and low response efficiency, and are difficult to meet the efficiency, accuracy and dynamic requirements of modern earth observation needs. Summary of the invention

[0003] The present application provides an artificial intelligence-based earth observation system performance evaluation method and platform, which is used to solve the technical problems of low evaluation accuracy and low response efficiency in the prior art.

[0004] In view of the above problems, this application provides an earth observation system performance evaluation method and platform based on artificial intelligence.

[0005] In a first aspect of the present application, a method for evaluating the performance of an earth observation system based on artificial intelligence is provided, the method comprising:

[0006] According to the earth observation system, multi-source observation data is obtained; according to multiple first-level evaluation indicators of the earth observation system, a first-level evaluation indicator system is established; based on the first-level evaluation indicator system, a multidimensional performance evaluation model of the earth observation system is established; the multi-source observation data is input into the multidimensional performance evaluation model to obtain multiple first-level performance evaluation results; based on the first-level evaluation indicator system, weights are assigned to establish a performance evaluation weighted model; the multiple first-level performance evaluation results are input into the performance evaluation weighted model to obtain the performance evaluation result of the earth observation system.

[0007] The second aspect of the present application provides an artificial intelligence-based earth observation system performance evaluation platform, the platform comprising:

[0008] An observation data acquisition module, wherein the observation data acquisition module acquires multi-source observation data according to an earth observation system; a first-level evaluation index system establishment module, wherein the first-level evaluation index system establishment module establishes a first-level evaluation index system according to multiple first-level evaluation indexes of the earth observation system; a performance evaluation model establishment module, wherein the performance evaluation model establishment module establishes a multi-dimensional performance evaluation model of the earth observation system based on the first-level evaluation index system; an evaluation result acquisition module, wherein the evaluation result acquisition module inputs the multi-source observation data into the multi-dimensional performance evaluation model to obtain multiple first-level performance evaluation results; a weighted model establishment module, wherein the weighted model establishment module performs weight allocation based on the first-level evaluation index system to establish a performance evaluation weighted model; a performance evaluation result acquisition module, wherein the performance evaluation result acquisition module inputs the multiple first-level performance evaluation results into the performance evaluation weighted model to obtain a performance evaluation result of the earth observation system.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0010] The present application obtains multi-source observation data according to the earth observation system; establishes a first-level evaluation index system according to multiple first-level evaluation indicators of the earth observation system; establishes a multidimensional performance evaluation model of the earth observation system based on the first-level evaluation index system; inputs the multi-source observation data into the multidimensional performance evaluation model to obtain multiple first-level performance evaluation results; performs weight distribution based on the first-level evaluation index system to establish a performance evaluation weighted model; inputs multiple first-level performance evaluation results into the performance evaluation weighted model to obtain the performance evaluation results of the earth observation system. The present invention solves the technical problems of low evaluation accuracy and low response efficiency in the prior art, and finally obtains the comprehensive performance evaluation results of the earth observation system by establishing a first-level evaluation index system, constructing a multidimensional performance evaluation model and a performance evaluation weighted model, and using multi-source observation data to input the model, thereby achieving the technical effect of improving evaluation accuracy and system response efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0012] Figure 1 A schematic diagram of a flow chart of an artificial intelligence-based earth observation system performance evaluation method provided in an embodiment of the present application;

[0013] Figure 2A schematic diagram of the structure of an artificial intelligence-based earth observation system performance evaluation platform provided in an embodiment of the present application.

[0014] Explanation of the reference numerals: observation data acquisition module 11, first-level evaluation index system establishment module 12, performance evaluation model establishment module 13, evaluation result acquisition module 14, weighted model establishment module 15, performance evaluation result acquisition module 16. DETAILED DESCRIPTION

[0015] This application provides an artificial intelligence-based earth observation system performance evaluation method and platform to solve the technical problems of low evaluation accuracy and low response efficiency in the existing technology. By establishing a primary evaluation indicator system, constructing a multidimensional performance evaluation model and a performance evaluation weighted model, and using multi-source observation data to input the model, the comprehensive performance evaluation results of the earth observation system are finally obtained, thereby achieving the technical effect of improving the evaluation accuracy and system response efficiency.

[0016] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0017] It should be noted that any variations of the terms "include" and "have" are intended to cover non-exclusive inclusions. For example, a process, method, platform, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.

[0018] Embodiment 1, as Figure 1 As shown, the present application provides an artificial intelligence-based earth observation system performance evaluation method, the method comprising:

[0019] Step S100: Obtain multi-source observation data according to the earth observation system.

[0020] In an embodiment of the present application, data is first collected from multiple observation platforms of the earth observation system, covering multi-dimensional information in space (such as satellite remote sensing images), ground (such as weather stations) and air (such as drones). Then, the collected raw data is loaded into the data processing system, and the formats of different data sources are standardized to ensure the consistency of the data. After loading, data cleaning is performed to remove noise, fill in missing data and ensure the consistency of data format. After cleaning, the data is fused, and information from different observation sources is integrated through spatial and temporal matching to generate a unified observation data set. Finally, the data is classified by content, format and purpose, such as being divided into environmental monitoring data, climate data and geographic data, etc., to obtain multi-source observation data.

[0021] Furthermore, in the method provided in the embodiment of the application, obtaining multi-source observation data according to the earth observation system also includes:

[0022] According to the earth observation system, an observation data set is loaded; data cleaning is performed according to the observation data set to obtain an observation data cleaning result; and classification is performed according to the observation data cleaning result to generate the multi-source observation data.

[0023] In the embodiment of the present application, the observation data set is first loaded, that is, the data set from different observation platforms is loaded into the data processing system. The observation data set includes satellite images, ground sensor data, meteorological data, etc. The loading process converts various types of data into a standardized format to ensure that they can be processed by the system compatibility. The purpose of loading is to achieve the unification of the format and structure between different data sources, and provide a consistent data basis for subsequent processing.

[0024] After loading, the data set is cleaned. Data cleaning is a crucial process, which mainly includes removing invalid data, repairing outliers, filling missing data, etc. to ensure the integrity and accuracy of the data. For example, satellite data may contain noise data or abnormal light spots, which will affect the accuracy of the evaluation model. Cleaning can remove these interfering data and improve data quality.

[0025] After the cleaning is completed, the data cleaning results are generated based on the cleaned data, and further classified. Classification is to group or divide the cleaned data into different categories or uses, such as environmental monitoring data, climate data, and geographic data.

[0026] Through the above loading, data cleaning and classification steps, a multi-source observation dataset is finally generated.

[0027] Step S200: establishing a primary evaluation indicator system according to a plurality of primary evaluation indicators of the earth observation system.

[0028] In the embodiment of the present application, firstly, a plurality of first-level evaluation indicators are determined based on the core requirements and functions of the earth observation system. These first-level evaluation indicators represent the key capability dimensions of the system, covering aspects such as hardware and software conditions, data processing capabilities, shared service levels, application results, competitive situation, and decision support. The determination of the first-level evaluation indicators is to ensure that the evaluation framework can fully reflect the core performance of the system, so as to facilitate the quantitative evaluation of the system in subsequent analysis.

[0029] After determining the indicators, the next step is to structure these indicators and build a first-level evaluation indicator system. This process organizes each first-level indicator according to its function or impact area in a hierarchical manner, making the system hierarchical and ensuring that the evaluation model can analyze system performance layer by layer. For example, hardware and software conditions include sub-items such as hardware performance and software compatibility, while data processing capabilities cover aspects such as data processing speed, accuracy, and reliability.

[0030] On this basis, evaluation criteria and measurement methods are set for each indicator in the primary indicator system to ensure that these indicators can be accurately measured in subsequent evaluations. Evaluation criteria are usually set by technical experts based on industry standards and actual needs, such as using quantitative indicators such as processing time and data accuracy to describe data processing capabilities.

[0031] Through the above steps, a complete first-level evaluation indicator system was finally established.

[0032] Furthermore, the method provided in the application embodiment also includes:

[0033] The multiple first-level evaluation indicators include hardware and software conditions, data processing capabilities, shared service levels, application results, competitive situation and decision support.

[0034] In the embodiment of the present application, multiple first-level evaluation indicators include software and hardware conditions, data processing capabilities, shared service levels, application results, competitive situation and decision support.

[0035] Among them, the hardware and software conditions measure the system's infrastructure, including the performance and reliability of hardware equipment, as well as the compatibility and stability of the software. Data processing capability evaluates the system's data processing speed, accuracy, and data storage and conversion capabilities, and is a key indicator for the system to efficiently process observation data. The shared service level examines the system's data sharing capabilities, user service levels, and data accessibility, reflecting the system's performance in terms of information openness and service coverage. The application results evaluate the value of the data and analysis results generated by the system in actual application scenarios, which is mainly reflected in the support effects on environmental monitoring and resource management. The position of the competitive situation analysis system in the industry or market and the comparison with other systems, covering market share, technological advantages, etc., to reflect the competitiveness of the system. Decision support examines the system's contribution to policy formulation and strategic planning, reflecting the actual value of its data results in decision support.

[0036] Step S300: Based on the first-level evaluation indicator system, a multi-dimensional performance evaluation model of the earth observation system is established.

[0037] In an embodiment of the present application, first, secondary correlation index mining is performed based on the primary evaluation index system. In this step, multiple secondary correlation indicators related to the primary indicators are refined, and multiple secondary correlation indicator distributions are formed to ensure the comprehensiveness and meticulousness of the evaluation. Next, based on the determined secondary correlation indicator distribution, performance evaluation records of the earth observation system are collected to obtain multiple performance evaluation record sets. These record sets contain the actual operating data of the system under various performance indicators, forming the data basis required for model training. Subsequently, these performance evaluation record sets are used to train the model using a deep learning method, and a primary evaluation model corresponding to each primary evaluation indicator is constructed one by one. These models can independently evaluate the performance of the earth observation system under various primary indicators.

[0038] Finally, all the first-level evaluation models are connected and integrated to generate the final multidimensional performance evaluation model.

[0039] Furthermore, in the method provided in the embodiment of the application, based on the primary evaluation index system, a multi-dimensional performance evaluation model of the earth observation system is established, which also includes:

[0040] According to the first-level evaluation indicator system, second-level correlation indicator mining is performed to obtain multiple second-level correlation indicator distributions; based on the multiple second-level correlation indicator distributions, performance evaluation records of the earth observation system are collected to obtain multiple performance evaluation record sets; according to the multiple performance evaluation record sets, deep learning is performed to obtain multiple first-level evaluation models; and the multiple first-level evaluation models are connected to generate the multidimensional performance evaluation model.

[0041] In the embodiment of the present application, firstly, the secondary correlation index mining is carried out according to the primary evaluation index system. The secondary correlation index is a refinement of the primary index, making the evaluation more accurate and in-depth. For example, for data processing capability, the secondary correlation index includes processing speed, accuracy, and storage efficiency. By identifying these secondary correlation indicators, multiple secondary correlation index distributions are formed.

[0042] After determining the distribution of secondary correlation indicators, the performance evaluation records of the earth observation system operation data are collected to form multiple performance evaluation record sets. These record sets contain the actual performance data of the system under each secondary indicator.

[0043] The collected data is then modeled using a feedforward neural network in deep learning. The feedforward neural network is an effective supervised learning algorithm suitable for processing high-dimensional feature data. By inputting the performance record data of the secondary correlation indicators, FNN extracts complex patterns and relationships to build multiple primary evaluation models. These models can evaluate the performance of the Earth Observation System under each primary indicator, such as the specific performance under the "data processing capability" indicator.

[0044] Finally, all the first-level evaluation models are connected and integrated to generate a multidimensional performance evaluation model.

[0045] Furthermore, in the method provided in the embodiment of the application, secondary correlation index mining is performed according to the primary evaluation index system to obtain multiple secondary correlation index distributions, and further includes:

[0046] The secondary indicators of the earth observation system are collected to obtain a secondary indicator set; based on each first-level evaluation indicator in the first-level evaluation indicator system, the secondary indicator set is respectively associated and evaluated to obtain multiple indicator associated evaluation sets; based on the indicator associated evaluation threshold, the secondary indicator set is screened according to the multiple indicator associated evaluation sets to generate the multiple secondary associated indicator distributions.

[0047] In the embodiment of the present application, various performance data of the system are first obtained through multi-source data collection methods in the earth observation system, such as satellite remote sensing, ground sensors and drone observation. These data are integrated through data collection and fusion technology, and data cleaning and format conversion algorithms are used to ensure the consistency of data from different sources, thereby constructing a complete set of secondary indicators.

[0048] After forming the secondary indicator set, the secondary indicator set under each primary evaluation indicator is evaluated for association based on the primary evaluation indicator system. In this process, correlation analysis and statistical regression analysis are used to evaluate the strength of association between the secondary indicators and the primary indicators. For example, the Pearson correlation coefficient or Spearman correlation coefficient is used to quantify the correlation between the primary indicators of data processing speed and data processing capacity, while statistical regression analysis further measures the contribution of the secondary indicators to the primary indicators. Through these analyses, multiple indicator association evaluation sets are generated, which clarify which secondary indicators have a strong correlation with the primary indicators.

[0049] Next, to ensure the efficiency and accuracy of the evaluation model, technical experts set the correlation evaluation threshold. Based on this threshold, a screening algorithm is used to screen the secondary indicators in each indicator correlation evaluation set. The screening process generally uses feature selection algorithms, such as variance threshold method or recursive feature elimination method, to remove secondary indicators with low correlation with primary indicators. These feature selection algorithms set evaluation thresholds to screen out secondary indicators that have strong explanatory power for primary indicators, thereby eliminating the interference of low-correlation data on model accuracy.

[0050] Finally, after screening, the highly correlated secondary indicators form multiple secondary correlation indicator distributions.

[0051] Step S400: inputting the multi-source observation data into the multi-dimensional performance evaluation model to obtain a plurality of primary performance evaluation results.

[0052] In an embodiment of the present application, first, the multi-source observation data is identified for associated data according to the distribution of secondary associated indicators. This step generates multiple associated data identification results by matching and analyzing the secondary indicators in different data sources, extracting data that is highly correlated with each primary indicator. Next, these associated data identification results are input into the corresponding primary evaluation models respectively. Each primary evaluation model analyzes and calculates the input data according to specific primary indicator requirements to generate corresponding primary performance evaluation results. These results represent the specific performance of the Earth Observation System in terms of hardware and software conditions, data processing capabilities, and shared service levels.

[0053] Furthermore, in the method provided in the embodiment of the application, the multi-source observation data is input into the multi-dimensional performance evaluation model to obtain multiple primary performance evaluation results, and further includes:

[0054] According to the multiple secondary correlation index distributions, the multi-source observation data are subjected to correlation data identification to obtain multiple correlation data identification results; the multiple correlation data identification results are input into the multiple primary evaluation models to generate the multiple primary performance evaluation results.

[0055] In an embodiment of the present application, first, based on the established secondary correlation index distribution, the multi-source observation data is identified for related data. This process uses the principal component analysis method to process and filter the main features in the data, extract the main components in the multi-source observation data that are highly correlated with the secondary correlation indicators, and generate multiple related data identification results.

[0056] Then, multiple associated data identification results are input into the corresponding first-level evaluation model. Each first-level evaluation model is specifically designed to perform performance analysis for different first-level indicators. In this process, a feedforward neural network model is applied to gradually process the input associated data identification results, pass the data layer by layer and identify the key patterns therein, and generate first-level evaluation results and multiple first-level performance evaluation results. These results are presented in the form of quantitative scores or indicators, reflecting the specific performance of the Earth Observation System under different first-level indicators, such as data processing capabilities, shared service levels, and application results.

[0057] Step S500: performing weight allocation based on the primary evaluation index system to establish a performance evaluation weighted model.

[0058] In the embodiment of the present application, the expert system is first used to weight the primary evaluation indicators and generate multiple weight allocation schemes. Next, the data is cleaned according to these weight allocation schemes to remove redundant and noisy data to generate multiple weight allocation matrices to ensure the accuracy and consistency of the data. Then, based on these weight allocation matrices, the final weight value of each indicator is determined by the centralized value calculation method to form a performance evaluation weighted model.

[0059] Furthermore, in the method provided in the embodiment of the application, weight allocation is performed based on the primary evaluation index system to establish a performance evaluation weighted model, and the method also includes:

[0060] The first-level evaluation index system is weighted according to the expert system to obtain multiple weight allocation schemes; data cleaning is performed according to the multiple weight allocation schemes to generate multiple weight allocation matrices; and the concentrated value of the weight of each indicator is calculated according to the multiple weight allocation matrices to generate the performance evaluation weighted model.

[0061] In the embodiment of the present application, the expert system is first used to weight the first-level evaluation indicators. The expert system supports decision-making through a domain expert knowledge base and a rule base. In this step, the expert system uses the hierarchical analysis method to compare each first-level evaluation indicator, such as data processing capacity, shared service level, etc., in pairs, and calculates the relative weight of each indicator. Through the hierarchical analysis method, the expert system generates multiple weight distribution schemes.

[0062] Next, the data is cleaned according to these weight allocation schemes. The goal of data cleaning is to remove noise, fill in missing values, and ensure the accuracy and consistency of the weight allocation scheme. Here, the data standardization method is used to make the data of each weight allocation scheme conform to a unified range and format. This standardization uses normalization to adjust the weight data to the same scale (such as between 0 and 1) to avoid affecting the calculation results due to unit or order of magnitude differences. The cleaned and standardized data are integrated into multiple weight allocation matrices, each of which represents all the primary indicators and their corresponding weights in a weight allocation scheme.

[0063] Then, based on these weight allocation matrices, the centralized value of each indicator weight is calculated to determine the final weighted value. In this stage, the weights in all weight allocation matrices are comprehensively calculated using the weighted average method. Specifically, for each first-level indicator, its values ​​in all weight allocation matrices are collected and a weighted average is calculated. This method can integrate the weight results of each scheme to make the final weight value more representative and robust. The calculated centralized value represents the final weight of each first-level indicator. The centralized value of the weight of each indicator is generated through this process.

[0064] Finally, all the concentrated value results are integrated to construct a performance evaluation weighted model.

[0065] Step S600: inputting the multiple primary performance evaluation results into the performance evaluation weighted model to obtain the performance evaluation result of the earth observation system.

[0066] In an embodiment of the present application, first, the multiple first-level performance evaluation results generated previously are input into the constructed performance evaluation weighted model. To ensure that each indicator can be scientifically integrated in the overall evaluation, the weighted model uses the final weight of each first-level indicator, that is, the weight value calculated by the previous concentrated value, to indicate the importance of each indicator in the comprehensive evaluation. Next, the performance evaluation weighted model applies its corresponding weight value to each first-level performance evaluation result. This process uses the weighted average method, multiplying each first-level evaluation result by its weight value, and then summing all weighted results.

[0067] Finally, the performance evaluation result of the Earth observation system is obtained by calculating the weighted total score. The result is presented in the form of a comprehensive score, which represents the overall performance level of the system.

[0068] In the embodiments of the present application, in summary, the embodiments of the present application have at least the following technical effects:

[0069] The present application obtains multi-source observation data according to the earth observation system; establishes a first-level evaluation index system according to multiple first-level evaluation indicators of the earth observation system; establishes a multidimensional performance evaluation model of the earth observation system based on the first-level evaluation index system; inputs the multi-source observation data into the multidimensional performance evaluation model to obtain multiple first-level performance evaluation results; performs weight distribution based on the first-level evaluation index system to establish a performance evaluation weighted model; inputs multiple first-level performance evaluation results into the performance evaluation weighted model to obtain the performance evaluation results of the earth observation system. The present invention solves the technical problems of low evaluation accuracy and low response efficiency in the prior art, and finally obtains the comprehensive performance evaluation results of the earth observation system by establishing a first-level evaluation index system, constructing a multidimensional performance evaluation model and a performance evaluation weighted model, and using multi-source observation data to input the model, thereby achieving the technical effect of improving evaluation accuracy and system response efficiency.

[0070] Embodiment 2 is based on the same inventive concept as the method for evaluating the performance of an earth observation system based on artificial intelligence in the above embodiment. Figure 2 As shown, the present application provides an artificial intelligence-based earth observation system performance evaluation platform, and the platform and method embodiments in the embodiments of the present application are based on the same inventive concept. Among them, the platform includes:

[0071] An observation data acquisition module 11, wherein the observation data acquisition module 11 acquires multi-source observation data according to an earth observation system; a first-level evaluation index system establishment module 12, wherein the first-level evaluation index system establishment module 12 establishes a first-level evaluation index system according to multiple first-level evaluation indexes of the earth observation system; a performance evaluation model establishment module 13, wherein the performance evaluation model establishment module 13 establishes a multidimensional performance evaluation model of the earth observation system based on the first-level evaluation index system; an evaluation result acquisition module 14, wherein the evaluation result acquisition module 14 inputs the multi-source observation data into the multidimensional performance evaluation model to obtain multiple first-level performance evaluation results; a weighted model establishment module 15, wherein the weighted model establishment module 15 performs weight allocation based on the first-level evaluation index system to establish a performance evaluation weighted model; a performance evaluation result acquisition module 16, wherein the performance evaluation result acquisition module 16 inputs the multiple first-level performance evaluation results into the performance evaluation weighted model to obtain a performance evaluation result of the earth observation system.

[0072] Furthermore, the platform is also used to implement the following functions:

[0073] According to the earth observation system, an observation data set is loaded; data cleaning is performed according to the observation data set to obtain an observation data cleaning result; and classification is performed according to the observation data cleaning result to generate the multi-source observation data.

[0074] Furthermore, the platform is also used to implement the following functions:

[0075] The multiple first-level evaluation indicators include hardware and software conditions, data processing capabilities, shared service levels, application results, competitive situation and decision support.

[0076] Furthermore, the platform is also used to implement the following functions:

[0077] According to the first-level evaluation indicator system, second-level correlation indicator mining is performed to obtain multiple second-level correlation indicator distributions; based on the multiple second-level correlation indicator distributions, performance evaluation records of the earth observation system are collected to obtain multiple performance evaluation record sets; according to the multiple performance evaluation record sets, deep learning is performed to obtain multiple first-level evaluation models; and the multiple first-level evaluation models are connected to generate the multidimensional performance evaluation model.

[0078] Furthermore, the platform is also used to implement the following functions:

[0079] The secondary indicators of the earth observation system are collected to obtain a secondary indicator set; based on each first-level evaluation indicator in the first-level evaluation indicator system, the secondary indicator set is respectively associated and evaluated to obtain multiple indicator associated evaluation sets; based on the indicator associated evaluation threshold, the secondary indicator set is screened according to the multiple indicator associated evaluation sets to generate the multiple secondary associated indicator distributions.

[0080] Furthermore, the platform is also used to implement the following functions:

[0081] According to the multiple secondary correlation index distributions, the multi-source observation data are subjected to correlation data identification to obtain multiple correlation data identification results; the multiple correlation data identification results are input into the multiple primary evaluation models to generate the multiple primary performance evaluation results.

[0082] Furthermore, the platform is also used to implement the following functions:

[0083] The first-level evaluation index system is weighted according to the expert system to obtain multiple weight allocation schemes; data cleaning is performed according to the multiple weight allocation schemes to generate multiple weight allocation matrices; and the concentrated value of the weight of each indicator is calculated according to the multiple weight allocation matrices to generate the performance evaluation weighted model.

[0084] Furthermore, the platform is also used to implement the following functions:

[0085] According to the page compensation parameters, the stretching coefficient, rotation angle, displacement vector, edge compensation direction, and edge compensation amount are analyzed, and the global compensation parameters and local compensation parameters are identified and separated; the control model of the tray position control parameters and the pressure control parameters are fitted using the historical test parameters, and the stretching coefficient, rotation angle, displacement vector, edge compensation direction, and edge compensation amount are calculated and converted through the control model to determine the control parameters; according to the global compensation parameters and the local compensation parameters, the adjustment area is located, and the control parameters are mapped and associated with the adjustment area to obtain the compensation control parameters.

[0086] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0087] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0088] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. An artificial intelligence-based earth observation system performance evaluation method, characterized in that: The method comprises: Obtain multi-source observation data based on the Earth Observation System; Establishing a primary evaluation indicator system according to the plurality of primary evaluation indicators of the earth observation system; Based on the first-level evaluation index system, a multi-dimensional performance evaluation model of the earth observation system is established; Inputting the multi-source observation data into the multi-dimensional performance evaluation model to obtain a plurality of primary performance evaluation results; Based on the first-level evaluation index system, weight allocation is performed to establish a performance evaluation weighted model; The multiple primary performance evaluation results are input into the performance evaluation weighted model to obtain the performance evaluation result of the earth observation system.

2. The method according to claim 1, characterized in that Based on the Earth Observation System, multi-source observation data is obtained, including: According to the earth observation system, loading an observation data set; Performing data cleaning according to the observation data set to obtain an observation data cleaning result; Classification is performed according to the observation data cleaning result to generate the multi-source observation data.

3. The method according to claim 1, characterized in that The multiple first-level evaluation indicators include hardware and software conditions, data processing capabilities, shared service levels, application results, competitive situation and decision support.

4. The method according to claim 1, characterized in that Based on the first-level evaluation index system, a multi-dimensional performance evaluation model of the earth observation system is established, including: Perform secondary correlation index mining according to the primary evaluation index system to obtain multiple secondary correlation index distributions; Based on the distribution of the multiple secondary correlation indicators, performance evaluation records of the earth observation system are collected to obtain multiple performance evaluation record sets; Perform deep learning based on the multiple performance evaluation record sets to obtain multiple primary evaluation models; The multiple primary evaluation models are connected to generate the multi-dimensional performance evaluation model.

5. The method according to claim 4, characterized in that Secondary correlation index mining is performed according to the primary evaluation index system to obtain multiple secondary correlation index distributions, including: Collecting secondary indicators of the earth observation system to obtain a secondary indicator set; Based on each first-level evaluation indicator in the first-level evaluation indicator system, respectively perform correlation evaluation on the second-level indicator set to obtain multiple indicator correlation evaluation sets; Based on the indicator association evaluation threshold, the secondary indicator set is screened according to the multiple indicator association evaluation sets to generate the multiple secondary association indicator distributions.

6. The method according to claim 4, characterized in that The multi-source observation data is input into the multi-dimensional performance evaluation model to obtain multiple primary performance evaluation results, including: Performing associated data identification on the multi-source observation data according to the distribution of the multiple secondary associated indicators to obtain multiple associated data identification results; The multiple associated data identification results are input into the multiple primary evaluation models to generate the multiple primary performance evaluation results.

7. The method according to claim 1, characterized in that Based on the first-level evaluation index system, weight allocation is performed to establish a performance evaluation weighted model, including: According to the expert system, weight allocation is performed on the first-level evaluation index system to obtain multiple weight allocation schemes; Perform data cleaning according to the multiple weight allocation schemes to generate multiple weight allocation matrices; The weight concentration value of each indicator is calculated according to the multiple weight allocation matrices to generate the performance evaluation weighted model.

8. The artificial intelligence-based earth observation system performance evaluation platform is characterized by: The platform includes: An observation data acquisition module, wherein the observation data acquisition module acquires multi-source observation data according to an earth observation system; A first-level evaluation indicator system establishment module, wherein the first-level evaluation indicator system establishment module establishes a first-level evaluation indicator system according to a plurality of first-level evaluation indicators of the earth observation system; A performance evaluation model establishment module, wherein the performance evaluation model establishment module establishes a multidimensional performance evaluation model of the earth observation system based on the primary evaluation indicator system; An evaluation result obtaining module, wherein the evaluation result obtaining module inputs the multi-source observation data into the multi-dimensional performance evaluation model to obtain a plurality of primary performance evaluation results; A weighted model building module, wherein the weighted model building module performs weight allocation based on the primary evaluation index system to build a performance evaluation weighted model; A performance evaluation result obtaining module is used to input the multiple primary performance evaluation results into the performance evaluation weighting model to obtain a performance evaluation result of the earth observation system.

Images