Statistical analysis-based engineering cost dynamic adjustment method and system

Abstract

The invention discloses a statistical analysis-based engineering cost dynamic adjustment method and system, and aims to solve the problems of many errors and poor accuracy and reliability when the engineering cost is changed due to factors such as material price fluctuation, the method comprises the steps of obtaining a price change message and extracting related data, matching an engineering material data table by using a preset template, and obtaining an engineering material data table; the system queries and processes price data according to a change material identifier, carries out source diversification processing on the change price data, and also relates to initial construction and dynamic updating of an engineering cost template, the system is composed of a price change module, an engineering material data table module, a to-be-changed module, a to-be-replaced module and other modules, and all the modules work cooperatively to realize corresponding functions. The method has the following advantages that on the aspect of data processing, multi-channel data are automatically integrated, prices are fused, and the operation risk is reduced; in the aspect of template construction and updating, historical data and algorithms are utilized, and the material matching accuracy is improved; during price adjustment, the confidence interval is calculated based on statistical analysis, and the accuracy and reliability of the project cost are effectively guaranteed.

Description

Technical Field

[0001] The present invention belongs to the technical field of project cost, and particularly relates to a dynamic adjustment method and system for project cost based on statistical analysis. Background Art

[0002] In the field of construction projects, especially during the bidding stage, the determined project cost often needs to be adjusted accordingly according to the fluctuations of market prices. This is because, over time, the costs of various materials and resources may change, thus affecting the overall project budget. In traditional practices, this adjustment process relies on manual operations, where cost engineers need to carefully search for materials with price changes in the project cost documents and manually update the prices. This method is not only cumbersome but also extremely inefficient and prone to errors. With the continuous development of the construction industry, the types of materials have become more and more, and the influencing factors of prices have become more and more complex, which puts higher requirements on the modification methods of project cost. Intelligence and precision have become urgent needs in order to improve work efficiency, reduce the occurrence of errors, and ensure the accuracy and reliability of project cost. Summary of the Invention

[0003] The purpose of the present invention is to improve work efficiency, reduce the occurrence of errors, and ensure the accuracy and reliability of project cost, and to provide a dynamic adjustment method and system for project cost based on statistical analysis;

[0004] To achieve the above purpose, the present invention adopts the following technical solutions: A dynamic adjustment method for project cost based on statistical analysis, including:

[0005] Step 1: Obtain the construction material price change message, and extract the construction material price change material identifier and the construction material change price data from it;

[0006] Step 2: Use a preset project cost template to match the project material data table from the project file data, and the project material data table includes project material identifiers;

[0007] Step 3: Use the construction material price change material identifier to match and query in the project material data table, and mark the successful ones as the material identifiers to be changed;

[0008] Step 4: Obtain the original price data of the construction materials. It is known that the change amount data follows a normal distribution N(μ,σ 2 ), and the values of μ and σ are obtained through the collection and statistical analysis of multiple construction material price change data. If P 矫正 is between [μ - σ, μ + σ], then use P 矫正 as the new price; otherwise, μ is the new price.

[0009] Further, in step 1, the source of the price change data of building materials is diversified, integrating the government project cost information network, building materials e-commerce platforms, and the report data of market research institutions, and the prices of the same material obtained from different sources are P1, P2, …, P n , and the corresponding credibility weights are w1, w2, …, w n Then the corrected price

[0010] Further, the corrected price P 矫正 The specific acquisition steps are as follows:

[0011] Step S1: For the collected building material price change messages, remove the obvious errors and duplicate data records, and for the missing values, use mean filling and median filling for processing. Then, for the manufacturers of the materials, use one-hot encoding to convert them into numerical forms. Finally, standardize all numerical data and map it to the interval [0, 1] or [-1, 1] to obtain the preprocessed data;

[0012] Step S2: Construct a multi-layer perceptron neural network model according to the number of features of the data;

[0013] Step S3: Train the multi-layer perceptron neural network model according to the preprocessed data to obtain a trained multi-layer perceptron neural network model;

[0014] Step S4: Input the prices of the same material obtained from different sources and the related feature data into the trained multi-layer perceptron neural network model, and the result output by the model is the corrected price P 矫正 .

[0015] Further, the specific steps of step S3 are as follows:

[0016] Step S31: Divide the preprocessed data according to the ratio of 70% training set, 20% validation set, and 10% test set;

[0017] Step S32: Input the training set data into the multi-layer perceptron neural network model, calculate the predicted output of the model through forward propagation, then calculate the error between the predicted value and the true value according to the loss function. Next, use the backpropagation algorithm to calculate the gradients of the error with respect to the weights and biases of each layer of the neural network, and update the weights and biases through the optimizer to gradually reduce the value of the loss function. During the training process, regularly evaluate the performance of the model on the validation set, record the loss value and accuracy of the validation set. When the loss value of the validation set no longer decreases within several consecutive training epochs, it is considered that the model has overfitted or fallen into a local optimal solution. At this time, stop the training to obtain a trained multi-layer perceptron neural network model.

[0018] Further, the specific steps of the preset project cost template in step 2 include: initial template construction and dynamic template update;

[0019] Initial template construction: Obtain historical project plan data and material type information. By matching and classifying material information in historical data, a project cost template is formed. When matching material information, corresponding project material information is obtained from each material type information and used as a matching string. The Levenshtein distance algorithm is used to measure the similarity between the matching string and the string in the historical project plan data. Then, let the project material information string be s1 and the comparison string in the historical project plan data be s2. The Levenshtein distance d(s1, s2) represents the minimum number of edit operations required to convert s1 into s2, and the similarity Set a similarity threshold T. When sim(s1, s2) ≥ T, it is considered a successful match, and the successfully matched material information is used to construct the template;

[0020] Dynamic template update: Regularly collect new data from industry authoritative databases, the latest building standard specifications, and large-scale engineering project case libraries, and then use text mining and data clustering algorithms for analysis to identify changes in material types, price ranges, and correlation relationships, and automatically update the project cost template.

[0021] The present invention also provides a project cost dynamic adjustment system based on statistical analysis, including: a price change module, an engineering material data table module, a to-be-changed module, and a replacement module;

[0022] Price change module: Used to obtain price change messages for building materials and extract building material price change material identifiers and building material change price data from them;

[0023] Engineering material data table module: Used to match an engineering material data table from engineering file data using the preset project cost template. The engineering material data table includes engineering material identifiers;

[0024] To-be-changed module: Used to match and query in the engineering material data table using the building material price change material identifier, and mark the successful ones as to-be-changed material identifiers;

[0025] Replacement module: Obtain the original price data of building materials. It is known that the change amount data follows a normal distribution N(μ, σ 2 )), and the values of μ and σ are obtained through statistical analysis of multiple collections of building material price change data. If P 矫正 is between [μ - σ, μ + σ], then use P 矫正 as the new price; otherwise, μ is the new price.

[0026] Further, in the price change module, specifically, the price data of building materials changes is processed with diversified sources. The government project cost information network, building materials e-commerce platforms, and market research institution report data are integrated, and the prices of the same material obtained from different sources are P1, P2, …, P n , and the corresponding credibility weights are w1, w2, …, w n Then the corrected price

[0027] Further, the engineering material data sheet module includes initial template construction and dynamic template update;

[0028] Among them, initial template construction: Obtain historical engineering plan data and material type information. By matching and classifying material information in historical data, a project cost template is formed; when matching material information, obtain the corresponding engineering material information from each material type information, use it as the matching string, and use the Levenshtein distance algorithm to measure the similarity with the string in the historical engineering plan data. Then set the engineering material information string as s1, the comparison string in the historical engineering plan data as s2, and the Levenshteir distance d(s1, s2) represents the minimum number of editing operations required to convert s1 into s2, and the similarity Set a similarity threshold T. When sim(s1, s2) ≥ T, it is considered a successful match, and the successfully matched material information is used to construct the template;

[0029] Dynamic template update: Regularly collect new data from industry authoritative databases, the latest building standards and specifications, and large-scale engineering project case libraries, and then use text mining and data clustering algorithms for analysis to identify changes in material types, price ranges, and association relationships, and automatically update the project cost template.

[0030] The present invention also provides a computer-readable medium including instructions. When the processing unit of the server executes the instructions, the control system of the server is enabled to execute the method according to any one of claims 1-5.

[0031] The present invention also provides a computer program product containing instructions. When it runs on the server, the server is enabled to execute the method according to any one of claims 1-5.

[0032] Beneficial effects:

[0033] 1. Improve data accuracy: Through the diversified processing of price data sources, the biases and errors that may exist in a single data source are avoided, so that the changed price can better reflect the actual market situation, provide more accurate basic data for project cost calculation, and reduce the cost errors caused by inaccurate price data.

[0034] 2. Enhance the applicability and accuracy of templates: Based on the template construction and update mechanism of the Levenshtein distance algorithm, compared with traditional empirical or simple statistical template construction methods, it can more accurately identify and cover the material information required for projects, and continuously optimize the template with the update of industry data, effectively reducing template mismatch problems caused by material omissions, incorrect selections, or industry changes, and improving the reliability of project cost calculation.

[0035] 3. Improve the scientificity and reliability of decision-making: Considering the statistical characteristics of change price data and calculating the confidence interval, cost engineers can judge the reasonableness of price fluctuations based on this, avoid making wrong decisions due to insufficient understanding of the randomness of price changes during the price adjustment process, enhance the scientificity and reliability of project cost adjustment, and contribute to better controlling project costs and risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is a logic block diagram of the present invention;

[0037] Figure 2 is a logic block diagram of the system. DETAILED DESCRIPTION OF THE INVENTION

[0038] The present invention will be further explained below with reference to the accompanying drawings.

[0039] Example 1

[0040] As Figure 1 shown, the present invention provides a dynamic adjustment method for project cost based on statistical analysis, including:

[0041] Step 1: Obtain the building material price change message, and extract the building material price change material identifier and the building material change price data from it.

[0042] In Step 1, perform source diversification processing on the building material change price data, integrate the government project cost information network, building material e-commerce platform, and market research institution report data, and the prices of the same material obtained from different sources are P1, P2,..., P n , and the corresponding credibility weights are w1, w2,..., w n Then the corrected price

[0043] Among them, the corrected price P 矫正 The specific acquisition steps include:

[0044] Step S1: For the collected building material price change messages, remove the obvious errors and duplicate data records. For the missing values, use mean filling and median filling for processing. Then, for the manufacturers of the materials, use one-hot encoding to convert them into numerical forms. Finally, perform standardization processing on all numerical data and map it to the interval [0, 1] or [-1, 1] to obtain the preprocessed data.

[0045] Step S2: Construct a multi-layer perceptron neural network model according to the number of features of the data. Among them, the hidden layers are connected in a fully connected manner, so that each node can receive the output information of all nodes in the previous layer. The number of nodes in the output layer is 1, that is, the fused material price. The ReLU (Rectified Linear Unit) activation function is used in the hidden layer. The ReLU function can effectively solve the problem of gradient disappearance, accelerate the training speed of the network, and at the same time has the characteristic of sparse activation, which can enable the network to learn more effective feature representations. The linear activation function is used in the output layer because the final fused price is a continuous numerical value, and the linear activation function can directly output the predicted value.

[0046] Step S3: Train the multi-layer perceptron neural network model according to the preprocessed data to obtain a trained multi-layer perceptron neural network model.

[0047] The specific steps for training the multi-layer perceptron neural network model in Step S3 include:

[0048] Step S31: Divide the preprocessed data according to the ratio of 70% training set, 20% validation set, and 10% test set. Among them, the training set is used to train the neural network to let the model learn the rules in the data. The validation set is used to evaluate the performance of the model during the training process and adjust the hyperparameters of the model (such as learning rate, number of hidden layer nodes, etc.) to prevent the model from overfitting. The test set is used to evaluate the generalization ability of the model after the model training is completed, that is, the performance of the model on unseen data.

[0049] Step S32: Input the training set data into the multi-layer perceptron neural network model, calculate the predicted output of the model through forward propagation, then calculate the error between the predicted value and the true value according to the loss function. Next, use the backpropagation algorithm to calculate the gradients of the error with respect to the weights and biases of each layer of the neural network, and update the weights and biases through the optimizer to gradually reduce the value of the loss function. During the training process, regularly evaluate the performance of the model on the validation set, record the loss value and accuracy of the validation set. When the loss value of the validation set no longer decreases within a continuous number of training rounds (such as 5 rounds), it is considered that the model has overfitted or fallen into a local optimal solution. At this time, stop the training to obtain a trained multi-layer perceptron neural network model.

[0050] Step S4: In the actual engineering cost modification scenario, input the same material prices and related feature data from different sources into the trained multi-layer perceptron neural network model, and the result output by the model is the corrected price P 矫正 。

[0051] Step 2: Use a preset engineering cost template to match the engineering material data table from the engineering document data. The engineering material data table includes engineering material identifiers.

[0052] Specifically, in this embodiment, the engineering cost template refers to a template that is preset and used to match the materials used in the construction plan. The engineering document data refers to the documents that need to be bid on. The engineering material data table refers to a data table that records the materials used in the engineering document data. The engineering material identifier refers to the identifier of the specific materials used in the engineering document.

[0053] The specific steps of the preset engineering cost template in Step 2 include: initial template construction and dynamic template update;

[0054] Initial template construction: Obtain historical engineering plan data and material type information, match and classify material information in the historical data to form an engineering cost template. When matching material information, obtain the corresponding engineering material information from each material type information as the matching string, use the Levenshtein distance algorithm to measure the similarity between the matching string and the string in the historical engineering plan data. Then, let the engineering material information string be s1, the comparison string in the historical engineering plan data be s2, and the Levenshteir distance d(s1, s2) represents the minimum number of editing operations required to convert s1 into s2. The similarity Set a similarity threshold T. When sim(s1, s2) ≥ T, it is considered a successful match, and the successfully matched material information is used to construct the engineering material data table;

[0055] When matching material information, even if there are differences in the character order between the engineering material information string and the comparison string in the historical engineering plan data, the similarity value can be obtained through this algorithm. For example, for the engineering material information string "cement No. 425" and the comparison string "No. 425 cement" in the historical engineering plan data, the similarity between the two can be accurately judged with the help of this algorithm, so that the successfully matched material information can be used to construct the template, avoiding omissions or misjudgments caused by differences in material name expressions, and improving the accuracy of material information recognition.

[0056] Dynamic template update: Regularly collect new data from industry authoritative databases, the latest building standards and specifications, and large-scale engineering project case databases, and then use text mining and data clustering algorithms for analysis to identify material information for automatically updating the engineering material data table.

[0057] Step 3: Use the building material price change material identification to match and query in the engineering material data sheet, and the successfully marked ones are the material identifications to be changed.

[0058] Step 4: Obtain the original price data of building materials. It is known that the change amount data follows a normal distribution N(μ,σ 2 ), and the values of μ and σ are obtained through statistical analysis of the collected building material price change data for many times. If P 矫正 is between [μ - σ, μ + σ], then use P 矫正 as the new price; otherwise, μ is the new price.

[0059] In summary, in terms of data processing, the limitation of obtaining prices from a single data source is broken through, and a multi-channel data fusion method is adopted to determine the change price. In terms of template construction, a historical data mining and dynamic update mechanism based on string similarity algorithm is introduced. In the price adjustment link, the change price is analyzed and processed in combination with statistical principles, providing a more scientific basis for the adjustment of project cost. Overall, a more comprehensive, intelligent and accurate project cost modification system is formed.

[0060] Embodiment 2

[0061] As Figure 2 shown, the present invention also provides a project cost dynamic adjustment system based on statistical analysis, including: a price change module, an engineering material data sheet module, a module to be changed, and a replacement module.

[0062] Price change module: used to obtain price change messages for building materials, and extract the building material price change material identification and building material change price data therefrom.

[0063] Specifically, in the price change module, the building material change price data is processed with diversified sources, integrating the government project cost information network, building materials e-commerce platforms and market research institution report data, and the prices of the same material obtained from different sources are P1, P2,..., P n , and the corresponding credibility weights are w1, w2,..., w n Then the corrected price

[0064] Engineering material data sheet module: used to match the engineering material data sheet from the engineering file data using a preset project cost template, and the engineering material data sheet includes engineering material identifications.

[0065] Specifically, in this embodiment, the project cost template refers to a template that is pre-set and matches the materials used in the construction plan in the construction plan. The project document data refers to the documents that need to be bid. The project material data table refers to a data table that records the materials used in the project document data. The project material identifier refers to the identifier of the materials specifically used in the project document.

[0066] The project material data table module includes initial template construction and dynamic template update;

[0067] Among them, for initial template construction: obtain historical project plan data and obtain material type information. By matching and classifying material information in historical data, a project cost template is formed; when matching material information, obtain the corresponding project material information from each material type information, use it as a matching string, and use the Levenshtein distance algorithm to measure the similarity with the string in the historical project plan data. Then, set the project material information string as s1, the comparison string in the historical project plan data as s2, and the Levenshtein distance d(s1, s2) represents the minimum number of editing operations required to convert s1 to s2. The similarity Set a similarity threshold T. When sim(s1, s2) ≥ T, it is considered a successful match, and the successfully matched material information is used to construct the template;

[0068] For dynamic template update: regularly collect new data from industry authoritative databases, the latest building standard specifications, and large-scale engineering project case libraries, and then use text mining and data clustering algorithms for analysis to identify changes in material types, price ranges, and correlation relationships, and automatically update the project cost template.

[0069] The module to be changed: used to match and query in the project material data table using the changed material identifier of building materials, and mark the successful ones as the material identifiers to be changed.

[0070] The replacement module: obtain the original price data of building materials. It is known that the changed quantity data follows a normal distribution N(μ, σ 2 ) and the values of μ and σ are obtained through statistical analysis of the collected building material price change data for many times. If P 矫正 is between [μ - σ, μ + σ], then use P 矫正 as the new price; otherwise, μ is the new price.

[0071] Combining the above two embodiments, the project cost dynamic adjustment method and system provided by the present invention have the following advantages:

[0072] (1) Improve data accuracy: By diversifying the sources of price data, the biases and errors that may exist in a single data source are avoided, enabling the changed prices to better reflect the actual market conditions, providing more accurate basic data for engineering cost calculation, and reducing cost errors caused by inaccurate price data.

[0073] (2) Enhance the applicability and accuracy of templates: The template construction and update mechanism based on the Levenshtein distance algorithm can more accurately identify and cover the material information required for the project compared to traditional empirical or simple statistical template construction methods. Moreover, the template is continuously optimized with the update of industry data, effectively reducing template mismatch problems caused by material omission, wrong selection, or industry changes, and improving the reliability of engineering cost calculation.

[0074] (3) Improve the scientificity and reliability of decision-making: By considering the statistical characteristics of the changed price data and calculating the confidence interval, cost engineers can use this to judge the reasonableness of price fluctuations, avoid making wrong decisions due to insufficient understanding of the randomness of price changes during the price adjustment process, enhance the scientificity and reliability of engineering cost adjustment, and help better control project costs and risks.

[0075] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art in this technical field, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A method for dynamic adjustment of construction cost based on statistical analysis, characterized in that: include: Step 1: Obtain the building material price change message, and extract the building material price change material identifier and building material price change data; Step 2: Use the preset engineering cost template to match the engineering material data sheet from the engineering file data, where the engineering material data sheet includes the engineering material identification; Step 3: Use the material identification of building material price change to match and query in the engineering material data table, and the successfully marked material identification is the material identification to be changed; Step 4: Obtain the original price data of building materials. It is known that the change data follows the normal distribution N(μ,σ 2 ), after collecting and analyzing the data of price changes of building materials for many times, we can get the values ​​of μ and σ. 矫正 Between [μ-σ,μ+σ], use P 矫正 is the new price; otherwise, μ is the new price.

2. The method for dynamic adjustment of construction cost based on statistical analysis according to claim 1 is characterized in that: In step 1, the source of building material price change data is diversified, integrating the government engineering cost information network, building materials e-commerce platform and market research agency report data, and the prices of the same material obtained from different sources are P1, P2, ..., P n , the corresponding credibility weight is The corrected price 3. The method for dynamic adjustment of construction cost based on statistical analysis according to claim 2 is characterized in that: The corrected price P 矫正 The specific acquisition steps include: Step S1: For the collected building material price change messages, remove the data records with obvious errors and duplications, and use mean filling and median filling for missing values. Then, for the material manufacturers, use the one-hot encoding method to convert them into numerical form. Finally, standardize all numerical data and map them to the interval [0,1] or [-1,1] to obtain preprocessed data. Step S2: construct a multi-layer perceptron neural network model according to the number of features of the data; Step S3: training the multi-layer perceptron neural network model according to the preprocessed data to obtain a trained multi-layer perceptron neural network model; Step S4: Input the prices of the same material from different sources and related feature data into the trained multi-layer perceptron neural network model, and the model output result is the corrected price P 矫正 .

4. The method for dynamic adjustment of construction cost based on statistical analysis according to claim 3 is characterized in that: The specific steps of step S3 include: Step S31: Divide the preprocessed data into 70% training set, 20% validation set, and 10% test set; Step S32: input the training set data into the multi-layer perceptron neural network model, calculate the predicted output of the model through forward propagation, and then calculate the error between the predicted value and the true value according to the loss function. Then, use the back propagation algorithm to calculate the gradient of the error to the weights and biases of each layer of the neural network, and update the weights and biases through the optimizer to gradually reduce the loss function value. During the training process, regularly evaluate the performance of the model on the validation set, record the loss value and accuracy of the validation set, and when the loss value of the validation set no longer decreases within multiple consecutive training rounds, it is considered that the model is overfitting or has fallen into a local optimal solution. At this time, stop training to obtain a trained multi-layer perceptron neural network model.

5. The method for dynamic adjustment of construction cost based on statistical analysis according to claim 1 is characterized in that: The specific steps of the engineering cost template preset in step 2 include: initial template construction and dynamic template update; Initial template construction: obtain historical engineering plan data and material type information, match and classify material information in historical data to form an engineering cost template, and when matching material information, obtain the corresponding engineering material information from each material type information, use it as a matching string, and use the Levenshtein distance algorithm to measure the similarity between the matching string and the string in the historical engineering plan data. Then, let the engineering material information string be s1, and the comparison string in the historical engineering plan data be s2. The Levenshtein distance d(s1,s2) represents the minimum number of editing operations required to convert s1 to s2, and the similarity Set the similarity threshold T. When sim(s1,s2)≥T, the match is considered successful, and the successfully matched material information is used to construct the template; Dynamic template update: Regularly collect new data from authoritative industry databases, the latest building standards and specifications, and large-scale engineering project case libraries, and then use text mining and data clustering algorithm analysis to identify changes in material types, price ranges, and correlation relationships, and automatically update the engineering cost template.

6. The dynamic adjustment system of engineering cost based on statistical analysis is characterized by: include: Includes price change module, engineering material data sheet module, pending change module and replacement module; Price change module: used to obtain price change messages for building materials, and extract the material identification and price change data of building materials from them; The engineering material data sheet module is used to match the engineering material data sheet from the engineering file data using a preset engineering cost template, and the engineering material data sheet includes an engineering material identification; To-be-changed module: used to match and query the material identification of building material price changes in the engineering material data table, and the successfully marked material identification is the to-be-changed material identification; Replacement module: Get the original price data of building materials. It is known that the change amount data follows the normal distribution N(μ,σ 2 ), after collecting and analyzing the data of price changes of building materials for many times, we can get the values ​​of μ and σ. 矫正 Between [μ-σ,μ+σ], use P 矫正 is the new price; otherwise, μ is the new price.

7. The construction cost dynamic adjustment system based on statistical analysis according to claim 6 is characterized in that: In the price change module, specifically, the source of building material price change data is diversified, integrating the government engineering cost information network, building materials e-commerce platform and market research agency report data, and the prices of the same material obtained from different sources are P1, P2, ..., P n , the corresponding credibility weight is The corrected price 8. The construction cost dynamic adjustment system based on statistical analysis according to claim 6 is characterized in that: The engineering material data table module includes initial template construction and dynamic template update; Among them, the initial template construction: obtain historical engineering plan data and material type information, and form an engineering cost template by matching and classifying the material information in the historical data; when matching the material information, obtain the corresponding engineering material information from each material type information, use it as a matching string, and use the Levenshtein distance algorithm to measure the similarity with the string in the historical engineering plan data. Then, let the engineering material information string be s1, and the comparison string in the historical engineering plan data be s2. The Levenshtein distance d(s1,s2) represents the minimum number of editing operations required to convert s1 to s2, and the similarity Set the similarity threshold T. When sim(s1,s2)≥T, the match is considered successful, and the successfully matched material information is used to construct the template; Dynamic template update: Regularly collect new data from authoritative industry databases, the latest building standards and specifications, and large-scale engineering project case libraries, and then use text mining and data clustering algorithm analysis to identify changes in material types, price ranges, and correlation relationships, and automatically update the engineering cost template.

9. A computer-readable medium comprising instructions, characterized in that: When the processing unit of the server executes the instruction, the control system of the server executes the method according to any one of claims 1-5.

10. A computer program product comprising instructions, characterized in that When the method is executed on a server, the server is enabled to execute the method according to any one of claims 1 to 5.

Images