Intelligent cost compiling and auditing method based on artificial intelligence in communication network

Abstract

The invention discloses an intelligent cost compilation and auditing method based on artificial intelligence in a communication network, and the method comprises the steps: carrying out the dynamic evaluation of network resources based on a multi-service demand analysis model, generating an optimized resource distribution scheme according to an evaluation result, carrying out the intelligent compilation of cost through a deep learning algorithm, and carrying out the auditing of the cost. Automatically auditing the compiled cost to ensure the rationality and accuracy, and adjusting a resource allocation strategy based on an auditing result; according to the intelligent cost compiling and auditing method based on artificial intelligence in the communication network, the problem that in the prior art, under the condition that multiple service requirements coexist, resource allocation is unbalanced in the method is solved, specifically, when multiple projects or tasks are carried out at the same time, the system cannot effectively process all the requirements due to unreasonable resource allocation, and the system is not easy to carry out. The resource allocation problem may cause reduction of service quality, and further affects the overall efficiency of project cost management and user experience in a communication network construction process.

Description

Technical Field

[0001] The present invention relates to the technical field of communication networks, and in particular to an intelligent cost compilation and review method based on artificial intelligence in a communication network. Background Art

[0002] In the process of modern communication network construction, the preparation and review of engineering cost is a complex and crucial link. The traditional cost preparation and review method usually relies on manual operation. The process is cumbersome and easily affected by human factors, resulting in low efficiency and frequent errors. In addition, with the rapid development of communication network technology, the projects involved are becoming more and more diversified and complex. Traditional methods can hardly meet the needs of today's communication projects. In recent years, the rapid progress of artificial intelligence technology has provided new solutions for the intelligent preparation and review of engineering cost. The intelligent cost preparation and review method based on artificial intelligence uses artificial intelligence technology. Through automated data analysis and intelligent algorithms, it can quickly and accurately calculate and review various expenses in the process of communication network construction, greatly improving the accuracy and efficiency of engineering cost. This intelligent method not only reduces human operation errors, but also significantly saves time and costs.

[0003] However, the existing AI-based intelligent cost compilation and auditing methods still have some limitations in practical applications. Especially when multiple business needs coexist, this method will encounter the problem of unbalanced resource allocation. Specifically, when multiple projects or tasks are carried out at the same time, the system cannot effectively handle all needs due to unreasonable resource allocation. This resource allocation problem may lead to a decline in service quality, which in turn affects the overall efficiency of engineering cost management and user experience during the construction of communication networks. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent cost compilation and auditing method based on artificial intelligence in a communication network, so as to solve the problem of unbalanced resource allocation encountered in the prior art when multiple business needs coexist. Specifically, when multiple projects or tasks are carried out at the same time, the system cannot effectively handle all needs due to unreasonable resource allocation. This resource allocation problem may lead to a decline in service quality, thereby affecting the overall efficiency of engineering cost management and user experience during the construction of the communication network.

[0005] To achieve the above object, the present invention provides the following technical solution: an intelligent cost compilation and review method based on artificial intelligence in a communication network, the method comprising: Dynamically evaluate network resources based on multi-service demand analysis models; Generate an optimized resource allocation plan based on the evaluation results; Use deep learning algorithms to intelligently compile construction costs; Automatically review the compiled cost to ensure its rationality and accuracy; Adjust resource allocation strategies based on the review results.

[0006] Preferably, the dynamic evaluation of network resources based on the multi-service demand analysis model includes the following steps: Collect and analyze business data based on the needs and characteristics of various businesses; Mapping business data to network resource demand models; Dynamically evaluate network resources to identify areas of low resource utilization and high load; Based on the evaluation results, optimized resource allocation suggestions are put forward to alleviate the problem of unbalanced resource allocation.

[0007] Preferably, mapping the service data to the network resource demand model comprises the following steps: Extract key business indicators based on historical business data, such as bandwidth requirements, transmission delay, etc. Use machine learning algorithms to extract features and normalize business data; Based on the feature extraction results, a mapping relationship between services and network resources is established; Based on the mapping relationship and resource evaluation results, a preliminary resource allocation plan is generated under the condition: Q = P1×A+P2×B / A+B, where P1 represents bandwidth demand, P2 represents transmission delay, and A and B represent weight coefficients respectively.

[0008] Preferably, establishing a mapping relationship between services and network resources based on the feature extraction result comprises the following steps: Select appropriate resource types based on the extracted business features; Matching the extracted business features with the availability of resource types; The fitness of resources is calculated by the formula F=P3×C+P4×D, where P3 represents the fitness of resource type, P4 represents the availability of resources, and C and D represent weight coefficients respectively; Based on the fitness results, the resource allocation plan is further refined.

[0009] Preferably, the calculation of the fitness of the resource by the formula F=P3×C+P4×D comprises the following steps: Based on resource type fitness P3 and resource availability P4, select appropriate weight coefficients C and D; Use the weighted sum method to calculate the total fitness F; Determine whether the fitness F meets the preset threshold T, that is, F>T. If so, the resource type matching is successful; Adjust the final resource allocation plan based on the successfully matched resource types.

[0010] Preferably, the step of determining whether the fitness F meets a preset threshold T comprises the following steps: Based on the importance and urgency of the business, set the fitness threshold T; The importance and urgency of the business are calculated by the formula G=E1×P5+E2×P6 / E1+E2, where P5 represents the importance of the business, P6 represents the urgency of the business, and E1 and E2 represent the weight coefficients respectively; If G>7, it is considered that the business requires higher service quality; The fitness threshold T is adjusted based on the priority of the service to ensure that high-priority services are allocated resources first.

[0011] Preferably, the calculation of the importance and urgency of the business by the formula G=E1×P5+E2×P6 / E1+E2 comprises the following steps: Based on historical data and real-time monitoring, obtain business importance and urgency parameters P5 and P6; Set reasonable weight coefficients E1 and E2 to reflect the characteristics of different businesses; The comprehensive score G is calculated using the weighted average method; The priority of resource allocation is adjusted based on the comprehensive score G.

[0012] It can be seen from the above technical solution that the present invention has the following beneficial effects: The intelligent cost preparation and auditing method based on artificial intelligence in this network dynamically evaluates network resources based on a multi-business demand analysis model, generates an optimized resource allocation plan according to the evaluation results, uses a deep learning algorithm to intelligently prepare the cost, automatically audits the prepared cost, and adjusts the resource allocation strategy based on the audit results, thereby overcoming the problem of unbalanced resource allocation in the existing technology, realizing the reasonable allocation of resources in the case of multiple concurrent tasks, and being able to effectively allocate computing resources when processing multiple projects or tasks, avoiding the decline in service quality due to improper resource allocation, thereby ensuring the accurate calculation and audit of various expenses in the process of communication network construction, and solving the problem of unbalanced resource allocation encountered by the existing technology when multiple business needs coexist. Specifically, when multiple projects or tasks are carried out simultaneously, the system cannot effectively handle all needs due to unreasonable resource allocation. This resource allocation problem may lead to a decline in service quality, which in turn affects the overall efficiency of engineering cost management and user experience during the construction of the communication network. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0014] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0015] like Figure 1 As shown, an intelligent cost preparation and auditing method based on artificial intelligence in a communication network achieves the goal of improving service quality and efficiency in the process of network resource allocation and cost preparation through the following key steps: First, a dynamic evaluation of network resources is performed based on a multi-service demand analysis model. The main purpose of this process is to evaluate the matching degree between the status of available resources in the network and the demand according to various current and predicted business needs, so as to identify areas with excess or shortage of resources. In this step, the algorithm establishes a model that accurately reflects the current network environment and future development trends by collecting and analyzing a large amount of network performance indicators, user behavior data, market dynamics and other information. This step lays a solid foundation for subsequent resource optimization and cost preparation. For example, in a specific embodiment, when a telecom operator needs to launch a new high-definition video streaming service, the system will collect historical traffic data, user preference evaluation, and competitive product performance report information in the area, analyze the bandwidth resources required to provide this new service and the user's demand for this service, and ensure the comprehensiveness and accuracy of the resource evaluation process. An optimized resource allocation plan is generated based on the evaluation results. After clarifying the resource requirements of each business area, the system further calculates the best resource allocation strategy. This step not only takes into account the needs of a single task, but also focuses on the coordination and balance between multiple tasks to maximize and optimize the utilization of the entire network resources. Advanced algorithm tools such as linear programming, dynamic programming, and genetic algorithms are used in the process to effectively improve the quality of the solution. Specifically, in the same instance, after comprehensive consideration of the needs of high-definition video streaming services and other key services such as existing voice calls and data transmission, the system has formulated a detailed resource allocation policy to reasonably allocate data center bandwidth resources and physical infrastructure resources shared by various services to ensure that each business can obtain the best support within the scope allowed by its priority, reducing the negative effects caused by mutual influence. Deep learning algorithms are used to intelligently compile construction costs. In order to more accurately estimate the costs required to implement the above resource allocation plan, deep learning technology is applied to cost compilation in this stage. By training neural network models on a large number of historical project data sets, the algorithm can capture the complex nonlinear relationship between cost factors and final expenses, and then predict the cost amount that is closer to the actual situation. This step greatly improves the automation level and work efficiency of budget compilation while ensuring the high quality of budget data. For example, in the above implementation case, by analyzing the actual return on investment, equipment procurement costs, operation and maintenance expenses of similar service launch projects in the past, the deep learning model can quickly and accurately calculate the expected total investment amount for the implementation of this project, and provide reliable data support for decision makers. Automatically audit the compiled cost to ensure its rationality and accuracy. This step is to verify the authenticity and reliability of the cost data and the logical consistency behind it. The audit procedures include but are not limited to cross-checking the amounts of each sub-item, comparing the difference analysis of similar projects, reviewing the cost composition ratio, etc. The purpose is to find out possible errors or potential risks and make corrections when necessary. For example, in our case, the automatic audit mechanism found that the equipment purchase price estimate was slightly high. After further inspection, it was determined that this was caused by not fully considering the price discount of the latest procurement channel. Timely adjustment of this value not only ensured the professional level of budget preparation but also saved the company a certain amount of cost. Adjust the resource allocation strategy based on the audit results to improve service quality. In the last step, review the initial resource allocation strategy based on the audit feedback information obtained in the previous step, and make appropriate adjustments accordingly. This is not only to further reduce costs and improve efficiency, but also to ensure that all businesses can get a better experience. In the above application scenario, it was found after adjustment that the computing power quota of some edge computing nodes can be slightly reduced while meeting the quality of video services. This not only frees up more valuable computing power to serve other business modules such as emerging businesses such as cloud games, but also promotes overall network load balancing, thereby effectively avoiding the problem of service quality degradation due to unbalanced resource allocation when multiple business needs coexist.

[0016] In summary, this intelligent cost preparation and review process based on AI technology not only simplifies the traditional manual labor intensity, but more importantly, it introduces a closed-loop management framework, which continuously improves the quality of communication services through continuous iteration, allowing it to respond flexibly to the increasingly complex multi-business environment.

[0017] Next, the present invention is described to dynamically evaluate network resources based on the multi-service demand analysis model, including the following steps: Collect and analyze business data based on the needs and characteristics of various businesses; Mapping business data to network resource demand models; Dynamically evaluate network resources to identify areas of low resource utilization and high load; Based on the evaluation results, optimized resource allocation suggestions are put forward to alleviate the problem of unbalanced resource allocation and thus improve service quality.

[0018] First, based on the demand characteristics of various businesses, collect and analyze business data. This step involves collecting various types of data related to different businesses, such as traffic, bandwidth requirements, delay sensitivity, etc. Business data can be collected with the help of network monitoring tools and business logs. Specifically, for intelligent cost compilation and review in communication networks, it may be necessary to collect network traffic data and user request records for different time periods. For example, in one embodiment, the network management software can record the amount of data transmitted per minute, as well as the specific requirements of different applications for network resources. These data will be used for subsequent analysis and modeling.

[0019] Secondly, the business data is mapped to the network resource demand model. The purpose of this step is to convert the collected business data into a demand forecast for network resources by analyzing it. The model can be built through a machine learning algorithm, such as a linear regression or decision tree model. Next, the network resources are dynamically evaluated to identify areas with low resource utilization and high load. In this step, the resource demand model built in the previous step is used to monitor the resource usage of each node in the network in real time. By comparing the actual usage with the predicted demand, it is possible to identify which areas have resource waste or shortage. For example, in a specific embodiment, a threshold can be set to determine the level of resource utilization. If the actual usage of a node is lower than, it is considered that the resource utilization of the node is low, otherwise it is considered to be high load. This allows the problem area to be quickly located.

[0020] Finally, based on the evaluation results, an optimized resource allocation suggestion is proposed to alleviate the problem of unbalanced resource allocation, thereby improving service quality. Through the dynamic evaluation in the previous step, it can be clarified which areas need to increase resource investment and which areas can appropriately reduce resource allocation. Optimization suggestions may include but are not limited to adjusting bandwidth allocation, migrating loads, or reconfiguring server resources. Specifically, in one embodiment, if it is detected that the resource utilization of certain servers is continuously low, it can be considered to migrate part of the load to other servers with higher utilization, thereby improving the overall resource utilization efficiency, reducing cost expenditures, and improving the stability and response speed of network services.

[0021] Through the above steps, dynamic management and optimization of network resources can be effectively achieved, ensuring the best service quality in a multi-service environment.

[0022] Next, the present invention is described to map the service data to the network resource demand model, including the following steps: Based on historical business data, extract key business indicators, such as bandwidth requirements, transmission delay, etc. Use machine learning algorithms to extract features and normalize business data; Based on the feature extraction results, a mapping relationship between services and network resources is established; Based on the mapping relationship and resource evaluation results, a preliminary resource allocation plan is generated, with the condition: Q=P1×A+P2×B / A+B.

[0023] Extract key business indicators based on historical business data. This step aims to determine the key factors that affect the performance of the communication network by analyzing past business activities. These indicators include but are not limited to bandwidth demand, transmission delay, packet loss rate, etc. In one embodiment, suppose a telecom operator collects traffic data for the past year and finds through statistical analysis that bandwidth demand surges in certain periods, resulting in increased transmission delay, so these indicators become key factors.

[0024] Use machine learning algorithms to extract features and normalize business data. This process involves further processing the extracted key business indicators to make them suitable for subsequent modeling. Machine learning algorithms, such as principal component analysis and K-means clustering, can be used for feature extraction, while normalization is to eliminate the dimensional differences between different indicators to make the model more robust. For example, the PCA algorithm can extract the principal components that best reflect business needs from multiple dimensions, and then use the minimum-maximum normalization method to scale these eigenvalues ​​to between 0 and 1.

[0025] Based on the feature extraction results, a mapping relationship between services and network resources is established. Based on the feature extraction results, a specific mapping relationship model is established. This model can be a linear regression, decision tree or other complex machine learning model. Specifically, in a telecommunications network, a random forest model can be trained to input features such as bandwidth requirements and transmission delays, and output corresponding network resource requirements, such as server resources and bandwidth allocation.

[0026] Based on the mapping relationship and the resource evaluation result, a preliminary resource allocation plan is generated, and the condition is Q = P1×A+P2×B / A+B. Next, the present invention describes the steps of establishing a mapping relationship between services and network resources based on the feature extraction result: Based on the extracted business features, select the appropriate resource type; in this step, by analyzing the demand characteristics of the business, such as bandwidth requirements, delay sensitivity, etc., select the resource type that can meet these requirements. Specifically, different types of network resources differ in transmission capacity, processing speed, and security, and these factors need to be considered when selecting. For example, in high-real-time video transmission services, low-latency, high-bandwidth transmission paths and resource nodes will be given priority. In one embodiment, for big data processing tasks that require high-speed data transmission, the system may select Fiber Channel with large-capacity transmission capabilities as the preferred resource.

[0027] Match the extracted business features with the availability of resource types; after selecting the appropriate resource type, the next step is to verify whether the selected resource is sufficient to meet the current business needs. This involves evaluating factors such as the current load of the resource, remaining bandwidth or storage space. Ensure that the resources not only match the business needs technically, but also provide sufficient support in time and space. Specifically, for an online trading system that is running at peak times, it may be necessary to check the available resources of related servers and network links to ensure that they can cope with the sudden surge in access requests.

[0028] The fitness of resources is calculated by the formula F=P3×C+P4×D, where P3 represents the fitness of the resource type, P4 represents the resource availability, and C and D represent weight coefficients respectively; this formula comprehensively considers the adaptability of the resource type and the actual availability of the resource to evaluate the overall fitness of the resource. Specifically, the value range of P3 is 0 to 1, and the higher the value, the more suitable the resource type is for the business needs; the value range of P4 is also 0 to 1, and the higher the value, the better the availability of the resource. The weight coefficients C and D are usually set according to the specific requirements of the business to adjust the importance of different types of resources. For example, in a real-time video streaming transmission scenario, if high-bandwidth transmission is more critical, C can be set to a higher value to increase the proportion of resource type fitness in the fitness calculation.

[0029] Based on the fitness results, the resource allocation plan is further refined; finally, according to the fitness results calculated in the previous step, the resource allocation plan is optimized and adjusted to ensure optimal resource utilization. Resources with high fitness will be allocated to the corresponding services first. Specifically, in a complex communication network, if the fitness of a path is significantly higher than other paths, the system will give priority to this path for data transmission to ensure optimal service performance. In this way, resource utilization and service quality can be maximized.

[0030] Finally, based on the successfully matched resource types, the final resource allocation plan is adjusted. When a resource type passes the fitness test, the system will give priority to allocating that resource to tasks or projects. This step ensures the effective use of resources and the smooth progress of the project. For example, in the above communication network project, if a frequency band resource passes the fitness test, then the frequency band will be given priority for critical communication tasks, thereby optimizing the overall network performance.

[0031] In one embodiment, suppose there is a communication network project, and the project leader needs to select the best frequency band for a new communication link. Assume there are three candidate frequency bands: A, B, and C. The fitness of frequency band A, P3 = 0.8, availability P4 = 0.7, the fitness of frequency band B, P3 = 0.9, availability, P4 = 0.4, the fitness of frequency band C, P3 = 0.7, availability P4 = 0.9, set weight coefficient C = 0.7, D ​​= 0.3, preset threshold T = 0.6. Calculate the fitness of the three frequency bands: Frequency band A: F = 0.8 × 0.7 + 0.7 × 0.3 = 0.77; Frequency band B: F = 0.9 × 0.7 + 0.4 × 0.3 = 0.75; Frequency band C: F = 0.7 × 0.7 + 0.9 × 0.3 = 0.7; The fitness of all frequency bands is greater than the preset threshold T=0.6, but since frequency band A has the highest fitness; F=0.77, the system will eventually select frequency band A as the new communication link Next, the present invention is described to determine whether the fitness F meets the preset threshold T, including the following steps: Based on the importance and urgency of the business, set the fitness threshold T; The importance and urgency of the business are calculated by the formula G = E1×P5+E2×P6 / E1+E2; where P5 represents the importance of the business, P6 represents the urgency of the business, and E1 and E2 represent weight coefficients respectively; If G>7, it is considered that the business requires higher service quality; The fitness threshold T is adjusted based on the priority of the service to ensure that high-priority services are allocated resources first.

[0032] First, the fitness threshold T is set according to the importance and urgency of the business. This step involves analyzing the impact of different businesses on the overall system and determining which businesses are more critical. For example, if a video conferencing service in a communication network is critical to business operations and its interruption will result in significant losses, then the fitness threshold T of this service may be set higher.

[0033] Next, the importance and urgency of the business are calculated using the formula G = E1×P5+E2×P6 / E1+E2. In the formula, P5 represents the score of the importance of the business, which usually ranges from 1 to 10, with 10 indicating very important; P6 represents the score of the urgency of the business, which also ranges from 1 to 10, with 10 indicating very urgent. The weight coefficients E1 and E2 are used to adjust the relative impact of the importance and urgency of the business, and the values ​​range from 0 to 1, indicating the importance of these two factors respectively. In the optimal case, if the business is both important and urgent, E1 and E2 are close to 1. The purpose of this formula is to comprehensively consider the multiple attributes of the business, so as to more accurately assess its overall needs.

[0034] If the calculated G value is greater than 7, it is considered that the service requires a higher quality of service. The threshold of 7 here is an empirical value that aims to balance the needs of different services and the resource limitations of the system. When G is greater than 7, it indicates that the importance or urgency of the service exceeds the average level and requires more resources and support to ensure service quality.

[0035] The fitness threshold T is adjusted based on the priority of the service to ensure that high-priority services are allocated resources first. For example, in one embodiment, a real-time voice call service of a communication network is considered to be a very important service. After calculation, the G value is 8.5, which exceeds the threshold 7. Therefore, the fitness threshold T of the service is increased accordingly to ensure its service quality. Specifically, assuming that the initial fitness threshold T is 0.8, it is adjusted to 0.9 according to the high priority of the service to ensure that the voice call service can still maintain good performance and stability when the network load is high.

[0036] These steps together constitute a complete process for determining whether the fitness F meets the preset threshold T, ensuring efficient and reliable operation of key services in the communication network.

[0037] Next, the present invention is described to calculate the importance and urgency of a service through the formula G=E1×P5+E2×P6 / E1+E2, including the following steps: Based on historical data and real-time monitoring, obtain business importance and urgency parameters P5 and P6; Set reasonable weight coefficients E1 and E2 to reflect the characteristics of different businesses; The comprehensive score G is calculated using the weighted average method; The priority of resource allocation is adjusted based on the comprehensive score G to ensure improved service quality.

[0038] First, based on historical data and real-time monitoring, obtain the importance and urgency parameters P5 and P6 of the service. This step involves collecting a large amount of historical service data and real-time monitoring data to evaluate the importance P5 and urgency P6 of each service. For example, in a communication network, the importance parameter P5 may be determined based on factors such as the number of users, traffic demand, and revenue contribution; while the urgency parameter P6 may be determined based on factors such as the frequency of failures, service interruption time, and user complaint rate. These parameters usually take values ​​in the range of 0 to 1, where 1 represents the highest importance or urgency.

[0039] Next, set reasonable weight coefficients E1 and E2 to reflect the characteristics of different services. The selection of weight coefficients is based on the service type and specific needs to ensure that the final calculated comprehensive score G can accurately reflect the actual situation of the service. For example, in a communication network, for key services such as medical and financial services, a higher E1 value may be set to indicate that these services are more important; while for ordinary services such as social media and entertainment services, a higher E2 value may be set to indicate that their urgency has a greater impact. The weight coefficients E1 and E2 are usually positive numbers, and E1+E2 should be greater than 0, but the specific optimal value needs to be adjusted and verified according to the actual application scenario.

[0040] Then, the comprehensive score G is calculated using the weighted average method. The comprehensive score G is calculated using the formula G=E1×P5 + E2×P6 / E1+E2. The weighted average in the formula takes into account the different contributions of importance P5 and urgency P6 to the comprehensive score of the business. This setting allows the final score to reflect not only the importance of the business itself, but also the urgency of the business, thereby more comprehensively evaluating the overall priority of the business.

[0041] Finally, the priority of resource allocation is adjusted based on the comprehensive score G to ensure the improvement of service quality. For example, in one embodiment, there are multiple services in the communication network of a certain telecommunications operator. The comprehensive score G of each service is calculated by the above formula, and the service with the highest score will obtain a higher resource allocation priority, such as a larger bandwidth, more computing resources, etc. In this way, services with high importance and high urgency can be processed more promptly and effectively, thereby improving the overall service quality. For example, the G value of a key financial transaction service is relatively high, and the system will give priority to ensuring the stable operation of the service, even at the expense of some low-scoring ordinary business resources. Specifically, this can be achieved by dynamically adjusting the routing path, adding cache nodes, etc. This resource optimization strategy not only improves user satisfaction, but also reduces unnecessary operating costs.

[0042] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent cost compilation and review method based on artificial intelligence in a communication network, characterized in that: The method comprises: S1. Dynamically evaluate network resources based on the multi-service demand analysis model. Specifically, divide the network resource demand into multiple indicator layers, use the fuzzy comprehensive calculation method to obtain the fuzzy weight vector of each indicator, and defuzzify it to obtain the deterministic weight. The specific formula is: ； in, It means that after defuzzification, the demand index The certainty weight value of , and Respectively indicate indicators The lower limit, middle value and upper limit of the fuzzy number; Based on the deterministic weight of the indicator and the current resource usage status, the comprehensive evaluation value of resource demand is calculated. ； in, Represents the comprehensive evaluation value of resource demand, It means that after defuzzification, the demand index The certainty weight value of Indicates The current status value of each demand indicator, Indicates the total number of demand indicators; S2. Generate an optimized resource allocation plan based on the evaluation results; S3. Use deep learning algorithms to intelligently compile construction costs; S4. Automatically review the compiled cost; S5. Adjust resource allocation strategy based on the audit results.

2. According to claim 1, the method for intelligent cost compilation and review based on artificial intelligence in a communication network is characterized in that: The S1 includes: Collect and analyze business data based on the needs and characteristics of various businesses; Mapping business data to network resource demand models; Dynamically evaluate network resources to identify areas of low resource utilization and high load; Based on the evaluation results, optimized resource allocation suggestions are put forward to alleviate the problem of unbalanced resource allocation.

3. According to claim 2, the method for intelligent cost compilation and review based on artificial intelligence in a communication network is characterized in that: Mapping the service data to the network resource demand model includes: Extract key business indicators such as bandwidth requirements and transmission delay based on historical business data; Use machine learning algorithms to extract features and normalize business data; Based on the feature extraction results, a mapping relationship between services and network resources is established; Based on the mapping relationship and resource evaluation results, a preliminary resource allocation plan is generated under the condition: Q = P1×A+P2×B / A+B, where P1 represents bandwidth demand, P2 represents transmission delay, and A and B represent weight coefficients respectively.

4. According to claim 3, the method for intelligent cost compilation and review based on artificial intelligence in a communication network is characterized in that: The step of establishing a mapping relationship between services and network resources based on the feature extraction result includes the following steps: Select appropriate resource types based on the extracted business features; Matching the extracted business features with the availability of resource types; The fitness of resources is calculated by the formula F=P3×C+P4×D, where P3 represents the fitness of resource type, P4 represents the availability of resources, and C and D represent weight coefficients respectively; Based on the fitness results, the resource allocation plan is further refined.

5. According to the method for intelligent cost compilation and review based on artificial intelligence in a communication network according to claim 4, it is characterized in that: The fitness of the resource calculated by the formula F=P3×C+P4×D includes: Based on resource type fitness P3 and resource availability P4, select appropriate weight coefficients C and D; Use the weighted sum method to calculate the total fitness F; Determine whether the fitness F meets the preset threshold T, that is, F>T. If so, the resource type matching is successful; Adjust the final resource allocation plan based on the successfully matched resource types.

6. According to claim 5, the method for intelligent cost compilation and review based on artificial intelligence in a communication network is characterized in that: The determination of whether the fitness F meets the preset threshold T includes: Based on the importance and urgency of the business, set the fitness threshold T; The importance and urgency of the business are calculated by the formula G=E1×P5+E2×P6 / E1+E2, where P5 represents the importance of the business, P6 represents the urgency of the business, and E1 and E2 represent the weight coefficients respectively; If G>7, it is considered that the business requires higher service quality; The fitness threshold T is adjusted based on the priority of the service to ensure that high-priority services are allocated resources first.

7. According to claim 6, the method for intelligent cost compilation and review based on artificial intelligence in a communication network is characterized in that: The calculation of the importance and urgency of the business by the formula G=E1×P5+E2×P6 / E1+E2 includes: Based on historical data and real-time monitoring, obtain business importance and urgency parameters P5 and P6; Set reasonable weight coefficients E1 and E2 to reflect the characteristics of different businesses; The comprehensive score G is calculated using the weighted average method; The priority of resource allocation is adjusted based on the comprehensive score G.

Images