A method for dynamic adjustment of machine parameters based on a closed-loop feedback mechanism for dust removal effect

Abstract

This invention specifically relates to the field of multimodal sensing technology, and discloses a method for dynamic adjustment of machine parameters based on a closed-loop feedback mechanism for dust removal effect. The method includes: S1: area determination and baseline parameter calibration; S2: real-time multimodal data acquisition; S3: comprehensive dust removal demand analysis; S4: dust removal execution and real-time status feedback; S5: dust removal effect verification; and S6: scheme optimization and equipment health management. By linking a high-definition industrial camera and a laser dust sensor in the sensing terminal, and combining environmental data from temperature and humidity sensors, a three-dimensional monitoring system is formed, encompassing surface condition, air particles, and environmental conditions. Based on the necessary dust removal values, the system determines whether to initiate operation, realizing a real-time dynamic monitoring mechanism based on multimodal fusion sensing. A dynamic adjustment mechanism is constructed, introducing an efficiency factor as the core quantitative indicator of equipment energy efficiency to objectively reflect the equipment energy efficiency of a single operation. The system also calculates the equipment health index for equipment maintenance, improving the real-time performance and accuracy of monitoring.

Description

Technical Field

[0001] This invention relates to the field of multimodal sensing technology, and more specifically, to a method for dynamic adjustment of machine parameters based on a closed-loop feedback mechanism for dust removal effect. Background Technology

[0002] In scenarios with clear requirements for environmental cleanliness, such as industrial production, warehousing management, and precision instrument maintenance, dust removal is a key link in ensuring production quality, equipment lifespan, and the health of workers. However, existing dust removal technology management systems have multiple shortcomings and are unable to meet the comprehensive needs of dust removal management in modern scenarios.

[0003] Existing dust removal technologies mostly rely on simple image observation to evaluate the effect after dust removal. Dust removal decisions are based on human intervention or thresholds, and dust removal equipment is maintained afterward. However, there are still some drawbacks in their use. First, the monitoring dimensions are limited. Existing dust removal systems generally rely on the post-event evaluation mode, which cannot dynamically perceive the dust removal effect during operation. Moreover, most dust removal systems only observe the surface dust coverage, failing to achieve coordinated monitoring of macroscopic surface conditions and microscopic air quality. In some scenarios, manual intervention is required to judge cleanliness, which affects the accuracy of the evaluation results. The reliance on human intervention is high, and the dust removal monitoring effect is relatively lagging, lacking real-time performance and accuracy.

[0004] The dust removal system lacks scientific basis. The start-up, shutdown and parameter adjustment of the existing system are mostly based on fixed thresholds or simple time-sequence control. The system only uses whether dust is detected as the start-up basis, without comprehensively considering the ambient temperature and humidity and the current status of the equipment. The decision-making dimensions are relatively limited. The power, working time and other parameters of the dust removal equipment are mostly preset fixed values, which cannot be dynamically adjusted according to the actual degree of pollution and the size of the area, resulting in insufficient safety risk management.

[0005] Third, there is a lack of preventative maintenance mechanisms. Existing dust removal systems focus on completing dust removal tasks and tend to neglect the monitoring and management of the health status of the dust removal equipment itself. They cannot quantify the degree of equipment performance degradation through operational data and only carry out repairs when the equipment is completely unusable. Furthermore, there is a lack of energy efficiency correlation analysis, which leads to increased equipment operation and maintenance costs and a greater risk of sudden failures. Summary of the Invention

[0006] In view of this, embodiments of the present invention provide a method for dynamic adjustment of machine parameters based on a closed-loop feedback mechanism for dust removal effect. By combining machine vision with dust sensors through multimodal fusion perception, and by introducing equipment health management, closed-loop learning optimization, and safety finite principle, the method effectively solves the problems of single monitoring dimension, lack of scientific basis for dust removal, and lack of preventive maintenance mechanism mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for dynamic adjustment of machine parameters based on a closed-loop feedback mechanism for dust removal effect, comprising a sensing terminal, an edge computing terminal, a cloud server cluster, and an execution terminal, the specific steps of which are as follows:

[0008] S1: Area Determination and Baseline Parameter Calibration: Identify the target dust removal physical area through the system software map of the edge computing terminal, collect initial dust-free parameters, and preset core threshold parameters;

[0009] S2: Multimodal real-time data acquisition: Multi-dimensional dust removal related data are collected synchronously through the sensing terminal and transmitted to the edge computing terminal to build a multi-dimensional dust removal related dataset, which is then temporarily stored in the local database;

[0010] S3: Comprehensive Dust Removal Requirements Analysis: The edge computing terminal calculates the necessary dust removal values ​​based on a multi-dimensional dust removal related dataset and formulates a safe startup strategy based on the necessary dust removal values;

[0011] S4: Dust removal execution and real-time status feedback: The edge computing terminal converts the security startup policy into a startup command and transmits it to the execution terminal. The execution terminal performs the dust removal action and triggers the dust removal effect verification.

[0012] S5: Dust removal effect verification: The sensing terminal collects multi-dimensional dust removal correlation data of the target dust removal physical area in real time after the dust removal action is completed, and transmits it to the edge computing terminal to calculate the dust removal effect value.

[0013] S6: Solution Optimization and Equipment Health Management: The edge computing terminal optimizes the dust removal solution based on the dust removal effect value and historical data in the cloud server cluster, and performs equipment health management.

[0014] The technical effects and advantages of this invention are as follows:

[0015] 1. This invention uses a high-definition industrial camera and a laser dust sensor in a sensing terminal to macroscopically quantify the surface dust coverage of the target dust removal physical area, reflecting the microscopic air quality. Combined with environmental data from temperature and humidity sensors, it forms a three-dimensional monitoring of surface condition, air particles, and environmental conditions. Based on the necessary dust removal values, it determines whether to start the operation, realizing a real-time dynamic monitoring mechanism of multi-modal fusion sensing, and improving the real-time performance and accuracy of monitoring.

[0016] 2. This invention uses an edge computing terminal to convert image difference, air dust concentration, and ambient temperature and humidity into necessary dust removal values, prioritizing environmental safety as the highest decision priority. This solves the problem of neglecting safety constraints in existing methods. Furthermore, during parameter adjustment, it combines the dust removal necessity values ​​with the decision to call the preset dust removal scheme or dynamically optimize parameters in the local database, thus constructing a dynamic adjustment mechanism that effectively avoids energy waste and incomplete dust removal.

[0017] 3. This invention introduces an efficiency factor as the core quantitative indicator of equipment energy efficiency, which is linked to dust removal effect, energy consumption and running time, to objectively reflect the equipment energy efficiency of a single operation. It also uses a moving average algorithm to calculate the equipment health index and performs equipment maintenance based on the equipment health index, which effectively avoids production interruptions caused by equipment failure and reduces energy consumption and operation and maintenance costs. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the method steps of the present invention.

[0019] Figure 2 This is a schematic diagram of the overall structure of the present invention.

[0020] Figure 3 This is a schematic diagram illustrating the steps involved in developing a secure boot strategy according to the present invention. Detailed Implementation

[0021] 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.

[0022] like Figure 1 The method for dynamically adjusting machine parameters based on a closed-loop feedback mechanism for dust removal effect is shown, including a sensing terminal, an edge computing terminal, a cloud server cluster, and an execution terminal.

[0023] In a more specific application of this invention, the sensing terminal is used to collect environmental conditions, dust characteristics, and equipment operation data of the target area. It includes a high-definition industrial camera, a laser dust sensor, a temperature and humidity sensor, and a vibration sensor. The high-definition industrial camera is used to collect image data of the target area, the laser dust sensor is used to quantitatively monitor the concentration of PM2.5 in the air of the target area, the temperature and humidity sensor is used to monitor the ambient temperature and relative humidity in real time, and the vibration sensor is installed on the fan housing / bearing of the dust removal equipment to monitor vibration acceleration. The sensing terminal is connected to the edge computing terminal through the industrial internet to provide raw data support for subsequent decision-making.

[0024] Edge computing terminals are used to receive data from sensing terminals, execute algorithm calculations, generate control commands, and store local policies and temporary data. They include an edge computing gateway and a local database. The edge computing gateway is used for data processing, algorithm execution, and command generation and distribution, while the local database is used to store local policies and historical data. Edge computing terminals are connected to a cloud server cluster via the Internet.

[0025] The cloud server cluster is responsible for long-term data storage, big data analysis, and model iteration. It builds a cloud database, stores all historical data, mines the correlation between dust removal parameters, environmental factors, and dust removal effects through big data, and continuously optimizes decision-making algorithms and equipment aging models. The cloud server cluster is connected to edge computing terminals via the Internet and accesses the factory intranet via dedicated lines.

[0026] The execution terminal receives instructions from the edge computing terminal and performs actions such as dust removal, cooling, and alarm activation, translating decisions into actual intervention. The execution terminal includes an intelligent dust removal robot, a fixed dust removal arm, an industrial air conditioner, and an audible and visual alarm. The intelligent dust removal robot receives instructions from the edge computing terminal, drives its walking mechanism to the target area, starts the dust removal operation, and provides real-time feedback on its operating status. The industrial air conditioner receives instructions to start cooling, sets the target temperature, and provides real-time feedback on the current temperature. The audible and visual alarm receives instructions such as high temperature warnings and low equipment health, alerting staff to handle abnormalities. The execution terminal connects to the edge computing terminal via a 5G network / RS485 interface.

[0027] For the connection methods of the aforementioned sensing terminals, edge computing terminals, cloud server clusters, and execution terminals, please refer to [link / reference]. Figure 2 .

[0028] The specific embodiments of the present invention include the following steps:

[0029] S1: Area Determination and Baseline Parameter Calibration: Identify the target dust removal physical area through the system software map of the edge computing terminal, collect initial dust-free parameters, and preset core threshold parameters;

[0030] Furthermore, the steps for identifying the target dust removal physical area are as follows:

[0031] S1.1: Control the high-definition industrial camera in the sensing terminal through the edge computing terminal to acquire panoramic images of the target dust removal scene, and transmit them to the edge computing terminal to generate a panoramic image of the target dust removal scene, automatically distinguishing between areas that need dust removal and areas that do not need dust removal.

[0032] In this embodiment, it is important to note that during the panoramic image acquisition process, it is necessary to ensure no blind spots according to the preset path, take one image every 1 meter, and record the shooting coordinates of each image. The generation of panoramic images requires the edge computing terminal to execute an image stitching algorithm to extract the overlapping area features of adjacent images and eliminate image distortion and overlap redundancy. Combining the shooting coordinate data, a two-dimensional electronic map corresponding to the target dust removal scene at a 1:1 scale is constructed in the system software map. Based on the two-dimensional electronic map, the edge computing terminal distinguishes the dust removal area and the non-dust removal area through an image recognition algorithm, where the image recognition algorithm can be contour detection and color recognition.

[0033] S1.2: Manually select the target physical area for dust removal within the area to be dusted, automatically capture the physical coordinates of the boundary corresponding to the selected area, take 3-5 high-resolution local images of the manually selected area, obtain feature labels, optimize the area boundary, and assign a unique name to each target physical area for dust removal.

[0034] In this embodiment, it should be specifically noted that manually selecting the target dust removal physical area requires the operator to do so through the software interface of the edge computing terminal. The unique image features of the target dust removal physical area are extracted as feature tags. The image features can be the texture pattern of the workstation floor, the shape of the equipment base, or the corner angle of the wall. The area boundary is optimized by comparing the local high-definition image and the panoramic image, eliminating redundant areas at the boundary, and correcting the boundary offset caused by manual operation. After optimization, the edge computing terminal stores the final boundary coordinates of the target dust removal physical area in the local database, which is convenient for subsequent dust removal equipment positioning and area matching.

[0035] S1.3: The edge computing terminal retrieves parameter information of all execution terminals in the local database, filters the execution terminals that are compatible with the target dust removal physical area, creates a target dust removal physical area-execution terminal ID association mapping table, and verifies the association validity.

[0036] In this embodiment, it should be specifically noted that the execution terminal can be an intelligent dust removal robot or a fixed dust removal arm. The parameter information includes dust removal radius, working range, and mobility. The execution terminal parameters are compared with the size and characteristics of the target dust removal physical area to select a suitable execution terminal. A target dust removal physical area-execution terminal ID association mapping table is created. For example, workstations 1-3 in workshop A are bound to intelligent dust removal robot ID001, and the robot's preset working path for this area is entered. At the same time, the association effective time and terminal status are recorded to ensure that the bound execution terminal is in an available state.

[0037] It needs to be specifically explained that the association validity verification refers to the edge computing terminal sending a regional positioning test command to the bound execution terminal. The robot moves to the starting point of the target dust removal physical area and feature labels and takes a local image. The captured features are compared with the feature labels stored in the edge computing terminal. If the matching degree is ≥95%, the association is determined to be valid. After the verification is passed, the edge computing terminal synchronously stores the mapping table in the local database and the cloud server cluster, and issues a dedicated regional identifier to the execution terminal to ensure that the execution terminal only responds to the dust removal command adapted to the target dust removal physical area.

[0038] Furthermore, the initial cleanroom parameters refer to the initial cleanroom parameter baseline vector B, and the core threshold parameters include the ambient temperature safety threshold Tmax and the dust removal necessary value start-up threshold N. T And the target value of dust removal effect E T .

[0039] In this embodiment, it should be specifically explained that the initial cleanroom parameter acquisition is achieved by controlling the high-definition industrial camera in the sensing terminal through the edge computing terminal to take 5-10 effective images of the confirmed clean target area under multiple angles and lighting conditions. The feature vector of each image is extracted through the edge computing terminal, and the average value of the vectors is calculated to obtain the initial cleanroom parameter benchmark vector B, which serves as the standard for subsequent dust removal effect evaluation and verification.

[0040] S2: Multimodal real-time data acquisition: Multi-dimensional dust removal related data are collected synchronously through the sensing terminal and transmitted to the edge computing terminal to build a multi-dimensional dust removal related dataset, which is then temporarily stored in the local database;

[0041] Furthermore, the multi-dimensional dust removal-related data includes image data, environmental data, and equipment data, among which the image data is real-time image feature vector R. img Environmental data includes real-time dust concentration P m Temperature T and humidity H; equipment data includes equipment operating vibration acceleration A and the cumulative operating time t of the dust removal equipment. r .

[0042] In this embodiment, it is necessary to specifically explain that multi-dimensional dust removal-related data are collected by the sensing terminal at a preset cycle of 30 seconds / time; real-time images of the target dust removal physical area are captured by a high-definition industrial camera to generate real-time image feature vectors; PM2.5 / PM1O concentrations in the air are quantitatively monitored by a laser dust sensor; ambient temperature and humidity are collected by a temperature and humidity sensor; and the vibration acceleration of the dust removal equipment is monitored by installing a vibration sensor on the dust removal equipment fan, and the cumulative running time is synchronously uploaded by the dust removal equipment controller.

[0043] It should be further explained that the high-definition industrial camera in the sensing terminal is connected to the edge computing terminal via gigabit Ethernet, the dust / temperature and humidity sensor is connected via RS485 bus, and the vibration sensor is connected via analog lines. The edge computing terminal preprocesses the received multi-dimensional dust removal related data, removes outliers, obtains a multi-dimensional dust removal related dataset, and temporarily stores it in the local database of the edge computing terminal.

[0044] S3: Comprehensive Dust Removal Requirements Analysis: The edge computing terminal calculates the necessary dust removal values ​​based on a multi-dimensional dust removal related dataset and formulates a safe startup strategy based on the necessary dust removal values;

[0045] Furthermore, obtaining the necessary dust removal values ​​requires setting a time window and acquiring the initial dust-free parameter baseline vector B of the target dust removal physical area stored in the local database within the target time window, as well as the real-time image feature vector R from the multi-dimensional dust removal correlation dataset. imgSubstitute the initial cleanroom parameter baseline vector and the real-time image feature vector into the formula:

[0046] ,

[0047] The image difference D was calculated. img To obtain the air dust concentration and ambient temperature and humidity from the multi-dimensional dust removal correlation dataset in the target dust removal physical area within the target time window, and to combine this with the image difference into the formula:

[0048] ,

[0049] The necessary value N for dust removal is calculated, where f(T,H) is the environmental constraint function, and P... m_max This represents the maximum allowable air dust concentration in an industrial setting. a1, a2, and a3 represent the weighting coefficients for image difference, air dust concentration, and environmental temperature and humidity constraints, respectively, with a total of 1.

[0050] In this embodiment, it should be specifically explained that the image difference degree is used to quantify the surface dust coverage of the target dust removal physical area. The initial dust-free parameter reference vector B contains the image features of the target dust removal physical area under clean conditions. The real-time image feature vector should be consistent with the dimension of the initial dust-free parameter reference vector. The image difference degree is the core foundation of the necessary dust removal value and is used to convert the surface dust degree from visual subjective judgment to a quantitative value in the range of [0,1] to avoid human observation error.

[0051] It should be further explained that the weighting coefficients are used to adjust the degree of influence of each parameter on the necessary value of dust removal. In industrial scenarios, the preset weighting coefficients are a1=0.4, a2=0.3 and a3=0.3. Dust removal should prioritize surface cleanliness, then focus on air quality and environmental safety. The weighting coefficients can be dynamically adjusted according to the requirements of the dust removal task. The final output of N is a specific value.

[0052] f(T,H) is the environmental constraint function, with temperature T as the core of safety control and humidity H as an auxiliary reference. The rules are set as follows:

[0053] When T≤Tmax, where Tmax is the maximum safe operating temperature of the equipment in an industrial setting, such as 40℃, f(T,H)=1, indicating that the environment is safe and there are no additional constraints on the dust removal decision.

[0054] When T > Tmax, f(T,H) increases sharply to 10. The environmental safety constraints are amplified by weight a3 to prevent the dust removal equipment from malfunctioning under high temperature.

[0055] Furthermore, such as Figure 3 As shown, the steps for formulating a secure boot strategy are as follows:

[0056] S2.1: Clearly define security constraint priorities, define security thresholds, formulate security threshold triggering rules, preset security response instructions, and match the execution terminals corresponding to the security response instructions;

[0057] In this embodiment, it is important to specify that the safety constraint priority refers to temperature as the highest priority safety constraint. Excessive temperature can easily lead to overheating and damage to the dust removal equipment, and may even cause safety hazards. The temperature safety threshold Tmax in the safety threshold is defined as 40℃, which can be flexibly adjusted according to the dust removal equipment model and the physical environment of the target dust removal area. The safety threshold triggering rule means that when the temperature T collected in real time by the sensing terminal is greater than Tmax, the highest level of safety response is triggered, regardless of whether the necessary dust removal value is met, environmental safety is prioritized. When T≤Tmax, the safety constraint is released. The safety response command refers to the preset safety response command for the scenario where T>Tmax, including the dust removal prohibition command, the cooling start command, and the alarm command. The dust removal prohibition command corresponds to all dust removal execution terminals, the cooling start command corresponds to the industrial air conditioner in the execution terminal, specifying the target temperature, and the alarm command corresponds to the audible and visual alarm in the execution terminal, with the alarm mode set.

[0058] S2.2: Formulate dust removal demand judgment rules, match dust removal solutions in the local database of the edge computing terminal based on the dust removal necessity value range, associate the dust removal solution with the execution terminal ID, and set a verification mechanism;

[0059] In this embodiment, it should be specifically noted that the formulation of the dust removal demand judgment rule requires obtaining the dust removal necessity value start threshold N from the local database of the edge computing terminal. T The necessary threshold for dust removal can be set to 0.5, referencing industrial cleanliness requirements. For scenarios with high cleanliness requirements, this can be lowered to 0.4, and for scenarios with lower requirements, it can be raised to 0.6. When T≤Tmax, if N≥N T This indicates that the pollution level of the target dust removal physical area has reached the required dust removal level, triggering the dust removal start-up process. If N < N T No dust removal is required.

[0060] The steps for matching a dust removal solution are as follows:

[0061] When 0.5≤N<0.7, it is considered mild pollution. A low-power solution is recommended, with the robot power at 60% and the duration at 3 minutes.

[0062] When 0.7 ≤ N < 0.9, it is considered moderate pollution. The standard solution is used, with the robot at 70% power and a duration of 5 minutes.

[0063] When N≥0.9, it is considered heavily polluted, so a high-intensity solution is used, with the robot power at 90% and the duration at 8 minutes.

[0064] Associate the dust removal scheme with the execution terminal ID to ensure that each dust removal scheme corresponds to a unique dust removal execution terminal, avoiding poor dust removal performance due to mismatch between the terminal and the dust removal scheme; the verification mechanism refers to the condition that N≥N T At that time, the edge computing terminal retrieves the basic scheme from the local database and automatically verifies the status of the execution terminal. If the terminal is available, it generates a dust removal start command containing the terminal ID, power, duration, and operation path. If the terminal is unavailable, it calls other execution terminals.

[0065] S2.3: Based on the safety threshold triggering rules and dust removal demand judgment rules, clarify the low power monitoring triggering conditions, filter unnecessary sensing terminals and adjust the collection cycle.

[0066] In this embodiment, it should be specifically noted that the low-power monitoring trigger condition refers to when T≤Tmax and N<N T When there is no need to activate dust removal and safety response, it enters a low-power monitoring state. Screening non-essential sensing terminals requires analyzing the functions and energy consumption of the sensing terminals. High-definition industrial cameras need to continuously shoot, resulting in high energy consumption. Laser dust sensors need to collect data in real time, resulting in medium energy consumption. Vibration sensors only need to be monitored when the equipment is running, resulting in low energy consumption. Adjusting the acquisition cycle specifically means turning off non-essential acquisition, retaining essential monitoring, and setting up a wake-up mechanism. When the temperature rises suddenly or the dust concentration increases suddenly in the low-power state, all sensing terminals are immediately woken up, and the normal acquisition cycle is restored.

[0067] S4: Dust removal execution and real-time status feedback: The edge computing terminal converts the security startup policy into a startup command and transmits it to the execution terminal. The execution terminal performs the dust removal action and triggers the dust removal effect verification.

[0068] Furthermore, the start command includes terminal identification information, operation parameter information, and safety control information. After receiving the start command, the terminal performs a self-check of the equipment status, executes the dust removal operation in stages, and automatically generates an operation completion signal after the dust removal operation is completed, which is then transmitted to the edge computing terminal.

[0069] In this embodiment, it should be specifically explained that the terminal identification information includes the unique ID of the execution terminal and the terminal type; the operation parameter information includes the operation path coordinates, the height of the suction port, and the operation interval; the safety control information includes the safety threshold and the emergency pause command triggering conditions; the equipment status self-check refers to checking the remaining power of the lithium battery of the dust removal equipment, the operating status of the fan, and the navigation system; the dust removal operation is divided into three stages: the start-up stage, the stable operation stage, and the closing stage. In the start-up stage, the execution terminal needs to start the equipment step by step according to the start command and start the walking mechanism to bring the execution equipment close to the target dust removal physical area; in the stable operation stage, the execution terminal maintains 70% fan power operation, moves according to the preset path, collects the process operation parameters in real time, and uploads them to the edge computing terminal once every 5 seconds; in the closing stage, when the operation is close to the preset duration, the execution terminal gradually reduces the fan power, returns to the starting position of the operation, and shuts down the equipment.

[0070] S5: Dust removal effect verification: The sensing terminal collects multi-dimensional dust removal correlation data of the target dust removal physical area in real time after the dust removal action is completed, and transmits it to the edge computing terminal to calculate the dust removal effect value.

[0071] Furthermore, the multi-dimensional dust removal correlation data after dust removal includes the image feature vector E after dust removal. img and the concentration of air dust P after dust removal MB Obtaining the dust removal effect value requires based on the feature vector E of the image after dust removal. img Substituting the initial cleanroom parameter reference vector B into the formula:

[0072] ,

[0073] The dust removal effect value E was calculated, and the dust removal effect value was cross-validated based on the air dust concentration after dust removal to obtain the cross-validation results.

[0074] In this embodiment, it should be specifically noted that cross-validation refers to validation based on the standardized dust concentration P in the air after dust removal. MB The dust removal efficiency value E was cross-validated using the following steps:

[0075] If E ≥ 0.85 and P MB ≤0.3, further confirming that the dust removal effect in the target physical area is good;

[0076] If E ≥ 0.85 but P MB A value >0.3 indicates that the surface of the target dust removal area is clean, but the airborne particles exceed the standard, and the cause needs to be analyzed.

[0077] If E < 0.85 but P MB ≤0.3 indicates that the air quality in the target dust removal area meets the standards, but the surface is not clean. The dust removal operation parameters need to be optimized.

[0078] S6: Solution Optimization and Equipment Health Management: The edge computing terminal optimizes the dust removal solution based on the dust removal effect value and historical data in the cloud server cluster, and performs equipment health management.

[0079] Furthermore, optimizing the dust removal solution requires judging whether the dust removal effect meets the standard based on the dust removal effect value, analyzing the reasons why the dust removal effect does not meet the standard based on the judgment results, formulating parameter adjustment rules based on the reasons analysis results, and optimizing the dust removal solution.

[0080] In this embodiment, it should be specifically explained that the dust removal effect compliance judgment refers to comparing the dust removal effect value with the dust removal effect target value. If the dust removal effect value is greater than the dust removal effect target value, the dust removal effect is judged to be compliant, the solution-effect mapping relationship is recorded, and the local database of the edge computing terminal is updated. If the dust removal effect value is less than the dust removal effect target value, the effect is judged to be non-compliant, and the solution optimization process needs to be initiated. The cause analysis results include insufficient dust removal intensity, insufficient operation time, and both insufficient. For insufficient dust removal intensity, the fan power is increased, and the increase is based on historical data. For insufficient operation time, the operation time is extended, and the extension is calculated based on the time gap ratio. For both insufficient, the power is increased and the operation time is extended.

[0081] It should be further noted that equipment health management needs to be based on the dust removal efficiency value E, the average operating power P, and the cumulative operating time t of the dust removal equipment. r Substituting into the formula D=E / (P×t) r The equipment efficiency factor D is calculated, where the average efficiency of operation is obtained through the sensing terminal. The higher the D, the better the equipment energy efficiency and the better the health status. The equipment health index is updated by the edge computing terminal using a moving average algorithm. Based on the equipment efficiency factor and the equipment health index, equipment maintenance alarm instructions are generated to perform equipment maintenance and synchronized to the database of the cloud server cluster to prevent the equipment from being assigned high-load operations until the equipment recovers its health.

[0082] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other.

[0083] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for dynamic adjustment of machine parameters based on a closed-loop feedback mechanism for dust removal effect, characterized in that, This includes sensing terminals, edge computing terminals, cloud server clusters, and execution terminals. The specific steps are as follows: S1: Area Determination and Baseline Parameter Calibration: Identify the target dust removal physical area through the system software map of the edge computing terminal, collect initial dust-free parameters, and preset core threshold parameters; S2: Multimodal real-time data acquisition: Multi-dimensional dust removal related data are collected synchronously through the sensing terminal and transmitted to the edge computing terminal to build a multi-dimensional dust removal related dataset, which is then temporarily stored in the local database; The multi-dimensional dust removal associated data includes image data, environmental data, and equipment data, where the image data is a real-time image feature vector R. img Environmental data includes real-time dust concentration P m Temperature T and humidity H; equipment data includes equipment operating vibration acceleration A and the cumulative operating time t of the dust removal equipment. r ; S3: Comprehensive Dust Removal Requirements Analysis: The edge computing terminal calculates the necessary dust removal values ​​based on a multi-dimensional dust removal related dataset and formulates a safe startup strategy based on the necessary dust removal values; The acquisition of the necessary dust removal values ​​requires setting a time window, and obtaining the initial dust-free parameter baseline vector B of the target dust removal physical area stored in the local database and the real-time image feature vector R from the multi-dimensional dust removal association dataset within the target time window. img Substitute the initial cleanroom parameter baseline vector and the real-time image feature vector into the formula: ， The image difference D was calculated. img To obtain the air dust concentration and ambient temperature and humidity from the multi-dimensional dust removal correlation dataset in the target dust removal physical area within the target time window, and to combine this with the image difference into the formula: ， The necessary value N for dust removal is calculated, where f(T,H) is the environmental constraint function, and P... m_max This represents the maximum allowable air dust concentration in an industrial setting. a1, a2, and a3 represent the weighting coefficients for image difference, air dust concentration, and environmental temperature and humidity constraints, respectively, with a total of 1. S4: Dust removal execution and real-time status feedback: The edge computing terminal converts the security startup policy into a startup command and transmits it to the execution terminal. The execution terminal performs the dust removal action and triggers the dust removal effect verification. S5: Dust removal effect verification: The sensing terminal collects multi-dimensional dust removal correlation data of the target dust removal physical area in real time after the dust removal action is completed, and transmits it to the edge computing terminal to calculate the dust removal effect value. The multi-dimensional dust removal correlation data after dust removal includes the feature vector E of the image after dust removal. img and the concentration of air dust P after dust removal MB Obtaining the dust removal effect value requires based on the feature vector E of the image after dust removal. img Substituting the initial cleanroom parameter reference vector B into the formula: ， The dust removal effect value E was calculated, and the dust removal effect value was cross-validated based on the air dust concentration after dust removal to obtain the cross-validation results. S6: Solution Optimization and Equipment Health Management: The edge computing terminal optimizes the dust removal solution based on the dust removal effect value and historical data in the cloud server cluster, and performs equipment health management.

2. The method for dynamic adjustment of machine parameters based on a closed-loop feedback mechanism for dust removal effect as described in claim 1, characterized in that: The steps for identifying the target dust removal physical area are as follows: S1.1: Control the high-definition industrial camera in the sensing terminal through the edge computing terminal to acquire panoramic images of the target dust removal scene, and transmit them to the edge computing terminal to generate a panoramic image of the target dust removal scene, automatically distinguishing between areas that need dust removal and areas that do not need dust removal. S1.2: Manually select the target physical area for dust removal within the area to be dusted, automatically capture the physical coordinates of the boundary corresponding to the selected area, take 3-5 high-resolution local images of the manually selected area, obtain feature labels, optimize the area boundary, and assign a unique name to each target physical area for dust removal. S1.3: The edge computing terminal retrieves parameter information of all execution terminals in the local database, filters the execution terminals that are compatible with the target dust removal physical area, creates a target dust removal physical area-execution terminal ID association mapping table, and verifies the association validity.

3. The method for dynamic adjustment of machine parameters based on a closed-loop feedback mechanism for dust removal effect as described in claim 1, characterized in that: The initial cleanroom parameters refer to the initial cleanroom parameter baseline vector B, and the core threshold parameters include the ambient temperature safety threshold Tmax and the dust removal necessary value start threshold N. T And the target value of dust removal effect E T .

4. The method for dynamic adjustment of machine parameters based on a closed-loop feedback mechanism for dust removal effect as described in claim 1, characterized in that: The steps for formulating the secure boot strategy are as follows: S2.1: Clearly define security constraint priorities, define security thresholds, formulate security threshold triggering rules, preset security response instructions, and match the execution terminals corresponding to the security response instructions; S2.2: Formulate dust removal demand judgment rules, match dust removal solutions in the local database of the edge computing terminal based on the dust removal necessity value range, associate the dust removal solution with the execution terminal ID, and set a verification mechanism; S2.3: Based on the safety threshold triggering rules and dust removal demand judgment rules, clarify the low power monitoring triggering conditions, filter unnecessary sensing terminals and adjust the collection cycle.

5. The method for dynamic adjustment of machine parameters based on a closed-loop feedback mechanism for dust removal effect as described in claim 1, characterized in that: The start command includes terminal identification information, operation parameter information, and safety control information. After receiving the start command, the terminal performs a self-check of the equipment status, performs dust removal operations in stages, and automatically generates an operation completion signal after the dust removal operation is completed, which is then transmitted to the edge computing terminal.

6. The method for dynamic adjustment of machine parameters based on a closed-loop feedback mechanism for dust removal effect as described in claim 1, characterized in that: The optimization of the dust removal scheme requires judging whether the dust removal effect meets the standard based on the dust removal effect value, analyzing the reasons why the dust removal effect does not meet the standard based on the judgment result, formulating parameter adjustment rules based on the reason analysis results, and optimizing the dust removal scheme.

Images