Intelligent optimization scheduling system and method based on multi-objective production

Abstract

The present application discloses an intelligent optimization scheduling system and method based on multi-objective production, which relates to the field of intelligent scheduling optimization technology. It solves the technical problem that the existing technology often adopts fixed weights or static priorities to generate scheduling plans, which makes it difficult for the scheduling plans to adapt to changes in the production environment, resulting in low accuracy of the multi-objective production scheduling plans; by constructing a product comprehensive function according to product data, equipment data, order data and inventory data; generating a scheduling plan according to the product comprehensive function and performing production scheduling based on it; generating an alarm signal according to the equipment data, constructing a product comprehensive function for the production of multiple products, and combining multiple objective functions through adaptive weights, it can dynamically adjust the optimal state according to real-time production data, thereby improving the accuracy of the production scheduling plan, and at the same time monitoring the equipment and order status in real time, and can issue early warning signals in time to avoid the occurrence of equipment failure or order processing failure.

Description

Technical Field

[0001] The present application belongs to the field of intelligent scheduling optimization technology, specifically an intelligent optimization scheduling system and method based on multi-objective production. Background Art

[0002] Manufacturing production refers to the process of transforming raw materials, components, or semi-finished products into final products through a series of processes, including processing, assembly, and testing. This process involves multiple steps, including production planning, resource scheduling, equipment operation, quality control, and logistics management. The core goal of manufacturing production is to complete product production efficiently and with high quality, while optimizing resource utilization, reducing costs, and meeting customer needs. In modern industrial production, production scheduling is a key step in improving efficiency and reducing costs.

[0003] Traditional production scheduling methods are usually optimized based on a single objective, such as minimizing costs or maximizing output, and are difficult to cope with complex production environments with multiple objectives and multiple constraints. With the development of intelligent manufacturing, there are also existing technologies that have implemented scheduling solutions for multi-objective production, but they often use fixed weights or static priorities to generate scheduling solutions, making it difficult for scheduling solutions to adapt to changes in the production environment, resulting in low accuracy of multi-objective production scheduling solutions. Therefore, the scheduling system for multi-objective production still needs further improvement. Summary of the Invention

[0004] The present application aims to solve at least one of the technical problems existing in the prior art; to this end, the present application proposes an intelligent optimization scheduling system and method based on multi-objective production, which is used to solve the technical problem that the prior art often uses fixed weights or static priorities to generate scheduling plans, making it difficult for the scheduling plans to adapt to changes in the production environment, resulting in low accuracy of the multi-objective production scheduling plans.

[0005] To achieve the above-mentioned purpose, the first aspect of the present application provides an intelligent optimization scheduling system based on multi-objective production, comprising: a data acquisition module, a data analysis module, a scheduling module, an early warning module and a database;

[0006] The data acquisition module is used to acquire production data in real time through data acquisition equipment; the production data includes product data, equipment data, order data and inventory data;

[0007] The data analysis module: constructs a product comprehensive function based on product data, equipment data, order data and inventory data; generates a scheduling plan based on the product comprehensive function; and generates an alarm signal based on the equipment data;

[0008] The scheduling module performs production scheduling according to the scheduling plan;

[0009] The early warning module: makes prompts according to the alarm signal and contacts the management personnel.

[0010] Through the above steps, this application constructs a comprehensive product function for the production of multiple products. This function is composed of multiple objective functions through adaptive weights, and can dynamically adjust the optimal state according to real-time production data, thereby significantly improving the accuracy of the production scheduling plan; at the same time, the system monitors the equipment and order status in real time, and can issue early warning signals in time to effectively avoid situations where equipment cannot work normally or orders are not processed in a timely manner.

[0011] Furthermore, the product comprehensive function is constructed based on product data, equipment data, order data and inventory data, including:

[0012] Acquire product data, equipment data, order data, and inventory data; the product data includes product ID, production cost, and production time; the equipment data includes equipment ID, equipment status, and historical equipment operation data; the order data includes order ID, order product quantity, order deadline, order customer level, and order total; the inventory data includes inventory product ID and its corresponding inventory quantity;

[0013] Calculate order priority based on order period, order customer level and order total amount;

[0014] Generate several target quantitative functions based on production cost, production time, equipment status and order priority;

[0015] By formula Construct a product comprehensive function FC; where ω1, ω2, ω3 and ω4 are dynamic weight coefficients, ω1∈(α1, α2), ω2∈(α3, α4), ω3∈(α5, α6), ω4∈(α7, α8); α1, α2, α3, α4, α5, α6, α7 and α8 are constants, and α1, α2, α3, α4, α5, α6, α7 and α8∈(0, 1); C is a quantitative function for minimizing production cost, DC is unit cost; T is a quantitative function for minimizing production time, DT is unit time; U is a quantitative function for maximizing equipment utilization, DU is unit utilization; P is a quantitative function for maximizing order priority, DP is unit priority.

[0016] Furthermore, the calculation of order priority based on order term, order customer level and order total amount includes:

[0017] Obtain the order ID and its corresponding order term DQ, order customer level DKD and order total DZJ; the order customer level is assessed by experts based on the customer's identity;

[0018] By formula Calculate order priority DYj ; Among them, β1, β2 and β3 are weight coefficients, β1, β2 and β3∈(0,1); DT is unit time, DD is unit level, and DJ is unit price.

[0019] This application calculates the order priorities corresponding to all order IDs through the above steps, and gives priority to high-priority order IDs when building product synthesis functions to ensure that high-priority orders are processed first. This not only helps to improve customer satisfaction, but also further optimizes production efficiency and resource utilization, thereby achieving more efficient production scheduling.

[0020] Furthermore, the method generates several target quantitative functions based on production cost, production time, equipment status and order priority, including:

[0021] Obtain production costs, production time, equipment status, and order priorities for several products; the production costs include raw material costs Energy costs and labor costs

[0022] By formula Construct a quantitative function C that minimizes production costs; where SL i Expressed as the production quantity of the i-th product;

[0023] By the formula T=max{ST i}Construct a quantization function T that minimizes production time; where ST i It is expressed as the production time of the i-th product;

[0024] By formula Construct the quantitative function U that maximizes equipment utilization; where YT k It is represented as the running time of the kth device, and ZT is represented as the total running time of all devices;

[0025] By formula Construct the maximum order priority quantitative function P; where WZ n It represents the order completion status corresponding to the high-priority order ID; GYDS represents the total number of high-priority orders; the GYDS is generated according to the order priority.

[0026] Furthermore, the GYDS is generated by order priority, including:

[0027] Get the order priority DY corresponding to several order IDs j ;

[0028] Prioritize your order j Order IDs greater than the priority threshold are marked as high-priority orders;

[0029] GYDS is obtained by summing the quantities corresponding to the order IDs labeled as high-priority orders.

[0030] Furthermore, the dynamic weight coefficient is dynamically adjusted according to production data, including:

[0031] Obtain production data in real time;

[0032] Inputting production data into a weight estimation model to obtain a number of estimated dynamic weights; the weight estimation model is constructed using a machine learning model;

[0033] The dynamic weight coefficients are updated by assigning several estimated dynamic weights to their corresponding dynamic weight coefficients.

[0034] Furthermore, the weight estimation model is constructed through a machine learning model, including:

[0035] Obtaining a number of historical production data and their corresponding dynamic weights;

[0036] Divide a number of historical production data and their corresponding dynamic weights into training data, verification data, and test data; perform data preprocessing on the training data, verification data, and test data to obtain a training set, a verification set, and a test set;

[0037] Select a machine learning model as the base model;

[0038] Train the basic model using the training set, and adjust the learning rate and hyperparameters on the validation set to obtain the pre-trained model;

[0039] By verifying the pre-trained model on the test set, we finally get the input production data and output a weight estimation model that estimates several dynamic weights.

[0040] Furthermore, generating a scheduling plan based on the product comprehensive function includes:

[0041] Real-time acquisition of production data and product comprehensive functions;

[0042] The production data and product comprehensive function are used through the deep reinforcement learning model to obtain the optimal solution of the product comprehensive function, and thus generate a scheduling plan.

[0043] Furthermore, generating an alarm signal according to the device data includes:

[0044] Obtaining device data and several order priorities; the device data includes the device ID and the device's historical working data; the device's historical working data includes the device's historical working hours and the corresponding working environment temperature and humidity;

[0045] By formula Calculate the damage index SSZ of several equipment k ; Among them, β4 and β5 are weight coefficients, β4 and β5∈(0,1); g is the proportional coefficient, g∈(0,π / 2); LS km W is the historical working time of the kth device in the mth time period; km and S km They represent the operating temperature and humidity of the k-th device in the m-th time period; ZW and ZS represent the optimal operating temperature and humidity; DW is the unit temperature, DS is the unit humidity, and DGT is the unit working time;

[0046] Determine the equipment damage index SSZ k Is it greater than the damage threshold? If yes, generate an equipment scrap alarm signal; if no, determine the equipment damage index SSZ k Is it greater than D times the damage threshold? If yes, generate an equipment maintenance warning signal; if no, do nothing. Where D is the proportional coefficient, D∈(0,1);

[0047] Determine whether the order priority is greater than the emergency processing threshold; if so, generate an order emergency processing prompt signal; if not, do nothing.

[0048] This application monitors equipment data and order data in real time. When equipment damage reaches a certain threshold, the system will immediately issue an alarm signal so that management personnel can repair it in time to prevent a decrease in production efficiency due to equipment problems. At the same time, the system tracks order data in real time and gives priority to high-priority orders, thereby improving customer satisfaction and overall work efficiency.

[0049] Another aspect of the present invention provides an intelligent optimization scheduling method based on multi-objective production, comprising:

[0050] S0: Real-time acquisition of production data; the production data includes product data, equipment data, order data and inventory data;

[0051] S1: Construct product comprehensive function based on product data, equipment data, order data and inventory data;

[0052] S2: Generate a scheduling plan based on the product comprehensive function; perform production scheduling based on the scheduling plan;

[0053] S3: Generate an alarm signal based on the device data; make a prompt based on the alarm signal and contact the management personnel.

[0054] Compared with the prior art, the present invention has the following advantages:

[0055] 1. This application constructs a product comprehensive function based on product data, equipment data, order data and inventory data; generates a scheduling plan based on the product comprehensive function; performs production scheduling based on the scheduling plan; generates an alarm signal based on equipment data, and constructs a product comprehensive function for the production of multiple products. The product comprehensive function is composed of multiple objective functions combined by adaptive weights, and can dynamically adjust the optimal state according to real-time production data, thereby improving the accuracy of the production scheduling plan. At the same time, it monitors the equipment and order status in real time, and can issue early warning signals in time to avoid equipment malfunctioning or untimely order processing.

[0056] 2. This application dynamically adjusts the dynamic weight coefficients corresponding to each objective function through real-time production data, so that the weight coefficients corresponding to the objective function can be changed when the production data is in a specific situation, making the production data more in line with the pre-set expectations, thereby improving the flexibility and adaptability of multi-objective scheduling.

[0057] 3. This application conducts real-time monitoring through equipment data and order data. When the equipment damage reaches a certain level, an alarm signal is issued to promptly remind management personnel to carry out repairs, thereby avoiding the situation where production efficiency is reduced. At the same time, order data is monitored and high-priority order data is processed in a timely manner to improve customer satisfaction and work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0059] Figure 1 This is a schematic diagram of the principle of the intelligent optimization scheduling system based on multi-objective production of this application;

[0060] Figure 2 This is a flow chart of the intelligent optimization scheduling method based on multi-objective production in this application. DETAILED DESCRIPTION

[0061] The following will clearly and completely describe the technical solutions of this application in conjunction with the embodiments. Obviously, the embodiments described are only a part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0062] See also Figure 1, the first embodiment of the present application provides an intelligent optimization scheduling system based on multi-objective production, including: a data acquisition module, a data analysis module, a scheduling module, an early warning module and a database;

[0063] Data acquisition module: acquires production data in real time through data acquisition equipment; production data includes product data, equipment data, order data and inventory data; data acquisition equipment includes various sensors, etc.

[0064] Data Analysis Module: Builds a product synthesis function based on product data, equipment data, order data, and inventory data. This function is designed to determine the optimal scheduling method for the current state. It also generates a scheduling plan based on the product synthesis function. This plan is a solution designed to optimize production results based on the current production state. It also generates alarm signals based on equipment data.

[0065] Scheduling module: performs production scheduling according to the scheduling plan;

[0066] Early warning module: Prompts are issued based on alarm signals and management personnel are contacted. Alarm signals include equipment scrapping alarm signals and equipment maintenance alarm signals.

[0067] In this embodiment, a product comprehensive function is constructed based on product data, equipment data, order data, and inventory data, including:

[0068] Obtain product data, equipment data, order data, and inventory data; product data includes product ID, production cost, and production time; equipment data includes equipment ID, equipment status, and historical equipment operation data; order data includes order ID, order product quantity, order deadline, order customer level, and order total; inventory data includes inventory product ID and its corresponding inventory quantity;

[0069] Order priority is calculated based on order deadline, order customer level and order total. Order priority means the order will be supplied first. The higher the priority, the higher the priority, the higher the priority of the product required for production.

[0070] Generate several target quantitative functions based on production cost, production time, equipment status and order priority; the target quantitative functions are used to make the scheduling plan move towards a certain goal, such as optimizing cost, optimizing equipment utilization, etc.

[0071] By formula Construct a product comprehensive function FC; where ω1, ω2, ω3 and ω4 are dynamic weight coefficients, ω1∈(α1, α2), ω2∈(α3, α4), ω3∈(α5, α6), ω4∈(α7, α8); α1, α2, α3, α4, α5, α6, α7 and α8 are constants, and α1, α2, α3, α4, α5, α6, α7 and α8∈(0, 1). The specific values ​​are set according to experience. The setting of constants is to make the dynamic weight coefficients adjustable within a certain range to avoid the situation where the objective function corresponding to a certain dynamic weight coefficient does not work; C is a quantitative function for minimizing production cost, DC is unit cost; T is a quantitative function for minimizing production time, DT is unit time; U is a quantitative function for maximizing equipment utilization, DU is unit utilization; P is a quantitative function for maximizing order priority, DP is unit priority.

[0072] This embodiment constructs a product comprehensive function through the above steps, aiming to schedule multiple products with multiple objectives, ensure that the scheduling between multiple products meets the predetermined objectives, improve the performance of the multi-objective production system, and thus improve the company's interests.

[0073] In this embodiment, the order priority is calculated based on the order period, order customer level, and order total amount, including:

[0074] Obtain the order ID and its corresponding order period DQ, order customer level DKD and order total DZJ; the order customer level is assessed by experts based on the customer's identity;

[0075] By formula Calculate order priority DY j ; Among them, β1, β2 and β3 are weight coefficients, β1, β2 and β3∈(0,1), and the specific values ​​are set according to experience. When more factors of order term are considered, the corresponding weight coefficient can be set larger. When more factors of order customer level are considered, the corresponding weight coefficient can be set larger. In this implementation, the weight coefficient is set to β1>β2>β3; DT is unit time, DD is unit level, DJ is unit price, and the specific values ​​are set according to experience; the shorter the order term, the faster the order needs to be produced, the higher the order customer level, the more it is necessary to produce its corresponding multiple products as soon as possible, and the higher the total order amount, the priority can be given to producing its corresponding products first, so the order priority increases accordingly.

[0076] This embodiment calculates the order priorities corresponding to all order IDs through the above steps, takes high-priority order IDs into account when constructing the product synthesis function, and gives priority to high-priority order IDs, which can not only improve customer satisfaction but also optimize production efficiency and resource utilization.

[0077] In this embodiment, several target quantization functions are generated based on production cost, production time, equipment status, and order priority, including:

[0078] Get production costs, production time, equipment status, and order priority for several products; production costs include raw material costs Energy costs and labor costs

[0079] By formula Construct a quantitative function C that minimizes production costs; where SL i It is expressed as the production quantity of the i-th product; its purpose is to schedule the production process so that the production cost is kept at the lowest cost;

[0080] By the formula T=max{ST i}Construct a quantization function T that minimizes production time; where ST i It is the production time of the i-th product. There is a condition that the production time of the i-th product must be less than the order deadline corresponding to its order to ensure that the product can be completed within the order deadline.

[0081] By formula Construct the quantitative function U that maximizes equipment utilization; where YT k It is represented as the operating time of the kth device, and ZT is represented as the total operating time of all devices. Each device can only produce one product. The quantitative function of maximizing equipment utilization is to ensure that the equipment utilization reaches the maximum while completing orders and saving energy consumption;

[0082] By formula Construct the maximum order priority quantitative function P; where WZ n It represents the order completion status corresponding to the high-priority order ID; GYDS represents the total number of high-priority orders; GYDS is generated by order priority; maximizing the order priority quantization function is to give priority to high-priority orders and ensure that high-priority order products can be produced as soon as possible.

[0083] In this embodiment, GYDS is generated based on order priority, including:

[0084] Get the order priority DY corresponding to several order IDs j ;

[0085] Prioritize your order j Order IDs greater than the priority threshold are marked as high-priority orders; the priority threshold is set based on experience;

[0086] GYDS is obtained by summing the quantities corresponding to the order IDs labeled as high-priority orders.

[0087] The dynamic weight coefficient in this embodiment is dynamically adjusted according to production data, including:

[0088] Obtain production data in real time;

[0089] Inputting production data into a weight estimation model to obtain a number of estimated dynamic weights; the weight estimation model is constructed using a machine learning model;

[0090] The dynamic weight coefficients are updated by assigning several estimated dynamic weights and their corresponding dynamic weight coefficients. When estimating the dynamic weight coefficients, the dynamic weight coefficient corresponding to each objective function will only fluctuate within its fixed range and will not exceed its fixed range.

[0091] The weight estimation model in this embodiment is constructed through a machine learning model, including:

[0092] Obtaining a number of historical production data and their corresponding dynamic weights;

[0093] Divide a number of historical production data and their corresponding dynamic weights into training data, validation data, and test data; perform data preprocessing on the training data, validation data, and test data to obtain a training set, a validation set, and a test set; the ratio of the training set, the test set, and the validation set is 7:2:1;

[0094] Select a machine learning model as the basic model; machine learning models include neural network models, etc.

[0095] Train the basic model using the training set, and adjust the learning rate and hyperparameters on the validation set to obtain the pre-trained model;

[0096] By verifying the pre-trained model on the test set, we finally get the input production data and output a weight estimation model that estimates several dynamic weights.

[0097] This embodiment dynamically adjusts the dynamic weight coefficients corresponding to each objective function through real-time production data, so that it can flexibly adjust the weight distribution when the production data is in a specific situation, thereby making the production data closer to the pre-set expected value, which can improve the flexibility and adaptability of multi-objective scheduling and ensure that the production scheduling plan can better cope with complex and changing production environments.

[0098] In this embodiment, the scheduling plan is generated according to the product comprehensive function, including:

[0099] Real-time acquisition of production data and product comprehensive functions;

[0100] The production data and product comprehensive function are used through a deep reinforcement learning model to obtain the optimal solution of the product comprehensive function, and a scheduling plan is generated based on this. The deep reinforcement learning model is a pre-trained model that can effectively obtain the optimal solution of the product comprehensive function and generate a scheduling plan based on this.

[0101] This embodiment uses a deep reinforcement learning model to obtain the optimal solution of the product comprehensive function and generates a multi-objective production scheduling plan based on it, thereby improving the intelligence level of scheduling decision-making.

[0102] In this embodiment, generating an alarm signal based on device data includes:

[0103] Obtain device data and several order priorities; device data includes device ID and historical device operating data; historical device operating data includes historical device operating hours and the corresponding operating environment temperature and humidity;

[0104] By formula Calculate the damage index SSZ of several equipment k ; Among them, β4 and β5 are weight coefficients, β4 and β5∈(0,1), and the specific values ​​are set according to experience. When the damage caused by the temperature of the equipment during operation is greater than the humidity factor, β4>β5 can be used. In this embodiment, both β4 and β5 are set to 0.5, that is, the degree of influence of the two on the damage of the equipment within the equipment operation time range is consistent; g is the proportional coefficient, g∈(0,π / 2), and the specific values ​​are set according to experience. In this embodiment, g is set to make the equipment damage index SSZ k ∈(0,1);LS km W is the historical working time of the kth device in the mth time period; km and S km The values ​​DW and DGT are the operating temperature and humidity of the kth device in the mth time period. ZW and ZS represent the optimal operating temperature and humidity, respectively. DW represents the unit temperature, DS represents the unit humidity, and DGT represents the unit operating time. The specific values ​​are set based on experience. In this embodiment, DW can be set to 1°C, DS to 1%, and DGT to 1 day. When the device is not in operation, temperature and humidity have little impact on device damage, so the device damage index is 0. Once the device is in operation, temperature and humidity will have an adverse effect on device damage. Therefore, as the operating time increases and the temperature and humidity deviate further from the optimal operating temperature and humidity, the device damage index increases.

[0105] Determine the equipment damage index SSZ k Is it greater than the damage threshold? If yes, generate an equipment scrap alarm signal; if no, determine the equipment damage index SSZ kIs it greater than D times the damage threshold? If yes, generate an equipment maintenance warning signal; if no, do nothing. The damage threshold is set based on experience. Where D is the proportionality coefficient, D∈(0,1). The specific value is set based on experience. In this embodiment, D is set to 0.6.

[0106] Determine whether the order priority is greater than the emergency processing threshold; if so, generate an order emergency processing prompt signal; if not, do nothing; the emergency processing threshold is set based on experience.

[0107] See also Figure 2 Another aspect of the present application provides an intelligent optimization scheduling method based on multi-objective production, including:

[0108] S0: Real-time acquisition of production data; production data includes product data, equipment data, order data and inventory data;

[0109] S1: Construct product comprehensive function based on product data, equipment data, order data and inventory data;

[0110] S2: Generate a scheduling plan based on the product comprehensive function; perform production scheduling based on the scheduling plan;

[0111] S3: Generate an alarm signal based on the device data; make a prompt based on the alarm signal and contact the management personnel.

[0112] Some of the data in the above formula are calculated by removing the dimensions and taking their numerical values. The formula is a formula that is closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0113] The working principle of this application is: by acquiring production data in real time; production data includes product data, equipment data, order data and inventory data; constructing a product comprehensive function based on product data, equipment data, order data and inventory data; generating a scheduling plan based on the product comprehensive function; performing production scheduling according to the scheduling plan; generating an alarm signal based on equipment data; making a prompt based on the alarm signal and contacting the management personnel, constructing a product comprehensive function for the production of multiple products, and the product comprehensive function is composed of multiple objective functions through adaptive weights, which can dynamically adjust the optimal state according to real-time production data, thereby improving the accuracy of the production scheduling plan, and at the same time monitoring the equipment and order status in real time, being able to issue early warning signals in time to avoid the equipment from failing to work properly or the order from being processed in time, thereby avoiding the problem that the existing technology often uses fixed weights or static priorities to generate scheduling plans, making it difficult for the scheduling plans to adapt to changes in the production environment, resulting in low accuracy of multi-objective production scheduling plans.

[0114] The above embodiments are only used to illustrate the technical method of the present application and are not intended to limit it. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present application.

Claims

1. Intelligent optimization scheduling system based on multi-objective production, characterized by: include: Data acquisition module, data analysis module, scheduling module, early warning module and database; The data acquisition module is used to acquire production data in real time through data acquisition equipment; the production data includes product data, equipment data, order data and inventory data; The data analysis module: constructs a product comprehensive function based on product data, equipment data, order data and inventory data; generates a scheduling plan based on the product comprehensive function; and generates an alarm signal based on the equipment data; The scheduling module performs production scheduling according to the scheduling plan; The early warning module: makes prompts according to the alarm signal and contacts the management personnel; The product comprehensive function is constructed based on product data, equipment data, order data and inventory data, including: The product data includes product ID, production cost and production time; the equipment data includes equipment ID, equipment status and equipment historical working data; the order data includes order ID, order product quantity, order period, order customer level and order total; the inventory data includes inventory product ID and its corresponding inventory quantity; Calculate order priority based on order period, order customer level and order total amount; Generate several target quantitative functions based on production cost, production time, equipment status and order priority; By formula Construct product comprehensive function FC; among them, 、 、 and is the dynamic weight coefficient, ∈( , ), ∈( , ), ∈( , ), ∈( , ); 、 、 、 、 、 、 and is a constant, and 、 、 、 、 、 、 and ∈(0,1); C is the quantitative function for minimizing production cost, DC is the unit cost; T is the quantitative function for minimizing production time, DT is the unit time; U is the quantitative function for maximizing equipment utilization, DU is the unit utilization; P is the quantitative function for maximizing order priority, DP is the unit priority; The dynamic weight coefficient is dynamically adjusted according to production data, including: Obtain production data in real time; Inputting production data into a weight estimation model to obtain a number of estimated dynamic weights; the weight estimation model is constructed using a machine learning model; The dynamic weight coefficients are updated by assigning several estimated dynamic weights to their corresponding dynamic weight coefficients.

2. The intelligent optimization scheduling system based on multi-objective production according to claim 1 is characterized in that: The calculation of order priority based on order term, order customer level and order total amount includes: Obtain the order ID and its corresponding order term DQ, order customer level DKD and order total DZJ; the order customer level is assessed by experts based on the customer's identity; By formula Calculating order priority ;in, is the weight coefficient, ∈(0, 1); DT is unit time, DD is unit level, and DJ is unit price.

3. The intelligent optimization scheduling system based on multi-objective production according to claim 1 is characterized in that: The method generates several target quantitative functions based on production cost, production time, equipment status and order priority, including: Obtain production costs, production time, equipment status, and order priorities for several products; the production costs include raw material costs Energy costs and labor costs ; By formula Construct a quantitative function C that minimizes production costs; where, Expressed as the production quantity of the i-th product; By formula Construct a quantization function T that minimizes production time; where, It is expressed as the production time of the i-th product; By formula Construct a quantitative function U that maximizes equipment utilization; where, It is represented as the running time of the kth device, and ZT is represented as the total running time of all devices; By formula Construct a maximization order priority quantization function P; where, It represents the order completion status corresponding to the high-priority order ID; GYDS represents the total number of high-priority orders; the GYDS is generated according to the order priority.

4. The intelligent optimization scheduling system based on multi-objective production according to claim 3 is characterized in that: The GYDS is generated by order priority, including: Get the order priorities corresponding to several order IDs ; Prioritize your order Order IDs greater than the priority threshold are marked as high-priority orders; GYDS is obtained by summing the quantities corresponding to the order IDs labeled as high-priority orders.

5. The intelligent optimization scheduling system based on multi-objective production according to claim 1 is characterized in that: The weight estimation model is constructed through a machine learning model, including: Obtaining a number of historical production data and their corresponding dynamic weights; Divide a number of historical production data and their corresponding dynamic weights into training data, verification data, and test data; perform data preprocessing on the training data, verification data, and test data to obtain a training set, a verification set, and a test set; Select a machine learning model as the base model; Train the basic model using the training set, and adjust the learning rate and hyperparameters on the validation set to obtain the pre-trained model; By verifying the pre-trained model on the test set, we finally get the input production data and output a weight estimation model that estimates several dynamic weights.

6. The intelligent optimization scheduling system based on multi-objective production according to claim 1 is characterized in that: Generating a scheduling plan based on the product comprehensive function includes: Real-time acquisition of production data and product comprehensive functions; The production data and product comprehensive function are used through the deep reinforcement learning model to obtain the optimal solution of the product comprehensive function, and thus generate a scheduling plan.

7. The intelligent optimization scheduling system based on multi-objective production according to claim 1 is characterized in that: Generating an alarm signal according to the device data includes: Obtaining device data and several order priorities; the device data includes the device ID and the device's historical working data; the device's historical working data includes the device's historical working hours and the corresponding working environment temperature and humidity; By formula Calculate several equipment damage indices ;in, is the weight coefficient, ∈(0, 1); g is the proportional coefficient, g∈(0, π / 2); It is represented by the historical working hours of the k-th device in the m-th time period; and They represent the operating temperature and humidity of the k-th device in the m-th time period; ZW and ZS represent the optimal operating temperature and humidity; DW is the unit temperature, DS is the unit humidity, and DGT is the unit working time; Determine equipment damage index Is it greater than the damage threshold? If yes, generate an equipment scrap alarm signal; if no, determine the equipment damage index Is it greater than D times the damage threshold? If yes, generate an equipment maintenance warning signal; if no, do nothing. Where D is the proportional coefficient, D∈(0,1); Determine whether the order priority is greater than the emergency processing threshold; if so, generate an order emergency processing prompt signal; if not, do nothing.

8. An intelligent optimization scheduling method based on multi-objective production, applied to an intelligent optimization scheduling system based on multi-objective production according to any one of claims 1 to 7, characterized in that: include: S0: Real-time acquisition of production data; the production data includes product data, equipment data, order data and inventory data; S1: Construct product comprehensive function based on product data, equipment data, order data and inventory data; The product comprehensive function is constructed based on product data, equipment data, order data and inventory data, including: The product data includes product ID, production cost and production time; the equipment data includes equipment ID, equipment status and equipment historical working data; the order data includes order ID, order product quantity, order period, order customer level and order total; the inventory data includes inventory product ID and its corresponding inventory quantity; Calculate order priority based on order period, order customer level and order total amount; Generate several target quantitative functions based on production cost, production time, equipment status and order priority; By formula Construct product comprehensive function FC; among them, 、 、 and is the dynamic weight coefficient, ∈( , ), ∈( , ), ∈( , ), ∈( , ); 、 、 、 、 、 、 and is a constant, and 、 、 、 、 、 、 and ∈(0,1); C is the quantitative function for minimizing production cost, DC is the unit cost; T is the quantitative function for minimizing production time, DT is the unit time; U is the quantitative function for maximizing equipment utilization, DU is the unit utilization; P is the quantitative function for maximizing order priority, DP is the unit priority; The dynamic weight coefficient is dynamically adjusted according to production data, including: Obtain production data in real time; Inputting production data into a weight estimation model to obtain a number of estimated dynamic weights; the weight estimation model is constructed using a machine learning model; Assigning a number of estimated dynamic weights and their corresponding dynamic weight coefficients to update the dynamic weight coefficients; S2: Generate a scheduling plan based on the product comprehensive function; perform production scheduling based on the scheduling plan; S3: Generate an alarm signal based on the device data; make a prompt based on the alarm signal and contact the management personnel.

Images