Distribution network engineering mechanical construction cost management and control method and system based on big data
By constructing deep interactive features and mechanical comprehensive efficiency indicators, the cost prediction model was optimized, which solved the problem of inaccurate cost prediction in the mechanized construction of power distribution network projects and achieved more accurate cost prediction.
Patent Information
- Application Number
- CN202511478064.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-16
- Publication Date
- 2025-11-11
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing technologies lack the ability to collect and integrate multi-source heterogeneous data in the mechanized construction of power distribution networks, which fails to deeply depict the impact of mechanical combinations, efficiency, and environmental losses on costs, resulting in unstable cost prediction results and limited accuracy.
By constructing deep interactive features and mechanical comprehensive efficiency indicators, and using a joint loss function to optimize the cost prediction model, multi-source construction data are integrated and mechanical comprehensive efficiency indicators are calculated. Deep interactive features representing the construction status are constructed, and the cost prediction model is optimized to improve prediction accuracy.
This approach enables cost predictions to be closer to actual engineering conditions, aligns with the common-sense engineering principle that low mechanical efficiency leads to increased costs, and improves the reliability and accuracy of the prediction results.
Smart Images
Figure CN120930893A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power grid mechanization cost control technology, and more specifically, to a method and system for controlling the cost of mechanized construction of distribution network projects based on big data. Background Technology
[0002] Distribution network construction is a crucial component of power grid development, and its cost forecasting directly impacts project budgeting, construction planning, and resource allocation efficiency. With the increasing mechanization of distribution network projects, the factors involved in the construction process are becoming increasingly complex. Cost variations are not only influenced by the amount of work and labor costs but also closely related to the combination and efficiency of construction machinery, geographical and environmental conditions, and fluctuations in material prices. Therefore, accurate and intelligent cost forecasting has become a key issue in power grid project management and cost control.
[0003] Currently, traditional forecasting methods mainly rely on reference quotas and engineering experience, lacking flexibility and real-time performance. They are unable to accurately reflect complex and ever-changing construction scenarios. Furthermore, traditional forecasting methods are mostly based on a single data source, such as sensor or management system data, lacking multi-source information fusion and easily affected by data gaps, noise, and spatiotemporal asynchrony. The forecast results are unstable and have limited accuracy. Existing forecasting models usually only consider the single objective of cost, with superficial feature construction and a lack of in-depth modeling of the interaction between mechanical working conditions and the environment, making it difficult to fully characterize the complexity of construction.
[0004] Traditional technical solutions have the following technical problems: Primarily targeting general engineering cost management, its data foundation and processing methods fail to fully consider the highly mechanized nature of engineering projects. It lacks the collection and fusion processing of multi-source heterogeneous data such as machine operating conditions, sensor data, and environmental geographic information, making it unable to depict the profound impact of machine combinations, efficiency, and environmental losses on costs. This results in a single-dimensional and incomplete data base for prediction.
[0005] Existing methods for feature construction mostly remain at the project attribute and macro-market level, lacking sufficient depth. They lack quantitative representation of the deep interactions between construction conditions and the environment, failing to reflect the true complexity and dynamic changes of construction, thus limiting model learning.
[0006] To address the above problems, this invention proposes a solution. Summary of the Invention
[0007] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a method and system for cost control of mechanized construction in power distribution networks based on big data. By constructing deep interactive features and comprehensive mechanical efficiency indicators for mechanized construction in power distribution networks, and constructing a joint loss function based on the deep interactive features and comprehensive mechanical efficiency indicators, the cost prediction model is trained and optimized to solve the problems of large gap between predicted and actual cost values and the disconnect between mechanical efficiency and cost in mechanized construction projects in power distribution networks.
[0008] To achieve the above objectives, the present invention provides the following technical solution: The big data-based method for controlling the cost of mechanized construction in power distribution network projects includes the following steps: Consistency processing and confidence assessment of acquired multi-source construction data, and fusion of the data to generate fused data based on the assessment results; Construction of deep interactive features characterizing the construction status based on the fused data, and calculation of the comprehensive mechanical efficiency index; Construction of a cost prediction model based on the deep interactive features, and construction of a joint loss function based on the output of the cost prediction model and the comprehensive mechanical efficiency index; Training and optimization of the cost prediction model using the joint loss function, and outputting the final predicted cost value.
[0009] In a preferred embodiment, the process of processing the acquired multi-source construction data for consistency specifically involves: aligning the multi-source construction data in the time dimension to generate time-aligned data; mapping the time-aligned data to a unified spatial coordinate system; and performing nearest-neighbor matching between the GPS coordinates of the construction machinery and the coordinates of environmental elements in the geographic information system based on the spherical distance function to obtain a consistent data source.
[0010] In a preferred embodiment, the deep interaction features include market features and working environment interaction features. The working environment interaction features include traction load index, turning angle and mechanical coordination index, hoisting precision positioning time ratio and terrain accessibility score. The comprehensive mechanical efficiency index is calculated by environmental reduction index, power load rate of construction machinery and time utilization rate of construction machinery.
[0011] In a preferred embodiment, the method for obtaining the environmental reduction index is as follows: weather indicators, slope indicators, and surface indicators are calculated based on fused data; the weather indicators, slope indicators, and surface indicators are weighted and synthesized according to preset weights to obtain an environmental risk index; the environmental risk index is converted into a baseline reduction index through linear mapping, and the baseline reduction index is corrected based on correction parameters to obtain the environmental reduction index, wherein the correction parameters are calculated based on historical comprehensive efficiency index data.
[0012] In a preferred embodiment, the construction of a cost prediction model based on deep interaction features, and the construction of a joint loss function based on the output of the cost prediction model and the comprehensive mechanical efficiency index, specifically involves: combining historical deep interaction features with historical comprehensive mechanical efficiency indexes to generate a historical feature dataset; based on the historical feature dataset, constructing a cost prediction model using machine learning algorithms to predict the cost, obtaining the predicted construction value, and obtaining the predicted comprehensive mechanical efficiency index through a linear regression function; using the minimum error between the actual construction value and the predicted construction value as the primary objective, and the minimum error between the comprehensive mechanical efficiency index and the predicted comprehensive mechanical efficiency index as the auxiliary objective, and introducing a model regularization term to construct a joint loss function.
[0013] The big data-based power distribution network mechanized construction cost control system includes: a data fusion module, used to perform consistency processing and confidence assessment on acquired multi-source construction data, and to fuse the data to generate fused data based on the assessment results; a feature construction module, used to construct deep interactive features representing the construction status based on the fused data, and to calculate the comprehensive mechanical efficiency index; a cost prediction module, used to construct a cost prediction model based on the deep interactive features, and to construct a joint loss function based on the output of the cost prediction model and the comprehensive mechanical efficiency index; and to train and optimize the cost prediction model using the joint loss function to output the final predicted cost value.
[0014] The technical effects and advantages of this invention, which is a method and system for controlling the cost of mechanized construction of power distribution network projects based on big data, are as follows: This invention incorporates the actual working state of machinery and environmental impact into the cost prediction model by constructing deep interaction features and a comprehensive mechanical efficiency index. This makes the predicted values closer to the actual engineering situation. The cost prediction model is constructed based on the deep interaction features, and a joint loss function is constructed based on the output of the cost prediction model and the comprehensive mechanical efficiency index. This makes the prediction logic of the model more consistent with the common sense of engineering that "low mechanical efficiency will inevitably lead to increased costs." The prediction results are no longer just mathematical optimal fits, but reliable outputs that conform to the physical laws of the field. This effectively solves the problems of large gaps between predicted and actual cost values and the disconnect between mechanical efficiency and cost. Attached Figure Description
[0015] Figure 1 This is a schematic diagram of the process for controlling the cost of mechanized construction of power distribution network projects based on big data, provided in an embodiment of the present invention.
[0016] Figure 2 This is a schematic diagram of the structure of a big data-based mechanized construction cost control system for power distribution network engineering, provided in an embodiment of the present invention.
[0017] Figure 3 A schematic diagram illustrating the calculation method of the comprehensive mechanical efficiency index provided in this embodiment of the invention. Detailed Implementation
[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0019] Example 1, Figure 1 This invention presents a method for controlling the cost of mechanized construction in power distribution network projects based on big data, comprising the following steps: S1, perform consistency processing and confidence assessment on the acquired multi-source construction data, and fuse them to generate fused data based on the assessment results; S2, construct deep interactive features representing the construction status based on the fused data, and calculate the comprehensive mechanical efficiency index; S3, a cost prediction model is constructed based on deep interaction features, and a joint loss function is constructed based on the output results of the cost prediction model and the comprehensive mechanical efficiency index; S4 uses the joint loss function to train and optimize the cost prediction model, and outputs the final predicted cost.
[0020] This embodiment incorporates the actual working state of machinery and environmental impact into the cost prediction model by constructing deep interaction features and a comprehensive mechanical efficiency index. This makes the predicted values closer to the actual engineering situation. The cost prediction model is constructed based on the deep interaction features, and a joint loss function is constructed based on the output of the cost prediction model and the comprehensive mechanical efficiency index. This makes the prediction logic of the model more consistent with the common sense of engineering that "low mechanical efficiency will inevitably lead to increased costs." The prediction results are no longer just mathematical optimal fits, but reliable outputs that conform to the physical laws of the field. This effectively solves the problems of large gaps between predicted and actual cost values and the disconnect between mechanical efficiency and cost.
[0021] S1 performs consistency processing and confidence assessment on the acquired multi-source construction data, and merges the data based on the assessment results to generate fused data.
[0022] In this embodiment, the consistency processing of the acquired multi-source construction data specifically includes: Align multi-source construction data along the time dimension to generate time-aligned data; Map time-aligned data to a unified spatial coordinate system; Based on the spherical distance function, the GPS coordinates of construction machinery are matched with the coordinates of environmental elements in the geographic information system to obtain a consistent data source.
[0023] It should be noted that GIS stands for Geographic Information System. In this embodiment, its function is to locate all geographical features, such as road nodes and construction area boundary points, in the GIS database based on the GPS positioning of the construction machinery.
[0024] In this embodiment, the acquired multi-source construction data includes internal management system data sources, IoT sensor data sources, and the required environmental and geographic data sources and market data sources; In the time dimension, various types of data are resampled according to a unified time benchmark, and missing time segments are filled in using a piecewise linear interpolation method to generate time-aligned data. The piecewise linear interpolation method uses the timestamps and values of adjacent sampling points as references and calculates an approximate value of the target time through linear proportion. Based on the time-aligned data, the time-aligned data is mapped to the same spatial coordinate system in the spatial dimension. The spherical distance formula is used to calculate the spherical distance between the actual GPS coordinates of the machinery and the candidate coordinates of the GIS environmental elements. Through nearest neighbor matching, the corresponding geographic element attributes are assigned to each piece of mechanized construction data to obtain a consistent data source.
[0025] The specific formula for piecewise linear interpolation is as follows:
[0026] In the formula, For the interpolation result, For the target timestamp, , The timestamps of the two most recent observation points before and after the target time. , These are the observations corresponding to the timestamp.
[0027] It should be noted that observed values refer to data variables that are collected at the construction site through sensor monitoring, manual recording, or engineering settlement systems, and that can truly reflect objective conditions such as construction costs, machinery usage efficiency, and environmental conditions. Their value range is a finite set of real numbers.
[0028] In this embodiment, the expression for nearest neighbor matching is:
[0029] In the formula, The environmental attributes matched to the nearest neighbor of the construction machinery. The attribute value of the selected environmental element. It is a spherical distance function. For machinery Latitude and longitude coordinates of the moment The latitude and longitude coordinates of the candidate environmental elements. For all candidate points Find the point closest to the machine's location.
[0030] In this embodiment, the confidence level of the consistent data source is evaluated, and the specific steps are as follows: Based on a consistent data source, calculate the field missing rate of the data source; Calculate the confidence level of each data source based on the field missing rate of the data source.
[0031] In this embodiment, after completing the data consistency processing, confidence assessment is performed on various data sources.
[0032] The specific formula for calculating the field missing rate is as follows:
[0033] In the formula, For the field missing rate, This represents the number of missing time periods. This represents the total number of samples in the data source field.
[0034] The specific formula for calculating the confidence level of a data source is as follows:
[0035] In the formula, Confidence level of the data source.
[0036] In this embodiment, based on the assessment results of the data source confidence, multi-source data is fused to generate fused data, specifically as follows: Based on the confidence level of the data source, the weight of the data source is obtained through normalization calculation; Based on the weights of the data sources, multi-source data is weighted and merged to obtain fused data.
[0037] In this embodiment, the formula for calculating the normalized confidence level of the data source is as follows:
[0038] In the formula, For the first Normalized weights of each data source, For the first Confidence level of each data source This represents the total number of data sources.
[0039] The specific formula for weighted fusion of multi-source data is as follows:
[0040] In the formula, To integrate data, For the first One data source.
[0041] S2, constructs deep interactive features representing the construction status based on fused data, and calculates the comprehensive mechanical efficiency index.
[0042] In this embodiment, the deep interaction features include market features and working environment interaction features. The working environment interaction features include traction load index, turning angle and mechanical coordination index, hoisting precision positioning time ratio and terrain accessibility score. The overall mechanical efficiency index is calculated using the environmental reduction index, the power load rate of the construction machinery, and the time utilization rate of the construction machinery.
[0043] In this embodiment, based on fused data, the basic characteristics and market characteristics of mechanized construction are extracted. The basic characteristics of mechanized construction include mechanical combination characteristics and environmental characteristics, while the market characteristics include the unit shift price of machinery and the actual value of completed projects. Based on the fundamental characteristics of mechanized construction, we construct interactive characteristics of the working environment and calculate the comprehensive efficiency index of machinery.
[0044] The specific features for constructing the working environment interaction characteristics are as follows: The specific formula for calculating the traction load index is as follows:
[0045] In the formula, The traction load index, This is the actual oil pressure. Rated oil pressure, To lay the slope of the path, This is an empirical value for the coefficient of friction.
[0046] The specific formula for calculating the turning angle and mechanical coordination index is as follows:
[0047] In the formula, This refers to the turning angle and mechanical coordination index. This refers to the actual number of conveyors deployed. The theoretical number of conveyors required. For experience weight, The turning angle of the path.
[0048] The specific formula for calculating the time required for precise positioning during hoisting is as follows:
[0049] In the formula, The time required for precise positioning during hoisting is reduced. To fine-tune the status duration of the aerial work platform vehicle Total working hours.
[0050] The specific formula for calculating terrain accessibility score is as follows:
[0051] In the formula, Rate the terrain accessibility. This is the standardized value for slope. This is the coefficient for the type of surface attachments. The shortest road distance. , , These are weighted parameters.
[0052] In this embodiment, the expression for the overall mechanical efficiency index is:
[0053] In the formula, This is a mechanical overall efficiency index. To improve the time utilization rate of construction machinery, The power load rate of the construction machinery. This is the environmental deduction index.
[0054] in,
[0055]
[0056] In the formula, For the effective working time of the machine, Total operation time The rated power of the machine. This represents the actual average output power of the construction machinery.
[0057] Figure 3 A schematic diagram illustrating the calculation method of the comprehensive mechanical efficiency index provided in this embodiment of the invention.
[0058] In this embodiment, the method for obtaining the environmental reduction index is as follows: Weather indicators, slope indicators, and surface indicators were calculated based on the fused data; The environmental risk index is obtained by weighting weather indicators, slope indicators, and surface indicators according to preset weights. The environmental risk index is converted into a baseline reduction index through linear mapping, and the baseline reduction index is corrected based on correction parameters to obtain the environmental reduction index. The correction parameters are calculated based on historical comprehensive efficiency index data.
[0059] In this embodiment, the rainfall and wind speed observed during construction by meteorological stations or IoT sensors are obtained from the fused data, the slope value is obtained from the construction site mapping data, and the construction surface type is determined through on-site investigation. The weather index is obtained by weighting the rainfall and wind speed during the construction period with the ratio of the empirically set reference values for rainfall and wind speed. The slope index is calculated based on the slope values obtained from the on-site survey data and using the maximum and minimum slopes within the project area as formula parameters. Based on the type of construction surface, surface indicators are obtained by assigning values according to engineering experience. The calculated weather indicators, slope indicators, and surface indicators are weighted and synthesized according to preset weights to obtain the environmental risk index. By using a linear mapping method and setting the reduction lower bound parameter and mapping intensity parameter, the environmental risk index is converted into a baseline reduction index; The environmental reduction index is obtained by correcting the baseline reduction index based on the correction parameters.
[0060] The specific formula for calculating weather indicators is as follows:
[0061] In the formula, As a weather indicator, The actual rainfall during the construction period. The measured average wind speed during construction. , These are the baseline values for rainfall and wind speed, respectively, set based on experience. , The weights are rainfall and wind speed, respectively.
[0062] The specific formula for calculating the slope index is as follows:
[0063] Among them, the cropping operation The expression is as follows:
[0064] In the formula, For slope index, The slope value is obtained from the survey data at the construction site. , These represent the maximum and minimum slopes within the project area, respectively.
[0065] The specific formula for calculating surface indicators is as follows:
[0066] Based on engineering experience, different surface types are assigned corresponding resistance coefficients, specifically:
[0067] In the formula, As a surface indicator, It is a land surface type.
[0068] The formula for calculating the environmental risk index is as follows:
[0069] In the formula, This is an environmental risk index. For the first One indicator, For the preset first Each indicator weight, ,and .
[0070] The specific formula for calculating the baseline reduction index is as follows:
[0071] In the formula, Baseline reduction index, To reduce the lower bound parameter, Mapping intensity In this embodiment, the expression for the environmental reduction index is:
[0072] In the formula, For the environmental depreciation index, Baseline reduction index, To correct the parameters, This is to reduce the lower bound parameter.
[0073] In this embodiment, the expression for the correction parameter is:
[0074] In the formula, This is the overall mechanical efficiency index obtained from historical calculations. This is a historical forecast of the overall mechanical efficiency index. This means taking the median of multiple data points when there are multiple data points.
[0075] It should be noted that if historical data is unavailable or inaccurate, it can be... .
[0076] S3 constructs a cost prediction model based on deep interaction features, and constructs a joint loss function based on the output of the cost prediction model and the mechanical comprehensive efficiency index.
[0077] S4 uses the joint loss function to train and optimize the cost prediction model, and outputs the final predicted cost.
[0078] In this embodiment, the construction of a cost prediction model based on deep interaction features, and the construction of a joint loss function based on the output of the cost prediction model and the comprehensive mechanical efficiency index, specifically involves: By combining historical deep interaction features with historical mechanical comprehensive efficiency indicators, a historical feature dataset is generated. Based on historical feature datasets, a cost prediction model is constructed using machine learning algorithms to predict costs, obtain predicted construction value, and obtain the predicted comprehensive mechanical efficiency index through a linear regression function. The primary objective is to minimize the error between the actual value created and the predicted value created, while the secondary objective is to minimize the error between the comprehensive mechanical efficiency index and the predicted comprehensive mechanical efficiency index. A model regularization term is introduced to construct a joint loss function.
[0079] In this embodiment, the actual value created is extracted from the market features contained in the historical deep interaction features.
[0080] In this embodiment, after training the cost prediction model based on the joint loss function, the feature dataset of the new project is input into the trained cost prediction model to obtain the final predicted cost value.
[0081] In this embodiment, the cost prediction model formula is as follows:
[0082]
[0083] In the formula, To create value through prediction The unit price per shift of machinery in the market characteristics. The number of machine shifts after taking into account the influence of the interaction characteristics of the working environment. This is an indicator of mechanical efficiency. The number of machine shifts required to complete the task under ideal working conditions. The normalized gravitational load index, The normalized turning angle and mechanical coordination index. The normalized ratio of hoisting precision positioning time. The normalized terrain accessibility score is... , , , For weights.
[0084] In this embodiment, the predicted comprehensive mechanical efficiency index is obtained through a linear regression function as follows:
[0085] In the formula, To predict the overall efficiency index of machinery, This is a historical indicator of overall mechanical efficiency. For bias terms, These are the weights of the regression function.
[0086] The formula for the joint loss function is as follows: In the formula, The total error metric that needs to be minimized during model optimization. To create real value To create value through prediction The mean squared error loss function is the primary objective. This refers to the overall mechanical efficiency index calculated in practice. For the predicted overall mechanical efficiency index The mean squared error loss function is used to assist the target. For regularization terms, Weight parameters for auxiliary targets.
[0087] Example 2, Figure 2 This invention presents a big data-based mechanized construction cost control system for power distribution network engineering, comprising: The data fusion module is used to perform consistency processing and confidence assessment on the acquired multi-source construction data, and to fuse them to generate fused data based on the assessment results; The feature construction module is used to construct deep interactive features representing the construction status based on the fused data and to calculate the comprehensive mechanical efficiency index. The cost prediction module is used to build a cost prediction model based on deep interactive features, and to construct a joint loss function based on the output of the cost prediction model and the mechanical comprehensive efficiency index; the joint loss function is used to train and optimize the cost prediction model, and output the final predicted cost value.
[0088] Example 3, Figure 3 A schematic diagram illustrating the calculation method of the comprehensive mechanical efficiency index provided in this embodiment of the invention.
[0089] In this embodiment, the overall mechanical efficiency index is calculated using the environmental reduction index, the power load rate of the construction machinery, and the time utilization rate of the construction machinery. The specific formula is as follows:
[0090] In the formula, This is a mechanical overall efficiency index. To improve the time utilization rate of construction machinery, The power load rate of the construction machinery. This is the environmental deduction index.
[0091] In this embodiment, the expression for the environmental reduction index is:
[0092] In the formula, For the environmental depreciation index, Baseline reduction index, To correct the parameters, This is to reduce the lower bound parameter.
[0093] In this embodiment, the expression for the correction parameter is:
[0094] In the formula, This is the overall mechanical efficiency index obtained from historical calculations. This is a historical forecast of the overall mechanical efficiency index. This means taking the median of multiple data points when there are multiple data points.
[0095] In this embodiment, the method for obtaining the environmental reduction index is as follows: Weather indicators, slope indicators, and surface indicators were calculated based on the fused data; The environmental risk index is obtained by weighting weather indicators, slope indicators, and surface indicators according to preset weights. The environmental risk index is converted into a baseline reduction index through linear mapping, and the baseline reduction index is corrected based on correction parameters to obtain the environmental reduction index. The correction parameters are calculated based on historical comprehensive efficiency index data.
[0096] This embodiment uses excavator construction in a power distribution line project as an example to demonstrate the entire process of calculating the comprehensive efficiency index of machinery. The machine's operating environment on that day involved a slight slope and farmland soil.
[0097] It should be noted that the data used in this embodiment are all multi-source data actually collected at the engineering site. After consistency processing and fusion as described in step S1, reliable fused data is obtained.
[0098] The data required for calculation in this embodiment is as follows: Time utilization The total operation time is obtained from the status logs of the mechanical vehicle-mounted sensors. The effective operating time of the machine is 8 hours. It takes 6.5 hours; In this embodiment, the time utilization rate is calculated according to the time utilization rate calculation formula. Specifically:
[0099] Power load rate The power sensor, mounted on the mechanical power output shaft, samples at a frequency of once per minute, recording the rated power for the day. The average output power is calculated as 100kW using a weighted average of data collected throughout the day. It is 65kW; In this embodiment, the power load factor is calculated according to the power load factor calculation formula. Specifically:
[0100] Weather indicators By accessing the local weather station API, hourly data during the construction period was obtained, confirming that the rainfall was 0mm and the average wind speed was 1.2m / s. Based on the thresholds (rainfall baseline 10mm, wind speed baseline 12m / s) and weights (rainfall weight 0.6, wind speed weight 0.4) set by the project, the weather index was calculated to be 0.04 according to the weather index calculation formula.
[0101] Slope index The data originates from high-precision DEM (Digital Elevation Model) data obtained during the project survey phase. The average slope of the machinery was calculated to be 7.2° by matching the machinery's daily work trajectory with the GIS system. Based on the overall slope range of the project area (0° to 21°), the slope index was calculated to be 0.343 according to the slope index calculation formula.
[0102] Surface Indicators Match the location of the machinery operation and confirm the land surface type as "farmland". Based on the preset engineering experience value according to the present invention, the farmland surface index is obtained as 0.6.
[0103] In this embodiment, the preset weights for weather indicators are 0.4, slope indicators are 0.3, and surface indicators are 0.3. Based on the environmental risk index calculation formula, the risk environmental index is calculated. Specifically:
[0104] In this embodiment, a lower bound parameter for the preset reduction is used. The baseline reduction index is 0.1. Based on the baseline reduction index calculation formula, the baseline reduction index is calculated. Specifically:
[0105] This embodiment lacks accurate historical data, therefore... The environmental reduction index Specifically:
[0106] In this embodiment, the mechanical comprehensive efficiency index is calculated according to the formula for calculating the mechanical comprehensive efficiency index as follows:
[0107] In this embodiment, the excavator's overall mechanical efficiency index under the current environment is calculated to be 0.497. This result quantifies the significant negative impact of the complex environment on construction efficiency. This overall mechanical efficiency index will be input into the cost prediction model to calculate the additional shift costs caused by low efficiency, thereby achieving accurate cost prediction.
[0108] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.
[0109] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.
[0110] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0111] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.
[0112] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
[0113] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for controlling the cost of mechanized construction in power distribution network projects based on big data, characterized in that, Includes the following steps: The acquired multi-source construction data is subjected to consistency processing and confidence assessment, and then fused to generate fused data based on the assessment results; Based on the fused data, a deep interactive feature characterizing the construction status is constructed, and the comprehensive mechanical efficiency index is calculated; A cost prediction model is constructed based on deep interaction features, and a joint loss function is constructed based on the output of the cost prediction model and the comprehensive mechanical efficiency index. The joint loss function is used to train and optimize the cost prediction model, and the final predicted cost is output.
2. The method for controlling the cost of mechanized construction of power distribution network projects based on big data as described in claim 1, characterized in that, The consistency processing of the acquired multi-source construction data specifically includes: Align multi-source construction data along the time dimension to generate time-aligned data; Map time-aligned data to a unified spatial coordinate system; Based on the spherical distance function, the GPS coordinates of construction machinery are matched with the coordinates of environmental elements in the geographic information system to obtain a consistent data source.
3. The method for controlling the cost of mechanized construction of power distribution network projects based on big data as described in claim 2, characterized in that, The expression for nearest neighbor matching is: In the formula, The environmental attributes matched to the nearest neighbor of the construction machinery. The attribute value of the selected environmental element. It is a spherical distance function. For machinery Latitude and longitude coordinates of the moment The latitude and longitude coordinates of the candidate environmental elements. For all candidate points Find the point closest to the machine's location.
4. The method for controlling the cost of mechanized construction of power distribution network projects based on big data as described in claim 3, characterized in that, The deep interaction features include market features and working environment interaction features. The working environment interaction features include traction load index, turning angle and mechanical coordination index, hoisting precision positioning time ratio and terrain accessibility score. The overall mechanical efficiency index is calculated using the environmental reduction index, the power load rate of the construction machinery, and the time utilization rate of the construction machinery.
5. The method for controlling the cost of mechanized construction of power distribution network projects based on big data as described in claim 4, characterized in that, The method for obtaining the environmental reduction index is as follows: Weather indicators, slope indicators, and surface indicators were calculated based on the fused data; The environmental risk index is obtained by weighting weather indicators, slope indicators, and surface indicators according to preset weights. The environmental risk index is converted into a baseline reduction index through linear mapping, and the baseline reduction index is corrected based on correction parameters to obtain the environmental reduction index. The correction parameters are calculated based on historical comprehensive efficiency index data.
6. The method for controlling the cost of mechanized construction of power distribution network projects based on big data as described in claim 5, characterized in that, The construction of the cost prediction model based on deep interaction features, and the construction of a joint loss function based on the output of the cost prediction model and the comprehensive mechanical efficiency index, are as follows: By combining historical deep interaction features with historical mechanical comprehensive efficiency indicators, a historical feature dataset is generated. Based on historical feature datasets, a cost prediction model is constructed using machine learning algorithms to predict costs, obtain predicted construction value, and obtain the predicted comprehensive mechanical efficiency index through a linear regression function. The primary objective is to minimize the error between the actual value created and the predicted value created, while the secondary objective is to minimize the error between the comprehensive mechanical efficiency index and the predicted comprehensive mechanical efficiency index. A model regularization term is introduced to construct a joint loss function.
7. The method for controlling the cost of mechanized construction of power distribution network projects based on big data as described in claim 6, characterized in that, The expression for the overall mechanical efficiency index is: In the formula, This is a mechanical overall efficiency index. To improve the time utilization rate of construction machinery, The power load rate of the construction machinery. This is the environmental deduction index.
8. The method for controlling the cost of mechanized construction of power distribution network projects based on big data as described in claim 7, characterized in that, The expression for the environmental reduction index is: In the formula, For the environmental depreciation index, Baseline reduction index, To correct the parameters, This is to reduce the lower bound parameter.
9. The method for controlling the cost of mechanized construction of power distribution network projects based on big data as described in claim 8, wherein the expression for the correction parameter is: In the formula, This is the overall mechanical efficiency index obtained from historical calculations. This is a historical forecast of the overall mechanical efficiency index. This means taking the median of multiple data points when there are multiple data points.
10. A system using the big data-based mechanized construction cost control method for power distribution network engineering as described in any one of claims 1-9, comprising: The data fusion module is used to perform consistency processing and confidence assessment on the acquired multi-source construction data, and to fuse them to generate fused data based on the assessment results; The feature construction module is used to construct deep interactive features representing the construction status based on the fused data and to calculate the comprehensive mechanical efficiency index. The cost prediction module is used to build a cost prediction model based on deep interactive features, and to construct a joint loss function based on the output of the cost prediction model and the mechanical comprehensive efficiency index; the joint loss function is used to train and optimize the cost prediction model, and output the final predicted cost value.