Intelligent auxiliary system for enterprise management consultation based on artificial intelligence

Abstract

The invention discloses an intelligent auxiliary system for enterprise management consultation based on artificial intelligence, and belongs to the technical field of enterprise management and artificial intelligence. Comprising a data access layer, a data processing and storage layer, an intelligent analysis layer, a decision suggestion generation layer and a user interaction layer. The data access layer collects multi-source data, the multi-source data is transmitted to the data processing and storage layer for deep cleaning and encryption after primary processing, and a knowledge graph is constructed and stored. The intelligent analysis layer carries out deep analysis on the data by using multiple technologies, the decision suggestion generation layer generates and optimizes suggestions according to analysis results, and finally, the suggestions are displayed in multiple modes through the user interaction layer. The system fuses multi-source data, applies technologies such as quantum computing simulation and knowledge graph, integrates multiple learning technologies for collaborative analysis, introduces emotion and intention recognition to assist decision making, and improves interaction experience. The data value can be mined accurately, a reliable decision basis is provided, the data security is guaranteed, the management consultation efficiency is improved, and the sustainable development of enterprises is promoted.

Description

Technical Field

[0001] The present invention relates to the technical fields of enterprise management and artificial intelligence, and particularly to an intelligent auxiliary system for enterprise management consulting based on artificial intelligence. Background Art

[0002] Currently, in the field of enterprise management consulting, the traditional service model mainly relies on manual experience and simple data statistical analysis. This method has many limitations: on the one hand, the efficiency of manually processing a large amount of data is low, and it is difficult to quickly respond to the influx of massive information, resulting in a long analysis cycle and being unable to provide decision-making support for enterprises in a timely manner; on the other hand, simple statistical analysis is difficult to dig out the deep potential value behind the data, lacking accuracy, and easily causing enterprises to make wrong decisions in the complex and changeable market environment.

[0003] With the acceleration of the digital transformation of enterprises, a large amount of operation data has been generated within enterprises, covering multiple aspects such as production, sales, finance, and customer relationships. At the same time, data such as industry dynamics and competitor information in the external market environment is also increasing continuously. These data contain huge value, but the traditional model cannot process and analyze them efficiently, and it is difficult to convert the data into valuable management decision-making basis, resulting in a lack of strong support in the enterprise management decision-making process and restricting the development of enterprises and the improvement of their market competitiveness. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent auxiliary system for enterprise management consulting based on artificial intelligence, which solves the problems of low efficiency in processing a large amount of data by workers, difficulty in quickly responding to the influx of massive information, resulting in a long analysis cycle, lack of accuracy, and easily causing enterprises to make wrong decisions in the complex and changeable market environment.

[0005] To achieve the above purpose, the present invention provides an intelligent auxiliary system for enterprise management consulting based on artificial intelligence, including a data access layer, a data processing and storage layer, an intelligent analysis layer, a decision-making suggestion generation layer, and a user interaction layer. Each layer works together to provide comprehensive management consulting services for enterprises. Among them, the data access layer, the data processing and storage layer, the intelligent analysis layer, the decision-making suggestion generation layer, and the user interaction layer are arranged and connected in sequence from top to bottom.

[0006] Preferably, the internal system access module, the Internet of Things data acquisition module, the satellite image acquisition module, and the external data scraping module of the data access layer summarize the collected data into a unified data buffer area. After completing the preliminary data format verification and simple cleaning in the buffer area, the data is transmitted to the data processing and storage layer through the enterprise internal high-speed local area network.

[0007] Preferably, the cleaning module in the data processing and storage layer receives the data processed by the data access layer and performs in-depth cleaning. The cleaned data is diverted according to types and uses. Sensitive data is transmitted to the quantum computing simulation encryption module for encryption, and the remaining data is transmitted to the knowledge graph construction module. The encrypted data and the constructed knowledge graph are both stored in the data storage module.

[0008] Preferably, the intelligent analysis layer obtains data from the data processing and storage layer, and first performs type judgment and task analysis on the data. Data with a small volume and suitable for pre-trained models enters the transfer learning module; data that needs to expand the volume or detect anomalies enters the GAN data augmentation and anomaly detection module. The data processed by these two modules is then distributed to the traditional machine learning analysis module and the deep learning analysis module for in-depth analysis according to the data type and analysis task, and the analysis results are transmitted to the decision-making and recommendation generation layer.

[0009] Preferably, the sentiment analysis module and the intent recognition module in the decision-making and recommendation generation layer process the text data transmitted from the intelligent analysis layer in parallel, and input the processing results into the rule engine and the expert system module to generate preliminary management recommendations in combination with preset rules and experience. The reinforcement learning optimization module adjusts the policy parameters of the rule engine and the expert system module according to the enterprise's feedback on the recommendations, optimizes the recommendation generation, and finally transmits the optimized recommendations to the user interaction layer.

[0010] Preferably, the VR experience module, the AR-assisted decision-making module, the voice interaction module, and the visualization interface module in the user interaction layer are all connected to the decision-making and recommendation generation layer to obtain the analysis results and recommendations. The user inputs instructions through the voice interaction module, and after the system parses the instructions, it calls the corresponding module to display the results in the visualization interface, VR, or AR scenario.

[0011] Preferably, the system adopts multi-source data fusion technology to fuse Internet of Things, satellite images, enterprise internal systems, and external network data.

[0012] Preferably, the system uses quantum computing simulation technology for data encryption and acceleration processing, and combines knowledge graph technology to integrate multi-source data.

[0013] Preferably, the system integrates transfer learning, GAN, traditional machine learning, and deep learning technologies for collaborative intelligent analysis.

[0014] Preferably, the system introduces sentiment analysis and intent recognition technologies to assist management decision-making, and adopts VR, AR, and voice interaction technologies to improve the user interaction experience.

[0015] Therefore, an intelligent auxiliary system for enterprise management consulting based on artificial intelligence with the above structure of the present invention has the following beneficial effects:

[0016] (1) Through multi-source data fusion and intelligent analysis technologies, the present invention deeply explores the data value, provides more accurate management decision-making basis for enterprises, reduces decision-making errors, and improves the competitiveness of enterprises in the market. And it adopts quantum computing simulation encryption technology to ensure the security of enterprise sensitive data, prevent data leakage, and provide a safe and reliable data environment for enterprise operation.

[0017] (2) The present invention has an automated data processing and analysis process, as well as various interaction methods, reduces manual intervention, improves the efficiency of management consulting services, and enables enterprises to quickly respond to market changes.

[0018] (3) The emotional analysis and intention recognition from a humanistic perspective of the present invention, combined with the analysis results of business data, provide more comprehensive and user-friendly management suggestions for enterprises, and promote the sustainable development of enterprises. And it adopts an immersive interaction experience and an intuitive data display method to enhance the convenience and experience of user operation, and facilitate enterprise managers to obtain information and make decisions.

[0019] The technical solution of the present invention will be further described in detail below through the accompanying drawings and embodiments. Description of the Drawings

[0020] Figure 1 It is a schematic structural diagram of an intelligent auxiliary system for enterprise management consulting based on artificial intelligence according to the present invention;

[0021] Figure 2 It is a schematic flow diagram of an intelligent auxiliary system for enterprise management consulting based on artificial intelligence according to the present invention. Detailed Embodiment

[0022] The technical solution of the present invention will be further described below through the accompanying drawings and embodiments.

[0023] Unless otherwise defined, the technical terms or scientific terms used in the present invention shall have the ordinary meanings understood by those of ordinary skill in the field to which the present invention pertains. The "first", "second" and similar terms used in the present invention do not denote any order, quantity or importance, but are only used to distinguish different components. The terms such as "including" or "comprising" mean that the elements or objects appearing before this term cover the elements or objects listed after this term and their equivalents, without excluding other elements or objects. The terms such as "connected" or "linked" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "up", "down", "left", "right" are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0024] Embodiment

[0025] As Figure 1 shown, the present invention provides an intelligent auxiliary system for enterprise management consulting based on artificial intelligence, specifically as follows:

[0026] Data access layer

[0027] Module composition: The internal system access module is responsible for establishing stable data interfaces with the enterprise's existing ERP, CRM, financial systems, etc., to achieve real-time data extraction and transmission, ensuring the timely acquisition of enterprise internal data. The Internet of Things data collection module deploys various sensors in the production workshop, such as temperature and humidity sensors, equipment status sensors, etc., and uses RFID tags and positioning devices in the logistics link to collect data such as equipment operation status, production environment parameters, and cargo transportation trajectories, obtaining more detailed real-time data from the production and logistics links. The satellite image acquisition module obtains satellite images of the enterprise's large-scale warehouses or production bases through the interfaces of professional satellite image data providers, obtaining enterprise site-related information from a macroscopic perspective. The external data scraping module uses web crawler technology to collect information such as industry trends and market trends from news websites, industry forums, social media platforms, etc., broadening the data source channels.

[0028] Among them, each collection module aggregates the collected data to the unified data buffer in the data access layer. In the buffer area, the data first undergoes preliminary data format verification to check whether the data conforms to the system's preset format specifications, eliminating data with obvious incorrect formats to ensure data format consistency. Then, simple cleaning is performed to remove duplicate data and correct obvious incorrect data, initially improving the data quality. After the preliminary processing is completed, the data enters the data processing and storage layer through the enterprise's internal high-speed local area network in a stable and high-speed transmission manner.

[0029] The Internet of Things technology uses the MQTT protocol for data transmission. This protocol has the characteristics of strong real-time performance and low power consumption, suitable for the data transmission requirements of Internet of Things devices. Satellite image analysis uses professional image analysis software such as ENVI and Erdas, combined with deep learning image recognition algorithms, to accurately process and extract information from satellite images. The web crawler is developed based on the Scrapy framework of Python, combined with anti-crawler strategies such as setting reasonable crawling frequencies and using proxy IPs, to ensure efficient and stable acquisition of external data while complying with website rules.

[0030] Data processing and storage layer

[0031] Module Composition: The data cleaning module uses data mining algorithms and rule libraries to perform operations such as duplicate removal, outlier handling, and missing value filling on the data, further improving data quality. The quantum computing simulation encryption module uses a quantum computing simulation library to simulate quantum encryption algorithms on a traditional computer to encrypt sensitive data and ensure data security. The knowledge graph construction module uses algorithms such as entity recognition and relationship extraction in natural language processing technology, combined with enterprise business knowledge, to construct an enterprise knowledge graph and achieve the associated integration of data. The data storage module uses the Hadoop distributed file system to store massive amounts of raw data, uses a NoSQL database to store unstructured data, and a relational database to store structured data, and performs reasonable storage according to the characteristics of different types of data.

[0032] Among them, the data cleaning module receives the preliminarily processed data from the data access layer, uses complex data mining algorithms and rule libraries to deeply clean the data, removes duplicate records, corrects outliers, and fills missing values. The cleaned data is diverted according to data type and usage. Sensitive data is transmitted to the quantum computing simulation encryption module, which uses a quantum computing simulation library to simulate quantum encryption algorithms on a traditional computer to encrypt the sensitive data. The remaining data is transmitted to the knowledge graph construction module, which uses algorithms such as entity recognition and relationship extraction in natural language processing technology, combined with enterprise business knowledge, to construct an enterprise knowledge graph. The encrypted data and the constructed knowledge graph are uniformly stored in the data storage module, where HDFS stores massive amounts of raw data, MongoDB stores unstructured data, and MySQL stores structured data.

[0033] Quantum computing simulation technology realizes a more complex and secure encryption method than traditional encryption algorithms by simulating the superposition and entanglement characteristics of quantum bits. The knowledge graph construction uses the Neo4j graph database for storage and management, and performs knowledge query and reasoning through graph traversal algorithms, which can quickly obtain associated information and provide strong support for subsequent analysis.

[0034] Intelligent Analysis Layer

[0035] Module Composition: The transfer learning module utilizes pre-trained deep learning models in other similar enterprises or industries (such as the BERT model pre-trained in the field of natural language processing, the ResNet model pre-trained in the field of image recognition, etc.), and combines a small amount of the enterprise's own data for fine-tuning training to improve the model training efficiency. The GAN data augmentation and anomaly detection module applies the generative adversarial network (GAN). Through the adversarial training of the generator and discriminator, it generates synthetic data to expand the training set and detects abnormal patterns in the data, enhancing the usability of the data and the ability to discover anomalies. The traditional machine learning analysis module uses classic algorithms such as decision trees, random forests, and support vector machines for data classification, clustering, and prediction, providing multi-dimensional analysis methods. The deep learning analysis module uses convolutional neural networks (CNNs) to process image data, recurrent neural networks (RNNs) to process time series data and text data, and conducts in-depth analysis for different types of data.

[0036] After obtaining data from the data processing and storage layer, the system first performs data type judgment and task analysis. For data with a small volume and similar to the applicable scenarios of the pre-trained model, it is transmitted to the transfer learning module, and a pre-trained deep learning model is used to combine a small amount of the enterprise's own data for fine-tuning training; for data that needs to expand the data volume or detect anomalies, it is transmitted to the GAN data augmentation and anomaly detection module, and synthetic data is generated through the adversarial training of the generator and discriminator to expand the training set and detect abnormal patterns. The data processed by the transfer learning module and the GAN data augmentation and anomaly detection module is distributed to the traditional machine learning analysis module and the deep learning analysis module for in-depth analysis according to the data type (image, time series, text, etc.) and analysis tasks (classification, clustering, prediction, etc.), and the final analysis results are transmitted to the decision-making and recommendation generation layer.

[0037] Transfer learning can quickly converge to a model suitable for the enterprise's specific tasks by transferring the network structure and some parameters of the pre-trained model, reducing the training time and cost. GAN uses the adversarial training mechanism to continuously optimize the generator and discriminator, improving the quality of synthetic data and the accuracy of anomaly detection.

[0038] Decision-making and Recommendation Generation Layer

[0039] Module Composition: The sentiment analysis module uses a pre-trained sentiment analysis model, specifically a BERT-based sentiment analysis model, to determine the sentiment polarity of text data such as internal corporate communication records, employee feedback, and customer evaluations, thereby understanding the attitudes within and outside the enterprise from a humanistic perspective. The intention recognition module uses the LSTM-CRF model of deep learning to understand the behavioral intentions of corporate managers and employees, providing forward-looking information for management decisions. The rule engine and expert system module generate preliminary management suggestions based on the analysis results and preset business rules and expert experience, and give suggestions by combining actual business and professional knowledge. The reinforcement learning optimization module uses the DQN reinforcement learning algorithm in the deep Q network to dynamically adjust the suggestion generation strategy according to the enterprise's feedback on the suggestions, making the suggestions more in line with the actual needs of the enterprise.

[0040] Among them, the sentiment analysis module and the intention recognition module process the text data transmitted from the intelligent analysis layer in parallel. The sentiment analysis module determines the sentiment polarity of the text data, and the intention recognition module understands the behavioral intentions. The processing results of both are jointly input into the rule engine and expert system module, which combines the preset business rules and expert experience to generate preliminary management suggestions. The reinforcement learning optimization module continuously receives the enterprise's feedback on the suggestions, converts the feedback information into a reward signal, and adjusts the policy parameters of the rule engine and expert system module through reinforcement learning algorithms such as the deep Q network (DQN) to optimize the suggestion generation. Finally, the optimized suggestions are transmitted to the user interaction layer.

[0041] Sentiment analysis and intention recognition utilize the semantic understanding ability of deep learning models for text to achieve high-precision sentiment judgment and intention prediction. Reinforcement learning optimizes the suggestion generation strategy through continuous trial and error and reward feedback, enabling the system to continuously improve the suggestions according to the actual situation.

[0042] User Interaction Layer

[0043] Module Composition: The VR experience module is developed based on game engines such as Unity or Unreal Engine, combined with VR devices, to provide users with an immersive enterprise operation simulation experience, allowing users to intuitively feel the enterprise operation status. The AR-assisted decision-making module is developed using ARCore (for Android) or ARKit (for iOS), which superimposes virtual analysis data and suggestions on the real scene, facilitating users to obtain information and make decisions in the actual scene. The voice interaction module uses technologies such as Baidu Speech Recognition and iFlytek Speech Recognition to achieve the recognition and parsing of voice commands, enhancing the convenience of interaction. The visual interface module develops Web and mobile applications, using visualization libraries such as Echarts and D3.js to display the analysis results and suggestions, presenting the data in an intuitive chart form.

[0044] Among them, the VR experience module, AR auxiliary decision-making module, voice interaction module, and visualization interface module are all connected to the decision-making recommendation generation layer to obtain analysis results and recommendations. The user inputs instructions through the voice interaction module. The voice instructions are first converted into text by voice recognition technology, and then the meaning of the instructions is parsed by natural language understanding technology. The system calls the corresponding module according to the meaning of the instructions. If the instruction requires immersive experience data display, the VR experience module is called; if real-world scenario auxiliary decision-making is needed, the AR auxiliary decision-making module is called; if intuitive data display is required, the visualization interface module is called, and finally the results are displayed in the corresponding module.

[0045] VR and AR technologies utilize technologies such as 3D modeling and spatial positioning to provide users with intuitive and highly interactive experiences. Voice interaction technology realizes natural human-machine interaction through technologies such as voice recognition and natural language understanding, improving the convenience and efficiency of user operations.

[0046] As Figure 2 shown, the present invention also discloses an implementation method of an intelligent auxiliary system for enterprise management consulting based on artificial intelligence, including the following steps:

[0047] S1. System deployment

[0048] Deployment of enterprise internal server clusters: The core components of the data processing and storage layer and the intelligent analysis layer have high demands for computing and storage resources, and enterprise internal server clusters can provide stable and powerful computing and storage capabilities. During deployment, server resources need to be reasonably planned. For example, according to the data volume and the scale of analysis tasks, sufficient disk space is allocated for data storage, and appropriate CPU and memory resources are allocated for model training and data processing. Install and configure software and tools such as Hadoop Distributed File System (HDFS), NoSQL databases (such as MongoDB), relational databases (such as MySQL), quantum computing simulation libraries (such as Qiskit), and various machine learning and deep learning frameworks (such as TensorFlow, PyTorch) to ensure that each component can operate stably and cooperate efficiently.

[0049] Cloud server deployment: Some external data collection modules of the data access layer need to have flexible network access capabilities to meet the collection requirements of different data sources; the applications of the user interaction layer must ensure that a large number of users can easily access them. Choose a suitable cloud service provider, such as Alibaba Cloud, Tencent Cloud, etc., and select the corresponding cloud server configuration according to the estimated access volume and data collection scale. Deploy external data collection modules such as web crawlers and satellite image data receiving and parsing programs on the cloud server, and use the elastic computing and storage services of the cloud server to easily adjust resources to cope with fluctuations in data volume. Develop and deploy web and mobile applications, and use the CDN (content distribution network) service of the cloud server to improve the access speed and stability of the user interaction layer applications to ensure that users can use the system smoothly.

[0050] S2. Data collection and preprocessing

[0051] Data collection: Based on the design of the data access layer, the internal system access module regularly extracts data from the company's ERP, CRM, financial system and other internal systems. The IoT data collection module collects data from production workshop equipment and logistics links at a set frequency. The satellite image acquisition module obtains satellite image data at an agreed time interval. The external data capture module collects data from news websites, industry forums, and social media platforms according to preset rules. In order to ensure the integrity and accuracy of the data, a strict data collection plan must be formulated to clarify the collection time, frequency, and data range, monitor the collection process, and promptly handle failed or abnormal data.

[0052] Data preprocessing: After the collected data is transferred to the data processing and storage layer, the data cleaning module uses data mining algorithms and rule bases to perform operations such as deduplication, outlier processing, and missing value filling. Use the Pandas library to write code to achieve data deduplication, set rules according to business logic to identify and correct outliers, and use the mean, median, or machine learning algorithm-based methods to fill in missing values. The quantum computing simulation encryption module encrypts sensitive data and selects appropriate quantum encryption algorithms and parameters, such as determining the number of quantum bits and encryption key length. The knowledge graph construction module uses entity recognition and relationship extraction algorithms in natural language processing technology, combined with corporate business knowledge, to build an enterprise knowledge graph and integrate data into a structured knowledge network for subsequent analysis and query.

[0053] S3, model training and optimization

[0054] Model training: The intelligent analysis layer is designed based on the algorithm and uses the collected and preprocessed data to train the model. The transfer learning module selects deep learning models that have been pre-trained in similar enterprises or industries, such as BERT or ResNet, and fine-tunes the training with the enterprise's own data, setting appropriate learning rate, number of training rounds and other parameters to adapt the model to the enterprise's specific tasks. The GAN data enhancement and anomaly detection module writes the generator and discriminator code, generates synthetic data through adversarial training to expand the training set, detect abnormal patterns, and adjust the network structure and training parameters to improve the quality of synthetic data and the accuracy of anomaly detection. The traditional machine learning analysis module uses algorithms such as decision trees and random forests, and selects appropriate algorithms and parameters according to the characteristics of the data to achieve data classification, clustering and prediction. The deep learning analysis module uses CNN and RNN to process different types of data, designs the network structure, and sets parameters such as the number of layers, number of neurons and activation functions for training.

[0055] Optimize suggestion generation strategy: The decision suggestion generation layer uses reinforcement learning to optimize the suggestion generation strategy. Reinforcement learning algorithms such as the Deep Q Network (DQN) use the company's feedback on the suggestions as a reward signal to adjust the strategy parameters of the rule engine and expert system modules. If the company adopts the suggestions and achieves positive results, it will be given a positive reward to encourage the system to strengthen the suggestion generation strategy; otherwise, it will be given a negative reward to guide the system to adjust the strategy. Through continuous trial and error and optimization, the suggestions are more in line with the actual needs of the enterprise.

[0056] S4. User interaction and feedback

[0057] User interaction: Users interact with the system in a variety of ways in the user interaction layer. When using the voice interaction module, voice commands are converted into text through technologies such as Baidu voice recognition and iFLYTEK voice recognition, and then the meaning is parsed through natural language understanding technology. The system calls the corresponding module according to the command and displays the results in the visual interface, VR or AR scene. In the VR experience module, users use VR devices to immersively view enterprise operation data and analysis results; in the AR decision-making assistance module, virtual analysis data and suggestions are superimposed on the real scene to assist decision-making; the visual interface module displays data in intuitive charts for easy viewing and operation by users.

[0058] Feedback collection and system optimization: The system collects user feedback during use, including evaluation of functions, feelings about operating experience, and opinions on analysis results and suggestions. Feedback information is collected by setting up feedback portals in the system, such as online questionnaires and feedback forms. Data analysis tools are developed to analyze feedback data and extract key issues and user needs. Based on the feedback analysis results, the system is optimized in a targeted manner, such as improving algorithms to improve analysis accuracy, optimizing interface design to improve user experience, and adjusting suggestion generation strategies to make suggestions more practical, so as to continuously improve the service quality of the system.

[0059] Therefore, the present invention adopts the above-mentioned intelligent auxiliary system for enterprise management consulting based on artificial intelligence. In terms of data processing, it integrates Internet of Things, satellite images, and data from internal and external enterprise systems to broaden the data dimension. After cleaning, encryption, and knowledge graph construction, the data quality and security are improved, facilitating in-depth analysis. At the intelligent analysis level, it integrates transfer learning, GAN, traditional and deep learning technologies to conduct collaborative analysis on different data, improving the analysis accuracy and efficiency and quickly mining the data value. In terms of generating decision suggestions, it introduces emotion and intention recognition technologies, generates suggestions by combining business data and expert experience, and optimizes according to feedback, making the decision more user-friendly and in line with the actual situation of the enterprise. In terms of user interaction, it uses VR, AR, and voice interaction technologies to provide an immersive and convenient interaction experience, facilitating managers to obtain information and make decisions. Overall, the system improves the decision-making accuracy, ensures data security, improves management efficiency, and promotes the sustainable development of the enterprise.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that they can still modify or equivalently replace the technical solutions of the present invention, and these modifications or equivalent replacements do not make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. An enterprise management consulting intelligent assistance system based on artificial intelligence, characterized in that: It includes a data access layer, a data processing and storage layer, an intelligent analysis layer, a decision suggestion generation layer and a user interaction layer; wherein the data access layer, the data processing and storage layer, the intelligent analysis layer, the decision suggestion generation layer and the user interaction layer are arranged and connected in sequence from top to bottom.

2. The enterprise management consulting intelligent assistance system based on artificial intelligence according to claim 1 is characterized by: The data access layer includes an internal system access module, an Internet of Things data acquisition module, a satellite image acquisition module and an external data capture module. Each module aggregates the collected data into a unified data cache area and transmits it to the data processing and storage layer after preliminary processing.

3. The enterprise management consulting intelligent assistance system based on artificial intelligence according to claim 1 is characterized by: The data processing and storage layer includes a data cleaning module, a quantum computing simulation encryption module, a knowledge graph construction module and a data storage module. The data cleaning module receives data from the data access layer, and the cleaned data is diverted to the quantum computing simulation encryption module and the knowledge graph construction module. The encrypted data and the constructed knowledge graph are stored in the data storage module.

4. The enterprise management consulting intelligent assistance system based on artificial intelligence according to claim 1 is characterized by: The intelligent analysis layer includes a transfer learning module, a GAN data enhancement and anomaly detection module, a traditional machine learning analysis module and a deep learning analysis module. After acquiring data from the data processing and storage layer, it is distributed to the corresponding modules according to the data type and analysis task. The processed data is then deeply analyzed and the results are transmitted to the decision recommendation generation layer.

5. The enterprise management consulting intelligent assistance system based on artificial intelligence according to claim 1 is characterized by: The decision suggestion generation layer includes a sentiment analysis module, an intention recognition module, a rule engine and expert system module, and a reinforcement learning optimization module. The sentiment analysis module and the intention recognition module process the results and input them into the rule engine and expert system module to generate preliminary suggestions. The reinforcement learning optimization module generates strategies based on enterprise feedback optimization suggestions, and the final suggestions are transmitted to the user interaction layer.

6. The enterprise management consulting intelligent assistance system based on artificial intelligence according to claim 1 is characterized by: The user interaction layer includes a VR experience module, an AR decision-making assistance module, a voice interaction module and a visualization interface module. Each module is connected to the decision suggestion generation layer and displays analysis results and suggestions according to user instructions.

7. An enterprise management consulting intelligent assistance system based on artificial intelligence according to any one of claims 1 to 6, characterized in that: The system adopts multi-source data fusion technology to integrate the Internet of Things, satellite images, enterprise internal systems and external network data.

8. An enterprise management consulting intelligent assistance system based on artificial intelligence according to any one of claims 1 to 6, characterized in that: The system uses quantum computing simulation technology to encrypt and accelerate data processing, and combines knowledge graph technology to integrate multi-source data.

9. An enterprise management consulting intelligent assistance system based on artificial intelligence according to any one of claims 1 to 6, characterized in that: The system integrates transfer learning, GAN, traditional machine learning and deep learning technologies for collaborative intelligent analysis.

10. An enterprise management consulting intelligent assistance system based on artificial intelligence according to any one of claims 1 to 6, characterized in that: The system introduces sentiment analysis and intent recognition technology to assist management decisions, and uses VR, AR and voice interaction technology to enhance user interaction experience.

Images