An adaptive high altitude machine control system and method

Abstract

This invention discloses an adaptive high-altitude robot control system and method, relating to the field of intelligent algorithm control for sandblasting. The adaptive sandblasting robot control system proposed in this invention analyzes the current environmental data and robot status through model analysis, which can provide intelligent and highly adaptive working mode judgment for the sandblasting robot, achieving higher operating efficiency, energy efficiency and reliability, while reducing human intervention and improving the level of automation.

Description

Technical Field

[0001] This invention relates to the field of intelligent algorithm control for sandblasting, specifically an adaptive high-altitude machine control system and method. Background Technology

[0002] An industrial robot is a machine device capable of automatically performing various industrial tasks. It typically consists of a front-end mechanical structure and a back-end control system. Industrial robots perform precise and efficient operations on production lines, replacing or assisting human labor in repetitive, tedious, or dangerous tasks. The applications of industrial robots are very broad, including automobile manufacturing, electronics assembly, metal processing, and logistics warehousing. They can perform various tasks such as welding, assembly, handling, painting, grinding, and quality inspection. Through IoT technology and data analysis, industrial robots can achieve highly flexible movements and precise operations, improving equipment intelligence, production efficiency, quality, and safety.

[0003] In the control system of high-altitude sandblasting robots, the selection of working modes is necessary, especially the switching between "sandblasting" and "sand return." This relies heavily on the operator's experience and judgment. The robot is typically programmed to perform predetermined tasks under specific conditions, including the sandblasting intensity, angle, and work area. Before starting the operation, the operator needs to manually set these parameters and decide whether the robot should perform sandblasting or sand return based on the estimated work requirements. During actual operation, the operator needs to monitor the working environment and the robot's status in real time to determine whether to switch working modes. For example, if the sandblasting effect is detected as poor or the sandblasting of a specific area is completed, the operator will manually instruct the robot to switch to sand return mode. This approach depends on the operator's understanding of the working environment and mastery of the robot's performance.

[0004] Existing high-altitude sandblasting robots rely too heavily on manual judgment in selecting working modes, which limits their efficiency and adaptability. In order to improve operational efficiency and adaptability, it is necessary to introduce a more intelligent working mode selection mechanism. Summary of the Invention

[0005] This invention provides an adaptive high-altitude machine control system and method, which aims to use intelligent algorithms to assist decision-making and select the most suitable working mode for the current conditions.

[0006] One type of adaptive high-altitude machine control system includes:

[0007] The data acquisition module is used to collect climate status data, machine status data, and material property data, and to standardize the data.

[0008] The model training module is used to train the GBM model based on standardized climate state data, machine state data, and material property data as features.

[0009] The model prediction module is used to predict the robot's working mode based on the output of the GBM model. The working mode includes sandblasting mode and sand return mode.

[0010] The decision output module is used to select the robot's working mode based on the prediction results output by the model; wherein, the model training module specifically includes:

[0011] The model initialization unit is used to set the initial predicted value, which is the average value or constant of the target variable; and to set the initial model: F0(x) = average(y);

[0012] The residual calculation unit is used to calculate the residual between the current model and the actual target value;

[0013] The decision tree fitting unit is used to fit a new decision tree and predict the current residual.

[0014] The model update unit is used to update the current model and integrate the output of the new decision tree;

[0015] The model iteration unit is used to repeatedly build new trees and update the model until a predetermined number of iterations is reached or the model performance is optimal.

[0016] Furthermore, in the residual calculation unit, the calculation process for calculating the residual between the current model and the actual target value is as follows:

[0017]

[0018] Wherein, the r ti L(y) represents the residual of the i-th data point in the t-th iteration. i F(x) i )) represents the loss function, used to measure the predictive model F(x) i ) and the actual target value y i The difference between them, the y i F(x) represents the actual target value of the i-th data point. i ) represents the model's predicted value at the i-th data point, where F t-1 (x) represents the model after the (t-1)th iteration, the... This represents the partial derivative.

[0019] Furthermore, in the decision tree fitting unit, the fitting process for fitting a new decision tree and predicting the current residual is as follows: Decision tree h t (x) By learning the dataset Perform residual fitting, where N represents the total number of data points in the dataset.

[0020] Furthermore, the model update unit updates the current model, and the update process for integrating the new decision tree output is as follows: based on F... t (x)=F t-1 (x)+α·h t (x) is updated, where α represents the learning rate.

[0021] Furthermore, the climate condition data includes at least temperature, humidity, and wind speed; the machine condition data includes at least sandblasting intensity and sandblasting speed; and the material property data includes at least sandblasting type and particle size.

[0022] An adaptive aerial work platform control method, which is based on an adaptive aerial work platform control system as described above, includes the following steps:

[0023] S1. Collect climate status data, machine status data, and material property data, and standardize the data;

[0024] S2. Use standardized climate state data, machine state data, and material property data as features to train the GBM model;

[0025] S3. Based on the output of the GBM model, predict the working mode of the robot, which includes sandblasting mode and sand return mode;

[0026] S4. Select the robot's working mode based on the prediction results output by the model;

[0027] Specifically, step S3 includes the following sub-steps:

[0028] S301. Set initial predicted values, wherein the initial predicted values ​​are the average value or constant of the target variable; and set the initial model: F0(x) = average(y);

[0029] S302. Calculate the residual between the current model and the actual target value;

[0030] S303. Fit a new decision tree and predict the current residuals;

[0031] S304. Update the current model and integrate the output of the new decision tree;

[0032] S305. Repeatedly build new trees and update the model until the predetermined number of iterations is reached or the model performance is optimal.

[0033] Furthermore, in step S302, the calculation process for the residual between the current model and the actual target value is as follows:

[0034]

[0035] Wherein, the r ti L(y) represents the residual of the i-th data point in the t-th iteration. i F(x) i )) represents the loss function, used to measure the predictive model F(x) i ) and the actual target value y i The difference between them, the y i F(x) represents the actual target value of the i-th data point. i ) represents the model's predicted value at the i-th data point, where F t-1 (x) represents the model after the (t-1)th iteration, the... This represents the partial derivative.

[0036] Furthermore, in step S303, the fitting process for fitting a new decision tree and predicting the current residual is as follows: Decision tree h t (x) By learning the dataset Perform residual fitting, where N represents the total number of data points in the dataset.

[0037] Furthermore, in step S304, the update process for updating the current model and integrating the output of the new decision tree is as follows: based on F t (x)=F t-1 (x)+α·h t (x) is updated, where α represents the learning rate.

[0038] Furthermore, step S1 specifically includes the following sub-steps:

[0039] S101. Collect climate state data, machine state data and material property data. The climate state data includes at least temperature, humidity and wind speed. The machine state data includes at least sandblasting intensity and sandblasting speed. The material property data includes at least sandblasting type and particle size.

[0040] S102. Standardize the individual parameters in the climate state data, machine state data, and material property data respectively;

[0041] S103. Perform weighted averages on the standardized climate state data, machine state data, and material property data respectively to obtain a comprehensive climate state index, a comprehensive machine state index, and a comprehensive material property index. Use the comprehensive climate state index, comprehensive machine state index, and comprehensive material property index as input features of the model.

[0042] The beneficial effects of the invention are:

[0043] (1) The adaptive sandblasting robot control system proposed in this invention analyzes the current environmental data and robot status through the GBM model, which can provide intelligent and highly adaptive working mode judgment for the sandblasting robot, achieve higher operating efficiency, energy efficiency and reliability, while reducing human intervention and improving the level of automation. Attached Figure Description

[0044] Figure 1 A flowchart of an adaptive high-altitude machine control method provided in an embodiment of the present invention;

[0045] Figure 2 This is a schematic diagram of the structure of a terminal device for adaptive high-altitude machine control proposed in an embodiment of the present invention;

[0046] Figure 3 This is a schematic diagram of the structure of an adaptive high-altitude machine control system proposed in an embodiment of the present invention. Detailed Implementation

[0047] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings, but the scope of protection of the present invention is not limited to the following description.

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention; that is, the described embodiments are only a part of the embodiments of the invention, and not all of them. The components of the embodiments of the invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0049] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention. It should be noted that relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0050] Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0051] The features and performance of the present invention will be further described in detail below with reference to embodiments.

[0052] An adaptive high-altitude machine control system includes:

[0053] The data acquisition module is used to collect climate status data, machine status data, and material property data, and to standardize the data.

[0054] The model training module is used to train the GBM model based on standardized climate state data, machine state data, and material property data as features.

[0055] The model prediction module is used to predict the robot's working mode based on the output of the GBM model. The working mode includes sandblasting mode and sand return mode.

[0056] The decision output module is used to select the robot's working mode based on the prediction results output by the model; wherein, the model training module specifically includes:

[0057] The model initialization unit is used to set the initial predicted value, which is the average value or constant of the target variable; and to set the initial model: F0(x) = average(y);

[0058] The residual calculation unit is used to calculate the residual between the current model and the actual target value;

[0059] The decision tree fitting unit is used to fit a new decision tree and predict the current residual.

[0060] The model update unit is used to update the current model and integrate the output of the new decision tree;

[0061] The model iteration unit is used to repeatedly build new trees and update the model until a predetermined number of iterations is reached or the model performance is optimal.

[0062] Furthermore, in the residual calculation unit, the calculation process for calculating the residual between the current model and the actual target value is as follows:

[0063]

[0064] Wherein, the r ti L(y) represents the residual of the i-th data point in the t-th iteration. i F(x) i )) represents the loss function, used to measure the predictive model F(x) i ) and the actual target value y i The difference between them, the y i F(x) represents the actual target value of the i-th data point. i ) represents the model's predicted value at the i-th data point, where F t-1 (x) represents the model after the (t-1)th iteration, the... This represents the partial derivative.

[0065] Furthermore, in the decision tree fitting unit, the fitting process for fitting a new decision tree and predicting the current residual is as follows: Decision tree h t (x) By learning the dataset Perform residual fitting, where N represents the total number of data points in the dataset.

[0066] Furthermore, the model update unit updates the current model, and the update process for integrating the new decision tree output is as follows: based on F... t (x)=F t-1 (x)+α·h t (x) is updated, where α represents the learning rate.

[0067] Furthermore, the climate status data includes at least temperature, humidity, and wind speed; the machine status data includes at least sandblasting intensity and sandblasting speed; and the material property data includes at least sandblasting type and particle size. It should be noted that the parameters in the data combination can be replaced and modified according to the actual situation. The types of parameters only indicate the degree of influence of the parameters on the operation of the sandblasting robot.

[0068] Furthermore, for the above embodiments, a possible output result is proposed as follows:

[0069] The GBM model outputs a probability score between 0 and 1, representing the likelihood of the "sandblasting" mode. A threshold (e.g., 0.3) is set to determine its suitable working mode, i.e.:

[0070] P 喷砂 =GBM (Comprehensive Climate State Index, Comprehensive Machine State Index, Comprehensive Machine State Index, Comprehensive Material Properties);

[0071] The decision-making logic for its working mode is as follows: when P 喷砂 If P is ≥0.3, then select the sandblasting working mode. 喷砂If the value is less than 0.3, then the sand return mode is selected. It should be noted that the above threshold is for illustrative purposes only, and the range of the threshold can be selected and modified according to the output type of the GBM model or the type of task used.

[0072] As a further preferred embodiment of the above, an adaptive high-altitude machine control method is proposed, which is implemented based on an adaptive high-altitude machine control system as described in any of the above embodiments.

[0073] like Figure 1 This includes the following steps:

[0074] S1. Collect climate status data, machine status data, and material property data, and standardize the data;

[0075] S2. Use standardized climate state data, machine state data, and material property data as features to train the GBM model;

[0076] S3. Based on the output of the GBM model, predict the working mode of the robot, which includes sandblasting mode and sand return mode;

[0077] S4. Select the robot's working mode based on the prediction results output by the model;

[0078] Specifically, step S3 includes the following sub-steps:

[0079] S301. Set initial predicted values, wherein the initial predicted values ​​are the average value or constant of the target variable; and set the initial model: F0(x) = average(y);

[0080] S302. Calculate the residual between the current model and the actual target value;

[0081] S303. Fit a new decision tree and predict the current residuals;

[0082] S304. Update the current model and integrate the output of the new decision tree;

[0083] S305. Repeatedly build new trees and update the model until the predetermined number of iterations is reached or the model performance is optimal.

[0084] Furthermore, in step S302, the calculation process for the residual between the current model and the actual target value is as follows:

[0085]

[0086] Wherein, the r ti L(y) represents the residual of the i-th data point in the t-th iteration. i F(x) i)) represents the loss function, used to measure the predictive model F(x) i ) and the actual target value y i The difference between them, the y i F(x) represents the actual target value of the i-th data point. i ) represents the model's predicted value at the i-th data point, where F t-1 (x) represents the model after the (t-1)th iteration, the... This represents the partial derivative.

[0087] Furthermore, in step S303, the fitting process for fitting a new decision tree and predicting the current residual is as follows: Decision tree h t (x) By learning the dataset Perform residual fitting, where N represents the total number of data points in the dataset.

[0088] Furthermore, in step S304, the update process for updating the current model and integrating the output of the new decision tree is as follows: based on F t (x)=F t-1 (x)+α·h t (x) is updated, where α represents the learning rate.

[0089] Furthermore, step S1 specifically includes the following sub-steps:

[0090] S101. Collect climate state data, machine state data and material property data. The climate state data includes at least temperature, humidity and wind speed. The machine state data includes at least sandblasting intensity and sandblasting speed. The material property data includes at least sandblasting type and particle size.

[0091] S102. Standardize individual parameters in climate state data, machine state data, and material property data respectively. For example, a standardization process is proposed as follows:

[0092] For climate state data, calculate the average values ​​μ of temperature, humidity, and wind speed respectively. 温度 μ 湿度 μ 风速 and standard deviation σ 温度 σ 湿度 σ 风速 Temperature, humidity, and wind speed are standardized using standardized formulas, namely:

[0093]

[0094]

[0095]

[0096] Where std represents the standardized value of the corresponding parameter.

[0097] For the machine status data, calculate the average values ​​μ of the sandblasting intensity and sandblasting speed respectively. 强度 μ 速度 and standard deviation σ 强度 σ 速度 The sandblasting intensity and sandblasting speed are standardized using standardized formulas, namely:

[0098]

[0099]

[0100] For material property standardization, the mean and standard deviation of particle size are calculated, and particle size is standardized using the standardization formula, i.e.:

[0101]

[0102] The sandblasting type is standardized using a unique thermal coding system.

[0103] S103. Weighted averages are calculated on the standardized climate state data, machine state data, and material property data to obtain a comprehensive climate state index, a comprehensive machine state index, and a comprehensive material property index. These indices are then used as input features to the model. An exemplary calculation process for the weighted average is proposed as follows:

[0104] Comprehensive Climate State Index = w1·Temperature std +w2·Humidity std +w3·Wind Speed std ;

[0105] Comprehensive machine condition index = w4·intensity std +w5 Speed std ;

[0106] Comprehensive material property index = w6·particle size std +w7 Sandblasting type std ;

[0107] Wherein, w1, w2, w3, w4, w5, w6, and w7 represent the weight values ​​of the corresponding parameters. Furthermore, w1 + w2 + w3 = 1, w4 + w5 = 1, and w6 + w7 = 1 to meet the weight requirements. In addition, in actual production, other parameters can be added as features. The selection method mainly depends on the degree of influence on the working mode of the sandblasting robot. For example, three-dimensional data (x, y, z coordinates in this solution), such as machine position, can be added, belonging to machine status data, or it can be a separate data type. The specific standardization process is as follows: standardize each dimension separately, transforming data of different scales to a unified scale to facilitate the training and prediction of machine learning models, i.e.:

[0108] Calculate the average value μ x μ y μ z and standard deviation σ x σ y σ z ;

[0109] Standardize each dimension separately:

[0110] As a further preferred implementation, an adaptive high-altitude machine control terminal device is proposed, such as... Figure 2 The terminal device 200 includes at least one memory 210, at least one processor 220, and a bus 230 connecting different platform systems.

[0111] The memory 210 may include a readable medium in the form of volatile memory, such as random access memory (RAM) 211 and / or cache memory 212, and may further include read-only memory (ROM) 213.

[0112] The memory 210 also stores a computer program, which can be executed by the processor 220 to perform any of the adaptive high-altitude machine control methods described in the embodiments of this application. The specific implementation and technical effects are consistent with those described in the embodiments of the above methods, and some details will not be repeated here. The memory 210 may also include a program / utility 214 having a set (at least one) of program modules 215. Such program modules include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment.

[0113] Accordingly, processor 220 can execute the aforementioned computer program, as well as executable program / utility 214.

[0114] Bus 230 can represent one or more of several types of bus structures, including a memory bus or memory controller, peripheral bus, graphics acceleration port, processor, or a local bus using any of the various bus structures.

[0115] Terminal device 200 can also communicate with one or more external devices 240, such as keyboards, pointing devices, Bluetooth devices, etc., and with one or more devices capable of interacting with it, and / or with any device that enables it to communicate with one or more other computing devices (e.g., routers, modems, etc.). This communication can be performed via input / output (I / O) interface 250. Furthermore, terminal device 200 can also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via network adapter 260. Network adapter 260 can communicate with other modules of terminal device 200 via bus 230. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with terminal device 200, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms.

[0116] As a further preferred embodiment, a computer-readable storage medium for adaptive high-altitude machine control is proposed. The computer-readable storage medium stores instructions that, when executed by a processor, implement any of the aforementioned adaptive high-altitude machine control methods. Its specific implementation and the achieved technical effects are consistent with those described in the embodiments of the above methods, and some details will not be repeated.

[0117] Figure 3The illustration shows a program product 300 for implementing the above-described method, which may employ a portable compact disc read-only memory (CD-ROM) and include program code, and may run on a terminal device, such as a personal computer. However, the program product 300 of the present invention is not limited thereto. In this embodiment, the readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, apparatus, or device. The program product 300 may employ any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0118] Computer-readable storage media may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable storage medium may also be any readable medium other than a readable storage medium, capable of sending, propagating, or transmitting a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof. Program code for performing operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on a user computing device, partially on a user device, as a standalone software package, partially on a user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing devices can be connected to user computing devices via any type of network, including local area networks (LANs) or wide area networks (WANs), or they can be connected to external computing devices (e.g., via the Internet using an Internet service provider).

[0119] The above description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the concept described herein through the above teachings or related technologies or knowledge. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. An adaptive high-altitude machine control system, characterized in that, include: The data acquisition module is used to collect climate status data, machine status data, and material property data, and to standardize the data. The model training module is used to train the GBM model based on standardized climate state data, machine state data, and material property data as features. The model prediction module is used to predict the robot's working mode based on the output of the GBM model. The working mode includes sandblasting mode and sand return mode. The decision output module is used to select the robot's working mode based on the prediction results output by the model. The output of the GBM model is a probability score between 0 and 1, representing the likelihood of the "sandblasting" mode, i.e.: P-blasting = GBM (Comprehensive Climate Condition Index, Comprehensive Machine Condition Index, Comprehensive Material Properties); The decision logic for the working mode is as follows: when P_blasting ≥ 0.3, the blasting working mode is selected; when P_blasting < 0.3, the sand return mode is selected. The model training module specifically includes: The model initialization unit is used to set initial predicted values, which are the average value or a constant of the target variable; and to set the initial model: The residual calculation unit is used to calculate the residual between the current model and the actual target value; The decision tree fitting unit is used to fit a new decision tree and predict the current residual. The model update unit is used to update the current model and integrate the output of the new decision tree; The model iteration unit is used to repeatedly build new trees and update the model until a predetermined number of iterations is reached or the model performance is optimal. The climate status data includes at least temperature, humidity, and wind speed; the machine status data includes at least sandblasting intensity and sandblasting speed; and the material property data includes at least sandblasting type and particle size. The data standardization includes: For climate state data, calculate the average values ​​μ of temperature, humidity, and wind speed respectively. 温度 μ 湿度 μ 风速 and standard deviation σ 温度 σ 湿度 σ 风速 Temperature, humidity, and wind speed are standardized using standardized formulas, namely: ； ； ； Where std represents the standardized value of the corresponding parameter; For the machine status data, calculate the average values ​​μ of the sandblasting intensity and sandblasting speed respectively. 强度 μ 速度 and standard deviation σ 强度 σ 速度 The sandblasting intensity and sandblasting speed are standardized using standardized formulas, namely: For material property standardization, the mean and standard deviation of particle size are calculated, and the particle size is standardized using a standardization formula. ; The sandblasting type is standardized using a unique thermal coding system. It also includes: performing weighted averages on the standardized climate state data, machine state data, and material property data respectively to obtain a comprehensive climate state index, a comprehensive machine state index, and a comprehensive material property index. These comprehensive climate state index, comprehensive machine state index, and comprehensive material property index are then used as input features of the model. The calculation process for the weighted average is as follows: Comprehensive Climate State Index = w1·Temperature std +w2·Humidity std +w3·Wind Speed std ; Comprehensive machine condition index = w4·intensity std +w5 Speed std ; Comprehensive material property index = w6·particle size std +w7 Sandblasting type std ; Wherein, w1, w2, w3, w4, w5, w6, and w7 represent the weight values ​​of the corresponding parameters. In addition, w1+w2+w3=1, w4+w5=1, and w6+w7=1 to meet the weight requirements.

2. The adaptive high-altitude machine control system as described in claim 1, characterized in that, The calculation process for the residual between the current model and the actual target value in the residual calculation unit is as follows: ; Among them, the Indicates the first In the first iteration The residuals of the data points, the This represents the loss function, used to evaluate the predictive model. Compared with the actual target value The differences between them, the Indicates the first The actual target value of each data point, the The model represents the first time. The predicted value of each data point, Indicates the first The model after rounds of iterations, the This represents the partial derivative.

3. The adaptive high-altitude machine control system as described in claim 1, characterized in that, In the decision tree fitting unit, the fitting process for fitting a new decision tree and predicting the current residual is as follows: Decision Tree By learning the dataset Perform residual fitting, wherein, the This represents the total number of data points in the dataset.

4. The adaptive high-altitude machine control system as described in claim 1, characterized in that, In the model update unit, the update process for updating the current model and integrating the output of the new decision tree is as follows: Based on... To update, wherein, the This represents the learning rate.

5. An adaptive high-altitude machine control method, the method being implemented based on an adaptive high-altitude machine control system according to any one of claims 1-4, characterized in that, Includes the following steps: S1. Collect climate status data, machine status data, and material property data, and standardize the data; S2. GBM training is performed using standardized climate state data, machine state data, and material property data as features; S3. Based on the output of GBM, predict the working mode of the robot, which includes sandblasting mode and sand return mode; S4. Select the robot's working mode based on the prediction results output by the model; Specifically, step S3 includes the following sub-steps: S301. Set initial predicted values, wherein the initial predicted values ​​are the average value or a constant of the target variable; and set the initial model: ; S302. Calculate the residual between the current model and the actual target value; S303. Fit a new decision tree and predict the current residuals; S304. Update and integrate the output of the new decision tree; S305. Repeat step S304 until the predetermined number of iterations is reached or the model performance is optimal.

6. The adaptive high-altitude machine control method as described in claim 5, characterized in that, In step S302, the calculation process for the residual between the current model and the actual target value is as follows: ; Among them, the Indicates the first In the first iteration The residual of each data point, the This represents the loss function, used to evaluate the predictive model. Compared with the actual target value The differences between them, the Indicates the first The actual target value of each data point, the The model represents the first time. The predicted value of each data point, Indicates the first The model after rounds of iterations, the This represents the partial derivative.

7. The adaptive high-altitude machine control method as described in claim 5, characterized in that, In step S303, the fitting process for fitting a new decision tree and predicting the current residual is as follows: Decision Tree By learning the dataset Perform residual fitting, wherein, the This represents the total number of data points in the dataset.

8. The adaptive high-altitude machine control method as described in claim 5, characterized in that, In step S304, the update process for updating the current model and integrating the output of the new decision tree is as follows: Based on... Line update, wherein, This represents the learning rate.

9. The adaptive high-altitude machine control method as described in claim 5, characterized in that, Step S1 specifically includes the following sub-steps: S101. Collect climate state data, machine state data and material property data. The climate state data includes at least temperature, humidity and wind speed. The machine state data includes at least sandblasting intensity and sandblasting speed. The material property data includes at least sandblasting type and particle size. S102. Standardize the individual parameters in the climate state data, machine state data, and material property data respectively; S103. Perform weighted averages on the standardized climate state data, machine state data, and material property data respectively to obtain a comprehensive climate state index, a comprehensive machine state index, and a comprehensive material property index. Use the comprehensive climate state index, comprehensive machine state index, and comprehensive material property index as input features of the model.

Images