Industrial control system and control device

Abstract

The invention discloses an industrial control system and a control device, and relates to the technical field of industrial control, a data acquisition module obtains spraying path precision data and spraying thickness consistency data, and a data processing module analyzes and extracts a spraying path precision deviation value and a thickness drift speed abnormal value; the anomaly detection module comprehensively calculates a health score by using a machine learning model, evaluates the health state of the spraying robot and predicts a potential fault; the early warning module performs real-time early warning when fault prediction exceeds a threshold value, and the control module dynamically adjusts control parameters of the spraying robot, corrects the path and thickness and triggers maintenance suggestions through health state change trend prediction, so that the self-diagnosis capability and the control precision of the industrial control system are remarkably improved, potential problems can be found in the early stage, and the safety of the industrial control system is improved. The shutdown maintenance risk and the product rejection rate are reduced, so that the production efficiency and the product quality are improved.

Description

Technical Field

[0001] The present invention relates to the field of industrial control technology, and in particular to an industrial control system and a control device. Background Art

[0002] Industrial control refers to the process of automating, optimizing and managing machines, equipment and processes in industrial production by using various technical means and equipment. Its core goal is to improve production efficiency, ensure product quality, reduce production costs and reduce human intervention. Industrial control usually relies on sensors, controllers, actuators and computer systems to collect and process data and execute preset control strategies to achieve efficient management of industrial production. For example, in the manufacturing industry, industrial control can be reflected in the automated production of assembly lines. A typical case is the spraying process in an automobile manufacturing plant. Through the industrial control system, sensors monitor the position and status of the car body, and the controller automatically adjusts the operation of the spraying robot according to the preset parameters to ensure that the spraying thickness of each car body is consistent, thereby improving product quality. In addition, in the operation of the reactor in a chemical plant, the industrial control system can monitor the temperature, pressure and reaction speed in real time, and automatically adjust the heating or cooling equipment based on these data to ensure that the chemical reaction is carried out under optimal conditions.

[0003] The prior art has the following deficiencies: Many existing industrial control systems have limited self-diagnostic capabilities and cannot detect potential problems at an early stage. For example, during the operation of a spraying robot, some mechanical parts may gradually wear out, resulting in reduced spraying accuracy. However, if the system lacks real-time health monitoring capabilities, the fault may not be discovered until it seriously affects production, resulting in long downtime for repairs. In addition, due to the reduction in control accuracy, the problem may not be discovered until after production is completed, resulting in the scrapping or recall of the entire batch of products. Summary of the invention

[0004] The purpose of the present invention is to provide an industrial control system and a control device to solve the deficiencies in the background technology.

[0005] In order to achieve the above-mentioned object, the present invention provides the following technical solutions: an industrial control system, including a data acquisition module, a data processing module, an abnormality detection module, an early warning module and a control module; A data acquisition module is used to obtain the spraying path accuracy data and spraying thickness consistency data during the operation of the spraying robot; A data processing module is used to analyze the spraying path accuracy data and calculate the spraying path accuracy deviation value, and to perform trend analysis on the spraying thickness consistency data and calculate the thickness drift velocity abnormal value; The anomaly detection module comprehensively calculates the spray path accuracy deviation value and thickness drift speed anomaly value based on the machine learning model, generates a health score based on the calculation results, evaluates the health status of the spray robot, and predicts potential failures; An early warning module is used to generate a real-time early warning signal when the fault prediction result exceeds a set threshold; The control module is used to predict the changing trend of the health status of the spray robot, dynamically adjust the control parameters of the spray robot according to the prediction results to correct the spray path and thickness, and trigger maintenance recommendations.

[0006] Preferably, in the data processing module, the spraying path accuracy data is analyzed and then the spraying path accuracy deviation value is calculated. The calculation method of the spraying path accuracy deviation value is: Enter the actual path data sequence , each point Represents the coordinates of the spray head in three-dimensional space ; Ideal path data sequence , each point Indicates the reference coordinates in the preset path ; Calculate the Euclidean distance between all points in the actual path A and the ideal path B, and generate the distance matrix D. The formula is: ; i=1,2,…,m represents the point index in the actual path, j=1,2,…,n represents the point index in the ideal path; define C(i,j) as the cumulative minimum distance of the optimal alignment path from (1,1) to (i,j), recursive formula: ; In the formula, The actual path point Align ideal path points ; The actual path point Align ideal path points ; The actual path point Align ideal path points , boundary conditions: C(1,1)=D(1,1), , ; Start from C(m,n), trace back to C(1,1), determine the optimal alignment path between the actual path and the ideal path, select the path based on the cumulative minimum distance, and record the point pairs aligned at each step ; Calculate the spray path accuracy deviation value, the expression is: ; Where L is the number of point pairs contained in the alignment path, is the cumulative distance of the optimal alignment path, and AGS is the spray path accuracy deviation value.

[0007] Preferably, in the data processing module, after trend analysis of the spraying thickness consistency data, the thickness drift velocity abnormal value is calculated, and the calculation method of the thickness drift velocity abnormal value is: Enter the thickness drift velocity sequence , represents the thickness drift velocity data at each time point, selects the neighborhood size k, calculates the k distance, and is the i-th data point The k-distance of is the distance to its k-neighborhood boundary points: ; In the formula, represents the k distance, It is the distance The kth closest point; ; The k-neighborhood includes all points satisfy: ; For any point and , and its reachable distance is: ; The local reachability density reflects the point For the density of its k-neighborhood , the expression is: ; In the formula, for The number of points in the neighborhood of ; calculate the thickness drift velocity anomaly, the expression is: ; Where EMP is the thickness drift velocity anomaly value.

[0008] Preferably, in the anomaly detection module, based on the machine learning model, the spray path accuracy deviation value and the thickness drift speed anomaly value are comprehensively calculated, and a health score is generated according to the calculation results to evaluate the health status of the spray robot, specifically: the spray path accuracy deviation value and the thickness drift speed anomaly value are converted into a comprehensive feature vector, and the comprehensive feature vector is used as the input of the machine learning model. The machine learning model uses each group of comprehensive feature vectors to predict the health score value label of the health status of the spray robot as the prediction target, and minimizes the sum of the prediction errors of the health score value labels of the health status of all spray robots as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence, and the model training is stopped. The health score value of the health status of the spray robot is determined according to the model output result, wherein the machine learning model is a polynomial regression model; Y=100: The health status is the best and no maintenance is required; Y=50 to 70: The health status is moderate and regular inspection is required; Y<50: The health status is poor and potential faults exist and immediate maintenance is required.

[0009] Preferably, the early warning module generates a real-time early warning signal when the fault prediction result exceeds a set threshold; The acquired health score value of the health status of the spray robot is compared with the health score value reference threshold pre-set based on historical data. If the health score value of the health status of the spray robot is less than the health score value reference threshold, an early warning signal is generated; if the health score value of the health status of the spray robot is greater than or equal to the health score value reference threshold, no early warning signal is generated.

[0010] Preferably, in the control module, the health status change trend of the spraying robot is predicted, and the control parameters of the spraying robot are dynamically adjusted according to the prediction results to correct the spraying path and thickness, specifically: Predict the changing trend of the health score of the spraying robot, the formula is: ; In the formula, is the predicted change of the health score at time t, For the moment The health score of is the spraying path accuracy deviation value at time t, is the thickness drift velocity abnormal value at time t, e is the weight parameter set of the control model, and f is the prediction function. When <0 and the absolute value is large, it indicates that the health status has deteriorated and the parameters need to be adjusted.

[0011] Preferably, the motion trajectory parameters of the spraying robot are dynamically adjusted by the control module, and the formula is: ; In the formula, is the path adjustment amount, which is used to correct the robot's motion trajectory. is the instantaneous rate of change of path deviation, reflecting the dynamic change of the path. , , The proportional, integral and differential control coefficients are optimized according to the spraying task; Dynamically adjust the spray flow Q and spray speed v to correct the thickness deviation. The formula is: ; ; In the formula, is the spray flow adjustment, which is used to correct the thickness change, Δv is the spray speed adjustment, and α, β, γ are the coefficients of the control model.

[0012] The present invention also provides an industrial control device, comprising: Integrated processors to perform data acquisition, processing, anomaly detection, and fault prediction functions; Communication interface, used for data exchange with the spraying robot and external monitoring equipment; Storage unit, used to store historical operation data, prediction models and warning records.

[0013] In the above technical solution, the technical effects and advantages provided by the present invention are: 1. The present invention realizes real-time monitoring, intelligent evaluation and dynamic adjustment of the operating status of the spray robot through efficient collaboration of data acquisition, processing, anomaly detection, early warning and control modules. Based on the accurate calculation of the spray path accuracy deviation value and the thickness drift speed anomaly value, combined with the predictive ability of the machine learning model for the health score value, the system can detect potential problems early, and adjust the control parameters according to the changing trend of the health status, correct the spray path and thickness, and ensure the stability of the spray accuracy and product quality.

[0014] 2. By introducing real-time health monitoring and prediction functions, the present invention overcomes the limitation of insufficient self-diagnosis capability in traditional industrial control systems, significantly reducing downtime for maintenance and product scrap rate. In addition, the system's early warning module timely warns of potential faults, while the control module dynamically optimizes the spraying parameters, effectively improving the operating efficiency and service life of the equipment, thereby providing reliable and efficient technical support for industrial automation production. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0016] Figure 1 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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.

[0018] Example 1, please refer to Figure 1 As shown, an industrial control system described in this embodiment includes a data acquisition module, a data processing module, an abnormality detection module, an early warning module and a control module; A data acquisition module is used to obtain the spraying path accuracy data and spraying thickness consistency data during the operation of the spraying robot; A data processing module is used to analyze the spraying path accuracy data and calculate the spraying path accuracy deviation value, and to perform trend analysis on the spraying thickness consistency data and calculate the thickness drift velocity abnormal value; The anomaly detection module comprehensively calculates the spray path accuracy deviation value and thickness drift speed anomaly value based on the machine learning model, generates a health score based on the calculation results, evaluates the health status of the spray robot, and predicts potential failures; An early warning module is used to generate a real-time early warning signal when the fault prediction result exceeds a set threshold; The control module is used to predict the changing trend of the health status of the spray robot, dynamically adjust the control parameters of the spray robot according to the prediction results to correct the spray path and thickness, and trigger maintenance recommendations.

[0019] In the data acquisition module, the spray path accuracy data reflects the deviation between the actual running trajectory of the spray robot and the preset trajectory. By accurately measuring the path data, the robot's operating status can be monitored and the stability and accuracy of motion control can be evaluated. Use high-precision optical sensors, laser rangefinders or visual cameras (such as 3D vision systems). The sensor is fixed to a reference point near the spray robot or in the spraying area to capture the real-time position of the robot's spray head. The sensor records its real-time trajectory in three-dimensional space (X, Y, Z) by tracking the actual position coordinates of the spray head. By comparing with the preset spray trajectory, the deviation value at each time point is calculated, including lateral deviation, longitudinal deviation and height deviation. The acquisition frequency is adjusted according to the robot's running speed, generally in milliseconds, to ensure accuracy.

[0020] The output data includes: Real-time coordinate data: the real-time position coordinates of the spray head in space. Path deviation data: the distance deviation between the actual path and the ideal path. Trajectory stability data: by analyzing the change trend of the deviation data, the stability of the robot operation is evaluated. Spraying thickness consistency data reflects the thickness uniformity of the spray layer on the body surface. Thickness data is the core indicator for measuring spraying quality, which helps to judge the stability of the spraying process and the operation status of the robot.

[0021] Use non-contact thickness measurement sensors, such as ultrasonic thickness gauges, optical interferometer thickness gauges, or electromagnetic induction thickness gauges. The sensor is fixed at the end of the spraying area or directly integrated into the robot spray head to monitor the thickness of the sprayed layer in real time. Real-time thickness measurement: The sensor collects thickness data of the sprayed surface coating through reflection or interference principles. The sensor samples the thickness at multiple points during the spraying process and generates a thickness distribution map. Compare thickness data at different time periods to monitor consistency changes.

[0022] The output data includes: Real-time thickness data: records the spray thickness value of each sampling point. Thickness distribution diagram: reflects the thickness uniformity of the entire spray surface. Consistency evaluation index: calculates the mean, standard deviation and drift trend based on the thickness data to evaluate the stability of the spray quality.

[0023] The hardware components include: Multiple types of sensors: High-precision optical or laser sensors for path accuracy acquisition. Ultrasonic or optical thickness measurement sensors for thickness consistency acquisition. Data interface: The sensor is connected to the data processing module through a high-speed data interface (such as Ethernet, CAN bus or RS485) to achieve real-time transmission. Mounting bracket: Fix the sensor in the best position to ensure the accuracy and stability of data acquisition.

[0024] The spraying process may generate dust or paint particles, which may interfere with sensor measurement. The high-speed movement of the spraying robot may cause data acquisition lag or errors. Select sensors with strong anti-interference ability and equip them with dust covers or air filtration systems. Use sensors with faster response speeds to adapt to the high-speed movement of robots. Use multi-sensor collaborative collection and use redundant data to correct errors.

[0025] The data processing module is used to analyze the spray path accuracy data and calculate the spray path accuracy deviation value, and at the same time perform trend analysis on the spray thickness consistency data and calculate the thickness drift speed abnormal value.

[0026] Before analysis, the collected spray path accuracy data needs to be preprocessed to improve data accuracy and analysis efficiency. Use filtering techniques (such as low-pass filtering and Kalman filtering) to remove data anomalies caused by environmental noise or sensor errors. Convert the path data to a standardized coordinate system relative to the ideal path for easy analysis. Ensure that the time interval of the data is consistent to avoid affecting the analysis of dynamic changes in the trajectory.

[0027] Calculate the path deviation at each time point: Lateral deviation: the distance between the actual path and the ideal path in the lateral direction. Longitudinal deviation: the distance between the actual path and the ideal path in the longitudinal direction. The total deviation is the Euclidean distance between the actual path and the ideal path.

[0028] After analyzing the spraying path accuracy data, the spraying path accuracy deviation value is calculated. The calculation method of the spraying path accuracy deviation value is: Enter the actual path data sequence , each point Represents the coordinates of the spray head in three-dimensional space ; Ideal path data sequence , each point Indicates the reference coordinates in the preset path .

[0029] Calculate the Euclidean distance between all points in the actual path A and the ideal path B, and generate the distance matrix D. The formula is: ; i=1,2,…,m represents the point index in the actual path, j=1,2,…,n represents the point index in the ideal path; define C(i,j) as the cumulative minimum distance of the optimal alignment path from (1,1) to (i,j), recursive formula: ; In the formula, The actual path point Align ideal path points ; The actual path point Align ideal path points ; The actual path point Align ideal path points , boundary conditions: C(1,1)=D(1,1), (accumulated along the first column), (accumulate along the first row); start from C(m,n), trace back to C(1,1), determine the optimal alignment path between the actual path and the ideal path, select the path based on the cumulative minimum distance, and record the point pairs aligned at each step ; Calculate the spray path accuracy deviation value, the expression is: ; Where L is the number of point pairs contained in the alignment path, is the cumulative distance of the optimal alignment path, and AGS is the spray path accuracy deviation value.

[0030] The larger the deviation value of the spray path accuracy, the more serious the spray robot deviates from the ideal trajectory, and there may be health problems. For example, a large deviation value may be caused by wear of mechanical parts, sensor failure, or unstable control system. In this case, it is difficult for the robot to accurately execute the predetermined path, which may cause uneven spraying and affect product quality. Continuous large deviations will not only increase rework costs, but may also cause further damage or downtime to the equipment.

[0031] On the contrary, the smaller the spray path accuracy deviation value, the closer the spray robot is to the ideal trajectory and its health is good. This shows that the robot components are operating normally and the control system can efficiently respond to changes in the external environment and maintain stability. Low deviation values ​​usually reflect proper equipment maintenance, excellent sensor performance, and high reliability of the control algorithm, thus ensuring uniformity and consistency of spray quality.

[0032] Trend analysis of spray thickness consistency data, including: De-noising: Use filtering technology (such as median filtering or Gaussian filtering) to remove measurement noise and ensure data quality. Normalization: Normalize thickness data collected at different locations to eliminate the impact caused by regional differences. Data grouping: Group thickness data by time period, such as by hour, shift or day, to facilitate subsequent trend analysis.

[0033] Calculate the average value of the spraying thickness in each time period to observe the change trend of the overall spraying thickness level: Calculate the standard deviation of the thickness in each time period to reflect the uniformity of the spraying thickness: Define the thickness drift rate as the rate of change of the average thickness over a continuous period of time: plot the time series curve of the average thickness and standard deviation, and observe its trend over time. Check whether there is periodic fluctuation or long-term drift.

[0034] Perform linear regression or polynomial regression on thickness data, fit trend lines, analyze overall change trends, and set normal ranges for thickness. Mark time points that exceed the normal range as abnormal points and analyze whether there are fluctuations or sudden drifts. Use sliding windows to calculate local average thickness and standard deviation to track local change trends in thickness consistency.

[0035] After trend analysis of the spraying thickness consistency data, the thickness drift velocity abnormal value is calculated. The calculation method of the thickness drift velocity abnormal value is: Enter the thickness drift velocity sequence , represents the thickness drift velocity data at each time point, select the neighborhood size k, usually the value of k is between 5-50, calculate the k distance, calculate the k distance, for the i-th data point The k distance of is the distance to its k-neighborhood boundary points, It is the distance The kth closest point; ; The k-neighborhood includes all points satisfy: ; For any point and , and its reachable distance is: ; The local reachability density reflects the point Relative to the density of its k-neighbors , the expression is: ; In the formula, for The number of points in the neighborhood of ; calculate the thickness drift velocity anomaly, the expression is: ; Where EMP is the thickness drift velocity anomaly value.

[0036] The larger the thickness drift velocity abnormality, the more serious problems may occur in the health of the spray robot. This usually indicates that there are abnormal and violent fluctuations in the spraying process, such as a blocked spray head, loose mechanical parts, or a malfunctioning control system. This drastic drift velocity abnormality will lead to extremely uneven coating thickness, which will directly affect product quality and may increase the rework rate. In addition, continued high abnormal values ​​may also cause increased equipment wear or production safety hazards, and the equipment needs to be checked and maintained immediately.

[0037] On the contrary, the smaller the thickness drift speed anomaly value, the better the health of the spray robot and the stable spray process. Low anomaly values ​​indicate that the thickness drift changes smoothly and the spray system can effectively adapt to environmental fluctuations and maintain uniform coating quality. In this case, the components of the spray robot are operating normally, the control algorithm performs well, the equipment does not require additional maintenance, and can maintain efficient production.

[0038] The anomaly detection module, based on the machine learning model, comprehensively calculates the spray path accuracy deviation value and thickness drift speed anomaly value, generates a health score based on the calculation results, evaluates the health status of the spray robot, and predicts potential failures.

[0039] The spray path accuracy deviation value and thickness drift speed anomaly value are converted into a comprehensive feature vector, and the comprehensive feature vector is used as the input of the machine learning model. The machine learning model predicts the health score value label of the health status of the spray robot with each group of comprehensive feature vectors as the prediction target, and minimizes the sum of the prediction errors of the health score value labels of the health status of all spray robots as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The health score value of the health status of the spray robot is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0040] The method for obtaining the health score value of the health status of the spraying robot is: from the comprehensive feature vector training data of the trained machine learning model, obtain the corresponding function expression: ; In the formula, is the output function of the model, AGS is the spray path accuracy deviation value, EMP is the thickness drift velocity abnormal value, The health score value of the spray robot's health status.

[0041] The value range of the health score value Y of the health status of the spray robot can be set according to business needs. The common range is: Y=100: the health status is the best and no maintenance is required; Y=50 to 70: the health status is moderate and regular inspection is required; Y<50: the health status is poor and potential faults exist, requiring immediate maintenance.

[0042] If the health score value Y shows a downward trend, it means that the spray robot may have gradually deteriorating components or control problems. Combined with time series analysis, the time window for the occurrence of failures can be predicted and prevented in advance.

[0043] An early warning module is used to generate a real-time early warning signal when the fault prediction result exceeds a set threshold; The acquired health score value of the health status of the spray robot is compared with the health score value reference threshold pre-set based on historical data. If the health score value of the health status of the spray robot is less than the health score value reference threshold, an early warning signal is generated; if the health score value of the health status of the spray robot is greater than or equal to the health score value reference threshold, no early warning signal is generated.

[0044] Warning signal types include: Visual warning: Intuitively convey warning information through indicator lights, screen displays, and color coding (such as red). Sound warning: Remind operators through buzzers or voice broadcasts. Digital warning: Transmit warning signals to the central control system or remote monitoring platform.

[0045] The control module is used to predict the changing trend of the health status of the spray robot, dynamically adjust the control parameters of the spray robot according to the prediction results to correct the spray path and thickness, and trigger maintenance recommendations.

[0046] Predict the changing trend of the health score of the spraying robot, the formula is: ; In the formula, is the predicted change of the health score at time t, For the moment The health score of is the spraying path accuracy deviation value at time t, is the thickness drift velocity anomaly at time t, e is the weight parameter set of the control model, which is determined by fitting historical data. f is the prediction function, usually a linear regression or machine learning model (such as polynomial regression, LSTM). When <0 and the absolute value is large, it indicates that the health status has deteriorated and parameters need to be adjusted or maintenance needs to be triggered.

[0047] The motion trajectory parameters of the spraying robot are dynamically adjusted through the control module. The formula is: ; In the formula, is the path adjustment amount, which is used to correct the robot's motion trajectory. is the instantaneous rate of change of path deviation, reflecting the dynamic change of the path. , , It is the proportional, integral and differential control coefficients (PID control parameters), which are optimized according to the spraying task.

[0048] Dynamically adjust the spray flow Q and spray speed v to correct the thickness deviation. The formula is: ; ; In the formula, is the spray flow adjustment amount, which is used to correct the thickness change, Δv is the spray speed adjustment amount, which is used to avoid overspray or underspray, and α, β, γ are the coefficients of the control model, which are optimized through experiments.

[0049] The conditions for triggering maintenance recommendations are based on the comparison of health score values ​​and parameter thresholds. When the health score value of the spray robot is lower than the set health score threshold, or the spray path accuracy deviation value exceeds the maximum allowable deviation range, or the thickness drift speed abnormal value exceeds the set limit, the system will trigger maintenance recommendations to remind the operator to check and maintain. If all the above parameters are within the normal range, the system will not trigger maintenance recommendations and the robot will continue to operate in its current state.

[0050] In this embodiment, the data acquisition module is responsible for acquiring the spray path accuracy data and the spray thickness consistency data. The data processing module analyzes the path data to calculate the spray path accuracy deviation value, and performs trend analysis on the thickness data to calculate the thickness drift velocity anomaly. The anomaly detection module comprehensively evaluates the above key indicators based on the machine learning model, generates a health score, and predicts the health status and potential faults of the spray robot. The early warning module sends a warning signal in real time when the fault prediction exceeds the set threshold. The control module dynamically adjusts the control parameters of the spray robot to correct the path and thickness by predicting the trend of health status changes, and triggers maintenance recommendations, thereby ensuring equipment stability and production quality.

[0051] Embodiment 2: An industrial control device described in this embodiment includes: Integrated processors to perform data acquisition, processing, anomaly detection, and fault prediction functions; Communication interface, used for data exchange with the spraying robot and external monitoring equipment; Storage unit, used to store historical operation data, prediction models and warning records.

[0052] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0053] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0054] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. An industrial control system, characterized in that: It includes data acquisition module, data processing module, abnormality detection module, early warning module and control module; A data acquisition module is used to obtain the spraying path accuracy data and spraying thickness consistency data during the operation of the spraying robot; A data processing module is used to analyze the spraying path accuracy data and calculate the spraying path accuracy deviation value, and to perform trend analysis on the spraying thickness consistency data and calculate the thickness drift velocity abnormal value; The anomaly detection module comprehensively calculates the spray path accuracy deviation value and thickness drift speed anomaly value based on the machine learning model, generates a health score based on the calculation results, evaluates the health status of the spray robot, and predicts potential failures; An early warning module is used to generate a real-time early warning signal when the fault prediction result exceeds a set threshold; The control module is used to predict the changing trend of the health status of the spray robot, dynamically adjust the control parameters of the spray robot according to the prediction results to correct the spray path and thickness, and trigger maintenance recommendations.

2. An industrial control system according to claim 1, characterized in that: In the data processing module, the spraying path accuracy data is analyzed and then the spraying path accuracy deviation value is calculated. The calculation method of the spraying path accuracy deviation value is: Enter the actual path data sequence , each point Represents the coordinates of the spray head in three-dimensional space ; Ideal path data sequence , each point Indicates the reference coordinates in the preset path ; Calculate the Euclidean distance between all points in the actual path A and the ideal path B, and generate the distance matrix D. The formula is: ; i=1,2,…,m represents the point index in the actual path, j=1,2,…,n represents the point index in the ideal path; define C(i,j) as the cumulative minimum distance of the optimal alignment path from (1,1) to (i,j), recursive formula: ; In the formula, The actual path point Align ideal path points ; The actual path point Align ideal path points ; The actual path point Align ideal path points , boundary conditions: C(1,1)=D(1,1), , ; Starting from C(m,n), backtrack to C(1,1), determine the optimal alignment path between the actual path and the ideal path, select the path based on the cumulative minimum distance, and record the point pairs aligned at each step. ; Calculate the spray path accuracy deviation value, the expression is: ; Where L is the number of point pairs contained in the alignment path, is the cumulative distance of the optimal alignment path, and AGS is the spray path accuracy deviation value.

3. An industrial control system according to claim 2, characterized in that: In the data processing module, the thickness drift velocity abnormal value is calculated after trend analysis of the spraying thickness consistency data. The calculation method of the thickness drift velocity abnormal value is: Enter the thickness drift velocity sequence , represents the thickness drift velocity data at each time point, selects the neighborhood size k, calculates the k distance, and is the i-th data point The k-distance of is the distance to its k-neighborhood boundary points: ; In the formula, represents the k distance, It is the distance The kth closest point; ; The k-neighborhood includes all points satisfy: ; For any point and , and its reachable distance is: ; The local reachability density reflects the point For the density of its k-neighborhood , the expression is: ; In the formula, for The number of points in the neighborhood of ; calculate the thickness drift velocity anomaly, the expression is: ; Where EMP is the thickness drift velocity anomaly value.

4. An industrial control system according to claim 3, characterized in that: In the anomaly detection module, based on the machine learning model, the spray path accuracy deviation value and the thickness drift speed anomaly value are comprehensively calculated, and a health score is generated according to the calculation results to evaluate the health status of the spray robot. Specifically, the spray path accuracy deviation value and the thickness drift speed anomaly value are converted into a comprehensive feature vector, and the comprehensive feature vector is used as the input of the machine learning model. The machine learning model uses each group of comprehensive feature vectors to predict the health score value label of the health status of the spray robot as the prediction target, and minimizes the sum of the prediction errors of the health score value labels of the health status of all spray robots as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence and the model training is stopped. The health score value of the health status of the spray robot is determined according to the model output result, wherein the machine learning model is a polynomial regression model; Y=100: The health status is the best and no maintenance is required; Y=50 to 70: The health status is moderate and regular inspection is required; Y<50: The health status is poor and potential faults exist and immediate maintenance is required.

5. An industrial control system according to claim 4, characterized in that: The early warning module generates a real-time early warning signal when the fault prediction result exceeds the set threshold; The obtained health score value of the health status of the spraying robot is compared with the health score value reference threshold value pre-set according to historical data. If the health score value of the health status of the spraying robot is less than the health score value reference threshold value, an early warning signal is generated at this time; If the health score value of the health status of the spray robot is greater than or equal to the health score value reference threshold, no warning signal is generated.

6. An industrial control system according to claim 1, characterized in that: In the control module, the health status change trend of the spraying robot is predicted, and the control parameters of the spraying robot are dynamically adjusted according to the prediction results to correct the spraying path and thickness. Specifically: Predict the changing trend of the health score of the spraying robot, the formula is: ; In the formula, is the predicted change of the health score at time t, For the moment The health score of is the spraying path accuracy deviation value at time t, is the thickness drift velocity abnormal value at time t, e is the weight parameter set of the control model, and f is the prediction function. When <0 and the absolute value is large, it indicates that the health status has deteriorated and the parameters need to be adjusted.

7. An industrial control system according to claim 6, characterized in that: The motion trajectory parameters of the spraying robot are dynamically adjusted through the control module. The formula is: ; In the formula, is the path adjustment amount, which is used to correct the robot's motion trajectory. is the instantaneous rate of change of path deviation, reflecting the dynamic change of the path. , , The proportional, integral and differential control coefficients are optimized according to the spraying task; Dynamically adjust the spray flow Q and spray speed v to correct the thickness deviation. The formula is: ; ; In the formula, is the change rate of thickness drift velocity, is the spray flow adjustment, which is used to correct the thickness change, Δv is the spray speed adjustment, and α, β, γ are the coefficients of the control model.

8. An industrial control device, used to implement the function of the industrial control system according to any one of claims 1 to 7, the device comprising: Integrated processors to perform data acquisition, processing, anomaly detection, and fault prediction functions; Communication interface, used for data exchange with the spraying robot and external monitoring equipment; Storage unit, used to store historical operation data, prediction models and warning records.

Images