Method and system for material demand forecasting and delivery route planning for a foundry machine

Abstract

The application discloses a method and system for material demand prediction and distribution path planning for a factory building machine, and relates to the technical field of demand prediction and path planning.The method comprises the following steps: obtaining design parameters of a target factory building; obtaining coordinates of the factory building machine in real time and calculating construction progress; running an adaptive speed prediction model to output future advance speed prediction; obtaining a future construction progress curve based on the predicted speed integral, combining with dynamic calculation of replenishment demand nodes to form a dynamic demand node set by mapping the predicted demand time and the factory building machine position; taking the dynamic demand node set as input, establishing a multi-objective vehicle path planning model, and solving an optimal distribution scheme; and finally issuing the scheme and periodically rolling and re-planning.The application realizes dynamic and accurate cooperation of material distribution and factory building machine construction, effectively improving the construction efficiency and industrialization level of the factory building machine.

Description

Technical Field

[0001] This invention relates to the field of demand forecasting and routing technology, specifically to a method and system for material demand forecasting and distribution routing for manufacturing machines. Background Technology

[0002] As a new type of mobile formwork equipment, the formwork machine can move along the track and realize the integrated continuous operation of formwork support, concrete pouring, curing and other processes. It is particularly suitable for long and narrow cast-in-place concrete structures with repetitive planar units, such as wineries, and can significantly reduce the amount of formwork used, simplify the process and shorten the construction period.

[0003] However, in current construction practices, the delivery of materials such as concrete and steel bars relies heavily on fixed plans or manual experience-based scheduling, which is disconnected from the dynamic progress of the construction machinery. This results in delivery vehicles either arriving too early and occupying limited construction space, or arriving too late and forcing the construction machinery to stop and wait, severely limiting the advantages of efficient and continuous operation of the construction machinery. The reason lies in the lack of accurate prediction of the construction machinery's construction status and a method for co-optimizing the prediction results with the delivery route.

[0004] Therefore, there is an urgent need for a technology that can predict the material demand of manufacturing equipment in real time and intelligently plan the delivery route accordingly, so as to ensure the continuity of the construction process and reduce resource waste. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to address the shortcomings of the prior art by providing a method and system for material demand forecasting and distribution route planning for manufacturing machines.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] A method for material requirements forecasting and distribution route planning for manufacturing equipment includes the following steps:

[0008] Step S1: Obtain the design parameters of the target plant. The design parameters include the total number of standard units that are continuously repeated along the direction of the plant construction machine, the length of each standard unit, and the standard consumption of each type of construction material in each standard unit.

[0009] Step S2: Obtain the coordinates of the plant-making machine in the long axis direction of the plant in real time, and calculate the real-time construction progress of the plant-making machine.

[0010] Step S3: Construct and run an adaptive speed prediction model. The adaptive speed prediction model outputs the predicted propulsion speed of the planter based on the historical speed sequence of the planter, the current key material availability index, and the time until the next technical interval.

[0011] Step S4: Calculate the future construction progress curve based on the predicted advancement speed, combine the safety stock of each material and its unit consumption, dynamically calculate the demand progress point that triggers the next replenishment, and determine the predicted demand time and the corresponding plant machine location through calculation to form a dynamic demand node set.

[0012] Step S5: Based on the set of dynamic demand nodes, establish a multi-objective vehicle route planning model and solve for the optimal delivery route.

[0013] Step S6: Issue the optimal delivery route plan to the corresponding delivery vehicles for execution, and trigger rolling replanning at fixed intervals to update delivery instructions based on the latest construction status and vehicle location.

[0014] Furthermore, step S2 specifically includes the following steps:

[0015] Step S2.1: The coordinates of the manufacturing machine along the long axis of the factory building are collected in real time through the positioning device integrated in the manufacturing machine;

[0016] Step S2.2: Obtain the number of complete units that have been cast and formed through visual recognition;

[0017] Step S2.3: Based on the coordinates and the number of complete units, calculate the real-time construction progress of the plant construction machine by dividing the distance difference between the current position and the starting point by the standard unit length and adding the number of completed complete units.

[0018] Furthermore, step S3 specifically includes the following steps:

[0019] Step S3.1: Collect the historical operating speed sequence of the manufacturing machine;

[0020] Step S3.2: Calculate the critical material availability rate index. The critical material availability rate index is obtained by dividing the inventory of the material with the smallest inventory among all critical materials on site by the sum of the unit standard consumption of all critical materials.

[0021] Step S3.3: Obtain the remaining time until the next technical interval;

[0022] Step S3.4: Input the historical speed sequence, the key material availability index, and the time until the next technical interval into the adaptive speed prediction model, and output the predicted propulsion speed.

[0023] Further, in step S3.4, the adaptive speed prediction model generates the predicted speed in the following way: the baseline speed term, the time series term obtained by learning the historical speed sequence through a long short-term memory network, and the context correction term, which is a linear combination of the key material availability index and the exponential decay function of the time to technical interval, are weighted and summed according to preset and dynamically adjustable weight coefficients.

[0024] Furthermore, step S4 specifically includes the following steps:

[0025] Step S4.1: Integrate the predicted advancement speed to obtain the future construction progress curve;

[0026] Step S4.2: For each material, add the construction progress at the time of the last replenishment delivery to the number of units that the safety stock can support to obtain the demand trigger progress point for the next replenishment. The number of units that the safety stock can support is obtained by dividing the safety stock quantity of the material by its standard unit consumption.

[0027] Step S4.3: Based on the future construction progress curve, calculate the predicted demand time corresponding to the demand triggering progress point using an inverse function;

[0028] Step S4.4: Determine the predicted position of the manufacturing machine corresponding to the predicted demand time based on the preset manufacturing machine propulsion model;

[0029] Step S4.5: Represent each demand node as a quadruple containing the predicted demand time, predicted location, material demand quantity, and allowable time window, forming a dynamic demand node set.

[0030] Further, in step S4.4, the plant construction machine propulsion model is a spatiotemporal mapping relationship based on the future construction progress curve, used to convert the predicted demand time into the corresponding predicted location of the plant construction machine. Its calculation method includes:

[0031] Query the future construction progress curve to obtain the predicted construction progress value corresponding to the predicted demand time;

[0032] Based on the known coordinates of the starting point of the factory construction, the length of the standard unit, and the predicted construction progress value, the predicted coordinates of the factory building machine in the long axis direction of the factory are calculated through a linear mapping relationship.

[0033] Based on the characteristic that the position of the plant-building machine is relatively fixed along the width direction during plant construction, and combined with the predicted coordinates along the long axis direction, its complete two-dimensional coordinates in the construction plane are determined as the predicted position of the plant-building machine under the predicted time requirement.

[0034] Furthermore, step S5 specifically includes the following steps:

[0035] Step S5.1: Construct a multi-objective optimization function that includes a total travel cost term, an early arrival penalty term, a late arrival penalty term, and a return time balancing penalty term;

[0036] Step S5.2: Add vehicle capacity constraints, constraints that each demand node is served only once, constraints that vehicles depart from the warehouse and eventually return to the warehouse, and dynamic time window constraints to the multi-objective optimization function; wherein, the dynamic time window constraint requires that the service start time of vehicles arriving at each demand node should be within an allowable time deviation range that is dynamically set according to material properties before and after the predicted demand time.

[0037] Step S5.3: Solve the multi-objective vehicle routing planning model, which includes the multi-objective optimization function and constraints, to obtain the optimal delivery route scheme, which includes vehicle assignment, route sequence, and service time of each node.

[0038] Furthermore, in step S5.3, the NSGA-II algorithm is used to solve the multi-objective constrained vehicle path planning model.

[0039] A material requirements forecasting and distribution route planning system for manufacturing equipment, used to implement any one of the material requirements forecasting and distribution route planning methods for manufacturing equipment, including:

[0040] The data acquisition module is used to acquire the real-time coordinates of the manufacturing machine, construction images, on-site material inventory data, and delivery vehicle status data.

[0041] The communication transmission module is used to establish a real-time data link between the data acquisition module, the factory machine control center, the delivery vehicle and the central processing unit;

[0042] The predictive calculation module is used to perform adaptive velocity prediction and material requirement node calculation;

[0043] The path planning module is used to establish and solve vehicle path planning models with multi-objective constraints.

[0044] The task issuance and monitoring module is used to convert planning schemes into specific instructions for issuance and to monitor the task execution status.

[0045] Furthermore, the system also includes a digital twin simulation module, which is used to construct a virtual construction scenario based on the building information model of the factory, the real-time status of the factory machinery and the status of the delivery vehicles, and to simulate the delivery route plan generated by the route planning module.

[0046] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0047] 1. The adaptive speed prediction model constructed in this invention overcomes the shortcomings of traditional time series prediction methods in responding to complex disturbances at the construction site by introducing dynamic factors such as the key material availability rate index and the time between technical intervals. It can more accurately capture the impact of actual factors such as material supply and process connection on the speed of the machine, thus providing a highly reliable input for subsequent material demand prediction and laying the foundation for dynamic collaborative decision-making.

[0048] 2. This invention transforms the predicted advancement speed integral into a future continuous construction progress curve and dynamically calculates the demand trigger point based on the safety stock strategy, realizing the mapping of material demand from time to space. It can transform the discrete unit replenishment logic into a dynamic calculation model based on continuous progress, and output the material demand time and plant machine location at any future time in real time, so that the material distribution plan can be coupled with the construction process, realizing accurate perception of demand.

[0049] 3. The delivery route planning model established in this invention transforms core considerations of construction coordination, such as early arrival, late arrival, and vehicle task balancing, into quantifiable penalty terms. Together with traditional driving costs, these constitute a multi-objective optimization function. Under complex constraints such as vehicle capacity and dynamic time windows, it can effectively balance multiple competing objectives and generate a set of Pareto optimal delivery solutions. Decision-makers can select the most suitable solution based on real-time working conditions, thus solving the problem of spatiotemporal conflict between delivery and construction activities.

[0050] 4. This invention triggers rolling replanning by setting a fixed period, feeding back the latest factory machine status, inventory information and vehicle location to the prediction and planning closed loop, giving the system strong dynamic adaptability and robustness. It can respond in a timely manner to uncertain events such as construction speed fluctuations, emergency order insertions, and vehicle abnormalities, and continuously output updated optimal delivery instructions. Attached Figure Description

[0051] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0052] Figure 1 This is a flowchart illustrating an embodiment of the present invention;

[0053] Figure 2 This is a system schematic diagram according to an embodiment of the present invention;

[0054] Figure 3 This is a flowchart illustrating the predictive planning collaboration process according to an embodiment of the present invention. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0056] like Figure 1 As shown, the material demand forecasting and distribution route planning method for manufacturing machines includes the following steps:

[0057] Step S1: Obtain the design parameters of the target plant. The design parameters include the total number of standard units that are continuously repeated along the direction of the plant construction machine, the length of each standard unit, and the standard consumption of each type of construction material in each standard unit.

[0058] Step S2: Obtain the coordinates of the plant-making machine in the long axis direction of the plant in real time, and calculate the real-time construction progress of the plant-making machine.

[0059] Step S3: Construct and run an adaptive speed prediction model. The adaptive speed prediction model outputs the predicted propulsion speed of the planter based on the historical speed sequence of the planter, the current key material availability index, and the time until the next technical interval.

[0060] Step S4: Calculate the future construction progress curve based on the predicted advancement speed, combine the safety stock of each material and its unit consumption, dynamically calculate the demand progress point that triggers the next replenishment, and determine the predicted demand time and the corresponding plant machine location through calculation to form a dynamic demand node set.

[0061] Step S5: Based on the set of dynamic demand nodes, establish a multi-objective vehicle route planning model and solve for the optimal delivery route.

[0062] Step S6: Issue the optimal delivery route plan to the corresponding delivery vehicles for execution, and trigger rolling replanning at fixed intervals to update delivery instructions based on the latest construction status and vehicle location.

[0063] The specific input parameters for step S1 include: the total number of repeating units in the plant; the length of each standard unit; the set of material types (such as concrete, steel bars, formwork, etc.); the standard consumption of each material in a standard unit; and the safety stock of each material at the construction site.

[0064] Step S2 specifically includes the following steps:

[0065] Step S2.1: The coordinates of the manufacturing machine along the long axis of the factory building are collected in real time through the positioning device integrated in the manufacturing machine;

[0066] Step S2.2: Obtain the number of complete units that have been cast and formed through visual recognition;

[0067] Step S2.3: Based on the coordinates and the number of complete units, calculate the real-time construction progress of the plant construction machine by dividing the distance difference between the current position and the starting point by the standard unit length and adding the number of completed complete units.

[0068] The specific formula for the real-time construction progress is as follows:

[0069] ;

[0070] in, This represents the construction progress at time t, and indicates the equivalent number of units where work has been completed. For example, 3.5 means that 3 complete units have been completed, and the 4th unit is 50% complete. The coordinate of the manufacturing machine at time t along the long axis of the factory building, i.e., the X-axis, is expressed in meters. This indicates the coordinates of the starting point (zero point) of the factory construction along its major axis, in meters. Indicates the design length of a single standard unit. This indicates the number of fully formed standard units that have been verified and confirmed by the vision system as of the current moment. The vision recognition system deployed on the manufacturing machine automatically determines and updates this value. When the system recognizes that the casting and curing of a standard unit has fully met the process specifications, this value increases by 1.

[0071] Step S3 specifically includes the following steps:

[0072] Step S3.1: Collect the historical operating speed sequence of the manufacturing machine;

[0073] Step S3.2: Calculate the critical material availability rate index. The critical material availability rate index is obtained by dividing the inventory of the material with the smallest inventory among all critical materials on site by the sum of the unit standard consumption of all critical materials.

[0074] Step S3.3: Obtain the remaining time until the next technical interval;

[0075] Step S3.4: Input the historical speed sequence, the key material availability index, and the time until the next technical interval into the adaptive speed prediction model, and output the predicted propulsion speed.

[0076] The specific formula for the key material availability index is as follows:

[0077] ;

[0078] in, This represents the key material availability index. This indicates the immediate available inventory of critical material j at time t. This represents the unit consumption of material j. This represents the set of critical material types, which are materials for which interruptions are not permitted in the construction process. These typically include concrete, prestressed steel bars, etc., and are pre-defined in the construction organization design and input into the system.

[0079] The remaining time for the next technical interval specifically refers to the remaining time for the next mandatory technical interval, such as prestressing tensioning or structural acceptance.

[0080] The specific formula for the adaptive speed prediction model is as follows:

[0081] ;

[0082] in, Indicates the prediction step size is The propulsion speed of the factory machine can be predicted by setting the prediction step size to 15 minutes. Each time the model runs, it outputs the predicted propulsion speed 15 minutes later. This indicates the theoretical benchmark speed set according to the current technological mode, such as continuous casting or intermittent curing. This represents the Long Short-Term Memory (LSTM) neural network processing historical velocity sequences. The time series prediction term output after learning, Indicates the remaining time for the next technical interval. , , The weighting coefficients are dynamically adjusted separately, and their sum is 1. The initial values ​​can be set to 0.4, 0.4, and 0.2 respectively. , These represent internal fixed weights, which can be set to 0.7 and 0.3 respectively, emphasizing that material supply has a greater immediate impact than process connection. This represents the exponential decay coefficient, which can be set to 0.5. It is used to control the decay rate of the effect of the adjacent technical intervals on the speed.

[0083] The LSTM network uses historical construction data, including speed sequences, corresponding material inventory sequences, and process schedules, for supervised training with the goal of minimizing prediction error.

[0084] The weighting coefficients are fine-tuned in reverse based on the root mean square error between the model's predicted speed and the actual speed over a past time window, such as one hour. If the introduction of the context correction term significantly reduces the error, the weighting coefficients are appropriately increased. The weight.

[0085] In step S3.4, the adaptive speed prediction model generates the predicted speed in the following way: the baseline speed term, the time series term obtained by learning the historical speed sequence through a long short-term memory network, and the context correction term, which is a linear combination of the key material availability index and the exponential decay function of the time to technical interval, are weighted and summed according to preset and dynamically adjustable weight coefficients.

[0086] Step S4 specifically includes the following steps:

[0087] Step S4.1: Integrate the predicted advancement speed to obtain the future construction progress curve;

[0088] Step S4.2: For each material, add the construction progress at the time of the last replenishment delivery to the number of units that the safety stock can support to obtain the demand trigger progress point for the next replenishment. The number of units that the safety stock can support is obtained by dividing the safety stock quantity of the material by its standard unit consumption.

[0089] Step S4.3: Based on the future construction progress curve, calculate the predicted demand time corresponding to the demand triggering progress point using an inverse function;

[0090] Step S4.4: Determine the predicted position of the manufacturing machine corresponding to the predicted demand time based on the preset manufacturing machine propulsion model;

[0091] Step S4.5: Represent each demand node as a quadruple containing the predicted demand time, predicted location, material demand quantity, and allowable time window, forming a dynamic demand node set.

[0092] The specific formula for the future construction progress curve is as follows:

[0093] ;

[0094] in, This represents the predicted construction progress at any future time t. Indicates the current time Real-time construction progress, Let represent the integral variable, indicating the time from the start of integration. Any point in time within the time interval until the end of integration, t. Indicates at a point in time Predicted propulsion speed of the manufacturing machine;

[0095] The specific formula for triggering the progress point of the requirement is:

[0096] ;

[0097] in, This represents the demand trigger progress point for the k-th replenishment of material j. This indicates the construction progress value at the time the material was last replenished, delivered, and confirmed as received. This indicates the safety stock level set for material j. This represents the unit consumption of material j; Among the settings When the construction progress is 0, meaning construction has not yet started, a virtual zero-time replenishment is considered to have occurred. The first actual demand trigger point thereafter will be when the construction progress reaches [a certain threshold]. Occurs during unit operation; The dimension of the construction progress value is the number of standard units, which is dimensionless and represents the number of equivalent units that have been completed. The physical dimension of the safety stock depends on the material. For example, concrete is in cubic meters and steel bars are in tons. Similarly, the unit consumption is in cubic meters per unit and steel bars are in tons per unit. The result of dividing the safety stock by the unit consumption is the number of standard units.

[0098] The formula for calculating the predicted demand time is:

[0099] ;

[0100] in, This indicates the predicted demand time for the k-th replenishment of material j. This represents the inverse function of the projected progress curve. It finds the progress value equal to the projected progress value by performing linear interpolation on discrete data points of the future construction progress curve. The corresponding time point is used as the predicted demand time. After obtaining the future construction schedule curve, the demand-triggered schedule value is a specific vertical axis value on this curve. It is necessary to find the corresponding horizontal axis time point. In computer implementations, the future construction schedule curve is usually represented by a series of discrete time points and their corresponding schedule values. Therefore, numerical methods such as linear interpolation can be used to quickly locate the time point that satisfies this equation; this time point is the predicted demand time. This means that the conversion from progress value to time has been completed;

[0101] The specific formula for predicting the location of the manufacturing machine is as follows:

[0102] ;

[0103] in, The predicted location of the demand node for the k-th replenishment of material j is represented by a two-dimensional coordinate. The factory machine has a fixed working coordinate in the short axis direction of the factory building, that is, on the Y-axis.

[0104] In step S4.4, the plant construction machine propulsion model is based on the spatiotemporal mapping relationship of the future construction progress curve, used to convert the predicted demand time into the corresponding predicted position of the plant construction machine. Its calculation method includes:

[0105] Query the future construction progress curve to obtain the predicted construction progress value corresponding to the predicted demand time;

[0106] Based on the known coordinates of the starting point of the factory construction, the length of the standard unit, and the predicted construction progress value, the predicted coordinates of the factory building machine in the long axis direction of the factory are calculated through a linear mapping relationship.

[0107] Based on the characteristic that the position of the plant-building machine is relatively fixed along the width direction during plant construction, and combined with the predicted coordinates along the long axis direction, its complete two-dimensional coordinates in the construction plane are determined as the predicted position of the plant-building machine under the predicted time requirement.

[0108] Step S5 specifically includes the following steps:

[0109] Step S5.1: Construct a multi-objective optimization function that includes a total travel cost term, an early arrival penalty term, a late arrival penalty term, and a return time balancing penalty term;

[0110] Step S5.2: Add vehicle capacity constraints, constraints that each demand node is served only once, constraints that vehicles depart from the warehouse and eventually return to the warehouse, and dynamic time window constraints to the multi-objective optimization function; wherein, the dynamic time window constraint requires that the service start time of vehicles arriving at each demand node should be within an allowable time deviation range that is dynamically set according to material properties before and after the predicted demand time.

[0111] Step S5.3: Solve the multi-objective vehicle routing planning model, which includes the multi-objective optimization function and constraints, to obtain the optimal delivery route scheme, which includes vehicle assignment, route sequence, and service time of each node.

[0112] The specific formula for the multi-objective optimization function is as follows:

[0113] ;

[0114] ;

[0115] ;

[0116] ;

[0117] ;

[0118] in, Represents a multi-objective optimization function. , , and These represent the total travel cost, early arrival penalty, late arrival penalty, and return time equalization penalty, respectively. This represents the set of available delivery vehicles, where i represents the vehicle index. Indicates the total number of nodes. This represents the travel cost from node a to node b, which can be converted into time or distance. This represents a binary decision variable. If vehicle i travels directly from node a to node b, the value is 1; otherwise, it is 0. Represents the set of all dynamic demand nodes. This indicates the actual start time of service for the vehicle at demand node a. This indicates the actual service completion time of the vehicle at demand node a. This represents the predicted demand time for demand node a. This indicates the time it takes for vehicle i to return to the warehouse after completing its task. This represents the average planned return time of all vehicles within this scheduling cycle. , , These represent the weighting coefficients, and their specific settings are as follows: , , , Yuanda This demonstrates that delays leading to downtime are far more serious than premature construction that merely occupies space. The resource balance is relatively small, while meeting the timeliness requirements.

[0119] In step S5.3, the NSGA-II algorithm is used to solve the multi-objective constrained vehicle routing planning model. This algorithm can effectively handle multi-objective optimization problems and output a set of Pareto optimal solutions. Decision-makers can select the final execution plan from these solutions based on real-time preferences, such as prioritizing cost or timeliness.

[0120] The NSGA-II algorithm uses a population size of 100, 500 iterations, a crossover probability of 0.9, and a mutation probability of 0.1. After outputting the Pareto front, the algorithm defaults to selecting the solution with the minimum total penalty term as the final delivery plan to ensure priority for construction continuity.

[0121] The optimal delivery route plan (vehicle-route-timetable) is distributed to the delivery vehicles for execution; the system initiates rolling time-domain optimization at fixed intervals, such as every 15 minutes.

[0122] Collect the latest factory machine location, progress, inventory, and vehicle location; using the current moment as the new initial time, repeat steps S2 to S5 to generate a new round of delivery instructions; the new instructions only cover a future time window, such as the plan for the next 2 hours, and inherit the reasonable tasks currently being executed. Compare the new and old plans, and only issue differential instructions for the changed parts, such as new demand nodes and vehicle delays, while maintaining the unaffected tasks.

[0123] like Figure 2As shown, a material requirements forecasting and distribution route planning system for manufacturing machines is used to implement any of the material requirements forecasting and distribution route planning methods for manufacturing machines, including:

[0124] The data acquisition module is used to acquire the real-time coordinates of the manufacturing machine, construction images, on-site material inventory data, and delivery vehicle status data.

[0125] The communication transmission module is used to establish a real-time data link between the data acquisition module, the factory machine control center, the delivery vehicle and the central processing unit;

[0126] The predictive calculation module is used to perform adaptive velocity prediction and material requirement node calculation;

[0127] The path planning module is used to establish and solve vehicle path planning models with multi-objective constraints.

[0128] The task issuance and monitoring module is used to convert planning schemes into specific instructions for issuance and to monitor the task execution status.

[0129] The system also includes a digital twin simulation module, which is used to construct a virtual construction scenario based on the building information model of the factory, the real-time status of the factory machinery and the status of the delivery vehicles, and to simulate the delivery route plan generated by the route planning module.

[0130] The specific architecture of this invention includes:

[0131] 1. Data Acquisition and Communication Layer: Composed of various sensors and 5G / Industrial IoT gateways, responsible for real-time acquisition and transmission of all data elements.

[0132] 2. The intelligent decision-making platform layer specifically includes: a predictive calculation engine, which encapsulates an adaptive speed prediction model and demand calculation logic; a route planning engine, which encapsulates a multi-objective vehicle route planning model and the NSGA-II solver; and a digital twin simulation module, which uses BIM models and real-time data to simulate and rehearse the delivery plan in a virtual environment before issuing instructions, detects potential spatial or temporal conflicts, and ensures the feasibility of the plan.

[0133] 3. Application Execution and Monitoring Layer: Responsible for converting optimization solutions into specific control commands, and monitoring the collaborative status of manufacturing machines and delivery vehicles in real time, triggering alarms and replanning.

[0134] like Figure 3As shown, the data input layer integrates real-time data from multiple sources, such as the location of the manufacturing machine, material inventory, and historical speed. The adaptive prediction model predicts the future advancement speed of the manufacturing machine based on historical data and real-time context, generating a continuous future construction progress curve. The intelligent material calculation combines progress prediction and safety stock strategy to calculate the time and location of material replenishment. The dynamic demand node organizes the calculation results into a set of structured demand points. Each node contains information on time, location, demand quantity, and time window. The path planning takes the demand node as input and constructs a multi-objective optimization model to balance delivery cost, timeliness, and resource balance. The delivery plan is rehearsed in the digital twin environment to detect spatiotemporal conflicts and feasibility issues. If the plan is feasible, it is executed. If a conflict is found, feedback is given to replan. Feasible plans are issued to delivery vehicles for execution. The status is monitored in real time and data is fed back to form a closed-loop control.

[0135] Any combination of one or more computer-readable media may be used. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus, or device.

[0136] The examples described herein are merely preferred embodiments of the invention and are not intended to limit the concept and scope of the invention. Any modifications and improvements made by those skilled in the art to the technical solutions of the invention without departing from the design concept of the invention should fall within the protection scope of the invention.

Claims

1. A method for material demand forecasting and distribution route planning for manufacturing machines, characterized in that, Includes the following steps: Step S1: Obtain the design parameters of the target plant. The design parameters include the total number of standard units that are continuously repeated along the direction of the plant construction machine, the length of each standard unit, and the standard consumption of each type of construction material in each standard unit. Step S2: Obtain the coordinates of the plant-making machine in the long axis direction of the plant in real time, and calculate the real-time construction progress of the plant-making machine. Step S3: Construct and run an adaptive speed prediction model. The adaptive speed prediction model outputs the predicted propulsion speed of the planter based on the historical speed sequence of the planter, the current key material availability index, and the time until the next technical interval. Step S4: Calculate the future construction progress curve based on the predicted advancement speed, combine the safety stock of each material and its unit consumption, dynamically calculate the demand progress point that triggers the next replenishment, and determine the predicted demand time and the corresponding plant machine location through calculation to form a dynamic demand node set. Step S5: Based on the set of dynamic demand nodes, establish a multi-objective vehicle route planning model and solve for the optimal delivery route. Step S6: Issue the optimal delivery route plan to the corresponding delivery vehicles for execution, and trigger rolling replanning at fixed intervals to update delivery instructions based on the latest construction status and vehicle location.

2. The method according to claim 1, characterized in that, Step S2 specifically includes the following steps: Step S2.1: The coordinates of the manufacturing machine along the long axis of the factory building are collected in real time through the positioning device integrated in the manufacturing machine; Step S2.2: Obtain the number of complete units that have been cast and formed through visual recognition; Step S2.3: Based on the coordinates and the number of complete units, calculate the real-time construction progress of the plant construction machine by dividing the distance difference between the current position and the starting point by the standard unit length and adding the number of completed complete units.

3. The method according to claim 2, characterized in that, Step S3 specifically includes the following steps: Step S3.1: Collect the historical operating speed sequence of the manufacturing machine; Step S3.2: Calculate the critical material availability rate index. The critical material availability rate index is obtained by dividing the inventory of the material with the smallest inventory among all critical materials on site by the sum of the unit standard consumption of all critical materials. Step S3.3: Obtain the remaining time until the next technical interval; Step S3.4: Input the historical speed sequence, the key material availability index, and the time until the next technical interval into the adaptive speed prediction model, and output the predicted propulsion speed.

4. The method according to claim 3, characterized in that, In step S3.4, the adaptive speed prediction model generates the predicted speed in the following way: the baseline speed term, the time series term obtained by learning the historical speed sequence through a long short-term memory network, and the context correction term, which is a linear combination of the key material availability index and the exponential decay function of the time to technical interval, are weighted and summed according to preset and dynamically adjustable weight coefficients.

5. The method according to claim 4, characterized in that, Step S4 specifically includes the following steps: Step S4.1: Integrate the predicted advancement speed to obtain the future construction progress curve; Step S4.2: For each material, add the construction progress at the time of the last replenishment delivery to the number of units that the safety stock can support to obtain the demand trigger progress point for the next replenishment. The number of units that the safety stock can support is obtained by dividing the safety stock quantity of the material by its standard unit consumption. Step S4.3: Based on the future construction progress curve, calculate the predicted demand time corresponding to the demand triggering progress point using an inverse function; Step S4.4: Determine the predicted position of the manufacturing machine corresponding to the predicted demand time based on the preset manufacturing machine propulsion model; Step S4.5: Represent each demand node as a quadruple containing the predicted demand time, predicted location, material demand quantity, and allowable time window, forming a dynamic demand node set.

6. The method according to claim 5, characterized in that, In step S4.4, the plant construction machine propulsion model is based on the spatiotemporal mapping relationship of the future construction progress curve, used to convert the predicted demand time into the corresponding predicted position of the plant construction machine. Its calculation method includes: Query the future construction progress curve to obtain the predicted construction progress value corresponding to the predicted demand time; Based on the known coordinates of the starting point of the factory construction, the length of the standard unit, and the predicted construction progress value, the predicted coordinates of the factory building machine in the long axis direction of the factory are calculated through a linear mapping relationship. Based on the characteristic that the position of the plant-building machine is relatively fixed along the width direction during plant construction, and combined with the predicted coordinates along the long axis direction, its complete two-dimensional coordinates in the construction plane are determined as the predicted position of the plant-building machine under the predicted time requirement.

7. The method according to claim 6, characterized in that, Step S5 specifically includes the following steps: Step S5.1: Construct a multi-objective optimization function that includes a total travel cost term, an early arrival penalty term, a late arrival penalty term, and a return time balancing penalty term; Step S5.2: Add vehicle capacity constraints, constraints that each demand node is served only once, constraints that vehicles depart from the warehouse and eventually return to the warehouse, and dynamic time window constraints to the multi-objective optimization function; wherein, the dynamic time window constraint requires that the service start time of vehicles arriving at each demand node should be within an allowable time deviation range that is dynamically set according to material properties before and after the predicted demand time. Step S5.3: Solve the multi-objective vehicle routing planning model, which includes the multi-objective optimization function and constraints, to obtain the optimal delivery route scheme, which includes vehicle assignment, route sequence, and service time of each node.

8. The method according to claim 7, characterized in that, In step S5.3, the NSGA-II algorithm is used to solve the multi-objective constrained vehicle path planning model.

9. A material demand forecasting and distribution route planning system for manufacturing machines, used to implement the material demand forecasting and distribution route planning method for manufacturing machines as described in any one of claims 1-8, characterized in that, include: The data acquisition module is used to acquire the real-time coordinates of the manufacturing machine, construction images, on-site material inventory data, and delivery vehicle status data. The communication transmission module is used to establish a real-time data link between the data acquisition module, the factory machine control center, the delivery vehicle and the central processing unit; The predictive calculation module is used to perform adaptive rate prediction and material requirement node calculation; The path planning module is used to establish and solve vehicle path planning models with multi-objective constraints. The task issuance and monitoring module is used to convert planning schemes into specific instructions for issuance and to monitor the task execution status.

10. The system according to claim 9, characterized in that, The system also includes a digital twin simulation module, which is used to construct a virtual construction scenario based on the building information model of the factory, the real-time status of the factory machinery and the status of the delivery vehicles, and to simulate the delivery route plan generated by the route planning module.

Images