Multi-industry complex engineering investment dynamic tracking management method

Abstract

The application provides a multi-industry complex engineering investment dynamic tracking management method, which is applied to the information technology field of engineering construction industry, and solves the investment tracking lag problem caused by data islands in traditional management by constructing a multi-source data acquisition module, acquiring real-time verification and pricing data, change design data and material price difference adjustment data, establishing a linkage mechanism among investment factor data, constructing a risk prediction model, calculating price difference influence weight and change design influence weight, inputting the price difference influence weight and the change design influence weight into a multi-target constraint analysis screening strategy library, and automatically generating an investment adjustment solution, replacing the manual experience-based adjustment strategy, providing a quantitative basis, avoiding investment control risks, and improving the response efficiency under investment early warning.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information technology in the construction industry, and in particular to a multi-industry complex engineering investment dynamic tracking management method. BACKGROUND

[0002] In the field of multi-industry complex engineering construction, investment management needs to consider cost, progress, resource allocation and other multi-dimensional dynamic changes. The traditional management mode is difficult to meet the fine control requirements due to problems such as data fragmentation and response lag. With the expansion of engineering scale and the increase of technical complexity, how to realize the real-time linkage of investment factors, intelligent early warning of risks and dynamic adjustment of strategies has become a key technical problem that needs to be solved in the industry.

[0003] In the prior art, engineering investment management mainly relies on manual statistics and experience judgment: first, construction investment, inspection pricing, design change and material price difference data are scattered in different business modules, lacking real-time linkage mechanism, leading to serious lag in investment dynamic tracking. When facing sudden factors such as design changes and material price fluctuations, the generation of adjustment strategies relies on manual experience, lacks intelligent analysis of multi-source data and deep mining of historical cases, resulting in high investment control risk and low response efficiency. SUMMARY

[0004] The present application provides a multi-industry complex engineering investment dynamic tracking management method to solve the defects of high investment management control risk and low response efficiency in the prior art.

[0005] On the one hand, the present application provides a multi-industry complex engineering investment dynamic tracking management method, comprising: integrating inspection pricing data acquisition interface, design change data acquisition interface, material price difference adjustment data acquisition interface, obtaining multi-source data acquisition module;

[0006] Receiving the real-time completion value of the engineering investment output by the multi-source data acquisition module;

[0007] Comparing and analyzing the real-time completion value of the engineering investment with the investment control baseline, generating a deviation rate, and performing abnormal data classification early warning according to the deviation rate;

[0008] Inputting the real-time completion value of the engineering investment into a risk prediction model, outputting a price difference influence weight and a design change influence weight;

[0009] Inputting the price difference influence weight and the design change influence weight into a multi-objective constraint analysis and screening strategy library, outputting an investment adjustment solution; the multi-objective constraint analysis and screening strategy library is a neural network model integrating multi-objective optimization algorithm, constraint condition rule and preset strategy template.

[0010] Optionally, the receiving the project investment real-time completion value output by the multi-source data collection module comprises:

[0011] Receiving the inspection valuation data output by the inspection valuation data collection interface;

[0012] Receiving the change design data output by the change design data collection interface after the reply is completed;

[0013] Receiving the material price difference adjustment data output by the material price difference adjustment data collection interface;

[0014] Calculating the sum of the inspection valuation data, the change design data, and the material price difference data to obtain the project investment real-time completion value.

[0015] Optionally, the comparing and analyzing the project investment real-time completion value with the investment control baseline to generate a deviation rate, and performing abnormal data classification early warning according to the deviation rate, comprises:

[0016] Determining the date of the project investment real-time completion value;

[0017] Obtaining the investment plan corresponding to the date;

[0018] Calculating the deviation rate based on the investment plan, the investment control baseline, and the project investment real-time completion value;

[0019] If the deviation rate is within a first preset range, a first-level early warning is generated;

[0020] If the deviation rate is within a second preset range, a second-level early warning is generated;

[0021] If the deviation rate is within a third preset range, a third-level early warning is generated.

[0022] Optionally, the inputting the project investment real-time completion value into a risk prediction model to output a price difference impact weight and a change design impact weight, comprises:

[0023] Using a long short-term memory network to train material price difference historical data to generate a material price difference prediction interval value;

[0024] Using an AI large model to perform semantic analysis on historical change design data to establish a mapping relationship between change types and investment increments;

[0025] According to the mapping relationship between the change types and the investment increments and the material price difference prediction interval value, using a causal inference method, a price difference impact weight and a change design impact weight are generated; the price difference impact weight and the change design impact weight are used to quantify the contribution proportion of price difference and change design to the risk prediction result.

[0026] Optionally, the method further includes:

[0027] Based on the predicted range of material price difference, and combined with the material price difference weighting coefficient, the first impact weight of material price difference on investment control is predicted.

[0028] By using an AI big data model to analyze the design change documents, the change type, change amount and schedule impact factors are extracted. Based on the mapping relationship between the change type and the investment increment, and combined with the design change weight coefficient, the second impact weight of the change data on investment control is predicted.

[0029] If the sum of the first influence weight and the second influence weight is greater than the preset influence weight, an engineering investment early warning event and an overall method for correcting engineering investment control are generated; the overall method for correcting engineering investment control includes material procurement plan optimization suggestions and design change optimization suggestions.

[0030] Optionally, the method further includes:

[0031] If the weight of the price difference is greater than the weight of the design change, then the material price difference data is input into the price difference prediction model, and the material price difference trend within a preset time period is output.

[0032] Optionally, the step of inputting the material price difference data into the price difference prediction model and outputting the material price difference change trend within a preset time period includes:

[0033] The material price difference data is cleaned to remove duplicate, missing, or abnormal data records, resulting in preprocessed data.

[0034] The preprocessed data is input into a pre-trained price difference prediction model, which outputs a predicted material price difference value.

[0035] The moving average method is used to generate a curve showing the change in the price difference over time for the predicted material price difference, thus obtaining the trend of the material price difference.

[0036] Optionally, the method further includes:

[0037] Calculate and predict the incremental investment in materials based on the trend of the material price difference;

[0038] Calculate the incremental investment for the design change based on the impact weights of the design change;

[0039] Calculate the ratio of the predicted increase in material investment to the increase in investment due to design changes to obtain the impact priority coefficient;

[0040] If the priority coefficient of the affected area is less than the preset impact coefficient, the design change is determined to be the main influencing factor, and a change risk warning report is generated.

[0041] If the priority coefficient of the affected material is greater than the preset impact coefficient, the material price difference is determined to be the main influencing factor, and a price difference fluctuation warning report is generated.

[0042] Optionally, the step of inputting the price difference impact weight and the design change impact weight into a multi-objective constraint analysis screening strategy library to output an investment adjustment solution includes:

[0043] An impact radiation analysis was performed on the impact weights of the design changes to obtain the impact radiation matrix;

[0044] The price difference influence weights are quantified to obtain the price difference quantification characteristics;

[0045] The influence radiation matrix and the price difference quantization features are input into a convolutional neural network to extract feature vectors.

[0046] The feature vector is input into a multi-objective constraint analysis screening strategy library to output an investment adjustment solution.

[0047] Optionally, inputting the feature vector into a multi-objective constraint analysis screening strategy library and outputting an investment adjustment solution includes:

[0048] Calculate the similarity weight between the feature vector and the historical engineering database;

[0049] Based on the aforementioned similarity weights, a weighted historical experience vector is generated;

[0050] The historical experience vector is processed using the Sigmoid function to output a device scheduling priority matrix;

[0051] The historical experience vector is processed using the Tanh function to output the control intensity value in the interval [-1,1].

[0052] The historical experience vector is processed using the ReLU function to output the number of days for schedule adjustment.

[0053] On the other hand, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the dynamic tracking and management method for investment in complex multi-industry projects as described above.

[0054] On the other hand, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the dynamic tracking and management method for investment in complex multi-industry projects as described above.

[0055] On the other hand, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the dynamic tracking and management method for investment in complex multi-industry projects as described above.

[0056] As can be seen from the above technical solutions, the multi-industry complex engineering investment dynamic tracking management method provided by the present invention includes: integrating a data acquisition interface for work verification and pricing, a data acquisition interface for design changes, and a data acquisition interface for material price difference adjustment to obtain a multi-source data acquisition module; receiving the real-time completion value of engineering investment output by the multi-source data acquisition module; comparing and analyzing the real-time completion value of engineering investment with the investment control baseline to generate a deviation rate, and performing abnormal data classification and early warning based on the deviation rate; inputting the real-time completion value of engineering investment into a risk prediction model to output the price difference impact weight and the design change impact weight; inputting the price difference impact weight and the design change impact weight into a multi-objective constraint analysis screening strategy library to output an investment adjustment solution; the multi-objective constraint analysis screening strategy library is a neural network model integrating multi-objective optimization algorithms, constraint rules, and preset strategy templates. By constructing a multi-source data acquisition module, real-time data on work verification and pricing, design changes, and material price difference adjustments are obtained, establishing a linkage mechanism among investment element data, and solving the problem of investment tracking lag caused by data silos in traditional management. By constructing a risk prediction model, the impact weights of price differences and design changes are calculated. These impact weights are then input into a multi-objective constraint analysis screening strategy library to automatically generate investment adjustment solutions. This replaces adjustment strategies dominated by manual experience, provides quantitative basis, avoids investment management risks, and improves response efficiency under investment early warning. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0058] Figure 1 This is a schematic diagram of the process for dynamic tracking and management of investment in complex multi-industry projects provided in this embodiment of the invention;

[0059] Figure 2 This is a schematic diagram of the output investment adjustment solution method provided in an embodiment of the present invention;

[0060] Figure 3 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present invention.

[0061] Figure label:

[0062] 310. Processor; 320. Communication interface; 330. Memory; 340. Communication bus. Detailed Implementation

[0063] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0064] like Figure 1 As shown, Figure 1 This is a flowchart of the dynamic tracking and management method for investment in complex multi-industry projects provided in this embodiment of the invention. The method mainly includes the following steps:

[0065] 101. Integrate the data acquisition interfaces for work verification and pricing, design change, and material price difference adjustment to obtain a multi-source data acquisition module;

[0066] Specifically, the multi-source data acquisition module integrates different data acquisition interfaces to achieve comprehensive collection of data on work completion and pricing, design changes, and material price difference adjustments. The work completion and pricing data acquisition interface collects project progress and cost information to ensure accurate calculation of project payments; the design change data acquisition interface monitors design changes and promptly captures the impact of design adjustments on investment and schedule; and the material price difference adjustment data acquisition interface tracks material market price fluctuations in real time and calculates material price differences, providing data support for cost control. Through the efficient integration of these interfaces, the multi-source data acquisition module comprehensively improves the comprehensiveness and accuracy of data collection, providing a foundation for subsequent investment management decisions.

[0067] 102. Receive the real-time completed value of engineering investment output from the multi-source data acquisition module.

[0068] Specifically, the engineering cost data received from the multi-source data acquisition module includes:

[0069] Receive the work verification and valuation data output from the work verification and valuation data collection interface;

[0070] Receive the revised design data after approval is completed from the revised design data acquisition interface;

[0071] Receive material price difference data output from the material price difference adjustment data acquisition interface;

[0072] The sum of the work completion and pricing data, design change data, and material price difference data is calculated to obtain the real-time completed value of the project investment.

[0073] Understandably, progress measurement and valuation data, design change data, and material price difference data are all key information in the dynamic management of engineering project investment. Progress measurement and valuation data reflects the actual progress and cost of the project, serving as an important basis for calculating project payments. Design change data, once approved, reflects the impact of design changes on investment and schedule, helping to adjust investment plans and construction schemes in a timely manner. Material price difference data reflects the market prices of construction materials; comparing current material prices with those stipulated in the contract plays a crucial role in cost control.

[0074] The real-time completion value of the project investment can be obtained by summing the work completion and pricing data, design change data, and material price difference data.

[0075] Calculating the real-time completed investment value of a project based on progress measurement data, design change data, and material price difference data requires consideration of multiple factors, including but not limited to the project's scale, complexity, material costs, labor costs, and potential risks. First, the investment amount for completed work is calculated based on the actual progress and costs reflected in the progress measurement data. Next, by combining the approved design change data, a reasonable estimate of the investment changes resulting from the design changes is made and incorporated into the calculation of the actual investment value. Simultaneously, close monitoring of material price data is conducted, and current market prices are compared with the material prices stipulated in the contract, allowing for dynamic adjustments to material costs. By comprehensively considering these factors, the calculated real-time completed investment value can be ensured to be more accurate and comprehensive, providing strong support for subsequent investment management decisions.

[0076] Therefore, in some embodiments, the specific steps for determining the real-time completion value of the engineering investment include:

[0077] 201. Input the verification and valuation data to the actual completed value of project investment, and output the actual completed value of project investment data;

[0078] The pricing relationship is (1):

[0079] (1)

[0080] This indicates the real-time completed value of the project investment. This represents real-time work completion and pricing data. This indicates incremental data related to design changes. This indicates the adjustment data for material price spreads;

[0081] 202. Input the changed design data into the changed design calculation formula, and output the changed design incremental data;

[0082] The change design calculation formula is (2):

[0083] ; (2)

[0084] in, Indicates the first For each change in the quantity of work, it is necessary to distinguish between the increased quantity and the decreased quantity, with the decreased quantity represented by a negative value; Indicates the first The unit price for the change, which includes costs for labor, materials, machinery, etc. Indicates the first The adjustment factor for the change, where the adjustment factor represents the urgency or process complexity, and the default adjustment factor is 1.

[0085] 203. Input material price difference adjustment data into the material price difference adjustment formula, and output material price difference data;

[0086] The material price difference adjustment formula is (3):

[0087] ; (3)

[0088] in, Indicates the first The actual unit price of the material; Indicates the first The contractual price for the materials Indicates the first The actual consumption of this material.

[0089] 204. Input the work completion and pricing data, design change data, and material price difference data into the cost relationship, and output the real-time completed value of the project investment.

[0090] For example, in the second quarter of a bridge project, the work completion and pricing work was carried out. First, the work completion cost was calculated. During this stage, 100 pile foundations were completed, with a contract unit price of 8,000 yuan per pile foundation. Without any other adjustment factors, the work completion cost was calculated as 100 × 8,000 = 800,000 yuan.

[0091] Next, the incremental cost of the design change was calculated. A new bridge abutment reinforcement project was added, with a volume of 50 cubic meters and a unit price of 1200 yuan / cubic meter. Due to the need for expedited work, the urgency factor was set to 1.1. Substituting these figures into the change cost relationship, the change cost is calculated as: 50 × 1200 × 1.1 = 66,000 yuan.

[0092] To calculate the material price difference adjustment, taking steel as an example, the real-time steel price is 5500 yuan / ton, the contract price is 5000 yuan / ton, and the steel consumption during this period is 20 tons. Therefore, the material price difference adjustment is calculated as 20 × (5500 - 5000) = 10000 yuan.

[0093] Finally, calculate the real-time completed value of the project investment. ,use , the previously calculated =800,000 yuan =66,000 yuan Substituting 10,000 yuan into the equation, we get... =800,000 + 66,000 + 10,000 = 867,000 yuan. Taking into account the cost of construction work, design changes, and material price fluctuations, the actual investment in the bridge project for the second quarter was 867,000 yuan.

[0094] 103. Compare and analyze the real-time completed value of project investment with the investment control baseline, generate the deviation rate, and conduct graded early warning of abnormal data based on the deviation rate.

[0095] Specifically, the steps for comparing and analyzing the real-time completed value of engineering investment with the investment control baseline, generating a deviation rate, and issuing graded early warnings for abnormal data based on the deviation rate include:

[0096] The date for determining the real-time completion value of the project investment;

[0097] Obtain the investment plan corresponding to the specified date;

[0098] The deviation rate is calculated based on the investment plan, investment control baseline, and real-time completion value of engineering investment.

[0099] The investment control baseline is the total investment amount of the project. Taking construction engineering as an example, the actual investment value in the second quarter was compared with the investment plan. The investment plan is a pre-set investment amount based on factors such as project schedule, historical data, and market forecasts, representing the expected investment level of the construction project at different stages. For example, if the project is divided into three stages, there are investment plans for each stage. Therefore, by drawing the investment plans for the three stages in chronological order, the investment control baseline can be obtained. By comparing the actual investment data with the investment plan at each stage, it can be ensured that the investment amount is within the preset range.

[0100] In practice, the investment control baseline was first displayed in chart form. Then, after the real-time completion value of engineering investment was generated, it was compared with the investment control baseline. The comparison revealed that the real-time completion value of engineering investment in the second quarter was close to the preset investment control baseline, with a relatively large deviation rate, but still within an acceptable range. This indicates that although the investment was slightly close to the preset investment, the overall situation was still under control. In response, a timely investment analysis was conducted to identify the reasons for the larger actual investment data and corresponding adjustment measures were proposed to ensure that investment in subsequent stages could be controlled within the preset range.

[0101] In addition, the classification and early warning of abnormal data based on the deviation rate specifically includes:

[0102] If the deviation rate is within the first preset range, a Level 1 warning will be generated;

[0103] If the deviation rate is within the second preset range, a level two warning will be generated;

[0104] If the deviation rate is within the third preset range, a level three warning will be generated.

[0105] In this embodiment of the invention, the first preset range is 80%-90%, the second preset range is 90%-95%, and the third preset range is greater than 95%. A level 1 warning indicates that the investment deviation is small but still needs attention; a level 2 warning indicates that the investment deviation is large and requires high attention from management; and a level 3 warning indicates that the investment deviation has seriously exceeded expectations and immediate action is needed to correct it.

[0106] For example, system pop-ups and SMS notifications can ensure that information is delivered to relevant personnel in a timely and accurate manner. Once a level 2 or level 3 warning is triggered, the system will automatically send pop-up reminders and SMS notifications to the management of the construction unit and the developer, reminding them to pay attention to the current investment deviation and take corresponding countermeasures according to the warning level.

[0107] 104. Input the real-time completed value of the project investment into the risk prediction model, and output the weights of the price difference impact and the weights of the design change impact;

[0108] The model inputs the real-time completed value of the engineering investment into the risk prediction model and outputs the weights of the price difference impact and the impact of design changes, specifically including:

[0109] The material price difference prediction interval is generated by training a long short-term memory network on historical material price difference data.

[0110] Using AI large-scale models to perform semantic analysis on historical design change data, a mapping relationship between change types and investment increments is established;

[0111] Based on the predicted range of material price differences and the mapping relationship between change type and investment increment, a causal inference method is used to generate the weights of price difference impact and design change impact.

[0112] Among them, the weighting of price difference and the weighting of design change are used to quantify the contribution ratio of price difference and design change to the risk prediction results.

[0113] Specifically, Long Short-Term Memory (LSTM) networks are a special type of recurrent neural network capable of processing sequential data and capturing dependencies over longer time series. In this embodiment of the invention, an LTM network is trained on historical material price spread data to learn the fluctuation patterns of material price spreads. Through training, the LTM network can generate a model that predicts the fluctuation range of material price spreads. Based on current market conditions and historical data, it can predict the fluctuation range of material price spreads over a future period, thus providing strong support for predicting the influence weight of price spreads.

[0114] Secondly, an AI big data model is used to conduct in-depth semantic analysis of historical design change data. During the analysis, the AI ​​big data model can identify and understand key information in historical design changes, such as the type of change, the reason for the change, and the investment increment before and after the change. Through comprehensive analysis of the type of change, the reason for the change, and the investment increment before and after the change, the AI ​​big data model can establish a mapping relationship between the type of change and the investment increment.

[0115] Then, using causal inference, the weights of price difference impact and design change impact are generated. Taking the generation of the price difference impact weight as an example, it is assumed that in the engineering project, historical material price difference data has been trained using a Long Short-Term Memory (LSTM) network to generate price difference prediction intervals. At this point, a hypothetical intervention condition is introduced: assuming that the price of a certain material will rise or fall by a certain percentage in the future. Then, the impact of the intervention condition on the overall investment risk of the project is calculated. By comparing the risk prediction results before and after adding the intervention condition, the contribution of the price difference to the risk change can be calculated, i.e., the price difference impact weight.

[0116] Similarly, to generate the impact weight of design change, we can assume a specific type of design change occurs and calculate its impact on the overall investment risk of the project. By comparing the risk prediction results before and after the change, we can calculate the degree of contribution of that design change type to the risk change, i.e., the impact weight of the design change.

[0117] By using causal inference methods, corresponding weight values ​​can be generated for different price differences and design change types. The weights of price difference impact and design change impact reflect the proportion of their contribution to the risk prediction results.

[0118] In some embodiments, the multi-industry complex engineering investment dynamic tracking and management method provided by the present invention further includes:

[0119] Based on the predicted range of material price difference and the weighting coefficient of material price difference, the weight of the first impact of material price difference on investment control is predicted.

[0120] By using AI big data models to analyze design change documents, we can extract change types, change amounts, and time-related factors. Based on the mapping relationship between change types and investment increments, and combined with the design change weight coefficients, we can predict the second impact weight of change data on investment control.

[0121] If the sum of the first and second impact weights is greater than the preset impact weight, then an early warning event for engineering investment and an overall method for correcting deviations in engineering investment management will be generated.

[0122] The overall approach to correcting deviations in engineering investment management includes suggestions for optimizing material procurement plans and suggestions for optimizing design changes.

[0123] Specifically, the material price difference weighting coefficient and the design change weighting coefficient are derived from historical data statistical analysis, reflecting the relative importance of material price difference changes and design changes on the overall investment.

[0124] After predicting the first and second impact weights, the weights of the first and second impact weights are summed. If the sum exceeds the preset impact weight threshold, it indicates that the current material price difference or design change has posed a significant risk to the project investment, requiring high attention from management.

[0125] At this point, an engineering investment early warning event is triggered, relevant personnel are promptly notified, and a set of targeted overall methods for engineering investment control and correction is automatically generated. These methods include optimization suggestions for material procurement plans, such as adjusting procurement timing and selecting alternative materials to reduce costs. For example, by combining historical information price and price difference data trends and using AI and other digital technologies, it is predicted that information prices will increase in the third quarter of 2025, causing a significant rise in price differences and posing a risk of exceeding the project investment control baseline. In this case, a timely warning will be issued to managers, suggesting that the procurement plan be brought forward to the second quarter of 2025 to mitigate the risk in advance.

[0126] It also includes optimization strategies for design changes, aiming to effectively control investment increases and ensure the achievement of project economic goals by adjusting design schemes and optimizing construction processes.

[0127] In some embodiments, the multi-industry complex engineering investment dynamic tracking management method provided by the present invention further includes: if the impact weight of price difference is greater than the impact weight of design change, then the material price difference data is input into the price difference prediction model, and the material price difference change trend within a preset time period is output.

[0128] The specific steps for obtaining the trend of material price difference changes are as follows:

[0129] The material price difference data is cleaned to remove duplicate, missing, or abnormal data records, resulting in preprocessed data.

[0130] The preprocessed data is input into the pre-trained price difference prediction model, which outputs the predicted material price difference value.

[0131] The moving average method is used to predict the material price difference, generating a curve showing the change in the price difference over time, thus obtaining the trend of the material price difference.

[0132] Because the raw material price difference data may contain duplicate, missing, or abnormal records, failing to clean this data will severely impact the accuracy of subsequent analyses. For example, data collection may result in missing data due to equipment malfunction, or data anomalies may occur due to human input errors. Abnormal data needs to be identified and corrected during the data cleaning stage. The cleaned data, i.e., preprocessed data, will serve as a reliable foundation for subsequent analyses.

[0133] Next, the preprocessed data is input into a pre-trained price spread prediction model. This model, trained on a large amount of historical data, is capable of capturing the patterns and trends in material price changes. Once the preprocessed data is input, the model calculates the predicted material price spread for a future period based on its internal algorithms and parameters. This predicted price spread will serve as a crucial basis for subsequent decision-making.

[0134] Then, the predicted material price difference is compared with the current material price difference to determine the accuracy of the prediction. If the predicted value differs significantly from the actual value, the prediction model needs to be retrained to improve its accuracy and reliability.

[0135] Finally, by using the moving average method, a curve showing the price difference over time is generated, revealing the trend of material price differences. The moving average method is a commonly used time series analysis technique that smooths data fluctuations by averaging data over a certain period, thereby revealing the long-term trend of the data. In this embodiment of the invention, the moving average method is applied to the price difference to generate a smooth curve representing the trend of material price differences over a future period. By observing this curve, the fluctuation of material prices can be intuitively understood, thus providing strong support for subsequent investment decisions.

[0136] In some embodiments, the multi-industry complex engineering investment dynamic tracking and management method provided by the present invention further includes:

[0137] Calculate and predict the incremental investment in materials based on the trend of material price difference changes;

[0138] Calculate the incremental investment for design changes based on the impact weights of the design changes;

[0139] Calculate the ratio of the predicted increase in material investment to the increase in investment due to design changes to obtain the impact priority coefficient;

[0140] If the impact priority coefficient is less than the preset impact coefficient, the design change is determined to be the main influencing factor, and a change risk warning report is generated.

[0141] If the priority coefficient of the affected material is greater than the preset impact coefficient, the material price difference is determined to be the main influencing factor, and a price difference fluctuation warning report is generated.

[0142] The calculation of the impact priority coefficient helps managers quickly identify the main factors affecting project investment. When the impact priority coefficient is less than the preset impact coefficient, it means that the design change has a more significant impact on project investment. In this case, a change risk warning report will be automatically generated, reminding managers to pay attention to the investment risks that design changes may bring and to take corresponding measures for risk control. Conversely, when the impact priority coefficient is greater than the preset impact coefficient, the fluctuation of material price differences becomes the dominant factor affecting project investment. A price difference fluctuation warning report will be generated, allowing managers to adjust material procurement strategies to reduce investment risks caused by price fluctuations. Through this method, the embodiments of the present invention can achieve dynamic tracking and management of complex project investments in multiple industries, improving the accuracy and efficiency of investment decisions.

[0143] 105. Input the weights of price difference impact and design change impact into the multi-objective constraint analysis screening strategy library, and output investment adjustment solutions.

[0144] The multi-objective constraint analysis screening strategy library is a neural network model that integrates multi-objective optimization algorithms, constraint rules, and preset strategy templates. Through training, the neural network model can identify the relationship between engineering cost data and material price trends. Combined with preset constraint rules, such as cost control and schedule requirements, it automatically screens and generates investment adjustment solutions that meet the requirements. This optimizes resource allocation, reduces cost risks, ensures dynamic adjustment of engineering investment within a preset range, and maximizes both economic and social benefits.

[0145] Specifically, the weights of price difference impact and design change impact are input into the multi-objective constraint analysis screening strategy library, and the output investment adjustment solutions include:

[0146] Perform an impact radiation analysis on the impact weights of the design changes to obtain the impact radiation matrix;

[0147] The impact weights of the price spread are quantified to obtain the price spread quantification characteristics.

[0148] The influence of the radiation matrix and price difference quantization features is input into a convolutional neural network to extract feature vectors;

[0149] The feature vectors are input into a multi-objective constraint analysis screening strategy library to output investment adjustment solutions.

[0150] Among them, the influence radiation matrix can clearly show the weight and degree of the impact of design changes on the overall project, which helps decision-makers quickly identify key change points.

[0151] Furthermore, by quantifying the impact weight of price differences, a quantitative characteristic of price differences is obtained, which can accurately assess the impact of material price differences on investment.

[0152] Finally, the influence radiation matrix and price difference quantification features are input into a convolutional neural network to extract feature vectors. The convolutional neural network can automatically learn and extract key features, providing strong support for subsequent multi-objective constraint analysis.

[0153] After inputting the feature vectors into the multi-objective constraint analysis screening strategy library, the library can output investment adjustment solutions. These solutions comprehensively consider factors such as project schedule, design changes, and material price fluctuations, ensuring the scientific validity and feasibility of the adjustment plans.

[0154] In some embodiments, an impact radiation analysis is performed on the design change data. This analysis, based on complex network theory, constructs an impact radiation matrix by simulating the propagation path and impact weight of design changes within an engineering project. Each element in the impact radiation matrix represents the degree of impact of the design change on a specific part of the project, thus clearly demonstrating the weight and extent of the design change's impact on the overall project. This helps decision-makers quickly identify key change points.

[0155] Quantifying price spread trends primarily involves collecting historical price data for relevant materials and employing statistical methods such as time series analysis and regression analysis to predict the potential trend of material price spreads over a future period. These predicted data are then converted into quantitative price spread characteristics, which can accurately assess the impact of material price spread changes on investment.

[0156] Subsequently, the influence radiation matrix and price spread quantification features are input into a convolutional neural network to extract feature vectors. A convolutional neural network is a deep learning model capable of automatically learning and extracting key features from input data. Taking the aforementioned four dimensions of data as input, the convolutional neural network extracts feature vectors that significantly impact investment decisions through multiple layers of convolution and pooling operations.

[0157] For example, in a large-scale bridge project, design changes caused the actual investment to exceed the preset range. By adopting a dynamic tracking and management method for complex multi-industry projects, an investment adjustment solution was automatically generated based on material price difference trends, work completion and valuation data, and design change data. The solution suggested adjusting the construction sequence of some non-critical paths to reduce the impact of design changes on the overall schedule. Ultimately, the solution successfully controlled the actual investment within the preset range, ensuring the smooth progress of the project.

[0158] In some embodiments, after inputting the feature vector into the multi-objective constraint analysis screening strategy library, the multi-objective constraint analysis screening strategy library can output an investment adjustment solution, wherein, for example... Figure 2 As shown, the specific steps for outputting an investment adjustment solution include:

[0159] 301. Calculate the similarity weight between the feature vector and the historical project database.

[0160] 302. Generate historical experience vectors based on similarity weights.

[0161] 303. Use the Sigmoid function to process the historical experience vector and output the device scheduling priority matrix.

[0162] 304. Use the Tanh function to process the historical experience vector and output the control intensity value in the interval [-1,1].

[0163] 305. Use the ReLU function to process the historical experience vector and output the number of days for adjusting the construction period.

[0164] Specifically, the similarity weights between the feature vectors and the historical project database are calculated. This is done by comparing the feature vectors of the current project with those in the historical project database, using cosine similarity and Euclidean distance algorithms to calculate the similarity, and then converting this into a weight value. A higher weight value indicates a higher similarity between the current project and historical projects, and thus greater reference value of the historical data for the current project.

[0165] For example, the current project is a commercial complex located in the city center, and its feature vector includes project scale, building height, and structural type. In the historical project database, a similar commercial complex project located in the city center with a similar scale and structural type is found. Calculations show that the similarity weight between the current project and this project is 0.85, indicating a high degree of similarity.

[0166] Next, a weighted historical experience vector is generated based on similarity weights. The experience data (such as investment amount, construction period, cost, etc.) in the historical project database are weighted according to similarity weights to obtain a weighted historical experience vector. This historical experience vector reflects how the experience data from historical projects can be applied to the current project under current project conditions.

[0167] Continuing with the aforementioned commercial complex project as an example, assuming the project's investment in the historical project database is 1 billion yuan and the construction period is 24 months, and using a similarity weight of 0.85, the weighted investment amount can be calculated to be 850 million yuan, and the weighted construction period to be 20.4 months. These weighted data constitute the historical experience vector for the current project.

[0168] Then, the historical experience vector is processed using the Sigmoid function to output a device scheduling priority matrix. The Sigmoid function maps input values ​​to the (0,1) interval, thus obtaining a matrix representing device scheduling priorities. The larger the element value in the matrix, the higher the scheduling priority of the corresponding device.

[0169] Taking equipment scheduling as an example, assume that the historical experience vector contains investment amounts and project duration information for various equipment. After processing the investment amount and project duration information using the Sigmoid function, an equipment scheduling priority matrix is ​​obtained. The element values ​​in the matrix reflect the scheduling priority of various equipment under the current project conditions. For example, an element value of 0.95 for a critical equipment indicates that this equipment has a high scheduling priority in the current project.

[0170] Next, the Tanh function is used to process the historical experience vector, outputting a control intensity value in the interval [-1, 1]. The Tanh function maps the input value to the interval [-1, 1], thus obtaining a control intensity value representing the direction and intensity of investment adjustments. A control intensity value greater than 0 indicates that investment needs to be increased or the construction period extended; a control intensity value less than 0 indicates that investment needs to be reduced or the construction period shortened.

[0171] Continuing with the investment adjustment example, assume the historical experience vector shows an investment amount of 850 million yuan, while the current project budget is 900 million yuan. After processing the investment amount using the Tanh function, a control intensity value of 0.3 is obtained. This indicates that under the current project conditions, an increase of approximately 3% (0.3 * 100%) in investment is needed to ensure the smooth progress of the project.

[0172] Finally, the ReLU function is used to process the historical experience vector and output the number of days for schedule adjustment. The ReLU function maps the input value to the interval [0, +∞), thus obtaining a value representing the number of days for schedule adjustment. When the input value is less than 0, the ReLU function outputs 0; when the input value is greater than or equal to 0, the ReLU function outputs the input value itself. This means that under the current project conditions, if the schedule needs to be adjusted, the number of days for adjustment should be greater than or equal to 0.

[0173] Continuing with the example of schedule adjustment, suppose the historical experience vector indicates a schedule duration of 20.4 months, while the current project's planned duration is 20 months. After processing the schedule using the ReLU function, we obtain a schedule adjustment period of 0.4 * 30 = 12 days. This means that under the current project conditions, the schedule needs to be extended by approximately 12 days to ensure the project's smooth progress.

[0174] Through the above steps, the multi-objective constraint analysis screening strategy library can comprehensively consider factors such as project progress, design changes, and material price fluctuations, and output a scientific and feasible investment adjustment solution.

[0175] Figure 3 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present invention.

[0176] like Figure 3 As shown, the electronic device may include a processor 310, a communications interface 320, a memory 330, and a communication bus 340. The processor 310, communications interface 320, and memory 330 communicate with each other via the communication bus 340. The processor 310 can call logical instructions from the memory 330 to execute a dynamic tracking and management method for complex multi-industry engineering investments.

[0177] Furthermore, the logical instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0178] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the multi-industry complex engineering investment dynamic tracking management method provided by the above methods.

[0179] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the multi-industry complex engineering investment dynamic tracking and management method provided by the above methods.

[0180] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0181] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0182] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for dynamic tracking and management of investment in complex multi-industry projects, characterized in that, include: By integrating the data acquisition interfaces for work verification and pricing, design change, and material price difference adjustment, a multi-source data acquisition module is obtained. Receive the work verification and pricing data output by the work verification and pricing data acquisition interface; Receive the revised design data after approval is completed, output by the revised design data acquisition interface; Receive material price difference data output from the material price difference adjustment data acquisition interface; The sum of the work completion and pricing data, the design change data, and the material price difference data is calculated to obtain the real-time completed value of the project investment. The real-time completed value of the project investment is compared and analyzed with the investment control baseline to generate a deviation rate, and abnormal data is graded and warned based on the deviation rate. Input the real-time completed value of the project investment into the risk prediction model, and output the weights of the price difference impact and the impact of the design change impact. An impact radiation analysis was performed on the impact weights of the design changes to obtain the impact radiation matrix; The price difference influence weights are quantified to obtain the price difference quantification characteristics; The influence radiation matrix and the price difference quantization features are input into a convolutional neural network to extract feature vectors. The feature vector is input into a multi-objective constraint analysis screening strategy library to output an investment adjustment solution. The feature vector is input into a multi-objective constraint analysis screening strategy library to output an investment adjustment solution, including: Calculate the similarity weight between the feature vector and the historical engineering database; Based on the aforementioned similarity weights, a weighted historical experience vector is generated; The historical experience vector is processed using the Sigmoid function to output a device scheduling priority matrix; The historical experience vector is processed using the Tanh function to output the control intensity value in the interval [-1,1]. The historical experience vector is processed using the ReLU function to output the number of days for schedule adjustment; the multi-objective constraint analysis screening strategy library is a neural network model that integrates multi-objective optimization algorithms, constraint rules and preset strategy templates.

2. The method for dynamic tracking and management of investment in complex multi-industry projects according to claim 1, characterized in that, The step of comparing and analyzing the real-time completed value of the project investment with the investment control baseline to generate a deviation rate, and then performing graded early warning for abnormal data based on the deviation rate, includes: The date on which the real-time completion value of the project investment is determined; Obtain the investment plan corresponding to the stated date; The deviation rate is calculated based on the investment plan, the investment control baseline, and the real-time completion value of the engineering investment. If the deviation rate is within a first preset range, a level one warning is generated; If the deviation rate is within the second preset range, a level two warning is generated; If the deviation rate is within the third preset range, a level three warning is generated.

3. The method for dynamic tracking and management of investment in complex multi-industry projects according to claim 1, characterized in that, The real-time completed value of the project investment is input into the risk prediction model, and the output weights for the impact of price differences and design changes are included: The material price difference prediction interval is generated by training a long short-term memory network on historical material price difference data. Using AI large-scale models to perform semantic analysis on historical design change data, a mapping relationship between change types and investment increments is established; Based on the predicted range of material price differences and the mapping relationship between the change type and investment increment, a causal inference method is used to generate the price difference impact weight and the change design impact weight; the price difference impact weight and the change design impact weight are used to quantify the contribution ratio of price difference and change design to the risk prediction results.

4. The method for dynamic tracking and management of investment in complex multi-industry projects according to claim 3, characterized in that, The method further includes: Based on the predicted range of material price difference, and combined with the material price difference weighting coefficient, the first impact weight of material price difference on investment control is predicted. By using an AI big data model to analyze the design change documents, the change type, change amount and schedule impact factors are extracted. Based on the mapping relationship between the change type and the investment increment, and combined with the design change weight coefficient, the second impact weight of the change data on investment control is predicted. If the sum of the first influence weight and the second influence weight is greater than the preset influence weight, an engineering investment early warning event and an overall method for correcting engineering investment control are generated; the overall method for correcting engineering investment control includes material procurement plan optimization suggestions and design change optimization suggestions.

5. The method for dynamic tracking and management of investment in complex multi-industry projects according to claim 1, characterized in that, The method further includes: If the weight of the price difference is greater than the weight of the design change, then the material price difference data is input into the price difference prediction model, and the material price difference trend within a preset time period is output.

6. The method for dynamic tracking and management of investment in complex multi-industry projects according to claim 5, characterized in that, The step of inputting the material price difference data into the price difference prediction model and outputting the material price difference change trend within a preset time period includes: The material price difference data is cleaned to remove duplicate, missing, or abnormal data records, resulting in preprocessed data. The preprocessed data is input into a pre-trained price difference prediction model, which outputs a predicted material price difference value. The moving average method is used to generate a curve showing the change in the price difference over time for the predicted material price difference, thus obtaining the trend of the material price difference.

7. The method for dynamic tracking and management of investment in complex multi-industry projects according to claim 6, characterized in that, Also includes: Calculate and predict the incremental investment in materials based on the trend of the material price difference; Calculate the incremental investment for the design change based on the impact weights of the design change; Calculate the ratio of the predicted increase in material investment to the increase in investment due to design changes to obtain the impact priority coefficient; If the priority coefficient of the affected area is less than the preset impact coefficient, the design change is determined to be the main influencing factor, and a change risk warning report is generated. If the priority coefficient of the affected material is greater than the preset impact coefficient, the material price difference is determined to be the main influencing factor, and a price difference fluctuation warning report is generated.

Images