Digital twin operation management and intelligent scheduling system

Abstract

The invention discloses a digital twinborn operation management and intelligent scheduling system, which comprises a physical perception and execution layer, a data transmission and processing layer, a digital twinborn core layer and an intelligent decision-making and scheduling layer. By constructing a quantitative equipment health index model and converting the quantitative equipment health index model into a scheduling constraint condition, the system can dynamically adjust load distribution according to the equipment health state, the crossing from'fixed capability scheduling 'to'adaptive capability scheduling' is realized, high-fidelity and super-real-time rehearsal is performed on a scheduling scheme in a virtual space, and the scheduling efficiency is improved. Potential path conflicts, resource contention and other problems are exposed in advance, a finally issued scheduling scheme better fits an actual production environment, field adjustment is reduced, execution deviation is monitored in real time through a virtual-real closed-loop control unit, rescheduling is automatically triggered when abnormality is detected, uncertainty of a production field can be dynamically responded, and the production efficiency is improved. The continuity and stability of the production process are guaranteed, and the intelligent level and robustness of the whole system are improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology, specifically to a digital twin operation management and intelligent scheduling system. Background Technology

[0002] In traditional Manufacturing Execution Systems (MES) or Advanced Planning and Scheduling Systems (APS), production scheduling is typically based on fixed equipment capacity models and theoretical working hours.

[0003] Scheduling systems typically do not consider the real-time health status, performance degradation, or predictive maintenance needs of equipment. This may result in high-load tasks being assigned to equipment on the verge of failure, causing unplanned downtime and disrupting production plans. When anomalies occur on the production floor (such as sudden equipment failure, material shortages, or processing timeouts), traditional systems struggle to detect and dynamically adjust scheduling strategies in a timely manner, resulting in insufficient system robustness. Summary of the Invention

[0004] The purpose of this invention is to provide a digital twin operation management and intelligent scheduling system to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a digital twin operation management and intelligent scheduling system, comprising:

[0006] Physical sensing and execution layer: Based on multi-source sensors, it collects real-time operating status data of physical devices and receives and executes control commands; and it includes an actuator for receiving and executing control commands.

[0007] Data transmission and processing layer: Connects the physical sensing and execution layer, used to clean and standardize multi-source heterogeneous data, transmit it through a low-latency network, and establish a time-series database for storage;

[0008] The core layer of the digital twin includes a virtual-real mapping module and a simulation engine. The virtual-real mapping module constructs a digital twin that is synchronized with the physical entity based on geometric models, physical mechanism models, and real-time data. The simulation engine is used to perform ultra-real-time simulation of the production process in a virtual environment and predict its execution process and results.

[0009] Intelligent Decision-Making and Scheduling Layer: Includes the Operations Management Module and the Intelligent Scheduling Module;

[0010] Operations Management Module: Used to analyze equipment lifecycle data, build a quantitative model of equipment health, and dynamically generate equipment availability constraints by calculating the comprehensive health index of the equipment.

[0011] Intelligent scheduling module: It is used to construct a multi-objective optimization function to generate candidate scheduling schemes based on production tasks and the dynamic equipment availability constraints, use the simulation engine to perform pre-evaluation of candidate schemes, select the optimal scheme to generate control commands for issuance;

[0012] Virtual-real closed-loop control unit: used to monitor the synchronization deviation between the physical execution process and the digital twin simulation state in real time. When the deviation continues to exceed the limit, it is determined to be an execution abnormality, triggering the intelligent scheduling module to start the rescheduling process.

[0013] Preferably, the process by which the operation management module generates equipment availability constraints is as follows:

[0014] Define the comprehensive health index of device k at time t. The formula is expressed as:

[0015]

[0016] in, The remaining lifespan of the device is predicted based on a deep learning model; The initial design life of the equipment; The key performance data amplitude values ​​are collected by the sensor in real time; This is the maximum allowable amplitude threshold for the device; This represents the average pass rate of products recently processed by this equipment. The preset ideal pass rate benchmark; , , These are the weighting coefficients for lifespan, operating status, and processing quality, respectively, and they satisfy... + + =1.

[0017] Based on health index Construct dynamic availability constraint functions :

[0018]

[0019] in, and The preset health threshold, For the equipment to be in a healthy state Minimum degraded operation coefficient at that time This is the equipment's rated maximum load.

[0020] Preferably, the process by which the intelligent scheduling module obtains the optimal solution is as follows:

[0021] Define a multi-objective optimization function for intelligent scheduling. Taking into account completion time, energy consumption, and equipment health deterioration, the formula is expressed as:

[0022]

[0023] in, This represents the maximum completion time of the current scheduling scheme; Used as the base time; This represents the total energy consumption of the plan; Baseline energy consumption; The equipment is expected to be affected by this batch of missions. The estimated health loss caused, and Determined by a function of task type, processing time, and historical loss rate; This is the baseline health loss value for a single device; Total number of devices; Let be the weight coefficients for each optimization objective, and satisfy . + + =1;

[0024] When solving the optimization function, the following set of constraints are defined:

[0025]

[0026] in, For 0-1 decision variables, the value is 1 if process j of workpiece i is assigned to equipment k; This refers to the start time of the process; This refers to the end time of the previous process. For logistics and transportation time; The task load assigned to device k.

[0027] Preferably, the logistics transportation time in the constraints The acquisition process is as follows:

[0028] In the simulation engine of the core layer of the digital twin, a virtual environment model consistent with the physical workshop is constructed;

[0029] In a virtual environment, AGVs perform transportation tasks, taking into account path congestion and obstacle avoidance logic, and perform accelerated speed simulation.

[0030] The AGV travel time during the statistical simulation process is used as... Feedback is sent to the set of constraints.

[0031] Preferably, the virtual-real closed-loop control unit is used to monitor abnormal execution processes as follows:

[0032] By using a virtual-real synchronization deviation detection model, the deviation between the actual operating data of the physical entity and the simulation data of the digital twin is calculated in real time. The formula is expressed as:

[0033]

[0034] in, For physical devices The first moment One state parameter; for The corresponding state parameters of the digital twin at any given time; This is the amount of compensation for latency caused by network communication and computing. For the first The effective range of variation of each state parameter is used to eliminate the influence of dimensions. These are the normalized weights for the state parameters;

[0035] when The time exceeding the preset tolerance ϵ exceeds the threshold. If an execution error is detected, the intelligent scheduling module is triggered to perform rescheduling.

[0036] Compared with the prior art, the beneficial effects of the present invention are:

[0037] 1. Achieved deep integration of production and operation and maintenance: by constructing a quantitative equipment health index model. And transform it into scheduling constraints. The system can dynamically adjust load distribution based on the health status of the equipment, realizing a leap from "fixed capacity scheduling" to "adaptive capacity scheduling", effectively extending equipment life and reducing unplanned downtime.

[0038] 2. Improved scheduling feasibility: Utilizing a digital twin simulation engine, scheduling schemes (especially dynamic logistics times) are analyzed in virtual space. High-fidelity, ultra-real-time simulations are conducted to expose potential path conflicts and resource contention issues in advance, making the final scheduling plan more in line with the actual production environment and reducing on-site adjustments.

[0039] 3. Possesses strong closed-loop error correction capabilities: The system monitors execution deviations in real time through a virtual-physical closed-loop control unit and automatically triggers rescheduling upon detecting anomalies. This gives the system strong self-healing and anti-interference capabilities, enabling it to dynamically respond to uncertainties in the production environment, ensuring the continuity and stability of the production process, and improving the overall system's intelligence and robustness. Attached Figure Description

[0040] Figure 1This is a system block diagram of the digital twin operation management and intelligent scheduling system proposed in this invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Please see Figure 1 The present invention provides a technical solution: a digital twin operation management and intelligent scheduling system, including a physical sensing and execution layer, a data transmission and processing layer, a digital twin core layer, and an intelligent decision-making and scheduling layer.

[0043] This system is applied in a discrete manufacturing workshop, which contains multiple CNC machine tools, AGV carts and corresponding sensor networks (vibration, temperature, current sensors, etc.).

[0044] Data Acquisition and Modeling:

[0045] Sensors in the physical sensing layer collect equipment data in real time. After preliminary processing by the edge gateway, the data is transmitted to the data processing layer via industrial Ethernet. In the core layer of the digital twin, a digital twin containing a three-dimensional geometric model and a physical behavior model is created for each machine tool and AGV, and bound to the real-time data stream to achieve virtual-real synchronization.

[0046] Dynamic health assessment and constraint generation:

[0047] In traditional scheduling, equipment capacity is usually considered a fixed value, and the operations management module periodically calculates the comprehensive health index of each machine tool. Define the comprehensive health index of device k at time t. .

[0048] The formula is:

[0049]

[0050] For example, for a machining center, an LSTM network can be used to predict its vibration based on historical vibration sequences. 200 hours (initial lifespan) (for 1000 hours); current vibration amplitude The threshold is 5.2 mm / s. The current velocity is 6.0 mm / s; the recent product qualification rate is... It is 98.5%. =0.4, =0.4, =0.2, calculated as follows ≈ 0.4*(200 / 1000)+0.4*(1-5.2 / 6.0)+0.2*0.985 = 0.3316 (Note: This is just an example value; the actual weights will be adjusted according to the device characteristics).

[0051] set up =0.7, =0.3, minimum load factor = 0.5. Because = 0.3316 satisfies Calculate its current available load capacity based on the dynamic constraint function:

[0052] = =0.5395×

[0053] This means that in subsequent scheduling, the device's workload must not exceed its rated maximum load. Approximately 54% of the system was in a protective downgrade mode, thus entering a protective downgrade operation mode.

[0054] Intelligent scheduling and simulation:

[0055] When a new production order arrives, the intelligent scheduling module constructs a multi-objective optimization function for intelligent scheduling, aiming to minimize the weighted sum of completion time, energy consumption, and equipment health deterioration. The aim is to minimize completion time, energy consumption, and equipment health deterioration.

[0056] The formula is:

[0057] When solving this function, the following constraints must be met: This means that if a task requires a certain amount of load Exceeding the current dynamic capabilities of the device This allocation scheme will be directly eliminated.

[0058] Meanwhile, constraints Logistics time It is not an estimated value, but a precise simulation time obtained by simulating the actual walking path and obstacle avoidance process of the AGV in the virtual workshop model in the core layer of the digital twin.

[0059] For each candidate solution, the system does not directly adopt its theoretical value, but instead calls upon a simulation engine to conduct "virtual trial production." In the virtual workshop, the engine strictly follows the sequence of the proposed solution, driving the digital twins of AGVs and machine tools. For example, it simulates the entire process of AGVs retrieving materials from the warehouse, navigating potentially congested aisles, and queuing in front of machine tools to wait for loading, thus calculating a much more accurate actual logistics timeline than theoretical distance calculations. ,this The feedback is sent back to the scheduling model for re-evaluation. After multiple rounds of "generation-simulation-evaluation" iterations, the system selects the scheme with the best overall performance index and converts it into specific control instructions (such as G-codes and AGV path instructions) and sends them to the physical layer for execution.

[0060] Closed-loop monitoring and rescheduling:

[0061] During the execution of scheduling instructions by physical devices, the system uses a virtual-real synchronization deviation detection model for monitoring.

[0062] The formula is: .

[0063] After the command is issued, the virtual-real closed-loop control unit continuously compares the actual spindle power of the machine tool. ) and the expected power of twin simulation ( Assuming that due to unexpected tool wear, the actual power consistently exceeds the simulated value, leading to deviation... If the threshold ϵ (e.g., 0.15) is exceeded and remains so for 5 seconds (e.g., 3 seconds), the system immediately determines that an execution abnormality has occurred, sends an alarm to the intelligent scheduling module, and transmits the latest status of all current equipment and work-in-process.

[0064] The intelligent scheduling module then interrupts the original scheduling plan, takes the latest workshop status as input, restarts the aforementioned optimization and simulation process, and quickly generates and issues a new adjustment plan, thereby achieving dynamic compensation and closed-loop control for production anomalies.

[0065] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0066] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A digital twin operation management and intelligent scheduling system, characterized in that, include: Physical sensing and execution layer: Based on multi-source sensors, it collects real-time operating status data of physical devices and receives and executes control commands; And an actuator, used to receive and execute control commands; Data transmission and processing layer: Connects the physical sensing and execution layer, used to clean and standardize multi-source heterogeneous data, transmit it through a low-latency network, and establish a time-series database for storage; The core layer of the digital twin includes a virtual-real mapping module and a simulation engine. The virtual-real mapping module constructs a digital twin that is synchronized with the physical entity based on geometric models, physical mechanism models, and real-time data. The simulation engine is used to perform ultra-real-time simulation of the production process in a virtual environment and predict its execution process and results. Intelligent Decision-Making and Scheduling Layer: Includes the Operations Management Module and the Intelligent Scheduling Module; Operations Management Module: Used to analyze equipment lifecycle data, build a quantitative model of equipment health, and dynamically generate equipment availability constraints by calculating the comprehensive health index of the equipment. Intelligent scheduling module: It is used to construct a multi-objective optimization function to generate candidate scheduling schemes based on production tasks and the dynamic equipment availability constraints, use the simulation engine to perform pre-evaluation of candidate schemes, select the optimal scheme to generate control commands for issuance; Virtual-real closed-loop control unit: used to monitor the synchronization deviation between the physical execution process and the digital twin simulation state in real time. When the deviation continues to exceed the limit, it is determined to be an execution abnormality, triggering the intelligent scheduling module to start the rescheduling process.

2. The digital twin operation management and intelligent scheduling system according to claim 1, characterized in that: The process by which the operation management module generates equipment availability constraints is as follows: Define the comprehensive health index of device k at time t. The formula is expressed as: ； in, The remaining lifespan of the device is predicted based on a deep learning model; The initial design life of the equipment; The key performance data amplitude values ​​are collected by the sensor in real time; This is the maximum allowable amplitude threshold for the device; This represents the average pass rate of products recently processed by this equipment. The preset ideal pass rate benchmark; , , These are the weighting coefficients for lifespan, operating status, and processing quality, respectively, and they satisfy... + + =1; Based on health index Construct dynamic availability constraint functions : ； in, and The preset health threshold, For the equipment to be in a healthy state Minimum degraded operation coefficient at that time This is the equipment's rated maximum load.

3. The digital twin operation management and intelligent scheduling system according to claim 1, characterized in that: The process by which the intelligent scheduling module obtains the optimal solution is as follows: Define a multi-objective optimization function for intelligent scheduling. Taking into account completion time, energy consumption, and equipment health deterioration, the formula is expressed as: ； in, This represents the maximum completion time of the current scheduling scheme; Used as the base time; This represents the total energy consumption of the plan; Baseline energy consumption; The equipment is expected to be affected by this batch of missions. The estimated health loss caused, and Determined by a function of task type, processing time, and historical loss rate; This is the baseline health loss value for a single device; Total number of devices; Let be the weight coefficients for each optimization objective, and satisfy . + + =1; When solving the optimization function, the following set of constraints are defined: ； in, For 0-1 decision variables, the value is 1 if process j of workpiece i is assigned to equipment k; This refers to the start time of the process; This refers to the end time of the previous process. For logistics and transportation time; The task load assigned to device k.

4. The digital twin operation management and intelligent scheduling system according to claim 3, characterized in that: The logistics transportation time in the constraints The acquisition process is as follows: In the simulation engine of the core layer of the digital twin, a virtual environment model consistent with the physical workshop is constructed; In a virtual environment, AGVs perform transportation tasks, taking into account path congestion and obstacle avoidance logic, and perform accelerated speed simulation. The AGV travel time during the statistical simulation process is used as... Feedback is sent to the set of constraints.

5. The digital twin operation management and intelligent scheduling system according to claim 1, characterized in that: The virtual-real closed-loop control unit is used to monitor abnormal execution processes as follows: By using a virtual-real synchronization deviation detection model, the deviation between the actual operating data of the physical entity and the simulation data of the digital twin is calculated in real time. The formula is expressed as: ； in, For physical devices The first moment One state parameter; for The corresponding state parameters of the digital twin at any given time; This is the amount of compensation for latency caused by network communication and computing. For the first The effective range of variation of each state parameter is used to eliminate the influence of dimensions. These are the normalized weights for the state parameters; when The time exceeding the preset tolerance ϵ exceeds the threshold. If an execution error is detected, the intelligent scheduling module is triggered to perform rescheduling.

Images