Automatic gate rust removal control system and rust removal method

Abstract

The invention provides an automatic gate rust removal control system and a rust removal method.The automatic gate rust removal control system comprises a detection unit, a control unit, a moving unit and a rust removal unit; the detection unit, the moving unit and the rust removal unit are all in signal connection with the control unit. The detection unit is used for acquiring a three-dimensional point cloud data set of the gate; the control unit is used for controlling the detection unit, the moving unit and the rust removal unit, and establishing a three-dimensional model and generating a rust removal path based on the three-dimensional point cloud data set; the moving unit is used for driving the derusting unit to move along the derusting path; the derusting unit is used for derusting. According to the automatic rust removal control system for the gate and the rust removal method, the rust removal device does not need to be manually operated to remove rust on the gate, and the rust removal convenience and safety are improved; the whole derusting process is automatically carried out, so that the derusting efficiency is improved, and the derusting effect is ensured; and the complex structure of the gate can be derusted conveniently, and the derusting effect is better.

Description

Technical Field

[0001] The invention relates to the technical field of gate rust removal, and in particular to a gate automatic rust removal control system and a rust removal method. Background Art

[0002] A ship lock is a box-shaped hydraulic structure that is used to ensure the smooth passage of ships and concentrate the water level difference on the channel. For example, the Three Gorges double-line continuous five-level ship lock is the main navigation facility of the Three Gorges hub. There are generally multiple gates in the ship lock for coordinated use. The gates are coated with a protective coating to ensure the long-term stable use of the gates. For example, the Three Gorges double-line five-level ship lock has been operating safely and stably for nearly 20 years.

[0003] However, after long-term use of metal structure equipment such as ship lock gates, the surface coating will gradually approach the design protection period, so it needs to be rust-removed;

[0004] The commonly used rust removal method is to use an aerial work platform or a bridge inspection vehicle to send workers to the location where the gate needs to be rusted, and then the workers use rust removal tools to perform rust removal operations;

[0005] However, in the actual rust removal process, since manual rust removal operations are required at the gate rust removal station, not only is the operation inconvenient, there are certain safety hazards, the work efficiency is also low, and manual operation is prone to incomplete rust removal, which affects the quality of rust removal. Summary of the invention

[0006] The main purpose of the present invention is to provide a gate automatic rust removal control system and a rust removal method to solve the problems of inconvenient operation and low efficiency in gate rust removal in the existing method.

[0007] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0008] A gate automatic rust removal control system, comprising a detection unit, a control unit, a moving unit and a rust removal unit;

[0009] The detection unit, the moving unit and the rust removal unit are all connected to the control unit by signal;

[0010] The detection unit is used to obtain a three-dimensional point cloud data set of the gate;

[0011] The control unit is used to control the detection unit, the moving unit and the rust removal unit, and to establish a three-dimensional model and generate a rust removal path based on the three-dimensional point cloud data set;

[0012] The moving unit is used to drive the rust removal unit to move along the rust removal path;

[0013] The rust removal unit is used for rust removal.

[0014] In a preferred solution, the detection unit includes a laser scanning module and a wireless transmission module, and the signal sending module is signal-connected to the detection unit.

[0015] In a preferred solution, the moving unit includes a lifting module, a front-rear moving module, a left-right moving module and a rotating module;

[0016] The lifting module is used to drive the rust removal unit to lift, the front and rear moving module is used to drive the rust removal unit to move front and rear, the left and right moving module is used to drive the rust removal unit to move left and right, and the rotating module is used to drive the rust removal unit to rotate 360 ​​degrees.

[0017] In a preferred solution, the rust removal unit includes a rust removal module and an image recording module;

[0018] The image recording module is used to record the rust removal process and the images after rust removal, and send the image signal to the control unit. The control unit receives the image signal and determines the rust removal effect according to the image signal.

[0019] In a preferred solution, the control unit is provided with a fault diagnosis module, an early warning module and a remote communication module;

[0020] The fault diagnosis module is used to monitor the working status of the detection unit, the mobile unit and the rust removal unit in real time, and compare and analyze whether a fault occurs;

[0021] The early warning module is used to send out alarm signals so that maintenance personnel can quickly locate and solve problems;

[0022] The remote communication module is used to transmit working signals and data.

[0023] A gate rust removal method comprises the following steps:

[0024] S1. Perform a comprehensive scan of the gate to obtain the gate surface morphology and generate a three-dimensional point cloud data set;

[0025] S2, processing the three-dimensional point cloud data set to determine the specific location and size that need to be rusted;

[0026] S3, calculating and obtaining the best rust removal path according to the specific location and size information of the rust removal;

[0027] S4, driving the rust removal unit to remove rust from the gate along the rust removal path by the mobile unit;

[0028] During the rust removal process, the rust-removed area is scanned to generate a new 3D point cloud data set;

[0029] S5, processing the three-dimensional point cloud data set to obtain the position and size that require secondary rust removal;

[0030] S6, calculating and obtaining the best rust removal path according to the position and size information in S5;

[0031] S7. Repeat S4-S6 until the rust is completely removed.

[0032] In the preferred solution, in S1 and S4, a laser scanner is used to perform a comprehensive scan of the target gate;

[0033] Scan 2-5 times to generate 2-5 sets of 3D point cloud data sets A n = {P1, P2, P3 ... P i}, A n represents the nth set of 3D point cloud data sets, P i Points representing a 3D point cloud dataset;

[0034] For the 3D point cloud dataset A n Processing, compared with P i (x i ,y i , z i ), P i (x i ,y i , z i ) is P i The coordinate value of

[0035] For the same position and the same P i (x i ,y i , z i ), included in A 0 , for different P at the same position i (x i ,y i , z i ), take the P with the largest number of occurrences i (x i ,y i , z i ) is included in A 0 , or calculate the average value and include it in A 0 , and finally get the complete A 0 .

[0036] In a preferred embodiment, in S2 and S5:

[0037] S21, denoising the 3D point cloud data set, for each point P in the point cloud set i (x i ,y i , z i ), the coordinate value calculation formula after filtering is as follows:

[0038]

[0039] Where N(i) represents point Pi The neighborhood point set, ω(p i , p j ) represents the point P i and P j The weight function is determined by the spatial distance and the normal vector angle;

[0040] S22, using Gaussian filtering to further smooth the data;

[0041] S23, calculating the local features of each point in the processed three-dimensional point cloud data set;

[0042] S24, using the calculated local features as input, constructing a random forest model;

[0043] S25. Obtain the location and size that require rust removal based on the random forest model.

[0044] In a preferred solution, in S23, the local features include the normal vector n of each point i and curvature k i .

[0045] In the preferred solution, in S3 and S6, the gate surface is divided into a plurality of small areas B1, B2, B3 ... B according to the location and size information of the rust removal required. n ,Then the optimal rust removal path is obtained based on the ant colony optimization algorithm;

[0046] In the divided area B n The following steps are included:

[0047] S31, dividing the gate surface into a number of continuous square areas, with the last area being set as B;

[0048] S32, starting from the first area, set to B m , and assume that the set of rust areas in the current region is {b1, b2, b3…b n};

[0049] where b n Represents the rust area in the current region;

[0050] S33, when Then calculate the next area B m+1 ;

[0051] when Then b x Move to next area B m+1 , until Then calculate the next area B m+1 ;

[0052] Assume B m =B n ;

[0053] where b x Represents the B m+1 rust area, b0 represents the maximum rust area allowed in the same area;

[0054] S34, let B m+1 =B m , the set of rust areas in the region is {b1, b2, b3…b n}, which includes the rust area included in the previous area;

[0055] S35, repeat S33-S34 until B m =B;

[0056] The above parts are divided into multiple small areas B1, B2, B3...B n , in S33-S35, B n Mark them one by one starting from B1.

[0057] The present invention provides a gate automatic rust removal control system and rust removal method. By adopting the above scheme, the following beneficial effects are achieved:

[0058] 1. There is no need to manually operate the rust removal device to remove rust from the gate, which increases the convenience and safety of rust removal.

[0059] 2. The entire rust removal process is carried out automatically, which is easy to operate, increases the efficiency of rust removal, and ensures the rust removal effect.

[0060] 3. It is convenient to remove rust from the complex structure of the gate, and the rust removal effect is better.

[0061] 4. It can accurately calculate and generate the best rust removal path, ensuring the efficiency and accuracy of rust removal.

[0062] 5. The rust removal process is carried out in different areas, which reduces the difficulty of rust removal and ensures the rust removal effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0064] Figure 1 The invention discloses a flow chart of an automatic gate rust removal control system and a rust removal method. DETAILED DESCRIPTION

[0065] Embodiment 1:

[0066] like Figure 1 As shown, a gate automatic rust removal control system includes a detection unit, a control unit, a moving unit and a rust removal unit, wherein:

[0067] Detection unit:

[0068] The detection unit of the present invention includes a laser scanning module and a wireless transmission module, which can perform a comprehensive scan of the gate and obtain a three-dimensional point cloud data set of the gate. The laser scanning module uses the principle of laser ranging to emit a laser beam to the surface of the gate and receive the reflected light. By measuring the flight time or phase change of the laser, the spatial position information of each point on the surface of the gate is accurately calculated, thereby generating three-dimensional point cloud data. The wireless transmission module is responsible for transmitting the collected three-dimensional point cloud data set to the control unit in real time, ensuring fast and stable transmission of data for subsequent processing and analysis.

[0069] Control unit:

[0070] The control unit is signal-connected with the detection unit, the moving unit and the rust removal unit, and plays a role in coordinating and controlling the work of each unit.

[0071] Based on the 3D point cloud data set received from the detection unit, the control unit uses a 3D modeling algorithm to build a 3D model of the gate. The 3D model can intuitively display the surface morphology and structural characteristics of the gate, providing accurate basic data for subsequent rust removal path planning. At the same time, the control unit uses a built-in intelligent algorithm to generate the best rust removal path based on the rust distribution reflected in the 3D model to ensure the efficiency and accuracy of rust removal.

[0072] In addition, the control unit is also equipped with a fault diagnosis module, an early warning module and a remote communication module. The fault diagnosis module uses real-time monitoring technology to monitor the working status of the detection unit, the mobile unit and the rust removal unit, and promptly determines whether a fault occurs by comparing and analyzing the normal working parameters and real-time operation data of each unit. Once a fault is detected, the early warning module will immediately issue an alarm signal, which can be in various forms such as sound and light alarms, so that maintenance personnel can quickly locate the fault location and take corresponding solutions, thereby minimizing system downtime and improving work efficiency. The remote communication module can realize information interaction between the system and external devices or remote control centers. It can be used to transmit working signals and data, facilitate operators to remotely monitor and manage the system, and is also conducive to collaborative work and data sharing between multiple devices.

[0073] Mobile unit:

[0074] The mobile unit includes a lifting module, a front and rear moving module, a left and right moving module and a rotating module, which can realize the flexible movement of the rust removal unit in three-dimensional space, ensuring that the rust removal unit can accurately perform comprehensive and efficient rust removal operations on the gate along the rust removal path generated by the control unit.

[0075] The lifting module, forward and backward movement module, left and right movement module and rotation module are all driven by motors or hydraulic transmission; the lifting module drives the rust removal unit to move up and down in the vertical direction to meet the rust removal needs of different height positions of the gate. The forward and backward movement module and the left and right movement module are responsible for controlling the forward and backward and left and right movement of the rust removal unit in the horizontal direction, respectively, so that the rust removal unit can be accurately positioned at the specific position where rust removal is required. The rotation module enables the rust removal unit to rotate 360 ​​degrees, further improving the flexibility and adaptability of the rust removal unit, and can effectively deal with rust of various complex shapes and angles on the gate surface.

[0076] Rust removal unit:

[0077] The rust removal unit includes a rust removal module and an image recording module. The rust removal module uses existing rust removal technologies, such as sandblasting and laser rust removal, to remove rust on the gate surface. During the rust removal process, the image recording module will record the rust removal process and the image after rust removal in real time, and send the image signal to the control unit. After receiving the image signal, the control unit uses image recognition and analysis technology to judge the rust removal effect based on the actual situation of the gate surface reflected in the image. If it is found that there is still rust that has not been completely removed, the control unit will adjust the rust removal path and parameters in time, and drive the rust removal unit again for secondary rust removal until the ideal rust removal effect is achieved.

[0078] Embodiment 2:

[0079] A gate rust removal method comprises the following steps:

[0080] S1. Perform a comprehensive scan of the gate to obtain the gate surface morphology and generate a three-dimensional point cloud data set;

[0081] S2, processing the three-dimensional point cloud data set to determine the specific location and size that need to be rusted;

[0082] S3, calculating and obtaining the best rust removal path according to the specific location and size information of the rust removal;

[0083] S4, driving the rust removal unit to remove rust from the gate along the rust removal path by the mobile unit;

[0084] During the rust removal process, the rust-removed area is scanned to generate a new 3D point cloud data set;

[0085] S5, processing the three-dimensional point cloud data set to obtain the position and size that require secondary rust removal;

[0086] S6, calculating and obtaining the best rust removal path according to the position and size information in S5;

[0087] S7. Repeat S4-S6 until the rust is completely removed.

[0088] In the preferred solution, in S1 and S4, a laser scanner is used to perform a comprehensive scan of the target gate;

[0089] Scan 2-5 times to generate 2-5 sets of 3D point cloud data sets A n = {P1, P2, P3...P i}, A n represents the nth set of 3D point cloud data sets, P i Points representing a 3D point cloud dataset;

[0090] For the 3D point cloud dataset A n Processing, compared with P i (x i ,y i , z i ), P i (x i ,y i , z i ) is P i The coordinate value of

[0091] For the same position and the same P i (x i ,y i , z i ), included in A 0 , for different P at the same position i (x i ,y i , z i ), take the P with the largest number of occurrences i (x i ,y i , z i ) is included in A 0 , or calculate the average value and include it in A 0 , and finally get the complete A 0 .

[0092] In a preferred embodiment, in S2 and S5:

[0093] S21, denoising the 3D point cloud data set, for each point P in the point cloud set i (x i ,y i , z i ), the coordinate value calculation formula after filtering is as follows:

[0094]

[0095] Where N(i) represents point P i The neighborhood point set, ω(p i , p j ) represents the point Pi and P j The weight function determined by the spatial distance and the normal vector angle:

[0096]

[0097] Among them, σ d and σ n Represent the parameters controlling the distance and normal vector weight, respectively, ||p i -p j || 2 Representative point P i and P j The square of the spatial distance, <n i , n j > 2 Representative point P i and P j The square of the normal vector angle;

[0098] v22, use Gaussian filtering to further smooth the data for point P i , and its coordinates after Gaussian filtering are:

[0099]

[0100] Among them, G(p i , p j ) represents the Gaussian kernel function:

[0101]

[0102] Here, ′ represents the standard deviation of the Gaussian kernel.

[0103] S23, calculating the local features of each point in the processed three-dimensional point cloud data set;

[0104] Local features include the normal vector n of each point i and curvature k i ;

[0105] Calculate the normal vector of each point: Use a method based on principal component analysis (PCA); for point P i The neighborhood point set N(i) of , construct the covariance matrix C:

[0106]

[0107] in, Represents the mean of the domain points, performs eigenvalue decomposition on the covariance matrix C, and the eigenvector corresponding to the minimum eigenvalue is point P i The normal vector n i .

[0108] Calculate the curvature: Use the method based on the change of the normal vector. For point P i , select two mutually perpendicular directions u and v in its neighborhood (which can be obtained by cross product of normal vectors, etc.), and calculate the partial derivatives of the normal vector in these two directions:

[0109]

[0110] Among them, point P i The curvature k i for:

[0111]

[0112] S24, using the calculated local features as input, constructing a random forest model;

[0113] From the preprocessed 3D point cloud data, a certain proportion (e.g., 20-70%) of points are randomly selected as training samples, and the coordinates (x, y, z), normal vector n and i and curvature k i As the eigenvector X:

[0114] Decision tree construction: In the feature vector of the training sample, a part of the features is randomly selected (for example, the number of selected features is determined by the feature selection parameter of the random forest), for example, m features are selected to form a new feature subset.

[0115] Based on these feature subsets, the best split features and split points are selected using criteria such as information gain and Gini index, and the samples are continuously divided into sub-nodes until the stopping condition is met (such as the number of samples in the node is less than a certain threshold, the depth of the tree reaches a preset value, etc.);

[0116] Repeat the above decision tree construction process to generate multiple decision trees, preferably 50-150 trees, to form a random forest model. During the training process, each decision tree is trained based on a different random feature subset and training sample subset to improve the generalization ability of the model.

[0117] S25. Obtain the location and size that require rust removal based on the random forest model.

[0118] The feature vector of the test sample is input into the trained random forest model, and each decision tree performs classification prediction on the sample to obtain a prediction result, preferably a category label of whether rust removal is needed, where 0 indicates that rust removal is not needed and 1 indicates that rust removal is needed;

[0119] For each sample point, the prediction results of all decision trees are combined and the final prediction category is determined by majority voting; if the majority of decision trees predict that the point needs rust removal, the point is marked as the location that needs rust removal, thereby determining the rust removal location and size of the entire gate surface.

[0120] In the preferred solution, in S3 and S6, the gate surface is divided into a plurality of small areas B1, B2, B3 ... B according to the location and size information of the rust removal required. n ,Then the optimal rust removal path is obtained based on the ant colony optimization algorithm;

[0121] Among them, each small area sets the corresponding pheromone concentration τ ij , the probability of an ant transferring from area i to area j when searching for a path is

[0122]

[0123] Among them, α and β represent parameters that control the relative importance of pheromone and heuristic information respectively. d ij Represents the distance from region i to j, allowed k Represents the set of areas that ant k can currently select; after the ant completes a traversal, it updates the pheromone on the path. The update formula is:

[0124]

[0125] Where ρ represents the volatility coefficient of pheromone, represents the pheromone increment left by ant k on path ij during this traversal, and its calculation formula is:

[0126]

[0127] Among them, Q represents a constant, L k Represents the total length of the path traversed by ant k this time; through multiple iterations, an optimal rust removal path with the shortest total path length and the highest efficiency is finally found, and the motion characteristics and constraints of the mobile unit are taken into consideration, the path is smoothed and optimized to ensure that the mobile unit can move smoothly and efficiently along the path.

[0128] In the divided area B n The following steps are included:

[0129] S31, dividing the gate surface into a number of continuous square areas, with the last area being set as B;

[0130] S32, starting from the first area, set to B m, and assume that the set of rust areas in the current region is {b1, b2, b3…b n};

[0131] where b n Represents the rust area in the current region;

[0132] S33, when Then calculate the next area B m+1 ;

[0133] when Then b x Move to next area B m+1 , until Then calculate the next area B m+1 ;

[0134] Assume B m =B n ;

[0135] where b x Represents the B m+1 rust area, b0 represents the maximum rust area allowed in the same area;

[0136] S34, let B m+1 =B m , the set of rust areas in the region is {b1, b2, b3…b n}, which includes the rust area included in the previous area;

[0137] S35, repeat S33-S34 until B m =B;

[0138] The above parts are divided into multiple small areas B1, B2, B3...B n , in S33-S35, B n Mark them one by one starting from B1.

[0139] The above embodiments are only preferred technical solutions of the present invention and should not be regarded as limiting the present invention. The protection scope of the present invention shall be the technical solutions recorded in the claims, including equivalent replacement solutions of the technical features in the technical solutions recorded in the claims. That is, equivalent replacement improvements within this scope are also within the protection scope of the present invention.

Claims

1. A gate automatic rust removal control system, characterized by: It includes a detection unit, a control unit, a moving unit and a rust removal unit; The detection unit, the moving unit and the rust removal unit are all connected to the control unit by signal; The detection unit is used to obtain a three-dimensional point cloud data set of the gate; The control unit is used to control the detection unit, the moving unit and the rust removal unit, and to establish a three-dimensional model and generate a rust removal path based on the three-dimensional point cloud data set; The moving unit is used to drive the rust removal unit to move along the rust removal path; The rust removal unit is used for rust removal.

2. According to claim 1, a gate automatic rust removal control system is characterized by: The detection unit includes a laser scanning module and a wireless transmission module, and the signal sending module is connected to the detection unit signal.

3. According to claim 1, a gate automatic rust removal control system is characterized by: The mobile unit includes a lifting module, a front-rear moving module, a left-right moving module and a rotating module; The lifting module is used to drive the rust removal unit to lift, the front and rear moving module is used to drive the rust removal unit to move front and rear, the left and right moving module is used to drive the rust removal unit to move left and right, and the rotating module is used to drive the rust removal unit to rotate 360 ​​degrees.

4. According to claim 1, a gate automatic rust removal control system is characterized by: The rust removal unit includes a rust removal module and an image recording module; The image recording module is used to record the rust removal process and the images after rust removal, and send the image signal to the control unit. The control unit receives the image signal and determines the rust removal effect according to the image signal.

5. According to claim 1, a gate automatic rust removal control system is characterized by: The control unit is equipped with a fault diagnosis module, a warning module and a remote communication module; The fault diagnosis module is used to monitor the working status of the detection unit, the mobile unit and the rust removal unit in real time, and compare and analyze whether a fault occurs; The early warning module is used to send out alarm signals so that maintenance personnel can quickly locate and solve problems; The remote communication module is used to transmit working signals and data.

6. A gate rust removal method, characterized by: The steps include: S1. Perform a comprehensive scan of the gate to obtain the gate surface morphology and generate a three-dimensional point cloud data set; S2, processing the three-dimensional point cloud data set to determine the specific location and size that need to be rusted; S3, calculating and obtaining the best rust removal path according to the specific location and size information of the rust removal; S4, driving the rust removal unit to remove rust from the gate along the rust removal path by the mobile unit; During the rust removal process, the rust-removed area is scanned to generate a new 3D point cloud data set; S5, processing the three-dimensional point cloud data set to obtain the position and size that require secondary rust removal; S6, calculating and obtaining the best rust removal path according to the position and size information in S5; S7. Repeat S4-S6 until the rust is completely removed.

7. A gate rust removal method according to claim 6, characterized in that: In S1 and S4, a laser scanner was used to fully scan the target gate; Scan 2-5 times to generate 2-5 sets of 3D point cloud data sets A n = {P1, P2, P3...P i }, A n represents the nth set of 3D point cloud data sets, P i Points representing a 3D point cloud dataset; For the 3D point cloud dataset A n Processing, compared with P i (x i ,y i ,z i ), P i (x i ,y i ,z i ) is P i The coordinate value of For the same position and the same P i (x i ,y i ,z i ), included in A 0 , for different P at the same position i (x i ,y i ,z i ), take the P with the largest number of occurrences i (x i ,y i ,z i ) is included in A 0 , or calculate the average value and include it in A 0 , and finally get the complete A 0 .

8. A gate rust removal method according to claim 6, characterized in that: In S2 and S5: S21, denoising the 3D point cloud data set, for each point P in the point cloud set i (x i ,y i ,z i ), the coordinate value calculation formula after filtering is as follows: Where N(i) represents point P i The neighborhood point set, ω(p i ,p j ) represents the point P i and P j The weight function is determined by the spatial distance and the normal vector angle; S22, using Gaussian filtering to further smooth the data; S23, calculating the local features of each point in the processed three-dimensional point cloud data set; S24, using the calculated local features as input, constructing a random forest model; S25. Obtain the location and size that require rust removal based on the random forest model.

9. A gate rust removal method according to claim 8, characterized in that: In S23, the local features include the normal vector n of each point i and curvature k i .

10. A gate rust removal method according to claim 6, characterized in that: In S3 and S6, the gate surface is divided into a plurality of small areas B1, B2, B3...B according to the location and size information of the rust removal. n ,Then the optimal rust removal path is obtained based on the ant colony optimization algorithm; In the divided area B n The following steps are included: S31, dividing the gate surface into a number of continuous square areas, with the last area being set as B; S32, starting from the first area, set to B m , and assume that the set of rust areas in the current area is {b1,b2,b3...b n }; where b n Represents the rust area in the current region; S33, when Then calculate the next area B m+1 ; when Then b x Move to next area B m+1 , until Then calculate the next area B m+1 ; Assume B m =B n ; where b x Represents the B m+1 rust area, b0 represents the maximum rust area allowed in the same area; S34, let B m+1 =B m , the set of rust areas in the region is {b1,b2,b3...b n },in Includes the rusted area included in the previous area; S35, repeat S33-S34 until B m =B; The above parts are divided into multiple small areas B1, B2, B3...B n , in S33-S35, B n Mark them one by one starting from B1.

Images