Enterprise management optimization method and system based on artificial intelligence technology

Abstract

The invention relates to the technical field of enterprise management analysis, and discloses an enterprise management optimization method and system based on an artificial intelligence technology, and the method comprises the steps: obtaining an enterprise project group architecture and historical business information, constructing a project group simulation unit, and analyzing the time relation of the historical business information; a service information flow is generated and mapped to a project team simulation unit, a service execution force curve is obtained, long-term trend characteristics and common fluctuation are extracted, a pressure simulation environment is constructed, the execution force curve is predicted based on pressure simulation and trend characteristics, a task distribution framework is constructed according to the execution force prediction curve, task distribution simulation is carried out, and an optimization management decision is obtained. According to the method, the accuracy and dynamic adaptability of task allocation are improved, the enterprise resource utilization efficiency is optimized, intelligent decision is supported, the overall management efficiency is improved, and the problem that in the prior art, task allocation depending on artificial experience is difficult to adapt to dynamic changes is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise management analysis, and in particular to an enterprise management optimization method and system based on artificial intelligence technology. Background Art

[0002] As modern enterprises operate in an increasingly complex market environment, task management and resource scheduling optimization have become crucial aspects of enterprise management, especially for large enterprises with multiple projects, multiple tasks and dynamic changes. Traditional task allocation and decision-making methods often fail to meet the needs of efficient and accurate management. Traditional task allocation methods usually rely on manual experience, simple rules or analysis based on static data, and are often unable to effectively cope with the complex and dynamically changing situations in business execution. Summary of the Invention

[0003] The purpose of the present invention is to provide an enterprise management optimization method and system based on artificial intelligence technology, aiming to solve the problem that the existing technology relies on manual experience to allocate tasks and is difficult to adapt to dynamic changes.

[0004] The present invention is implemented as follows: In a first aspect, the present invention provides an enterprise management optimization method based on artificial intelligence technology, comprising: Obtaining the architecture information and historical business information of each project group of the enterprise, and building a project group simulation unit based on the architecture information; Analyzing the time relationship of the historical business information to generate a business information flow, and mapping the project group simulation unit information according to the business information flow to obtain a business execution curve; extracting long-term trend characteristics of each of the business execution capability curves and common fluctuation characteristics between the business execution capability curves to construct a pressure simulation environment; Performing an expected simulation on the business execution capability curve based on the pressure simulation environment and the long-term trend characteristics to generate an execution capability prediction curve; An enterprise task allocation framework is constructed based on each of the execution prediction curves to simulate task allocation for current business information and obtain optimized management decisions.

[0005] In a second aspect, the present invention provides an enterprise management optimization system based on artificial intelligence technology, which is used to implement the enterprise management optimization method based on artificial intelligence technology described in any one of the first aspects, including: The architecture simulation module is used to obtain the architecture information and historical business information of each project group of the enterprise and build a project group simulation unit based on the architecture information; A capability analysis module is used to analyze the time relationship of the historical business information to generate a business information flow, and to perform information mapping on the project group simulation unit according to the business information flow to obtain a business execution capability curve; A pressure simulation module, configured to extract the long-term trend characteristics of each of the business execution curves and the common fluctuation characteristics between the business execution curves, so as to construct a pressure simulation environment; A capability prediction module, configured to perform an expected simulation on the business execution capability curve based on the pressure simulation environment and the long-term trend characteristics to generate an execution capability prediction curve; The task allocation module is used to construct an enterprise task allocation framework based on each of the execution prediction curves to simulate task allocation for current business information and obtain optimized management decisions.

[0006] The present invention provides an enterprise management optimization method based on artificial intelligence technology, which has the following beneficial effects: The present invention obtains the enterprise project team architecture and historical business information, constructs a project team simulation unit, analyzes the time relationship of historical business information, generates a business information flow and maps it to the project team simulation unit, obtains a business execution curve, uses a graph neural convolutional network to extract long-term trend characteristics and common fluctuations, constructs a pressure simulation environment, and uses a long short-term memory network to predict the execution curve based on pressure simulation and trend characteristics. A task allocation framework is constructed according to the execution prediction curve, and task allocation simulation is performed to obtain optimized management decisions. This method improves the accuracy and dynamic adaptability of task allocation, optimizes the efficiency of enterprise resource utilization, supports intelligent decision-making, improves overall management efficiency, and solves the problem in the existing technology that tasks allocated by relying on manual experience are difficult to adapt to dynamic changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 This is a schematic diagram of the steps of an enterprise management optimization method based on artificial intelligence technology provided by an embodiment of the present invention; Figure 2 This is a structural diagram of an enterprise management optimization system based on artificial intelligence technology provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0008] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0009] The implementation of the present invention is described in detail below with reference to specific embodiments.

[0010] Reference Figure 1 、 Figure 2As shown, a preferred embodiment of the present invention is provided.

[0011] In a first aspect, the present invention provides an enterprise management optimization method based on artificial intelligence technology, comprising: S1: Obtain the architecture information and historical business information of each project group of the enterprise, and build a project group simulation unit based on the architecture information; S2: parsing the time relationship of the historical business information to generate a business information flow, and performing information mapping on the project group simulation unit according to the business information flow to obtain a business execution curve; S3: extracting the long-term trend characteristics of each of the business execution capability curves and the common fluctuation characteristics between the business execution capability curves through a graph neural convolutional network model to construct a pressure simulation environment; S4: performing an expected simulation on the business execution capability curve based on the pressure simulation environment and the long-term trend characteristics using a long short-term memory network model to generate an execution capability prediction curve; S5: Constructing an enterprise task allocation framework based on each of the execution prediction curves to simulate task allocation for current business information and obtain optimized management decisions.

[0012] Specifically, in step S1 of the embodiment provided by the present invention, the basic architecture of the enterprise is collected, and the internal architecture information and parallel architecture information of each project group in the enterprise are collected, which includes the organizational structure, function allocation, staffing and other information of each project group. By fully understanding the architecture of the enterprise, the internal structure of each project group and its collaborative relationship with other groups can be clearly identified. The architecture information is the basis for subsequent simulation and optimization, ensuring the accuracy and comprehensiveness of the simulation. This step provides basic data support for subsequent simulation and analysis, ensuring that the simulation unit can fully reflect the complexity of the enterprise architecture, thereby providing reliable input for optimization management.

[0013] More specifically, based on the company's historical database, the architectural information of each project team is traced, and the architectural change records and historical business information of the project team are obtained. This step includes retrieving past organizational changes, project progress, and the management model adopted. The company's architecture and business model are dynamically changing. Understanding past architectural changes is critical to understanding the work dynamics of the project team and how it evolves over time. By tracing historical changes, more accurate time series data can be provided for the simulation unit. This step helps to establish a time evolution model of the architecture, so that the simulation unit not only reflects the current status, but also takes into account past business impacts. This can improve the accuracy of the model and help simulate and predict future development trends.

[0014] More specifically, based on the project team's internal architecture information, basic simulation units are constructed for each project team. Each simulation unit represents the project team's functions and task configuration at a specific point in time. As an abstract model, the simulation unit can represent complex organizational structures and business processes in a computable form. This allows for rapid analysis and optimization of enterprise resource usage and allocation without excessive intervention. This step converts abstract architectural information into simulation units, providing a foundation for further analysis and optimization. The simulation units provide reliable computational support for subsequent task allocation and decision optimization.

[0015] More specifically, according to the parallel architecture information of the project team, the collaborative relationship of the basic simulation units is adjusted, that is, by considering the interdependence and collaboration between project teams, the interactive relationship between each simulation unit is adjusted. The various project teams in the enterprise often need to cooperate with each other to complete the task. By accurately modeling this collaborative relationship, the dynamic interaction between project teams can be better simulated. Correcting the collaborative relationship is a key step to improve the accuracy of the simulation. The corrected simulation unit can more accurately reflect the collaboration mechanism between project teams, ensure that the simulation results are more in line with the actual operation situation, and avoid ignoring important interaction factors.

[0016] More specifically, based on the architecture change records, the architecture simulation units of the current period are deduced from past periods to generate architecture simulation units for each past period. Through deduction, the impact of historical architecture changes is reflected in the current simulation. The company's historical architecture changes have a profound impact on current and future work. Deducing historical architecture changes helps identify the long-term impact of past decisions on the operation of the current project team and avoids blindly adopting the current architecture for decision-making. Through architectural deduction of past periods, we can better understand the impact of different architecture choices, which can help decision makers identify which historical operations are most effective and which changes may have negative impacts, thereby improving the quality of decision-making.

[0017] More specifically, the deduced historical architecture simulation unit is combined with the architecture simulation unit of the current period to generate a complete project team simulation unit. The combined simulation unit can span different time periods and comprehensively display the architectural evolution of the project team. Combining the architectures of the past and current periods can provide a full-perspective model that takes into account both the current architectural configuration and past business models and decisions. This can provide a more complete management view and ensure that all influencing factors are considered when making decisions. This step enables the simulation unit to not only represent the current architecture, but also reflect the dynamic changes of the architecture, helping managers understand how past decisions affect current operations. Through a comprehensive time perspective, optimized management decisions will be more forward-looking and accurate.

[0018] It is understandable that the entire chain of steps forms an accurate simulation unit covering historical and current status by gradually collecting, tracing and deducing architectural information. The technical effect of each step helps to improve the accuracy and practicality of the simulation model, ensuring that the management optimization of the project team is based on complete and accurate information. This method can help enterprises efficiently formulate task allocation, resource optimization and decision-making processes.

[0019] Specifically, in step S2 of the embodiment provided by the present invention, business information is collected and organized from the enterprise's historical business database. The historical business information includes multiple dimensions such as project content, start time, end time, funds used, task completion status, staffing, task evaluation, etc. The purpose of obtaining historical business information is to understand the business execution status of the project team in different time periods. This information provides original data for subsequent time relationship analysis and execution analysis. Through comprehensive historical data, a specific business background can be provided for each simulation unit. This step provides detailed historical data support for subsequent analysis. The collected information not only reflects the completion status of the project, but also reveals the efficiency and problems of project execution, providing key basis for subsequent analysis.

[0020] More specifically, the collected historical business information is analyzed for temporal relationships to determine the relative order and dependencies between various business information, including the comparison of the start and end time of tasks, the dependencies between tasks, the delivery time of each link, etc. Project execution usually has a temporal sequence and dependencies. Understanding these temporal relationships can help identify bottlenecks, delays or inefficient processes that occur during project execution. Through time analysis, a structured time dimension can be provided for business processes to ensure the orderliness of business information flow. Time relationship analysis helps establish an accurate business time series. This step ensures an in-depth understanding of the time scheduling, priority sorting and delivery cycle of each project group and task, and can truly reflect the time logic of project execution in the simulation.

[0021] More specifically, based on the results of the time relationship analysis, a business information flow is generated. The business information flow refers to organizing the tasks, activities and execution status of the project team into a linear or nonlinear flow data chain in chronological order. The business information flow includes information on the start and end time of the task, the participants, the resources used, the task completion rate, the budget usage and other dimensions. The business information flow helps to comprehensively present the key activities, resource usage and project progress in the project execution process. By constructing the information flow, the progress of task completion, time consumption, resource waste and other issues can be more clearly identified. This step organizes various historical business information into an easy-to-understand flow data chain, which provides basic data for the subsequent analysis of execution capabilities, can intuitively display the actual execution status of the project, and help better optimize management decisions.

[0022] More specifically, the generated business information flow is mapped to the simulation unit of the project group. Specifically, the historical business information flow is matched with the architecture, task allocation, staffing, etc. in the current project group simulation unit to reflect the actual performance of each project group in the business execution process. The mapping process needs to consider the project group's time window, resource allocation, personnel task participation, etc. to ensure the correlation and accuracy between the information flow and the simulation unit. The project group simulation unit needs to be combined with real business data. Only through mapping can the theoretical simulation be compared with the actual execution situation, thereby improving the accuracy and practicality of the model. Through the mapping of information flow and simulation unit, the model can accurately reflect the actual performance of each project group in the business execution process, which ensures that the simulation unit can not only describe the project architecture, but also reflect its actual task execution capabilities, providing a solid foundation for execution analysis.

[0023] More specifically, based on the mapping results of information flow and simulation units, the business execution curve of the project team is calculated. The business execution curve is a dynamic curve that reflects the efficiency and effectiveness of the project team during the execution process. The curve is usually generated by analyzing factors such as task completion, time consumption, resource utilization efficiency, and personnel performance to comprehensively derive the execution performance of the project team. Commonly used algorithms include weighted average method, scoring model or machine learning model. The business execution curve is the core indicator for evaluating the performance of the project team. Through this curve, the execution performance of the project team at each time point can be intuitively observed, thereby helping decision makers to determine which parts need to be optimized and which resources need to be reconfigured. The generated business execution curve provides the enterprise with a dynamic performance evaluation tool. It not only shows the execution status of the project team at each stage, but also reveals the efficiency bottlenecks and potential problems in task completion, providing decision support for subsequent resource optimization and task allocation.

[0024] More specifically, the generated business execution curve is further analyzed to find fluctuations or bottlenecks in the curve and identify stages or tasks with weak execution. Data mining, regression analysis and other technologies can be used to analyze the trend of the business execution curve. Based on the analysis results, optimization suggestions are made for the project team's task allocation, staffing, resource utilization and other aspects. The analysis and optimization stage helps management better understand the practical significance of the business execution curve. Through in-depth analysis of the execution curve, problems in project execution can be discovered and practical optimization plans can be proposed, thereby improving the overall efficiency of the enterprise. The analysis and optimization process at this stage enables the enterprise to identify potential management or execution problems, so as to take targeted measures to improve the overall business execution and project team efficiency.

[0025] It is understandable that this entire process ultimately results in a business execution curve through temporal relationship analysis of historical business information, generation of business information flows, and mapping with project team simulation units. In each step, accurate data analysis and mapping provide a solid foundation for subsequent analysis and optimization, thereby helping managers better understand project execution and provide effective optimization suggestions.

[0026] Specifically, in step S3 of the embodiment provided by the present invention, each business execution curve is regarded as a node in the graph; the characteristic vector of the node is its execution time series (or its embedding); edges are established according to business relevance (such as department commonality, task similarity, time overlap, etc.) to form a business execution graph. The advantage of GCN is that it can mine the "graph" relationship of structured data. The graph structure must be constructed first before the graph convolution operation can be performed; the edges between nodes can transmit related information, mine the resonance, coupling, dependency and other relationships between execution forces, and construct a graph structure that reflects the coupling and collaborative mode between business execution forces, providing structural information for downstream GCN training.

[0027] More specifically, each node in the graph (i.e., the business execution curve) is input into a standard GCN; GCN calculates the representation of each node (i.e., long-term trend features) through graph structure propagation and convolution; the output is the global trend embedding vector of each curve. Through the hierarchical propagation mechanism, GCN can integrate the features of each node and its neighboring nodes, thereby extracting information reflecting the collective long-term trend; this trend is usually a manifestation of stability, such as business cyclicality, growth or decline patterns, generating deep feature vectors reflecting the long-term behavior patterns of business groups or tasks, which can be used to judge the sensitivity of the impact of pressure changes on trends.

[0028] More specifically, a time-series graph neural network (ST-GCN / T-GCN / Gated GCN) is constructed. Each time slice constitutes a layer of the graph, and the node represents the execution force value at that point in time. The model propagates through the time dimension to capture the common patterns of fluctuations in each curve (such as sudden declines, periodic oscillations, etc.). It is difficult to model dynamic fluctuations in the time dimension using only ordinary GCN. The time-series graph neural network can learn the coordinated fluctuation patterns of execution force over time, identify nodes that may resonate due to external or systemic pressure, and extract the common fluctuation characteristics of multiple business curves that may be affected by the same pressure factors in the same time period. This is conducive to building a realistic and controllable pressure scenario simulation model.

[0029] More specifically, the long-term trend characteristics (static) and fluctuation characteristics (dynamic) are integrated through feature splicing, attention mechanism, contrast loss function, etc.; cluster analysis is performed on the curves of different industries / departments / task types to find highly similar groups. Constructing a stress simulation environment requires referring to both static trends (long-term stress tolerance) and dynamic fluctuations (short-term stress sensitivity); cluster analysis can help discover which project groups have similar stress resistance or stress transmission paths, and obtain a set of business curve models with strong interpretability and obvious differences in stress response; support differentiated pressure and personalized response strategy design in the simulation environment.

[0030] More specifically, based on the extracted features, we design pressure application strategies (such as simulating task surges, staff reductions, budget fluctuations, etc.); map pressure as a disturbance factor of the business execution curve, and dynamically adjust the status of the simulation unit; design feedback indicators, such as the execution efficiency decline rate, fluctuation response delay, etc. The high-fidelity simulation environment can help companies detect potential risks in advance under "virtual pressure"; the curve disturbance reflects the stress resistance mode of different project teams, which can test the resilience of the organization under high pressure in advance and build a data-driven, controllable, and predictable stress testing platform; support companies to conduct system testing under stress scenarios before making structural adjustments, resource allocation or task arrangements.

[0031] More specifically, run a simulation environment to monitor business responses under different pressure strategies; compare simulation outputs with historical real data to optimize the GCN model and feature construction method; integrate the results into the enterprise management system through model distillation, transfer learning, etc., and the model needs to be continuously optimized in actual use; incorporating a feedback mechanism can improve the model's adaptability and explanatory power for real-world business scenarios, and improve the reliability and credibility of the simulation environment; support enterprises in automated decision-making for pressure forecasting, load balancing, and personnel scheduling.

[0032] Specifically, in step S4 of the embodiment provided by the present invention, long-term trend characteristics and pressure simulation data are collected, including historical execution curves, various pressure application conditions, time series data, interference factors, etc.; these data are prepared into a format suitable for LSTM input, that is, the feature vector of each time point (including long-term trend characteristics and simulated environmental impact) is used as the input of LSTM. LSTM is a powerful model for processing time series data. It requires continuous historical data as input and relies on time context to capture long-term trends and short-term fluctuations; long-term trend characteristics and pressure simulation data provide LSTM with key information related to changes in actual business execution. This step provides the LSTM model with high-quality, structured input data, ensuring that the impact of long-term trends and pressure environment on execution can be effectively learned during the model training process.

[0033] More specifically, a standard LSTM model architecture is designed, which usually includes: an input layer, multiple LSTM layers (which can be deepened), a fully connected layer, and an output layer; the input layer receives long-term trend characteristics and pressure simulation data, the LSTM layer learns time dependencies, and finally outputs future execution values ​​through the fully connected layer. LSTM can effectively capture dependencies in long time series data and can handle long-term time dependency problems, which is suitable for analyzing changes in business execution. In practical applications, LSTM can identify patterns in time series, such as seasonal changes, emergencies, etc., to help capture dynamic fluctuations in business execution, and provide an end-to-end time series prediction framework that can accurately generate the expected execution curve under the input of pressure changes and long-term trends.

[0034] More specifically, the long-term trend characteristics and pressure simulation data are properly standardized / normalized to ensure that the data meets the numerical stability requirements when input into the LSTM; the data is converted into time step form through the sliding window method, so that the input data at each time point includes the execution characteristics and external pressure factors of the previous time points. LSTM is sensitive to the numerical range of the input data, and data standardization must be used to eliminate dimensional differences and improve training efficiency; sliding window is a common method in time series prediction, which can help the model predict future data points based on historical data points. Data preprocessing ensures stable training and good generalization ability of the LSTM model, which can better handle time series under pressure and long-term trend changes.

[0035] More specifically, the LSTM model is trained using training set data, and model parameters are optimized by minimizing the loss function (such as MSE); the validation set is used to evaluate the model's effectiveness to avoid overfitting, and model performance is optimized through hyperparameter tuning (such as learning rate, number of LSTM layers, etc.). Training is the core process of model learning. LSTM gradually optimizes internal parameters by returning errors, so that it can accurately capture the patterns of changes in execution power; the use of the validation set can ensure that the model has good predictive capabilities on unseen data and avoid overfitting on the training set. The training process enables LSTM to effectively learn the inherent patterns of time series and the influence of external disturbance factors. The validation process ensures the stability and generalization ability of the model.

[0036] More specifically, after training, the LSTM model is used to predict future time points, that is, to generate a future execution curve based on the current execution data and the pressure simulation environment. The prediction results can be obtained through the output layer of the LSTM model, which is usually the predicted value of business execution at each future time step. The trained LSTM can generate a future execution prediction curve based on historical data, pressure factors, and changes in long-term trends. The generated prediction curve is the result of the model's understanding of historical trends and pressure changes, and is a simulation of future business execution performance. The generated execution prediction curve can help the project team identify potential high-pressure periods in advance and provide data support for decision-making.

[0037] More specifically, the prediction results are post-processed, such as smoothing, denoising, and trend analysis. If the generated execution curve deviates significantly from the actual historical data, technologies such as transfer learning or model integration can be used to further optimize the model. The prediction results may have large short-term volatility, and post-processing can help smooth the prediction curve and improve the stability and accuracy of the prediction. If the prediction error is large, the model performance can be improved by integrating multiple models or introducing additional data to make the prediction curve smoother and in line with actual business fluctuations. At the same time, the accuracy of the model can be improved to ensure the feasibility of the prediction results in business applications.

[0038] Specifically, in step S5 of the embodiment provided by the present invention, multiple execution prediction curves output from the LSTM model are imported into the task allocation system as a time-dynamic input of the task allocation capability. The execution prediction value reflects the available resource intensity and task processing capability of each department, position or team in a certain period of time in the future, providing a dynamic allocation basis for subsequent task scheduling, so that task allocation is no longer static, but can be intelligently matched according to the carrying capacity of different time periods.

[0039] More specifically, an enterprise task model is established to map the tasks to be assigned with resource elements (people, time, tools, budget), usually represented as a task-resource bipartite graph or relationship table. Task allocation needs to be reasonably matched based on resource capabilities (such as execution prediction). The establishment of the model is the basis for realizing intelligent allocation, achieving a computable correspondence between tasks and resources, and opening up a logical path for subsequent algorithmic task allocation.

[0040] More specifically, set priority, deadline, completion quality requirements, resource restrictions and other constraint parameters for different tasks. In real business scenarios, tasks often have time, sequence and resource constraints, which need to be clearly modeled to ensure that the simulation results are realistic and feasible, enhance the constraint adaptability of the task scheduling algorithm, and achieve "optimal allocation without violating constraints."

[0041] More specifically, based on the execution prediction results and task models, we construct task allocation algorithm rules, such as maximum efficiency matching, load balancing distribution, and minimum total cost distribution. The technologies that can be used include: heuristic algorithms (such as greedy and genetic algorithms), linear programming / integer programming, and multi-objective optimization (considering efficiency, cost, and risk). Different companies have different decision-making goals, and it is necessary to build flexible task allocation rules to adapt to diverse business strategies, achieve precise matching + optimized scheduling, and match the tasks in each time period to the actual capabilities of the team to maximize resource value.

[0042] More specifically, simulated task allocation is performed in a virtual environment, that is, scheduling is performed according to predicted values ​​and task models, simulating the future business execution process. By rehearsing scheduling results, bottlenecks, congestion, or insufficient capacity can be discovered, and strategic adjustments can be made in advance to improve the foresight and robustness of task allocation, helping decision makers to predict problems, allocate resources, and arrange buffers.

[0043] More specifically, key indicators in the simulation process (such as task completion rate, time utilization, resource vacancy rate, and bottleneck nodes) are evaluated, and these results are fed back to the scheduling model for optimization and iteration. Problems in the simulation results can reveal deviations in the task model or execution evaluation. It is necessary to continuously iterate the scheduling strategy through a feedback mechanism to achieve the adaptive learning ability of the task scheduling model and continuously improve the allocation accuracy and strategy optimization level.

[0044] More specifically, based on the simulation and evaluation results, task allocation plans, resource allocation suggestions, predictive risk warnings and other content are output to managers as the basis for decision-making on optimized management. The ultimate goal is to assist enterprise managers in scientific, efficient and predictive task scheduling and decision-making.

[0045] The present invention provides an enterprise management optimization method based on artificial intelligence technology, which has the following beneficial effects: The present invention obtains the enterprise project team architecture and historical business information, constructs a project team simulation unit, analyzes the time relationship of historical business information, generates a business information flow and maps it to the project team simulation unit, obtains a business execution curve, uses a graph neural convolutional network to extract long-term trend characteristics and common fluctuations, constructs a pressure simulation environment, and uses a long short-term memory network to predict the execution curve based on pressure simulation and trend characteristics. A task allocation framework is constructed according to the execution prediction curve, and task allocation simulation is performed to obtain optimized management decisions. This method improves the accuracy and dynamic adaptability of task allocation, optimizes the efficiency of enterprise resource utilization, supports intelligent decision-making, improves overall management efficiency, and solves the problem in the existing technology that tasks allocated by relying on manual experience are difficult to adapt to dynamic changes.

[0046] Preferably, the steps of obtaining the architecture information and historical business information of each project group of the enterprise and constructing a project group simulation unit according to the architecture information include: S11: Collecting the basic architecture of the enterprise as the optimization target to obtain internal architecture information and parallel architecture information of several project teams in the enterprise; S12: Tracing back the internal architecture information and parallel architecture information of each project team based on the company's historical database to obtain the architecture change records and historical business information of each project team; S13: constructing basic simulation units of each project group according to the internal architecture information of each project group, and modifying the collaborative relationship of each basic simulation unit according to the parallel architecture information of each project group to obtain the architecture simulation unit of each project group in the current period; S14: Deducing the architecture simulation units of the current period to past periods based on the architecture change records to generate architecture simulation units of each past period, and combining the architecture simulation units of each past period with the architecture simulation units of the current period to generate a project group simulation unit of the project group.

[0047] Specifically, we collect architectural information from the enterprise that is the optimization target, and obtain the internal architectural information and parallel architectural information of each project team in the enterprise. Collecting the basic architecture of the enterprise lays the foundation for subsequent task allocation and resource scheduling, clarifies the internal organizational structure, team division of labor, and the work content and interaction methods of each project team, and obtains a complete architectural view to facilitate understanding of how each project team collaborates and how resources are allocated, and provides basic data for subsequent simulation and decision support.

[0048] More specifically, through the enterprise historical database, the architectural information changes of each project team and related historical business information can be traced to obtain the architectural development trajectory of the project team. The architectural change records can help analyze the evolution of the project team, including how to respond to different business challenges and resource allocation changes. Historical business information provides the project team's past business performance data, ensuring that the project team's architectural simulation unit can reflect historical change trends and understand the impact of past adjustments on the existing architecture and business performance.

[0049] More specifically, based on the internal architecture information of each project group, basic simulation units are constructed for each project group. These simulation units should be able to reflect the project group's workflow, task allocation, role responsibilities, etc. The basic simulation units help convert each project group's work process, resource allocation, and interaction mode into a quantifiable model to facilitate subsequent simulation and optimization analysis. They provide an independent and operational model for each project group, help identify resource bottlenecks, uneven workloads, and other issues, and provide a basis for optimization decisions.

[0050] More specifically, based on the parallel architecture information of the project team, the collaborative relationship between the basic simulation units is corrected so that the project teams can work together and adapt to the dynamic coordination between organizations. The project teams do not operate independently, and their work often overlaps and depends on each other. Correcting the collaborative relationship can ensure the overall collaboration and efficiency of resource sharing, and ensure that the basic simulation units of each project team can accurately reflect the dependencies between each other, thereby achieving overall optimization of the system, rather than just local task allocation.

[0051] More specifically, based on the architecture change records, we can deduce the architecture simulation units of past periods to generate architecture simulation units of various past stages. This way, we can obtain the changes in the project team architecture at different historical nodes. Deducing past architecture changes can help us identify the effects and shortcomings of architecture adjustments, draw experience from historical data, and optimize current and future architectures. Through historical deduction, we can obtain information on the evolution of the project team architecture and enhance our understanding of the current architecture, especially how to respond to different business needs and changes in the external environment.

[0052] More specifically, historical architecture simulation units are combined with current architecture simulation units to form comprehensive project team simulation units. This step helps integrate past experience with the existing architecture, providing more perspectives and evidence for decision-making optimization. Combining historical and current architecture simulation units can help management identify successful practices and past problems in the architecture's evolution and implement improvements. This provides project teams with a full lifecycle view of the architecture, helping managers better understand the long-term impact of architectural decisions and make more informed adjustments and optimizations.

[0053] It can be understood that through these steps, the entire process can comprehensively collect enterprise architecture information, trace historical changes, build simulation units for the project team, and combine historical and current architectures to generate comprehensive simulation units, providing data support for subsequent resource scheduling, task allocation, and management decisions. Ultimately, the optimized project team simulation units will effectively improve project execution efficiency, optimize resource allocation, and enhance organizational collaboration capabilities.

[0054] Preferably, the steps of parsing the time relationship of the historical business information to generate a business information flow, and mapping the project group simulation unit information according to the business information flow to obtain a business execution curve include: S21: parsing the historical business information in terms of business project content, business allocation time, business completion time, business funds used, and business results and benefits, to obtain the business project content, business allocation time, business completion time, business funds used, and business results and benefits of each business project completed by the project team in the past; S22: Perform multi-dimensional feature coding on the business project content, business allocation time, business completion time, business usage funds, and business result income of the business project to generate business characteristic information of each business project; S23: constructing a time coordinate axis, and deploying each piece of service characteristic information to a corresponding position on the time coordinate axis according to the service allocation time corresponding to each piece of service characteristic information, so as to generate a service information flow; S24: parsing the time correspondence of the project group simulation unit according to the business information flow to construct an information mapping relationship between the business information flow and the architecture simulation unit of each period in the project group simulation unit; S25: allocating each piece of service characteristic information in the service information flow to a corresponding architecture simulation unit based on the information mapping relationship, so as to generate a service characteristic sequence of the architecture simulation unit in each period; S26: performing a multi-dimensional evaluation of the business execution capability of the architecture simulation unit according to the business characteristic sequence to generate a business execution capability feature of the architecture simulation unit; S27: Connecting the business execution capability characteristics of the architecture simulation units in each period to generate a business execution capability curve.

[0055] Specifically, historical business information is analyzed, including business project content, business allocation time, business completion time, business funds used, and business results and benefits, to obtain relevant data on various business projects completed by the project team in the past. Analyzing this information helps to understand the basic situation of past business and provide data support for subsequent simulation and analysis. Business project content and time factors directly affect the efficiency and results of project execution. Providing a detailed view of historical business facilitates the construction of subsequent business information flows and provides basic data for further analysis of the project team's business execution capabilities.

[0056] More specifically, multi-dimensional feature coding is performed on various information of business projects (business project content, business allocation time, business completion time, business funds used, and business results and benefits). This step converts different business data into standardized features to facilitate subsequent processing. Multi-dimensional feature coding can convert complex and heterogeneous business data into a unified format, so that business data from different projects can be compared, analyzed and processed in a standard manner. Through coding, quantifiable business characteristic data can be obtained, which can effectively construct and analyze subsequent business information flows and architecture simulation units.

[0057] More specifically, a time coordinate axis is constructed, and each business characteristic information is deployed on the coordinate axis in chronological order according to the business allocation time corresponding to the business characteristic information. The time coordinate axis helps to combine business information with the time dimension, ensuring that all business characteristic information can be arranged in chronological order, which is convenient for observing the dynamic changes and flow trends of the business. The business information flow generated by the time coordinate axis can time-sequence historical business information, providing a clear framework for subsequent time mapping and analysis.

[0058] More specifically, based on the business information flow, the time correspondence between it and the architecture simulation unit of the project group simulation unit is analyzed, and a mapping relationship between the business information flow and the project group architecture simulation unit is constructed. By constructing the time correspondence, the business information flow can be accurately connected with the changes in the project group architecture, thereby ensuring that the correlation between the business characteristic information and the architecture simulation unit is fully reflected, and generating a mapping relationship under the time dimension, so that the business information can be accurately reflected in the project group architecture model, thereby providing a basis for subsequent optimization and adjustment.

[0059] More specifically, based on the established information mapping relationship, each business characteristic information in the business information flow is allocated to the architecture simulation unit of the corresponding period. This step ensures that the business characteristic information can be effectively transmitted to the architecture simulation unit of each stage, thereby accurately reflecting the business execution status at each time point, completing the allocation of business characteristics to architecture simulation units, ensuring that the architecture simulation unit can present the multi-dimensional characteristics of business execution, and further supporting execution capability evaluation.

[0060] More specifically, based on the business characteristic sequence of each period, a multi-dimensional business execution capability assessment is conducted on the architecture simulation unit. This step takes into account the architecture adjustments and business execution status of the project team in different periods, and conducts quantitative and qualitative assessments. Assessing the business execution capability of the architecture simulation unit helps to identify the effectiveness and deficiencies of the architecture in each period, and provides specific guidance for subsequent improvements and optimizations. Through multi-dimensional assessments, the differences in the execution capabilities of the architecture in different time periods can be revealed, providing more detailed insights for business decision-making and improving the business adaptability of the overall architecture.

[0061] More specifically, the business execution capabilities of the architecture simulation units in each period are connected to generate a business execution curve to reflect the changes in the execution capabilities of the project team architecture in each period. The generated business execution curve can intuitively display the changing trend of the project team architecture execution capabilities and help identify key business nodes and optimization opportunities. The final business execution curve provides a clear and quantifiable business execution capability dynamics, which can serve as a basis for optimization decisions and help managers make more precise architectural adjustments in the future.

[0062] It can be understood that through these steps, historical business information is effectively converted into business information flows that can be analyzed and simulated, and a business execution curve is generated by mapping it with the project team architecture. The entire process enables a comprehensive and systematic analysis of the project team's architectural adjustments and business execution capabilities, thereby providing strong support for the company's architectural optimization, resource scheduling, and business strategy decision-making.

[0063] Preferably, the step of performing a multi-dimensional evaluation of the business execution capability of the architecture simulation unit according to the business characteristic sequence to generate the business execution capability characteristics of the architecture simulation unit includes: S261: Analyzing the business execution efficiency and business revenue multiple of the architecture simulation unit according to each business characteristic information in the business characteristic sequence, so as to generate a shallow evaluation feature of the architecture simulation unit corresponding to each business characteristic information in the business characteristic sequence; S262: performing a comprehensive analysis of the business processing capabilities of the architecture simulation unit corresponding to each type of business content based on the shallow evaluation features, so as to obtain deep evaluation features of the business content categories corresponding to each business characteristic information of the architecture simulation unit; S263: Dividing the service characteristic sequence into service content categories to obtain characteristic subsequences of several service content categories; S264: Analyzing the change relationship of the deep evaluation features of each business characteristic information of the same business content category according to each characteristic subsequence, so as to obtain the execution capability change curve of the architecture simulation unit corresponding to each business content category; S265: Perform a time-sequence-based interactive analysis on the execution capability change curves of the architecture simulation unit corresponding to each business content category to generate a baseline execution capability curve and a business category adaptability distribution curve of the architecture simulation unit in the time period corresponding to the business characteristic sequence, which together serve as the business execution capability characteristics of the architecture simulation unit in that period.

[0064] Specifically, based on the business characteristic information of each business characteristic sequence, the business execution efficiency and business revenue multiple of the architecture simulation unit are analyzed respectively to generate shallow evaluation features of the architecture simulation unit corresponding to each business characteristic information. Before conducting a deep evaluation, the shallow evaluation can quickly identify the efficiency and effectiveness of the architecture unit in specific business tasks, and provide preliminary performance indicators for subsequent analysis. By evaluating the execution efficiency and revenue multiple, we can have a preliminary understanding of how different business characteristics affect the basic performance of the architecture unit. Through the shallow evaluation features, we can obtain preliminary feedback on business execution capabilities, which helps to locate which business characteristics are most critical to the execution efficiency and revenue of the architecture. This helps to quickly identify the direction of business improvement and lay the foundation for subsequent deep analysis.

[0065] More specifically, based on shallow evaluation features, a comprehensive analysis of the business processing capabilities of the architecture simulation unit under various business contents is conducted to obtain deep evaluation features of the architecture simulation unit corresponding to each business characteristic information. The shallow evaluation gives the performance of a single characteristic in a certain dimension, while the comprehensive analysis comprehensively evaluates the processing capabilities of the architecture simulation unit under different business categories by combining multiple dimensions. Through comprehensive analysis, we can have a more detailed understanding of the business processing capabilities of the architecture simulation unit, and then obtain a more in-depth evaluation. Through deep evaluation features, we can understand the performance of the architecture under different business types in a more refined way, revealing the adaptability and potential of the architecture to various tasks and loads, and providing a basis for subsequent performance optimization.

[0066] More specifically, different business contents in the business characteristic sequence are classified, and based on the characteristics of each category, characteristic subsequences of several business content categories are obtained. Different business contents may have different characteristics and need to be divided according to the type of business content. By dividing the business categories, the performance characteristics of each category can be further analyzed, which helps to provide a detailed analysis perspective for the performance change trend of each type of business content. The characteristic subsequences obtained by division can more clearly observe the differences and similarities between the business content categories, thereby more accurately evaluating the execution capability and adaptability of the architecture simulation unit under different business types.

[0067] More specifically, based on each characteristic subsequence, the change relationship of each business characteristic information of the same business content category is analyzed, so as to obtain the execution change curve of the architecture simulation unit in each business content category. The execution change curve reflects the adaptability and execution effect of the architecture simulation unit for different business categories in different time periods. These change curves help to identify business fluctuations, discover bottlenecks, and provide decision support for subsequent architecture optimization and adjustment. The execution change curve can reveal the performance of the architecture over time when facing specific business types. This not only provides an intuitive display of the dynamic changes in business capabilities, but also helps to identify which time periods and business types pose the greatest challenges to the architecture.

[0068] More specifically, a time-series-based interactive analysis is conducted on the execution capability change curves of each business content category to generate a baseline execution capability curve and a business category adaptability distribution curve for the architecture simulation unit in the time period corresponding to the business characteristic sequence. Through the time-series interactive analysis, we can fully understand the overall execution capability of the architecture simulation unit in different time periods, and by comparing the adaptability of each business category, we can reveal the architecture's responsiveness to different business categories. This helps to analyze the stability and resilience of the architecture under various business change conditions. The interactive analysis will generate two key curves: a baseline execution capability curve and a business category adaptability distribution curve. These curves show how the architecture's execution capability changes with time and business categories, and intuitively reflect the architecture's comprehensive execution capability characteristics.

[0069] More specifically, the benchmark execution curve and the business category adaptability distribution curve are combined to ultimately generate the business execution characteristics of the architecture simulation unit during that period. Generating business execution characteristics is the ultimate goal of the entire analysis. It summarizes the comprehensive performance of the architecture simulation unit in multiple dimensions (such as time, business category, etc.). This characteristic can provide management with a comprehensive perspective to support architecture optimization decisions. By generating business execution characteristics, we can clearly understand the execution capabilities of the architecture under different conditions, and based on this, we can further optimize the architecture design and enhance the architecture's business adaptability and execution.

[0070] It is understandable that these steps generate detailed business execution curves by analyzing the execution capabilities of the architecture simulation unit in multiple dimensions, combining time, business category, and various characteristics. These curves not only help reveal the performance of the architecture in different business environments, but also provide strong data support for future architecture optimization, resource allocation, and decision-making.

[0071] Preferably, the steps of extracting the long-term trend characteristics of each of the business execution curves and the common fluctuation characteristics between the business execution curves through a graph neural convolutional network model, and simultaneously constructing a pressure simulation environment include: S31: performing time alignment processing on the business execution capability curves of each project group so that the business execution capability curves of each project group are in a time-aligned state; S32: Performing dynamic topology analysis based on the business execution capability curves in the time-aligned state to generate a parallel execution capability topology map of each business group; S33: Inputting the parallel execution force topology map into a pre-trained graph neural convolutional network model, causing the graph neural convolutional network model to perform graph structure convolution processing on the parallel execution force topology map in the form of independent mode and parallel mode variations, so as to generate long-term trend characteristics of each of the business execution force curves and common fluctuation characteristics between the business execution force curves; S34: constructing a time axis, and performing expansion processing on the common fluctuation characteristics based on the time axis to visually express the common fluctuation characteristics on the time axis, thereby generating a fluctuation value sequence; S35: dividing the fluctuation value sequence into several types based on the time axis, so as to divide the fluctuation value sequence into several fluctuation distribution forms; wherein the fluctuation distribution form is used to divide the fluctuation value sequence into several sections according to a specified form; S36: Tracing the causes of fluctuations for each of the fluctuation distribution forms to obtain pressure factors corresponding to each of the fluctuation distribution forms; S37: performing interactive confidence analysis on each of the pressure factors, and combining each of the pressure factors based on the confidence analysis result to obtain a pressure simulation environment.

[0072] Specifically, the business execution curves from different project groups are time-aligned to ensure that the business execution curves of each project group are in a unified time series. This allows subsequent analysis to compare the business performance of different project groups within the same time period. The business execution curves of each project group may have different time scales or time intervals. Through time alignment, these differences can be eliminated, making the analysis more comparable and unified. The execution of this step ensures the consistency and comparability of the data, provides standardized input data for subsequent dynamic topology analysis, and avoids analysis errors caused by time deviation.

[0073] More specifically, based on the time-aligned business execution curve, a dynamic topology analysis is performed to generate a parallel execution topology map for each business group. The topology analysis will reveal the interdependence between different project groups in different time periods and the dynamic changes in business execution. Through topology analysis, the correlation and interaction between project groups can be identified, and the trends and patterns of business execution over time can be revealed. This is of great significance for understanding the collaborative effects of different business groups and their business performance. The dynamic topology map provides a multi-dimensional perspective to help identify and visualize the interaction patterns, commonalities and differences of each project group, providing useful structural information for subsequent model input.

[0074] More specifically, the obtained parallel execution topology map is input into the trained graph convolutional network (GCN) model, and the model performs graph structure convolution processing on the topology map. During this process, the GCN model will learn the mutual influence and patterns between different project groups based on the structure of the graph, including the changes in independent mode and parallel mode. The graph convolutional network (GCN) is particularly suitable for processing graph structure data. It can capture complex spatial dependencies and dynamic changes, and learn the implicit patterns between data through graph convolution. The GCN model can automatically extract long-term trend characteristics and common fluctuation characteristics when processing the map. Through the graph convolution processing of the GCN model, accurate long-term trend characteristics and common fluctuation characteristics can be obtained. These characteristics reveal the potential laws and changing trends between the business execution curves of each project group, providing a scientific basis for subsequent pressure simulation.

[0075] More specifically, based on the aforementioned common fluctuation characteristics, a timeline is constructed, and these characteristics are expanded according to the timeline. This can achieve a visual expression of the common fluctuation characteristics and generate a corresponding fluctuation numerical sequence for further analysis. The timeline can help associate the fluctuation characteristics with actual time, thereby revealing the time patterns and trends of the fluctuations. Through visual expression, the fluctuation characteristics can be made more intuitive and provide a basis for analysis. The implementation of this step enables the dynamic changes of the common fluctuation characteristics to be presented in the time dimension, making it easier to observe their evolution trends and helping subsequent analysis to clarify the causes and evolution patterns of the fluctuations.

[0076] More specifically, based on the time axis, the fluctuation value sequence is divided into several categories to generate different fluctuation distribution forms. Each fluctuation feature distribution represents the fluctuation characteristics of business execution within a specific period of time. By dividing the fluctuation value sequence, the different stages, cycles or patterns of the fluctuation characteristics can be identified. In this way, the fluctuation characteristics can be divided into different categories, which is convenient for analyzing the performance and fluctuation reasons of each category. The implementation of this step helps to refine the fluctuation characteristics and provide a clear distribution for each fluctuation pattern, which helps to understand the changes in business execution behind each fluctuation segment.

[0077] More specifically, the causes of fluctuations are traced for each fluctuation distribution form to identify the stress factors associated with each fluctuation segment. Stress factors include external or internal factors that may cause specific fluctuation patterns. Tracing the causes of fluctuations can reveal the root causes of various fluctuations, thereby helping managers understand which factors lead to fluctuations in business execution, which is crucial for future business adjustments and stress testing. By tracing the causes of fluctuations, clear stress simulation factors can be provided for each fluctuation segment. These factors help simulate and predict similar fluctuations that may occur in the future.

[0078] More specifically, an interactive confidence analysis is conducted on each stress factor, and they are combined based on the analysis results to ultimately construct a comprehensive stress simulation environment. The interactive confidence analysis helps to evaluate the mutual influence and intensity of each factor, so as to determine which factors play a dominant role in the overall environment. After combining these factors, a more accurate and practical stress simulation environment can be constructed. Through interactive confidence analysis, reliable factor weights can be provided for the stress simulation environment to ensure the accuracy of the simulation environment. The stress simulation environment finally constructed can help enterprises predict the performance of business execution under different pressures and provide data support for decision-making.

[0079] It can be understood that through the above steps, the graph neural convolutional network model is used to extract long-term trend characteristics and common fluctuation characteristics, and these characteristics are visualized and analyzed through the timeline, and finally the pressure factor is obtained and a pressure simulation environment is constructed. The technical effect of this process is that through precise analysis and simulation, the common fluctuation patterns between project teams can be revealed, and corporate managers can make more scientific decisions and plans.

[0080] Preferably, the step of performing an expected simulation on the business execution capability curve based on the pressure simulation environment and the long-term trend characteristics by using a long short-term memory network model to generate an execution capability prediction curve includes: S41: Deploying model parameters of the long short-term memory network model according to the pressure simulation environment to set constraints for the long short-term memory network model; S42: Substitute the long-term trend feature into the long short-term memory network model, and let the long short-term memory network model perform original prediction on the long-term trend feature to obtain the original prediction feature of the long-term trend feature, and at the same time correct the original prediction feature according to the set constraints to obtain the execution prediction curve.

[0081] Specifically, based on the constructed stress simulation environment, the model parameters of the long short-term memory network (LSTM) model are deployed. At this stage, the constraints and parameter settings of the LSTM model are configured mainly based on the analysis results of the stress factors to ensure that the model can be effectively trained and predicted under specific stress conditions. The stress simulation environment can provide additional background information for the LSTM model by identifying the stress factor set, ensuring that it can be appropriately trained according to different stress conditions. These constraints help improve the accuracy of the model's predictions and ensure that the simulation can reflect the actual business performance in actual applications. By deploying parameters according to the stress simulation environment, the LSTM model can be trained under the influence of multi-dimensional stress factors, thereby enhancing its adaptability to fluctuations in complex business execution. This step provides a constrained training basis for subsequent predictions, enabling the model to produce more reliable prediction results in different scenarios.

[0082] More specifically, the long-term trend features extracted above are input into the LSTM model for original prediction. The LSTM model, through its unique structure, captures long-term time dependencies and predicts the business execution curve based on these long-term trend features. The long-term trend features reflect the long-term pattern of changes in business execution over time. The LSTM model, through its recurrent neural network (RNN) structure, can effectively learn and memorize these long-term trends and generate original prediction features on this basis. The LSTM model can capture the long-term dependencies of time series. By analyzing the long-term trend features, the generated original prediction results have strong time adaptability and trend. This step provides preliminary prediction features for the subsequent correction steps and enables the model to use historical data to predict future trends.

[0083] More specifically, after obtaining the original prediction features of the LSTM model, the original prediction results are corrected according to the previously set constraints. The constraints are based on the pressure factor and require that the output of the LSTM model be consistent with the actual business scenario under a specific pressure environment. Therefore, the correction process aims to optimize the LSTM prediction results through these constraints so that the generated execution prediction curve is more in line with reality. Although the LSTM model can effectively capture long-term trends, the original prediction results may be affected by errors or overfitting during model training. By imposing constraints, these problems can be avoided and reasonable corrections can be made to make the prediction results more consistent with the actual business execution fluctuations. The corrected execution prediction curve is more accurate and can better conform to the actual business scenario and pressure environment. This step can significantly improve the interpretability and accuracy of the prediction, ensuring that the prediction curve can adapt to the business execution fluctuations under actual operating conditions.

[0084] More specifically, ultimately, based on the revised prediction features, an execution prediction curve is generated. This prediction curve is the expected value of future business execution. Through dynamic analysis and the revised LSTM model, the prediction curve can reflect the performance of the business under different pressure environments and long-term trends. The execution prediction curve is the final product of the entire process. By learning from past data and considering pressure factors, it can help decision makers foresee future trends in business execution. The prediction curve provides data support for subsequent business adjustments, resource planning, and response measures. The execution prediction curve generated by the LSTM model can accurately reflect possible future fluctuations and provide management with data-driven decision-making basis. This prediction curve not only provides the company with strategic direction, but also helps the company respond quickly when facing external pressure.

[0085] It's clear that this entire process, by combining the stress simulation environment with long-term trend characteristics, provides accurate input to the LSTM model and imposes necessary constraints to ensure the model's effective predictions within specific business environments. By modifying the original prediction characteristics, the resulting execution prediction curve can help companies make more accurate business plans and decisions. The technical benefit of this method lies in improving the model's accuracy and practicality by introducing external constraints and dynamic adjustments, providing strong data support for business forecasting.

[0086] Preferably, the steps of constructing an enterprise task allocation framework based on each of the execution capability prediction curves to simulate task allocation for current business information and obtain optimized management decisions include: S51: performing parallel analysis on the execution capability prediction curves to obtain the execution capability weight of each project team; S52: performing business type adaptability analysis on each of the execution capability prediction curves to obtain business category adaptability distribution characteristics of each project team corresponding to each business category; S53: Based on the execution capability weight of each project team and the adaptability distribution characteristics of the business categories, the execution capability characteristics of each project team are coded to generate execution simulation units corresponding to each project team; S54: configuring the execution simulation units with several types of overall business information to be simulated through the historical database, and performing execution resource scheduling strategy analysis and evaluation on each execution simulation unit based on the overall business information of each type, so as to obtain the execution resource scheduling strategy corresponding to each type of overall business information under the current execution simulation unit; S55: performing unit connection processing on each of the execution simulation units according to the execution resource scheduling strategy of each type of overall business information to construct an enterprise task allocation framework; S56: Obtain current business information, and substitute the current business information into the enterprise task allocation framework, so that the enterprise task allocation framework performs task allocation processing on the current business information, so as to divide the current business information into tasks to be executed corresponding to each of the execution simulation units, and the tasks to be executed of each of the execution simulation units are collectively used as optimization management decisions.

[0087] Specifically, all generated execution prediction curves are analyzed in parallel. Through comprehensive analysis of the execution prediction curves of each project group, the execution weight of each project group is calculated, that is, the relative execution ability of each project group in the overall task. The purpose of this step is to quantify the execution level of each project group, so as to provide a reasonable basis for subsequent task allocation. If the execution weight of a project group is higher, then the project group will be allocated more task resources or responsibilities. Through parallel analysis, the execution weight of each project group can be obtained, making the task allocation process targeted and accurate, which can avoid waste of resources, allocate more tasks to project groups with stronger execution capabilities, and improve overall work efficiency.

[0088] More specifically, a business type adaptability analysis is performed on the execution prediction curve of each project group, that is, the adaptability of each project group to different types of business (such as production tasks, sales tasks, technical support, etc.) is analyzed to obtain the adaptability distribution characteristics of each project group in different business categories. The purpose of business type adaptability analysis is to evaluate whether the project group can effectively cope with different types of tasks. Some project groups may perform well in certain business categories and poorly in other categories. Therefore, this analysis helps to accurately allocate tasks. This step ensures that tasks can be allocated according to the adaptability distribution characteristics of each project group, avoids the mismatch between tasks and project group capabilities, and improves the efficiency of task completion. Through adaptability analysis, a more refined match between business types and project groups can be achieved.

[0089] More specifically, based on the execution capability weights of each project group and the adaptability distribution characteristics of the business categories, feature coding of the project group's execution capability is performed for each project, and corresponding execution simulation units are generated based on these feature codes. Feature coding helps to create a digital representation for each project group that can be directly used in subsequent simulations. These feature codes include the project group's execution capability level and the adaptability of the business category, which can effectively support subsequent resource scheduling and task allocation. The execution simulation units generated in this step provide standardized input data for subsequent simulation analysis, enhancing the accuracy of decision-making. Through feature coding, effective differentiation of different project groups and intelligent matching of corresponding business types can be achieved.

[0090] More specifically, by accessing the historical database, each execution simulation unit is configured with the expected simulated business information. Next, the execution resource scheduling strategy is analyzed and evaluated for each type of business information, resulting in a resource scheduling plan for each execution simulation unit. This step uses historical business data to configure actual task information for the execution simulation unit, ensuring that the scheduling strategy reflects actual business needs. Resource scheduling strategy analysis helps predict resource requirements for different task types within each project team and allocate resources appropriately. With the support of the historical database, a resource scheduling strategy can be tailored for each execution simulation unit, ensuring optimal allocation of execution resources. This step helps improve resource utilization and ensure that tasks are completed on time and efficiently.

[0091] More specifically, according to the execution resource scheduling strategy, each execution simulation unit is subjected to unit connection processing so that each unit can be organically connected together, and finally a complete enterprise task allocation framework is constructed. This framework can coordinate the work tasks of each project group and ensure that the task allocation is reasonable and smooth. The unit connection processing helps to ensure smooth task handover between different project groups, avoiding task delays or waste of resources. After the task allocation framework is constructed, global coordination and optimized management can be achieved. Through unit connection, the enterprise task allocation framework finally constructed can achieve seamless connection between the execution simulation units, improve the coordination efficiency across project groups, and ensure that the task division and execution between project groups can proceed smoothly, reducing management costs and time waste.

[0092] More specifically, the current business information is obtained and substituted into the established enterprise task allocation framework, which will then process the current business information. Specifically, tasks will be divided into corresponding execution simulation units, and each unit will undertake corresponding tasks, ultimately forming an overall optimized management decision. The substitution of current business information enables the enterprise task allocation framework to perform task allocation based on actual conditions, ensuring that task processing meets actual needs. By allocating tasks to appropriate execution simulation units, business processing efficiency can be improved. This stage enables the framework to be actually applied to current business scenarios, achieving business goals through task allocation, and task allocation can more accurately consider the actual execution capabilities of the project team, thereby ensuring that tasks can be completed efficiently and orderly.

[0093] More specifically, all tasks to be executed are divided into various execution simulation units. Ultimately, these tasks serve as the basis for optimizing management decisions, helping managers make the best resource allocation and scheduling decisions. Through reasonable task allocation and optimal resource scheduling, managers can make more reasonable decisions based on the task status of each execution simulation unit, optimize resource utilization, and improve overall business execution. This decision is based on the task division of the execution simulation unit, which can improve overall business efficiency and resource utilization, and provide the company with data-supported efficient management decisions.

[0094] It is understandable that through these steps, enterprises can build a dynamic and optimized task allocation framework based on the execution prediction curve. This method analyzes multi-dimensional factors such as business adaptability, resource scheduling, and historical data support, so that task allocation is not only based on the project team's execution ability, but also takes into account the actual business needs and execution environment. Ultimately, enterprises can allocate tasks in a more refined and intelligent manner, thereby achieving the goal of optimizing management and improving efficiency.

[0095] Reference Figure 2 As shown, in a second aspect, the present invention provides an enterprise management optimization system based on artificial intelligence technology, which is used to implement the enterprise management optimization method based on artificial intelligence technology described in any one of the first aspects, including: The architecture simulation module is used to obtain the architecture information and historical business information of each project group of the enterprise and build a project group simulation unit based on the architecture information; A capability analysis module is used to analyze the time relationship of the historical business information to generate a business information flow, and to perform information mapping on the project group simulation unit according to the business information flow to obtain a business execution capability curve; A pressure simulation module is used to extract the long-term trend characteristics of each of the business execution curves and the common fluctuation characteristics between the business execution curves through a graph neural convolutional network model to build a pressure simulation environment; A capability prediction module, configured to perform an expected simulation of the business execution capability curve based on the pressure simulation environment and the long-term trend characteristics using a long short-term memory network model to generate an execution capability prediction curve; The task allocation module is used to construct an enterprise task allocation framework based on each of the execution prediction curves to simulate task allocation for current business information and obtain optimized management decisions.

[0096] In this embodiment, for the specific implementation of each module in the above system embodiment, please refer to the above method embodiment, which will not be repeated here.

[0097] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An enterprise management optimization method based on artificial intelligence technology, characterized in that: include: Obtaining the architecture information and historical business information of each project group of the enterprise, and building a project group simulation unit based on the architecture information; Analyzing the time relationship of the historical business information to generate a business information flow, and mapping the project group simulation unit information according to the business information flow to obtain a business execution curve; extracting long-term trend characteristics of each of the business execution capability curves and common fluctuation characteristics between the business execution capability curves to construct a pressure simulation environment; Performing an expected simulation on the business execution capability curve based on the pressure simulation environment and the long-term trend characteristics to generate an execution capability prediction curve; An enterprise task allocation framework is constructed based on each of the execution prediction curves to simulate task allocation for current business information and obtain optimized management decisions.

2. The enterprise management optimization method based on artificial intelligence technology according to claim 1, characterized in that: The steps of obtaining the architecture information and historical business information of each project group of the enterprise and constructing a project group simulation unit according to the architecture information include: Collect basic enterprise architecture and business information to obtain architecture information, architecture change records, and historical business information for several project teams; Constructing an architecture simulation unit for each project group in the current period based on the architecture information of each project group; The architecture simulation unit is deduced in past periods according to the architecture change record to generate architecture simulation units of each past period, so as to combine and generate a project group simulation unit.

3. The enterprise management optimization method based on artificial intelligence technology according to claim 2, characterized in that: The steps of parsing the time relationship of the historical business information to generate a business information flow, and mapping the project group simulation unit information according to the business information flow to obtain a business execution curve include: parsing the historical business information based on business item characteristics to generate business characteristic information for each business item; Constructing a time coordinate axis and deploying the service characteristic information to a corresponding position on the time coordinate axis to generate a service information flow; Parsing the time correspondence of the project group simulation units according to the business information flow, and allocating each business characteristic information to the corresponding architecture simulation unit based on the parsing result, so as to generate a business characteristic sequence of the architecture simulation unit in each period; Evaluating the business execution capability of the architecture simulation unit according to the business characteristic sequence to generate a business execution capability feature of the architecture simulation unit; The business execution capability characteristics of the architecture simulation units in each period are connected to generate a business execution capability curve.

4. The enterprise management optimization method based on artificial intelligence technology according to claim 1, characterized in that: The steps of extracting the long-term trend characteristics of each of the business execution capability curves and the common fluctuation characteristics between the business execution capability curves, and simultaneously constructing a pressure simulation environment include: Performing time alignment processing on the business execution capability curves of each project group to perform dynamic topological analysis on each of the business execution capability curves and generate a parallel execution capability topology map of each business group; Inputting the parallel execution force topology map into a pre-trained graph neural convolutional network model, and causing the graph neural convolutional network model to perform graph structure convolution processing on the parallel execution force topology map to generate long-term trend characteristics of each of the business execution force curves and common fluctuation characteristics between the business execution force curves; Constructing a time axis, and performing expansion processing on the common fluctuation characteristics based on the time axis to generate a fluctuation value sequence; Dividing the fluctuation value sequence into several types based on the time axis to divide the fluctuation value sequence into several fluctuation distribution forms; wherein the fluctuation distribution form is used to divide the fluctuation value sequence into several sections according to a specified form; The causes of fluctuations are traced for each of the fluctuation distribution forms, and confidence analysis and combination of the obtained fluctuation causes are performed to obtain a pressure simulation environment.

5. The enterprise management optimization method based on artificial intelligence technology according to claim 1, characterized in that: The step of performing an expected simulation on the business execution capability curve based on the pressure simulation environment and the long-term trend characteristics to generate an execution capability prediction curve includes: Deploying model parameters of a pre-trained long short-term memory network model according to the pressure simulation environment to set constraints for the long short-term memory network model; The long-term trend feature is substituted into the long short-term memory network model, and the long short-term memory network model is made to make an original prediction on the long-term trend feature to obtain the original prediction feature of the long-term trend feature. At the same time, the original prediction feature is corrected according to the set constraints to obtain the execution prediction curve.

6. The enterprise management optimization method based on artificial intelligence technology according to claim 1, characterized in that: The steps of constructing an enterprise task allocation framework based on the execution capability prediction curves to simulate task allocation for current business information and obtain optimized management decisions include: Analyzing the execution capacity ratio and business type adaptability of each execution capacity prediction curve, and encoding the execution capacity characteristics of each project group based on the analysis results to generate an execution simulation unit for each project group; Configure several types of overall business information expected to be simulated for each execution simulation unit through the historical database, and perform execution resource scheduling strategy analysis and evaluation on each execution simulation unit based on each type of overall business information, so as to obtain the execution resource scheduling strategy corresponding to each type of overall business information under the current execution simulation unit; Perform unit connection processing on each of the execution simulation units according to the execution resource scheduling strategy of each type of overall business information to construct an enterprise task allocation framework; Obtain current business information and substitute the current business information into the enterprise task allocation framework so that the enterprise task allocation framework performs task allocation processing on the current business information to divide the current business information into tasks to be executed corresponding to each of the execution simulation units, and the tasks to be executed of each of the execution simulation units are jointly used as optimization management decisions.

7. An enterprise management optimization system based on artificial intelligence technology, characterized in that: An enterprise management optimization method based on artificial intelligence technology for implementing any one of claims 1 to 6, comprising: The architecture simulation module is used to obtain the architecture information and historical business information of each project group of the enterprise and build a project group simulation unit based on the architecture information; A capability analysis module is used to analyze the time relationship of the historical business information to generate a business information flow, and to perform information mapping on the project group simulation unit according to the business information flow to obtain a business execution capability curve; A pressure simulation module, configured to extract the long-term trend characteristics of each of the business execution curves and the common fluctuation characteristics between the business execution curves, so as to construct a pressure simulation environment; A capability prediction module, configured to perform an expected simulation on the business execution capability curve based on the pressure simulation environment and the long-term trend characteristics to generate an execution capability prediction curve; The task allocation module is used to construct an enterprise task allocation framework based on each of the execution prediction curves to simulate task allocation for current business information and obtain optimized management decisions.

Images