An ai-based industrial device control system and method

Abstract

The application discloses an AI-based industrial equipment control system and method, and belongs to the technical field of industrial automation control. The system collects multi-dimensional operation data of industrial equipment in real time through a data acquisition module, including appearance data parameters and working condition parameters. After abnormal value elimination and standardization processing by a data processing module, the system inputs a pre-trained DRL model to generate multiple sets of control parameter adjustment suggestions. The adjustment suggestions are verified by Monte Carlo simulation in a digital twin simulation environment, and the optimal scheme is selected to generate control instructions, which are issued to the industrial equipment for execution through an OPC UA protocol. Meanwhile, the model parameters are dynamically optimized through an adaptive learning module. The application solves the problems of poor adaptability of static strategies, missing verification mechanism and weak self-learning ability in traditional industrial control, improves the dynamic response capability, safety and adaptability of the control system, and realizes the collaborative improvement of equipment energy efficiency optimization and production quality stability.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of industrial automation control, and specifically discloses an AI-based industrial equipment control system and method. BACKGROUND

[0002] With the promotion of Industry 4.0, the current industrial equipment control technology supports a dynamic intelligent reconstruction network architecture, which can quickly adapt to different production scene requirements, allows flexible adjustment of control strategies through programming, and adapts to diversified production processes; the industrial computer collects and analyzes data in real time to provide decision support for production optimization and reduce human misjudgment; and the industrial equipment control technology, by integrating hardware reliability, software intelligence and interactive friendliness, has become the core driving force for promoting the upgrading of industrial automation.

[0003] However, the traditional industrial equipment control technology has the following limitations:

[0004] (1) Defects of static control strategy: existing systems mostly use pre-programmed PID control or rule-based control logic, which is difficult to adapt to dynamic changes in complex working conditions; especially in multi-physical field coupling scenarios, such as pressure holding time, injection speed, cooling speed, and multi-parameter coordinated regulation of pressure holding pressure, the control parameter adjustment lags behind, resulting in a decrease in equipment energy efficiency and even safety hazards;

[0005] (2) Lack of verification mechanism: traditional methods directly issue control instructions to physical devices, lacking a virtual verification link; although some control systems introduce simulation modules, they only rely on a single physical model for prediction, without integrating a dynamic correction mechanism driven by real-time data, resulting in a large deviation between simulation results and actual working conditions;

[0006] (3) Weak self-adaptation: existing AI control schemes require a large amount of labeled data and have a long model update cycle; when facing long-period variables such as equipment aging and changes in raw material characteristics, the control strategy is difficult to achieve autonomous evolution and still requires frequent human intervention;

[0007] In view of the above problems, an AI-based industrial equipment control system and method are urgently needed. SUMMARY

[0008] In order to overcome the prior art defects mentioned above, the present application provides an AI-based industrial equipment control system and method, which collects multi-dimensional operation data of industrial equipment in real time through a data acquisition module, including appearance data parameters and working condition parameters, and after abnormal value elimination and standardization processing by a data processing module, inputs a pre-trained DRL model to generate multiple sets of control parameter adjustment suggestions, which are verified by Monte Carlo simulation in a digital twin simulation environment to generate control instructions after screening the optimal scheme, and are issued to industrial equipment for execution through OPC UA protocol, while the model parameters are dynamically optimized through an adaptive learning module, effectively solving the problems mentioned in the background art.

[0009] To achieve the above object, the present application provides the following technical scheme: an AI-based industrial equipment control system and method, characterized by comprising a data acquisition module, a data processing module, an AI control core module, an adaptive learning module and industrial equipment, wherein the data acquisition module is installed on the industrial equipment; specifically comprising the following steps:

[0010] S1, the data acquisition module collects multi-dimensional operation data of industrial equipment in real time, and the data processing module performs abnormal value elimination and standardization processing on the multi-dimensional operation data to obtain preprocessed data;

[0011] S2, input the preprocessed data into a pre-trained DRL model to output multiple sets of control parameter adjustment suggestions;

[0012] S3, verify the feasibility of the control parameter adjustment suggestions in the digital twin simulation environment, and select the optimal control parameter adjustment suggestion according to the verification result;

[0013] S4, generate control instructions according to the optimal control parameter adjustment suggestion after verification;

[0014] S5, issue the control instructions to the industrial equipment through OPC UA protocol to adjust the production process and feed back the execution data to the adaptive learning module.

[0015] In combination with all the above technical schemes, the present application has the following positive effects:

[0016] (1) Improved dynamic adaptability: through the combination of DRL model and digital twin simulation environment, the system can respond to dynamic changes of complex working conditions in real time, replacing traditional static control strategy and reducing the delay of control parameter adjustment;

[0017] (2) Enhanced safety and reliability: through Monte Carlo simulation verification and multi-physical field coupling simulation in the digital twin environment, the feasibility of the control instructions is verified before being issued, effectively avoiding equipment failure caused by parameter misadjustment;

[0018] (3) Optimized self-learning ability: the adaptive learning module supports online incremental learning and fault diagnosis, the model update cycle is shortened to the hour level, and can adapt to long-period variables such as device aging and raw material property changes, reducing the frequency of manual intervention. BRIEF DESCRIPTION OF DRAWINGS

[0019] The application will be further described with the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the application, and other drawings can be obtained by those of ordinary skill in the art without creative labor on the basis of the following drawings.

[0020] Figure 1 The method of the application is implemented in the flowchart.

[0021] Figure 2 The flowchart of the output control parameter adjustment suggestion in the application. DETAILED DESCRIPTION

[0022] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only a part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the application.

[0023] Referring to Figure 1 The application proposes an AI-based industrial equipment control system and method, characterized in that it comprises a data acquisition module, a data processing module, an AI control core module, an adaptive learning module and industrial equipment, wherein the data acquisition module is installed on the industrial equipment.

[0024] In a more specific application of the application, the data acquisition module is used to acquire appearance data parameters and working condition parameters. The appearance data parameters specifically include the size, warping degree and surface defect data of the product. The working condition parameters specifically include the qualified rate, mold internal pressure and injection speed. The data acquisition module specifically can be a combination of a machine vision unit and a sensor array. The machine vision unit is a combination of a high-resolution industrial camera and a laser scanner, which realizes millimeter-level detection of the product appearance size, warping degree and surface defects. The sensor array is composed of pressure sensors and speed sensors deployed at key positions of the production equipment. The pressure sensor is specifically an embedded piezoelectric sensor, and the speed sensor is installed on the injection molding machine screw to measure the instantaneous speed during the injection molding stage.

[0025] It should be further explained that the qualified rate statistics can be realized by integrating data management software in the unit to record the number of qualified / unqualified products in real time and calculate the qualified rate.

[0026] The data processing module is used for noise reduction, abnormal value elimination and standardization processing of the data collected by the data collection module. The abnormal value elimination method is to calculate the mean μ and the standard deviation σ based on the historical data distribution in the sliding time window, and to eliminate the real-time data points exceeding the range of mean ± 3σ. The standardization processing method is to perform Min-Max normalization on the multi-source heterogeneous data, to unify the dimension to the [0, 1] interval, and to form a standardized data matrix for the model.

[0027] The AI control core module comprises a deep reinforcement learning model and a digital twin simulation environment, which are used to generate control parameter adjustment suggestions after inputting preprocessed data, and to verify the feasibility of the parameter adjustment suggestions.

[0028] The adaptive learning module comprises an online incremental learning unit. The online incremental learning unit dynamically updates the DRL model parameters according to the actual feedback data after the industrial equipment executes the control instruction.

[0029] It should be further explained that the specific implementation steps of the online incremental learning unit for dynamically updating the DRL model parameters are as follows:

[0030] (1) Data feedback and collection: After the industrial equipment executes the control instruction, the multi-dimensional running data and working condition parameters after execution are collected in real time through the data collection module. Specifically, the size, warping degree, surface defect, qualified rate, mold pressure and injection speed are included.

[0031] (2) Feedback data preprocessing: The data processing module performs abnormal value elimination by dynamic threshold method and Min-Max normalization on the feedback data to generate a standardized data matrix.

[0032] (3) Parameter updating: An online policy gradient algorithm is adopted. The feedback data is taken as a new training sample, the policy network of the DRL model is optimized by small-batch gradient descent combined with the historical experience replay pool, and the update formula is: wherein θ is the policy network parameter, α is the learning rate, R(s, a) is the reward function calculated by the energy efficiency ratio, the qualified rate and the safety score, ∇θ is the gradient operator of the parameter, and represents the gradient direction of the expected return, is the expected return, that is, the expected value of the reward obtained by the policy in the state s by taking the action a.

[0033] It should be further explained that in practice, the expected return cannot be directly calculated and needs to be approximated by sampling:

[0034] In the formula, N represents the number of sampled trajectories, T represents the time step of a single trajectory, represents the probability of the policy network selecting the action in the state . denotes the cumulative discounted reward from time step t, where γ is the discount factor, r k is the immediate reward obtained at time step k;

[0035] Model verification and deployment: the updated DRL model needs to be verified through Monte Carlo simulation in the digital twin simulation environment to ensure the feasibility of its control recommendations. After verification, the new model parameters are deployed to the AI control core module in real time.

[0036] Cumulative discounted reward calculation example:

[0037] Parameter setting: the rewards obtained at time steps t to t+2 are r t =1, r t+1 =2, r t+2 =3, and the discount factor γ = 0.9;

[0038] Calculation process: G t =1+0.9×2+0.9 2 ×3=1+1.8+2.43=5.23;

[0039] Parameter update calculation example:

[0040] Parameter setting: energy efficiency ratio = 82%, pass rate = 95%, safety score = 90,

[0041] Reward function: R(s, a) = 0.4 × energy efficiency ratio + 0.5 × pass rate + 0.1 × safety score,

[0042] Policy network parameter θ old [0.5, -0.3], sample trajectory N = 2, time step T = 3, , ∂π / ∂θ = [0.2, -0.1];

[0043] Calculation process:

[0044] Trajectory 1:

[0045] At t = 0, the gradient term = ∇θlog(0.7) * 10 = (1 / 0.7) * [∂π / ∂θ] * 10 ≈ 2.857;

[0046] At t = 1, the gradient term = ∇θlog(0.6) * 8 = (1 / 0.6) * [∂π / ∂θ] * 8 ≈ 2.0;

[0047] At t = 2, the gradient term = ∇θlog(0.5) * 5 = (1 / 0.5) * [∂π / ∂θ] * 5 = 1.0;

[0048] Total gradient of trajectory 1: 2.857+2.0+1.0=5.857;

[0049] Trajectory 2:

[0050] At t=0, gradient term = ∇θlog(0.8)*12=(1 / 0.8)*[∂π / ∂θ]*12≈3.75;

[0051] At t=1, gradient term = ∇θlog(0.4)*6=(1 / 0.4)*[∂π / ∂θ]*6≈3.0;

[0052] At t=2, gradient term = ∇θlog(0.3)*3=(1 / 0.3)*[∂π / ∂θ]*3≈0.5;

[0053] Total gradient of trajectory 1: 3.75+3.0+0.5=7.25;

[0054] Average gradient: (5.857+7.25) / 2≈6.553;

[0055] =[0.5,−0.3]+0.01x6.553≈[0.5655,−0.2345];

[0056] Conclusion: adjust the policy network parameters from [0.5,−0.3] to [0.5655,−0.2345];

[0057] The industrial equipment can be intelligent manufacturing equipment production equipment, specifically can be an injection molding machine;

[0058] The specific implementation of the present application is shown in Figure 1 Specifically includes the following steps:

[0059] S1, the data acquisition module acquires the multi-dimensional operation data of the industrial equipment in real time, the data processing module carries out outlier rejection and standardization processing to the multi-dimensional operation data, and obtains the pretreatment data;

[0060] The data acquisition module cooperates with the laser scanner through a high-resolution industrial camera to collect appearance data parameters of the product, including size, warpage, yield rate and surface defect data, and obtains the pressure in the mold and the injection speed through the pressure sensor and the speed sensor; the data processing module performs noise reduction, outlier rejection and standardization processing on the data collected by the data acquisition module, the outlier rejection method is to calculate the mean μ and the standard deviation σ based on the historical data distribution in the sliding time window, and the real-time data points exceeding the mean ± 3σ range are rejected, and the missing data is filled based on the linear interpolation algorithm of time series; the standardization processing method is to perform Min-Max normalization on the multi-source heterogeneous data, unify the dimension to the [0, 1] interval, and form a standardized data matrix for the model to use.

[0061] It needs to be further explained that the method of using dynamic threshold value for noise reduction and outlier rejection is as follows: taking pressure data as an example, the mean and standard deviation of the pressure data within 10 minutes are calculated, if a certain pressure data deviates from the mean by more than 3 times the standard deviation, it is determined as an outlier and is rejected, assuming that the pressure data is 105MPa, 120MPa, 104MPa, 103MPa, 101MPa, the mean is 103MPa, and the standard deviation is 1.43MPa, 120MPa deviates from the mean by more than 3 times the standard deviation, and is rejected.

[0062] The linear interpolation algorithm based on time series fills in the missing data, which is to calculate the missing data by linear interpolation according to the parameter values of the adjacent time points, for example, if the pressure data is missing at a certain time, the average of the pressure values at the previous and next time points is used to fill in, taking the pressure data at a certain time as an example, assuming that t1, t2, t3 are three consecutive time points, the pressure measured at t1 is 103MPa, the data at t2 is missing, and the pressure data measured at t3 is 105MPa, then the pressure data at t2 is filled in as (103+105) / 2=104MPa.

[0063] The Min-Max normalization processing method for data is as follows, taking pressure data as an example: set P max =110MPa, P min =100MPa, the pressure data P at a certain time is 105MPa, after normalization P norm =(105-100) / (110-100)=0.5MPa.

[0064] S2, input the preprocessed data into the pre-trained DRL model, and output a plurality of control parameter adjustment suggestions;

[0065] It needs to be explained that the control parameter adjustment suggestions in the above step S2 include the adjustment of injection speed, holding pressure and cooling speed, and the specific adjustment formula is as follows:

[0066] Injection speed adjustment formula: ; in the formula V new is the adjusted injection speed; V base is the reference injection speed; ΔL is the difference between the measured size and the target size; N defect is the number of surface defects per unit area; k1, k2 are material-related correction coefficients, set according to the characteristics of the injection material; L target is the theoretical size of the plastic part under ideal molding conditions.

[0067] Injection speed adjustment example:

[0068] Scenario: When a certain injection molding machine produces plastic parts, the measured size is 2mm smaller than the target size, i.e. ΔL = -2mm, and the surface defect density is 5 / m²;

[0069] Parameter setting: reference injection speed V base = 80mm / s, target size L target = 100mm, material correction coefficients k1 = 0.1, k2 = 0.05

[0070] Calculation process:

[0071] Conclusion: The adjusted injection speed is increased to 100.2mm / s to compensate for the size deficiency and reduce the defect rate.

[0072] Holding pressure adjustment formula: ; in the formula P hold is the holding pressure; P max is the maximum holding pressure, determined by the material viscosity; t hold is the holding time; t gate is the gate freezing time; ΔD is the difference between the measured shrinkage and the theoretical value; λ, μ are process coupling coefficients, determined through DOE experiments.

[0073] Holding pressure adjustment example:

[0074] Scenario: In a certain injection molding process, the measured shrinkage is 0.5% higher than the theoretical value, i.e. ΔD = 0.5%, and the holding pressure needs to be optimized.

[0075] Parameter setting: maximum holding pressure P max = 100MPa, holding time t hold = 10s, gate freezing time t gate = 5s, process coupling coefficients λ = 0.2, μ = 0.1;

[0076] Calculation process:

[0077] Conclusion: The holding pressure is adjusted to 33.02 MPa to suppress the shrinkage deviation.

[0078] Cooling speed adjustment formula: ; in the formula, V cool is the cooling speed; W meas is the measured warpage; W crit is the critical warpage threshold, ΔT mold is the temperature difference between the mold cavity and the core, and η is the heat conduction correction coefficient; W crit , ΔT mold , and η are set according to expert opinions;

[0079] It should be further explained that the calculation method of the warpage is: W = h / (2L) × 100%, where h is the maximum warping height, and L is the measurement reference length, which is the diagonal length of the plastic part;

[0080] Warpage calculation example:

[0081] Parameter setting: the maximum warping height of a plastic part is 2 mm, and the diagonal length of the plastic part is 200 mm;

[0082] Calculation process: the warpage W of the plastic part = 2 / 2 × 200 × 100% = 0.5%;

[0083] Cooling speed adjustment example:

[0084] Scene description: in a certain injection molding production line, the measured warpage of the plastic part is 0.6%, i.e. W meas = 0.6%, the temperature difference between the mold cavity and the core is 2℃, i.e. ΔT mold = 2℃, and the cooling speed needs to be optimized to further improve the quality stability.

[0085] Parameter setting: reference cooling speed: V ref = 120 L / min, critical warpage threshold: W crit = 1%, heat conduction correction coefficient: η = 0.03;

[0086] Calculation process:

[0087] Conclusion: the cooling speed is adjusted to 50.88 L / min.

[0088] S3, verify the feasibility of the control parameter adjustment suggestion in the digital twin simulation environment, and select the optimal control parameter adjustment suggestion according to the verification result;

[0089] For example Figure 2As shown, in step S3, the method of feasibility verification is to input the control parameter adjustment suggestion into the digital twin simulation environment and simulate at least one complete production cycle. If the simulation result shows that the equipment energy efficiency ratio has improved significantly and the control parameter does not exceed the safety threshold, it is determined to pass the verification, and the control parameter adjustment suggestion is output. If the simulation result does not pass the verification, the last stable control parameter adjustment suggestion is output and an alarm log is generated, and an artificial intervention channel is started.

[0090] The selection method of the optimal control parameter adjustment suggestion is as follows:

[0091] In the digital twin simulation environment, the system performs multi-dimensional performance evaluation on the simulation results of each control parameter adjustment suggestion, including: energy efficiency ratio improvement rate, product qualification rate change, equipment safety index, process stability; through a weighted scoring model , the scores of each scheme are calculated and the suggestion with the highest total score is selected; in the above scoring model: E is the energy efficiency ratio improvement rate, Q is the qualification rate change, S is the equipment safety score, P is the process stability score, k is the weight coefficient, and A is the score, wherein k is determined through historical experience and expert suggestion;

[0092] An example of calculating the score of each scheme is as follows:

[0093] Parameter setting: energy efficiency ratio improvement rate weight k E =0.4, qualification rate change weight k Q =0.4, equipment safety weight k S =0.1, process stability weight k P =0.1;

[0094] Scheme 1: energy efficiency ratio improvement rate E=15%, qualification rate change Q=+2%, equipment safety score S=1.0, process stability score P=1.0;

[0095] Scheme 2: energy efficiency ratio improvement rate E=12%, qualification rate change Q=+3%, equipment safety score S=1.0, process stability score P=0.625;

[0096] Scheme 3: energy efficiency ratio improvement rate E=10%, qualification rate change Q=+1%, equipment safety score S=0.5, process stability score P=0.417;

[0097] Scheme 4: energy efficiency ratio improvement rate E=8%, qualification rate change Q=+0%, equipment safety score S=1.0, process stability score P=0.333;

[0098] Calculation process:

[0099] Scheme 1 score A1=(15 / 15*0.4)+(2 / 3*0.4)+(1*0.1)+(1*0.1)≈0.866;

[0100] Scheme 2 score A2=(12 / 15*0.4)+(3 / 3*0.4)+(1*0.1)+(0.625*0.1)=0.8825;

[0101] Scheme 3 score A3=(10 / 15*0.4)+(1 / 3*0.4)+(0.5*0.1)+(0.417*0.1)=0.4917;

[0102] Scheme 4 score A4=(8 / 15*0.4)+(0 / 3*0.4)+(1*0.1)+(0.333*0.1)≈0.346;

[0103] Conclusion: Scheme 2 has the highest score, and scheme 2 is selected as the optimal control parameter adjustment suggestion;

[0104] S4, generating a control instruction according to the control parameter adjustment suggestion after verification;

[0105] In the above step S4, the control parameter is one-to-one corresponding to the control variable of the equipment, for example: the injection speed is mapped to the screw advance rate parameter of the injection molding machine, the holding pressure is mapped to the pressure control node of the hydraulic system, and the cooling speed is mapped to the opening control node of the cooling water valve; and the instruction is packaged according to the data format of the OPC UA protocol.

[0106] S5, transmitting the control instruction to the industrial equipment through the OPC UA protocol to adjust the production process and feeding back the execution data to the adaptive learning module.

[0107] The control instruction is transmitted to the injection molding machine through the security channel of the OPC UA protocol; in specific implementation, the system first parses the control instruction into device recognizable data points, and then writes the device recognizable data points into the injection molding machine to execute the instruction, at the same time, the system monitors the instruction execution state, if the device feedback execution success signal, then record to the log; if timeout or return error code, trigger the adaptive learning module to roll back to the last stable parameter set, and push the alarm to the man-machine interface.

[0108] The above is only an example and description of the structure of the present application, and those skilled in the art can make various modifications or supplements to the described specific embodiments or use similar ways to replace, as long as it does not deviate from the structure of the present application or exceed the scope defined by the present application, which shall belong to the protection scope of the present application.

Claims

1. An AI-based industrial device control method, characterized by, Specifically comprising the following steps: S1, the data acquisition module collects multi-dimensional running data of the industrial equipment in real time, and the data processing module performs outlier rejection and standardization processing on the multi-dimensional running data to obtain preprocessed data; S2, inputting the preprocessed data into the pre-trained DRL model to output multiple sets of control parameter adjustment suggestions, the control parameter adjustment suggestions specifically including adjusting the injection speed, cooling speed and holding pressure; S3, verifying the feasibility of the control parameter adjustment suggestions in the digital twin simulation environment, and selecting the optimal control parameter adjustment suggestion according to the verification result; S4, generating a control instruction according to the optimal control parameter adjustment suggestion after verification; S5, issuing the control instruction to the industrial equipment through the OPC UA protocol to adjust the production process and feed back the execution data to the adaptive learning module; The formula for adjusting the injection speed is: ; in the formula, V new is the adjusted injection speed; V base is the reference injection speed; ΔL is the difference between the measured size and the target size; N defect is the number of surface defects per unit area; k1 and k2 are material-related correction coefficients, which are set according to the characteristics of the injection material; and L target is the theoretical size of the plastic part under ideal molding conditions. The formula for adjusting the holding pressure is: ; in the formula, P hold is the holding pressure; P max is the maximum holding pressure, determined by the viscosity of the material; t hold is the holding time; t gate is the gate freezing time; ΔD is the difference between the measured shrinkage rate and the theoretical value; λ and μ are process coupling coefficients, which are determined through DOE experiments; The formula for adjusting the cooling speed is: ; in the formula, V cool is the cooling speed; W meas is the measured warping degree; W crit is the critical warping degree threshold, ΔT mold is the temperature difference between the mold cavity and the core, and η is the heat conduction correction coefficient; W crit , ΔT mold , and η are set according to expert opinions.

2. The AI-based industrial device control method of claim 1, wherein: The preprocessed data specifically includes appearance data parameters and working condition parameters; the appearance data parameters specifically include product size, warping degree, and surface defect data; the working condition parameters specifically include yield, mold internal pressure, and injection speed. 3.The AI-based industrial device control method of claim 1, wherein: The outlier rejection adopts a dynamic threshold method to eliminate abnormal data points caused by sensor noise and communication packet loss; the standardization processing method is to perform Min-Max normalization on multi-source heterogeneous data, unify the dimension to the [0, 1] interval, and form a standardized data matrix for model use.

4. The AI-based industrial device control method of claim 1, wherein: The digital twin simulation environment is constructed by combining physical modeling and data driving, and supports Monte Carlo simulation verification of control strategies.

5. The AI-based industrial device control method of claim 1, wherein: The feasibility verification method is: inputting the control parameter adjustment suggestion into the digital twin simulation environment, simulating running at least one complete production cycle; if the simulation result shows that the equipment energy efficiency ratio is significantly improved and the control parameter does not exceed the safety threshold, it is determined to pass the verification, and the control parameter set is output; if the simulation result does not pass the verification, the last stable control parameter set is output and an alarm log is generated, and an artificial intervention channel is started.

6. The AI-based industrial device control method of claim 1, wherein: The adaptive learning module includes an online incremental learning unit which dynamically updates the DRL model parameters according to the actual feedback data after the industrial equipment executes the control instruction.

7. An AI-based industrial equipment control system applied to the AI-based industrial equipment control method of any one of claims 1-6, characterized in that It comprises a data acquisition module, a data processing module, an AI control core module, an adaptive learning module, and an industrial equipment, wherein the data acquisition module is installed on the industrial equipment.

Images