A path obstacle avoidance method and system for runner blade passage detection

Abstract

The application discloses a path obstacle avoidance method and system for runner blade flow passage detection, and belongs to the technical field of runner blade surface detection. According to the movement space in the runner flow passage, a plurality of groups of degree of freedom variable groups corresponding to the snake arm joints are generated in combination with the Monte Carlo method. The coordinate data groups of the ends of the snake arm joints in each group are determined according to the degree of freedom variable groups. The coordinate data groups are effectively checked by using the movement space, and the distance of the detection end of the snake arm robot in any two groups of effective coordinate data groups is determined. Whether the number of the effective coordinate data groups meets the requirements is determined according to the distance. If the number of the effective coordinate data groups does not meet the requirements, the number of the effective coordinate data groups is increased. Iterative calculation is performed on the effective coordinate data groups in combination with the particle swarm algorithm, and the optimal running track of the snake arm robot is obtained. The problem that the snake arm robot collides with the runner flow passage wall during detection is solved, and free movement of the snake arm robot in the narrow space of the runner flow passage is realized.

Description

Technical Field

[0001] This invention relates to the field of turbine runner blade surface inspection technology, specifically to a path obstacle avoidance method and system for runner blade flow channel inspection. Background Technology

[0002] Currently, the inspection of turbine runner blades mainly relies on manual photography or video recording to detect surface erosion and wear. Especially for runners in operation, inspectors need to reach an inspection platform beneath the runner to visually inspect the blade surface by taking photos or videos. The inspection platform is located beneath the runner, and the inspection instruments can only enter the runner flow channel through the runner's outlet, which is typically narrow, making it difficult for larger instruments to access. Therefore, auxiliary tools such as extension poles are generally used to carry smaller inspection instruments to photograph or record the blade surface; 3D scanning instruments can also be used for manual scanning of the runner blades. However, all of these methods are limited by the spatial curvature of the blades, making it difficult to achieve a comprehensive inspection of the entire blade.

[0003] A combined robot consisting of a snake-arm robot and an articulated arm robot is used for comprehensive inspection of impeller blades. The snake-arm robot enters the impeller flow channel to inspect the blade surface. However, due to the narrow impeller outlet and the fact that the impeller flow channel space is not easily described by mathematical formulas, the movement space of the snake-arm robot is limited. Without proper design of the movement space of the snake-arm robot and the articulated arm robot, the snake-arm robot is prone to collision with the wall of the impeller flow channel during the inspection process, which may cause damage to the snake-arm robot and the inspection instrument or prevent the completion of the inspection task. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention provides a path obstacle avoidance method and system for detecting the flow channel of a rotor blade. By optimizing the running trajectory of the snake-arm robot during detection, the problem of collision between the joints of the snake arm and the wall of the rotor flow channel during the detection process is solved, thus achieving safe detection of the rotor blade.

[0005] This invention is achieved through the following technical solution: A path obstacle avoidance method for detecting flow channels in a rotor blade includes the following steps: Step 1: Based on the motion space within the rotary flow channel, and combined with the Monte Carlo method, generate multiple sets of degree-of-freedom variables corresponding to the serpentine arm joints; Step 2: Determine the coordinate data set of the end joint of each snake arm according to the set of degrees of freedom variables. Use the motion space to verify the validity of the coordinate data set, discard the coordinate data set that exceeds the motion space, and keep the coordinate data set as the valid coordinate data set to obtain multiple sets of valid coordinate data sets. Step 3: Based on the valid coordinate data set obtained in Step 2, determine the distance between the detection ends of the snake-arm robot in any two valid coordinate data sets; Step 4: Determine whether the number of multiple sets of valid coordinate data meets the requirements based on the distance: If the distance between the detection ends of any two sets of valid coordinate data is greater than the set value, then repeat steps 1-3 and increase the number of valid coordinate data sets until the distance between the detection ends of any two sets of valid coordinate data is less than the set value. Then the number of multiple sets of valid coordinate data meets the requirements. Step 5: Based on the multiple sets of valid coordinate data obtained in Step 4, and combined with the particle swarm optimization algorithm, perform iterative calculations to obtain the optimal running trajectory of the snake-arm robot.

[0006] Preferably, in step 1, the motion space within the impeller flow channel is determined based on the impeller's structural parameters.

[0007] Preferably, the method for determining the motion space within the impeller channel is as follows: Based on the runner outlet opening value, runner inlet pitch, and runner inlet radius, the inlet and outlet profiles of the blades are restricted to obtain the blade profile restriction dimensions. Then, based on the blade profile restriction dimensions and the detection distance requirements between the detection instrument and the blade surface, the motion space within the runner flow channel is determined.

[0008] Preferably, the method for generating the set of degrees of freedom variables of the serpentine arm joint in step 1 is as follows: The random sampling range for the Monte Carlo method is set based on the motion space within the impeller channel; Within the random sampling range, a one-dimensional vector is generated by a random function. The joint step length of each serpentine arm joint is determined based on the one-dimensional vector. The degree of freedom variables of each serpentine arm joint are determined based on the joint step length, thus obtaining a set of degree of freedom variables.

[0009] Preferably, the degree of freedom of the serpentine arm joint is variable The expression is as follows:

[0010] in, R It is a one-dimensional vector. For the first i Minimum rotation angle of each serpentine arm joint For the first i The maximum rotation angle of the serpentine arm joint.

[0011] Preferably, the method for determining the coordinate data set in step 2 is as follows: Substituting each set of degrees of freedom variables into the kinematic equations and solving them, we obtain the coordinate data set of the end joint of the snake arm corresponding to each set of degrees of freedom variables.

[0012] Preferably, the kinematic equations are either forward kinematic equations or inverse kinematic equations.

[0013] Preferably, the method for determining the optimal running trajectory of the snake-arm robot in step 5 is as follows: The multiple sets of valid coordinate data obtained in step 4 are used as the initial particle swarm. The forward kinematics equations are used to calculate the posture and end-effector position of each snake arm when it is turned off. The particle swarm fitness function is calculated, and the posture and end-effector position of the snake arm joints are iteratively optimized to obtain the optimal running trajectory of the snake arm robot.

[0014] Preferably, during iterative optimization, the optimal running trajectory of the snake-arm robot is obtained when the particle swarm fitness function is less than the threshold or the population iteration exceeds the set number.

[0015] A path obstacle avoidance system for detecting flow channels in turbine blades, comprising, The degree-of-freedom variable module is used to generate multiple sets of degree-of-freedom variables corresponding to the serpentine arm joints based on the motion space within the rotary flow channel and in conjunction with the Monte Carlo method. The joint coordinate update module is used to determine the coordinate data set of the end of each snake arm joint based on the set of degrees of freedom variables. The coordinate data set is validated using the motion space. The coordinate data set that exceeds the motion space is discarded, and the retained coordinate data set is taken as the valid coordinate data set, thus obtaining multiple sets of valid coordinate data sets. The detection end distance determination module is used to determine the distance between the detection ends of the snake-arm robot in any two sets of valid coordinate data. The detection end distance determination module is used to determine whether the number of multiple sets of valid coordinate data meets the requirements based on the distance: when the distance between the detection ends of any two sets of valid coordinate data is greater than the set value, the operation of the degree of freedom variable module, the joint coordinate update module and the detection end distance determination module is executed in a loop, and the number of valid coordinate data sets is increased until the distance between the detection ends of any two sets of valid coordinate data is less than the set value, then the number of multiple sets of valid coordinate data meets the requirements. The trajectory optimization module is used to perform iterative calculations based on the effective coordinate data set and the particle swarm optimization algorithm to obtain the optimal running trajectory of the snake-arm robot.

[0016] Compared with the prior art, the present invention has the following beneficial technical effects: This invention provides a path obstacle avoidance method for detecting impeller blades in a flow channel. Based on the motion space within the impeller flow channel, and combined with the Monte Carlo method, multiple sets of degrees of freedom variables for the serpentine arm joints are generated. The coordinate data sets for each set of serpentine arm joints are determined based on these sets of variables, and the validity of the coordinate data sets is verified. The valid coordinate data sets are used as the initial particle swarm, and the particle swarm algorithm is used to iteratively calculate the initial particle swarm to obtain the optimal running trajectory of the serpentine arm robot. The serpentine arm robot uses the optimal running trajectory to detect the impeller blades in the flow channel. This method solves the problem of collisions between the serpentine arm joints and the impeller flow channel wall during the detection process, enabling the serpentine arm robot to move freely within the confined space of the impeller flow channel. This method allows the serpentine arm robot to move along the blade surface even without knowing the three-dimensional geometric dimensions of the blade. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the combined robot and its rotating wheel of the present invention; Figure 2 This is a schematic diagram of the structure of the snake-arm robot of the present invention; Figure 3 This is a flowchart of the path obstacle avoidance method of the present invention; Figure 4 This is a flowchart illustrating the calculation of the optimal path of the serpentine arm joint in motion space according to the present invention.

[0018] In the image: 1. Snake-arm robot; 2. Articulated arm robot; 3. Rotary wheel; 4. Snake-arm joint; 5. Base. Detailed Implementation

[0019] The present invention will now be described in further detail with reference to the accompanying drawings. These descriptions are intended to explain the invention and not to limit it.

[0020] See Figure 3 A path obstacle avoidance method for detecting impeller blade flow channels includes the following steps: Step 1: Build a combined robot for blade surface inspection.

[0021] S1.1 Select an articulated arm robot based on the weight of the testing instrument and the snake-arm robot, and use the teaching method to make the end of the articulated arm robot enter the flow channel of the impeller along the outlet of the impeller and move from top to bottom, thereby determining the movement space of the end of the articulated arm robot in the flow channel of the impeller.

[0022] S1.2 Construct a connecting base based on the structural parameters of the snake-arm robot. The end of the articulated arm robot is connected to the base 5 of the snake-arm robot through the connecting base to form a combined robot.

[0023] See Figure 1 and 2The combined robot includes a snake-arm robot 1 and an articulated arm robot 2. One end of the connecting base is connected to the end of the articulated arm robot 2, and the other end of the connecting base is connected to the base 5 of the snake-arm robot. The snake-arm robot includes multiple snake-arm joints 4 connected in sequence. The detection instrument is installed at the detection end of the snake-arm robot, that is, the front end of the first snake-arm joint, and the end of the last snake-arm joint is connected to the base 5.

[0024] Step 2: Determine the impeller flow channel, as well as the impeller outlet and inlet, based on the blade structure.

[0025] During the inspection process, the snake-arm robot 1 inspects the water outlet of the impeller 3 from top to bottom. During the blade inspection, the snake-arm robot 1 uses a non-contact method to inspect the blade, that is, the snake-arm robot 1 cannot collide or contact with any wall of the impeller channel.

[0026] Step 3: Determine the movement space of the testing instrument within the rotor flow channel based on the structural parameters of the rotor.

[0027] Based on the required detection distance between the instrument and the blade surface, the movement space of the instrument within the runner channel is determined. The space including the blade is obtained by subtracting the space occupied by the upper crown and lower ring from the overall space of the runner. Since this space is difficult to describe mathematically, the runner's outlet opening value, inlet pitch, and inlet radius are used to constrain the inlet and outlet profiles of the blades, resulting in blade profile constraint dimensions. Based on these constraint dimensions and the required detection distance on the blade surface, the movement space of the instrument within the runner channel is calculated.

[0028] Step 4: Using a combination of Monte Carlo and particle swarm optimization algorithms, determine the optimal trajectory of the snake-arm robot in the motion space. (See [reference needed]) Figure 4 This includes the following processes: S4.1. Based on the motion space within the impeller channel, set the random sampling range for the Monte Carlo method; S4.2. Within the random sampling range, a one-dimensional vector is generated by a random function. The joint step length of the serpentine arm joint is determined based on the one-dimensional vector. The degree of freedom variables of each serpentine arm joint are determined based on the joint step length, resulting in a set of free variables, as follows: Within the random sampling range, determined by a random function RAND(j)(j=1,2,3…,N) Generate includes N A one-dimensional vector with elements R One-dimensional vector R It takes values ​​between 0 and 1, based on a one-dimensional vector. R Determine the joint step length of each serpentine arm joint. This leads to a set of degrees of freedom variables for each serpentine arm joint, the first... iThe degrees of freedom variables of the serpentine arm joint are as follows:

[0029] in, R It is a one-dimensional vector. For the first i Minimum rotation angle of each serpentine arm joint For the first i The maximum rotation angle of the serpentine arm joint.

[0030] S4.3 Substitute the set of degrees of freedom variables into the kinematic equations to solve for the coordinate data set of the ends of each serpentine arm joint.

[0031] The kinematic equations can be either forward kinematic equations or inverse kinematic equations, preferably forward kinematic equations.

[0032] S4.4. The validity of the coordinate data set is verified by using the motion space within the rotary flow channel: if any coordinate data in the coordinate data set exceeds the motion space, the coordinate data set is discarded; otherwise, the coordinate data set is retained as a valid coordinate data set.

[0033] S4.5 Repeat steps S4.2-S4.4 to obtain a set number of valid coordinate data sets, that is, to obtain the different spatial positions of the serpentine arm joint in the motion space, forming a certain number of reachable spatial points of the serpentine arm joint in the motion space.

[0034] S4.6. Based on the set number of valid coordinate data sets obtained in step S4.5, determine the distance between the detection ends of the snake-arm robot in any two sets of valid coordinate data sets, and determine whether the number of valid coordinate data sets meets the requirements based on the distance: if the distance between the detection ends of the snake-arm robot in any two sets of valid coordinate data sets is greater than the set value, repeat step S4.5 to increase the number of valid coordinate data sets until the distance between the detection ends of the snake-arm robot in any two sets of valid coordinate data sets is less than the set value, then the number of valid coordinate data sets meets the requirements, and multiple sets of valid coordinate data sets are obtained.

[0035] In this embodiment, if the distance between the detection ends of any two sets of valid coordinate data is less than 2mm, it is determined that the number of reachable spatial points of the snake arm joint in the motion space has reached a sufficient number, that is, the number of valid coordinate data sets meets the requirements; if the distance is greater than 2mm, step S4.5 is executed to supplement the number of reachable spatial points of the snake arm joint in the motion space until the number of valid coordinate data sets meets the requirements.

[0036] S4.7. Using the multiple sets of valid coordinate data obtained in step S4.6 as the initial particle swarm, and using the forward kinematics equations to calculate the attitude and end-effector position of each serpentine arm joint, and calculate the particle swarm fitness function. The posture and end position of the serpentine arm joint are iteratively optimized. S4.8 When the set iteration conditions are met, the optimal posture and end position of the snake arm joint are obtained.

[0037] Specifically, the iteration condition is that the particle swarm fitness function is less than a threshold, or the population iteration exceeds a set number. In this embodiment, when the particle swarm fitness function... Alternatively, if the population iteration exceeds a set number, update the population to obtain the optimal posture and end-effector position of the snake arm joint, i.e., the optimal running trajectory of the snake arm robot in the motion space, and perform blade detection based on the optimal running trajectory of the snake arm robot.

[0038] This invention also provides a path obstacle avoidance system for impeller blade flow channel detection, comprising, The degree-of-freedom variable module is used to generate multiple sets of degree-of-freedom variables corresponding to the serpentine arm joints based on the motion space within the rotary flow channel and in conjunction with the Monte Carlo method. The joint coordinate update module is used to determine the coordinate data set of the end of each snake arm joint based on the set of degrees of freedom variables. The coordinate data set is validated using the motion space. The coordinate data set that exceeds the motion space is discarded, and the retained coordinate data set is taken as the valid coordinate data set, thus obtaining multiple sets of valid coordinate data sets. The detection end distance determination module is used to determine the distance between the detection ends of the snake-arm robot in any two sets of valid coordinate data. The detection end distance determination module is used to determine whether the number of multiple sets of valid coordinate data meets the requirements based on the distance: when the distance between the detection ends of any two sets of valid coordinate data is greater than the set value, the operation of the degree of freedom variable module, the joint coordinate update module and the detection end distance determination module is executed in a loop, and the number of valid coordinate data sets is increased until the distance between the detection ends of any two sets of valid coordinate data is less than the set value, then the number of multiple sets of valid coordinate data meets the requirements. The trajectory optimization module is used to perform iterative calculations based on the effective coordinate data set and the particle swarm optimization algorithm to obtain the optimal running trajectory of the snake-arm robot.

[0039] This invention provides a path obstacle avoidance method for impeller blade flow channel detection. During blade detection, an articulated arm robot sends a snake-arm robot from the impeller outlet into the impeller flow channel. A combination of Monte Carlo method and particle swarm optimization algorithm is used to calculate the optimal motion trajectory of the snake-arm robot in the motion space, enabling the snake-arm robot to perform obstacle avoidance detection on the blades within the impeller flow channel. This solves the problem of obstacle avoidance for the snake-arm joint movement in the narrow space of the impeller flow channel, and enables the detection instrument to safely detect the blades.

[0040] The above content is only for illustrating the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of protection of the claims of this invention.

Claims

1. A path obstacle avoidance method for detecting flow channels in impeller blades, characterized in that, Includes the following steps: Step 1: Based on the motion space within the rotary flow channel, and combined with the Monte Carlo method, generate multiple sets of degree-of-freedom variables corresponding to the serpentine arm joints; Step 2: Determine the coordinate data set of the end joint of each snake arm according to the set of degrees of freedom variables. Use the motion space to verify the validity of the coordinate data set, discard the coordinate data set that exceeds the motion space, and keep the coordinate data set as the valid coordinate data set to obtain multiple sets of valid coordinate data sets. Step 3: Based on the valid coordinate data set obtained in Step 2, determine the distance between the detection ends of the snake-arm robot in any two valid coordinate data sets; Step 4: Determine whether the number of multiple sets of valid coordinate data meets the requirements based on the distance: If the distance between the detection ends of any two sets of valid coordinate data is greater than the set value, then repeat steps 1-3 and increase the number of valid coordinate data sets until the distance between the detection ends of any two sets of valid coordinate data is less than the set value. Then the number of multiple sets of valid coordinate data meets the requirements. Step 5: Based on the multiple sets of valid coordinate data obtained in Step 4, and combined with the particle swarm optimization algorithm, perform iterative calculations to obtain the optimal running trajectory of the snake-arm robot.

2. The path obstacle avoidance method for impeller blade flow channel detection according to claim 1, characterized in that, In step 1, the motion space within the impeller flow channel is determined based on the impeller's structural parameters.

3. The path obstacle avoidance method for impeller blade flow channel detection according to claim 2, characterized in that, The method for determining the motion space within the impeller channel is as follows: Based on the runner outlet opening value, runner inlet pitch, and runner inlet radius, the inlet and outlet profiles of the blades are restricted to obtain the blade profile restriction dimensions. Then, based on the blade profile restriction dimensions and the detection distance requirements between the detection instrument and the blade surface, the motion space within the runner flow channel is determined.

4. The path obstacle avoidance method for impeller blade flow channel detection according to claim 1, characterized in that, The method for generating the set of degrees of freedom variables for the serpentine arm joint described in step 1 is as follows: The random sampling range for the Monte Carlo method is set based on the motion space within the impeller channel; Within the random sampling range, a one-dimensional vector is generated by a random function. The joint step length of each serpentine arm joint is determined based on the one-dimensional vector. The degree of freedom variables of each serpentine arm joint are determined based on the joint step length, thus obtaining a set of degree of freedom variables.

5. The path obstacle avoidance method for impeller blade flow channel detection according to claim 4, characterized in that, The degree of freedom variables of the snake-arm joint The expression is as follows: in, R It is a one-dimensional vector. For the first i Minimum rotation angle of each serpentine arm joint For the first i The maximum rotation angle of the serpentine arm joint.

6. The path obstacle avoidance method for impeller blade flow channel detection according to claim 1, characterized in that, The method for determining the coordinate data set in step 2 is as follows: Substituting each set of degrees of freedom variables into the kinematic equations and solving them, we obtain the coordinate data set of the end joint of the snake arm corresponding to each set of degrees of freedom variables.

7. The path obstacle avoidance method for impeller blade flow channel detection according to claim 6, characterized in that, The kinematic equations are either forward kinematic equations or inverse kinematic equations.

8. The path obstacle avoidance method for impeller blade flow channel detection according to claim 1, characterized in that, The method for determining the optimal trajectory of the snake-arm robot in step 5 is as follows: The multiple sets of valid coordinate data obtained in step 4 are used as the initial particle swarm. The forward kinematics equations are used to calculate the posture and end-effector position of each snake arm when it is turned off. The particle swarm fitness function is calculated, and the posture and end-effector position of the snake arm joints are iteratively optimized to obtain the optimal running trajectory of the snake arm robot.

9. The path obstacle avoidance method for impeller blade flow channel detection according to claim 8, characterized in that, During iterative optimization, the optimal running trajectory of the snake-arm robot is obtained when the particle swarm fitness function is less than the threshold or the population iteration exceeds the set number.

10. A path obstacle avoidance system for detecting flow channels in impeller blades, characterized in that, include, The degree-of-freedom variable module is used to generate multiple sets of degree-of-freedom variables corresponding to the serpentine arm joints based on the motion space within the rotary flow channel and in conjunction with the Monte Carlo method. The joint coordinate update module is used to determine the coordinate data set of the end of each snake arm joint based on the set of degrees of freedom variables. The coordinate data set is validated using the motion space. The coordinate data set that exceeds the motion space is discarded, and the retained coordinate data set is taken as the valid coordinate data set, thus obtaining multiple sets of valid coordinate data sets. The detection end distance determination module is used to determine the distance between the detection ends of the snake-arm robot in any two sets of valid coordinate data. The detection end distance determination module is used to determine whether the number of multiple sets of valid coordinate data meets the requirements based on the distance: when the distance between the detection ends of any two sets of valid coordinate data is greater than the set value, the operation of the degree of freedom variable module, the joint coordinate update module and the detection end distance determination module is executed in a loop, and the number of valid coordinate data sets is increased until the distance between the detection ends of any two sets of valid coordinate data is less than the set value, then the number of multiple sets of valid coordinate data meets the requirements. The trajectory optimization module is used to perform iterative calculations based on the effective coordinate data set and the particle swarm optimization algorithm to obtain the optimal running trajectory of the snake-arm robot.

Images