Multi-motor cooperative angle adjusting method for concrete grinding equipment

Abstract

The invention discloses a multi-motor cooperative angle adjusting method for concrete grinding equipment, and relates to the technical field of industrial automation control. Concrete surface data are collected through multiple sensors, data accuracy is ensured through data fusion and Kalman filtering, then the angle of a grinding head and the rotating speed of a motor are dynamically adjusted through inverse kinematics and fuzzy control, the grinding head is made to be in the optimal working state, and the grinding quality and efficiency are improved; meanwhile, a reinforcement learning algorithm optimizes polishing parameters, and the adaptability to a complex surface is enhanced; and an intelligent adjusting mechanism reduces manual intervention, reduces equipment abrasion, prolongs the service life, reduces the construction cost, and provides an efficient and intelligent grinding scheme for the building industry.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial automation control, and in particular to a method for adjusting the coordinated angles of multiple motors in concrete grinding equipment. Background Art

[0002] Concrete grinding technology occupies a core position in the construction of buildings and infrastructure. Its quality and efficiency directly affect the durability and aesthetics of the project. With the increasing requirements of modern buildings for surface flatness and precision, intelligent and precise grinding equipment has become a key direction for the development of the industry. Traditional grinding equipment mostly relies on manual operation or simple mechanical adjustment, and generally has problems such as low efficiency, uneven grinding effects, and poor adaptability to complex surfaces. These methods are usually unable to respond to the dynamic changes of the concrete surface in real time, resulting in unstable grinding quality, high manual intervention costs, and increased equipment wear. The core challenge lies in how to achieve dynamic coordinated adjustment of the grinding head angle and motor speed, and how to perform adaptive optimization based on different concrete surface characteristics. Existing technologies make it difficult to accurately establish a mapping relationship between grinding effects and equipment parameters based on real-time monitoring of contact angles, pressures, and wear conditions. They also lack data-driven self-learning capabilities to cope with diverse construction scenarios. Summary of the Invention

[0003] The purpose of the present invention is to provide a multi-motor coordinated angle adjustment method for concrete grinding equipment, which significantly improves the quality and efficiency of concrete grinding through multi-sensor data fusion, dynamic parameter adjustment and reinforcement learning algorithm.

[0004] The purpose of the present invention can be achieved through the following technical solutions:

[0005] This application provides a method for adjusting the angle of multiple motors in a concrete grinding machine, comprising the following steps:

[0006] The concrete surface roughness, hardness, grinding head contact angle and pressure data are acquired through multiple sensors, and a data fusion algorithm is used to generate real-time surface characteristic distribution to obtain comprehensive surface state parameters.

[0007] Based on the comprehensive surface state parameters, the contact angle and pressure data are denoised using the Kalman filter algorithm to generate the first optimized parameter set and determine the preliminary adjustment range of the grinding head angle and motor speed;

[0008] When the contact angle deviation in the first optimization parameter set exceeds a preset threshold, the adjustment amount of the grinding head joint is calculated by an inverse kinematics algorithm, a first angle adjustment instruction is generated, and accurate grinding head posture parameters are obtained;

[0009] When the uneven pressure distribution in the first optimization parameter set exceeds a preset threshold, the mapping relationship between the pressure distribution and the speed is analyzed by a fuzzy control algorithm to generate a first speed adjustment instruction and determine a dynamic adjustment scheme for the motor speed;

[0010] The first angle adjustment instruction and the first speed adjustment instruction are iteratively optimized through a reinforcement learning algorithm to obtain a mapping relationship between historical polishing data and real-time feedback, generate a second optimization parameter set, and obtain an adaptive adjustment strategy;

[0011] According to the second optimized parameter set, a motion control algorithm is used to drive the grinding head joint and motor to perform dynamic adjustments, generate real-time execution instructions, and obtain the actual motion trajectory and speed state of the grinding head;

[0012] Multi-sensors are used to collect surface flatness and equipment wear data in real time during the grinding process. A data fusion algorithm is used to update the surface property distribution and equipment status, generate the first feedback data set, and determine the real-time change trend of the grinding effect and equipment wear.

[0013] Furthermore, comprehensive surface state parameters are obtained, including:

[0014] The concrete surface roughness, hardness, grinding head contact angle and pressure data are acquired through multiple sensors to generate a raw sensor data set. The roughness, hardness, contact angle and pressure data are then fused through a weighted average algorithm to generate a real-time surface property distribution.

[0015] Based on the real-time surface characteristic distribution, the principal component analysis algorithm is used to extract the main surface characteristic parameters and determine the comprehensive surface state parameters. When the deviation between the comprehensive surface state parameters and the preset standard parameters exceeds the threshold, the contact angle and pressure of the grinding head are adjusted according to the deviation value to generate the adjusted grinding parameters;

[0016] The polishing state is updated by adjusting the polishing parameters, new sensor data is obtained, and updated surface state parameters are generated. Then, based on the degree of matching between the updated surface state parameters and the standard parameters, it is determined whether the surface treatment is completed to obtain the final surface state.

[0017] Furthermore, determine the initial adjustment range of the grinding head angle and motor speed, including:

[0018] The contact angle data and pressure data in the comprehensive surface state parameters are obtained through the sensor, and the Kalman filter algorithm is used to denoise the data to obtain a first optimized parameter set. When the variance of the contact angle data in the first optimized parameter set exceeds a preset threshold, the grinding head angle is preliminarily adjusted to determine the adjustment range.

[0019] When the variance is lower than the threshold, the current angle is maintained and the preliminary adjustment range is obtained. When the pressure data fluctuation in the first optimization parameter set exceeds the preset threshold, the motor speed is preliminarily adjusted to determine the adjustment range. When the fluctuation is lower than the threshold, the current speed is maintained and the preliminary adjustment range is obtained according to the preliminary adjustment range.

[0020] Furthermore, accurate grinding head posture parameters are obtained, including:

[0021] When the contact angle deviation exceeds the preset threshold, the real-time contact angle data is obtained through the sensor, and the data is filtered to obtain a smooth contact angle value. Then, the inverse kinematics algorithm is used to calculate the adjustment amount of the grinding head joint based on the smooth contact angle value and determine the joint motion control parameters;

[0022] Generate angle adjustment instructions based on joint motion control parameters and output them to the grinding head drive system to obtain the adjusted joint state;

[0023] Obtain the adjusted joint state, use posture error analysis to calculate the deviation between the grinding head posture and the target posture, and determine the posture error value. When the posture error value exceeds the preset threshold, iterative optimization is performed through the algorithm to update the inverse kinematics algorithm parameters and obtain the optimized adjustment amount;

[0024] Based on the optimized adjustment amount, a new angle adjustment instruction is generated and output to the grinding head drive system to determine the precise grinding head posture parameters. Then, through parameter calibration processing, the precise grinding head posture parameters are verified to obtain the final posture control data.

[0025] Furthermore, a dynamic adjustment scheme for the motor speed is determined, specifically including:

[0026] When the uneven pressure distribution exceeds the preset threshold, the real-time pressure distribution data is obtained through the sensor to obtain the pressure distribution matrix. The mapping relationship between the pressure distribution matrix and the speed is then analyzed through the fuzzy control algorithm to determine the fuzzy rule set.

[0027] Generate speed adjustment instructions based on fuzzy rule sets and real-time pressure distribution data to obtain instruction sequences, and then use the instruction sequences to dynamically adjust the motor speed to determine the adjusted speed parameters;

[0028] By monitoring the pressure distribution after adjustment in real time, it is determined whether it still exceeds the preset threshold value and obtains the monitoring result. When the monitoring result shows that the uneven pressure distribution still exceeds the threshold value, the fuzzy rule set is adjusted through the iterative optimization algorithm to obtain the optimized rule set;

[0029] The speed adjustment instructions are regenerated according to the optimized rule set to determine the final motor speed adjustment plan.

[0030] Furthermore, an adaptive adjustment strategy is obtained, which specifically includes:

[0031] The reinforcement learning algorithm is used to obtain historical polishing data and real-time feedback data to generate a first mapping relationship. When the convergence of the first mapping relationship is lower than a preset threshold, the reward function of the reinforcement learning algorithm is adjusted to regenerate a second mapping relationship.

[0032] Extracting characteristic distributions of the first angle adjustment instruction and the first speed adjustment instruction according to the second mapping relationship, determining the parameter adjustment direction, and generating a second optimized parameter set;

[0033] The second optimized parameter set is used to update the first angle adjustment instruction and the first speed adjustment instruction to obtain an adaptive adjustment instruction set, and then real-time polishing effect data is obtained to generate a data feedback loop and update the mapping relationship.

[0034] Furthermore, the actual motion trajectory and speed state of the grinding head are obtained, specifically including:

[0035] The input data is obtained from the second optimized parameter set, converted into the initial parameters of the motion control algorithm through a preset mapping relationship, and the basic configuration driven by the algorithm is obtained. The drive signal of the grinding head joint is then generated through the motion control algorithm to obtain the real-time execution instruction of the joint movement;

[0036] According to the real-time execution instructions, the drive control module is used to adjust the operating status of the grinding head joint and the motor to obtain the original data of the actual motion trajectory;

[0037] The coordinates of key points are extracted from the raw data of the actual motion trajectory, and a smooth trajectory curve is generated through an interpolation algorithm to obtain the optimized motion trajectory. When the deviation between the optimized motion trajectory and the preset trajectory exceeds a threshold, the motor speed is adjusted through a feedback control algorithm to obtain an updated speed state;

[0038] According to the updated speed status, the state monitoring module is used to obtain the real-time operation data of the motor and obtain the dynamic operation status of the grinding head. Then, the characteristic parameters are extracted from the dynamic operation status, and new real-time execution instructions are generated through the preset mapping relationship to obtain the adjusted motion trajectory and speed status of the grinding head.

[0039] Furthermore, the real-time trend of grinding effect and equipment wear is determined, including:

[0040] The surface flatness and equipment wear data of the grinding process are collected in real time by multiple sensors to obtain a first sensor data set. The first sensor data set is then fused using a Kalman filter algorithm to update the surface property distribution and equipment status to obtain a first fused data set.

[0041] When the surface flatness value in the first fused data set exceeds the preset flatness threshold, the flatness data is smoothed using mean filtering to obtain a second fused data set. The time series change rate of the surface characteristic distribution is then calculated to determine the real-time change trend of the polishing effect.

[0042] The support vector machine algorithm is used to classify the equipment wear level using the equipment status data in the second fused dataset to obtain the equipment wear status sequence. The time series change rate of the wear level is then calculated to determine the real-time change trend of equipment wear.

[0043] Based on the changing trends of polishing effect and equipment wear, a weighted average method is used to generate a comprehensive feedback data set to obtain real-time process optimization parameters.

[0044] Furthermore, it also includes: based on the first feedback data set, using a reinforcement learning algorithm to update the second optimization parameter set online, generating a third optimization parameter set, and obtaining a long-term adaptive polishing parameter optimization solution.

[0045] Furthermore, it also includes: converting the third optimized parameter set into the final execution instruction through the motion control algorithm, driving the grinding head and the motor to work together, generating the final grinding trajectory and speed sequence, and obtaining an efficient and uniform grinding effect and a stable equipment operation state.

[0046] The beneficial effects of the present invention are:

[0047] The present invention uses multiple sensors to collect concrete surface characteristic data in real time, and uses data fusion and denoising algorithms to generate accurate comprehensive surface state parameters. Based on this data, the grinding head angle and motor speed are dynamically adjusted to ensure that the grinding head is always in the optimal working state. This effectively solves the problems of uneven grinding effect and low efficiency caused by the inability of traditional grinding equipment to monitor and adjust grinding parameters in real time, significantly improves the grinding quality and efficiency, and ensures that the flatness and precision of the concrete surface meet the requirements of modern construction.

[0048] By iteratively optimizing grinding parameters using a reinforcement learning algorithm, the system can automatically adjust grinding strategies based on varying concrete surface characteristics. By real-time monitoring of surface flatness and equipment wear during the grinding process, combined with data fusion and posture error analysis, the system can adaptively optimize the grinding head angle and motor speed to better adapt to complex and changing concrete surfaces. This effectively addresses the poor adaptability of traditional grinding equipment to complex surfaces, avoids unstable grinding quality caused by varying surface characteristics, and improves the versatility and reliability of the equipment in diverse construction scenarios.

[0049] Through intelligent parameter adjustment and adaptive learning mechanism, the angle of the grinding head and the motor speed can be dynamically optimized according to real-time data, avoiding equipment damage caused by excessive wear or improper operation. At the same time, this method reduces the frequency of manual intervention, reduces the error and labor intensity of manual operation, and analyzes historical grinding data and real-time feedback data through reinforcement learning algorithms. The equipment can continuously optimize the grinding strategy and generate long-term adaptive grinding parameter optimization solutions, effectively solving the problems of increased wear of traditional grinding equipment and high cost of manual intervention, extending the service life of the equipment, improving construction efficiency, and reducing overall construction costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] For better understanding and implementation, the technical solution of the present application is described in detail below with reference to the accompanying drawings.

[0051] Figure 1 A flow chart of a method for adjusting the angle of multiple motors in a concrete grinding machine provided in this application;

[0052] Figure 2 A schematic diagram of a flow chart of a multi-motor coordinated angle adjustment method for concrete grinding equipment to obtain accurate grinding head posture parameters provided in this application;

[0053] Figure 3 A flowchart of an adaptive adjustment strategy for a multi-motor coordinated angle adjustment method for concrete grinding equipment provided in this application. DETAILED DESCRIPTION

[0054] To further illustrate the technical means and effectiveness of the present invention in achieving its intended purpose, exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present application. Rather, they are merely examples of methods and systems consistent with certain aspects of the present application, as detailed in the appended claims.

[0055] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. As used in this application and the appended claims, the singular forms "a," "an," "the," and "the" are intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0056] The following describes in detail the specific implementation methods, features and effects of the present invention in conjunction with the accompanying drawings and preferred embodiments.

[0057] See also Figure 1-Figure 3 This embodiment provides a method for adjusting the angle of multiple motors in a concrete grinding device, including the following steps:

[0058] S1. Use multiple sensors to obtain data on the concrete surface roughness, hardness, and contact angle and pressure of the grinding head. Use data fusion algorithms to generate real-time surface characteristic distribution and obtain comprehensive surface state parameters.

[0059] Furthermore, in step S1, comprehensive surface state parameters are obtained, specifically including:

[0060] The concrete surface roughness, hardness, grinding head contact angle and pressure data are acquired through multiple sensors to generate a raw sensor data set. The roughness, hardness, contact angle and pressure data are then fused through a weighted average algorithm to generate a real-time surface property distribution.

[0061] Based on the real-time surface characteristic distribution, the principal component analysis algorithm is used to extract the main surface characteristic parameters and determine the comprehensive surface state parameters. When the deviation between the comprehensive surface state parameters and the preset standard parameters exceeds the threshold, the contact angle and pressure of the grinding head are adjusted according to the deviation value to generate the adjusted grinding parameters;

[0062] The polishing state is updated by adjusting the polishing parameters, new sensor data is obtained, and updated surface state parameters are generated. Then, based on the degree of matching between the updated surface state parameters and the standard parameters, it is determined whether the surface treatment is completed to obtain the final surface state.

[0063] Specifically, in the weighted average algorithm, the importance of different data can be trained and analyzed through machine learning algorithms (such as neural networks and support vector machines) based on the historical data and actual working conditions of concrete grinding, so as to determine reasonable weights and ensure that the real-time surface characteristic distribution is more in line with the actual situation; in the principal component analysis algorithm, the cumulative contribution rate of the principal component is set to 85%-95% as the standard for extracting the number of principal components to ensure that the main surface characteristic parameters can effectively reflect the surface state; and the setting of the threshold can refer to a large amount of concrete grinding project case data, and combine statistical methods to calculate the mean and standard deviation of the data, and scientifically determine the threshold range by adding or subtracting a certain multiple of the standard deviation from the mean.

[0064] By collaboratively collecting data from multiple sensors, combined with weighted average and principal component analysis algorithms, comprehensive and precise perception of the concrete surface condition is achieved, enabling rapid and accurate acquisition of comprehensive surface condition parameters. Dynamic adjustment of the grinding head angle and pressure based on parameter deviations, combined with real-time determination of surface treatment completion, creates a closed-loop control system that effectively avoids under- or over-grinding, significantly improving the quality and efficiency of concrete surface grinding while also reducing equipment energy consumption and wear, and increasing operational stability and service life.

[0065] S2. Based on the comprehensive surface state parameters, the contact angle and pressure data are denoised using the Kalman filter algorithm to generate a first optimized parameter set and determine the preliminary adjustment range of the grinding head angle and motor speed;

[0066] Furthermore, in step S2, the initial adjustment range of the grinding head angle and the motor speed is determined, specifically including:

[0067] The contact angle data and pressure data in the comprehensive surface state parameters are obtained through the sensor, and the Kalman filter algorithm is used to denoise the data to obtain a first optimized parameter set. When the variance of the contact angle data in the first optimized parameter set exceeds a preset threshold, the grinding head angle is preliminarily adjusted to determine the adjustment range.

[0068] When the variance is lower than the threshold, the current angle is maintained and the preliminary adjustment range is obtained. When the pressure data fluctuation in the first optimization parameter set exceeds the preset threshold, the motor speed is preliminarily adjusted to determine the adjustment range. When the fluctuation is lower than the threshold, the current speed is maintained and the preliminary adjustment range is obtained according to the preliminary adjustment range.

[0069] Specifically, using the Kalman filter algorithm to denoise contact angle and pressure data can effectively remove noise interference from sensor-collected data, improving data accuracy and reliability. Based on the denoised data, preliminary adjustments to the grinding head angle and motor speed are made based on appropriately set thresholds. This allows the grinding equipment to quickly adapt to changes in the concrete surface condition. While ensuring grinding quality, the appropriate operating parameter range is planned in advance, reducing unnecessary adjustments and improving grinding efficiency. This also reduces equipment wear and energy consumption caused by inaccurate parameters, laying a solid foundation for subsequent more precise grinding parameter optimization and stable equipment operation.

[0070] S3. When the contact angle deviation in the first optimization parameter set exceeds a preset threshold, the adjustment amount of the grinding head joint is calculated by an inverse kinematics algorithm, a first angle adjustment instruction is generated, and accurate grinding head posture parameters are obtained;

[0071] Furthermore, in step S3, accurate grinding head posture parameters are obtained, specifically including:

[0072] S31. When the contact angle deviation exceeds a preset threshold, real-time contact angle data is obtained through a sensor, and data filtering is performed to obtain a smoothed contact angle value. Then, an inverse kinematics algorithm is used to calculate the adjustment amount of the grinding head joint based on the smoothed contact angle value to determine the joint motion control parameters;

[0073] S32. Generate an angle adjustment instruction based on the joint motion control parameters and output it to the grinding head drive system to obtain the adjusted joint state;

[0074] S33, obtaining the adjusted joint state, using posture error analysis to calculate the deviation between the grinding head posture and the target posture, and determining the posture error value. When the posture error value exceeds a preset threshold, iterative optimization of the algorithm is performed to update the inverse kinematics algorithm parameters to obtain the optimized adjustment amount;

[0075] S34. Generate a new angle adjustment instruction based on the optimized adjustment amount, output it to the grinding head drive system, determine the precise grinding head posture parameters, and then verify the precise grinding head posture parameters through parameter calibration to obtain the final posture control data.

[0076] Specifically, in terms of threshold setting, the contact angle deviation threshold can be determined according to the concrete grinding process standards and equipment accuracy requirements. For example, for high-precision grinding tasks, the contact angle deviation threshold can be set to ±0.5; the attitude error value threshold comprehensively considers the grinding surface flatness requirements and the equipment motion accuracy to ensure that the deviation between the grinding head posture and the ideal state will not affect the grinding quality, such as setting the attitude error value threshold to ±1°; when calculating the adjustment amount of the grinding head joint, based on the inverse kinematics algorithm, it is necessary to establish a kinematic model of the grinding head, and use the smooth contact angle value as the target output. Combined with the current joint angle and position information of the grinding head, the inverse solution of the kinematic equation is solved to obtain each The angle or distance that the joint needs to rotate or move is used to determine the joint motion control parameters; angle adjustment instructions are generated based on the joint motion control parameters. A mapping relationship between the control parameters and the drive signal needs to be established to convert the joint motion control parameters (such as angle, speed, etc.) into electrical signal instructions that can be recognized by the grinding head drive system, such as pulse signals, voltage signals, etc., to drive the motor or hydraulic device to drive the joint movement; when calculating the deviation between the grinding head posture and the target posture through posture error analysis, the three-dimensional space vector method can be used to represent the current posture and target posture of the grinding head with space vectors respectively, and the posture deviation value can be quantified by calculating parameters such as the angle between the two vectors and the translation distance.

[0077] When the contact angle deviation exceeds the preset threshold, the real-time contact angle data is first filtered, which can effectively reduce noise interference and provide more accurate data for subsequent calculations. The inverse kinematics algorithm is used to calculate the joint adjustment amount and generate adjustment instructions, which can quickly adjust the grinding head posture. Through posture error analysis and iterative optimization, the grinding head posture can be continuously corrected to make it closer to the target posture. Finally, after parameter calibration and verification, the accuracy of the final posture control data is ensured, which can significantly improve the accuracy of the grinding head posture control, enable the grinding equipment to better adapt to the complex conditions of the concrete surface, improve the grinding quality and efficiency, and reduce grinding defects caused by inaccurate posture.

[0078] S4. When the uneven pressure distribution in the first optimization parameter set exceeds a preset threshold, a mapping relationship between the pressure distribution and the speed is analyzed by a fuzzy control algorithm to generate a first speed adjustment instruction and determine a dynamic adjustment scheme for the motor speed;

[0079] Furthermore, in step S4, a dynamic adjustment scheme for the motor speed is determined, specifically including:

[0080] When the uneven pressure distribution exceeds the preset threshold, the real-time pressure distribution data is obtained through the sensor to obtain the pressure distribution matrix. The mapping relationship between the pressure distribution matrix and the speed is then analyzed through the fuzzy control algorithm to determine the fuzzy rule set.

[0081] Generate speed adjustment instructions based on fuzzy rule sets and real-time pressure distribution data to obtain instruction sequences, and then use the instruction sequences to dynamically adjust the motor speed to determine the adjusted speed parameters;

[0082] By monitoring the pressure distribution after adjustment in real time, it is determined whether it still exceeds the preset threshold value and obtains the monitoring result. When the monitoring result shows that the uneven pressure distribution still exceeds the threshold value, the fuzzy rule set is adjusted through the iterative optimization algorithm to obtain the optimized rule set;

[0083] The speed adjustment instructions are regenerated according to the optimized rule set to determine the final motor speed adjustment plan.

[0084] Specifically, by analyzing the mapping relationship between pressure distribution and speed through fuzzy control algorithm and dynamically adjusting the motor speed, the problem of uneven pressure distribution during concrete grinding can be quickly improved, avoiding local excessive or insufficient grinding; combining real-time monitoring and iterative optimization, the fuzzy rule set is continuously adjusted so that the system can adapt to different grinding conditions, improve the stability of the grinding process, ensure the consistency of grinding quality, reduce equipment loss, and realize efficient and intelligent grinding operations.

[0085] S5. Iteratively optimize the first angle adjustment instruction and the first speed adjustment instruction using a reinforcement learning algorithm to obtain a mapping relationship between historical polishing data and real-time feedback, generate a second optimization parameter set, and obtain an adaptive adjustment strategy;

[0086] Furthermore, in step S5, an adaptive adjustment strategy is obtained, which specifically includes:

[0087] S51. Obtain historical polishing data and real-time feedback data through a reinforcement learning algorithm to generate a first mapping relationship. When the convergence of the first mapping relationship is lower than a preset threshold, adjust the reward function of the reinforcement learning algorithm and regenerate a second mapping relationship.

[0088] S52: extracting characteristic distributions of the first angle adjustment instruction and the first speed adjustment instruction according to the second mapping relationship, determining a parameter adjustment direction, and generating a second optimized parameter set;

[0089] S53: Use the second optimization parameter set to update the first angle adjustment instruction and the first speed adjustment instruction to obtain an adaptive adjustment instruction set, then obtain real-time polishing effect data, generate a data feedback loop, and update the mapping relationship.

[0090] Specifically, when making a convergence judgment, the convergence judgment criterion can be obtained by calculating the change amplitude of the loss function value in multiple consecutive iterations (such as the change in 10 consecutive iterations is less than 0.01); when adjusting the reward function, a heuristic search algorithm such as a genetic algorithm is used to search for the optimal combination in the parameter space, or the degree of influence of each parameter on the strategy is calculated through policy gradient, and the reward function is adjusted in a targeted manner; for feature extraction, a convolutional neural network or a recurrent neural network is used to mine the spatiotemporal characteristics of the angle and speed adjustment instructions, and an algorithm such as a decision tree and a support vector machine is used to train the model based on historical data. The parameter adjustment direction is determined according to the extracted features, and the implementation path of each link is clarified.

[0091] By building a mapping relationship between historical and real-time data through reinforcement learning and dynamically optimizing the reward function based on convergence, the algorithm can quickly adapt to complex and changeable grinding conditions; by extracting instruction features to determine the direction of parameter adjustment, the angle and speed adjustment instructions can be accurately optimized to form an adaptive adjustment instruction set; combined with the feedback loop of real-time grinding effect data, the mapping relationship is continuously updated to achieve dynamic optimization of grinding parameters, effectively improving the adaptability of grinding equipment to different concrete surfaces, significantly improving grinding quality and efficiency, reducing manual intervention, and ensuring the stability and continuity of the grinding process.

[0092] S6. Based on the second optimized parameter set, a motion control algorithm is used to drive the grinding head joint and the motor to perform dynamic adjustment, generate real-time execution instructions, and obtain the actual motion trajectory and speed state of the grinding head;

[0093] Furthermore, in step S6, the actual motion trajectory and speed state of the grinding head are obtained, which specifically includes:

[0094] The input data is obtained from the second optimized parameter set, converted into the initial parameters of the motion control algorithm through a preset mapping relationship, and the basic configuration driven by the algorithm is obtained. The drive signal of the grinding head joint is then generated through the motion control algorithm to obtain the real-time execution instruction of the joint movement;

[0095] According to the real-time execution instructions, the drive control module is used to adjust the operating status of the grinding head joint and the motor to obtain the original data of the actual motion trajectory;

[0096] The coordinates of key points are extracted from the raw data of the actual motion trajectory, and a smooth trajectory curve is generated through an interpolation algorithm to obtain the optimized motion trajectory. When the deviation between the optimized motion trajectory and the preset trajectory exceeds a threshold, the motor speed is adjusted through a feedback control algorithm to obtain an updated speed state;

[0097] According to the updated speed status, the state monitoring module is used to obtain the real-time operation data of the motor and obtain the dynamic operation status of the grinding head. Then, the characteristic parameters are extracted from the dynamic operation status, and new real-time execution instructions are generated through the preset mapping relationship to obtain the adjusted motion trajectory and speed status of the grinding head.

[0098] Specifically, by converting the second optimized parameter set into the initial parameters of the motion control algorithm, the grinding head joint drive signal and real-time execution instructions are accurately generated to ensure the accuracy of the initial movement of the grinding head; the drive control module is used to adjust the joint and motor operating status, obtain the actual motion trajectory original data, and optimize the trajectory curve through the interpolation algorithm to effectively improve the smoothness and continuity of the motion trajectory; when the motion trajectory deviation exceeds the threshold, the motor speed is adjusted in time with the help of the feedback control algorithm to achieve coordinated correction of the motion trajectory and speed state; combined with the state monitoring module to collect motor operating data in real time, and dynamically generate new instructions according to the preset mapping relationship to form a closed-loop control, so that the grinding head can quickly respond to changes in working conditions and accurately match the preset trajectory, effectively improving the stability of the grinding process, trajectory accuracy and grinding efficiency, and ensuring the high quality and uniformity of concrete surface grinding.

[0099] S7. Using multiple sensors to collect real-time surface flatness and equipment wear data during the grinding process, using a data fusion algorithm to update surface property distribution and equipment status, generating a first feedback data set, and determining real-time trends in grinding effects and equipment wear;

[0100] Furthermore, in step S7, the real-time change trend of the grinding effect and equipment wear is determined, specifically including:

[0101] The surface flatness and equipment wear data of the grinding process are collected in real time by multiple sensors to obtain a first sensor data set. The first sensor data set is then fused using a Kalman filter algorithm to update the surface property distribution and equipment status to obtain a first fused data set.

[0102] When the surface flatness value in the first fused data set exceeds the preset flatness threshold, the flatness data is smoothed using mean filtering to obtain a second fused data set. The time series change rate of the surface characteristic distribution is then calculated to determine the real-time change trend of the polishing effect.

[0103] The support vector machine algorithm is used to classify the equipment wear level using the equipment status data in the second fused dataset to obtain the equipment wear status sequence. The time series change rate of the wear level is then calculated to determine the real-time change trend of equipment wear.

[0104] Based on the changing trends of polishing effect and equipment wear, a weighted average method is used to generate a comprehensive feedback data set to obtain real-time process optimization parameters.

[0105] Specifically, multi-sensor real-time collection of surface flatness and equipment wear data during the grinding process, and the Kalman filter algorithm is used for data fusion to effectively improve the accuracy and reliability of the data, laying the foundation for accurately grasping the surface property distribution and equipment status; the data exceeding the flatness threshold is smoothed by mean filtering, and combined with the time series change rate analysis, it can keenly capture the dynamic change trend of the grinding effect; the support vector machine algorithm is used to scientifically classify the equipment wear level, and by calculating the wear level time series change rate, the equipment wear trend can be accurately predicted; finally, a comprehensive feedback data set is generated through weighted averaging, and the grinding effect and equipment wear trend are organically combined to obtain real-time process optimization parameters, which not only ensures the stability of the grinding quality, but also can plan equipment maintenance in advance, reduce sudden equipment failures, and achieve efficient and sustainable operation of the grinding process.

[0106] Furthermore, it also includes: based on the first feedback data set, using a reinforcement learning algorithm to update the second optimization parameter set online, generating a third optimization parameter set, and obtaining a long-term adaptive polishing parameter optimization solution.

[0107] Furthermore, a long-term adaptive polishing parameter optimization solution is obtained, which specifically includes:

[0108] Obtain the first feedback data set, use data preprocessing technology to extract feature vectors, generate an initial input data set, then process the initial input data set through the reinforcement learning algorithm, combine it with the second optimization parameter set, calculate the reward function value, and obtain the updated parameter adjustment direction;

[0109] When the parameter adjustment direction meets the preset convergence threshold, the online update mechanism is used to update the second optimization parameter set to generate a third optimization parameter set. Then, based on the third optimization parameter set, the polishing parameter optimization scheme is adjusted to generate an adaptive control strategy.

[0110] Through dynamic parameter adjustment technology, the execution effect of the adaptive control strategy is monitored to obtain real-time feedback data. Then, a data-driven adjustment method is used to analyze the real-time feedback data, update the first feedback data set, and generate a new feature vector;

[0111] The new feature vector is processed by the reinforcement learning algorithm, and the third optimization parameter set is iteratively optimized to obtain a long-term adaptive polishing parameter optimization solution.

[0112] Specifically, in the reinforcement learning algorithm, the design of the reward function enables effective optimization of the polishing parameters. From the perspective of polishing effect, surface flatness can be used as an important indicator. The closer the surface flatness is to the preset ideal value, the higher the positive reward is given. Conversely, when the flatness exceeds the preset flatness threshold, a negative reward is given, and the greater the deviation, the greater the negative reward. This incentivizes the algorithm to adjust parameters to improve polishing effect. For equipment wear, the reward is set according to the equipment wear level classified by the support vector machine algorithm. Low wear levels correspond to high positive rewards. As the wear level increases, the reward gradually decreases or even becomes negative, guiding the algorithm to avoid excessive wear of the equipment. At the same time, energy consumption factors must also be considered. If energy consumption is at a reasonably low level during the polishing process, a certain positive reward is given; if energy consumption is too high, a negative reward is given, prompting the algorithm to optimize parameters to reduce energy consumption. Taking these factors into consideration, a reward function can be constructed through a weighted summation method. Appropriate weights are assigned to each factor based on the actual polishing focus. For example, when focusing on polishing effect, the weight of the surface flatness indicator can be appropriately increased. This will form a comprehensive and reasonable reward function, providing a clear optimization direction for the reinforcement learning algorithm.

[0113] Long-term adaptive optimization of polishing parameters is achieved. Effective features are extracted from the first feedback data set. The reward function value is calculated with the help of reinforcement learning algorithm combined with the second optimization parameter set. The parameter adjustment direction can be accurately determined to ensure that the parameter adjustment is in a more optimal direction. When the convergence threshold is met, the third optimization parameter set is updated and generated, and the polishing plan is adjusted to form an adaptive control strategy. During the strategy execution process, dynamic parameter adjustment technology and data-driven adjustment methods are used to monitor and analyze feedback data in real time, continuously update the first feedback data set, and then iteratively optimize the third optimization parameter set. This series of operations enables the polishing parameters to be dynamically adjusted according to the actual polishing conditions, effectively improving the polishing quality and efficiency, reducing equipment wear and energy consumption, and improving the stability and adaptability of the entire polishing process to achieve long-term stable polishing effects.

[0114] Furthermore, it also includes: converting the third optimized parameter set into the final execution instruction through the motion control algorithm, driving the grinding head and the motor to work together, generating the final grinding trajectory and speed sequence, and obtaining an efficient and uniform grinding effect and a stable equipment operation state.

[0115] Furthermore, an efficient and uniform polishing effect and a stable equipment operation state are obtained, specifically including: obtaining a third optimized parameter set, decomposing it into trajectory parameters and speed parameters through a motion control algorithm, obtaining an initial execution instruction set, and when the initial execution instruction set meets the preset trajectory smoothness threshold, generating a polishing trajectory sequence through an interpolation algorithm; when it does not meet the threshold, adjusting the trajectory parameters and re-decomposing them to obtain a smooth polishing trajectory sequence.

[0116] According to the smooth grinding trajectory sequence, the PID control algorithm is used to generate a speed sequence that matches the trajectory to obtain a speed instruction sequence. Then, through the collaborative drive mechanism, the speed instruction sequence and the grinding trajectory sequence are integrated into the final execution instruction set to obtain the drive signal set.

[0117] Extract the grinding head drive signal from the drive signal set and determine whether the signal meets the equipment stable operation threshold. If it does, generate a grinding head control sequence. If it does not, optimize the signal parameters to obtain a stable grinding head control sequence.

[0118] The motor collaborative working parameters are obtained, the motor drive signal is adjusted according to the stable grinding head control sequence, and the motor control sequence is generated. Then, through the real-time feedback mechanism, the operating status data is extracted from the grinding head control sequence and the motor control sequence to determine whether a uniform grinding effect and a stable equipment state are achieved, and the final execution result is obtained.

[0119] Specifically, the motion control algorithm deeply decomposes and precisely regulates the third optimized parameter set, effectively converting the parameters into trajectory parameters and speed parameters that meet the actual grinding requirements. When the trajectory smoothness threshold is met, the interpolation algorithm is used to generate a continuous and stable grinding trajectory sequence. When it is not met, timely adjustment and optimization are performed to ensure the smoothness and accuracy of the grinding trajectory. The PID control algorithm is combined to generate a matching speed sequence, which is integrated into the final execution instruction set through the collaborative drive mechanism to achieve the collaborative operation of the grinding head and the motor. Through strict verification of the drive signal, the grinding head control sequence is guaranteed to meet the stable operation requirements of the equipment, and the motor drive signal is optimized based on the motor collaborative working parameters to form a closed-loop real-time feedback mechanism. Ultimately, this process can achieve efficient coordination between the grinding head and the motor, not only generating a precise and uniform final grinding trajectory and speed sequence, effectively improving grinding efficiency and surface flatness, but also ensuring the stability of the equipment's operating state, reducing abnormal equipment wear, and reducing energy consumption, achieving efficient, stable, and high-quality grinding operations.

[0120] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as a preferred embodiment as above, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to equivalent embodiments using the technical contents disclosed above without departing from the scope of the technical solution of the present invention. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A method for adjusting the angle of multiple motors in a concrete grinding machine, characterized by: The steps include: The concrete surface roughness, hardness, grinding head contact angle and pressure data are acquired through multiple sensors, and a data fusion algorithm is used to generate real-time surface characteristic distribution to obtain comprehensive surface state parameters. Based on the comprehensive surface state parameters, the contact angle and pressure data are denoised using the Kalman filter algorithm to generate the first optimized parameter set and determine the preliminary adjustment range of the grinding head angle and motor speed; When the contact angle deviation in the first optimization parameter set exceeds a preset threshold, the adjustment amount of the grinding head joint is calculated by an inverse kinematics algorithm, a first angle adjustment instruction is generated, and accurate grinding head posture parameters are obtained; When the uneven pressure distribution in the first optimization parameter set exceeds a preset threshold, the mapping relationship between the pressure distribution and the speed is analyzed by a fuzzy control algorithm to generate a first speed adjustment instruction and determine a dynamic adjustment scheme for the motor speed; The first angle adjustment instruction and the first speed adjustment instruction are iteratively optimized through a reinforcement learning algorithm to obtain a mapping relationship between historical polishing data and real-time feedback, generate a second optimization parameter set, and obtain an adaptive adjustment strategy; According to the second optimized parameter set, a motion control algorithm is used to drive the grinding head joint and motor to perform dynamic adjustments, generate real-time execution instructions, and obtain the actual motion trajectory and speed state of the grinding head; Multi-sensors are used to collect surface flatness and equipment wear data in real time during the grinding process. A data fusion algorithm is used to update the surface property distribution and equipment status, generate the first feedback data set, and determine the real-time change trend of the grinding effect and equipment wear.

2. A method for adjusting the angle of multiple motors in a concrete grinding device according to claim 1, characterized in that: Obtain comprehensive surface state parameters, including: The concrete surface roughness, hardness, grinding head contact angle and pressure data are acquired through multiple sensors to generate a raw sensor data set. The roughness, hardness, contact angle and pressure data are then fused through a weighted average algorithm to generate a real-time surface property distribution. Based on the real-time surface characteristic distribution, the principal component analysis algorithm is used to extract the main surface characteristic parameters and determine the comprehensive surface state parameters. When the deviation between the comprehensive surface state parameters and the preset standard parameters exceeds the threshold, the contact angle and pressure of the grinding head are adjusted according to the deviation value to generate the adjusted grinding parameters; The polishing state is updated by adjusting the polishing parameters, new sensor data is obtained, and updated surface state parameters are generated. Then, based on the degree of matching between the updated surface state parameters and the standard parameters, it is determined whether the surface treatment is completed to obtain the final surface state.

3. The method for adjusting the angle of multiple motors in a concrete grinding device according to claim 1, characterized in that: Determine the initial adjustment range of the grinding head angle and motor speed, including: The contact angle data and pressure data in the comprehensive surface state parameters are obtained through the sensor, and the Kalman filter algorithm is used to denoise the data to obtain a first optimized parameter set. When the variance of the contact angle data in the first optimized parameter set exceeds a preset threshold, the grinding head angle is preliminarily adjusted to determine the adjustment range. When the variance is lower than the threshold, the current angle is maintained and the preliminary adjustment range is obtained. When the pressure data fluctuation in the first optimization parameter set exceeds the preset threshold, the motor speed is preliminarily adjusted to determine the adjustment range. When the fluctuation is lower than the threshold, the current speed is maintained and the preliminary adjustment range is obtained according to the preliminary adjustment range.

4. The method for adjusting the angle of multiple motors in a concrete grinding device according to claim 1, wherein: Get accurate grinding head posture parameters, including: When the contact angle deviation exceeds the preset threshold, the real-time contact angle data is obtained through the sensor, and the data is filtered to obtain the smooth contact angle value. Then, the inverse kinematics algorithm is used to calculate the adjustment amount of the grinding head joint based on the smooth contact angle value and determine the joint motion control parameters; Generate angle adjustment instructions based on joint motion control parameters and output them to the grinding head drive system to obtain the adjusted joint state; Obtain the adjusted joint state, use posture error analysis to calculate the deviation between the grinding head posture and the target posture, and determine the posture error value. When the posture error value exceeds the preset threshold, iterative optimization is performed through the algorithm to update the inverse kinematics algorithm parameters and obtain the optimized adjustment amount; Based on the optimized adjustment amount, a new angle adjustment instruction is generated and output to the grinding head drive system to determine the precise grinding head posture parameters. Then, through parameter calibration processing, the precise grinding head posture parameters are verified to obtain the final posture control data.

5. The method for adjusting the angle of multiple motors in a concrete grinding device according to claim 1, characterized in that: Determine the dynamic adjustment scheme of the motor speed, including: When the uneven pressure distribution exceeds the preset threshold, the real-time pressure distribution data is obtained through the sensor to obtain the pressure distribution matrix. The mapping relationship between the pressure distribution matrix and the speed is then analyzed through the fuzzy control algorithm to determine the fuzzy rule set. Generate speed adjustment instructions based on fuzzy rule sets and real-time pressure distribution data to obtain instruction sequences, and then use the instruction sequences to dynamically adjust the motor speed to determine the adjusted speed parameters; By monitoring the pressure distribution after adjustment in real time, it is determined whether it still exceeds the preset threshold value and obtains the monitoring result. When the monitoring result shows that the uneven pressure distribution still exceeds the threshold value, the fuzzy rule set is adjusted through the iterative optimization algorithm to obtain the optimized rule set; The speed adjustment instructions are regenerated according to the optimized rule set to determine the final motor speed adjustment plan.

6. The method for adjusting the angle of multiple motors in a concrete grinding device according to claim 1, characterized in that: Get the adaptive adjustment strategy, including: The reinforcement learning algorithm is used to obtain historical polishing data and real-time feedback data to generate a first mapping relationship. When the convergence of the first mapping relationship is lower than a preset threshold, the reward function of the reinforcement learning algorithm is adjusted to regenerate a second mapping relationship. Extracting characteristic distributions of the first angle adjustment instruction and the first speed adjustment instruction according to the second mapping relationship, determining the parameter adjustment direction, and generating a second optimized parameter set; The second optimized parameter set is used to update the first angle adjustment instruction and the first speed adjustment instruction to obtain an adaptive adjustment instruction set, and then real-time polishing effect data is obtained to generate a data feedback loop and update the mapping relationship.

7. The method for adjusting the angle of multiple motors in a concrete grinding device according to claim 1, characterized in that: Get the actual motion trajectory and speed state of the grinding head, including: The input data is obtained from the second optimized parameter set, converted into the initial parameters of the motion control algorithm through a preset mapping relationship, and the basic configuration driven by the algorithm is obtained. The drive signal of the grinding head joint is then generated through the motion control algorithm to obtain the real-time execution instruction of the joint movement; According to the real-time execution instructions, the drive control module is used to adjust the operating status of the grinding head joint and the motor to obtain the original data of the actual motion trajectory; The coordinates of key points are extracted from the raw data of the actual motion trajectory, and a smooth trajectory curve is generated through an interpolation algorithm to obtain the optimized motion trajectory. When the deviation between the optimized motion trajectory and the preset trajectory exceeds a threshold, the motor speed is adjusted through a feedback control algorithm to obtain an updated speed state; According to the updated speed status, the state monitoring module is used to obtain the real-time operation data of the motor and obtain the dynamic operation status of the grinding head. Then, the characteristic parameters are extracted from the dynamic operation status, and new real-time execution instructions are generated through the preset mapping relationship to obtain the adjusted motion trajectory and speed status of the grinding head.

8. The method for adjusting the angle of multiple motors in a concrete grinding device according to claim 1, characterized in that: Determine the real-time trend of grinding performance and equipment wear, including: The surface flatness and equipment wear data of the grinding process are collected in real time by multiple sensors to obtain a first sensor data set. The first sensor data set is then fused using a Kalman filter algorithm to update the surface property distribution and equipment status to obtain a first fused data set. When the surface flatness value in the first fused data set exceeds the preset flatness threshold, the flatness data is smoothed using mean filtering to obtain a second fused data set. The time series change rate of the surface characteristic distribution is then calculated to determine the real-time change trend of the polishing effect. The support vector machine algorithm is used to classify the equipment wear level using the equipment status data in the second fused dataset to obtain the equipment wear status sequence. The time series change rate of the wear level is then calculated to determine the real-time change trend of equipment wear. Based on the changing trends of polishing effect and equipment wear, a weighted average method is used to generate a comprehensive feedback data set to obtain real-time process optimization parameters.

9. The method for adjusting the angle of multiple motors in a concrete grinding device according to claim 1, characterized in that: Also includes: According to the first feedback data set, the second optimization parameter set is updated online using a reinforcement learning algorithm to generate a third optimization parameter set, thereby obtaining a long-term adaptive polishing parameter optimization solution.

10. The method for adjusting the angle of multiple motors in a concrete grinding device according to claim 1, characterized in that: Also includes: The third optimized parameter set is converted into the final execution instruction through the motion control algorithm, driving the grinding head and motor to work together to generate the final grinding trajectory and speed sequence, thereby obtaining an efficient and uniform grinding effect and a stable equipment operation state.

Images