A resource scheduling system for post-casting belt engineering projects
Through real-time data acquisition and deep learning models, combined with regression analysis and demand prediction models, the inventory update frequency of steel bar materials is dynamically adjusted, which solves the problem of insufficient inventory update frequency in traditional resource scheduling, ensures that materials are in time, avoids construction stagnation, and optimizes supply Chain management to reduce operating costs.
Patent Information
- Application Number
- CN202411237316.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-05
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-09-05
AI Technical Summary
In the traditional post-pouring belt construction resource scheduling, the inventory update frequency of steel bar materials is insufficient, and it is impossible to flexibly match the changes in construction progress, resulting in material shortages and construction stagnation, affecting the progress of the project.
Using real-time data acquisition and deep learning models combined with regression analysis and demand prediction models, dynamically adjust the inventory update frequency of steel bar materials, adjust the inventory update frequency through fuzzy logic, ensure that the materials are in place in time during peak demand periods, and flexibly adjust the supply time according to construction needs. and batch.
Effectively avoid construction stagnation, reduce inventory backlog and operational costs, improve resource utilization efficiency, optimize supply chain management, and ensure that project progress is carried out as planned.
Smart Images

Figure CN118761603B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of resource scheduling for a post-cast strip engineering project, and in particular to a resource scheduling system for a post-cast strip engineering project. Background Art
[0002] The post-casting project resource scheduling system is a digital tool specifically used to manage and optimize post-casting construction resources in construction projects. Post-casting is a construction joint set up in construction to prevent cracks in concrete structures due to uneven settlement or temperature stress, and pouring is carried out after a certain period of time. Due to the particularity of post-casting construction, the management of its construction progress, resource allocation, and process connection is highly complex and dynamic. The resource scheduling system ensures the efficient use of various resources and reduces waste and delays in construction through real-time monitoring and scheduling of resources such as manpower, materials, and equipment, combined with the actual situation and progress plan of the construction site. At the same time, the system can handle coordination issues between multiple processes, improve overall construction efficiency, and ensure project quality. Through data analysis and dynamic adjustment, the system can also predict and avoid possible risks, provide decision-making support for construction management, and ultimately ensure that the post-casting project is completed on time and with quality.
[0003] The prior art has the following deficiencies:
[0004] In the traditional resource scheduling process of post-casting construction, the inventory of steel bars is usually updated regularly. However, due to the complex and changeable progress of post-casting construction, the construction demand may not be able to fully match the regularly updated inventory frequency, resulting in inflexible resource scheduling. If the inventory update frequency of steel bars is low, and the actual construction progress and demand change rapidly, it may lead to insufficient materials on site. This shortage will directly lead to construction stagnation and affect the progress of the entire project.
[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention
[0006] The purpose of the present invention is to provide a resource scheduling system for post-casting strip engineering projects, which utilizes real-time data collection and deep learning models, combined with regression analysis and demand forecasting models, to dynamically adjust the inventory update frequency of steel bar materials, to ensure that materials are in place in time during peak demand periods, to avoid construction stagnation, and at the same time, through intelligent inventory management strategies, to extend the update cycle when demand is stable, to reduce inventory backlogs and related costs, and to flexibly adjust delivery time and batches according to construction needs, to optimize supply chain management, to reduce risks, to improve resource utilization efficiency, and to reduce operating costs overall, so as to solve the problems in the above-mentioned background technology.
[0007] In order to achieve the above-mentioned object, the present invention provides the following technical solutions: a post-casting strip engineering project resource scheduling system, including a real-time data acquisition module, a construction complexity assessment module, a demand forecasting model module, an inventory update dynamic adjustment module, and a supply time and batch optimization module;
[0008] Real-time data acquisition module, establish an efficient data acquisition system to obtain the progress information of post-casting belt construction in real time;
[0009] The construction complexity assessment module analyzes the collected construction progress data, extracts features that reflect the construction complexity, inputs the extracted features into a pre-trained deep learning model, outputs the predicted results through the deep learning model, and assesses the complexity of the construction process;
[0010] The demand forecasting model module uses regression analysis to establish a demand forecasting model based on the results of the progress complexity analysis, and predicts the demand for steel bars in the future based on historical data and current progress;
[0011] The inventory update dynamic adjustment module obtains the future demand fluctuation information and demand surge information of steel bars according to the output of the demand forecasting model, and dynamically adjusts the inventory update frequency of steel bars based on the demand fluctuation information and demand surge information using fuzzy logic;
[0012] The delivery time and batch optimization module dynamically adjusts the inventory update frequency and flexibly adjusts the delivery time and batch of steel bar materials according to changes in construction site demand.
[0013] Preferably, the trained deep learning model is used to analyze the construction progress data and evaluate the construction complexity. The specific steps are as follows:
[0014] Pre-process the collected construction progress data;
[0015] After the preprocessing is completed, the key features related to the construction complexity are selected;
[0016] After selecting features, feature engineering is performed to further enhance the expressiveness of the features, where feature engineering includes combining different features, generating new features, and performing multi-order interactions of features;
[0017] Convert the processed feature data into the format for input into the deep learning model, and organize the data into multi-dimensional tensor or sequence data format according to the model's requirements;
[0018] The sorted data is input into a pre-trained deep learning model, and the model will output an evaluation result based on the input features. The evaluation result is a construction complexity score, which indicates the complexity of the construction process in the future.
[0019] Preferably, the demand for steel bars in the future is predicted, and the specific steps are as follows:
[0020] Collect historical data related to reinforcement material requirements;
[0021] After data collation, key features related to reinforcement material requirements were selected;
[0022] The features extracted from the construction progress data are input into the pre-trained complexity assessment model to assess the complexity of the current construction stage;
[0023] Based on the selected features, the multivariate linear regression model is selected to predict the demand for steel reinforcement materials;
[0024] After the model training is completed, the model is calibrated and verified, and the model output is adjusted to make it more in line with the actual demand level;
[0025] After calibration and verification, the model is deployed in the actual application environment to perform real-time steel material demand forecasting.
[0026] Preferably, according to the output of the demand forecasting model, the demand fluctuation index is calculated by the future demand for steel bars, and the specific steps are as follows:
[0027] The time series modeling of steel material demand is performed, and the demand time series is set as , where t represents time, It represents the actual demand for steel bar materials at time t. The difference method is used to eliminate non-stationarity. A nonlinear autoregressive model is constructed to capture the time dynamic characteristics of demand. The constructed expression is: , where represents the demand forecast value of steel bar materials at time t, is a nonlinear function; is a random noise term; n is the time lag;
[0028] Based on time series data, the acceleration of steel material demand is calculated to measure the rate and trend of demand change. The calculation expression is: , where It represents the acceleration of steel material demand at time t, and measures the trend of steel material demand change. Here, t represents the time interval, The positive or negative value indicates the acceleration or deceleration trend of the demand for steel bar materials;
[0029] Introducing the demand elasticity coefficient , which represents the sensitivity of steel material demand to external influences, and the demand elasticity coefficient is calculated by the following formula: , where represents the demand elasticity coefficient at time t, Indicates the sensitivity of reinforcement material demand to changes in external factors;
[0030] The demand fluctuation index is calculated by combining the demand acceleration and demand elasticity coefficient. The calculation expression of the demand fluctuation index is: , where Represents the demand volatility index, which is used to quantify and evaluate the volatility of steel bar demand at time t. It is the cumulative demand in the past T time, which is used to balance the relationship between current demand and historical demand.
[0031] Preferably, according to the output of the demand forecasting model, the demand surge index is calculated by the future demand for steel bars, and the specific steps are as follows:
[0032] Define the future demand forecast value obtained from the demand forecast model, and assume represents the forecast value of steel material demand at time t, , , , , , , respectively represent the demand forecast values at the next m moments;
[0033] Calculate the demand growth rate in the future time period. The calculation expression of demand growth rate is: ,in, is the demand forecast value at time t+1, represents the growth rate of demand at time t;
[0034] In order to reflect the impact of demand growth at different time points on future demand, a time weight coefficient is introduced to weight demand growth rate. Defined as: ,in, is the weight coefficient at time t+i, which can usually be defined as: , so that the weight decreases over time;
[0035] The demand growth rate, demand volatility and weighted demand growth rate are combined to calculate the demand surge index. The calculation expression is: , where represents the demand surge index at time t, represents the demand volatility, , , and Represents the adjustment coefficient.
[0036] Preferably, fuzzy logic is used to dynamically adjust the inventory update frequency of steel bar materials. The specific steps are as follows:
[0037] Fuzzify the demand fluctuation index and demand sudden increase index. Fuzzification is to transform specific values into fuzzy sets to reflect uncertainty and ambiguity.
[0038] Based on the fuzzy set of demand fluctuation index and demand surge index, fuzzy rules are set. The fuzzy rules stipulate how the inventory update frequency should be adjusted under different demand fluctuation and surge conditions.
[0039] Through the fuzzy reasoning mechanism, the input fuzzy set is reasoned by applying the set fuzzy rules. The result of fuzzy reasoning is a fuzzy set, which represents the recommended adjustment degree of inventory update frequency.
[0040] The fuzzy output obtained by fuzzy reasoning is defuzzified and the fuzzy set is converted into specific inventory update frequency adjustment amount.
[0041] Preferably, the inventory update frequency of steel bar materials is dynamically adjusted according to the results of the defuzzification processing. If the defuzzification results show that the update cycle needs to be shortened, the corresponding inventory replenishment plan is immediately executed to ensure that the materials can be in place in time; if the results show that the update cycle needs to be extended, the next replenishment is delayed to reduce the frequency and cost of inventory management.
[0042] In the above technical solution, the technical effects and advantages provided by the present invention are:
[0043] Through the application of real-time data collection and deep learning models, the present invention can accurately evaluate the complexity of the construction progress, and combine regression analysis and demand forecasting models to effectively predict the future demand for steel bars. According to the output of the demand forecasting model, the demand fluctuation index and the demand surge index are calculated through the future demand for steel bars. Based on the fuzzy logic adjustment mechanism of the demand fluctuation index and the demand surge index, the inventory update frequency can dynamically respond to the actual demand changes at the construction site, ensuring that the steel bars are in place in time during the peak demand period, avoiding construction stagnation due to material shortages, and thus ensuring that the project progress is carried out as planned.
[0044] The present invention uses intelligent inventory management strategies to extend the update cycle when demand is stable, avoid excessive inventory backlogs, reduce unnecessary inventory holding costs, procurement costs and logistics management costs, and flexibly adjust delivery times and batches according to real-time changes in construction needs to further optimize supply chain management, reduce the risk of supply chain disruptions, and overall improve resource utilization efficiency and reduce operating costs. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.
[0046] Figure 1 The present invention is a module schematic diagram of a post-casting strip engineering project resource scheduling system. DETAILED DESCRIPTION
[0047] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.
[0048] The present invention provides Figure 1 A post-casting belt project resource scheduling system shown includes a real-time data acquisition module, a construction complexity assessment module, a demand forecasting model module, an inventory update dynamic adjustment module, and a delivery time and batch optimization module;
[0049] Real-time data acquisition module, establish an efficient data acquisition system to obtain the progress information of post-casting belt construction in real time;
[0050] This can be done through on-site sensors, feedback systems from construction personnel, or project management software. Real-time data collection ensures that the construction management team can accurately grasp the actual progress of the current construction.
[0051] The construction complexity assessment module analyzes the collected construction progress data, extracts features that reflect the construction complexity, inputs the extracted features into a pre-trained deep learning model, outputs the predicted results through the deep learning model, and assesses the complexity of the construction process;
[0052] Use the trained deep learning model to analyze the construction progress data and evaluate the construction complexity. The specific steps are as follows:
[0053] Pre-process the collected construction progress data;
[0054] This includes cleaning the data (such as dealing with missing values and outliers), standardizing or normalizing the data, and formatting the data to match the model input requirements. For example, the data may need to be converted into a specific format, such as a tensor or matrix, to fit the input structure of a deep learning model. In addition, time series data needs to be processed to ensure the consistency of the time order, which is particularly important when analyzing construction progress.
[0055] After the preprocessing is completed, the key features related to the construction complexity are selected;
[0056] These features may include schedule delay time, resource utilization, dependencies between processes, weather conditions at the construction site, number of workers, etc. Through expert experience or data-based feature importance analysis, determine which features are most important for the model's prediction, and exclude redundant or irrelevant features to improve the model's prediction efficiency and accuracy.
[0057] After selecting features, feature engineering is performed to further enhance the expressiveness of features. Feature engineering includes combining different features, generating new features (such as differential features between processes), and performing multi-order interactions of features.
[0058] The expressiveness of features is enhanced by combining different features, generating new features (such as differential features between processes), and performing multi-order interactions of features. The purpose of this step is to enable the input features to better reflect the complexity of the construction process, thereby improving the prediction performance of the model.
[0059] Convert the processed feature data into the format for input into the deep learning model, and organize the data into multi-dimensional tensor or sequence data format according to the model's requirements;
[0060] For example, for time series data, you may need to organize the data into a three-dimensional tensor containing the batch size, time steps, and feature dimensions. This step ensures that the data can be seamlessly input into the deep learning model for prediction.
[0061] Input the organized data into a pre-trained deep learning model;
[0062] This step involves preparing batches of data and feeding them into the model, ensuring that the model can process the data one by one and output predictions. Typically, the input process can be done through automated pipeline processing to ensure efficient data flow and stable model predictions.
[0063] Once the data is input into the model, the model will output an evaluation result based on the input features. The evaluation result is a construction complexity score, which indicates the complexity of the construction process in the future period of time.
[0064] After obtaining these prediction results, they can be further converted into more intuitive indicators or levels, such as construction risk level, key processes that need attention, etc., to facilitate the project management team to understand and use.
[0065] The demand forecasting model module uses regression analysis to establish a demand forecasting model based on the results of the progress complexity analysis, and predicts the demand for steel bars in the future based on historical data and current progress;
[0066] To predict the demand for steel bars in the future, the specific steps are as follows:
[0067] Collect historical data related to reinforcement material requirements;
[0068] This includes historical construction progress data, steel material consumption records, project duration, construction process schedule, and external factors that affect the construction progress (such as weather conditions, labor availability, etc.). At the same time, the data is sorted and cleaned to remove noise, fill missing values, and handle outliers to ensure the integrity and accuracy of the data. The sorted data is divided into different time periods and construction stages to provide a basis for subsequent modeling.
[0069] After data collation, key features related to reinforcement material requirements were selected;
[0070] These features may include the current construction progress, the complexity of the construction, the completed construction tasks, the expected workload in the next few days, the number of workers, and the weather forecast. Through the feature selection process, redundant or less relevant variables are eliminated, and those features that have a significant impact on the demand for steel bars are retained. Statistical methods (such as correlation analysis) or machine learning algorithms (such as LASSO regression) can be used to assist in feature selection.
[0071] The features extracted from the construction progress data are input into the pre-trained complexity assessment model to assess the complexity of the current construction stage;
[0072] Complexity assessment can reflect the difficulty of construction, the potential risk of construction delays, and the degree of fluctuation in demand for steel materials. The complexity assessment result will be used as one of the input features of the regression model to help the model better capture the changing trend of demand.
[0073] Based on the selected features, the multivariate linear regression model is selected to predict the demand for steel reinforcement materials;
[0074] After model selection, use historical data to train the model and adjust the model's parameters so that it can accurately fit the historical data and capture the pattern of demand changes. Through cross-validation or leaving out a validation set, evaluate the model's generalization ability to ensure its prediction accuracy for unseen data.
[0075] After the model training is completed, the model is calibrated and verified, and the model output is adjusted to make it more in line with the actual demand level;
[0076] For example, if the forecast demand output by the model is usually too low or too high, this can be corrected by adjusting the forecast value or adding a calibration factor. The validation step includes applying the model to a test data set, evaluating the model's predictive performance, and calculating error metrics (such as mean square error, mean absolute error, etc.) to ensure that the model's predictive ability meets expectations.
[0077] After calibration and verification, the model is deployed in the actual application environment to perform real-time steel material demand forecasting;
[0078] The deployed model will receive data input from the construction site in real time and generate a forecast of steel demand in the future. The output of the model will be used to guide inventory management and material procurement decisions. In order to ensure the real-time and accuracy of the forecast, it is necessary to establish an automated data transmission and processing process to ensure that the construction progress data can be delivered to the model in a timely manner.
[0079] The inventory update dynamic adjustment module obtains the future demand fluctuation information and demand surge information of steel bars according to the output of the demand forecasting model, and dynamically adjusts the inventory update frequency of steel bars based on the demand fluctuation information and demand surge information using fuzzy logic;
[0080] When the model predicts that the future demand for steel bars will fluctuate greatly or have a sudden increase trend, it may mean that the construction will enter a critical stage and material consumption will increase significantly. At this time, it is crucial to shorten the inventory update cycle. By speeding up the update frequency, it can ensure that the steel bar materials arrive at the construction site in time and avoid construction stagnation or delays caused by material shortages. This approach can effectively cope with peak demand periods and ensure the continuous advancement of the project. Conversely, when the model prediction shows that future demand is relatively stable or demand fluctuations are small, it means that the construction progress is relatively stable and the demand for materials is relatively balanced. In this case, you can choose to maintain the current inventory update cycle, or appropriately extend the update cycle when the demand is particularly stable. This not only helps to reduce the cost of frequent procurement and inventory management, but also avoids excessive backlogs of materials, reduces capital occupation and storage pressure.
[0081] By dynamically adjusting the inventory update frequency, construction projects can better balance the timeliness of material supply and the optimization of inventory costs, and achieve efficient and economical resource management. This can ensure that construction needs are fully met and avoid waste caused by excessive reserves, thus improving the management efficiency and economic benefits of the entire project.
[0082] According to the output of the demand forecasting model, the demand fluctuation index is calculated through the future demand for steel bars. The specific steps are as follows:
[0083] The time series modeling of steel material demand is performed, and the demand time series is set as , where t represents time, It represents the actual demand for steel bar materials at time t. The difference method is used to eliminate non-stationarity. A nonlinear autoregressive model is constructed to capture the time dynamic characteristics of demand. The constructed expression is: , where represents the demand forecast value of steel bar materials at time t, is a nonlinear function that can be fitted by a neural network or support vector regression (SVR); is a random noise term; n is the time lag;
[0084] Based on time series data, the acceleration of steel material demand is calculated to measure the rate and trend of demand change. The calculation expression is: , where It represents the acceleration of steel material demand at time t, and measures the trend of steel material demand change. Here, t represents the time interval (usually taken as 1), The positive or negative value of indicates the acceleration or deceleration trend of the demand for steel bars. The change of acceleration can further capture the sudden change or smooth transition of demand.
[0085] Introducing the demand elasticity coefficient , which represents the sensitivity of steel material demand to external influences, and the demand elasticity coefficient is calculated by the following formula: , where It represents the demand elasticity coefficient at time t, which measures the sensitivity of steel material demand to external influences. It indicates the sensitivity of steel material demand to changes in external factors. The larger the demand elasticity coefficient, the more sensitive the demand is to external factors.
[0086] Represents external factors affecting the demand for steel bars. These factors may include economic indicators (such as GDP growth rate), material market prices, construction environment conditions (such as weather, labor availability), etc. The value of should be selected according to the actual value of these external factors at time t, and converted into a value that can reflect the impact of external conditions on demand through standardization or feature extraction. These data can be obtained through channels such as economic reports, market data and environmental monitoring, and used as key variables affecting demand changes in the model.
[0087] The demand fluctuation index is calculated by combining the demand acceleration and demand elasticity coefficient. The calculation expression of the demand fluctuation index is: , where Represents the demand volatility index, which is used to quantify and evaluate the volatility of steel bar demand at time t. It is the cumulative demand in the past T time, which is used to balance the relationship between current demand and historical demand. The higher the demand fluctuation index, the greater the demand fluctuation.
[0088] According to the output of the demand forecasting model, the demand surge index is calculated through the future demand for steel bars. The specific steps are as follows:
[0089] Define the future demand forecast value obtained from the demand forecast model, and assume represents the forecast value of steel material demand at time t, , , , , , , respectively represent the demand forecast values at the next m moments;
[0090] Calculate the demand growth rate in the future time period. The calculation expression of demand growth rate is: ,in, is the demand forecast value at time t+1, represents the growth rate of demand at time t;
[0091] In order to reflect the impact of demand growth at different time points on future demand, a time weight coefficient is introduced so that the demand growth rate closer to the current moment has a greater weight. Defined as: ,in, is the weight coefficient at time t+i, which can usually be defined as: , so that the weight decreases over time;
[0092] The demand growth rate, demand volatility and weighted demand growth rate are combined to calculate the demand surge index. The calculation expression is: , where represents the demand surge index at time t, represents the demand volatility, , , and Represents the scaling factor used to control the sensitivity of the demand surge index based on demand.
[0093] Fuzzy logic is used to dynamically adjust the inventory update frequency of steel bars. The specific steps are as follows:
[0094] Fuzzify the demand fluctuation index and demand sudden increase index. Fuzzification is to transform specific values into fuzzy sets to reflect uncertainty and ambiguity.
[0095] For the demand fluctuation index, several fuzzy sets can be defined, such as "low fluctuation", "medium fluctuation", and "high fluctuation". For the demand surge index, similar fuzzy sets can also be defined, such as "no surge", "small surge", and "large surge". Fuzzification allows these indices to more flexibly reflect the actual situation and provide input for fuzzy reasoning.
[0096] Based on the fuzzy set of demand fluctuation index and demand surge index, fuzzy rules are set. The fuzzy rules stipulate how the inventory update frequency should be adjusted under different demand fluctuation and surge conditions.
[0097] For example, you can set the following rules:
[0098] If "demand fluctuations are high fluctuations" and "demand surges are large surges", then "inventory update frequency should be significantly increased".
[0099] If "demand fluctuation is low fluctuation" and "demand surge is no surge", then "inventory update frequency should be maintained or reduced". The setting of fuzzy rules is based on actual experience and understanding of the system and is the core decision-making mechanism of the fuzzy logic system.
[0100] Through the fuzzy reasoning mechanism, the input fuzzy set is reasoned by applying the set fuzzy rules. The result of fuzzy reasoning is a fuzzy set, which represents the recommended adjustment degree of inventory update frequency.
[0101] The reasoning process takes into account demand fluctuations and sudden increases, and weights the results of different rules to obtain a comprehensive fuzzy output.
[0102] Defuzzify the fuzzy output obtained by fuzzy reasoning and convert the fuzzy set into a specific inventory update frequency adjustment amount;
[0103] Commonly used defuzzification methods include the centroid method, the maximum membership method, etc. The result of defuzzification is a clear value, indicating how the inventory update cycle should be adjusted (such as "shorten the update cycle by 2 days" or "extend the update cycle by 1 day").
[0104] According to the results of the defuzzification process, the inventory update frequency of steel bar materials is dynamically adjusted. If the defuzzification results show that the update cycle needs to be shortened, the corresponding inventory replenishment plan is immediately executed to ensure that the materials can be in place in time; if the results show that the update cycle needs to be extended, the next replenishment is delayed to reduce the frequency and cost of inventory management.
[0105] The delivery time and batch optimization module dynamically adjusts the inventory update frequency and flexibly adjusts the delivery time and batch of steel bars according to the changes in construction site demand;
[0106] After dynamically adjusting the inventory update frequency, it is necessary to further flexibly adjust the supply arrangement of steel bar materials according to the actual demand changes at the construction site. Specifically, if the construction progress suddenly accelerates or there is an unexpected increase in demand, the supply time of steel bar materials should be adjusted in time so that the materials can be delivered to the site earlier or more frequently. At the same time, the batch and quantity of each supply can be adjusted according to the size of the demand to ensure that the steel bar materials required for construction can be supplied on time and in sufficient quantity to avoid construction delays caused by insufficient materials. In this way, not only can the construction needs be met, but also the logistics and inventory management can be optimized to reduce costs.
[0107] Through the application of real-time data collection and deep learning models, the present invention can accurately evaluate the complexity of the construction progress, and combine regression analysis and demand forecasting models to effectively predict the future demand for steel bars. According to the output of the demand forecasting model, the demand fluctuation index and the demand surge index are calculated through the future demand for steel bars. Based on the fuzzy logic adjustment mechanism of the demand fluctuation index and the demand surge index, the inventory update frequency can dynamically respond to the actual demand changes at the construction site, ensuring that the steel bars are in place in time during the peak demand period, avoiding construction stagnation due to material shortages, and thus ensuring that the project progress is carried out as planned.
[0108] The present invention uses intelligent inventory management strategies to extend the update cycle when demand is stable, avoid excessive inventory backlogs, reduce unnecessary inventory holding costs, procurement costs and logistics management costs, and flexibly adjust delivery times and batches according to real-time changes in construction needs to further optimize supply chain management, reduce the risk of supply chain disruptions, and overall improve resource utilization efficiency and reduce operating costs.
[0109] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.
[0110] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
[0111] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.
Claims
1. A post-casting belt project resource scheduling system, characterized in that: It includes real-time data collection module, construction complexity assessment module, demand forecasting model module, inventory update dynamic adjustment module and delivery time and batch optimization module; Real-time data acquisition module, establish an efficient data acquisition system to obtain the progress information of post-casting belt construction in real time; The construction complexity assessment module analyzes the collected construction progress data, extracts features that reflect the construction complexity, inputs the extracted features into a pre-trained deep learning model, outputs the predicted results through the deep learning model, and assesses the complexity of the construction process; The demand forecasting model module uses regression analysis to establish a demand forecasting model based on the results of the progress complexity analysis, and predicts the demand for steel bars in the future based on historical data and current progress; The inventory update dynamic adjustment module obtains the future demand fluctuation information and demand surge information of steel bars according to the output of the demand forecasting model, and dynamically adjusts the inventory update frequency of steel bars based on the demand fluctuation information and demand surge information using fuzzy logic; According to the output of the demand forecasting model, the demand fluctuation index is calculated through the future demand for steel bars. The specific steps are as follows: The time series modeling of steel material demand is performed, and the demand time series is set as , where t represents time, It represents the actual demand for steel bar materials at time t. The difference method is used to eliminate non-stationarity. A nonlinear autoregressive model is constructed to capture the time dynamic characteristics of demand. The constructed expression is: , where represents the demand forecast value of steel bar materials at time t, is a nonlinear function; is a random noise term; n is the time lag; Based on time series data, the acceleration of steel material demand is calculated to measure the rate and trend of demand change. The calculation expression is: , where It represents the acceleration of steel material demand at time t, and measures the trend of steel material demand change. Here, t represents the time interval, The positive or negative value indicates the acceleration or deceleration trend of the demand for steel bar materials; Introducing the demand elasticity coefficient , which represents the sensitivity of steel material demand to external influences, and the demand elasticity coefficient is calculated by the following formula: , where represents the demand elasticity coefficient at time t, Indicates the sensitivity of reinforcement material demand to changes in external factors; The demand fluctuation index is calculated by combining the demand acceleration and demand elasticity coefficient. The calculation expression of the demand fluctuation index is: , where Represents the demand volatility index, which is used to quantify and evaluate the volatility of steel bar demand at time t. It is the cumulative demand in the past T time, which is used to balance the relationship between current demand and historical demand; According to the output of the demand forecasting model, the demand surge index is calculated through the future demand for steel bars. The specific steps are as follows: Define the future demand forecast value obtained from the demand forecast model, and assume represents the forecast value of steel material demand at time t, , , , , , They represent the demand forecast values at the next m moments respectively; Calculate the demand growth rate in the future time period. The calculation expression of demand growth rate is: ,in, is the demand forecast value at time t+1, represents the growth rate of demand at time t; In order to reflect the impact of demand growth at different time points on future demand, a time weight coefficient is introduced to weight demand growth rate. Defined as: ,in, is the weight coefficient of time t+i; The demand growth rate, demand volatility and weighted demand growth rate are combined to calculate the demand surge index. The calculation expression is: , where represents the demand surge index at time t, represents the demand volatility, , , and represents the adjustment coefficient; Fuzzy logic is used to dynamically adjust the inventory update frequency of steel bars. The specific steps are as follows: Fuzzify the demand fluctuation index and demand sudden increase index. Fuzzification is to transform specific values into fuzzy sets to reflect uncertainty and ambiguity. Based on the fuzzy set of demand fluctuation index and demand surge index, fuzzy rules are set. The fuzzy rules stipulate how the inventory update frequency should be adjusted under different demand fluctuation and surge conditions. Through the fuzzy reasoning mechanism, the input fuzzy set is reasoned by applying the set fuzzy rules. The result of fuzzy reasoning is a fuzzy set, which represents the recommended adjustment degree of inventory update frequency. Defuzzify the fuzzy output obtained by fuzzy reasoning and convert the fuzzy set into a specific inventory update frequency adjustment amount; The delivery time and batch optimization module dynamically adjusts the inventory update frequency and flexibly adjusts the delivery time and batch of steel bar materials according to changes in construction site demand.
2. A post-casting strip engineering project resource scheduling system according to claim 1, characterized in that: Use the trained deep learning model to analyze the construction progress data and evaluate the construction complexity. The specific steps are as follows: Pre-process the collected construction progress data; After the preprocessing is completed, the key features related to the construction complexity are selected; After selecting features, feature engineering is performed to further enhance the expressiveness of the features, where feature engineering includes combining different features, generating new features, and performing multi-order interactions of features; Convert the processed feature data into the format for input into the deep learning model, and organize the data into multi-dimensional tensor or sequence data format according to the model's requirements; The sorted data is input into a pre-trained deep learning model, and the model will output an evaluation result based on the input features. The evaluation result is a construction complexity score, which indicates the complexity of the construction process in the future.
3. A post-casting strip engineering project resource scheduling system according to claim 2, characterized in that: To predict the demand for steel bars in the future, the specific steps are as follows: Collect historical data related to reinforcement material requirements; After data collation, key features related to reinforcement material requirements were selected; The features extracted from the construction progress data are input into the pre-trained complexity assessment model to assess the complexity of the current construction stage; Based on the selected features, the multivariate linear regression model is selected to predict the demand for steel reinforcement materials; After the model training is completed, the model is calibrated and verified, and the model output is adjusted to make it more in line with the actual demand level; After calibration and verification, the model is deployed in the actual application environment to perform real-time steel material demand forecasting.
4. A post-casting strip engineering project resource scheduling system according to claim 1, characterized in that: According to the results of the defuzzification process, the inventory update frequency of steel bar materials is dynamically adjusted. If the defuzzification results show that the update cycle needs to be shortened, the corresponding inventory replenishment plan is immediately executed to ensure that the materials can be in place in time; if the results show that the update cycle needs to be extended, the next replenishment is delayed to reduce the frequency and cost of inventory management.
Citation Information
Patent Citations
Artificial aging method for detecting true mechanical properties of steel wire rods
CN104749045A
Electric power material demand prediction method and device, terminal and storage medium
CN118195224A
Inventory management device
US20220284382A1
