TWM613466U - System for anticipating delivery time - Google Patents

Abstract

A system for anticipating a delivery time is provided. The system relies on production history data in an enterprise resource planning system, present data about materials and production equipment, and the data learning from the above data through a machine-learning algorithm to anticipate a delivery time. For example, the system can create a prediction model from the machine-learning algorithm for predicting the delivery time. The prediction model provides a weight-setting function that allows a decision maker to change production combination by modifying weights applied to parameters for material planning and schedule planning. Therefore, the system can be used to simulate a decision-making process so as to obtain an appropriate solution of production such as one or more production combinations achieving the delivery time.

Description

Intelligent delivery forecast system

The disclosure book discloses an intelligent system for predicting the delivery date, especially a system for predicting the delivery date based on historical data and production information through intelligent means.

In the product manufacturing process, companies need to prepare production lines for the needs of raw materials, production equipment and personnel. At the factory side, the delivery rate and production choices greatly affect the internal operation and financial status of the company. If the above key can be accurately controlled The subject will be able to effectively control the operation of the factory and improve the operation and financial status. However, the process of making products in the current factory is often caused by many uncontrollable factors and cannot deliver as expected, nor can it quickly decide how to respond when the production line changes.

Furthermore, in response to the trend of a small number of diversified and just-in-time production technologies (Just In Time, JIT), the delivery time of companies responding to customers is closely related to customer satisfaction. The current industry norm is a relatively short-term (such as 72 hours to 1 week) delivery time. The replies often cause problems for the front-line business.

In order to accurately provide customers with delivery dates, the disclosure paper proposes an intelligent delivery forecasting system. The intelligent delivery forecasting system includes software using computer technology and functional modules implemented with hardware to implement intelligent delivery forecasting methods. The intelligent delivery forecast system connects an enterprise resource planning system, a production system and a manufacturing execution system. The system includes a production scheduling simulation system. The manufacturing execution system collects instant messages on the process and receives data on the actual production line, and Obtain the data of production equipment, materials and products from the production system; including a computer system with a processor, through which the processor executes a program to achieve an artificial intelligence multi-objective optimization core, which uses one or more intelligent algorithms in an algorithm module After performing machine learning calculations and modeling on the data provided by the enterprise resource planning system, it provides scheduling parameters, material personnel planning parameters, and a variety of delivery planning combinations to the production scheduling simulation system. The production scheduling simulation system is based on the obtained data The calculated data is fed back to the core of artificial intelligence multi-objective optimization, providing multi-configuration results to optimize the establishment of a predictive model; the system includes a decision-making system implemented by a computer system, and multiple sets of product order delivery dates and production costs provided by the predictive model With batch production options, a final decision is made, including a delivery date.

Among them, it mainly provides a data collection module that extracts enterprise resource planning system (ERP) data in entities (such as companies, factories) realized by computers, and uses machine learning algorithms to learn the data obtained by the data collection module, and learn the data The relevance of various information to form a model training module for predicting the delivery date. When the data collection module receives a production demand data, it can predict the delivery date through the established forecasting model, and realize one or more production combinations on this delivery date to form a decision, and finally provide the company's boss decision.

Furthermore, the intelligent delivery forecasting system may also include other functional modules, such as a data cleaning module, used to clean the data obtained through the enterprise resource planning system to form data that meets the data format required for machine learning algorithm learning; It can also include an automatic optimization module. When the model training module uses one or more machine learning algorithms to learn data to obtain multiple prediction models, it can select through the automatic optimization module or form a better-performing prediction model by itself. , As a predictive model for predicting the delivery date; it can also include an automatic learning module. When new data is obtained through the data collection module, the model training module can use machine learning algorithms to learn new data, so that subsequent learning The new data derives an optimized prediction model for the correlation of various information.

Moreover, a weight setting function can be provided in the forecast model, allowing the decision maker of the entity to modify the weight of the parameters in the material planning and/or scheduling planning in the production combination to change the production sequence; and the system also includes a production combination The selection module uses another machine learning algorithm to learn the production combination selected by the decision maker of the entity in the past and modify the historical data of the weight of the parameters in each production combination to form a production combination selection model to provide the entity for production re-planning.

Preferably, the data obtained through the enterprise resource planning system is historical data of the entity’s past production combinations. The historical data at least includes at least one customer order and at least one factory shipment order, and the historical data is used to learn the entity’s past using a machine learning algorithm. The combination of production and a forecasting model for forecasting the delivery date. Furthermore, the data obtained through the enterprise resource planning system also includes the production data of the factory and the production time of the machines therein. The parameters of the material planning include the inventory of raw materials, components and/or semi-finished products required to produce a product; the parameters of the scheduling planning include the delivery ratio and quantity of the products produced according to the date.

In order to further understand the features and technical content of the present invention, please refer to the following detailed descriptions and drawings about the present invention. However, the drawings provided are only for reference and explanation, and are not used to limit the present invention.

The following are specific specific examples to illustrate the implementation of this creation, and those skilled in the art can understand the advantages and effects of this creation from the content disclosed in this specification. This creation can be implemented or applied through other different specific embodiments, and various details in this specification can also be based on different viewpoints and applications, and various modifications and changes can be made without departing from the concept of this creation. In addition, the drawings of this creation are merely schematic illustrations, and are not depicted in actual size, and are stated in advance. The following implementations will further describe the related technical content of this creation in detail, but the disclosed content is not intended to limit the scope of protection of this creation.

It should be understood that although terms such as "first", "second", and "third" may be used herein to describe various elements or signals, these elements or signals should not be limited by these terms. These terms are mainly used to distinguish one element from another, or one signal from another signal. In addition, the term "or" used in this document may include any one or a combination of more of the associated listed items depending on the actual situation.

In order to provide companies with an accurate delivery date before the product is manufactured, the manual discloses an intelligent delivery forecasting system, which uses machine learning (or with deep learning) to learn the production data provided by the company to generate forecasts for the delivery date. Model to simulate the decision-making process of the enterprise. In particular, in the process of simulating enterprise decision-making, historical data of the enterprise is used, and boundary conditions such as cost can be quoted without affecting the operation of the real factory. The method of Robotic Process Automation (RPA) is used for many times. Simulate, and then use machine learning to optimize the predictive model. Further, when a suitable solution for the final decision is obtained, a suitable safety stock solution can also be reversed.

According to the embodiment of the intelligent delivery forecasting system, the main body is a software solution running on a computer system. The system includes a computer system with one or more processors and memories. The processors execute programs to realize various modules, including Apply artificial intelligence (AI) multi-objective optimization core system, a parameterized production target decision-making system, and a production scheduling simulation system that can be customized for the production line. After the production parameters are given according to customer needs, the intelligent delivery forecast system can simulate the relationship between equipment, materials, semi-finished products and personnel in each process of the manufacturing process to calculate the completion time of each batch of products. The intelligent delivery forecast system can be realized by the Advanced Planning and Scheduling (APS). The computer system of the system also includes an interface to the enterprise resource planning system (Enterprise Resource Planning, ERP), which can obtain the financial, personnel, equipment, materials and other data of the enterprise, as well as the manufacturing execution system (Manufacturing Execution System, MES). The interface is used to read historical production data, and the intelligent delivery forecast system can actively initiate execution commands.

You can refer to the conceptual diagram of the operation of the intelligent delivery forecasting system proposed in the disclosure book shown in Figure 1. This figure shows that the intelligent delivery forecasting system is based on intelligent means as the core and can build a model based on historical data 101 obtained from the enterprise, which proposes the system Model 103 includes delivery model 131 and capacity model 132. In other words, the intelligent delivery forecasting system can predict the delivery date (ie delivery date) based on customer needs and the company’s resources, as well as estimate the production capacity and Production related parameters.

Under the concept of the operation of the intelligent delivery forecasting system, the delivery model 131 obtains an expected delivery date (that is, the number of days 105) according to the information provided by the enterprise, and can also include the delivery probability of the delivery date. The production capacity model 132 is based on the production combination selection of past historical data. Let the machine learn the production combination selected by the decision maker in the past. Related production data such as product classification, including existing products, existing quantities, quantity changes, similar new products, etc., propose a set of production standards 107. Then, in accordance with the batch time 109 in production and the production rule 111, the decision maker makes a production selection 113, and finally produces a production selection suggestion 115.

It is mentioned here that when the system provides a production selection suggestion 115, it can let enterprise decision-makers (such as business, boss, etc.) know that the demand for new orders may affect the production of existing orders, and make them forecast and produce according to the delivery date. The combination determines the delivery date of the order to be received, the various possibilities of materials and substitution, and finally allows the decision maker to choose the final solution.

Figure 2 shows an example of the architecture of the intelligent delivery forecasting system.

The figure shows a schematic diagram of the system architecture that realizes the delivery date forecasting system 20 and its surroundings to provide data. The intelligent delivery date forecasting system 20 is connected to an enterprise resource planning system 24, a production system 25, and a manufacturing execution system 27. One of them is the enterprise resource planning system 24, which is a computer system that is traditionally used for resource planning in enterprises. It connects the management servers of various departments in the enterprise, manages and records various resources and information in a unified manner, and records the sales and production of various levels of the enterprise. Purchase inventory and other data to form a huge historical production database, and these data can be used as the learned data in the intelligent delivery forecasting system 20 to form a forecasting model.

The delivery forecasting system 20 includes functional modules implemented by computer technology, and the processor of the computer system executes software programs to implement an artificial intelligence multi-objective optimization core 21, a production scheduling simulation system 22, and a decision-making system 23. The artificial intelligence multi-objective optimization core 21 provides a variety of intelligent means implemented by computer software and hardware, such as one or more intelligent algorithms described in the algorithm module 211, which indicates that there are several modules, such as genetic algorithm A, Particle swarm optimization algorithm B and decision tree algorithm C, but the actual operation is not limited by the several algorithms shown in the figure.

The algorithm module 211 represents several intelligent algorithms applicable to the system, such as genetic algorithm (GA) A. Genetic algorithm A is a computer simulation method that can be used in the system to simulate production time scheduling, which can solve Problems in the actual production process.

The algorithm module 211 proposes a Particle Swarm Optimization (PSO) algorithm B. The particle swarm optimization algorithm B is used in the delivery forecast system 20 to obtain the best solution to provide decision-making based on historical data, especially for production The optimization problem during the purchase of raw materials simulates the best solution.

The algorithm module 211 may include a decision tree algorithm (RF, XGB) C. The decision tree algorithm C is a machine learning algorithm that uses classification and regression methods to sample and randomly select features from the data to obtain the data features. , Can simulate the results of decision-making, the purpose is to produce a variety of decision-making directions to provide decision-makers.

After one or more intelligent algorithms in the algorithm module 211 perform machine learning calculations and modeling on the data provided by the enterprise resource planning system 24, it can provide scheduling parameters, material personnel planning parameters, and multiple delivery planning combinations to intelligent delivery. The production scheduling simulation system 22 in the forecasting system 20. The production scheduling simulation system 22 is also used to collect instant messages on the manufacturing process by adding a self-manufacturing execution system 27 to receive data on the actual production line. The production scheduling simulation system 22 also obtains data such as production equipment 251, materials 252, and products 253 in the production system 25.

In this way, the production scheduling simulation system 22 can calculate the production-related schedules, raw materials, delivery schedules, actual historical production data, and the real data on the production line calculated based on historical data, and the data generated by the production scheduling simulation system 22 It can feed back to the artificial intelligence multi-objective optimization core 21, provide multi-configuration results to optimize the prediction model established therein, and provide the decision-making system through the prediction model 23 multi-group product order delivery, production costs and batch production options and other multiple solutions , The decision-making system 23 is a functional module implemented by the computer system, providing multiple solutions to the business executive to make the final decision, including the delivery date provided to the business end 29 to the customer, and the provision of the manufacturing execution system 27 production resources and resources Production instructions to carry out production. When the decision system 23 makes the final delivery date judgment, it can also retrospectively provide the best solution for the inventory that the enterprise should have.

Through the intelligent delivery forecasting system 20, business decision makers can obtain simulated delivery and production information. One of the main purposes is to provide the business end 29 that can quickly and relatively accurately respond to the customer’s expected delivery date, and establish it according to it. The predictive model of the company extends the production portfolio selection model, which can assist factory managers to quickly determine how to plan production and estimate the possibility of cost changes under multiple choices.

FIG. 3 shows various functional means of the intelligent delivery forecasting system using a computer system with software means to realize the intelligent delivery forecasting system 20, and the various functional means are described by the functional modules shown in the figure.

The main intelligent means in the intelligent delivery forecasting system 20 include the data collection module 301. The way to achieve data collection is to obtain data from an entity (such as a company, a factory) through a computer system through a wired or wireless communication method through a network or a specific connection. Data, such as the data provided by the enterprise resource planning system (ERP). The data is mainly the historical data of the entity's past production combination, and it can also include the production data of the factory and the production time of the machines in it, which can then be established in the intelligent delivery forecast system 20 For the customized database of this company, the model training module 307 uses a machine learning algorithm to learn the collected data. The data is coded so that the machine learning algorithm finds association rules based on the provided data, and learns the data. The relevance of various information can finally form a forecast model for predicting the delivery date. Among them, when the data of the production demand is received by the communication means of the data collection module 301, the delivery date can be predicted through the predictive model, and one or more production combinations that realize the delivery date can be used to provide enterprise decision makers with decision making.

According to the embodiment of the intelligent delivery forecasting system 20, it may also include other data processing and learning software combined with hardware to achieve functional modules, such as a data cleaning module 303. The implemented data cleaning means is used to clean the data through the enterprise resource planning system. The acquired data is used to form data in a data format required by the machine learning algorithm for learning, so that subsequent methods can use this format as data for model training.

The intelligent delivery forecasting system 20 may also include an automatic optimization module 309. When the model training module 307 learns data through one or more machine learning algorithms to obtain multiple prediction models, the automatic optimization module 309 can achieve The software means compares various models with each other, selects or composes a better-performing forecasting model by itself, as a forecasting model for forecasting the delivery date.

When new data is obtained through the above-mentioned data collection module 301, the system uses the automatic learning module 305 to learn the new data with the machine learning algorithm for the model training module 307, and automatically extracts new production processes and learns new data. , In particular, the relevance of various information can be obtained based on learning new data, and fine-tuned to optimize the prediction model.

The intelligent delivery forecasting system 20 can be further provided with a production combination selection module 311. This is after the intelligent delivery forecasting system 20 obtains the forecast delivery date, the software method of production combination selection can be used to learn from another machine learning algorithm. The historical data of the production combination selected by the decision maker of the entity in the past and the weight of the parameters in each production combination are revised, and then a production combination selection model is formed to provide the entity for a production re-planning. According to an embodiment, the formed production combination selection model is used to propose a suggested combination of physical partial delivery or relocation, forming a new production combination of new material planning and/or new scheduling planning.

According to a contextual application applicable to the intelligent delivery forecasting system 20. When a company performs production, the sales personnel need to reply to the expected delivery date of the customer's batch when the customer places an order in the factory. However, it is common that the sales personnel need to communicate with the production line personnel and the material management personnel and confirm the current factory capacity status, and give them based on experience. The pre-delivery date, but this pre-delivery date is often much different from the actual delivery date. Therefore, the intelligent delivery forecasting system 20 proposed in the disclosure uses the intelligent means of the above-mentioned functional modules to obtain historical data from the enterprise, such as customer orders, factory shipment orders, etc., and then use machine learning algorithms to learn the data and train Develop a forecasting model to forecast the delivery of existing products. And because this prediction model refers to past data, it can deal with new products or a small number of products that lack complete information in the future through intelligent means. The main method is to collect the company's production information and machine production time and other data, such as material advance Time, machine capacity and parameters, working time between processes, etc., learn to meet the needs of the company's delivery date and production combination.

Next, FIG. 4 shows a flowchart of an embodiment of the intelligent delivery forecasting method executed by the intelligent delivery forecasting system, wherein the description of the action can be referred to FIG. 3.

The intelligent delivery forecast method is implemented in a computer system, and various learning algorithms are realized through the processing circuit, memory, and database in the computer system. The process includes, at the beginning, the computer system obtains the enterprise resource plan through the internal or external network The data of the system includes at least one customer order and at least one factory shipment (step S401), and the data is cleaned by software means, mainly to clean the data obtained through the enterprise resource planning system to form a learning algorithm that conforms to the machine learning algorithm. The data in the required data format can also be filtered for unnecessary, invalid or performance-affecting data (step S403), and then start to use machine learning algorithms to learn the information relevance in the data (step S405) to form a predictive model ( Step S407). Among them, if multiple prediction models are obtained by learning data through one or more machine learning algorithms, the computer system can also be used to select or form a better-performing prediction model as a prediction model for predicting delivery dates. After that, according to an embodiment, if new data is generated, the machine learning algorithm can be used to continue to learn the new data, and the relevance of various information can be obtained based on the learned new data, which is fed back to the system to optimize the prediction model.

After the prediction model is completed, the production data is received (step S409). For example, the business end receives data such as customer orders, product items, unit quantities, and expected delivery dates for orders, and can accept multiple orders at the same time. The details include the materials, procedures, semi-finished products and finished products required for the manufacture of each product item. These data can also be derived from the enterprise resource planning system (ERP) and the manufacturing execution system (MES).

Then the weight can be set (step S411). The predictive model provides a weight setting function, which provides decision makers to adjust production parameters. In this step, the decision makers of the entity can be allowed to modify the material planning and/or scheduling planning in the production mix. The weight of the parameter changes the production sequence. For example, if this customer is more important than other customers, or if this semi-finished product is a key component of many other products, these parameters that have not been considered by the system can be adjusted to increase their priority by adjusting their weights.

Finally, the delivery date, production combination, etc. are obtained through the predictive model (step S413), where each production combination can record material planning and scheduling planning, and/or machine production time, and the parameters of material planning can include production one. The inventory of raw materials, components and/or semi-finished products required by the product, and the parameters of the scheduling plan can include the delivery ratio and quantity of the product produced according to the date. In one embodiment, afterwards, another machine learning algorithm can be used to learn the production combination selected by the decision maker of the entity in the past, and to modify the historical data of the weight of the parameters in each production combination to form a production combination selection model , Provide entities for a production redrawing.

For example, taking aluminum forging as an example, forging is divided into two to several processes, such as rough forging and precision forging. Rough forging and precision forging use different forging equipment, and some equipment can also be used in different processes. The raw material for forging is aluminum. There are also molds that can be reused many times. The semi-finished product forged is the rough blank after rough forging, and the finished product is the aluminum forging formed by precision forging. The man-hours required for each process, the required staffing, and the required raw materials , The data of finished products, scraps and mass production yields, as well as the statistical data of its dependence on different equipment, are imported from the above-mentioned data sources. If the forecast target is an item that has been executed by the system, it can also be the same The historical data of the item is directly brought in, and the final complete production model data is confirmed by the engineering and production management personnel.

Fig. 5 shows a flowchart of another embodiment of the intelligent delivery forecasting method, which expresses the processing procedure of the proposed intelligent delivery forecasting system for existing products and new products. The first part is aimed at the company’s existing products. In the method, first obtain the existing product information (step S501), perform data processing, including data cleaning and filtering, with the purpose of obtaining modelable data (step S503), and then use the machine The learning algorithm learns the characteristics and relevance of the data (step S505), and establishes a predictive model (step S507). On the other hand, for new products, obtain new product information (step S509), including the introduction of feed time, machine parameters and production time, process procedures and other data (step S511), and also through data processing to obtain modelable After the data (step S513), machine learning algorithms are used to learn the features and associations in the new data (step S515) to establish a prediction model (step S517).

What is mentioned here is that for new products, this part often lacks the complete content as provided by the enterprise resource planning system. Therefore, similar raw materials, machines, equipment, and similar manufacturing methods can be used as related learning targets. , After the data is cleaned, a forecasting model can be established for each project and integrated into a forecasting model.

It can be seen that the system has formed a predictive model for both existing products and new products, and can obtain multiple predictive models through multiple machine learning algorithms. In step S519, the system can continue to use the model algorithm to obtain respective results , And then select one of the prediction models (step S521), the prediction model is used to generate the delivery date (step S525), similarly, the newly generated data will continue to automatically optimize the model (step S523).

The system architecture that implements the above method flow can continue to refer to the schematic diagram of the implementation of the delivery model and the production capacity model of the intelligent delivery forecasting method shown in Figure 6. The architecture of this figure covers the delivery and production standard dual system, the upper half of the figure The department schematically shows a delivery model architecture 61, which is concerned with time data 611. According to customer needs obtained by the enterprise resource planning system, order information 612, sales information 613, and other information 614 related to personnel in the enterprise are obtained for data cleaning 615, and subsequent steps such as modeling and training 616, finally produce a prediction model 617 that provides delivery time.

On the other hand, the capacity model architecture 62 described in the following half, capacity includes the lead time 621 that is not easy to change, and the parameter model 622 required for production. Similarly, data cleaning (623,624), modeling, and training (625,626) In other steps, a pre-processing model 627 for pre-processing and a parameter model 628 for processing production parameters are established. In this way, material planning and scheduling 629 are started, and after the above-mentioned prediction model is imported, a delivery time 630 is generated.

FIG. 7 then schematically shows an embodiment of the details of processing new and old product data in the intelligent delivery forecasting method, and how to establish data through feedback and an embodiment of an optimization model.

According to the figure, for the existing product 71, the enterprise resource planning system (ERP) provides factory data 711, including shipment orders 712 and orders 713. After data cleaning, modelable data 714 is obtained, and learning and Train to establish model 715. The predictive model is used to predict the delivery date 716 until the system gets the real delivery date 717. The predicted delivery date and actual delivery date formed by this part can be fed back to the enterprise resource planning system and become the existing Part of the factory data 711 in the product 71.

For the new product 72, after the customer opens the specifications, the product data 721 is obtained. It also includes various in-plant data 722 required for production, such as machine 723, input and output 724, and process 725. After data cleaning, the data is obtained Model data (726, 727, 728), and then use machine learning algorithms to learn the data of each project, and build models for each project (729, 730, 731). Through the software means of the system, each project model can be integrated, Establish an integrated model 732, which can propose a new product delivery forecast 733 for new products. When the real delivery date 734 is obtained, the new product delivery forecast 733 and the real delivery date 734 will form the new product data in the system, which is also the future optimization model in accordance with.

Finally, the delivery date predicted by the system will also be compared with the actual delivery date. For related procedures, please refer to the example diagram of using enterprise data in the intelligent delivery date forecasting method shown in FIG. 8.

The figure shows that the order data that the intelligent delivery date forecasting system can obtain from the enterprise resource planning system 80 includes orders 801 with accurate pre-delivery dates, shipment orders 802 with accurate pre-delivery dates, orders 803 with inaccurate pre-delivery dates, and pre-delivery dates. For the inaccurate shipment order 804, the accurate and inaccurate data on the pre-delivery date are separately data happy, and the modelable data (805, 807) are obtained respectively, and the forecast model 806 is established. After the delivery date is predicted, it can be compared to the real The delivery date 808 is used as the basis for confirming the model 809.

According to the above embodiment, the intelligent delivery forecasting system adopts the data provided by the artificial intelligence multi-objective optimization core (refer to Figure 2, 21), which can provide a production scheduling simulation system (refer to Figure 2, 22) to construct a production model , The production model considers the collaborative work of the production line to produce multiple items and batches at the same time. Among them, the allocation and use of equipment and raw materials, the priority of each item batch using limited equipment resources, and the splitting or merging of production orders Financial planning such as the time and cost of raw material acquisition are all adjustable parameters for multi-objective optimization.

For example, in the aluminum forging process, the preheating and pressure forging time required for different workpieces may be different. At the same time, it is necessary to consider factors such as the proficiency of the operator and the yield rate of the equipment processing in each process. The process parameters and historical data generated by the process can be used to obtain statistical models of process time and materials. According to the embodiment, using the worst case, typical case, and best case (Worst Case, Typical Case, Best Case) analysis of the statistical model, or using more complete statistical sampling, the Monte Carlo method (MC) can be calculated in parallel for simulation .

The model built by the core of the multi-objective optimization core can generate multiple sets of adjustable production parameters. After being brought into the production scheduling simulation system for simulation, a parallel calculation method can simulate a large number of parameters at the same time to obtain an optimized solution for delivery and finance. . Related artificial intelligence machine learning algorithms, such as the genetic algorithm (GA), particle swarm optimization (PSO), and decision tree algorithm (RF, GB, XGB) mentioned in Figure 2, can be used in a high-dimensional wide-area parameter space Continuous iteration, converging to the best parameter combination with high efficiency and precision.

The intelligent delivery forecasting system finally forecasts the delivery date after the updated model will get closer and closer to the actual delivery date, so that the delivery date provided by the intelligent delivery forecasting system can be an accurate basis for the business end to negotiate the order delivery date with the customer , And the production parameters that can be obtained at the same time, after confirmation by the engineering and production management personnel, based on the interface of the intelligent delivery forecast system, the enterprise resource planning system, and the manufacturing execution system, you can directly order raw materials and execute production orders, and then Ensure that the delivery date of the final completed order matches the delivery date of the forecast plan.

To sum up, the above embodiments describe an intelligent delivery forecasting system. The proposed intelligent delivery forecasting system first obtains data provided by the enterprise, such as historical data of past production and current data related to materials and production equipment. The method learns data to form a predictive model for predicting the delivery date, thereby simulating the decision-making process, and providing suitable solutions for the enterprise, and finally the decision-maker makes production planning and delivery judgments. In this way, the intelligent delivery forecasting system can provide business decision-makers and business executives with answering the need to meet customer orders when production capacity is tight, and to decide the priority order without considering the short-term labor and the addition of machinery and equipment, and provide a comparison between moving orders and alternative materials. The most suitable solution.

The content disclosed above is only a preferred and feasible embodiment of the present model, and does not limit the scope of the patent application of the present model. Therefore, all equivalent technical changes made by using the description and schematic content of the present model are included in the application of the present model. Within the scope of the patent.

101: historical data 103: System Model 131: Delivery Model 132: Capacity Model 105: days 107: Standard 109: Batch time 111: Production Rules 113: Production Selection 115: Production Selection Suggestions 24: Enterprise Resource Planning System 20: Delivery forecast system 21: Artificial intelligence multi-objective optimization core 211: Algorithm module A: Genetic algorithm B: particle swarm optimization algorithm C: Decision tree algorithm 22: Production scheduling simulation system 23: Decision System 25: Production system 251: Production Equipment 252: Materials 253: Products 27: Manufacturing Execution System 29: Business end 301: Data Collection Module 303: data cleaning module 305: Automatic learning module 307: Model Training Module 309: Automatic optimization module 311: Production Combination Selection Module 61: Delivery model architecture 611: Time Data 612: order information 613: Sales Information 614: Additional Information 615: data cleaning 616: Modeling, Training 617: Delivery time prediction model 62: Capacity model architecture 621: lead time 622: Parametric Model 623,624: Data cleaning 625,626: Modeling, training 627: front model 628: Parametric Model 629: Planning materials and scheduling 630: delivery time 71: Existing products 711: Factory Data 712: Shipment 713: order 714: Modelable data 715: Model 716: Delivery Forecast 717: actual delivery date 72: new products 721: product data 722: In-plant data 723: Machine 724: Incoming Material 725: process 726, 727, 728: Modelable data 729, 730, 731: Build a model 732: Establish a unified model 733: New product delivery forecast 734: actual delivery date 80: Enterprise Resource Planning System 801: Orders with accurate pre-delivery date 802: Shipment orders with accurate pre-delivery date 803: Orders with inaccurate pre-delivery date 804: Shipment orders with inaccurate pre-delivery date 805, 807: Modelable data 806: Build a predictive model 808: actual delivery date 809: Confirm Model Steps S401 to S413: One of the processes of the embodiment of intelligent delivery forecasting Steps S501~S525: Second process of the embodiment of intelligent delivery forecasting

Figure 1 shows the conceptual diagram of the operation of the intelligent delivery forecast system;

Figure 2 shows an embodiment diagram of the architecture of the intelligent delivery forecasting system;

Figure 3 shows an embodiment diagram of functional modules of the intelligent delivery forecasting system;

Fig. 4 shows a flow chart of one of the embodiments of the method executed by the intelligent delivery date forecasting system;

FIG. 5 shows a flowchart of the second embodiment of the method executed by the intelligent delivery date forecasting system;

FIG. 6 shows a schematic diagram of an embodiment of the architecture of a delivery model and a production capacity model that implements an intelligent delivery forecast method;

FIG. 7 shows a schematic diagram of an embodiment of processing product data in an intelligent delivery forecast method; and

Fig. 8 shows an embodiment diagram of using enterprise data in the intelligent delivery date forecasting method.

24: Enterprise Resource Planning System

20: Delivery forecast system

21: Artificial intelligence multi-objective optimization core

211: Algorithm module

A: Genetic algorithm

B: particle swarm optimization algorithm

C: Decision tree algorithm

22: Production scheduling simulation system

23: Decision System

25: Production system

251: Production Equipment

252: Materials

253: Products

27: Manufacturing Execution System

29: Business end

Claims (10)

An intelligent delivery forecasting system that connects an enterprise resource planning system, a production system and a manufacturing execution system, including: a production scheduling simulation system, a computer system and a processor executing a software program to realize the production scheduling simulation system , Collect the instant messages on the manufacturing process and receive the data on the actual production line from the manufacturing execution system, and obtain the data of production equipment, materials and products from the production system; an artificial intelligence multi-objective optimization core, by the computer system The processor executes a software program to realize the artificial intelligence multi-objective optimization core. One or more intelligent algorithms in an algorithm module are used to perform machine learning calculations and modeling on the data provided by the enterprise resource planning system to provide scheduling parameters , Material personnel planning parameters and a variety of delivery planning are combined into the production scheduling simulation system, where the production scheduling simulation system calculates the data based on the obtained data and feeds it back to the artificial intelligence multi-objective optimization core, providing multiple configurations As a result, a predictive model is optimized and established. The enterprise resource planning system connects to the management servers of various departments in an enterprise, manages and records various resources and information uniformly, records the data of the enterprise’s sales, production and purchase inventory, and forms a historical production database as The data learned in the intelligent delivery forecasting system is used to form the forecasting model; and a decision-making system in which the processor of the computer system executes a software program to implement the decision-making system, according to the multiple sets of product orders provided by the forecasting model Make a final decision, including a delivery date, for the period, production cost, and batch production options. The intelligent delivery forecasting system according to claim 1, wherein the module for realizing the intelligent delivery forecasting system by the processor of the computer system executing a software program further includes: a data collection module for extracting the Number of enterprise resource planning systems According to; a model training module, using a machine learning algorithm to learn the data obtained by the data collection module, learning the correlation of various information in the data, to form the prediction model for predicting the delivery date; a data cleaning model Group, used to clean the data obtained through the enterprise resource planning system to form data conforming to the data format required by the machine learning algorithm learning; and an automatic optimization module, when the model training module passes one or more When the machine learning algorithm learns data to derive multiple prediction models, the automatic optimization module selects or forms a better-performing prediction model as the prediction model for predicting the delivery date; wherein, when the data collection module receives When the data of a production demand is reached, the delivery date is predicted through the prediction model, and one or more production combinations that realize the delivery date. The intelligent delivery forecast system according to claim 2, which further includes an automatic learning module, when new data is obtained through the data collection module, the model training module is made to learn the new data with the machine learning algorithm. Optimize the prediction model based on the relevance of various information obtained by learning the new data. The intelligent delivery forecast system according to claim 2, wherein the data obtained through the enterprise resource planning system is historical data of the entity’s past production combinations, and the historical data includes at least one customer order and at least one factory shipment order, The machine learning algorithm learns the past production combination of the entity through the historical data, and obtains the prediction model for predicting the delivery date. The intelligent delivery forecast system according to claim 4, wherein the data obtained through the enterprise resource planning system also includes the production data of the factory and the production time of the machines therein. The intelligent delivery forecast system according to any one of claims 1 to 5, wherein each The production portfolio records material planning and scheduling planning, and/or machine production time. The intelligent delivery forecast system according to claim 6, wherein the parameters of the material planning include the inventory of raw materials, components and/or semi-finished products required to produce a product; the parameters of the scheduling planning include the production of the product according to the date Delivery ratio and quantity. The intelligent delivery forecast system according to claim 7, wherein the forecast model provides a weight setting function to allow the decision maker of the entity to modify the weight of the parameter in the material plan and/or the schedule plan in the production combination Change the production sequence. The intelligent delivery forecast system according to claim 8, which further includes a production combination selection module, wherein another machine learning algorithm is used to learn the production combination selected by the decision maker of the entity in the past and to modify the parameters in each production combination The historical data of the weights forms a production portfolio selection model, which provides the entity for a production re-planning. The intelligent delivery forecast system according to claim 9, wherein the production combination selection model is used to propose a suggested combination of partial delivery or relocation of the entity to form a new material plan and/or a new schedule A new production mix planned.

Images