Automatic production line task scheduling and optimizing method and system based on artificial intelligence

Abstract

The invention belongs to the technical field of production line task scheduling and optimization, and discloses an automatic production line task scheduling and optimization method and system based on artificial intelligence, and the method comprises the following steps: collecting historical production data of each production line, the historical production data comprising production comprehensive data and production efficiency, the production comprehensive data comprises an equipment state, a yield, a product qualification rate, work order completion time and a raw material supply rate; and training a machine learning model for predicting the production efficiency based on the production comprehensive data, collecting real-time production comprehensive data, and inputting the real-time production comprehensive data into the trained machine learning model to predict the production efficiency. By comprehensively considering the production efficiency of the production line, the fault condition of the production equipment, the stability of the production line and other factors, the production tasks are reasonably distributed and scheduled, so that the stable operation of the production line is ensured, and meanwhile, the production efficiency is maximized.

Description

Technical Field

[0001] The present invention belongs to the technical field of production line task scheduling and optimization, and specifically relates to an automated production line task scheduling and optimization method and system based on artificial intelligence. Background Art

[0002] With the in-depth development of industrialization, automated production lines have become an important means for manufacturing companies to improve production efficiency and reduce production costs. However, traditional production line scheduling methods often rely on manual experience and fixed rules, which are difficult to cope with complex and changing production environments and demands. Therefore, the introduction of artificial intelligence technology to intelligently schedule and optimize production line tasks has become a hot topic in current research.

[0003] For example, a Chinese patent with authorization announcement number CN113780747B discloses a system and method for automatically dividing, grasping and scheduling production line tasks, including a production line system that can operate independently and a workstation system that can operate independently, the production line system and the workstation system are communicatively connected; and including a process model construction module arranged in the production line system, for constructing the processes required for product production; a production task grasping module arranged in the workstation system, for obtaining process information required for product production; a production task division module arranged in the production line system, for dividing the production task into multiple subtasks according to the process; a scheduling task grasping module arranged in the workstation system, for grasping subtasks and executing them until all production tasks are completed, etc.; the invention optimizes the production line task division and scheduling process, improves the efficiency of the production line task process, realizes flexible production, lean production, and reduces costs.

[0004] For another example, a Chinese patent with publication number CN112686471A discloses a method for optimizing and scheduling steel production based on a greedy strategy, which includes the following steps: S1: obtaining the total production task from the production purchase and sales contract, and then decomposing each production task into a number of production subtasks according to the production process; S2: generating a production line information table based on the relevant information of the workshop production line, and each production line completes one process; S3: creating a structure array T and a process information array W, wherein the structure array T is a collection of decomposed production subtasks, and the process information array W stores the process information, and establishes a connection between the production subtask and the corresponding process information in sequence according to the processing order, so as to find the predecessor of the production subtask; S4: adopting a greedy algorithm, preferentially selecting the production subtask that makes the production line most efficient to be dispatched to each production line.

[0005] The above solution still has the following problems:

[0006] When scheduling the production tasks of the production line, the specific conditions of the production line are not taken into consideration, such as the production efficiency of the production line, the failure of the production equipment, the stability of the production line, etc., which will affect the scheduling and optimization of the production tasks. The present invention aims to reasonably allocate and schedule the production tasks through comprehensive consideration of various factors to maximize production efficiency. Summary of the invention

[0007] The purpose of the present invention is to provide an automated production line task scheduling and optimization method and system based on artificial intelligence, so as to solve the technical problems raised in the existing background technology.

[0008] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solution: an automated production line task scheduling and optimization method based on artificial intelligence, comprising the following steps:

[0009] Collect historical production data of each production line. Historical production data includes comprehensive production data and production efficiency. Comprehensive production data includes equipment status, output, product qualification rate, work order completion time, and raw material supply rate.

[0010] Based on comprehensive production data, train a machine learning model to predict production efficiency. Collect real-time comprehensive production data and input it into the trained machine learning model to predict production efficiency.

[0011] Compare and analyze the predicted production efficiency with the preset production efficiency threshold, determine and generate the production efficiency level, and allocate production tasks according to the production efficiency level;

[0012] Obtaining historical comprehensive data of equipment on each production line, the historical comprehensive data including comprehensive parameters of equipment operation status and equipment failure rate, and predicting the equipment failure rate at a future time based on the comprehensive parameters of equipment operation status;

[0013] Compare and analyze the predicted equipment failure rate at a future moment with the preset equipment failure rate threshold, and determine the maintenance frequency level based on the analysis results;

[0014] Generate a production line stability factor based on the predicted production efficiency, equipment failure rate at future times, and mean time to repair failures;

[0015] The production line stability coefficient is compared and analyzed with the preset production line stability coefficient threshold, the corresponding production line stability level is generated, and the production tasks are dynamically adjusted based on the production line stability level.

[0016] Preferably, the method of training the machine learning model for predicting production efficiency is as follows:

[0017] Convert the collected comprehensive production data into a corresponding set of feature vectors;

[0018] Each set of feature vectors is used as the input of the machine learning model. The machine learning model takes the production efficiency corresponding to each set of comprehensive production data as output, the actual production efficiency corresponding to each set of comprehensive production data as the prediction target, and minimizing the loss function value of the machine learning model as the training target; training is stopped when the loss function value of the machine learning model is less than or equal to the preset target loss value.

[0019] Preferably, the method of determining and generating the production efficiency level is as follows:

[0020] The production efficiency levels include: primary production efficiency, secondary production efficiency and tertiary production efficiency, wherein the primary production efficiency, secondary production efficiency and tertiary production efficiency decrease in descending order;

[0021] The predicted production efficiency is marked as Yx, and the preset production efficiency thresholds are Yx1 and Yx2, where Yx1<Yx2;

[0022] When Yx≤Yx1, at this time, the third-level production efficiency is generated;

[0023] When Yx2≥Yx>Yx1, at this time, the secondary production efficiency is generated;

[0024] When Yx>Yx2, at this time, the first-level production efficiency is generated.

[0025] Preferably, the method of allocating production tasks according to production efficiency levels is as follows:

[0026] Obtain the total production task from the production purchase and sales contract, and then prioritize the production tasks based on the total production task, according to the urgency of the order, order size and product production cycle.

[0027] Preferably, the training method of the machine learning model for predicting the equipment failure rate at a future moment includes:

[0028] Preset the prediction time step K, sliding step P and sliding window length N; use the sliding window method to convert the historical comprehensive data into multiple training samples, one training sample corresponds to a label, and constitutes a group of training data, and multiple groups of training data constitute a training set. The training set is used as the input of the machine learning model, and the equipment failure rate at the future moment after the prediction time step K is used as the output. The subsequent equipment failure rate of each training set is used as the prediction target. The prediction accuracy is used as the training target to train the machine learning; generate a machine learning model that predicts the equipment failure rate at the future moment based on the comprehensive parameters of the equipment operating status.

[0029] Preferably, the method of determining the maintenance frequency level is as follows:

[0030] The maintenance frequency levels include primary maintenance frequency, secondary maintenance frequency and tertiary maintenance frequency, wherein the primary maintenance frequency, secondary maintenance frequency and tertiary maintenance frequency decrease in sequence;

[0031] The predicted equipment failure rate is marked as Sg, and the preset equipment failure rate thresholds are Sg1 and Sg2, where Sg1<Sg2;

[0032] When Sg≤Sg1, at this time, the third-level maintenance frequency is generated;

[0033] When Sg2≥Sg>Sg1, at this time, the secondary maintenance frequency is generated;

[0034] When Sg>Sg2, at this time, the first-level maintenance frequency is generated.

[0035] Preferably, the average time to repair a fault is calculated as follows:

[0036]

[0037] In the formula, Gx represents the average time of fault repair, i represents the total number of fault repairs, and t j represents the time taken to repair the jth fault, j∈(1,2,…,i).

[0038] Preferably, the production line stability coefficient Sw is obtained by weighted calculation based on production efficiency, equipment failure rate, and average time for fault repair.

[0039] Preferably, the method of generating a corresponding production line stability level by combining the production line stability coefficient with a preset production line stability coefficient threshold is as follows:

[0040] The preset production line stability coefficient thresholds are Sw1 and Sw2, where Sw1 < Sw2;

[0041] When Sw≤Sw1, the stability of the production line is low;

[0042] When Sw2≥Sw>Sw1, the stability of the production line is medium;

[0043] When Sw>Sw2, at this time, the production line stability is high.

[0044] The automated production line task scheduling and optimization system based on artificial intelligence is implemented based on the above-mentioned automated production line task scheduling and optimization method based on artificial intelligence, including:

[0045] The data collection module is used to collect the historical production data of each production line. The historical production data includes comprehensive production data and production efficiency. The comprehensive production data includes equipment status, output, product qualification rate, work order completion time, and raw material supply rate;

[0046] Model training module, used to train a machine learning model for predicting production efficiency based on comprehensive production data, collect real-time comprehensive production data, and input it into the trained machine learning model to predict production efficiency;

[0047] The first analysis module is used to compare and analyze the predicted production efficiency with a preset production efficiency threshold, determine and generate a production efficiency level, and allocate production tasks according to the production efficiency level;

[0048] A data acquisition module is used to acquire the historical comprehensive data of each production line equipment, wherein the historical comprehensive data includes comprehensive parameters of equipment operation status and equipment failure rate, and predict the equipment failure rate at a future time based on the comprehensive parameters of equipment operation status;

[0049] The second analysis module is used to compare and analyze the predicted equipment failure rate at a future moment with a preset equipment failure rate threshold, and determine the maintenance frequency level according to the analysis result;

[0050] A data processing module is used to generate a production line stability coefficient based on the predicted production efficiency, equipment failure rate at future moments, and mean time to repair failures;

[0051] The third analysis module is used to compare and analyze the production line stability coefficient with a preset production line stability coefficient threshold, generate a corresponding production line stability level, and dynamically adjust the production tasks based on the production line stability level.

[0052] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0053] 1. The present invention predicts the production efficiency of the production line based on the comprehensive production data, and can reasonably allocate production tasks according to the production efficiency of each production line, provide theoretical support for the formulation of more accurate production plans and scheduling arrangements, avoid the problem that the production plan is inconsistent with the actual situation, ensure the reasonable allocation of tasks in each link, balance the workload, improve the utilization rate and output of the production line, increase product output, and enhance the profitability of the enterprise.

[0054] 2. The present invention can arrange maintenance in advance by predicting the equipment failure rate in the future, avoiding unexpected shutdown of the equipment during the production process, thereby improving the continuity and stability of the production line. By reasonably arranging the maintenance frequency based on the predicted equipment failure rate, preventive maintenance can be performed before the failure occurs, reducing the high maintenance costs caused by sudden failures. Through reasonable maintenance frequency and timely preventive maintenance, the damage to the equipment during operation can be reduced, the aging of the equipment can be delayed, and the service life of the equipment can be extended.

[0055] 3. The present invention can understand the specific situation of the production line in real time by real-time monitoring and evaluating the stability of the production line, and then identify potential production bottlenecks and causes of fluctuations, take measures to reduce production fluctuations, and maintain production continuity. By reasonably allocating tasks based on the stability coefficient of the production line, it can avoid overload production, improve production efficiency, and ensure that the production line always runs under the optimal load. By dynamically adjusting production tasks, it can balance the production load, reduce production fluctuations, ensure the continuity and stability of the production process, reasonably allocate production tasks, reduce delays caused by equipment failures and production fluctuations, and improve the on-time completion rate of work orders. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0057] Figure 1 The module structure diagram of the embodiment 1 of the present invention is shown;

[0058] Figure 2 The module structure diagram of the second embodiment of the present invention is shown;

[0059] Figure 3 The module structure diagram of the third embodiment of the present invention is shown;

[0060] Figure 4 A method flow chart of embodiment 4 of the present invention is shown. DETAILED DESCRIPTION

[0061] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] Example 1

[0063] Reference Figure 1 As shown, the automated production line task scheduling and optimization system based on artificial intelligence includes: a data acquisition module, a model training module and a first analysis module, and the modules are connected by wired / wireless means to realize data transmission;

[0064] The data collection module is used to collect the historical production data of each production line. The historical production data includes comprehensive production data and production efficiency. The comprehensive production data includes equipment status, output, product qualification rate, work order completion time, and raw material supply rate.

[0065] Specifically, the historical production data may be production data within the past year;

[0066] Equipment status includes equipment normal operation time and equipment failure time; equipment normal operation time refers to the time when the equipment works normally in the past year, which can be obtained through the equipment management software records; the longer the equipment normal operation time, the better the equipment status and the higher the corresponding production efficiency;

[0067] Equipment failure time refers to the equipment failure downtime in the past year, which can be obtained through the data recorded during the production process. Frequent failures and long downtime will affect the production efficiency of the production line;

[0068] Output refers to the number of products produced by a production line within a unit of time (such as a day, a week, or a month). The number of products produced can directly reflect the production capacity of the production line, and then reflect the production efficiency, which can be obtained through the data recorded during the production process;

[0069] The product qualification rate refers to the ratio of the number of qualified products to the total production quantity. A high qualification rate means less waste and higher production efficiency, as follows:

[0070]

[0071] In the formula, Ch represents the product qualification rate, Hg represents the number of qualified products, and Zc represents the total production quantity.

[0072] The work order completion time refers to the entire period from the issuance of a production task to the completion and delivery of the task. It is an important indicator for measuring production efficiency and process management. The work order completion time includes the sum of the time for each link, including preparation time, actual production time, quality inspection time, and packaging time.

[0073] The raw material supply rate is an important indicator to measure whether the raw material supply is timely and sufficient. It reflects the reliability of suppliers and the efficiency of supply chain management. The raw material supply rate can help companies evaluate the stability and responsiveness of their supply chains, as follows:

[0074]

[0075] In the formula, Yc represents the raw material supply rate, Sj represents the actual amount of raw materials received, and Jh represents the planned amount of raw materials received.

[0076] Production efficiency refers to the output volume completed per unit time during the production process of the production line. It is a key indicator to measure the efficiency and production capacity of the production system or equipment. The level of production efficiency directly affects the cost, profit and market competitiveness of the enterprise. Production efficiency can be reflected by the output volume per unit time, as follows:

[0077]

[0078] In the formula, Sx represents production efficiency, Cl represents output per unit time, and Ds represents unit time.

[0079] Model training module, used to train a machine learning model for predicting production efficiency based on comprehensive production data, collect real-time comprehensive production data, and input it into the trained machine learning model to predict production efficiency;

[0080] The way to train a machine learning model to predict production efficiency is as follows:

[0081] Convert the collected comprehensive production data into a corresponding set of feature vectors;

[0082] Each set of feature vectors is used as the input of the machine learning model. The machine learning model uses the production efficiency corresponding to each set of comprehensive production data as the output, the actual production efficiency corresponding to each set of comprehensive production data as the prediction target, and minimizing the loss function value of the machine learning model as the training target; when the loss function value of the machine learning model is less than or equal to the preset target loss value, the training is stopped;

[0083] The machine learning model can be one of the models such as SVM regression, random forest regression or neural network regression.

[0084] The loss function value of the machine learning model is the mean square error;

[0085] Mean square error is one of the commonly used loss functions. Minimization is used as the goal to train the model so that the machine learning model can better fit the data, thereby improving the performance and accuracy of the model;

[0086] In the loss function, MSE is the loss function value of the machine learning model, x is the feature vector group number; m is the number of feature vector groups; y is the number of feature vector groups; x is the production efficiency predicted by the x-th group of feature vectors, is the actual production efficiency corresponding to the xth group of eigenvectors;

[0087] Other model parameters of the machine learning model, target loss value, optimization algorithm, training set test set validation set ratio, and loss function optimization are all achieved through actual engineering implementation and continuous experimental tuning.

[0088] The first analysis module compares and analyzes the predicted production efficiency with the preset production efficiency threshold, determines and generates a production efficiency level, and allocates production tasks according to the production efficiency level;

[0089] The method for determining the generation production efficiency level is as follows:

[0090] The production efficiency levels include: primary production efficiency, secondary production efficiency and tertiary production efficiency, wherein the primary production efficiency, secondary production efficiency and tertiary production efficiency decrease in descending order;

[0091] The predicted production efficiency is marked as Yx, and the preset production efficiency thresholds are Yx1 and Yx2, wherein Yx1<Yx2, and the production efficiency thresholds are determined by those skilled in the art based on experience;

[0092] When Yx≤Yx1, it means that the production efficiency is low. At this time, the third-level production efficiency is generated;

[0093] When Yx2≥Yx>Yx1, it means that the production efficiency is moderate, and at this time, the secondary production efficiency is generated;

[0094] When Yx>Yx2, it means that the production efficiency is relatively high. At this time, the first-level production efficiency is generated.

[0095] The way to allocate production tasks according to production efficiency level is as follows:

[0096] Obtain the total production task from the production purchase and sales contract, and then prioritize the production tasks based on the total production task, according to the urgency of the order, order size and product production cycle;

[0097] For orders with high urgency, large order size and long product production cycle, they are allocated to production lines with first-level production efficiency;

[0098] For orders with moderate urgency, moderate order size and moderate product production cycle, they are assigned to production lines with production efficiency level 2;

[0099] For orders that are not very urgent, have a small order size, and have a short production cycle, they are assigned to production lines with a third-level production efficiency.

[0100] In this embodiment, by predicting the production efficiency of the production line based on the comprehensive production data, the production tasks can be reasonably allocated according to the production efficiency of each production line, providing theoretical support for the formulation of more accurate production plans and scheduling arrangements, avoiding the problem of production plans not being consistent with actual conditions, and reasonably allocating work order tasks according to the status and capabilities of different production lines or equipment to ensure that tasks in each link are reasonably allocated and the workload is balanced. By reasonably allocating work order tasks, the punctuality of work order completion can be improved, delays can be reduced, the utilization rate and output of production lines can be improved, product output can be increased, and the profitability of the enterprise can be improved.

[0101] Example 2

[0102] Reference Figure 2 As shown, the present invention further includes, based on Example 1:

[0103] A data acquisition module is used to acquire the historical comprehensive data of each production line equipment, wherein the historical comprehensive data includes comprehensive parameters of equipment operation status and equipment failure rate, and predict the equipment failure rate at a future time based on the comprehensive parameters of equipment operation status;

[0104] Specifically, the comprehensive parameters of the equipment operation status include the equipment's operating temperature, pressure, vibration, voltage, and current;

[0105] Among them, high operating temperature of the equipment may indicate cooling system problems or excessive wear; for hydraulic systems or pneumatic systems, changes in pressure can reflect the health of the equipment; by monitoring the vibration of the equipment, abnormal vibration patterns may indicate bearing problems, imbalance or other mechanical failures; monitoring the voltage and current changes of the equipment, abnormal fluctuations may indicate problems with the electrical system;

[0106] The operating temperature, pressure, vibration, voltage and current of the equipment can be monitored in real time through temperature sensors, pressure sensors, acceleration sensors, current sensors and voltage sensors.

[0107] The training method of the machine learning model for predicting the equipment failure rate at a future moment includes;

[0108] The comprehensive parameters of the equipment operation status are used as the input of the multi-feature time series prediction machine learning model, and the equipment failure rate at future moments is used as the output of the multi-feature time series prediction machine learning model. The equipment failure rate at future moments is taken as the prediction target, and minimizing the sum of the prediction accuracies of all training data is taken as the training target. The multi-feature time series prediction machine learning model is trained until the sum of the prediction accuracies converges and the training is stopped.

[0109] The loss function value of the machine learning model is the mean square error.

[0110] Mean square error is one of the commonly used loss functions. The model is trained with minimization as the goal, so that the machine learning model can better fit the data, thereby improving the performance and accuracy of the model.

[0111] In the loss function, MSE1 is the loss function value of the machine learning model, r is the feature vector group number; t is the number of feature vector groups; w r is the predicted equipment failure rate at the future moment corresponding to the rth group of eigenvectors, is the equipment failure rate at the future moment actually corresponding to the rth group of eigenvectors.

[0112] Other model parameters of the machine learning model, target loss value, optimization algorithm, training set test set validation set ratio, and loss function optimization are all achieved through actual engineering implementation and continuous experimental tuning.

[0113] Methods for predicting equipment failure rates at future times include:

[0114] Preset the prediction time step K, sliding step P and sliding window length N; use the sliding window method to convert the historical comprehensive data into multiple training samples, one training sample corresponds to a label, and constitutes a group of training data, and multiple groups of training data constitute a training set. The training set is used as the input of the machine learning model, and the equipment failure rate at the future moment after the prediction time step K is used as the output. The subsequent equipment failure rate of each training set is used as the prediction target. The prediction accuracy is used as the training target to train the machine learning; generate a machine learning model that predicts the equipment failure rate at the future moment based on the comprehensive parameters of the equipment operating status.

[0115] Specifically, a simple example of the sliding window method is as follows: assuming that you want to use historical comprehensive data packets 1, 2, 3, 4, 5, 6 to train a time prediction model to predict the value of one time step in the future, you can use a sliding window of length 3 and a sliding step of length 1 to generate a predicted future training set and prediction target. For example: the training set is 1, 2, 3, 2, 3, 4 and 3, 4, 5, and the prediction targets are 4, 5 and 6.

[0116] The sliding window method can be used to accurately and in advance predict the equipment failure rate at a future moment based on the existing historical comprehensive data packets in the selected scheduling area. Therefore, when monitoring the equipment failure rate on the production line, it can be discovered in advance that a high equipment failure rate may occur on the production line, thereby reminding the staff in advance and providing accurate data support for subsequent maintenance work, without any lag in equipment failure rate monitoring.

[0117] The second analysis module compares and analyzes the predicted equipment failure rate at a future moment with a preset equipment failure rate threshold, and determines the maintenance frequency level according to the analysis result;

[0118] The maintenance frequency levels include primary maintenance frequency, secondary maintenance frequency and tertiary maintenance frequency, wherein the primary maintenance frequency, secondary maintenance frequency and tertiary maintenance frequency decrease in sequence;

[0119] The predicted equipment failure rate is marked as Sg, and the preset equipment failure rate thresholds are Sg1 and Sg2, wherein Sg1<Sg2, and the equipment failure threshold is determined by those skilled in the art based on experience;

[0120] When Sg≤Sg1, it means that the equipment failure rate is low, and at this time, the third-level maintenance frequency is generated;

[0121] When Sg2≥Sg>Sg1, it means that the equipment failure rate is moderate, and at this time, the secondary maintenance frequency is generated;

[0122] When Sg>Sg2, it means that the equipment failure rate is high. At this time, a first-level maintenance frequency is generated.

[0123] Exemplarily, the first-level maintenance frequency, the second-level maintenance frequency and the third-level maintenance frequency can be customized. For example, the first-level maintenance frequency can be once every 40 minutes, the second-level cleaning frequency can be once every 120 minutes, and the third-level maintenance frequency can be once every 240 minutes. The maintenance frequency can also be customized according to specific production conditions.

[0124] In this embodiment, by predicting the equipment failure rate at future times, maintenance can be arranged in advance to avoid unexpected equipment shutdowns during the production process, thereby improving the continuity and stability of the production line. Based on the predicted equipment failure rate, the maintenance frequency can be reasonably arranged, and preventive maintenance can be performed before the failure occurs to reduce the high maintenance costs caused by sudden failures. Through reasonable maintenance frequency and timely preventive maintenance, the damage to the equipment during operation can be reduced, the aging of the equipment can be delayed, and the service life of the equipment can be extended. By predicting and handling potential failures in advance, safety accidents caused by equipment failures can be reduced, and the safety of the production line and operators can be improved.

[0125] Example 3

[0126] Reference Figure 3 As shown, the present invention further includes, based on Embodiment 1 and Embodiment 2:

[0127] The data processing module generates a production line stability coefficient based on the predicted production efficiency, equipment failure rate at future times, and the average time to repair failures;

[0128] Specifically, the mean time to repair a fault refers to the average time from when the equipment fails to when it is successfully repaired. The shorter the mean time to repair, the higher the reliability of the equipment and the smaller the impact on the stability of the production line. It is obtained through data recorded during the production process;

[0129]

[0130] In the formula, Gx represents the average time of fault repair, i represents the total number of fault repairs, and t j represents the time taken to repair the jth fault, j∈(1,2,…,i);

[0131] The production line stability factor is generated as follows:

[0132]

[0133] In the formula, Sw represents the stability coefficient of the production line, Yx represents the production efficiency, Sg represents the equipment failure rate, and Gx represents the average time for fault repair. represents the weight coefficient, Both are greater than 0.

[0134] It should be noted that the higher the production efficiency, the lower the equipment failure rate and the shorter the average time for fault repair, the higher the stability coefficient of the production line, and the corresponding production line is relatively stable; conversely, the lower the production efficiency, the higher the equipment failure rate and the longer the average time for fault repair, the lower the stability coefficient of the production line, and the corresponding production line is also susceptible to fluctuations.

[0135] It should be noted that the formulas involved in the above-mentioned are all calculated by removing dimensions and taking numerical values. They are a formula that is closest to the actual situation obtained by collecting a large amount of data and performing software simulation. The weight coefficients in the formulas and the various preset thresholds in the analysis process are set by technical personnel in this field according to actual conditions or obtained by simulating a large amount of data; the size of the weight coefficient is to quantify each parameter to obtain a specific value for subsequent comparison. The size of the weight coefficient depends on the amount of sample data and the corresponding processing coefficients initially set by technical personnel in this field for each group of sample data; as long as it does not affect the proportional relationship between the parameters and the quantized values.

[0136] The third analysis module generates a corresponding production line stability level by comparing the production line stability coefficient with a preset production line stability coefficient threshold, and dynamically adjusts the production task based on the production line stability level;

[0137] Specifically, the production line stability levels include: high production line stability, medium production line stability, and low production line stability;

[0138] The preset production line stability coefficient thresholds are Sw1 and Sw2, wherein Sw1<Sw2, and the production line stability coefficient thresholds are determined by those skilled in the art based on experience;

[0139] When Sw≤Sw1, it means that the production line stability coefficient is low. At this time, the stability of the generated production line is low;

[0140] When Sw2≥Sw>Sw1, it means that the production line stability coefficient is moderate. At this time, the stability of the generated production line is medium;

[0141] When Sw>Sw2, it means that the production line stability coefficient is high. At this time, the generated production line has high stability.

[0142] The method of dynamically adjusting production tasks based on the stability level of the production line is as follows:

[0143] For production lines with high stability coefficients, appropriately increase the task volume in the production plan and increase the running time of the production line to fully utilize its stability performance;

[0144] For production lines with moderate stability coefficients, reasonably arrange the production line's operating time and downtime for maintenance to ensure stable production and complete the tasks in the production plan on time;

[0145] For production lines with low stability coefficients, appropriately reduce the amount of tasks in the production plan and appropriately reduce the operating time of the production line to ensure stable production.

[0146] In this embodiment, by real-time monitoring and evaluating the stability of the production line, the specific situation of the production line can be understood in real time, and then potential production bottlenecks and causes of fluctuations can be identified, measures can be taken to reduce production fluctuations, maintain production continuity, and quickly respond to and handle emergencies, reduce the impact on production, and improve emergency response capabilities.

[0147] By reasonably allocating tasks based on the stability coefficient of the production line, we can avoid overload production, improve production efficiency, and ensure that the production line always runs under the optimal load. By dynamically adjusting production tasks, we can balance the production load, reduce production fluctuations, and ensure the continuity and stability of the production process. We can avoid excessive use of the production line, reduce the wear and aging of equipment, and extend equipment life. By reasonably allocating production tasks, we can reduce delays caused by equipment failures and production fluctuations and improve the on-time completion rate of work orders.

[0148] Example 4

[0149] Reference Figure 4 As shown, the task scheduling and optimization method of the automated production line based on artificial intelligence is applied to the above-mentioned task scheduling and optimization system of the automated production line based on artificial intelligence, and includes the following steps:

[0150] Collect historical production data of each production line. Historical production data includes comprehensive production data and production efficiency. Comprehensive production data includes equipment status, output, product qualification rate, work order completion time, and raw material supply rate.

[0151] Based on comprehensive production data, train a machine learning model to predict production efficiency. Collect real-time comprehensive production data and input it into the trained machine learning model to predict production efficiency.

[0152] Compare and analyze the predicted production efficiency with the preset production efficiency threshold, determine and generate the production efficiency level, and allocate production tasks according to the production efficiency level;

[0153] Obtaining historical comprehensive data of equipment on each production line, the historical comprehensive data including comprehensive parameters of equipment operation status and equipment failure rate, and predicting the equipment failure rate at a future time based on the comprehensive parameters of equipment operation status;

[0154] Compare and analyze the predicted equipment failure rate at a future moment with the preset equipment failure rate threshold, and determine the maintenance frequency level based on the analysis results;

[0155] Generate a production line stability factor based on the predicted production efficiency, equipment failure rate at future times, and mean time to repair failures;

[0156] The production line stability coefficient is compared and analyzed with the preset production line stability coefficient threshold, the corresponding production line stability level is generated, and the production tasks are dynamically adjusted based on the production line stability level.

[0157] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

[0158] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. An automated production line task scheduling and optimization method based on artificial intelligence, characterized in that: The steps include: Collect historical production data of each production line. Historical production data includes comprehensive production data and production efficiency. Comprehensive production data includes equipment status, output, product qualification rate, work order completion time, and raw material supply rate; Based on comprehensive production data, train a machine learning model to predict production efficiency. Collect real-time comprehensive production data and input it into the trained machine learning model to predict production efficiency. Compare and analyze the predicted production efficiency with the preset production efficiency threshold, determine and generate the production efficiency level, and allocate production tasks according to the production efficiency level; Obtaining historical comprehensive data of equipment on each production line, the historical comprehensive data including comprehensive parameters of equipment operation status and equipment failure rate, and predicting the equipment failure rate at a future time based on the comprehensive parameters of equipment operation status; Compare and analyze the predicted equipment failure rate at a future moment with the preset equipment failure rate threshold, and determine the maintenance frequency level based on the analysis results; Generate a production line stability factor based on the predicted production efficiency, equipment failure rate at future times, and mean time to repair failures; The production line stability coefficient is compared and analyzed with the preset production line stability coefficient threshold, the corresponding production line stability level is generated, and the production tasks are dynamically adjusted based on the production line stability level.

2. The method for scheduling and optimizing automated production line tasks based on artificial intelligence according to claim 1, characterized in that: The method of training a machine learning model to predict production efficiency is as follows: Convert the collected comprehensive production data into a corresponding set of feature vectors; Each set of feature vectors is used as the input of the machine learning model. The machine learning model uses the production efficiency corresponding to each set of comprehensive production data as the output, the actual production efficiency corresponding to each set of comprehensive production data as the prediction target, and minimizing the loss function value of the machine learning model as the training target; Stop training when the machine learning model loss function value is less than or equal to the preset target loss value.

3. The method for scheduling and optimizing automated production line tasks based on artificial intelligence according to claim 2, characterized in that: The method for determining and generating the production efficiency level is as follows: The production efficiency levels include: primary production efficiency, secondary production efficiency and tertiary production efficiency, wherein the primary production efficiency, secondary production efficiency and tertiary production efficiency decrease in descending order; The predicted production efficiency is marked as Yx, and the preset production efficiency thresholds are Yx1 and Yx2, where Yx1<Yx2; When Yx≤Yx1, at this time, the third-level production efficiency is generated; When Yx2≥Yx>Yx1, at this time, the secondary production efficiency is generated; When Yx>Yx2, at this time, the first-level production efficiency is generated.

4. The method for scheduling and optimizing automated production line tasks based on artificial intelligence as claimed in claim 3, characterized in that: The method of allocating production tasks according to production efficiency levels is as follows: Obtain the total production task from the production purchase and sales contract, and based on the total production task, prioritize the production tasks according to the urgency of the order, order size and product production cycle.

5. The method for scheduling and optimizing automated production line tasks based on artificial intelligence according to claim 4, characterized in that: The training method of the machine learning model for predicting the equipment failure rate at a future moment includes: Preset the prediction time step K, sliding step P and sliding window length N; The historical comprehensive data is converted into multiple training samples using the sliding window method. One training sample corresponds to one label and constitutes a set of training data. Multiple sets of training data constitute a training set. The training set is used as the input of the machine learning model, and the equipment failure rate at the future moment after the predicted time step K is used as the output. The subsequent equipment failure rate of each training set is used as the prediction target. The prediction accuracy is used as the training target to train the machine learning. Generate a machine learning model that predicts the equipment failure rate at future times based on comprehensive parameters of the equipment's operating status.

6. The method for scheduling and optimizing automated production line tasks based on artificial intelligence according to claim 5, characterized in that: The method for determining the maintenance frequency level is as follows: The maintenance frequency levels include primary maintenance frequency, secondary maintenance frequency and tertiary maintenance frequency, wherein the primary maintenance frequency, secondary maintenance frequency and tertiary maintenance frequency decrease in sequence; The predicted equipment failure rate is marked as Sg, and the preset equipment failure rate thresholds are Sg1 and Sg2, where Sg1<Sg2; When Sg≤Sg1, at this time, the third-level maintenance frequency is generated; When Sg2≥Sg>Sg1, at this time, the secondary maintenance frequency is generated; When Sg>Sg2, at this time, the first-level maintenance frequency is generated.

7. The method for scheduling and optimizing automated production line tasks based on artificial intelligence according to claim 6, characterized in that: The mean time to repair a fault is calculated as follows: In the formula, Gx represents the average time of fault repair, i represents the total number of fault repairs, and t j represents the time taken to repair the jth fault, j∈(1,2,…,i).

8. The method for scheduling and optimizing automated production line tasks based on artificial intelligence according to claim 7, characterized in that: The production line stability coefficient Sw is obtained by weighted calculation based on production efficiency, equipment failure rate, and average time for fault repair.

9. The method for scheduling and optimizing automated production line tasks based on artificial intelligence according to claim 8, characterized in that: The method of generating the corresponding production line stability level by combining the production line stability coefficient with the preset production line stability coefficient threshold is as follows: The preset production line stability coefficient thresholds are Sw1 and Sw2, where Sw1 < Sw2; When Sw≤Sw1, the stability of the production line is low; When Sw2≥Sw>Sw1, at this time, the stability of the production line is medium; When Sw>Sw2, at this time, the production line stability is high.

10. An automated production line task scheduling and optimization system based on artificial intelligence, used to implement the automated production line task scheduling and optimization method based on artificial intelligence as described in any one of claims 1 to 9, characterized in that: include: The data collection module is used to collect the historical production data of each production line. The historical production data includes comprehensive production data and production efficiency. The comprehensive production data includes equipment status, output, product qualification rate, work order completion time, and raw material supply rate; Model training module, used to train a machine learning model for predicting production efficiency based on comprehensive production data, collect real-time comprehensive production data, and input it into the trained machine learning model to predict production efficiency; The first analysis module is used to compare and analyze the predicted production efficiency with a preset production efficiency threshold, determine and generate a production efficiency level, and allocate production tasks according to the production efficiency level; A data acquisition module is used to acquire the historical comprehensive data of each production line equipment, wherein the historical comprehensive data includes comprehensive parameters of equipment operation status and equipment failure rate, and predict the equipment failure rate at a future time based on the comprehensive parameters of equipment operation status; The second analysis module is used to compare and analyze the predicted equipment failure rate at a future moment with a preset equipment failure rate threshold, and determine the maintenance frequency level according to the analysis result; A data processing module is used to generate a production line stability coefficient based on the predicted production efficiency, equipment failure rate at future moments, and mean time to repair failures; The third analysis module is used to compare and analyze the production line stability coefficient with a preset production line stability coefficient threshold, generate a corresponding production line stability level, and dynamically adjust the production tasks based on the production line stability level.

Images