Metering method, device and system for setting production by sales
By building a sales forecast model and an inventory capacity calculation model, and combining it with a linear programming algorithm to optimize production capacity allocation, the problems of production continuity interruption and resource waste in open-pit sand and gravel aggregate mine production were solved, and scientific and efficient production planning was achieved.
Patent Information
- Application Number
- CN202510711166.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-29
- Publication Date
- 2025-09-09
AI Technical Summary
In open-pit sand and gravel aggregate mine production, the traditional full-load production model leads to interruptions in production continuity, increased energy consumption and waste of resources. In addition, the lack of scientific quantitative model support makes it difficult to accurately balance the relationship between inventory, sales and production capacity.
Adopting a sales-based production measurement method, by collecting and calibrating production data, building a sales forecast model and inventory capacity calculation model, and combining linear programming algorithms to optimize the production capacity allocation of finished materials of multiple specifications, the company realizes the automatic generation and dynamic adjustment of production plans.
It realizes the scientificity and flexibility of production planning, avoids inventory overflow, improves resource utilization, reduces production costs, and improves production line operation efficiency.
Smart Images
Figure CN120612043A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of accounting methods, and in particular relates to a sales-based production measurement method, device and system. Background Art
[0002] In open-pit sand and gravel aggregate mining, traditional production planning generally adopts a full-load production model, which arranges production capacity according to fixed production hours and lacks dynamic response to market demand and inventory status. This model has significant flaws: when the inventory of a certain specification of aggregate in the finished product silo reaches the upper limit of capacity (i.e., "storage overflow"), the production line needs to be shut down urgently to avoid material overflow, resulting in interruption of production continuity. This not only seriously affects production efficiency, but also increases energy consumption and maintenance costs due to frequent equipment startup and shutdown. In addition, full-load production does not take into account the proportional coordination of multiple specifications of finished materials, which may lead to a backlog of some specifications and a shortage of other specifications, resulting in waste of resources and delayed market response.
[0003] While some companies have attempted to adjust production plans through manual experience, this lacks scientific quantitative models, making it difficult to accurately balance inventory, sales, and production capacity. For example, traditional sales forecasts often rely on simple statistical averages, failing to capture sales fluctuations; capacity allocation is based on fixed ratios and unable to adapt to market demand fluctuations. These issues are particularly prominent in the production of multi-specification finished materials. There is an urgent need for an intelligent measurement method that can integrate sales data, inventory status, and production capacity constraints in real time to achieve dynamic optimization based on sales and improve the scientific nature and flexibility of production planning. Summary of the Invention
[0004] The technical problem to be solved by the present invention is to provide a sales-based production measurement method, device and system, which can dynamically match sales volume and production capacity, avoid inventory overflow, improve resource utilization, reduce production costs, and is suitable for production scenarios of finished materials with multiple specifications.
[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: A method, device and system for measuring production based on sales, comprising the following steps: S1: Collect and calibrate production-related data: Collect real-time and historical data including inventory data, sales data, and production capacity parameters; S2: Calculate the maximum production plan for a single-specification finished material: Calculate the available inventory capacity (RL) based on the inventory data obtained in S1. Use S1's historical sales data to train a quadratic polynomial regression model to predict tomorrow's sales (XL). Combined with S1's production capacity parameters, the minimum value of the "inventory and sales constraint" and the "capacity constraint" is taken to determine the maximum feasible production plan quantity (SCmax) for a single-specification finished material. S3: Optimize the ratio of finished materials with multiple specifications and generate a total plan: Based on the maximum production plan quantity SCmax of a single specification obtained in S2, optimize the capacity allocation ratio of each specification through a linear programming algorithm under the process constraints of the finished material ratio, and calculate the planned quantity SCi of each specification and the total planned quantity SC 总 , and deduce the actual production time H=SC 总 / Designed production capacity; S4: Execute the production plan and adjust it dynamically: The total planned quantity, output of each specification, and production time generated by S3 are converted into production instructions and issued. By monitoring inventory and production progress in real time, if any abnormality is found, the S2-S3 process is re-triggered to update the plan, forming a closed-loop management of "data collection-calculation-execution-feedback".
[0006] The inventory data includes current inventory and maximum inventory; The sales data includes historical daily sales data and today's actual sales; The production capacity parameters include designed production capacity and planned production time.
[0007] Preferably, in S4, the abnormality detected includes inventory approaching an upper limit or a sudden change in sales volume.
[0008] Preferably, the sub-steps of S1 are: S1.1: Use the warehouse intelligent scanning device to scan the finished material storage location in real time; obtain the current inventory quantity KC of each specification of finished material; and compare and calibrate it with the maximum inventory quantity recorded in the warehouse management system; S1.2: Extract today’s actual sales data for each specification of finished materials from the weighing sensor; simultaneously export the historical daily sales data for the past 30 days to form a time series dataset ; is the time serial number; For sales volume; Eliminate outliers in historical data due to equipment failures and holidays; outliers include single-day sales that exceed ±3 times the mean standard deviation; S1.3: Read the designed production capacity and preset planned production time in the production line control system; confirm that the parameters have not been illegally modified.
[0009] Preferably, the sub-steps of S2 are: S2.1: Calculation of available inventory capacity RL: Available inventory capacity RL = maximum inventory - (yesterday's inventory + today's production CL - today's actual sales XL); Yesterday's inventory comes from historical archives, and today's production CL comes from the production management system; S2.2: Tomorrow's sales XL forecast modeling: Using quadratic polynomial regression model y= ax 2+ bx + c Fit the historical sales data of S1.2; solve the parameters using the least squares method a, b 、 c ; Minimize the sum of squares of prediction errors; Substitute tomorrow's time sequence number into the quadratic polynomial regression model; calculate the sales forecast; and manually correct it based on today's actual sales trend from S1.2; S2.3: Capacity calculation under dual constraints: Inventory and sales constraints: SCmax1=RL+XL; Capacity constraint: SCmax2 = designed capacity × planned production time; Maximum feasible production plan quantity: SCmax=min(SCmax1;SCmax2).
[0010] Preferably, the sub-steps of S3 are: S3.1 Set the proportion parameters of each specification of finished materials K 1- K 4; satisfy: K 1+ K 2+ K 3+ K 4=10; 1≤ K 1; K 2; K 3; K 4≤7; S3.2: Calculation of single specification planned quantity and total planned quantity: Maximum production plan quantity for each specification: SC i =SCmax×( Ki / 10); Total planned volume: S3.3: Linear programming optimization of scale parameters: Objective function: maximize the minimum value of the planned quantity of each specification; that is, max(min(SC1; SC2; SC3; SC4)) Constraint: SC i ≤SCmax; K 1- K 4 is a positive integer and satisfies the proportional sum constraint Solving the optimal solution by the simplex method K 1- K 4 combinations; recalculate SC i and SC 总 ; S3.4: Derivation and verification of production time: Calculate actual production time: H=SC 总 / Designed production capacity like H >Plan production time; compress specifications proportionally Ki Until H ≤ Planned production time; or trigger manual decision-making.
[0011] Preferably, S4.1: Synchronize tomorrow's production order to the control system of mining excavators, crushers, conveyor belts and other equipment through the industrial Internet of Things system; tomorrow's production order includes the production plan quantity SC of each specification of finished material i , total planned quantity SC 总 and production time H ; S4.2: Collect inventory data and production progress data every hour and compare them with planned values. If the remaining inventory capacity of a certain specification is less than 10% of the maximum inventory, the capacity switching logic will be automatically triggered. If the actual sales growth rate is greater than the forecast value by 20%, immediately rerun steps S2.2-S3.3 and generate an emergency production plan; S4.3: At the end of each day, actual production, sales, inventory data, and plan execution deviations are stored in the historical database to form a closed-loop record. The model training process is automatically triggered every Sunday night. The sales forecast model of S2.2 is refitted with the latest 7 days of data. The parameters are updated. a, b 、 c ; Improve the accuracy of next week's forecast.
[0012] Preferably, in S1.1, if the scan data deviates from the system record by more than 5%, a calibration anomaly occurs, triggering a manual review process.
[0013] A measurement system for determining production based on sales adopts the measurement method for determining production based on sales. A metering device for determining production based on sales, comprising: one or more processors; The processor is configured to store one or more programs; When the one or more programs are executed by the one or more processors, a sales-based production measurement method as described in any one of claims 1 to 8 is implemented.
[0014] The present invention can achieve the following beneficial effects: 1. By building sales forecast models, inventory capacity calculation models and multi-specification production capacity optimization models, the production plan can be automatically generated and dynamically adjusted.
[0015] 2. Through real-time inventory capacity calculation and sales forecast, the production plan of single-specification finished materials is dynamically constrained to ensure that the production scale does not exceed the sum of warehouse capacity and market demand. Emergency shutdowns caused by overstocking are eliminated from the source, thereby improving production line operation efficiency and reducing unplanned downtime.
[0016] 3. Optimize the proportion parameters of finished materials of various specifications based on the linear programming algorithm, balance the production capacity distribution under process constraints, and avoid backlogs or shortages of a single specification. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] The present invention will be further described below with reference to the accompanying drawings and examples: Picture 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION
[0018] The preferred solution is Picture 1 As shown in the figure, a measurement method based on sales volume is as follows: S1: Collect and calibrate production-related data: Collect real-time and historical data including inventory data, sales data, and production capacity parameters; S2: Calculate the maximum production plan for a single-specification finished material: Calculate the available inventory capacity (RL) based on the inventory data obtained in S1. Use S1's historical sales data to train a quadratic polynomial regression model to predict tomorrow's sales (XL). Combined with S1's production capacity parameters, the minimum value of the "inventory and sales constraint" and the "capacity constraint" is taken to determine the maximum feasible production plan quantity (SCmax) for a single-specification finished material. S3: Optimize the ratio of finished materials with multiple specifications and generate a total plan: Based on the maximum production plan quantity SCmax of a single specification obtained in S2, optimize the capacity allocation ratio of each specification through a linear programming algorithm under the process constraints of the finished material ratio, and calculate the planned quantity SCi of each specification and the total planned quantity SC 总 , and deduce the actual production time H=SC 总 / Designed production capacity; S4: Execute the production plan and adjust it dynamically: The total planned quantity, output of each specification, and production time generated by S3 are converted into production instructions and issued. By monitoring inventory and production progress in real time, if any abnormality is found, the S2-S3 process is re-triggered to update the plan, forming a closed-loop management of "data collection-calculation-execution-feedback".
[0019] Anomalies detected include inventory approaching the upper limit or sudden changes in sales.
[0020] The inventory data includes the current inventory and the maximum inventory. The inventory data is used to calculate the available inventory capacity to avoid inventory overflow. The sales data includes historical daily sales data and today's actual sales data. Sales data (historical + real-time) is used to train forecasting models and dynamically adjust plans. The capacity parameters mentioned above include the designed capacity and the planned production duration. These parameters (designed capacity and planned duration) are rigid constraints on production capacity and ensure the feasibility of the plan.
[0021] Preferably, the sub-steps of S1 are: S1.1: Use the warehouse intelligent scanning device to scan the finished material storage location in real time; obtain the current inventory quantity KC of each specification of finished material; and compare and calibrate it with the maximum inventory quantity recorded in the warehouse management system; Smart devices (such as LiDAR) enable contactless inventory counting, improving data real-time availability. Comparison and calibration ensure inventory data accuracy. If the discrepancy exceeds 5% (e.g., the system records 1,000 tons, but the scanned data measures 950 tons), a manual review is triggered to prevent data errors from causing planning errors. If the scanned data deviates by more than 5% from the system record, a calibration anomaly is detected, triggering a manual review process.
[0022] S1.2: Extract today’s actual sales data for each specification of finished materials from the weighing sensor; simultaneously export the historical daily sales data for the past 30 days to form a time series dataset ; is the time serial number; For sales volume; Eliminate outliers in historical data due to equipment failures and holidays; outliers include single-day sales that exceed ±3 times the mean standard deviation; The weighing sensor is a floor scale, which is used to record sales in real time to ensure the authenticity of the data.
[0023] S1.3: Read the designed production capacity and preset planned production time in the production line control system; confirm that the parameters have not been illegally modified.
[0024] Preferably, the sub-steps of S2 are: S2.1: Calculation of available inventory capacity RL: Available inventory capacity RL = maximum inventory - (yesterday's inventory + today's production CL - today's actual sales XL); Yesterday's inventory comes from historical archives, and today's production CL comes from the production management system; S2.2: Tomorrow's sales XL forecast modeling: Using quadratic polynomial regression model y = ax 2+ bx + c Fit the historical sales data of S1.2; solve the parameters using the least squares method a, b 、 c; Minimize the sum of squares of prediction errors; Substitute tomorrow's time sequence number into the quadratic polynomial regression model; calculate the sales forecast; and manually correct it based on today's actual sales trend from S1.2; The quadratic polynomial model is suitable for scenarios where sales trends accelerate or decelerate (e.g., sales growth accelerates during peak season); The least squares method finds the best fitting curve through mathematical optimization to minimize the error between the model prediction value and the historical data; Manual corrections (such as adjustments based on recent order volumes) combined with empirical judgment can improve forecast robustness.
[0025] S2.3: Capacity calculation under dual constraints: Inventory and sales constraints: SCmax1=RL+XL; ensure that the production plan does not exceed the sum of inventory capacity RL and market demand (XL).
[0026] Capacity constraint: SCmax2 = designed capacity × planned production duration; ensure that the production plan does not exceed the maximum possible output of the production line on that day (capacity × duration).
[0027] Maximum feasible production plan: SCmax = min(SCmax1; SCmax2). Taking the minimum value reflects the "barrel effect," with the strictest constraint serving as the upper limit for production capacity.
[0028] The available inventory capacity RL reflects the remaining storage capacity of the warehouse to avoid production stoppage caused by overproduction. The aforementioned Sales Forecast XL: predicts market demand through historical data modeling, aligning production with actual sales; The dual constraints described above take minimum values: ensuring that the production plan does not exceed either the inventory capacity or the actual production capacity of the production line.
[0029] Preferably, the sub-steps of S3 are: S3.1 Set the proportion parameters of each specification of finished materials K 1- K 4; satisfy: K 1+ K 2+ K 3+ K 4=10; 1≤ K 1; K 2; K 3; K 4≤7; the total proportion is 10 for ease of calculation (e.g. K1=3 represents a 30% proportion); the proportion of a single specification is ≥1 and ≤7 to avoid extreme allocation (e.g. a certain specification accounts for too high a proportion, resulting in the inability to produce other specifications).
[0030] S3.2: Calculation of single specification planned quantity and total planned quantity: Maximum production plan quantity for each specification: SC i =SCmax×( Ki / 10); Total planned volume: S3.3: Linear programming optimization of scale parameters: Objective function: maximize the minimum value of the planned quantity of each specification; that is, max(min(SC1; SC2; SC3; SC4)) Constraint: SC i ≤SCmax; K 1- K 4 is a positive integer and satisfies the proportional sum constraint Solving the optimal solution by the simplex method K 1- K 4 combinations; recalculate SC i and SC 总 ; S3.4: Derivation and verification of production time: Calculate actual production time: H =SC 总 / Designed production capacity like H >Plan production time; compress specifications proportionally Ki Until H ≤ Planned production time; or trigger manual decision-making.
[0031] The process constraints K1-K4 are: setting the output ratio of each specification according to the process characteristics of the production line (such as the crushing ratio) to ensure a smooth production process; The linear programming optimization mentioned above: balances the production capacity of each specification under the constraints, avoids underproduction or overproduction of certain specifications, and improves overall efficiency; The production time H is calculated based on the total planned quantity and the designed production capacity of the production line to guide production scheduling.
[0032] If the production time exceeds the planned time (for example, H=9 hours is calculated and the plan is 8 hours), the ratio needs to be adjusted to adapt to the time limit, or a manual decision is made whether to work overtime to ensure that the plan can be executed.
[0033] Preferably, S4.1: Synchronize tomorrow's production order to the control system of mining excavators, crushers, conveyor belts and other equipment through the industrial Internet of Things system; tomorrow's production order includes the production plan quantity SC of each specification of finished material i , total planned quantity SC 总 and production time H ; S4.2: Collect inventory data and production progress data every hour and compare them with planned values. If the remaining inventory capacity of a certain specification is less than 10% of the maximum inventory, the capacity switching logic will be automatically triggered. If the actual sales growth rate is greater than the forecast value by 20%, immediately rerun steps S2.2-S3.3 and generate an emergency production plan; When the remaining inventory capacity is less than 10%, the production of that specification will be automatically suspended and adjusted to other specifications; if the sales growth rate exceeds expectations (for example, the predicted daily sales volume is 100 tons, and the actual growth rate reaches 120 tons / day), the emergency process will be triggered to re-optimize the plan to avoid supply and demand imbalance.
[0034] S4.3: At the end of each day, actual production, sales, inventory data, and plan execution deviations are stored in the historical database to form a closed-loop record. The model training process is automatically triggered every Sunday night. The sales forecast model of S2.2 is refitted with the latest 7 days of data. The parameters are updated. a, b 、 c ; Improve the accuracy of next week's forecast.
[0035] A measurement system for determining production based on sales adopts the measurement method for determining production based on sales. A metering device for determining production based on sales, comprising: one or more processors; The processor is configured to store one or more programs; When the one or more programs are executed by the one or more processors, a sales-based production measurement method as described in any one of claims 1 to 8 is implemented.
[0036] The above embodiments are merely preferred technical solutions of the present invention and should not be construed as limiting the present invention. The scope of protection of the present invention shall be the technical solutions set forth in the claims, including equivalent alternatives to the technical features of the technical solutions set forth in the claims. In other words, equivalent alternatives and improvements within this scope are also within the scope of protection of the present invention.
Claims
1. A sales-based production measurement method characterized by The following steps are involved: S1: Collect and calibrate production-related data: Collect real-time and historical data including inventory data, sales data, and production capacity parameters; S2: Calculate the maximum production plan for a single-specification finished material: Calculate the available inventory capacity (RL) based on the inventory data obtained in S1. Use S1's historical sales data to train a quadratic polynomial regression model to predict tomorrow's sales volume (XL). Combined with S1's production capacity parameters, the minimum value of the "Inventory and Sales Constraint" and the "Capacity Constraint" is taken to determine the maximum feasible production plan quantity (SCmax) for a single-specification finished material. S3: Optimize the ratio of finished materials with multiple specifications and generate a total plan: Based on the maximum production plan quantity SCmax of a single specification obtained in S2, optimize the capacity allocation ratio of each specification through a linear programming algorithm under the process constraints of the finished material ratio, and calculate the planned quantity SCi of each specification and the total planned quantity SC 总 , and deduce the actual production time H=SC 总 / Designed production capacity; S4: Execute the production plan and adjust it dynamically: The total planned quantity, output of each specification, and production time generated by S3 are converted into production instructions and issued. By monitoring inventory and production progress in real time, if any anomalies are found, the S2-S3 process is re-triggered to update the plan, forming a closed-loop management of "data collection - calculation - execution - feedback".
2. The method of measuring production based on sales according to claim 1, characterized in that: The inventory data includes current inventory and maximum inventory; The sales data includes historical daily sales data and today's actual sales; The production capacity parameters include designed production capacity and planned production time.
3. The method of measuring production based on sales according to claim 1, characterized in that: In S4, anomalies found include inventory approaching the upper limit or sudden changes in sales.
4. The method of measuring production based on sales according to claim 1, characterized in that: The sub-steps of S1 are: S1.1: Use the warehouse intelligent scanning device to scan the finished material storage location in real time; obtain the current inventory quantity KC of each specification of finished material; and compare and calibrate it with the maximum inventory quantity recorded in the warehouse management system; S1.2: Extract today’s actual sales data for each specification of finished materials from the weighing sensor; simultaneously export the historical daily sales data for the past n days to form a time series dataset ; is the time serial number; For sales volume; Eliminate outliers in historical data due to equipment failures and holidays; outliers include single-day sales that exceed ±3 times the mean standard deviation; S1.3: Read the designed production capacity and preset planned production time in the production line control system; confirm that the parameters have not been illegally modified.
5. The method of measuring production based on sales according to claim 1, characterized in that: The sub-steps of S2 are: S2.1: Calculation of available inventory capacity RL: Available inventory capacity RL = maximum inventory - (yesterday's inventory + today's production CL - today's actual sales XL); Yesterday's inventory comes from historical archives, and today's production CL comes from the production management system; S2.2: Tomorrow's sales XL forecast modeling: Using quadratic polynomial regression model y = ax 2+ bx + c Fit the historical sales data of S1.2; solve the parameters using the least squares method a, b 、 c ; Minimize the sum of squares of prediction errors; Substitute tomorrow's time serial number into the quadratic polynomial regression model; calculate the sales forecast value; And make manual corrections based on the actual sales trend of S1.2 today; S2.3: Capacity calculation under dual constraints: Inventory and sales constraints: SCmax1=RL+XL; Capacity constraint: SCmax2 = designed capacity × planned production time; Maximum feasible production plan quantity: SCmax=min(SCmax1;SCmax2).
6. The method of measuring production based on sales according to claim 1, characterized in that: The sub-steps of S3 are: S3.1 Set the proportion parameters of each specification of finished materials K 1- K 4; satisfy: K 1+ K 2+ K 3+ K 4=10; 1≤ K 1; K 2; K 3; K 4≤7; S3.2: Calculation of single specification planned quantity and total planned quantity: Maximum production plan quantity for each specification: SC i =SCmax×( Ki / 10); Total planned volume: S3.3: Linear programming optimization of scale parameters: Objective function: maximize the minimum value of the planned quantity of each specification; that is, max(min(SC1; SC2; SC3; SC4)) Constraint: SC i ≤SCmax; K 1- K 4 is a positive integer and satisfies the proportional sum constraint Solving the optimal solution by the simplex method K 1- K 4 combinations; recalculate SC i and SC 总 ; S3.4: Derivation and verification of production time: Calculate actual production time: H =SC 总 / Designed production capacity like H >Planned production time; Compress each specification proportionally Ki Until H ≤Planned production time; Or trigger manual decision-making.
7. The method of measuring production based on sales according to claim 1, characterized in that: S4.1: Synchronize tomorrow’s production order to the control systems of mining excavators, crushers, conveyors and other equipment through the Industrial Internet of Things system; tomorrow’s production order includes the production plan quantity SC of each specification of finished material i , total planned quantity SC 总 and production time H ; S4.2: Collect inventory data and production progress data every hour and compare them with planned values. If the remaining inventory capacity of a certain specification is less than 10% of the maximum inventory, the capacity switching logic will be automatically triggered. If the actual sales growth rate is greater than the forecast value by 20%; Immediately rerun steps S2.2-S3.3; Generate emergency production plans; S4.3: At the end of each day, actual production, sales, inventory data, and plan execution deviations are stored in a historical database to form a closed-loop record. The model training process is automatically triggered every Sunday night. Refit the sales forecast model of S2.2 using the latest m days of data; update the parameters a, b 、 c ; Improve the accuracy of next week's forecast.
8. The method of measuring production based on sales according to claim 1, characterized in that: In S1.1, if the scan data deviates from the system record by more than 5%, a calibration anomaly occurs, triggering a manual review process.
9. A metering system for production based on sales, characterized by: A sales-based production measurement method according to any one of claims 1-8 is adopted.
10. A metering device for determining production based on sales, characterized in that: include: one or more processors; The processor is configured to store one or more programs; When the one or more programs are executed by the one or more processors, a sales-based production measurement method as described in any one of claims 1 to 8 is implemented.