Electronic information intelligent analysis and decision-making system based on artificial intelligence

Abstract

The invention discloses an electronic information intelligent analysis and decision-making system based on artificial intelligence, and particularly relates to the technical field of electronic information processing and artificial intelligence, and the system specifically comprises a multi-modal data fusion module, a dynamic knowledge graph construction and updating module, an intelligent decision-making module, and a decision-making verification and feedback module. According to the electronic information intelligent analysis and decision-making system based on artificial intelligence, efficient fusion of multi-modal data is achieved through a pulse neural network and a data electrocardiogram mechanism, dynamic weight distribution adapts to high-speed data, and the signal recognition accuracy is improved in a complex environment; a self-evolution knowledge graph is constructed, a concept entropy model drives minute-level automatic reconstruction, and manual intervention is avoided; the processing efficiency, the decision real-time performance and the reliability are remarkably improved in scenes such as spectrum management and emergency communication, and the method has innovation and engineering values.

Description

Technical Field

[0001] The present invention relates to the field of electronic information processing and artificial intelligence technology, and in particular to an electronic information intelligent analysis and decision-making system based on artificial intelligence. Background Art

[0002] In the field of electronic information processing, with the explosive growth of data volumes and the increasing complexity of application scenarios, existing analysis and decision-making systems face numerous challenges. Traditional systems suffer from low data fusion efficiency when processing multi-source, heterogeneous data, failing to fully tap into the data's value. Knowledge graph construction and updates rely on manual intervention, resulting in poor timeliness. The decision-making process lacks explainability, making it difficult to meet the demands of scenarios requiring high security and reliability. Furthermore, existing technologies lack the accuracy and real-time nature of decision-making in complex and changing environments, making it difficult to quickly adapt to changing circumstances and make informed decisions. Summary of the Invention

[0003] The main purpose of the present invention is to provide an electronic information intelligent analysis and decision-making system based on artificial intelligence, which can effectively solve the problems in the existing technology of low data fusion efficiency, untimely knowledge graph updating, unexplainable decision-making process, and insufficient decision-making accuracy and real-time performance.

[0004] To achieve the above object, the technical solution adopted by the present invention is:

[0005] An electronic information intelligent analysis and decision-making system based on artificial intelligence, comprising:

[0006] Multimodal data fusion module, used to obtain electronic information data of multiple modes and divide it into high-frequency data and low-frequency data according to the processing frequency, and generate data fusion distribution results;

[0007] Dynamic knowledge graph construction and update module, used to analyze knowledge relevance and importance based on data fusion allocation results and generate optimized dynamic knowledge graph;

[0008] Intelligent decision-making module, used to extract decision node information from the dynamic knowledge graph and generate decision execution results;

[0009] Decision verification and feedback module, used to monitor changes in decision data flow and reallocate risk node paths, generating decision verification feedback results;

[0010] The adaptive adjustment module is used to dynamically adjust the decision execution priority according to the feedback results and generate a decision optimization report.

[0011] Preferably, the multimodal data fusion module includes:

[0012] Data classification submodule, used to divide data into high-frequency data and low-frequency data according to processing frequency;

[0013] The feature fusion allocation submodule uses the formula

[0014]

[0015] Calculate the fusion allocation result; where S represents the fusion capability of high-performance fusion nodes, H represents the fusion capability of ordinary fusion nodes, and F H With F L Represent the processing frequencies of high-frequency and low-frequency data respectively, L S With L H They represent the load conditions of high-performance fusion nodes and ordinary fusion nodes respectively, D represents the fusion demand, and R represents the fusion allocation result.

[0016] The fusion adjustment submodule is used to monitor node load in real time and dynamically adjust data allocation results.

[0017] Preferably, the dynamic knowledge graph construction and update module includes:

[0018] The knowledge association analysis submodule is used to calculate the knowledge association matching degree between nodes;

[0019] The knowledge importance evaluation submodule uses the formula

[0020]

[0021] Evaluate the node knowledge importance, where M represents the degree of match between the node knowledge importance and the task requirements, C represents the node storage capacity, F represents the node load rate, D represents the task processing time requirement, and U represents the task priority;

[0022] The graph optimization submodule is used to adjust the graph node storage resources and knowledge allocation.

[0023] Preferably, the intelligent decision-making module includes:

[0024] Decision status extraction submodule, used to obtain the execution progress and risk assessment of decision nodes;

[0025] The execution capability and progress comparison submodule is used to calculate the execution capability and progress gap;

[0026] The decision path adjustment submodule is used to adjust the decision execution path in real time.

[0027] Preferably, the decision verification and feedback module includes:

[0028] Data acquisition submodule, used to collect decision-making data flow information in real time;

[0029] Data analysis submodule, used to calculate the traffic fluctuation percentage;

[0030] The path adjustment submodule is used to optimize the execution path of risk nodes.

[0031] Preferably, the adaptive adjustment module includes:

[0032] Decision fluctuation monitoring submodule, used to evaluate the amplitude of decision fluctuation;

[0033] The path load analysis submodule uses the formula

[0034]

[0035] Calculate the node load index, where L k is the load index of the kth node, F k is the decision fluctuation amplitude value of the kth node, F j Decision flow for nodes, is the average value of decision traffic, n is the total number of nodes;

[0036] The decision scheduling optimization submodule is used to adjust the priority of decision tasks.

[0037] An electronic information intelligent analysis and decision-making method, based on the above-mentioned electronic information intelligent analysis and decision-making system based on artificial intelligence, includes the following steps:

[0038] S1. Data fusion and classification steps: divide data by processing frequency and dynamically adjust fusion allocation;

[0039] S2. Knowledge graph construction and optimization steps: Analyze knowledge relevance and importance and update the graph;

[0040] S3. Decision execution and evaluation step: Real-time evaluation of decision execution capabilities and adjustment of paths;

[0041] S4. Decision verification and feedback step: Monitor data traffic fluctuations and optimize risk node paths;

[0042] S5. Adaptive adjustment step: dynamically adjust decision priorities and generate optimization reports.

[0043] Preferably, the knowledge graph construction and optimization steps include:

[0044] When the difference between the knowledge entropy of new information and the existing knowledge entropy exceeds a threshold, the graph reconstruction is automatically triggered;

[0045] The graph evolution algorithm is used to adjust the graph structure according to the knowledge relevance.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] The present invention significantly improves the efficiency of electronic information processing and the accuracy of decision-making through multi-dimensional technological innovation. At the data fusion level, the use of biological neuron-inspired pulse neural networks and the "data electrocardiogram" mechanism achieves efficient processing of multimodal heterogeneous data, solving the problem of insufficient adaptability of fixed window processing modes to dynamic flow rates in existing technologies. The dynamic knowledge graph engine introduces a "concept entropy" evaluation model and a dual-channel architecture, which enables the self-update speed of the knowledge graph to reach the minute level, completely changing the lag of traditional methods that rely on manual intervention, ensuring that the system can absorb new information and reconstruct the knowledge system in real time, and provide more accurate knowledge support for decision-making. Through implementation verification in scenarios such as emergency communications and industrial Internet of Things, the system has demonstrated significant advantages in key indicators such as resource scheduling efficiency, fault prediction accuracy, and anti-interference ability, providing a solution that is both innovative and engineering practical for the field of intelligent analysis and decision-making of electronic information. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a block diagram of the working modules of the present invention. DETAILED DESCRIPTION

[0049] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below in conjunction with specific implementation methods.

[0050] like Figure 1 As shown, the present invention discloses an electronic information intelligent analysis and decision-making system based on artificial intelligence, comprising:

[0051] Multimodal data fusion module, used to obtain electronic information data of multiple modes and divide it into high-frequency data and low-frequency data according to the processing frequency, and generate data fusion distribution results;

[0052] Dynamic knowledge graph construction and update module, used to analyze knowledge relevance and importance based on data fusion allocation results and generate optimized dynamic knowledge graph;

[0053] Intelligent decision-making module, used to extract decision node information from the dynamic knowledge graph and generate decision execution results;

[0054] The decision verification and feedback module is used to monitor changes in decision data flows and reallocate risk node paths, generating decision verification feedback results.

[0055] Among them, the multimodal data fusion module includes:

[0056] The data classification submodule is used to obtain the various modal data types and characteristics of electronic information, analyze and compare the processing frequency of each type of data, and divide each type of data into high-frequency data and low-frequency data based on the processing frequency to generate data classification results. It uses a frequency statistics algorithm to determine the processing frequency by analyzing the number of times data is processed within a period of time;

[0057] The feature fusion allocation submodule evaluates the load and fusion capability of each fusion node based on the data classification results, and uses the formula

[0058]

[0059] Calculate the fusion allocation results, combine the node load, fusion capability and data processing frequency differences, and generate preliminary fusion allocation results;

[0060] Among them, S represents the fusion capability of high-performance fusion nodes, H represents the fusion capability of ordinary fusion nodes, and F H With F L Represent the processing frequencies of high-frequency and low-frequency data respectively, L S With L H They represent the load conditions of high-performance fusion nodes and ordinary fusion nodes respectively, D represents the fusion demand, and R represents the fusion allocation result.

[0061] The fusion adjustment submodule monitors the real-time load changes of fusion nodes based on the preliminary fusion allocation results, analyzes the discrepancy between current fusion capabilities and demand, performs dynamic fusion allocation adjustments, and generates data fusion allocation results. It uses a dynamic programming algorithm to dynamically adjust data distribution across fusion nodes based on changes in node load and fusion capabilities.

[0062] As a further preferred embodiment of the present invention, the dynamic knowledge graph construction and update module includes:

[0063] The knowledge association analysis submodule, based on the data fusion allocation results, calls the data processing frequency and fusion characteristics of each fusion node, combines the priority and progress of the knowledge graph construction and update tasks, analyzes the node knowledge association and data processing requirements, calculates and obtains the knowledge association matching degree of each node, and generates the knowledge association matching degree. The cosine similarity algorithm is used to calculate the similarity of knowledge features between nodes to measure knowledge association;

[0064] The knowledge importance evaluation submodule builds and updates the priority and progress of tasks based on the knowledge graph, evaluates the knowledge carrying capacity of each fusion node, analyzes the storage capacity and load status of the node, and calculates the degree of match between the node knowledge importance and the task requirements.

[0065]

[0066] The calculation obtains the matching degree between the node knowledge importance and the task requirements, and generates the node knowledge importance evaluation result by combining the storage capacity, load status and task requirements;

[0067] Among them, M represents the matching degree between the importance of node knowledge and task requirements, C represents the storage capacity of the node, F represents the load rate of the node, D represents the processing time requirement of the task, and U represents the priority of the task;

[0068] The graph optimization submodule optimizes nodes in the dynamic knowledge graph based on the knowledge association matching and node knowledge importance assessment results. It analyzes the knowledge association and processing power matching between nodes, adjusts the storage resources and knowledge allocation of nodes in the graph, and generates an optimized dynamic knowledge graph. Using graph optimization algorithms, such as a variant of the PageRank algorithm, the graph structure is adjusted based on the knowledge importance and association of nodes.

[0069] As a further preferred embodiment of the present invention, the intelligent decision-making module includes:

[0070] The decision status extraction submodule is used to extract the decision status, decision progress information, and risk assessment of each decision node in the optimized dynamic knowledge graph, obtain the current execution progress of each decision, monitor the progress of node decisions, and obtain the decision execution processing status. Decision status extraction is achieved by regularly querying the status information and progress records of decision nodes.

[0071] The execution capability and progress comparison submodule compares the execution capability of the decision with the current progress according to the decision execution processing status, evaluates whether each node can complete the decision task on time, and calculates the gap between the execution capability and progress of the node using the formula:

[0072]

[0073] The calculation obtains the execution capability and progress gap of each node and generates a decision-making progress matching result.

[0074] Among them, Δ represents the gap between execution capability and progress, C represents the storage capacity of the node, L represents the load rate of the node, T represents the processing time requirement of the decision, and C T Represents the current progress of the decision, P T Represents the overall progress of decision-making;

[0075] The decision path adjustment submodule adjusts the decision path of each node in real time based on the decision progress matching results and the environmental status of the electronic information processing process (such as network conditions and device performance). Based on each node's execution capabilities and decision progress, the decision path is replanned and executed to generate the decision execution results. Reinforcement learning algorithms, such as the Deep Q Network (DQN), are used to dynamically adjust the decision path based on environmental status and decision progress.

[0076] As a further preferred embodiment of the present invention, the decision verification and feedback module includes:

[0077] The data collection submodule is used to monitor the decision data flow information of each node in the decision execution results at different time periods. Based on the decision processing capability and risk assessment, it collects flow information in real time and generates real-time decision data flow information. By setting data collection points on the decision execution path, decision data flow information is collected in real time.

[0078] The data analysis submodule analyzes the decision data flow change trend of each node based on the real-time decision data flow information, using the formula:

[0079]

[0080] The calculation obtains the traffic fluctuation percentage, compares it with the preset fluctuation threshold, filters out the risk decision nodes whose fluctuation exceeds the threshold, and generates a decision traffic fluctuation trend analysis report. Among them, ΔL represents the traffic fluctuation percentage, X i represents the node decision data flow at the current moment, X i-1 represents the node decision data flow at the previous moment, L avg represents the average processing delay of the node, Q T represents the total amount of decisions, and m represents the length of the time period;

[0081] The path adjustment submodule replans the execution paths of risky decision nodes based on the decision traffic fluctuation trend analysis report. It analyzes the decision traffic fluctuations at each risk node, optimizes the execution order between nodes, and generates decision verification feedback. Using a greedy algorithm, it selects the optimal execution path for adjustment based on the traffic fluctuations at the risk node.

[0082] As a further preferred embodiment of the present invention, the adaptive adjustment module includes:

[0083] The decision fluctuation monitoring submodule detects the decision fluctuation amplitude and trend of the current path node based on the decision verification feedback results, obtains the decision difference and volatility of each path node, and evaluates the decision fluctuation amplitude of the path node to generate the decision path flow assessment results. Statistical analysis methods are used to calculate the standard deviation and coefficient of variation of the decision data to evaluate the decision fluctuation amplitude.

[0084] The path load analysis submodule analyzes the load of the path nodes based on the decision path flow evaluation results, calculates the load index of each node, compares the relationship between the load and the decision fluctuation amplitude, and selects the path nodes with heavier loads to obtain the load ratio of the path nodes.

[0085]

[0086] Calculate the node load index and generate path node load data;

[0087] Among them, L k is the load index of the kth node, F k is the decision fluctuation amplitude value of the kth node, F j Decision flow for nodes, is the average value of decision traffic, n is the total number of nodes;

[0088] The decision scheduling optimization submodule analyzes the decision execution priority of the current path node based on path node load data and decision path traffic assessment results, calculates the need to adjust the decision scheduling sequence, adjusts the node decision scheduling sequence, determines whether the decision needs to be replanned, and generates a report on the adjusted decision optimization. It uses a priority queue algorithm to adjust the priority of decision tasks based on the load index and decision fluctuation range.

[0089] Based on the above, the present invention discloses an electronic information intelligent analysis and decision-making method; the method comprises the following steps:

[0090] S1. Data fusion and classification steps: divide data by processing frequency and dynamically adjust fusion allocation;

[0091] S2. Knowledge graph construction and optimization steps: Analyze knowledge relevance and importance and update the graph;

[0092] S3. Decision execution and evaluation step: Real-time evaluation of decision execution capabilities and adjustment of paths;

[0093] S4. Decision verification and feedback step: Monitor data traffic fluctuations and optimize risk node paths;

[0094] S5. Adaptive adjustment step: dynamically adjust decision priorities and generate optimization reports.

[0095] The knowledge graph construction and optimization steps include:

[0096] When the difference between the knowledge entropy of new information and the existing knowledge entropy exceeds a threshold, the graph reconstruction is automatically triggered;

[0097] The graph evolution algorithm is used to adjust the graph structure according to the knowledge relevance.

[0098] The present invention is further disclosed below with reference to specific examples:

[0099] Example: Real-time spectrum management and resource scheduling system for emergency communications scenarios

[0100] During sudden natural disasters or major incidents, emergency communication networks face challenges such as complex electromagnetic interference, coexistence of multiple signal standards, and dynamic resource demands. This embodiment deploys the system of the present invention to achieve real-time analysis and intelligent scheduling of spectrum resources, ensuring the stability and reliability of emergency communications.

[0101] System deployment architecture

[0102] Front-end perception layer

[0103] Deploy a distributed RF sensing node array (100+ nodes), support 20MHz-6GHz full-band signal acquisition, integrate a spike neural network (SNN) front-end preprocessing module, and extract signal characteristics such as time domain, frequency domain, and modulation mode in real time (sampling rate ≥ 100kHz).

[0104] The node has a built-in edge computing unit to perform preliminary noise reduction and feature compression on the original signal (compression ratio ≥ 80%), and transmit it to the middle station through the 5G private network.

[0105] Middle platform processing layer

[0106] Multimodal data fusion module:

[0107] The receiving perception layer receives RF signal characteristics (such as power spectrum density and modulation type), geographic information (base station location and terrain data), and emergency service requirements (voice and video transmission priority), and divides them into high-frequency real-time signals (such as emergency command voice) and low-frequency historical data (such as interference pattern library) according to the processing frequency.

[0108] The "feature fusion allocation submodule" formula is used to calculate the fusion node resource allocation. For example, high-frequency signals are allocated to high-performance GPU nodes (fusion capability S = 1000 MFLOPS), and low-frequency data are allocated to the CPU cluster (fusion capability H = 200 MFLOPS), dynamically adjusting the node load to below 70%.

[0109] Dynamic knowledge graph engine:

[0110] Build an "emergency communication knowledge graph" that includes static domain knowledge (spectrum allocation rules, equipment parameters) and dynamic situational memory (real-time interference source location, base station load status).

[0111] When the conflict degree between the newly detected interference signal and the existing knowledge (calculated by the "concept entropy" model) exceeds the threshold (such as the entropy value difference > 0.8), the map reconstruction is triggered and the interference source positioning and avoidance strategy is updated (reconstruction time < 20s).

[0112] Intelligent decision-making module:

[0113] Hierarchical decision-making unit execution scheduling:

[0114] Intuitive reaction layer (100ms level): Immediately trigger spectrum avoidance for sudden strong interference signals and switch to the backup frequency band (such as switching from 4G 1800MHz to 5G 3.5GHz).

[0115] Logical analysis layer (1s level): Optimizes base station channel allocation based on a historical interference pattern library and real-time load data. For example, video transmission services are prioritized to frequency bands with higher signal-to-noise ratios.

[0116] Strategic assessment layer (1-minute level): Combined with the evolution trend of disasters (such as the movement path of typhoons), the spectrum demand in the next 30 minutes is predicted and satellite communication resources are pre-allocated.

[0117] Back-end decision-making layer

[0118] Deploy digital twin emergency communication scenarios, map the front-end equipment status and spectrum usage in real time, and output decision explainability reports through the credibility verification module (such as "Due to continuous interference detected in the 5GHz band, base station No. 3 switched to the 2.4GHz band, in accordance with Article 4.3.2 of the GJB-9001C standard").

[0119] Technical Effect: Improved spectrum utilization: Through dynamic resource allocation, spectrum utilization efficiency is improved by 45% compared with traditional solutions, and the emergency service access success rate is increased from 70% to 95%.

[0120] Interference response speed: In sudden interference scenarios, the system's delay from signal perception to policy execution is less than 500ms, which is 80% shorter than manual intervention.

[0121] Decision traceability: The credibility verification module supports 72-hour decision path backtracking, meeting the audit requirements of military emergency communications.

[0122] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. An electronic information intelligent analysis and decision-making system based on artificial intelligence, characterized in that: include: Multimodal data fusion module, used to obtain electronic information data of multiple modes and divide it into high-frequency data and low-frequency data according to the processing frequency, and generate data fusion distribution results; Dynamic knowledge graph construction and update module, used to analyze knowledge relevance and importance based on data fusion allocation results and generate optimized dynamic knowledge graph; Intelligent decision-making module, used to extract decision node information from the dynamic knowledge graph and generate decision execution results; Decision verification and feedback module, used to monitor changes in decision data flow and reallocate risk node paths, generating decision verification feedback results; The adaptive adjustment module is used to dynamically adjust the decision execution priority according to the feedback results and generate a decision optimization report.

2. The electronic information intelligent analysis and decision-making system based on artificial intelligence according to claim 1 is characterized in that: The multimodal data fusion module includes: Data classification submodule, used to divide data into high-frequency data and low-frequency data according to processing frequency; The feature fusion allocation submodule uses the formula Calculate the fusion allocation result; where S represents the fusion capability of high-performance fusion nodes, H represents the fusion capability of ordinary fusion nodes, and F H With F L Represent the processing frequencies of high-frequency and low-frequency data respectively, L S With L H They represent the load conditions of high-performance fusion nodes and ordinary fusion nodes respectively, D represents the fusion demand, and R represents the fusion allocation result. The fusion adjustment submodule is used to monitor node load in real time and dynamically adjust data allocation results.

3. The electronic information intelligent analysis and decision-making system based on artificial intelligence according to claim 1 is characterized by: The dynamic knowledge graph construction and update module includes: The knowledge association analysis submodule is used to calculate the knowledge association matching degree between nodes; The knowledge importance evaluation submodule uses the formula Evaluate the node knowledge importance, where M represents the degree of match between the node knowledge importance and the task requirements, C represents the node storage capacity, F represents the node load rate, D represents the task processing time requirement, and U represents the task priority; The graph optimization submodule is used to adjust the graph node storage resources and knowledge allocation.

4. The artificial intelligence-based electronic information intelligent analysis and decision-making system according to claim 1, characterized in that: The intelligent decision-making module includes: Decision status extraction submodule, used to obtain the execution progress and risk assessment of decision nodes; The execution capability and progress comparison submodule is used to calculate the execution capability and progress gap; The decision path adjustment submodule is used to adjust the decision execution path in real time.

5. The electronic information intelligent analysis and decision-making system based on artificial intelligence according to claim 1 is characterized in that: The decision verification and feedback module includes: Data collection submodule, used to collect decision-making data flow information in real time; Data analysis submodule, used to calculate the traffic fluctuation percentage; The path adjustment submodule is used to optimize the execution path of risk nodes.

6. The electronic information intelligent analysis and decision-making system based on artificial intelligence according to claim 1 is characterized by: The adaptive adjustment module includes: Decision fluctuation monitoring submodule, used to evaluate the amplitude of decision fluctuation; The path load analysis submodule uses the formula Calculate the node load index, where L k is the load index of the kth node, F k is the decision fluctuation amplitude value of the kth node, F j Decision flow for nodes, is the average value of decision traffic, n is the total number of nodes; The decision scheduling optimization submodule is used to adjust the priority of decision tasks.

7. A method for intelligent analysis and decision-making of electronic information, based on the artificial intelligence-based intelligent analysis and decision-making system for electronic information according to any one of claims 1 to 6, characterized in that: The steps include: S1. Data fusion and classification steps: divide data by processing frequency and dynamically adjust fusion allocation; S2. Knowledge graph construction and optimization steps: Analyze knowledge relevance and importance and update the graph; S3. Decision execution and evaluation step: Real-time evaluation of decision execution capabilities and adjustment of paths; S4. Decision verification and feedback step: Monitor data traffic fluctuations and optimize risk node paths; S5. Adaptive adjustment step: dynamically adjust decision priorities and generate optimization reports.

8. The electronic information intelligent analysis and decision-making method according to claim 7, characterized in that: The knowledge graph construction and optimization steps include: When the difference between the knowledge entropy of new information and the existing knowledge entropy exceeds a threshold, the graph reconstruction is automatically triggered; The graph evolution algorithm is used to adjust the graph structure according to the knowledge relevance.

Images