Dispatching method and device, storage medium and electronic equipment
By obtaining equipment production data information in the foundry, predicting the demand for the conversion process, and optimizing the dispatch plan using the dispatch decision model, the problem of wasting time in the equipment conversion process is solved, and equipment utilization and output efficiency are improved.
Patent Information
- Application Number
- CN202310953849.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-07-31
- Publication Date
- 2025-02-14
AI Technical Summary
In wafer foundry, the equipment conversion process leads to the wasted time of the equipment utilization, affecting the equipment output. How to control the frequency of the conversion process through reasonable dispatch methods, reduce the time wasted in the equipment conversion process, improve the equipment utilization rate, and achieve the realization of the equipment utilization rate. Maximize output.
By obtaining the production data information of each equipment in the target equipment group, we predict whether the next production unit needs to convert the process, the optimal output and the optimal availability, and use the preset dispatch decision model to calculate the total output At the maximum value, each device executes the output value of each process and performs product dispatch.
By predicting and optimizing the dispatch plan, rationally allocate the process, reducing the frequency of conversion processes, improving equipment utilization, maximizing output, and avoiding wasting equipment time.
Smart Images
Figure CN119443538A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of semiconductor technology, and more specifically to a dispatching method and apparatus, a storage medium, and an electronic device. Background Technology
[0002] Wafer foundries need to continuously expand their markets to improve capacity and equipment utilization, which is crucial for the sustainable development of the company.
[0003] Due to the complexity of wafer fabrication processes, even products manufactured using the same process may have different equipment requirements, including gas type, temperature, and pressure. Therefore, when processing different wafer products on the same equipment, a process transfer operation is necessary. This operation allows for the readjustment of the equipment's internal environment. However, this readjustment process wastes equipment time, impacting output. Given limited equipment resources, the most pressing issue is how to rationally allocate work to control the frequency of process transfers for equipment groups requiring such operations, thereby reducing wasted time and maximizing equipment utilization to achieve maximum output.
[0004] In the current wafer manufacturing process, the dispatching of products for equipment groups that require process change operations does not fully consider the time wasted caused by the process change, which leads to the underutilization of equipment time and thus output loss. Therefore, the dispatching scheme made under such circumstances is often not optimal. Summary of the Invention
[0005] The problem this invention aims to solve is: how to allocate tasks more rationally in order to improve equipment utilization and maximize output.
[0006] To address the above problems, embodiments of the present invention provide a work dispatching method, the method comprising:
[0007] Obtain production data information of each device in the target device group in the current production unit; the target device group consists of a group of devices with the same function.
[0008] Based on the acquired production data, predict whether each piece of equipment in the target equipment group needs to change processes in the next production unit, the optimal output value of each piece of equipment in the target equipment group in the next production unit, and the optimal utilization rate.
[0009] Based on the prediction results and the types of processes that the equipment in the target equipment group can execute, a preset dispatching decision model is used to calculate the output value of each equipment in the target equipment group when executing each process at the maximum total output of the target equipment group; the types of processes include: processes that require process conversion and processes that do not require process conversion;
[0010] Product dispatch is based on the calculation results.
[0011] Optionally, the acquired production data information includes:
[0012] The optimal output value of each device in the target equipment group in the current production unit;
[0013] The optimal availability of each piece of equipment in the target equipment group within the current production unit;
[0014] The quantity of different products to be manufactured in the target equipment group;
[0015] Whether each piece of equipment in the target equipment group is switching processes in the current production unit.
[0016] Optionally, the acquired production data information may also include:
[0017] The number of emergency workpieces for different products in the target equipment group.
[0018] Optionally, the step of predicting whether each piece of equipment in the target equipment group needs to change processes in the next production unit, the optimal output value and the optimal utilization rate of each piece of equipment in the target equipment group in the next production unit, based on the acquired production data information, includes:
[0019] Using an extreme learning machine model, it is predicted whether each piece of equipment in the target equipment group needs to change processes in the next production unit, the optimal output value of each piece of equipment in the target equipment group in the next production unit, and the optimal utilization rate.
[0020] Optionally, the extreme learning machine model is obtained by learning from historical data using the extreme learning algorithm.
[0021] Optionally, the preset dispatch decision model is a linear programming model, and the linear programming model uses the prediction result as a constraint.
[0022] Optionally, the constraints include:
[0023] The sum of the outputs of each process on the same equipment is less than or equal to the predicted optimal output value;
[0024] The available time for each process on the same non-transfer process equipment is less than or equal to the predicted optimal available time.
[0025] The available time and lost time of each process on the same transfer equipment are less than or equal to the predicted optimal available time.
[0026] The production quantity of a single process in the next production unit is greater than or equal to the minimum production quantity.
[0027] Optionally, the linear programming model is:
[0028]
[0029] Where Cap represents the total output of the target equipment group, i represents the i-th equipment in the target equipment group, N1 represents the total number of equipment in the target equipment group that does not need to be converted to a new process, N2 represents the total number of equipment in the target equipment group that needs to be converted to a new process, j represents the j-th process among the processes that the equipment in the target equipment group can execute, H1 represents the total number of processes among the executable processes in the target equipment group that do not need to be converted to a new process, H2 represents the total number of processes among the executable processes in the target equipment group that need to be converted to a new process, and W ij This represents the output when the i-th device performs the j-th process.
[0030] This invention also provides a dispatching device, the device comprising:
[0031] The data acquisition unit is adapted to acquire production data information of each device in the target device group in the current production unit, wherein the target device group consists of a group of devices with the same function;
[0032] The prediction unit is adapted to predict, based on the acquired production data information, whether each piece of equipment in the target equipment group needs to change processes in the next production unit, the optimal output value of each piece of equipment in the target equipment group in the next production unit, and the optimal utilization rate.
[0033] The decision-making unit is adapted to calculate the output value of each device in the target equipment group when executing each process, based on the prediction results and the types of processes that the devices in the target equipment group can execute, using a preset dispatch decision model; the types of processes include: processes that require process conversion and processes that do not require process conversion;
[0034] The dispatching unit is suitable for dispatching products based on calculation results.
[0035] This invention also provides a computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor to implement the steps of any of the methods described above.
[0036] This invention also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the steps of any of the methods described above when running the computer program.
[0037] Compared with the prior art, the technical solution of the embodiments of the present invention has the following advantages:
[0038] The present invention, based on the production data of the current production unit, predicts the optimal values of production attributes for the next production unit. This includes whether each piece of equipment in the target equipment group needs to switch processes in the next production unit, the optimal output value of each piece of equipment in the target equipment group, and the optimal utilization rate. Then, based on the prediction results and the types of processes that the equipment in the target equipment group can execute, a preset dispatching decision model is used to calculate the output value of each piece of equipment executing each process when the total output of the target equipment group reaches its maximum. This allows for product dispatching based on the calculation results. Since the prediction of whether each piece of equipment needs to switch processes in the next production unit is obtained, a dispatching scheme is subsequently determined based on the prediction results and the types of processes that the equipment in the target equipment group can execute. This allows for the reasonable allocation of executable processes to the corresponding equipment, controlling the frequency of process switching as much as possible to maximize the total output of the target equipment group. Attached Figure Description
[0039] Figure 1 This is a flowchart of a dispatching method according to an embodiment of the present invention;
[0040] Figure 2 This is a schematic diagram of the structure of an extreme learning machine algorithm;
[0041] Figure 3 This is a schematic diagram of the structure of a dispatching device in an embodiment of the present invention. Detailed Implementation
[0042] In the current wafer manufacturing process, the dispatching of products for equipment groups that require process change operations does not fully consider the time wasted caused by the process change, which leads to the underutilization of equipment time and thus output loss. Therefore, the dispatching scheme made under such circumstances is often not optimal.
[0043] To address this problem, the present invention provides a dispatching method that predicts whether each device will need to switch processes in the next production unit. Subsequently, based on the prediction results and the types of processes that can be executed by the devices in the target device group, a dispatching scheme is determined. This allows the executable processes to be rationally allocated to the corresponding devices for execution, thereby controlling the frequency of process switching as much as possible and maximizing the total output of the target device group.
[0044] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.
[0045] Reference Figure 1 This invention provides a dispatching method, which may include the following steps:
[0046] Step 11: Obtain production data information of each device in the target device group in the current production unit. The target device group consists of a group of devices with the same function.
[0047] In specific implementation, the target equipment group may include two or more devices. Each device in the target equipment group has the same function, and therefore can execute at least one identical process. The dispatching method in this embodiment of the invention is used to allocate products among multiple devices executing the same process to maximize total output. For example, the target equipment group may include three devices, namely device E1, device E2, and device E3, which can execute three processes, namely process Re1, process Re2, and process Re3. Through the dispatching method in this embodiment of the invention, it can be determined whether devices E1 to E3 execute processes Re1 to Re3, and the output of each device when executing the corresponding process.
[0048] In practice, a 24-hour period can be considered a production unit. The current production unit is the 24 hours preceding the current moment. The next production unit is the 24 hours following the current moment. By acquiring the production data of each device in the target equipment group within the current production unit, the optimal production attribute values for each device in the target equipment group in the next production unit can be predicted.
[0049] In other embodiments, other durations (such as 48 hours) can also be used as a production unit, and there is no limitation here.
[0050] In practice, production data can be pre-stored on the corresponding platform, and the dispatching device can access the platform and read the relevant information to provide data support for subsequent dispatching.
[0051] In one embodiment of the present invention, the production data information may include: the optimal output value of each device in the target equipment group in the current production unit, the optimal utilization rate of each device in the target equipment group in the current production unit, the quantity of different products to be manufactured in the target equipment group, and whether each device in the target equipment group has changed processes in the current production unit.
[0052] For example, a target equipment group may include three devices, namely device E1, device E2, and device E3. The target equipment group needs to produce a total of five products. Then, the production data information may include: the optimal output of devices E1 to E3 in the current production unit, the optimal availability of devices E1 to E3 in the current production unit, whether devices E1 to E3 have changed processes, and the number of work-in-process (WIP) for each of the five products in the target equipment group.
[0053] Based on the above production data, the optimal values of the production attributes of the next production unit can be predicted. The optimal values of the production attributes of the next production unit obtained at this time can minimize the frequency of process changeovers and avoid wasting production capacity.
[0054] In another embodiment of the present invention, the production data information may include: the number of urgent workpieces for different products in the target equipment group. For example, if the target equipment group needs to produce a total of 5 products, the number of urgent workpieces for each of these 5 products in the target equipment group. Urgent workpieces refer to the number of workpieces with short delivery periods, and the delivery period for urgent workpieces can be set according to actual circumstances.
[0055] By predicting the optimal production attribute values for the next production unit based on the number of urgent workpieces of the same product in the target equipment group, the optimal production attribute values for the next production unit can be made to meet the urgency of the workpieces, thereby balancing the urgency of the workpieces with capacity waste. In the event of an urgent workpiece situation, even if it results in some capacity waste, the delivery deadline should be met as much as possible.
[0056] Step 12: Based on the acquired production data, predict whether each piece of equipment in the target equipment group needs to change processes in the next production unit, the optimal output value of each piece of equipment in the target equipment group in the next production unit, and the optimal utilization rate.
[0057] In practical implementation, a pre-trained Extreme Learning Machine model can be used to predict whether each device in the target equipment group needs to change its process in the next production unit, the optimal output value of each device in the target equipment group in the next production unit, and the optimal utilization rate. In this embodiment of the invention, whether each device in the target equipment group needs to change its process in the next production unit, the optimal output value of each device in the target equipment group in the next production unit, and the optimal utilization rate are collectively referred to as the optimal production attribute values.
[0058] In practice, the extreme learning machine model is obtained by learning from historical data using the extreme learning algorithm. (Refer to...) Figure 2 Extreme Learning (ELM) is a feedforward neural network learning method that mainly consists of three layers: an input layer, a hidden layer, and an output layer. The input weights of this algorithm (such as W...) j ) and hidden layer threshold (e.g., β) jk The values can be randomly generated and do not need to be adjusted during training; only the number of neurons in the hidden layer needs to be set to obtain a unique optimal solution. Where x1 to x... n This represents the input to the Extreme Learning algorithm, y1 to y2. m This represents the output of the Extreme Learning algorithm.
[0059] In practice, the number of neurons in the hidden layer of the Extreme Learning Machine (ELM) algorithm can be determined first. Using historical data as the sample set, the Sigmold function is selected as the activation function for model training. The resulting output weight matrix indicates that the model has been successfully trained. The historical data includes production data from several production units preceding the current production unit.
[0060] In practice, the Extreme Learning Machine model can be used to predict whether each piece of equipment needs to change processes in the next production unit, the optimal output value of each piece of equipment in the target equipment group in the next production unit, and the optimal utilization rate.
[0061] For example, if "0" indicates that the equipment does not need to change processes in the next production unit, and "1" indicates that the equipment needs to change processes in the next production unit, then for equipment i, the following output information can be predicted:
[0062] move i，t+1 ;
[0063] UT i，t+1 ;
[0064] and Zzc i，t+1 .
[0065] Among them, move i，t+1 UT represents the optimal output of device i in the next production unit. i，t+1 Zzc represents the optimal availability of equipment i in the next production unit. i，t+1 Indicates whether device i needs to change processes in the next production unit, Zzc i，t+1 =0 or Zzc i，t+1 =1.
[0066] The Extreme Learning Machine model takes the production data information of the current production unit t as input and predicts the optimal production attribute value of the next 24-hour production unit t+1, thus obtaining the output parameter value, which serves as the data input for dispatching decisions.
[0067] Step 13: Based on the prediction results and the types of processes that the equipment in the target equipment group can execute, use a preset dispatch decision model to calculate the output value of each equipment in the target equipment group when executing each process when the total output of the target equipment group is at its maximum value.
[0068] The types of processes include: processes that require process conversion and processes that do not require process conversion.
[0069] In one embodiment of the present invention, the preset dispatch decision model can be a linear programming model, which uses the prediction result as a constraint to maximize production forecast.
[0070] Specifically, the process and equipment transition relationship can be determined first, and then a linear programming model can be used to solve the final output allocation scheme by taking the output of each product on the equipment as the variable and the output maximization as the objective.
[0071] When determining the process and equipment transition relationships, a relationship matrix can be constructed based on the equipment requirements of different processes. Processes can then be categorized based on this matrix. The relationship matrix can include two types: ① Relationship matrix Re, consisting of processes that do not require transition, with a total number of processes H1; ② Relationship matrix Re', consisting of processes that require transition, with a total number of processes H2. A process requiring transition refers to a process whose equipment requirements change relative to the current process, such as changes in temperature or pressure. By categorizing processes, subsequent process allocation can prioritize assigning processes requiring transition to the necessary equipment, while processes that do not require transition are not limited to specific equipment.
[0072] Based on the prediction results of whether each piece of equipment in the target equipment group needs to change its process in the next production unit, the equipment in the target equipment group is divided into two categories: ①E, which consists of equipment that does not need to change its process in the next production unit, and the total number of equipment that does not need to change its process in the next production unit is N1; ②E', which consists of equipment that needs to change its process in the next production unit, and the total number of equipment that needs to change its process in the next production unit is N2.
[0073] Assuming Cap represents the total output of the target equipment group, the linear programming model can be:
[0074]
[0075] Where Cap represents the total output of the target equipment group, i represents the i-th equipment in the target equipment group, N1 represents the total number of equipment in the target equipment group that does not need to be converted to a new process, N2 represents the total number of equipment in the target equipment group that needs to be converted to a new process, j represents the j-th process among the processes that the equipment in the target equipment group can execute, H1 represents the total number of processes among the executable processes in the target equipment group that do not need to be converted to a new process, H2 represents the total number of processes among the executable processes in the target equipment group that need to be converted to a new process, and Wij represents the output of the i-th equipment when executing the j-th process.
[0076] The constraints of the above linear programming model can be:
[0077] ① The sum of the outputs of each process on the same equipment is less than or equal to the predicted optimal output. Taking equipment i as an example, i represents any device in the target equipment group, which can be either a device that does not require process switching in the next production unit or a device that does require process switching in the next production unit.
[0078] ② The available time for each process on the same non-transfer equipment is less than or equal to the predicted optimal available time. Taking the following production unit that does not require transfer equipment i as an example... UT i，t+1 This represents the optimal availability of device i in the next production unit, 24 * UT. i，t+1 P represents the optimal available time for device i in the next production unit. ij This indicates the availability of process j on equipment i.
[0079] ③ The available time and lost time for each process on the same process transfer equipment are less than or equal to the predicted optimal available time. Taking the next production unit requiring process transfer equipment i as an example... Los indicates the time lost due to the need to switch to new process equipment.
[0080] ④ The production quantity of a single process in the next production unit is greater than or equal to the minimum production quantity. Taking equipment i as an example, The production quantity of process j in the next production unit must not be lower than the minimum production quantity S. j Minimum production quantity S j It is known.
[0081] Step 14: Dispatch products based on the calculation results.
[0082] Assuming the target equipment group consists of 3 devices, the optimal production attribute values for the next production unit are predicted using the Extreme Learning Machine model, as shown in Table 1.
[0083] Table 1
[0084] equipment Optimal output Is process conversion necessary? Optimal availability E1 500 0 95％ E2 600 0 93％ E3 650 1 88％
[0085] Where 0 indicates that the equipment does not need to change processes in the next production unit, and 1 indicates that the equipment needs to change processes in the next production unit.
[0086] Table 2
[0087] Re 1 Re 2 Re 3 E1 313 25 E2 25 364 E3 326 25 25
[0088] Using the parameters output in Table 1 as input parameters for the linear programming model, a linear programming model is constructed. It is assumed that the next production unit executes three processes: Re1, Re2, and Re3. Process Re3 requires a process change, with a loss time of 1.5 hours, and the minimum requirement for each process is 25 pieces. The optimal work assignment decision for the next production unit is obtained using the linear programming model, as shown in Table 2.
[0089] Using the work assignment scheme in Table 2, the Cap is 1103pcs.
[0090] The dispatching method in this embodiment of the invention can reasonably control the number of process transfer equipment, rationally allocate the products, reduce equipment efficiency loss, and significantly improve equipment output.
[0091] To enable those skilled in the art to better understand and implement the present invention, the apparatus, computer-readable storage medium, and electronic device corresponding to the above method are described in detail below.
[0092] Reference Figure 3 This invention provides a dispatching device 30, which may include: a data acquisition unit 31, a prediction unit 32, a decision-making unit 33, and a dispatching unit 34. Wherein:
[0093] The data acquisition unit 31 is adapted to acquire production data information of each device in the target device group in the current production unit, wherein the target device group consists of a group of devices with the same function.
[0094] The prediction unit 32 is adapted to predict, based on the acquired production data information, whether each piece of equipment in the target equipment group needs to change processes in the next production unit, the optimal output value of each piece of equipment in the target equipment group in the next production unit, and the optimal utilization rate.
[0095] The decision unit 33 is adapted to calculate the output value of each device in the target equipment group when executing each process, based on the prediction results and the types of processes that the devices in the target equipment group can execute, using a preset dispatch decision model; the types of processes include: processes that require process conversion and processes that do not require process conversion.
[0096] The dispatching unit 34 is adapted to dispatch products based on calculation results.
[0097] In practice, prediction unit 32 is trained using historical data and takes the production data of the current production unit as input to predict the optimal value of the attributes for the next production unit. Decision unit 33 uses the predicted optimal value of the production attributes as a constraint to optimize the allocation of processing quantities.
[0098] The dispatching device 30 can be a chip or a chip module. The data acquisition unit 31, prediction unit 32, decision-making unit 33, and dispatching unit 34 can be implemented with reference to the above description of the method steps, and will not be repeated here.
[0099] Regarding the modules / units included in the various devices and products described in the above embodiments, they can be software modules / units, hardware modules / units, or a combination of both. For example, for various devices and products applied to or integrated into a chip, all of their modules / units can be implemented using hardware methods such as circuits, or at least some modules / units can be implemented using software programs that run on a processor integrated within the chip, while the remaining (if any) modules / units can be implemented using hardware methods such as circuits; for various devices and products applied to or integrated into a chip module, all of their modules / units can be implemented using hardware methods such as circuits, and different modules / units can be located in the same component (e.g., chip, circuit module, etc.) or different components of the chip module, or at least some modules / units can be implemented using hardware methods such as circuits. The components can be implemented using software programs that run on a processor integrated within the chip module. The remaining (if any) modules / units can be implemented using hardware methods such as circuits. For various devices and products applied to or integrated into the terminal, all their modules / units can be implemented using hardware methods such as circuits. Different modules / units can be located in the same component (e.g., chip, circuit module, etc.) or different components within the terminal. Alternatively, at least some modules / units can be implemented using software programs that run on a processor integrated within the terminal, while the remaining (if any) modules / units can be implemented using hardware methods such as circuits. This embodiment of the invention also provides another computer-readable storage medium storing computer instructions. When these computer instructions are executed, they perform the control method steps of any of the electronic devices described in the above embodiments, which will not be repeated here.
[0100] This invention also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, which is executed by a processor to implement the steps of any of the methods described above.
[0101] In specific implementations, the computer-readable storage medium may include ROM, RAM, disk, or optical disk, etc.
[0102] This invention also provides an electronic device, including a memory and a processor. The memory stores a computer program that can run on the processor. When the processor runs the computer program, it performs the steps of any of the methods described above, which will not be repeated here.
[0103] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of the invention; therefore, the scope of protection of the present invention should be determined by the scope defined in the claims.
Claims
1. A method for dispatching work, characterized in that: include: Acquire production data information of each device in a target device group in a current production unit, wherein the target device group is composed of a group of devices having the same function; Based on the acquired production data information, predict whether each device in the target device group needs to change processes in the next production unit, the optimal value of the output of each device in the target device group in the next production unit, and the optimal value of the availability; Based on the prediction results and the types of processes executable by the equipment in the target equipment group, the output value of each equipment in the target equipment group when executing each process is calculated using a preset dispatch decision model when the total output of the target equipment group is the maximum value; The types of processes include: processes that require process conversion and processes that do not require process conversion; Product dispatch is carried out based on the calculation results.
2. The method for dispatching work according to claim 1, characterized in that: The acquired production data information includes: the optimal value of each device in the target device group output in the current production unit; The optimal value of the availability of each device in the target device group in the current production unit; The number of different products to be manufactured in the target equipment group; Whether each device in the target device group is changing processes in the current production unit.
3. The method for dispatching work according to claim 2, characterized in that: The acquired production data also includes: The number of urgent workpieces for different products in the target equipment group.
4. The method for dispatching work according to claim 1, characterized in that: The method of predicting whether each device in the target device group needs to switch processes in the next production unit, the optimal output value and the optimal utilization rate of each device in the target device group in the next production unit based on the acquired production data information includes: The extreme learning machine model is used to predict whether each device in the target device group needs to change processes in the next production unit, the optimal value of the output of each device in the target device group in the next production unit, and the optimal value of the availability.
5. The method for dispatching work according to claim 4, characterized in that: The extreme learning machine model is obtained by learning historical data using an extreme learning algorithm.
6. The method for dispatching work according to claim 1, characterized in that: The preset work dispatch decision model is a linear programming model, and the linear programming model uses the prediction results as constraints.
7. The method for dispatching work according to claim 6, characterized in that: The constraints include: The sum of the outputs of each process of the same equipment is less than or equal to the predicted optimal output value; The available time of each process of the same non-conversion process equipment is less than or equal to the optimal value of the predicted available time; The available time and loss time of each process of the same process conversion equipment are less than or equal to the optimal value of the predicted available time; The production quantity of a single process in the next production unit is greater than or equal to the minimum production quantity.
8. The method for dispatching work according to claim 6, characterized in that: The linear programming model is: Wherein, Cap represents the total output of the target equipment group, i represents the i-th equipment in the target equipment group, N1 represents the total number of equipment in the target equipment group that does not need to be converted, N2 represents the total number of equipment in the target equipment group that needs to be converted, j represents the j-th process in the executable processes of the equipment in the target equipment group, H1 represents the total number of processes in the executable processes in the target equipment group that do not need to be converted, H2 represents the total number of processes in the executable processes in the target equipment group that need to be converted, and W ij Represents the output when the i-th equipment executes the j-th process.
9. A work dispatching device, characterized in that: include: A data acquisition unit, adapted to acquire production data information of each device in a target device group in a current production unit, wherein the target device group is composed of a group of devices having the same function; A prediction unit, adapted to predict, based on the acquired production data information, whether each device in the target device group needs to switch processes in the next production unit, the optimal value of the output of each device in the target device group in the next production unit, and the optimal value of the availability; A decision-making unit, adapted to calculate, based on the prediction result and the types of processes executable by the devices in the target device group, the output value of each device in the target device group when executing each process when the total output of the target device group is the maximum value by using a preset dispatch decision model; The types of processes include: processes that require process conversion and processes that do not require process conversion; The dispatching unit is suitable for dispatching products based on calculation results.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The computer program is executed by a processor to implement the steps of the method according to any one of claims 1 to 8.
11. An electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, wherein: When the processor runs the computer program, the steps of the method according to any one of claims 1 to 8 are performed.