Resource scheduling method, system and device, storage medium and program product

Abstract

The invention discloses a resource scheduling method, system and device, a storage medium and a program product, and relates to the technical field of project management, and the method comprises the steps: carrying out the semantic analysis of an electronic bill input by a user, and generating a standardized demand abstract and a corresponding demand feature; evaluating and screening the demand features through a pre-trained multi-modal demand evaluation model and a preset weight configurator to obtain demand features meeting preset conditions; generating a resource scheduling strategy based on a preset dynamic scheduling algorithm in combination with the demand characteristics meeting the preset conditions and pre-collected resource information; and executing a resource allocation task based on the resource scheduling strategy. According to the scheme, by introducing the multi-modal demand evaluation model and the weight configuration mechanism, the accuracy and comprehensiveness of understanding the user demand by the system can be improved. And meanwhile, real-time optimal configuration of the resources is realized based on a dynamic scheduling algorithm, so that resource waste is avoided, and the overall performance of the system and the user experience are improved.

Description

Technical Field

[0001] This application relates to the field of project management technology, and in particular to resource scheduling methods, systems, devices, storage media, and program products. Background Technology

[0002] With the rapid development of information technology, various business systems face increasingly complex and ever-changing user needs and business scenarios. Traditional demand management methods often rely on single-dimensional data for analysis and decision-making, making it difficult to comprehensively and accurately understand users' true needs and effectively cope with complex and ever-changing business environments. Furthermore, in terms of resource allocation, static resource scheduling strategies cannot be adjusted in real time according to dynamic changes in actual needs, easily leading to resource waste or shortages, thus reducing overall system performance and user experience. Summary of the Invention

[0003] The main objective of this application is to provide a resource scheduling method, system, device, storage medium, and program product, which aims to solve the technical problem of how to improve the accuracy of the system's understanding of user needs and realize dynamic optimization of resource allocation to improve overall performance and user experience.

[0004] To achieve the above objectives, this application proposes a resource scheduling method, which includes:

[0005] Semantic parsing is performed on the electronic submission form input by the user to generate a standardized requirement summary and corresponding requirement features;

[0006] The demand features are evaluated and filtered by a pre-trained multimodal demand assessment model and a preset weight configurator to obtain demand features that meet preset conditions.

[0007] Based on the demand characteristics that meet the preset conditions and the pre-collected resource information, a resource scheduling strategy is generated according to a preset dynamic scheduling algorithm.

[0008] Resource allocation tasks are executed based on the resource scheduling strategy.

[0009] In one embodiment, after the step of semantically parsing the electronic submission form input by the user, the method further includes:

[0010] If a vague description is detected in the electronic submission form, the vague description is completed using natural language processing technology.

[0011] In one embodiment, the step of evaluating and filtering the demand features using a pre-trained multimodal demand assessment model and a preset weight configurator to obtain demand features that meet preset conditions includes:

[0012] The demand features are input into a pre-trained multimodal demand assessment model, and the pre-trained multimodal demand assessment model outputs the demand assessment results.

[0013] Based on a preset multi-dimensional evaluation index system, the evaluation results of the requirements are weighted and calculated using a preset weight configurator to generate a comprehensive score.

[0014] If the overall score is greater than or equal to the preset target score threshold, then the requirement feature is determined to meet the preset conditions.

[0015] In one embodiment, the step of evaluating and filtering the demand features using a pre-trained multimodal demand assessment model and a preset weight configurator to obtain demand features that meet preset conditions includes:

[0016] Based on historical multimodal demand data and corresponding demand result labels, an initial multimodal demand assessment model including a feature fusion layer is constructed.

[0017] The demand features are input into the initial multimodal demand assessment model. The parameters of the multimodal demand assessment model are trained and optimized using deep learning algorithms and regularization strategies to obtain a pre-trained multimodal demand assessment model.

[0018] In one embodiment, before the step of evaluating and filtering the demand features using a pre-trained multimodal demand assessment model and a preset weight configurator to obtain demand features that meet preset conditions, the method further includes:

[0019] Based on a pre-defined three-level circuit breaker review protocol, the compliance of the aforementioned requirements is verified.

[0020] In one embodiment, the step of combining the approved demand characteristics with pre-collected resource information and generating a resource scheduling strategy based on a preset dynamic scheduling algorithm includes:

[0021] Based on the approved requirements characteristics, a requirements-resource matching model is constructed.

[0022] Based on the demand and resource matching model, semantic matching is performed on the approved demand features and the pre-collected resource information to obtain an initial resource scheduling strategy.

[0023] The final resource allocation strategy is generated by combining the initial resource scheduling strategy and the preset dynamic scheduling algorithm.

[0024] In one embodiment, the step of semantically matching the approved demand features and pre-collected resource information based on the demand and resource matching model to obtain an initial resource scheduling strategy includes:

[0025] Based on the demand and resource matching model, semantic matching is performed on the approved demand features and the pre-collected resource information to obtain the specific requirements corresponding to the demand features;

[0026] Using a pre-defined planning algorithm, the optimal resource allocation solution within the predicted time frame is calculated based on the current resource information status and demand forecast.

[0027] By introducing a priority queue and a load balancing algorithm, the demand characteristics are prioritized to obtain a priority-sorted task queue.

[0028] Based on the specific requirements corresponding to the demand characteristics, the optimal solution for resource allocation, and the priority-sorted task queue, an initial resource scheduling strategy is obtained.

[0029] In one embodiment, the step of generating a resource allocation strategy by combining the resource scheduling strategy and the preset dynamic scheduling algorithm includes:

[0030] Based on the workload and urgency of the demand characteristics, the demand characteristics are classified using the preset dynamic scheduling algorithm to obtain scheduling suggestions;

[0031] Based on the scheduling recommendations, the initial resource allocation strategy is optimized to obtain the final resource allocation strategy.

[0032] In one embodiment, the step of classifying the demand characteristics according to their workload and urgency using the preset dynamic scheduling algorithm to obtain scheduling suggestions includes:

[0033] If the aforementioned demand characteristics trigger the emergency recruitment scheduling conditions, an emergency demand scheduling suggestion will be generated.

[0034] If the workload of the aforementioned requirement features exceeds a preset threshold, a project-level requirement scheduling suggestion will be generated.

[0035] If the workload of the required features does not exceed a preset threshold, an iterative requirement scheduling suggestion is generated.

[0036] In one embodiment, the step of performing the resource allocation task based on the resource allocation strategy includes:

[0037] During the execution of resource allocation tasks, an abnormality is alerted through a light-up warning mechanism, and an illegal request is blocked through a veto mechanism.

[0038] In one embodiment, after the step of performing the resource allocation task based on the resource allocation strategy, the method further includes:

[0039] Feedback data collected during the execution of resource allocation tasks is input into the multimodal demand assessment model and dynamic resource scheduling module;

[0040] By comparing the feedback data with the prediction results of the multimodal demand assessment model and the execution results of the resource scheduling strategy, the parameter configuration of the multimodal demand assessment model and the resource scheduling strategy are adjusted.

[0041] Furthermore, to achieve the above objectives, this application also proposes a resource scheduling system, which includes:

[0042] The data acquisition and preprocessing module performs semantic parsing and content completion on the electronic submission form input by the user, generating a standardized requirement summary and corresponding requirement features;

[0043] The evaluation module is used to evaluate and filter the demand features through a pre-trained multimodal demand evaluation model and a preset weight configurator to obtain demand features that meet preset conditions.

[0044] The resource scheduling module is used to combine the approved demand characteristics with the pre-collected resource information and generate a resource allocation strategy based on a preset dynamic scheduling algorithm.

[0045] The task execution module is used to execute resource allocation tasks based on the resource allocation strategy, and to provide abnormal prompts through a light-up warning mechanism and to block illegal requests through a veto mechanism during the execution of resource allocation tasks.

[0046] In addition, to achieve the above objectives, this application also proposes a resource scheduling device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the resource scheduling method described above.

[0047] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and which, when executed by a processor, implements the steps of the resource scheduling method described above.

[0048] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the resource scheduling method described above.

[0049] This application proposes a resource scheduling method, system, device, storage medium, and program product. The method includes: semantically parsing an electronic submission form input by a user to generate a standardized demand summary and corresponding demand features; evaluating and filtering the demand features through a pre-trained multimodal demand evaluation model and a preset weight configurator to obtain demand features that meet preset conditions.

[0050] This solution combines demand characteristics that meet preset conditions with pre-collected resource information and generates a resource scheduling strategy based on a preset dynamic scheduling algorithm; then, it executes resource allocation tasks based on this strategy. By introducing a multimodal demand assessment model and a weight configuration mechanism, this approach improves the accuracy and comprehensiveness of the system's understanding of user needs. Simultaneously, the real-time optimization of resource allocation based on the dynamic scheduling algorithm helps avoid resource waste and improves overall system performance and user experience. Attached Figure Description

[0051] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0052] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0053] Figure 1 This is a flowchart illustrating an embodiment of the resource scheduling method of this application.

[0054] Figure 2 The flowchart of the scheduling decision algorithm provided in Embodiment 1 of this application;

[0055] Figure 3 This is a flowchart illustrating Embodiment 2 of the resource scheduling method of this application;

[0056] Figure 4 This is a flowchart illustrating Embodiment 3 of the resource scheduling method of this application;

[0057] Figure 5 This is a schematic diagram of the module structure of the resource scheduling system according to an embodiment of this application;

[0058] Figure 6 This is a schematic diagram of the device structure of the hardware operating environment involved in the resource scheduling method in the embodiments of this application. Detailed Implementation

[0059] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0060] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0061] The main solution of this application embodiment is as follows: semantic parsing of the electronic submission form input by the user to generate a standardized requirement summary and corresponding requirement features; evaluation and screening of the requirement features through a pre-trained multimodal requirement assessment model and a preset weight configurator to obtain requirement features that meet preset conditions; combining the requirement features that meet the preset conditions with the pre-collected resource information and generating a resource scheduling strategy based on a preset dynamic scheduling algorithm; and executing resource allocation tasks based on the resource scheduling strategy.

[0062] In this embodiment, for ease of description, the resource scheduling system will be used as the execution subject in the following description.

[0063] In the current requirements management process, user-input requirements lack structured constraints and often rely on free text descriptions, resulting in a high rate of missing key information (such as value analysis and scope of application), vague development goals, frequent requirement changes, and increased rework rates. The rigid rules of the static management framework lead to inefficient manual verification of submitted content, making it impossible to verify field completeness and format compliance in real time, and unable to dynamically respond to business changes. This allows invalid requirements to enter the process, resulting in a high rate of wasted review resources. There is a significant non-technical driving force in decision-making, relying primarily on empirical judgments rather than quantitative evaluation models. Evaluation standards are subjective and non-configurable, making it difficult to adapt to different types of business scenarios (such as strategic projects and efficiency optimization projects). This leads to a high rate of misjudgment of requirement priorities, resulting in scheduling chaos, unbalanced resource allocation, low resource utilization, and a high rate of missed detection of critical risk requirements. Furthermore, the update of requirement status relies heavily on manual synchronization, frequently switching between emails, Excel spreadsheets, and various project management tools, resulting in data fragmentation and information lag. It lacks a unified real-time visualization and early warning capability, making it impossible to identify progress deviations (such as development timeouts, test blockages, and other abnormal situations) in a timely manner. The data generated during the evaluation process cannot be automatically accumulated into a reusable knowledge base, which seriously restricts the intelligent upgrading and continuous optimization capabilities of the resource scheduling system.

[0064] This application provides a solution that deeply couples a workflow engine, an evaluation decision model, and a real-time data lake to replace fragmented tools (such as Excel / Lanhu / email) and achieve closed-loop data-driven management of requirements from submission to deployment. By designing an electronic submission template with logical validation capabilities, it supports dynamic field expansion based on requirement type, improving the completeness and consistency of information collection. The system integrates Natural Language Processing (NLP) technology to automatically complete fuzzy user input and generate standardized requirement summaries. Furthermore, a semantic integrity scoring algorithm is introduced to perform compliance pre-checks on unstructured text, achieving preliminary quality control of requirement content and significantly reducing the workload and error rate of manual review. Simultaneously, this application proposes a dynamic scheduling algorithm based on dimensional feature evaluation, combining resource status and requirement priority to achieve intelligent generation of resource allocation strategies. A configurable weight configurator is introduced, allowing flexible adjustment of evaluation parameters according to different business scenarios, shortening the iteration cycle of the evaluation model from months in traditional methods to hours, significantly improving system response speed and adaptability. This mechanism effectively alleviates the mismatch between development resources and requirements, significantly improving resource utilization. A dual risk control mechanism is introduced during resource scheduling and execution: on the one hand, a veto mechanism is used to intercept non-compliant requests in real time, ensuring that all requests entering the process meet compliance requirements, with an interception rate of 100%; on the other hand, a warning mechanism is built to monitor and alert key nodes (such as development timeouts, test blockages, etc.) in real time, greatly improving the efficiency of identifying progress deviations and reducing the risk of project delays.

[0065] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device capable of performing the above functions. The following description uses a personal computer as an example to illustrate this embodiment and the subsequent embodiments.

[0066] Based on this, embodiments of this application provide a resource scheduling method, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the resource scheduling method of this application.

[0067] In this embodiment, the resource scheduling method includes steps S10 to S40:

[0068] Step S10: Semantic parsing of the electronic submission form input by the user to generate a standardized requirement summary and corresponding requirement features;

[0069] Understandably, since the content entered by users in the electronic submission form may contain ambiguities, incomplete information, or non-standard terminology, step S10 is executed. This involves using Natural Language Processing (NLP) technology to semantically analyze and complete the ambiguous content in the user-entered electronic submission form, generating a standardized requirement summary and corresponding requirement features. This approach improves the completeness and accuracy of requirement information, reduces manual review costs, and provides structured data support for subsequent requirement assessment, resource scheduling, and process automation.

[0070] Specifically, the system first collects structured electronic submission forms filled out by users. These forms include, but are not limited to, the following fields: requirement description, current situation analysis, expected goals, urgency level, scope of application, and value quantification. To improve the completeness and accuracy of the submissions, the system uses natural language processing technology to perform semantic understanding and content completion on ambiguous content in the electronic submissions, and generates standardized requirement summaries.

[0071] During the form completion process, the system activates its intelligent form engine, dynamically generating fields based on the user's selected requirement type and loading the corresponding subforms. For example, when the requirement is identified as AI-related, a "Data Security Level" field is added to enhance the targeting of information collection.

[0072] Simultaneously, the system performs format validation and integrity checks on the electronic submission form to ensure that all required fields are complete and that data types are verified. After validation, the electronic submission form will be archived in the requirements database to trigger subsequent review processes.

[0073] In addition, the system also has a historical demand feedback mechanism. Through a similarity matching algorithm, it recommends historical demand items and their standardized descriptions that are similar to the current demand to users, helping users to optimize the content of their submissions and improve the quality and consistency of their demands.

[0074] After the standardized summary is submitted, the system, based on the extracted requirement characteristics, calls a multi-source data acquisition interface to collect multimodal data related to the requirement from multiple internal and external data sources for subsequent evaluation and resource scheduling decisions. The collected data includes, but is not limited to: text data (such as user feedback and product descriptions), image data (such as product images and user interface screenshots), behavioral data (such as user operation logs and browsing history), and time-series data (such as changes in business traffic and resource usage).

[0075] In addition, the system can also access external public large model services to obtain semantically enhanced data related to the required content, further enriching the evaluation dimensions and improving prediction accuracy.

[0076] To ensure the accuracy of subsequent assessment and modeling, the system performs unified cleaning and preprocessing on the collected multimodal data. Specifically, the collected multimodal data is cleaned to remove noise, duplicate data, and outliers; data standardization is performed, and the data format and range are unified to ensure comparability and compatibility; timestamp alignment technology is used to ensure that data from different modalities are consistent in the time dimension, facilitating subsequent joint modeling and analysis; finally, feature extraction algorithms are applied to extract key demand feature vectors from the original multimodal data, forming a structured demand feature dataset, which serves as the input basis for the multimodal demand assessment model.

[0077] Through the above steps, the entire process from unstructured demand input to structured data output is automated; multimodal data acquisition capabilities are built, supporting the fusion processing of various data types such as text, images, behavior, and time series; at the same time, high-quality preprocessed data is improved, providing solid data support for subsequent demand assessment, resource scheduling, and risk control; and the intelligence level, response efficiency, and resource allocation capabilities of the resource scheduling system are effectively improved.

[0078] Step S20: The demand features are evaluated and screened using a pre-trained multimodal demand evaluation model and a preset weight configurator to obtain demand features that meet preset conditions.

[0079] It should be noted that the weight configurator refers to a component used to dynamically adjust the weight parameters of each feature dimension in the evaluation model according to the business scenario. This weight configurator can flexibly adjust the priority evaluation criteria according to different project types, thereby improving the adaptability and accuracy of resource allocation strategies.

[0080] Understandably, since the focus of demand evaluation differs in different business scenarios, the traditional fixed weight mechanism is difficult to meet the diverse requirements of demand management. Therefore, in step S20, a configurable weight mechanism can be introduced in combination with a multimodal evaluation model to achieve dynamic evaluation and intelligent screening of demand features, ensuring that the output set of demand features has a higher degree of business matching and decision-making reference value.

[0081] In one feasible embodiment, step S20 may include steps S21 to S23:

[0082] Step S21: Input the demand features into the pre-trained multimodal demand assessment model, and output the demand assessment result from the pre-trained multimodal demand assessment model;

[0083] Specifically, before inputting demand features into the pre-trained multimodal demand assessment model, the model needs to be trained. Specifically, a multimodal fusion neural network deep learning architecture is adopted. Preprocessed multimodal data is used as input, and by designing appropriate fusion layers, features from different modalities are fused. The model is trained using a large amount of historical multimodal data, with actual user demand results as labels. Model parameters are adjusted by minimizing a loss function (such as cross-entropy loss function) to improve the model's prediction accuracy. During training, regularization techniques are used to prevent overfitting, and optimization algorithms such as stochastic gradient descent are used to accelerate model convergence.

[0084] The multimodal data collected and preprocessed in real time is input into the trained multimodal demand assessment model, which then outputs the demand assessment results for the current user, including but not limited to demand type, urgency, importance, and potential demand trends, providing a basis for subsequent resource scheduling and priority ranking.

[0085] Step S22: Based on the preset multi-dimensional evaluation index system, the demand evaluation results are weighted and calculated using a preset weight configurator to generate a comprehensive score.

[0086] It should be noted that the multi-dimensional evaluation index system includes five dimensions: strategic alignment, direct value, implementation feasibility, technological maturity, and investment cost. Each dimension has its specific scoring logic and veto mechanism, as shown in Table 1:

[0087]

[0088]

[0089] Table 1

[0090] In this embodiment, a preset weight configurator is used in conjunction with a preset multi-dimensional evaluation index system to assign weighted scores to the evaluation results. Specifically, different weights are assigned to the five dimensions in the multi-dimensional evaluation index system: W1, W2, W3, W4, and W5, totaling 1.

[0091] Subsequently, the comprehensive score of the demand assessment results is calculated as follows: W1 * Strategic Alignment Score + W2 * Direct Value Score + W3 * Implementation Feasibility Score + W4 * Technology Maturity Score + W5 * Input Cost Score.

[0092] For example, suppose requirement A scores the following across all dimensions:

[0093] Strategic fit score: 80;

[0094] Direct value score: 75;

[0095] Feasibility score: 90;

[0096] Technology readiness score: 85;

[0097] The weights are allocated as follows:

[0098] W1 = 0.2;

[0099] W2 = 0.3;

[0100] W3 = 0.25;

[0101] W4 = 0.15;

[0102] W5 = 0.1;

[0103] The overall score for this demand assessment is:

[0104] Overall score = 0.2×80 + 0.3×75 + 0.25×90 + 0.15×85 + 0.1×60 = 80.25;

[0105] By collecting relevant data and performing calculations based on the above algorithm model, the scores and comprehensive evaluation results of the intelligent resource scheduling system and method based on multimodal evaluation and dynamic resource scheduling in various dimensions can be obtained.

[0106] Step S23: If the overall score is greater than or equal to the preset target score threshold, then it is determined that the demand feature meets the preset conditions.

[0107] After calculating the comprehensive score of the demand, the comprehensive score is compared with a preset target score threshold. If the comprehensive score is greater than or equal to the preset target score threshold, the demand is determined to meet the resource scheduling criteria and serves as the basic input for generating the resource allocation strategy in the next step.

[0108] For example, if the preset threshold is 60 points, the above requirements (with a comprehensive score of 80.25) will be selected and enter the subsequent resource scheduling process.

[0109] Through the steps described above, by introducing a pre-trained multimodal requirement assessment model, the extracted requirement features can be analyzed and scored automatically and quantitatively, transforming vague business requirements into calculable technical indicators. Simultaneously, by introducing a weight configurator, the weight parameters of each assessment dimension can be dynamically adjusted according to different business scenarios, making the assessment system more flexible and adaptable. By weighting and scoring requirement features and setting screening thresholds, the system can automatically identify high-quality requirements that meet current resource allocation goals from a massive pool of requirements. This provides a scientific basis for subsequent resource allocation and scheduling, avoiding resource waste on low-priority or non-compliant tasks.

[0110] Step S30: Combine the demand characteristics that meet the preset conditions with the pre-collected resource information, and generate a resource allocation strategy based on the preset dynamic scheduling algorithm;

[0111] It should be noted that the pre-collected resource information includes, but is not limited to, computing resources (number of CPU cores, clock speed, utilization rate), storage resources (total capacity, used capacity, remaining capacity, read / write speed), network resources (bandwidth, latency, packet loss rate), and human resources (total number of developers, available development man-days, development types, and estimated availability date).

[0112] In this embodiment, multi-dimensional resource information is collected in real time through the system's built-in sensors, monitoring agents, and resource management interfaces. Simultaneously, historical resource usage data is collected to form a resource usage history database for analyzing resource usage patterns.

[0113] Meanwhile, using technologies such as message queues and publish-subscribe patterns, the collected resource information is transmitted to the resource scheduling decision-making unit in real time. The real-time status of resources is then displayed on the system management interface in the form of visual charts (such as line charts, bar charts, and dashboards), allowing administrators to intuitively understand resource usage. Furthermore, a resource status threshold alarm mechanism is set up to promptly issue alarms to relevant personnel when resource utilization exceeds a set threshold (e.g., CPU utilization exceeds 80%) or when a resource malfunctions.

[0114] Based on the above, step S30 may include steps S31 to S33:

[0115] Step S31: Based on the demand characteristics that meet the preset conditions, construct a demand-resource matching model;

[0116] Specifically, the system first receives demand characteristics that meet preset conditions, including but not limited to demand type, urgency, importance, and potential trends. Based on this information, the system constructs a demand-resource matching model to describe the mapping relationship between different types of demands and available resources.

[0117] In this embodiment, the demand-resource matching model is constructed using semantic analysis and knowledge graph technology. This enables semantic matching of demand features and resource attributes, clarifying the specific resource requirements for different demands (e.g., high computing performance demands correspond to high-frequency CPUs, and large data storage demands correspond to large-capacity storage devices). Simultaneously, the model considers the correlation and dependencies between tasks to optimize resource allocation combinations.

[0118] Step S32: Based on the demand and resource matching model, perform semantic matching on the approved demand features and the pre-collected resource information to obtain an initial resource scheduling strategy.

[0119] In this embodiment, the system inputs the reviewed demand features and real-time collected resource information into the aforementioned demand and resource matching model for semantic-level matching, outputting preliminary resource scheduling suggestions. Furthermore, the system combines reinforcement learning algorithms and dynamic programming mechanisms, integrating priority ranking and load balancing strategies, to ultimately generate an initial resource scheduling strategy.

[0120] Specifically, step S32 may also include steps S321 to S324:

[0121] Step S321: Based on the demand and resource matching model, perform semantic matching on the approved demand features and the pre-collected resource information to obtain the specific requirements corresponding to the demand features;

[0122] First, the system uses semantic analysis technology to convert the matching relationship between demand characteristics and resource attributes into specific resource configuration requirements. For example, for an AI training task, the system will identify the required GPU model, memory size, storage space, and concurrent computing power, and compare them with the available resources in the current resource pool.

[0123] Step S322: Calculate the optimal resource allocation solution within the predicted time using a preset planning algorithm based on the current resource information status and demand forecast.

[0124] Meanwhile, the system also incorporates a dynamic programming algorithm to calculate the optimal path for resource allocation based on the current resource usage status and demand forecasts for a future period. This algorithm comprehensively considers multiple objective functions, such as maximizing resource utilization and minimizing task completion time, ensuring that the resource scheduling strategy is forward-looking and stable.

[0125] Step S323: Introduce a priority queue and load balancing algorithm to prioritize the demand characteristics and obtain a priority-sorted task queue.

[0126] To ensure priority resource allocation for critical needs, this embodiment also introduces a priority queue mechanism. Different needs are prioritized based on factors such as urgency, impact, and evaluation score, ensuring that high-priority needs receive resources first. Simultaneously, a load balancing algorithm is applied to rationally distribute resources across task nodes, preventing some nodes from becoming overloaded while others remain idle.

[0127] Step S324: Based on the specific requirements corresponding to the demand characteristics, the optimal solution for resource allocation, and the priority-sorted task queue, an initial resource scheduling strategy is obtained.

[0128] Finally, by combining the specific requirements corresponding to the demand characteristics obtained from the above steps, the optimal solution for resource allocation, and the priority-ordered task queue, an initial resource scheduling strategy is generated.

[0129] Through the above steps, a demand-resource matching model based on semantic understanding and knowledge graphs is constructed to improve the accuracy of resource allocation. Simultaneously, dynamic programming and reinforcement learning algorithms are combined to achieve intelligent generation of resource scheduling strategies. Furthermore, priority ranking and load balancing mechanisms are introduced to ensure that high-priority tasks are executed first and to achieve efficient resource utilization.

[0130] Step S33: Combine the resource scheduling strategy and the preset dynamic scheduling algorithm to generate the final resource allocation strategy.

[0131] It should be noted that the preset dynamic scheduling algorithm refers to an intelligent scheduling mechanism that dynamically adjusts the task execution order and time nodes based on real-time task status, resource usage, and changes in demand priority.

[0132] It is understandable that since the resource scheduling strategy only describes the type and quantity of resources required by the task, but does not involve the specific time arrangement and execution order, the execution step S33 can optimize the task execution rhythm through the intelligent scheduling algorithm, thereby improving the balance between resource utilization and task completion rate.

[0133] In another feasible embodiment, step S33 may further include steps S331 to S332:

[0134] Step S331: Based on the workload and urgency of the demand characteristics, the demand characteristics are classified using the preset dynamic scheduling algorithm to obtain scheduling suggestions;

[0135] It should be noted that the scheduling suggestion refers to a task execution time arrangement and priority ranking suggestion generated by comprehensively considering the characteristics of demand, resource availability, and organizational strategy requirements. This suggestion aims to provide clear time-based guidance for subsequent resource allocation, ensuring that tasks can be executed efficiently while meeting constraints.

[0136] In this step, the system first classifies the demand characteristics based on their workload and urgency, combined with the current resource pool status and organizational strategy, using a preset dynamic scheduling algorithm to generate specific scheduling suggestions.

[0137] As shown in Table 2:

[0138]

[0139] Table 2

[0140] This embodiment classifies different demand characteristics using a preset dynamic scheduling algorithm and proposes corresponding scheduling suggestions, as follows:

[0141] Project-level scheduling: suitable for requirements with a large workload (>5 person-days) and many independent modules, accounting for 70% of the total requirements;

[0142] Iterative scheduling: suitable for needs requiring rapid response and short iteration cycles, accounting for 20% of total needs;

[0143] Urgent Recruitment Schedule: Applicable to needs with high urgency and low workload (≤1 person-day), accounting for 10% of the total demand.

[0144] Please refer to Figure 2 , Figure 2 The flowchart shows the scheduling decision algorithm. First, the system determines whether the urgency of the demand characteristics triggers the emergency recruitment scheduling conditions. If the urgency of the demand characteristics triggers the emergency recruitment scheduling conditions, a dedicated resource pool is allocated through the emergency scheduling channel, and an emergency demand scheduling suggestion is generated. If the urgency of the demand characteristics does not trigger the emergency recruitment scheduling conditions, the system further determines whether the workload in the demand characteristics exceeds a preset threshold.

[0145] If the workload of the required feature exceeds a preset threshold (e.g., 5 person-days), resources are allocated from the quarterly project resource pool through the project-level scheduling channel to generate a project-level requirement scheduling suggestion; if the workload of the required feature does not exceed the preset threshold, the requirement is inserted into the monthly iteration queue through the iteration-level scheduling channel to generate an iteration-level requirement scheduling suggestion.

[0146] The output scheduling recommendations include, but are not limited to, the planned start date (the precise date based on the resource pool schedule), the expected delivery date (a conservative timeframe considering dependency links), the resource allocation list (development / testing personnel and their contribution percentage), and conflict warning markers (red: severe dependency conflict, yellow: potential risk).

[0147] Step S332: Based on the scheduling suggestion, optimize the initial resource scheduling strategy to obtain the final resource scheduling strategy.

[0148] Finally, the system further optimizes and adjusts the initial strategy by using the generated scheduling suggestions as time-dimensional constraints, so as to improve its executability and stability.

[0149] Specifically, the system compares the task execution times in the initial resource scheduling strategy with the planned start date and expected delivery date in the scheduling suggestions to ensure tasks are completed within the specified time. If resource usage is concentrated or conflicting within a certain time period, the system will automatically adjust the resource allocation ratio and introduce a load balancing strategy for optimization. For tasks marked with red / yellow flags, the system will reassess their priority and adjust the resource allocation order, inserting emergency resources or extending the execution cycle if necessary. Based on the above optimization results, the final resource scheduling strategy is output, including resource configuration details, execution time arrangements, priority ranking, and exception handling mechanisms.

[0150] Through the above steps, a dynamic scheduling algorithm is introduced to adjust the scheduling strategy in real time according to the dynamic changes in actual needs, thereby improving the flexibility, response speed and execution efficiency of resource scheduling.

[0151] Step S40: Execute the resource allocation task based on the resource allocation strategy.

[0152] In this embodiment, the resource allocation interface is used to perform resource allocation tasks on computing resources according to the resource scheduling strategy obtained above. These tasks include, but are not limited to: allocating computing resources to processes and scheduling threads; allocating storage space and allocating file storage paths for storage resources; and allocating bandwidth, establishing and managing network connections, etc.

[0153] Furthermore, to ensure timely progress and identify potential problems during resource allocation, the system employs a warning mechanism. This mechanism compares the actual completion status of task nodes with preset time standards, using different colors to indicate the current status. For example, in the requirements gathering phase, if the system detects that the reviewer can complete the review within the stipulated time, the node is displayed in green. In the requirements assessment phase, if the system detects that the requirements management committee has failed to output the scheduling time within the stipulated time, but there is still time remaining, the node is displayed in yellow, reminding relevant personnel to expedite the process. In the development phase, if the system detects that the development engineer has failed to complete the task before the expected delivery date, the node is displayed in red, triggering an emergency response process.

[0154] Meanwhile, to ensure all requirements comply with organizational standards, business objectives, and compliance requirements, the system introduces a veto mechanism to block non-compliant requests. This mechanism, as a crucial tool for quality control and risk prevention, effectively enhances the rigor and efficiency of requirement review.

[0155] Furthermore, during task execution, the system monitors the task's progress, resource usage, and system performance metrics in real time. When resource requirements change (e.g., task size increases, execution priority changes) or resource status becomes abnormal (e.g., hardware failure, network interruption), a resource rescheduling mechanism is triggered to dynamically adjust resource allocation, ensuring the task can continue to execute normally. Simultaneously, resource usage logs are recorded during task execution, providing data support for subsequent resource scheduling optimization.

[0156] Through the above steps, an early warning mechanism is introduced, enabling the system to provide real-time status indicators for each key node, allowing project managers to detect task delays or blockages immediately. Combined with a veto mechanism, this not only effectively avoids resource waste caused by low-quality or non-compliant requirements but also improves the standardization and consistency of internal requirement governance.

[0157] The above-described method performs semantic parsing on user-input electronic submissions to generate standardized requirement summaries and corresponding requirement features. A pre-trained multimodal requirement assessment model and a preset weight configurator are used to evaluate and filter these requirement features, resulting in those that meet preset conditions. These requirements, combined with pre-collected resource information, are then used to generate a resource scheduling strategy based on a preset dynamic scheduling algorithm. Finally, resource allocation tasks are executed based on this strategy. This solution, by introducing a multimodal requirement assessment model and a weight configuration mechanism, improves the accuracy and comprehensiveness of the system's understanding of user needs. Furthermore, the real-time optimization of resource allocation based on the dynamic scheduling algorithm helps avoid resource waste and improves overall system performance and user experience.

[0158] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in Embodiment 1 above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 3 Before step S20, the resource scheduling method further includes step S020:

[0159] Step S020: Based on the preset three-level circuit breaker review protocol, perform compliance verification on the required characteristics.

[0160] Compared to the first embodiment, this embodiment also introduces a three-level circuit breaker review protocol to verify the compliance of demand characteristics, so as to improve the system's accuracy in identifying and intercepting illegal demands, while taking into account both review efficiency and business flexibility.

[0161] It should be noted that the aforementioned three-level circuit breaker review agreement is an effective mechanism for conducting multi-level, progressive compliance verification of requirements.

[0162] Specifically, before the demand characteristics formally enter the resource scheduling process, AI reviewers will complete the intelligent review of the demand within a preset timeframe (e.g., no more than 24 hours) and provide three types of results: The first type is approval, in which case the system will assign an evaluation queue number to the demand, preparing it for subsequent scheduling; the second type requires supplementary information, in which case the system will automatically mark the missing fields and prompt the submitter to supplement them; the third type is rejection, in which the system will clearly explain the compliance reasons, such as involving sensitive words or violating policy regulations. Simultaneously, the system will automatically compile various key indicators for the demand throughout its entire lifecycle, including the annual cumulative demand volume, the number completed, and the number of cancelled items, and generate a visual dashboard for management to monitor and support decision-making in real time.

[0163] In this embodiment, the review process performed by the AI ​​reviewer strictly follows a preset three-level circuit breaker review protocol. When a request is submitted, the system first performs L1-level basic format verification. If basic formatting issues such as missing fields or syntax errors are found, the first-level circuit breaker mechanism is immediately triggered, intercepting the request and highlighting the error location on the interface, prompting the user to correct and resubmit. If the L1 review passes, it enters the L2-level compliance risk screening stage. The system uses keyword recognition and semantic analysis to determine whether the request involves potential compliance risks, such as data privacy leaks, security vulnerabilities, or violations of regulatory policies. Once relevant risk keywords are identified, the system immediately triggers a real-time intervention mechanism from the legal department, with professionals conducting further review. For requests that pass L2 review, if their urgency score exceeds a set threshold (e.g., greater than 4 points, out of a maximum of 5), they enter the L3-level executive channel push mechanism. The system directly pushes the request to senior management for rapid approval and priority processing.

[0164] By applying the three-level circuit breaker review protocol described in the above embodiment, this embodiment achieves comprehensive compliance verification of requirements from form to substance. This not only effectively improves the ability to identify non-compliant requirements but also significantly optimizes the flow efficiency of efficient requirements. Furthermore, all review records, reasons for rejection, and modification suggestions are fully archived by the system, forming a closed-loop governance mechanism and providing data support for subsequent model evaluation and resource scheduling.

[0165] Based on the first embodiment of this application, in the third embodiment of this application, the content that is the same as or similar to that in the first embodiment described above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 4 After step S40, the resource scheduling method further includes steps S401 to S402:

[0166] Step S401: Input the feedback data collected during the execution of the resource allocation task into the multimodal demand assessment model and dynamic resource scheduling module;

[0167] Step S402: By comparing the differences between the feedback data and the prediction results of the multimodal demand assessment model and the execution results of the resource scheduling strategy, the parameter configuration of the multimodal demand assessment model and the resource scheduling strategy are adjusted.

[0168] Specifically, the system first collects user feedback data on the system services, including but not limited to user satisfaction ratings, actual business performance data (such as order completion rate, task success rate, etc.), and actual resource usage and performance metrics during system operation.

[0169] Then, the feedback data is input into the multimodal demand assessment model and the dynamic resource scheduling module. By analyzing the differences between the feedback data and the model's prediction results and the execution results of the scheduling strategy, the model parameters and scheduling strategies are optimized and adjusted. For example, if it is found that the model's assessment of certain types of demand is inaccurate, corresponding training data can be added or the model structure can be adjusted; if it is found that a certain resource scheduling strategy leads to low resource utilization or a decrease in system performance, the parameters of the scheduling algorithm can be adjusted or the scheduling algorithm can be replaced to continuously improve the system's intelligent demand management capabilities.

[0170] By continuously receiving real-world execution feedback data, the parameter configuration of the multimodal demand assessment model is constantly optimized, enhancing its ability to identify and predict different types of demands, thereby better supporting resource scheduling decisions. Simultaneously, the scheduling strategy is adjusted in reverse based on actual execution results, making resource allocation more closely aligned with real-world business rhythms and avoiding resource waste or task delays caused by rigid preset strategies.

[0171] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the resource scheduling method of this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0172] This application also provides a resource scheduling system; please refer to [reference needed]. Figure 5 The resource scheduling system includes:

[0173] The data acquisition and preprocessing module 10 performs semantic parsing on the electronic submission form input by the user, and generates a standardized requirement summary and corresponding requirement features;

[0174] Evaluation module 20 is used to evaluate and filter the demand features through a pre-trained multimodal demand evaluation model and a preset weight configurator to obtain demand features that meet preset conditions.

[0175] The resource scheduling module 30 is used to combine the approved demand characteristics with the pre-collected resource information and generate a resource scheduling strategy based on a preset dynamic scheduling algorithm.

[0176] The task execution module 40 is used to execute resource allocation tasks based on the resource scheduling strategy.

[0177] The resource scheduling device provided in this application, employing the resource scheduling method in the above embodiments, can solve the technical problem of demand management. Compared with the prior art, the beneficial effects of the demand management device provided in this application are the same as those of the resource scheduling method provided in the above embodiments, and other technical features in the demand management device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0178] This application provides a resource scheduling device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the resource scheduling method in the first embodiment described above.

[0179] The following is for reference. Figure 6 The diagram illustrates a structural schematic of a resource scheduling device suitable for implementing embodiments of this application. The resource scheduling device in these embodiments may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 6 The resource scheduling device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0180] like Figure 6As shown, the resource scheduling device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in the read-only memory 1002 or a program loaded from the storage device 1003 into the random access memory 1004. The random access memory 1004 also stores various programs and data required for the operation of the resource scheduling device. The processing unit 1001, the read-only memory 1002, and the random access memory 1004 are interconnected via a bus 1005. An input / output interface 1006 is also connected to the bus. Typically, the following systems can be connected to the input / output interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the resource scheduling device to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows resource scheduling devices with various systems, it should be understood that it is not required to implement or possess all of the systems shown. More or fewer systems may be implemented alternatively.

[0181] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0182] The resource scheduling device provided in this application, employing the resource scheduling method described in the above embodiments, can solve the technical problem of how to achieve dynamic resource scheduling capabilities for demand management and improve the intelligence level and response efficiency of demand management. Compared with the prior art, the beneficial effects of the resource scheduling device provided in this application are the same as those of the resource scheduling method provided in the above embodiments, and other technical features of this resource scheduling device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0183] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0184] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0185] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the resource scheduling method in the above embodiments.

[0186] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0187] The aforementioned computer-readable storage medium may be included in the resource scheduling device; or it may exist independently and not be assembled into the resource scheduling device.

[0188] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by the resource scheduling device, the resource scheduling device performs the following: semantic parsing of the user-input electronic submission form to generate a standardized demand summary and corresponding demand features; evaluates and filters the demand features using a pre-trained multimodal demand assessment model and a preset weight configurator to obtain demand features that meet preset conditions; combines the demand features that meet the preset conditions with pre-collected resource information and generates a resource scheduling strategy based on a preset dynamic scheduling algorithm; and executes resource allocation tasks based on the resource scheduling strategy.

[0189] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0190] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0191] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0192] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described resource scheduling method. This solves the technical problem of how to achieve dynamic resource scheduling capabilities for demand management and improve the intelligence level and response efficiency of demand management. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the resource scheduling method provided in the above embodiments, and will not be repeated here.

[0193] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the resource scheduling method described above.

[0194] The computer program product provided in this application can solve the technical problem of how to realize dynamic resource scheduling capability for demand management and improve the intelligence level and response efficiency of demand management. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as the beneficial effects of the resource scheduling method provided in the above embodiments, and will not be repeated here.

[0195] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A resource scheduling method, characterized in that, The resource scheduling method includes: Semantic parsing is performed on the electronic submission form input by the user to generate a standardized requirement summary and corresponding requirement features; The demand features are evaluated and filtered by a pre-trained multimodal demand assessment model and a preset weight configurator to obtain demand features that meet preset conditions. Based on the demand characteristics that meet the preset conditions and the pre-collected resource information, a resource scheduling strategy is generated according to a preset dynamic scheduling algorithm. Resource allocation tasks are executed based on the resource scheduling strategy.

2. The resource scheduling method as described in claim 1, characterized in that, After the step of semantically parsing the electronic submission form input by the user, the method further includes: If a vague description is detected in the electronic submission form, the vague description is completed using natural language processing technology.

3. The resource scheduling method as described in claim 1, characterized in that, The step of evaluating and filtering the demand features using a pre-trained multimodal demand assessment model and a preset weight configurator to obtain demand features that meet preset conditions includes: The demand features are input into a pre-trained multimodal demand assessment model, and the pre-trained multimodal demand assessment model outputs the demand assessment results. Based on a preset multi-dimensional evaluation index system, the evaluation results of the requirements are weighted and calculated using a preset weight configurator to generate a comprehensive score. If the overall score is greater than or equal to the preset target score threshold, then the requirement feature is determined to meet the preset conditions.

4. The resource scheduling method as described in claim 3, characterized in that, Before the step of evaluating and filtering the demand features using a pre-trained multimodal demand assessment model and a preset weight configurator to obtain demand features that meet preset conditions, the following steps are included: Based on historical multimodal demand data and corresponding demand result labels, an initial multimodal demand assessment model including a feature fusion layer is constructed. The demand features are input into the initial multimodal demand assessment model. The parameters of the multimodal demand assessment model are trained and optimized using deep learning algorithms and regularization strategies to obtain a pre-trained multimodal demand assessment model.

5. The resource scheduling method as described in claim 3, characterized in that, Before the step of evaluating and filtering the demand features using a pre-trained multimodal demand assessment model and a preset weight configurator to obtain demand features that meet preset conditions, the method further includes: Based on a pre-defined three-level circuit breaker review protocol, the compliance of the aforementioned requirements is verified.

6. The resource scheduling method as described in claim 1, characterized in that, The step of combining the demand characteristics that meet the preset conditions with the pre-collected resource information and generating a resource allocation strategy based on a preset dynamic scheduling algorithm includes: Based on the demand characteristics that meet the preset conditions, a demand-resource matching model is constructed. Based on the demand and resource matching model, semantic matching is performed on the approved demand features and the pre-collected resource information to obtain an initial resource scheduling strategy. The final resource scheduling strategy is generated by combining the resource scheduling strategy and the preset dynamic scheduling algorithm.

7. The resource scheduling method as described in claim 6, characterized in that, Based on the demand and resource matching model, the steps of semantically matching the approved demand features and pre-collected resource information to obtain the initial resource scheduling strategy include: Based on the demand and resource matching model, semantic matching is performed on the approved demand features and the pre-collected resource information to obtain the specific requirements corresponding to the demand features; Using a pre-defined planning algorithm, the optimal resource allocation solution within the predicted time frame is calculated based on the current resource information status and demand forecast. By introducing a priority queue and a load balancing algorithm, the demand characteristics are prioritized to obtain a priority-sorted task queue. Based on the specific requirements corresponding to the demand characteristics, the optimal solution for resource allocation, and the priority-sorted task queue, an initial resource scheduling strategy is obtained.

8. The resource scheduling method as described in claim 6, characterized in that, The step of generating the final resource scheduling strategy by combining the resource scheduling strategy and the preset dynamic scheduling algorithm includes: Based on the workload and urgency of the demand characteristics, the demand characteristics are classified using the preset dynamic scheduling algorithm to obtain scheduling suggestions; Based on the scheduling suggestions, the initial resource scheduling strategy is optimized to obtain the final resource scheduling strategy.

9. The resource scheduling method as described in claim 8, characterized in that, The step of classifying demand characteristics based on workload and urgency using the preset dynamic scheduling algorithm to obtain scheduling suggestions includes: If the aforementioned demand characteristics trigger the emergency recruitment scheduling conditions, an emergency demand scheduling suggestion will be generated. If the workload of the aforementioned requirement features exceeds a preset threshold, a project-level requirement scheduling suggestion will be generated. If the workload of the required features does not exceed a preset threshold, an iterative requirement scheduling suggestion is generated.

10. The resource scheduling method as described in claim 1, characterized in that, The steps of performing resource allocation tasks based on the resource allocation strategy include: During the execution of resource allocation tasks, an abnormality is alerted through a light-up warning mechanism, and an illegal request is blocked through a veto mechanism.

11. The resource scheduling method as described in claim 1, characterized in that, Following the step of executing the resource allocation task based on the resource allocation strategy, the method further includes: Feedback data collected during the execution of resource allocation tasks is input into the multimodal demand assessment model and dynamic resource scheduling module; By comparing the feedback data with the prediction results of the multimodal demand assessment model and the execution results of the resource scheduling strategy, the parameter configuration of the multimodal demand assessment model and the resource scheduling strategy are adjusted.

12. A resource scheduling system, characterized in that, The resource scheduling system includes: The data acquisition and preprocessing module performs semantic parsing on the electronic submission form input by the user, generating a standardized requirement summary and corresponding requirement features; The evaluation module is used to evaluate and filter the demand features through a pre-trained multimodal demand evaluation model and a preset weight configurator to obtain demand features that meet preset conditions. The resource scheduling module is used to combine the approved demand characteristics with the pre-collected resource information and generate a resource scheduling strategy based on a preset dynamic scheduling algorithm. The task execution module is used to execute resource allocation tasks based on the resource scheduling strategy.

13. A resource scheduling device, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the resource scheduling method as described in any one of claims 1 to 11.

14. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the resource scheduling method as described in any one of claims 1 to 11.

15. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the steps of the resource scheduling method as described in any one of claims 1 to 11.

Images