Sand blasting process control system and method based on data driving

Abstract

The invention discloses a sand blasting process control system and method based on data driving, and relates to the field of intelligent control, the sand blasting process control system comprises a real-time sensing module, a workpiece identification module, a control decision module, a control execution module and a memory module, the sand blasting process and parameter information of a workpiece to be subjected to sand blasting are rapidly obtained, and sand blasting parameters are formulated according to the obtained information; then, a control signal is generated according to the sand blasting parameters, automatic sand blasting machining operation is achieved, on the basis that the state and result of the sand blasting process are monitored in real time, an improved FAHP analytic hierarchy process is adopted to dynamically adjust a control instruction, and a historical control scheme and operation parameters of the sand blasting process are recorded, so that quick response to similar fire accidents is achieved; the problems that in the prior art, machining quality is poor, efficiency is low and limitation is high are solved, sand blasting parameters can be automatically adjusted according to the states of the surfaces of different workpieces, and therefore the efficiency and quality of sand blasting machining are improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control, and in particular to a data-driven sandblasting process control system and method thereof. Background Art

[0002] With the continuous upgrading of power equipment and the development of technology, the requirements for surface cleaning of power equipment are getting higher and higher, and sandblasting technology has emerged. The outer line sandblasting machine is a device used to clean and process the surface of power equipment such as wires or signal lines. It is widely used in industrial production. It uses high-speed spraying of sand or other granular media to impact and rub the surface of the workpiece to remove dirt, oxide layer, coating or scratches on the surface, thereby achieving the purpose of surface cleaning, removing rust, increasing roughness, etc.

[0003] Traditional sandblasting machines usually use a fixed preset sandblasting pressure to clean the workpiece surface. However, different workpiece surfaces have different materials and uneven conditions. The fixed pressure processing method cannot adapt to the working conditions of different workpiece surfaces. For complex workpiece surfaces, the manual preset method cannot achieve the best processing effect, and there are problems of poor processing quality and low efficiency, which has limitations in use.

[0004] Therefore, a data-driven sandblasting process control system and method thereof are needed to solve the above problems. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention discloses a data-driven sandblasting process control system and method thereof, which can achieve efficient and accurate sandblasting operations through real-time perception, workpiece identification and adaptive decision-making control.

[0006] The present invention adopts the following technical solutions:

[0007] A data-driven sandblasting process control system, comprising:

[0008] A real-time sensing module, which includes a sandblasting parameter sensing unit and a workpiece parameter sensing unit. The sandblasting parameter sensing unit uses a multi-point sensor group to obtain the process parameters of the sandblasting process in real time, and the workpiece parameter sensing unit uses a visual sensor to obtain the surface parameters of the workpiece to be sandblasted in real time. The real-time sensing module converts the obtained parameter information into a digital signal through a signal conditioner and transmits it to the cloud for processing;

[0009] A workpiece recognition module, wherein the workpiece recognition module extracts workpiece surface feature information through a cloud processor, and uses a nearest interpolation algorithm to standardize the extracted workpiece surface feature points, and the cloud processor uses an edge detection algorithm to identify the geometric shape features, texture features, and material features of the workpiece surface;

[0010] A control decision module, wherein the control decision module uses an adaptive control decision model to make control decisions on the sandblasting process, wherein the adaptive control decision model establishes a model of the workpiece to be sandblasted based on the extracted surface feature information of the workpiece, and generates a sandblasting process control instruction according to the model of the workpiece to be sandblasted and the sandblasting parameters acquired in real time;

[0011] A control execution module, wherein the control execution module obtains a sandblasting process control signal according to the sandblasting process control instruction generated by the control decision module, and sends the sandblasting process control signal to the controller through the electrical interface to control the sandblasting equipment to perform an operation. When the sandblasting equipment performs an operation, the control execution module uses a closed-loop feedback mechanism to feed back the state of the workpiece during the sandblasting process in real time. When the feedback signal differs from the preset target signal, the improved FAHP hierarchical analysis algorithm is used to dynamically adjust the sandblasting process control signal;

[0012] Memory module, used to record the historical control scheme of the sandblasting process and the surface parameters of the sandblasted workpiece.

[0013] Furthermore, the multi-point sensor group includes at least a pressure sensor, a flow sensor, an angle sensor and a temperature sensor, the process parameters of the sandblasting process include at least sandblasting pressure, sandblasting medium flow, sandblasting angle and sandblasting temperature, and the surface parameters of the sandblasting workpiece include shape, size, surface unevenness and surface roughness.

[0014] Furthermore, the nearest interpolation algorithm obtains the feature point that is closest to the extracted workpiece surface feature point in a predefined standardized space according to the Euclidean distance measurement method, and performs polynomial interpolation calculation based on the coordinates and attributes of the nearest feature point to generate a standardized feature point. The edge detection algorithm uses an edge detection operator to perform edge detection on the surface of the workpiece to be sandblasted to generate an edge image, so as to capture the area with obvious brightness changes in the image, and extract the edge contour of the workpiece to be sandblasted according to the edge image.

[0015] Furthermore, the adaptive control decision model includes an input layer, an adaptive parameter adjustment layer, a particle update layer, an adaptive weight layer, a multi-neighborhood search layer and an output layer. The operation of the adaptive control decision model includes the following steps:

[0016] S1. Initialization input. In the control decision module, the size of the initial control strategy set, the maximum number of iterations, the inertia weight and the neighborhood search parameter are set through the input layer. The inertia weight is used to control the speed and stability of the strategy adjustment. The extracted workpiece surface feature information is input into the adaptive control decision model to establish a workpiece model to be sandblasted. The neighborhood search parameter at least includes the control strategy field search range and the search step. The adaptive control decision model is initialized according to the sandblasting parameters obtained in real time and the objective function to be optimized. The objective function to be optimized includes at least sandblasting uniformity and surface roughness.

[0017] S2, adaptive parameter adjustment, the adaptive parameter adjustment layer sets the control parameters, objective function and adaptive algorithm according to the historical optimal solution of the workpiece model to be sandblasted and the current optimal solution of the particle swarm, and dynamically adjusts the neighborhood search range based on different neighborhood search strategies;

[0018] S3, particle update: the particle update layer calculates the evaluation index of the sandblasting effect according to the current control strategy, wherein the evaluation index includes the sandblasting uniformity and the surface roughness, and updates the position and speed of the particles according to the evaluation index to generate a new control strategy;

[0019] S4, adaptive inertia weight adjustment, the adaptive weight layer adaptively updates the inertia weight according to the current number of iterations and the global historical optimal fitness value, and adjusts the change trend of the inertia weight according to the change of the objective function to accelerate or slow down the adjustment speed of the control strategy;

[0020] S5, multi-neighborhood search, formulating multiple neighborhood search strategies through the multi-neighborhood search layer, wherein the multiple neighborhood search strategies include global neighborhood search, local neighborhood search and chaotic search;

[0021] S6. Output the optimal decision plan. When the maximum number of iterations is reached or the termination condition is met, the optimal control strategy is output through the output layer as the control instruction of the sandblasting process. Otherwise, repeat steps S2 to S5.

[0022] Furthermore, when there is a difference between the feedback signal and the preset target signal, the improved FAHP hierarchical analysis algorithm calculates the deviation between the current sandblasting process and the desired target according to the preset target signal and the feedback signal, and the output difference function formula is:

[0023]

[0024] In formula (1), F represents the deviation between the current sandblasting process and the expected target, z represents the feedback data of the workpiece surface parameters during the sandblasting process, y represents the calibration difference of the control signal during the operation of the sandblasting equipment, μ represents the output control signal of the sandblasting process, Δ(z) represents the change value of the surface parameters of the sandblasting workpiece during the operation of the sandblasting equipment, A(z) represents the change amplitude of the surface parameters of the sandblasting workpiece during the operation of the sandblasting equipment, and the measurement function formula of the change value of the surface parameters of the sandblasting workpiece during the operation of the sandblasting equipment is:

[0025] Δ i (z) = w i (p i -s i ) (2)

[0026] In formula (2), Δ i (z) represents the change value of the surface parameter of the i-th workpiece during the operation of the sandblasting equipment, i represents the ordinal number of the workpiece surface parameter variable during the sandblasting process, 1≤i≤N, N represents the number of workpiece surface parameter variables during the sandblasting process, p i represents the actual variable value of the surface parameter of the i-th workpiece during the sandblasting process, s i represents the self-loss variable of the surface parameters of the i-th workpiece during the sandblasting process, 0≤s≤1, w i represents the weight of the i-th workpiece surface parameter variable during the sandblasting process; during the operation of the sandblasting equipment, the output function formula of the variation range of the sandblasting workpiece surface parameter during the operation of the sandblasting equipment is:

[0027]

[0028] In formula (3), A(z) is the variation range of the surface parameters of the blasting workpiece during the operation of the blasting equipment. is the average value of all parameter variable changes.

[0029] Furthermore, the electrical interface performs noise reduction processing through a frequency selective filter, and uses a low-distortion and high-speed analog-to-digital converter to convert the analog control signal into a digital control signal for transmission. During the transmission process, automatic error detection and correction processing is achieved through wide amplitude modulation, high-speed modulation, code division multiplexing and error correction coding.

[0030] Furthermore, the memory module uses a cloud database to store historical control schemes of the sandblasting process and the corresponding surface parameters of the sandblasted workpiece. The cloud database uses an access counter to record the number of times the stored information is called, and uses a timer to trigger a clearing operation. The timer periodically clears the historical control schemes of the sandblasting process and the corresponding surface parameters of the sandblasted workpiece based on a least used cache strategy. The least used cache strategy sorts the stored information items from large to small weights according to the number of times the stored information items are called and the calling time, and clears the stored information items in reverse order. Before performing the operation, the user saves the historical control schemes of the sandblasting process and the corresponding surface parameters of the sandblasted workpiece to a local storage device or uploads them to a storage platform for long-term storage.

[0031] Furthermore, a data-driven sandblasting process control method comprises the following steps:

[0032] Step 1: Use a multi-point sensor group to obtain the process parameters in the sandblasting process in real time, and use a visual sensor to obtain the surface parameters of the workpiece to be sandblasted in real time, and then convert the obtained parameter information into a digital signal through a signal conditioner and transmit it to the cloud for processing;

[0033] Step 2: extracting workpiece surface feature information using a cloud processor, and using a nearest interpolation algorithm to standardize the extracted workpiece surface feature points, and then identifying the workpiece surface feature information using an edge detection algorithm, wherein the workpiece surface feature information includes at least geometric shape features, texture features, and material features;

[0034] Step 3: using an adaptive control decision model to make a control decision on the sandblasting process, wherein the adaptive control decision model establishes a model of the workpiece to be sandblasted based on the extracted surface feature information of the workpiece, and generates a sandblasting process control instruction according to the model of the workpiece to be sandblasted and the sandblasting parameters obtained in real time;

[0035] Step 4: Obtain a control signal according to the generated sandblasting process control instruction, and send the sandblasting process control signal to the controller through the electrical interface to control the sandblasting equipment to perform the operation. When the sandblasting equipment performs the operation, a closed-loop feedback mechanism is used to feed back the status of the workpiece in the sandblasting process in real time. When the feedback signal differs from the preset target signal, the improved FAHP hierarchical analysis algorithm is used to dynamically adjust the sandblasting process control signal.

[0036] Step 5: Record the historical control scheme of the sandblasting process and the surface parameters of the sandblasted workpiece to facilitate the subsequent call of similar control operations.

[0037] The beneficial effects of the present invention are:

[0038] 1. The present invention obtains workpiece surface feature information through a real-time perception module and a workpiece identification module. The adaptive control decision model can establish a workpiece model to be sandblasted according to the extracted workpiece surface feature information, and generate control instructions according to the sandblasting parameters obtained in real time. It can control the sandblasting process for workpieces of different shapes, textures and materials, ensure the accuracy and stability of the sandblasting process, reduce the processing cycle and resource consumption, and improve the sandblasting efficiency.

[0039] 2. The present invention adopts a closed-loop feedback mechanism to provide real-time feedback on the status of the workpiece when the sandblasting equipment is operating. This mechanism can promptly detect and correct deviations in the sandblasting process to ensure the stability and consistency of the sandblasting quality. When the feedback signal differs from the preset target signal, the improved FAHP hierarchical analysis algorithm is used to dynamically adjust the sandblasting process control signal to further improve the sandblasting quality and efficiency.

[0040] 3. The present invention efficiently records and manages the historical control scheme and workpiece surface parameters of the sandblasting process through the cloud database, intelligently optimizes the storage space by using the access counter and the timer of the least used cache strategy, and supports users to back up locally or upload to the storage platform for long-term storage, which effectively improves the convenience of information retrieval and the efficiency of storage management. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic diagram of the overall system process architecture of the present invention;

[0042] Figure 2 It is a structural diagram of the adaptive control decision model in the present invention;

[0043] Figure 3 It is a schematic diagram of the overall method flow of the present invention;

[0044] Figure 4 This is a schematic diagram of the architecture of the real-time perception module in the present invention;

[0045] Figure 5 It is a schematic diagram of the architecture of the control execution module in the present invention. DETAILED DESCRIPTION

[0046] The following will be combined with the attached embodiment of the present invention Figure 1 To Attachment Figure 5 , the technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. 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.

[0047] The embodiment of the present invention discloses a data-driven sandblasting process control system, which aims to achieve precise control of the sandblasting process through real-time perception, accurate identification, intelligent decision-making and precise execution. The system mainly consists of five parts: real-time perception module, workpiece identification module, control decision module, control execution module and memory module.

[0048] The real-time sensing module includes a sandblasting parameter sensing unit and a workpiece parameter sensing unit;

[0049] The sandblasting parameter sensing unit uses a multi-point sensor group, such as pressure sensors, flow sensors, temperature sensors, etc., to obtain key process parameters in the sandblasting process in real time, such as sandblasting pressure, abrasive flow, sandblasting temperature, etc. The workpiece parameter sensing unit uses visual sensors, such as high-resolution cameras, to capture the surface image of the workpiece to be sandblasted in real time and obtain the surface parameters of the workpiece, such as surface roughness, shape, size, etc. The real-time sensing module converts the raw signal collected by the sensor into a digital signal through a signal conditioner, and transmits it to the cloud through the network for further processing.

[0050] The workpiece recognition module uses a cloud processor to use image processing technology to extract feature information on the workpiece surface, including geometric shape features, texture features, and material features. The nearest interpolation algorithm is used to standardize the extracted feature points to ensure the accuracy and consistency of subsequent analysis. The edge detection algorithm is used to identify the contour and detail features of the workpiece surface, providing a basis for subsequent control decisions.

[0051] The control decision module builds a digital model of the workpiece to be sandblasted based on the extracted workpiece surface feature information. And according to the workpiece model to be sandblasted and the sandblasting parameters obtained in real time, the adaptive control decision model generates sandblasting process control instructions, including sandblasting pressure, sandblasting time, sandblasting angle, etc.

[0052] The control execution module receives the sandblasting process control instructions generated by the control decision module and converts them into specific control signals. The control signal is sent to the controller of the sandblasting equipment through the electrical interface to control the sandblasting equipment to perform operations according to the instructions. During the sandblasting process, a closed-loop feedback mechanism is used to monitor the workpiece status in real time. When the feedback signal differs from the preset target signal, the improved FAHP hierarchical analysis algorithm is used to dynamically adjust the control signal to ensure that the sandblasting effect meets the expectations.

[0053] The memory module records the historical control scheme of each sandblasting process, including control parameters, execution time, execution results, etc. The surface parameters of each sandblasted workpiece, such as material, shape, size, etc., are recorded to provide a reference for subsequent analysis and optimization.

[0054] Implementation steps

[0055] System initialization: Start the system, ensure that all modules are working properly, and connect to the cloud processor.

[0056] Real-time perception and recognition: The real-time perception module collects sandblasting parameters and workpiece parameters, and the workpiece recognition module extracts and standardizes workpiece features to perform edge detection and recognition.

[0057] Control decision generation: The control decision module generates sandblasting process control instructions based on the recognition results and real-time parameters.

[0058] Control execution and feedback: The control execution module receives instructions and executes them, while using a closed-loop feedback mechanism to monitor the workpiece status in real time and adjust the control signal as needed.

[0059] Data recording and analysis: The memory module records the historical control scheme and workpiece surface parameters of the sandblasting process, providing data support for subsequent analysis and optimization.

[0060] Through the above implementation steps, the system can achieve precise control of the sandblasting process, improve the sandblasting quality and efficiency, and reduce production costs.

[0061] In the data-driven sandblasting process control system, the nearest interpolation algorithm is used to standardize the extracted feature points on the workpiece surface. The following are the specific implementation steps of the algorithm:

[0062] 1. Predefined standardized space construction: First, a predefined standardized space is constructed based on historical data and experience. This space contains the standardized coordinates and attributes of various possible feature points on the workpiece surface. Each feature point in the standardized space should have clear coordinates (such as position in two-dimensional or three-dimensional space) and attributes (such as color, texture, material, etc.).

[0063] 2. Feature point extraction: Through the image processing technology in the workpiece recognition module, feature points are extracted from the surface image of the workpiece to be sandblasted. These feature points represent the key geometric shape, texture and material characteristics of the workpiece surface.

[0064] 3. Euclidean distance metric: For each extracted feature point, calculate its Euclidean distance to all feature points in the predefined standardized space. Euclidean distance is a commonly used metric to measure the straight-line distance between two points.

[0065] 4. Find the nearest feature point: In the predefined standardized space, find the feature point with the smallest Euclidean distance to the extracted feature point, that is, the nearest feature point.

[0066] 5. Polynomial interpolation calculation: Based on the coordinates and attributes of the nearest feature points, the polynomial interpolation method is used to generate standardized feature points. Polynomial interpolation is a mathematical method used to generate smooth curves or surfaces based on known data points. Here, polynomial interpolation is used to generate standardized feature points corresponding to the extracted feature points based on the information of the nearest feature points.

[0067] 6. Standardized feature point generation: After polynomial interpolation calculation, standardized feature points are obtained, which have coordinates and properties similar to those of the feature points in the predefined standardized space.

[0068] The edge detection algorithm is used to identify the edge contour of the surface of the workpiece to be sandblasted. The following are the specific implementation steps of the algorithm:

[0069] 1. Edge detection operator selection: Select appropriate edge detection operators, such as Canny operator, Sobel operator, Prewitt operator, etc. These operators can detect areas with obvious brightness changes in the image, that is, edges.

[0070] 2. Edge image generation: Apply the selected edge detection operator to perform edge detection on the surface image of the sandblasted workpiece. The result of edge detection is an edge image, which only contains the edge information in the image.

[0071] 3. Edge contour extraction: Extract the edge contour of the workpiece to be sandblasted from the edge image. This usually involves further processing of the edge image, such as contour tracking, contour refinement, etc.

[0072] 4. Edge profile optimization: The extracted edge profile is optimized to remove noise and unnecessary details. The optimized edge profile will be used for subsequent control decisions and sandblasting process execution.

[0073] Through the above specific implementation steps, the nearest interpolation algorithm and edge detection algorithm can accurately adjust the surface feature points of the workpiece and identify the edge contour, providing strong support for the precise control of the sandblasting process.

[0074] The adaptive control decision model is a multi-level decision system designed to optimize the sandblasting process and ensure that key indicators such as sandblasting uniformity and surface roughness are optimal. The following is a detailed implementation description of the model's working steps:

[0075] S1: Initialization input

[0076] In the control decision module, the size of the initial control strategy set (i.e., the size of the particle swarm), the maximum number of iterations (i.e., the maximum number of cycles for the algorithm to run), the inertia weight (a coefficient that affects the particle velocity adjustment), and the neighborhood search parameters (including the search range and the search step) are set through the input layer. The inertia weight is used to balance the capabilities of global search and local search to ensure the speed and stability of strategy adjustment. The surface feature information of the workpiece to be sandblasted, such as shape, size, material, etc., is extracted and input into the adaptive control decision model. Based on this information, a model of the workpiece to be sandblasted is established for subsequent simulation and optimization. The adaptive control decision model is initialized according to the real-time acquired sandblasting parameters (such as sandblasting pressure, sandblasting speed, sandblasting angle, etc.) and the objective function to be optimized (sandblasting uniformity and surface roughness).

[0077] S2: Adaptive parameter adjustment

[0078] According to the historical optimal solution of the workpiece model to be sandblasted and the current optimal solution of the particle swarm, the adaptive parameter adjustment layer sets the control parameters (such as the range and step size of the sandblasting parameters), the objective function (the mathematical expression of the sandblasting uniformity and surface roughness) and the adaptive algorithm (such as the particle swarm optimization algorithm). Based on different neighborhood search strategies (such as distance-based search, similarity-based search, etc.), the neighborhood search range is dynamically adjusted to meet the search needs at different stages.

[0079] S3: Particle Update

[0080] The particle update layer calculates the evaluation index of the sandblasting effect, including sandblasting uniformity and surface roughness, based on the current control strategy (i.e., the position and velocity of the particles). The position and velocity of the particles are updated according to the evaluation index to generate a new control strategy. This usually involves mathematical operations on the particle velocity and position to move it towards a more optimal solution.

[0081] S4: Adaptive inertia weight adjustment

[0082] The adaptive weight layer adaptively updates the inertia weight according to the current number of iterations and the global historical optimal fitness value. This helps to balance the capabilities of global search and local search during the search process. According to the changes in the objective function (such as whether it tends to be stable, whether there is oscillation, etc.), the change trend of the inertia weight is adjusted. This can speed up or slow down the adjustment speed of the control strategy to improve the search efficiency.

[0083] S5: Multi-Neighborhood Search

[0084] The multi-neighborhood search layer formulates multiple neighborhood search strategies, including global neighborhood search (looking for a better solution in the entire search space), local neighborhood search (looking for a better solution near the current solution), and chaos search (using chaos theory for random search). According to the current search stage and search requirements, select the appropriate neighborhood search strategy to perform the search operation.

[0085] S6: Output the optimal decision solution

[0086] When the maximum number of iterations is reached or the termination condition is met (such as finding a solution that meets the requirements, the fitness value no longer increases significantly, etc.), the search process is stopped. The optimal control strategy is output through the output layer as the control instructions for the sandblasting process. These instructions include specific setting values ​​of the sandblasting parameters to ensure that the sandblasting uniformity and surface roughness are optimized.

[0087] If the termination condition is not met, steps S2 to S5 are repeated to continue searching for a better solution.

[0088] Through the cyclic iteration of the above steps, the adaptive control decision model can continuously optimize the sandblasting process to ensure that the workpiece surface achieves the expected sandblasting effect.

[0089] When there is a difference between the feedback signal and the preset target signal, the improved FAHP hierarchical analysis algorithm calculates the deviation between the current sandblasting process and the expected target according to the preset target signal and the feedback signal, and the output difference function formula is:

[0090]

[0091] In formula (1), F represents the deviation between the current sandblasting process and the expected target, z represents the feedback data of the workpiece surface parameters during the sandblasting process, y represents the calibration difference of the control signal during the operation of the sandblasting equipment, μ represents the output control signal of the sandblasting process, Δ(z) represents the change value of the surface parameters of the sandblasting workpiece during the operation of the sandblasting equipment, A(z) represents the change amplitude of the surface parameters of the sandblasting workpiece during the operation of the sandblasting equipment, and the measurement function formula of the change value of the surface parameters of the sandblasting workpiece during the operation of the sandblasting equipment is:

[0092] Δ i (z) = w i (p i -s i ) (2)

[0093] In formula (2), Δ i (z) represents the change value of the surface parameter of the i-th workpiece during the operation of the sandblasting equipment, i represents the ordinal number of the workpiece surface parameter variable during the sandblasting process, 1≤i≤N, N represents the number of workpiece surface parameter variables during the sandblasting process, p irepresents the actual variable value of the surface parameter of the i-th workpiece during the sandblasting process, s i represents the self-loss variable of the surface parameters of the i-th workpiece during the sandblasting process, 0≤s≤1, w i represents the weight of the i-th workpiece surface parameter variable during the sandblasting process; during the operation of the sandblasting equipment, the output function formula of the variation range of the sandblasting workpiece surface parameter during the operation of the sandblasting equipment is:

[0094]

[0095] In formula (3), A(z) is the variation range of the surface parameters of the blasting workpiece during the operation of the blasting equipment. is the average value of all parameter variable changes.

[0096] The hardware working environment of the improved FAHP (Fuzzy Analytic Hierarchy Process) analytic hierarchy process algorithm usually involves computing equipment and related auxiliary equipment to ensure that the algorithm can run efficiently and accurately. The hardware working environment of the improved FAHP analytic hierarchy process algorithm includes:

[0097] High-performance computers: Equipped with powerful processors, such as multi-core CPUs or high-performance GPUs, to support complex computing tasks. Sufficient memory and storage space to ensure that algorithms are not resource-constrained when processing large amounts of data.

[0098] Dedicated servers: For application scenarios that require continuous operation or handle a large number of concurrent requests, dedicated servers can provide higher stability and reliability. Servers are usually equipped with redundant power supplies, hot backup and other mechanisms to ensure that they can quickly resume operation in the event of hardware failure.

[0099] Input and output devices: Input devices such as keyboards and mice are used to operate the algorithm and set parameters. Output devices such as monitors are used to display the algorithm's operating results and status information.

[0100] Storage devices: Storage devices such as external hard disks and solid-state drives (SSDs) are used to store the input data and output results required by the algorithm. Network storage devices (such as NAS) can also be used as a solution for data backup and sharing.

[0101] Network equipment: Network devices such as network cards, routers, and switches are used to implement network communication between algorithms and other systems or devices. In a distributed computing environment, network equipment is also responsible for allocating computing tasks to different computing nodes.

[0102] Power supply: Stable power supply is the basis for ensuring the normal operation of the algorithm. Emergency power supply equipment such as uninterruptible power supply (UPS) or generator should be equipped to cope with sudden power outages.

[0103] It is generally recommended to place computer equipment in an environment with moderate temperature and good ventilation, and to equip it with equipment such as air conditioners or dehumidifiers to maintain appropriate temperature and humidity conditions. Appropriate electromagnetic shielding and grounding measures should be taken to reduce the impact of electromagnetic interference.

[0104] In summary, the hardware working environment of the improved FAHP analytic hierarchy process algorithm needs to comprehensively consider multiple aspects such as core computing equipment, auxiliary equipment, environmental requirements, and other considerations. By rationally configuring and optimizing hardware resources, the algorithm can be ensured to be efficient, accurate, stable, and reliable.

[0105] The laboratory configuration uses a Core i964+128G memory computer, and the simulation environment is established using simulation software. The on-site experimental environment is set up, the simulation data accuracy is 95%, and the algorithm operation error does not exceed 2.5%. The improved FAHP hierarchical analysis algorithm (Group A) and the FAHP hierarchical analysis algorithm (Group B) are used for comparative experiments. Five scenarios are set up manually, including different workpiece surface materials, different workpiece surface finishes, different sandblasting equipment angles, different sandblasting equipment distances from the workpiece, and different sandblasting media. The improved FAHP hierarchical analysis algorithm (Group A) and the FAHP hierarchical analysis algorithm (Group B) are run respectively, and the running time and error rate are compared, which are recorded in Table 1.

[0106] Table 1 Results Statistics

[0107]

[0108] As can be seen from Table 1, the improved FAHP AHP algorithm (group A) shows better running time and lower error rate in different scenarios compared with the FAHP AHP algorithm (group B). In five different scenarios, the average error rate of the algorithm in group A is 1.52%, while the average error rate of the algorithm in group B is 2.68%. The average running time of the algorithm in group A is 2.56 seconds, and the average running time of the algorithm in group B is 3.4 seconds. Considering the running efficiency and accuracy, it can be seen that the improved FAHP AHP algorithm is more applicable in the sandblasting process.

[0109] The electrical interface uses a frequency selective filter to reduce noise, and adopts a low-distortion, high-speed analog-to-digital converter to convert the analog control signal into a digital control signal for transmission. During the transmission process, automatic error detection and correction are achieved through wide amplitude modulation, high-speed modulation, code division multiplexing and error correction coding.

[0110] First, the analog control signal to be transmitted is pre-processed and then subjected to noise reduction by a frequency selective filter. A frequency selective filter is a filter that can filter out unwanted signal frequencies based on a specific frequency, thereby improving the accuracy and stability of the signal. Secondly, the analog control signal after noise reduction is converted into a digital control signal for transmission using a low-distortion, high-speed analog-to-digital converter. A low-distortion, high-speed analog-to-digital converter is a device that can quickly and efficiently convert analog signals into digital signals. It uses extremely low distortion and high speed to ensure the accuracy and response speed of the control signal.

[0111] During the transmission process, in order to further improve the transmission efficiency and stability, a variety of transmission technologies such as wide amplitude modulation, high-speed modulation and code division multiplexing are adopted. Wide amplitude modulation is to achieve information transmission by changing the amplitude of the digital signal, which can improve the anti-noise ability and signal coverage; high-speed modulation is to achieve information transmission by changing the frequency of the digital signal, which can improve the transmission rate and stability; code division multiplexing is to combine multiple digital signals into one digital signal for transmission through different encoding methods, which can improve the transmission bandwidth and data capacity. Finally, in order to ensure the accuracy and integrity of the transmitted data, the present invention adopts error correction coding to realize automatic error detection and correction processing. Error correction coding is a digital processing technology that can encode and decode the transmitted information to detect and correct errors, which can greatly improve the reliability and controllability of data transmission.

[0112] When the memory module is performing sandblasting control, the system will automatically record the control scheme used in the sandblasting process and the surface parameters of the sandblasted workpiece, and store them in a local storage device or upload them to a cloud database for long-term storage.

[0113] The memory module stores the historical control scheme of the sandblasting process and the corresponding surface parameters of the sandblasted workpiece in the cloud database to facilitate the subsequent call of similar control operations. The access counter is used to record the number of times the storage information item is called, so that useless data in the storage information library can be regularly cleared later. The timer is used to trigger the clearing operation, and the historical control scheme of the sandblasting process and the corresponding surface parameters of the sandblasted workpiece are regularly cleared to ensure the efficient use of the storage information library. The least used cache strategy is used for clearing operations, and the storage information items are sorted according to the number of times the storage information items are called and the calling time (the more times they are called, the higher the weight), and the storage information items are cleared in reverse order.

[0114] When similar or identical workpieces to be sandblasted or control requirements are detected subsequently, the stored corresponding control scheme is directly called to perform control operations to improve the control response speed.

[0115] In summary, the memory module of the present invention adopts a cloud database to store the historical control scheme of the sandblasting process and the corresponding surface parameters of the sandblasting workpiece, and adopts an access counter, a timer and a least-used cache strategy to manage the storage information library, which is efficient, reliable and intelligent.

[0116] A data-driven sandblasting process control method comprises the following steps:

[0117] Step 1: Real-time data acquisition and transmission: a multi-point sensor group is used to obtain the process parameters of the sandblasting process in real time, and a visual sensor is used to obtain the surface parameters of the workpiece to be sandblasted in real time. The acquired parameter information is then converted into a digital signal through a signal conditioner and transmitted to the cloud for processing;

[0118] Step 2: Extracting and standardizing workpiece surface feature information, using a cloud processor to extract workpiece surface feature information, and using a nearest interpolation algorithm to standardize the extracted workpiece surface feature points, and then using an edge detection algorithm to identify the workpiece surface feature information, the workpiece surface feature information includes at least geometric shape features, texture features and material features;

[0119] Step 3: Establishing an adaptive control decision model and generating instructions. An adaptive control decision model is used to make control decisions on the sandblasting process. The adaptive control decision model establishes a model of the workpiece to be sandblasted based on the extracted surface feature information of the workpiece, and generates sandblasting process control instructions according to the model of the workpiece to be sandblasted and the sandblasting parameters obtained in real time.

[0120] Step 4: Control signal transmission and closed-loop feedback adjustment: obtain the control signal according to the generated sandblasting process control instruction, and send the sandblasting process control signal to the controller through the electrical interface to control the sandblasting equipment to perform the operation. When the sandblasting equipment performs the operation, the closed-loop feedback mechanism is used to feed back the status of the workpiece in the sandblasting process in real time. When the feedback signal is different from the preset target signal, the improved FAHP hierarchical analysis algorithm is used to dynamically adjust the sandblasting process control signal;

[0121] Step 5: Historical data recording and subsequent call, record the historical control scheme of the sandblasting process and the surface parameters of the sandblasted workpiece, so as to facilitate the subsequent call of similar control operations.

[0122] When the control execution module controls the sandblasting equipment to perform operations, the operation method of the sandblasting equipment generally includes the following steps, which are based on the general operation process of the sandblasting machine and the specific requirements of the data-driven control system:

[0123] 1. Preparation

[0124] Operators should wear protective glasses, masks, gloves and other protective equipment to ensure safety. Check whether the power supply, air source, conveyor belt, spray gun, etc. of the sandblasting machine are working properly. Ensure that the dust removal system, ventilation ducts, etc. are well sealed to prevent dust leakage. Select the appropriate abrasive according to the sandblasting requirements, start the sand stirring device, and slowly pour the abrasive into the sandblasting machine.

[0125] 2. Setup and Startup

[0126] Set the sandblasting parameters, such as sandblasting pressure, sandblasting time, abrasive flow rate, etc. through the control system. These parameters should be determined according to the material, shape and surface requirements of the workpiece to be sandblasted. Start the power supply and air supply of the sandblasting machine to ensure the normal operation of all components. Start the conveyor belt to send the workpiece to be sandblasted into the sandblasting room.

[0127] 3. Sand blasting operation

[0128] When the workpiece enters the sandblasting room, the automatic sandblasting button is turned on and the sandblasting machine starts working. During the sandblasting process, the spray gun will spray the workpiece according to the preset parameters. The operator should monitor the sandblasting process in real time and observe the changes on the workpiece surface. If the sandblasting effect is not good or there is an abnormality on the workpiece surface, the sandblasting parameters should be adjusted in time or the sandblasting should be stopped. After automatic sandblasting, some dead corners or hard-to-reach areas may not be completely sandblasted. At this time, the operator should use a manual spray gun to re-spray these areas to ensure the quality of sandblasting.

[0129] 4. Shutdown and Cleanup

[0130] When the workpiece is finished sandblasting, turn off the automatic sandblasting button to stop the sandblasting machine. Turn off the conveyor belt and remove the workpiece from the sandblasting room. After stopping, the abrasive and dust inside the sandblasting machine should be cleaned in time. Check and clean the spray gun, conveyor belt and other parts to ensure that the equipment can work normally the next time it is used. Finally, turn off the power and air source of the sandblasting machine to ensure the safety of the equipment.

[0131] 5. Data Recording and Analysis

[0132] The control execution module records the historical data of the sandblasting process, including sandblasting parameters, workpiece surface parameters, etc. These data can be used for subsequent analysis and optimization. Regularly analyze the recorded data to find potential improvement points. According to the analysis results, optimize the sandblasting parameters, abrasive type, etc. to improve the sandblasting quality and efficiency.

[0133] In summary, the operation method of sandblasting equipment is a complex and delicate process, which requires operators to master and strictly abide by the operating specifications. At the same time, through the data-driven control system, the sandblasting process can be accurately controlled and optimized, and the sandblasting quality and efficiency can be improved.

[0134] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A data-driven sandblasting process control system, characterized in that: include: A real-time sensing module, which includes a sandblasting parameter sensing unit and a workpiece parameter sensing unit. The sandblasting parameter sensing unit uses a multi-point sensor group to obtain the process parameters of the sandblasting process in real time, and the workpiece parameter sensing unit uses a visual sensor to obtain the surface parameters of the workpiece to be sandblasted in real time. The real-time sensing module converts the obtained parameter information into a digital signal through a signal conditioner and transmits it to the cloud for processing; A workpiece recognition module, wherein the workpiece recognition module extracts workpiece surface feature information through a cloud processor, and uses a nearest interpolation algorithm to standardize the extracted workpiece surface feature points, and the cloud processor uses an edge detection algorithm to identify the geometric shape features, texture features, and material features of the workpiece surface; A control decision module, wherein the control decision module uses an adaptive control decision model to make control decisions on the sandblasting process, wherein the adaptive control decision model establishes a model of the workpiece to be sandblasted based on the extracted surface feature information of the workpiece, and generates a sandblasting process control instruction according to the model of the workpiece to be sandblasted and the sandblasting parameters acquired in real time; A control execution module, wherein the control execution module obtains a sandblasting process control signal according to the sandblasting process control instruction generated by the control decision module, and sends the sandblasting process control signal to the controller through the electrical interface to control the sandblasting equipment to perform an operation. When the sandblasting equipment performs an operation, the control execution module uses a closed-loop feedback mechanism to feed back the state of the workpiece during the sandblasting process in real time. When the feedback signal differs from the preset target signal, the improved FAHP hierarchical analysis algorithm is used to dynamically adjust the sandblasting process control signal; Memory module, used to record the historical control scheme of the sandblasting process and the surface parameters of the sandblasted workpiece.

2. A data-driven sandblasting process control system according to claim 1, characterized in that: The multi-point sensor group includes at least a pressure sensor, a flow sensor, an angle sensor and a temperature sensor. The process parameters of the sandblasting process include at least sandblasting pressure, sandblasting medium flow, sandblasting angle and sandblasting temperature. The surface parameters of the sandblasting workpiece include shape, size, surface concavity and surface roughness.

3. A data-driven sandblasting process control system according to claim 1, characterized in that: The nearest interpolation algorithm obtains the feature point that is closest to the extracted workpiece surface feature point in a predefined standardized space according to the Euclidean distance measurement method, and performs polynomial interpolation calculation based on the coordinates and attributes of the nearest feature point to generate a standardized feature point. The edge detection algorithm uses an edge detection operator to perform edge detection on the surface of the workpiece to be sandblasted to generate an edge image, so as to capture the area with obvious brightness changes in the image, and extract the edge contour of the workpiece to be sandblasted according to the edge image.

4. A data-driven sandblasting process control system according to claim 1, characterized in that: The adaptive control decision model includes an input layer, an adaptive parameter adjustment layer, a particle update layer, an adaptive weight layer, a multi-neighborhood search layer and an output layer. The operation of the adaptive control decision model includes the following steps: S1. Initialization input. In the control decision module, the size of the initial control strategy set, the maximum number of iterations, the inertia weight and the neighborhood search parameter are set through the input layer. The inertia weight is used to control the speed and stability of the strategy adjustment. The extracted workpiece surface feature information is input into the adaptive control decision model to establish a workpiece model to be sandblasted. The neighborhood search parameter at least includes the control strategy field search range and the search step. The adaptive control decision model is initialized according to the sandblasting parameters obtained in real time and the objective function to be optimized. The objective function to be optimized includes at least sandblasting uniformity and surface roughness. S2, adaptive parameter adjustment, the adaptive parameter adjustment layer sets the control parameters, objective function and adaptive algorithm according to the historical optimal solution of the workpiece model to be sandblasted and the current optimal solution of the particle swarm, and dynamically adjusts the neighborhood search range based on different neighborhood search strategies; S3, particle update: the particle update layer calculates the evaluation index of the sandblasting effect according to the current control strategy, wherein the evaluation index includes the sandblasting uniformity and the surface roughness, and updates the position and speed of the particles according to the evaluation index to generate a new control strategy; S4, adaptive inertia weight adjustment, the adaptive weight layer adaptively updates the inertia weight according to the current number of iterations and the global historical optimal fitness value, and adjusts the change trend of the inertia weight according to the change of the objective function to accelerate or slow down the adjustment speed of the control strategy; S5, multi-neighborhood search, formulating multiple neighborhood search strategies through the multi-neighborhood search layer, wherein the multiple neighborhood search strategies include global neighborhood search, local neighborhood search and chaotic search; S6. Output the optimal decision plan. When the maximum number of iterations is reached or the termination condition is met, the optimal control strategy is output through the output layer as the control instruction of the sandblasting process. Otherwise, repeat steps S2 to S5.

5. The data-driven sandblasting process control system according to claim 1, characterized in that: When there is a difference between the feedback signal and the preset target signal, the improved FAHP hierarchical analysis algorithm calculates the deviation between the current sandblasting process and the expected target according to the preset target signal and the feedback signal, and the output difference function formula is: In formula (1), F represents the deviation between the current sandblasting process and the expected target, z represents the feedback data of the workpiece surface parameters during the sandblasting process, y represents the calibration difference of the control signal during the operation of the sandblasting equipment, μ represents the output control signal of the sandblasting process, Δ(z) represents the change value of the surface parameters of the sandblasting workpiece during the operation of the sandblasting equipment, A(z) represents the change amplitude of the surface parameters of the sandblasting workpiece during the operation of the sandblasting equipment, and the measurement function formula of the change value of the surface parameters of the sandblasting workpiece during the operation of the sandblasting equipment is: Δ i (z)=w i (p i -s i ) (2) In formula (2), Δ i (z) represents the change value of the surface parameter of the i-th workpiece during the operation of the sandblasting equipment, i represents the ordinal number of the workpiece surface parameter variable during the sandblasting process, 1≤i≤N, N represents the number of workpiece surface parameter variables during the sandblasting process, p i represents the actual variable value of the surface parameter of the i-th workpiece during the sandblasting process, s i represents the self-loss variable of the surface parameters of the i-th workpiece during the sandblasting process, 0≤s≤1, w i represents the weight of the i-th workpiece surface parameter variable during the sandblasting process; During the operation of the sandblasting equipment, the output function formula of the change range of the surface parameters of the sandblasting workpiece during the operation of the sandblasting equipment is: In formula (3), A(z) is the variation range of the surface parameters of the blasting workpiece during the operation of the blasting equipment. is the average value of all parameter variable changes.

6. A data-driven sandblasting process control system according to claim 1, characterized in that: The electrical interface performs noise reduction processing through a frequency selective filter, and uses a low-distortion, high-speed analog-to-digital converter to convert analog control signals into digital control signals for transmission. During the transmission process, automatic error detection and correction processing is achieved through wide amplitude modulation, high-speed modulation, code division multiplexing, and error correction coding.

7. A data-driven sandblasting process control system according to claim 1, characterized in that: The memory module uses a cloud database to store historical control schemes of the sandblasting process and the corresponding surface parameters of the sandblasted workpiece. The cloud database uses an access counter to record the number of times the stored information is called, and uses a timer to trigger a clearing operation. The timer periodically clears the historical control schemes of the sandblasting process and the corresponding surface parameters of the sandblasted workpiece based on a least-used cache strategy. The least-used cache strategy sorts the stored information items from large to small weights according to the number of times the stored information items are called and the calling time, and clears the stored information items in reverse order. Before performing the operation, the user saves the historical control schemes of the sandblasting process and the corresponding surface parameters of the sandblasted workpiece to a local storage device or uploads them to a storage platform for long-term storage.

8. A data-driven sandblasting process control method, applied to a data-driven sandblasting process control system according to any one of claims 1 to 7, characterized in that: The following steps are involved: Step 1: Use a multi-point sensor group to obtain the process parameters in the sandblasting process in real time, and use a visual sensor to obtain the surface parameters of the workpiece to be sandblasted in real time, and then convert the obtained parameter information into a digital signal through a signal conditioner and transmit it to the cloud for processing; Step 2: extracting workpiece surface feature information using a cloud processor, and using a nearest interpolation algorithm to standardize the extracted workpiece surface feature points, and then identifying the workpiece surface feature information using an edge detection algorithm, wherein the workpiece surface feature information includes at least geometric shape features, texture features, and material features; Step 3: using an adaptive control decision model to make a control decision on the sandblasting process, wherein the adaptive control decision model establishes a model of the workpiece to be sandblasted based on the extracted surface feature information of the workpiece, and generates a sandblasting process control instruction according to the model of the workpiece to be sandblasted and the sandblasting parameters obtained in real time; Step 4: Obtain a control signal according to the generated sandblasting process control instruction, and send the sandblasting process control signal to the controller through the electrical interface to control the sandblasting equipment to perform the operation. When the sandblasting equipment performs the operation, a closed-loop feedback mechanism is used to feed back the status of the workpiece in the sandblasting process in real time. When the feedback signal differs from the preset target signal, the improved FAHP hierarchical analysis algorithm is used to dynamically adjust the sandblasting process control signal. Step 5: Record the historical control scheme of the sandblasting process and the surface parameters of the sandblasted workpiece to facilitate the subsequent call of similar control operations.

Images