Production load balancing control system and control method thereof
By employing hierarchical control and signal processing technology, the problem of production load imbalance under the MTO mode was solved, enabling efficient and balanced production of multiple varieties and small batches of products, thereby improving production efficiency and equipment utilization.
Patent Information
- Application Number
- CN202610065409.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-19
- Publication Date
- 2026-05-05
AI Technical Summary
Existing production scheduling systems in MTO mode suffer from problems such as equipment conflicts, load imbalance, delayed disturbance response, and poor algorithm adaptability, resulting in low production plan executability, uneven equipment workload, and high order delay rates.
By employing hierarchical control theory, constraint programming algorithm and signal processing technology, Fourier load smoothing algorithm is used to suppress high-frequency fluctuations, three-dimensional evaluation matrix is used to optimize order sequence and dynamic rule base engine is used to match equipment, so as to achieve balanced control of production load.
It has achieved load balancing in the production of multiple varieties and small batches of products, improved production efficiency and equipment utilization, shortened order response time, and reduced production costs.
Smart Images
Figure CN121979140A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of production scheduling technology, and in particular to a method for calculating production load. Background Technology
[0002] Existing production scheduling systems have significant shortcomings when dealing with the MTO (Made to Order) model in the pump and valve industry: 1) Static scheduling defects: Traditional ERP / MRP relies on the assumption of unlimited capacity and cannot handle dynamic constraints such as equipment conflicts and mold-specific matching, resulting in low plan executability; 2) Load imbalance problem: Single-level scheduling strategies (such as only workshop dispatching or order placement control) lack global coordination, resulting in uneven equipment workload (volatility > 30%) and an average order pool retention of more than 48 hours; 3) Delayed response to disturbances: Events such as order insertion or equipment failure trigger a full schedule rescheduling, with adjustment time exceeding 2 hours, resulting in an order delay rate of ≥25%; 4) Poor algorithm adaptability: General APS tools do not model the characteristics of "long process chain + multiple resource constraints", and cannot optimize core bottlenecks in valve production such as mold switching and material matching. Summary of the Invention
[0003] In view of the load imbalance problem in the MTO production scheduling, it is necessary to propose a production load balancing control system and its control method. By integrating hierarchical control theory, constraint programming algorithm and signal processing technology, the load imbalance problem in the production of multi-variety, small-batch products is solved.
[0004] A production load balancing control system includes a top-level module, a middle-level module, and a bottom-level module, and the three modules can communicate with each other, realizing the information flow of the entire balancing control system. The top-level module is used for order input and analysis. That is, customer orders are input into the top-level module, and the order load is converted to the frequency domain through the Fourier load smoothing algorithm to achieve the purpose of low-pass filtering and suppressing high-frequency fluctuations. The middle-layer module is used to arrange the optimal order of orders. Under the premise of the ORR optimal rule sorting, the three-dimensional evaluation matrix method is used to arrange the optimal order of orders and generate the order delivery sequence, which is also the order delivery priority. The bottom-level module is used for dispatching work at the workshop level, enabling the distribution of order tasks generated by the top-level module and intermediate modules to various work equipment.
[0005] This technical solution achieves balanced production scheduling for multiple varieties and small batches of products. It uses the Fourier load smoothing algorithm to suppress the problem of over-scheduling at the source, and uses the three-dimensional evaluation matrix method to determine the optimal production sequence of orders, thereby improving production efficiency. Through a hierarchical control architecture, it achieves balanced load across the entire chain from order acceptance to delivery to work assignment.
[0006] A control method for a production load balancing control system, comprising: The top-level module inputs the customer order set and generates the admission order set based on the Fourier load smoothing algorithm, thus suppressing the risk of production system overload from the source. The admission order set is input into the middle layer module, and an order placement sequence is generated based on the three-dimensional evaluation matrix, which is the production order priority of the orders in the admission order set. In the underlying module, the process route of the placed orders is input according to the order placement sequence. Based on the dynamic rule base engine, the orders are matched with the operating equipment to generate machine-level operation instructions, thereby achieving balanced production scheduling of equipment load.
[0007] By implementing the control method, the idle rate of equipment and the actual waiting time are reduced, thereby improving operational efficiency and reducing production costs.
[0008] Furthermore, in order to successfully implement the control method of the equilibrium control system, a coordination logic for the control method is also provided in the equilibrium control system. The coordination logic includes a positive condition transmission path and a negative state feedback path. The positive condition transmission path includes the top-level module transmitting order information to the middle-level module, the middle-level module transmitting the order placement priority to the bottom-level module, and the bottom-level module assigning production tasks to various operating equipment; thus realizing the smooth issuance and transmission of instructions at all levels. The directional status feedback path includes the bottom module feeding back the load of the operating equipment and the status of the operating device to the middle module, and the bottom module feeding back the inventory of work-in-process to the top module; thus realizing the feedback of actual production information to the balance control system.
[0009] By setting the aforementioned collaborative logic in the control method, the interaction of information and data flow between various modules in the balanced control system are realized, providing a mechanism to ensure the production and adjustment of multi-variety, small-batch products, and also enabling convenient order insertion, etc.
[0010] Furthermore, the method for generating the admission order set in the top-level module is as follows: Input a customer order set O, which includes information such as product specifications, required quantity, and expected delivery date. Denote the customer order set O = {01, 02, ..., 0n}. The CTP delivery commitment algorithm is used to calculate the committable delivery date of the customer order set O based on the historical capacity regression model; The Fourier load smoothing algorithm is used to transform the load sequence of the customer order set into the frequency domain, and the high-frequency fluctuation components are suppressed by low-pass filtering to achieve balanced load distribution. Output or generate the admission order set O´ and promised delivery date T0 of the customer order set O.
[0011] Furthermore, the method for generating the order delivery sequence in the middle-layer module is as follows: Input the admission order set O´ and the real-time manufacturing resource status, which includes information such as equipment load, material inventory, and mold status, as a prerequisite for order priority ranking. The deployment priority of each access order set O' is calculated using a three-dimensional evaluation matrix according to Formula 1. The evaluation of the three-dimensional evaluation matrix includes three dimensions: operating equipment, materials, and molds. Specifically, by dynamically calculating the weighted evaluation values of the three evaluation dimensions of operating equipment load margin, material availability rate, and mold readiness, the deployment sequence of access order sets O' can be precisely controlled, achieving seamless connection between new loads and in-process tasks. P=α·(1-L_util)+β·R_mat+γ·S_mold ............................. (Formula 1) Where α, β and γ∈[0,1] are adjustable weighted evaluation values, L_util is the equipment load margin, R_mat is the material kitting rate, and S_mold is the mold readiness. The FCP finite capacity planning algorithm is adopted, with the peak load of a single operating equipment as the objective, to solve the constraints of order and equipment matching, and generate order placement sequence and placement time.
[0012] Furthermore, the method for generating the machine-level operation instructions in the underlying module is as follows: Enter the process route for the orders that have been placed. The orders that have been placed refer to the set of admission orders O´ that have already been scheduled for production. The order task is assigned using a dynamic rule base engine. If the work equipment queue is empty, the EDD rule is selected. If there is a bottleneck process, the CPM critical path optimization is applied, or the SPT rule is enabled. The machine-level operation instructions are generated, which include information such as the ID of the operation equipment, the start time of the process, the end time of the process, and the work group of the process. This achieves the purpose of decomposing the order tasks to specific operation equipment and specific operation time, thus ensuring the smooth production of products.
[0013] Furthermore, when order insertion is required, a perturbation response mechanism is adopted to initiate a variable neighborhood search algorithm to rearrange the affected processes, thereby inserting new orders into the original order delivery sequence and forming a new order delivery sequence after the insertion. In other words, a new order delivery sequence is generated, consisting of the inserted order and the subsequent undelivered orders, thus achieving convenience in order insertion and ease of adjustment.
[0014] The beneficial effects of the technical solution of this invention are: it achieves uniform load distribution for multi-variety, small-batch production scheduling, convenient order insertion, high scheduling efficiency, and compact production. Furthermore, the Fourier load smoothing algorithm can suppress systemic overload from the source, the three-dimensional evaluation matrix quantifies ORR decision-making to improve the scientific nature of deployment, and the hierarchical collaboration mechanism realizes global to local load linkage balance. The modular algorithm library supports rapid adaptation to different production line scales. Attached Figure Description
[0015] Figure 1 This is a schematic diagram of the hierarchical control mechanism of the present invention. Detailed Implementation
[0016] To more clearly illustrate the technical solution of the present invention, the technical solution of the invention will be described in detail with reference to the accompanying drawings. Obviously, the following description is some typical embodiments of the present invention. For those skilled in the art, other solutions can be obtained based on these embodiments without creative effort.
[0017] The technical solution of this patent will now be illustrated using valve production scheduling as an example.
[0018] In one implementation, the production load balancing control system includes a top-level module, a middle-level module, and a bottom-level module, and the three modules can communicate with each other, realizing the information flow of the entire balancing control system; it realizes a hierarchical collaborative mechanism for production planning and scheduling, improving the efficiency and feasibility of balancing production scheduling; The top-level module is used for order admission, that is, customer orders are input into the top-level module, and the order load is converted to the frequency domain through the Fourier load smoothing algorithm to achieve the purpose of low-pass filtering and suppressing high-frequency fluctuations; The middle-layer module is used to arrange the optimal order of orders. Under the premise of the ORR optimal rule sorting, the optimal order of orders is arranged by a three-dimensional evaluation matrix method, that is, the order delivery priority is generated. The bottom-level module is used for dispatching work at the workshop level, enabling the distribution of order tasks generated by the top-level module and intermediate modules to various work equipment.
[0019] Another implementation method, a control method for a production load balancing control system, includes, Input the customer order set into the top-level module, and generate the admission order set based on the Fourier load smoothing algorithm; The admission order set is input into the middle layer module, and an order placement sequence is generated based on the three-dimensional evaluation matrix, which is the production order priority of the orders in the admission order set. In the underlying module, the process route of the placed orders is input according to the order placement sequence. Based on the dynamic rule base engine, the orders are matched with the operating equipment to generate machine-level operation instructions, thereby achieving balanced production scheduling of equipment load.
[0020] As a supplement to this embodiment, in order to smoothly implement the control method of the equalization control system, a coordination logic for the control method is also provided in the equalization control system. The coordination logic includes a positive condition transmission path and a reverse state feedback path. The positive condition transmission path includes the top-level module transmitting order information to the middle-level module, the middle-level module transmitting the order placement priority to the bottom-level module, and the bottom-level module assigning production tasks to various operating equipment; thus realizing the smooth issuance and transmission of instructions at all levels. The directional status feedback path includes the bottom module feeding back the load of the operating equipment and the status of the operating device to the middle module, and the bottom module feeding back the inventory of work-in-process to the top module; thus realizing the feedback of actual production information to the balance control system.
[0021] As a supplement to this embodiment, the method for generating the admission order set in the top-level module is as follows: Input a customer order set O, which includes information such as product specifications, required quantity, and expected delivery date. Denote the customer order set O = {01, 02, ..., 0n}. The CTP delivery commitment algorithm is used to calculate the committable delivery date of the customer order set O based on the historical capacity regression model; The Fourier load smoothing algorithm is used to transform the load sequence of the customer order set into the frequency domain, and the high-frequency fluctuation components are suppressed by low-pass filtering to achieve balanced load distribution. Output or generate the admission order set O´ and promised delivery date T0 of the customer order set O.
[0022] As a supplement to this embodiment, the method for generating the order delivery sequence in the middle-layer module is as follows: Input the admission order set O´ and the real-time manufacturing resource status, which includes information such as equipment load, material inventory, and mold status, as a prerequisite for order priority ranking. The deployment priority of each access order set O' is calculated using a three-dimensional evaluation matrix according to Formula 1. The evaluation of the three-dimensional evaluation matrix includes three dimensions: operating equipment, materials, and molds. Specifically, by dynamically calculating the weighted evaluation values of the three evaluation dimensions of operating equipment load margin, material availability rate, and mold readiness, the deployment sequence of access order sets O' can be precisely controlled, achieving seamless connection between new loads and in-process tasks. P=α·(1-L_util)+β·R_mat+γ·S_mold ............................. (Formula 1) Where α, β and γ∈[0,1] are adjustable weighted evaluation values, L_util is the equipment load margin, R_mat is the material kitting rate, and S_mold is the mold readiness. The FCP finite capacity planning algorithm is adopted, with the peak load of a single operating equipment as the objective, to solve the constraints of order and equipment matching, and generate order placement sequence and placement time.
[0023] As a supplement to this embodiment, the method for generating the machine-level operation instructions in the underlying module is as follows: Enter the process route for the orders that have already been placed; The order task is assigned using a dynamic rule base engine. If the work equipment queue is empty, the EDD rule is selected. If there is a bottleneck process, the CPM critical path optimization is applied, or the SPT rule is enabled. The machine-level operation instructions are generated, which include information such as the ID of the operation equipment, the start time of the process, the end time of the process, and the work group of the process. This achieves the purpose of decomposing the order tasks to specific operation equipment and specific operation time, thus ensuring the smooth production of products.
[0024] As a further supplement to this implementation method, when it is necessary to insert an order, a disturbance response mechanism is adopted to initiate a variable neighborhood search algorithm to rearrange the affected processes, thereby achieving the convenience of order insertion and the ease of adjustment.
[0025] Through the implementation of this invention, compared with the traditional MRP model, significant improvements have been achieved in the following indicators (as shown in Table 1), effectively ensuring stable and efficient production of multi-variety, small-batch products.
[0026] Table 1. Comparison of indicators between the patented technical solution and the traditional MRP model. Serial Number index This invention Traditional system Increase Remark 1 Equipment utilization volatility ≤10% 25%~35% ≥60%↓ 2 Order insertion response time 8.2 hours 48 hours 83%↓ 3 Emergency order response time 9.5 minutes 130 minutes 93%↓ 4 Delivery time achievement rate 97.3% 78.5% 24%↑ 5 Work-in-process turnover 15.2 times / year 8.7 times / year 75%↑ The above embodiments are merely descriptions of a typical application of the technical solution of the present invention. Reasonable extensions can be made without requiring creative effort.
Claims
1. A production load balancing control system, characterized in that, It includes a top-level module, a middle-level module, and a bottom-level module, and the three modules can communicate with each other; The top-level module is used for order input and analysis, and transforms the order load to the frequency domain through the Fourier load smoothing algorithm; The middle-layer module is used to arrange the optimal order of orders. Under the premise of the ORR optimal rule sorting, the optimal order of orders is arranged by a three-dimensional evaluation matrix method to generate an order delivery sequence. The bottom-level module is used for dispatching work at the workshop level, enabling the distribution of order tasks generated by the top-level module and intermediate modules to various work equipment.
2. A control method for a production load balancing control system as described in claim 1, characterized in that, include, Input the customer order set into the top-level module, and generate the admission order set based on the Fourier load smoothing algorithm; Input the admission order set into the middle-layer module, and generate an order delivery sequence based on the three-dimensional evaluation matrix; In the underlying module, the process route of the placed orders is input according to the order placement sequence, and the order is matched with the working equipment to generate machine-level operation instructions based on the dynamic rule base engine.
3. The control method of the production load balancing control system as described in claim 2, characterized in that, The equalization control system also includes a coordination logic for the control method, which includes a positive condition transmission path and a negative state feedback path. The positive condition transmission path includes the top-level module transmitting order information to the middle-level module, the middle-level module transmitting the order placement priority to the bottom-level module, and the bottom-level module assigning production tasks to various operating equipment; thus realizing the smooth issuance and transmission of instructions at all levels. The directional status feedback path includes the bottom module feeding back the load of the working equipment and the status of the working device to the middle module, and the bottom module feeding back the inventory of finished products to the top module. This enables the feedback of actual production information to the balance control system.
4. The control method for the production load balancing control system as described in claim 2, characterized in that, The method for generating the admission order set in the top-level module is as follows: Enter customer order set O; The CTP delivery commitment algorithm is used to calculate the committable delivery date of the customer order set O based on the historical capacity regression model; The Fourier load smoothing algorithm is used to transform the load sequence of the customer order set into the frequency domain; Generate the admission order set O´ and the promised delivery date T0 for the customer order set O.
5. The control method of the production load balancing control system as described in claim 4, characterized in that, The method for generating the order delivery sequence in the middle-layer module is as follows: Input the admission order set O´ and the real-time manufacturing resource status as prerequisites for order priority ranking; The deployment priority of each access order set O´ is calculated using a three-dimensional evaluation matrix according to P=α·(1-L_util)+β·R_mat+γ·S_mold, where α, β and γ∈[0,1] are adjustable weighted evaluation values, L_util is the equipment load margin, R_mat is the material kitting rate and S_mold is the mold readiness. The FCP finite capacity planning algorithm is adopted, with the peak load of a single operating equipment as the objective, to solve the constraints of order and equipment matching, and generate order placement sequence and placement time.
6. The control method for the production load balancing control system as described in claim 5, characterized in that, The method for generating the machine-level operation instructions in the underlying module is as follows. Enter the process route for the orders that have already been placed; The order task is assigned using a dynamic rule base engine. If the work equipment queue is empty, the EDD rule is selected. If there is a bottleneck process, the CPM critical path optimization is applied, or the SPT rule is enabled. Generate machine-level operation instructions.
7. The control method for the production load balancing control system as described in claim 6, characterized in that, When order insertion is required, a perturbation response mechanism is adopted to initiate a variable neighborhood search algorithm to rearrange the affected processes and generate a new order placement sequence of the inserted order and subsequent unplaced orders.