Interference detection device

Abstract

The device models the shapes of the robot and surrounding obstacles with less data and improves the accuracy of interference detection. The interference detection device performs interference detection between the robot and surrounding obstacles. The device includes: a cuboid assembly conversion unit that converts the robot and surrounding obstacles into three-dimensional models of cuboid assemblies; and an interference determination unit that determines whether there is interference between the three-dimensional models of the robot and the surrounding obstacles by simulating the movements of the aforementioned three-dimensional models based on motion programs.

Description

Technical Field

[0001] Invention Field

[0002] This invention relates to an interference detection device. Prior Technology

[0003] Background of the Invention

[0004] To explore the robot's movement path, interference checks were performed on the robot in a virtual environment with surrounding machines or obstacles such as safety fences. In this regard, a known technique uses triangular mesh data of a three-dimensional model comprising the robot and surrounding obstacles to perform interference detection. See, for example, Patent Document 1. Furthermore, a known technique utilizes a polyhedron equivalent to a simplified convex hull to perform high-speed processing involving computational interference with three-dimensional shapes as objects. These three-dimensional shapes are composed of free-form surfaces or numerous free-form surfaces expressed by high-order surface equations. See, for example, Patent Document 2. Furthermore, a known technique uses a voxel model to detect interference between a robot and surrounding obstacles. This voxel model has a plurality of voxels, or voxelized spheres and cylinders. See, for example, Patent Document 3. Prior technology documents Patent documents

[0005] Patent Document 1: Japanese Patent Application Publication No. 2020-179441 Patent Document 2: Japanese Patent Application Publication No. 2002-342395 Patent Document 3: Japanese Patent Application Publication No. 2012-232408 Summary of the Invention

[0006] Invention Summary The problem the invention aims to solve

[0007] In Patent Document 1, when performing interference checks on a 3D model using triangular meshes, although correct judgment results can be obtained, it has the problem of very high computational cost (especially when the number of triangular meshes is large). Furthermore, in Patent Document 2, depending on the shape of the surrounding obstacles, the space occupied by the convex hull is sometimes larger than the actual shape, which may reduce the accuracy of the interference inspection. Furthermore, in Patent Document 3, when the sphere or cylinder contained in the stereo pixel is larger than the actual robot or surrounding obstacles, there will be a problem that the robot is always in an interference state when it approaches the surrounding obstacles.

[0008] Therefore, it is desirable to model the shape of the robot and surrounding obstacles with less data and improve the accuracy of interference detection. The means to solve the problem

[0009] The interference detection device disclosed herein performs interference detection between a robot and surrounding obstacles. The interference detection device includes: a cuboid set conversion unit that converts the robot and the surrounding obstacles into three-dimensional models of cuboid sets; and an interference determination unit that determines whether there is interference between the three-dimensional models of the robot and the surrounding obstacles by simulating the movements of the three-dimensional models of the robot and the surrounding obstacles based on the motion program. Invention Effects

[0010] Based on a single-state sample, the shape of the robot and surrounding obstacles can be modeled with less data, and the accuracy of interference detection can be improved. Simple Explanation of the Diagram

[0011] Figure 1 is a functional block diagram showing a functional configuration example of an interference detection device in one embodiment. Figure 2 is an example of 3D CAD data showing surrounding obstacles. Figure 3 is a diagram showing an example of a cut surface in the case of a robot. Figure 4 is an example of a three-dimensional model of a set of cuboids representing the surrounding obstacles shown in Figure 2. Figure 5 is an example of a three-dimensional model of the convex hull of the surrounding obstacle shown in Figure 2. Figure 6 is an example of a set of cuboids of surrounding obstacles shown in Figure 4 with a set margin. Figure 7 is a flowchart illustrating the interference detection process of the interference detection device. Figure 8 is a flowchart illustrating the three-dimensional model conversion process of the interference detection device. Implementation

[0012] Forms used to implement inventions

[0013] <I. Implementation Form> Figure 1 is a functional block diagram showing a functional configuration example of an interference detection device in one embodiment. As shown in Figure 1, the interference detection device 1 is a conventional computer, including a control unit 10, an input unit 11, a display unit 12, and a memory unit 13. The control unit 10 includes a cuboid set conversion unit 101 and an analog execution unit 102. The cuboid set conversion unit 101 includes a simplified range setting unit 111, a point cloud data conversion unit 112, a gene point cloud data segmentation unit 113, a minimum volume search unit 114, and a margin setting unit 115. The analog execution unit 102 includes a coordinate system conversion unit 121 and an interference determination unit 122. Furthermore, the interference detection device 1 can also be connected to the robot control device (not shown) that controls the robot's (not shown) movements via a network such as a LAN (Local Area Network) or the Internet. Alternatively, the interference detection device 1 can also be directly connected to the robot control device (not shown) via a connection interface (not shown).

[0014] <Input Section 11> The input unit 11 is, for example, a keyboard or a touch keyboard disposed on the display unit 12 described later. As described later, the input unit 11 receives from users such as operators the number of cuboids when a three-dimensional model of a robot (not shown), or a surrounding machine (not shown), or a safety fence or other surrounding obstacle is set as a collection of cuboids.

[0015] <Display Unit 12> The display unit 12 is, for example, a liquid crystal display, which displays the determination results of the interference detection device 1.

[0016] <Memory Section 13> The memory unit 13 is an SSD (Solid State Drive) or HDD (Hard Disk Drive), etc., and can also store various motion programs that cause the robot or peripheral machine (not shown) to move. In addition, the memory unit 13 includes a motion record (log) memory unit 131 and a shape data memory unit 132. As described later, the motion recording memory unit 131 stores three-dimensional coordinate values ​​and time as motion records. The aforementioned three-dimensional coordinate values ​​are the coordinate values ​​representing the position and posture of the robot and the surrounding obstacles' respective sets of cuboids, which are transformed by the cuboid set conversion unit 101 based on shape data, when the interference determination unit 122 performs motion program simulation. The aforementioned shape data are 3D CAD data of the robot (not shown) and its surrounding machines or safety fences, etc., which are represented in three dimensions. The shape data memory unit 132 stores shape data such as 3D CAD data of the robot (not shown) and 3D CAD data of surrounding obstacles (not shown).

[0017] <Control Unit 10> The control unit 10 is a control unit known to those skilled in the art, and includes CPU (Central Processing Unit), ROM (Read Only Memory), RAM (Random Access Memory), CMOS (Complementary Metal-Oxide-Semiconductor) memory, etc., and these components are configured to communicate with each other via a bus. The CPU is a processor that controls the entire interference detection device 1. The CPU reads the system program and application program stored in ROM via a bus and controls the entire interference detection device 1 according to the system program and application program. Therefore, as shown in FIG1, the control unit 10 is configured to implement the functions of a cuboid set conversion unit 101 and an analog execution unit 102. Furthermore, the cuboid set conversion unit 101 is configured to implement the functions of a simplified range setting unit 111, a point cloud data conversion unit 112, a gene point cloud data segmentation unit 113, a minimum total volume search unit 114, and a margin setting unit 115. Furthermore, the analog execution unit 102 is configured to implement the functions of a coordinate system conversion unit 121 and an interference determination unit 122. Various data, such as temporary calculation data or display data, are stored in RAM. Furthermore, the CMOS memory is configured as non-volatile memory, which is backed up by a battery (not shown), and can maintain its memory state even when the power supply to the interference detection device 1 is disconnected.

[0018] <Includes cuboid set transformation part 101> For example, as a preprocessing step of the main processing of the simulation execution unit 102 described later, the cuboid set conversion unit 101 converts the robot and surrounding obstacles into a three-dimensional model of a set of cuboids from the shape data (3D CAD data) of the robot and surrounding obstacles (not shown) stored in the shape data memory unit 132 of the memory unit 13. The following description of the operation of the cuboid set conversion unit 101 will focus on the functions of the simplified range setting unit 111, point cloud data conversion unit 112, gene point cloud data segmentation unit 113, minimum total volume search unit 114, and margin setting unit 115 that constitute the cuboid set conversion unit 101.

[0019] <Simplified Range Setting Section 111> The simplified range setting unit 111 displays 3D CAD data as shape data of surrounding obstacles on the display unit 12, and sets the range of the shape data of the surrounding obstacles to be used for interference detection based on the input operation of the input unit 11 performed by the user. Figure 2 is an example of the shape data of a surrounding obstacle. Furthermore, in Figure 2, the shape data (3D CAD data) of the surrounding obstacle shows the shape of the surrounding obstacle in the XY plane as viewed from the positive Z-axis direction. As shown in Figure 2, the simplified range setting unit 111 accepts the specification of the range shown by the dashed line in the shape data (3D CAD data) of the surrounding obstacles based on the input operation of the input unit 11 performed by the user. The simplified range setting unit 111 sets the accepted range as the specified range. In this way, interference detection device 1 only performs interference detection on the actual range of the robot's (not shown) movements, which simplifies interference detection.

[0020] <Point Cloud Data Conversion Section 112> The point cloud data conversion unit 112 uses, for example, a conventional conversion method to convert to point cloud data. Based on the 3D CAD data of the robot and the surrounding obstacles (not shown) stored in the shape data memory unit 132, it converts them into point cloud data of the robot and point cloud data of the surrounding obstacles within a specified range set by the simplified range setting unit 111.

[0021] <Gene Point Cloud Data Segmentation Section 113> For example, the gene point cloud data segmentation unit 113 calculates an evaluation value for each combination of multiple cut surfaces. The evaluation value of each combination of multiple cut surfaces regenerated based on a known gene algorithm is repeated a maximum number of times. The aforementioned multiple cut surfaces are cut surfaces of the number of cuboids input by the user, which are the point cloud data of the robot and surrounding obstacles converted by the point cloud data conversion unit 112. Specifically, for example, when the user inputs the number of cuboids n through the input unit 11, the gene point cloud data segmentation unit 113 randomly generates k combinations of n (n-1) cut surfaces (n and k are integers of 2 or more) that divide the point cloud data of the robot (or surrounding obstacles) not shown into n. Figure 3 is an example of a cut surface in the case of a robot. In Figure 3, the robot is represented by solid lines, and the three-dimensional model of the robot's set of cuboids is represented by dashed lines. Also in Figure 3, the three-dimensional model of the robot's set of cuboids shows the case where the number of cuboids n input by the user is 8, and the 7 cut surfaces are represented by thick solid lines. The gene point cloud data segmentation unit 113 calculates the volume of the cuboid containing each of the n point cloud data, and calculates the total volume of the set of cuboids of the robot (or surrounding obstacles) as the evaluation value of each combination of (n-1) cut surfaces. The aforementioned n point cloud data are segmented by (n-1) cut surfaces of each of the k combinations. The gene point cloud data segmentation unit 113 repeats the recalculation of the evaluation values ​​of each of the (n-1) cut surfaces of the k combinations regenerated by gene manipulation based on the gene algorithm, with a preset maximum repetition number. Furthermore, the number of cut surfaces into which the point cloud data is divided is not limited to (n-1), and can also be less than (n-1).

[0022] <Minimum Total Volume Search Unit 114> The minimum total volume search unit 114 searches for the (n-1) cut surfaces with the highest evaluation values ​​among the evaluation values ​​of each of the k combinations, that is, the set of cuboids with the smallest total volume of the 3D model of the robot (or surrounding obstacles) not shown, and selects them as the optimal cut surfaces. As shown in Figure 3, the minimum total volume search unit 114 selects the set of cuboids divided by the optimal (n-1) cut surfaces as the 3D model of the robot (or surrounding obstacles) not shown.

[0023] Figure 4 is an example of a three-dimensional model of a set of cuboids representing the surrounding obstacles shown in Figure 2. As shown in Figure 4, the three-dimensional model of the surrounding obstacles shown in Figure 2 is a collection of two cuboids 200(1) and 200(2) shown by thick dashed lines within the specified range set by the simplified range setting unit 111. The cuboids 200(1) and 200(2) are cut off from each other. Figure 5 is an example of a three-dimensional model of the convex hull of the surrounding obstacle shown in Figure 2. As shown in Figure 5, the 3D model of the convex hull of the surrounding obstacles, indicated by the thick dashed line, is larger than the actual shape of the surrounding obstacles, as mentioned above. In contrast, the 3D model of the set of cuboids of the surrounding obstacles in Figure 4 shows a size almost equal to the actual shape of the surrounding obstacles. In this way, the interference detection device 1 can perform interference detection with good accuracy.

[0024] <Balance Setting Section 115> The margin setting unit 115 sets the margin for a three-dimensional model of a collection of cuboids of robots and surrounding obstacles (not shown). Specifically, the margin setting unit 115 sets a pre-set margin d in each direction of the X-axis, Y-axis, and Z-axis for each of the cuboids of the robot and surrounding obstacles (not shown), as shown in FIG6.

[0025] By performing the preprocessing described above by the cuboid assembly conversion unit 101, the use of the triangular mesh of the original 3D CAD data of the robot and surrounding obstacles can be avoided, thereby reducing the computational cost of interference inspection.

[0026] <Simulation Execution Department 102> The simulation execution unit 102 performs simulation as the main processing based on the motion program, and checks whether there is any interference between the robot's three-dimensional model and the three-dimensional models of the surrounding obstacles. The aforementioned simulation is to make the three-dimensional models of the robot and the surrounding obstacles, which are converted by the cuboid set conversion unit 101 as preprocessing, move. The following explanation of the operation of the analog execution unit 102 will focus on the functions of the coordinate system conversion unit 121 and the interference determination unit 122 that constitute the analog execution unit 102.

[0027] <Coordinate System Conversion Unit 121> The coordinate system conversion unit 121 updates the position and pose of a three-dimensional model of a set of cuboids of the robot and surrounding obstacles (not shown) in response to the movements of the robot and surrounding obstacles, for example, when the simulation of the motion program has been executed by the simulation execution unit 102.

[0028] <Interference Detection Unit 122> Interference determination unit 122 determines whether there is interference between the robot's three-dimensional model and the three-dimensional models of surrounding obstacles by means of motion simulation performed by simulation execution unit 102 based on motion program. The aforementioned motion simulation is a motion simulation of the three-dimensional model of the robot and the cuboid collection of surrounding obstacles (not shown). Specifically, the interference determination unit 122, for example, in the simulation of the motion program performed by the simulation execution unit 102, stores the three-dimensional coordinate values ​​and time as motion records in the motion record memory unit 131. The aforementioned three-dimensional coordinate values ​​represent the coordinate values ​​of the position and posture of the set of cuboids of the robot and the surrounding obstacles. The interference determination unit 122 detects cuboids that interfere with each other in space at the same time during the motion record of the robot and the surrounding obstacles, and determines whether there is interference. The interference determination unit 122 can also display the determination result on the display unit 12. In this way, users can grasp the positional relationship between the robot and surrounding obstacles.

[0029] Interference detection and processing of interference detection device 1 Next, referring to Figure 7, the interference detection process of interference detection device 1 will be explained. Figure 7 is a flowchart illustrating the interference detection process of the interference detection device 1. The process shown here is executed whenever an interference detection instruction is received from a user via the input unit 11 from a robot (not shown).

[0030] In step S1, when the user receives a conversion instruction for a 3D model of a set of cuboids of the robot through the input unit 11, the cuboid set conversion unit 101 performs a 3D model conversion process. This 3D model conversion process converts the robot's shape data (3D CAD data) stored in the shape data memory unit 132 into a 3D model of the robot's set of cuboids. Furthermore, the detailed process of the 3D model conversion process will be described later.

[0031] In step S2, the cuboid set conversion unit 101 performs the three-dimensional model conversion process in the same way as the robot in step S1. The aforementioned three-dimensional model conversion process converts the shape data (3D CAD data) of the surrounding obstacles stored in the shape data memory unit 132 into a three-dimensional model of the cuboid set of the surrounding obstacles.

[0032] In step S3, the simulation execution unit 102 (interference determination unit 122) performs simulations of the three-dimensional model of the robot's set of cuboids converted in step S1 and the three-dimensional model of the surrounding obstacles' set of cuboids converted in step S2, based on the motion program, to determine whether there is any interference between the robot's three-dimensional model and the three-dimensional model of the surrounding obstacles.

[0033] In step S4, the simulation execution unit 102 (interference determination unit 122) displays the determination result on the display unit 12.

[0034] <Three-dimensional model conversion processing of interference detection device 1> Figure 8 is a flowchart illustrating the detailed processing steps of the 3D model conversion process shown in steps S1 and S2 of Figure 7.

[0035] In step S21, the cuboid set conversion unit 101 reads the shape data (3D CAD data) of the accepted robot (or surrounding obstacles) from the shape data memory unit 132 through the input unit 11, based on the conversion instruction of the three-dimensional model of the cuboid set of the robot (or surrounding obstacles) accepted by the user.

[0036] In step S22, the cuboid set conversion unit 101 determines whether the shape data read in step S21 is a surrounding obstacle. If the shape data is a surrounding obstacle, the process proceeds to step S23. Alternatively, if the shape data is not a surrounding obstacle, that is, if it is a robot, the process proceeds to step S24.

[0037] In step S23, the simplified range setting unit 111 displays, for example, the shape data (3D CAD data) of the surrounding obstacles on the display unit 12, and accepts and sets the specified range of the shape data of the surrounding obstacles to be used for interference detection based on the input operation of the input unit 11 performed by the user.

[0038] In step S24, the point cloud data conversion unit 112 converts the shape data (3D CAD data) of the robot (or the point cloud data of the surrounding obstacles within the specified range set in step S3) into the point cloud data of the robot.

[0039] In step S25, the gene point cloud data segmentation unit 113 randomly generates k combinations of (n-1) cut surfaces. The aforementioned (n-1) cut surfaces are the point cloud data of the robot (or surrounding obstacles) converted in step S24, which are segmented into cut surfaces of cuboid number n input by the user through the input unit 11. The gene point cloud data segmentation unit 113 calculates the total volume of the set of cuboids of the robot (or surrounding obstacles) segmented by the (n-1) cut surfaces of each of the k combinations, and uses it as the evaluation value of each combination of (n-1) cut surfaces.

[0040] In step S26, the gene point cloud data segmentation unit 113 recalculates the evaluation values ​​of the (n-1) cut surfaces of each of the k combinations regenerated by gene manipulation based on the gene algorithm.

[0041] In step S27, the gene point cloud data segmentation unit 113 determines whether the maximum number of repetitions m has been repeated to obtain m candidate point cloud sets. If m candidate point cloud sets have been obtained, the process proceeds to step S28. Otherwise, if no m candidate point cloud sets have been obtained, the process returns to step S26.

[0042] In step S28, the minimum total volume search unit 114 searches for the (n-1) cut surfaces of each of the k combinations, and selects the (n-1) cut surfaces with the highest evaluation values, i.e., the set of cuboids with the smallest total volume of the 3D model of the robot (or surrounding obstacle) not shown, as the optimal cut surfaces. The minimum total volume search unit 114 selects the set of cuboids divided by the optimal (n-1) cut surfaces as the 3D model of the robot (or surrounding obstacle) not shown.

[0043] In step S29, the margin setting unit 115 sets a margin for the three-dimensional model of the set of cuboids of the robot (or surrounding obstacles) selected in step S28.

[0044] Based on the above, an interference detection device 1 in one embodiment converts the robot and surrounding obstacles into a three-dimensional model of a set of cuboids based on the shape data (3D CAD data) of the robot and surrounding obstacles. In this way, the interference detection device 1 can model the shape of the robot and surrounding obstacles with less data and improve the accuracy of interference detection. Furthermore, by using a three-dimensional model of a set of cuboids, the interference detection device 1 can reduce the computational cost of interference detection and achieve interference detection at high speed.

[0045] The above describes one embodiment, but the interference inspection device 1 is not limited to the above embodiment, and also includes variations and improvements within the scope of achieving the purpose.

[0046] <Variation Example 1> In the above embodiment, although it is assumed that the interference detection device 1 is a different device from the robot control device (not shown), it is not limited thereto. For example, the interference detection device 1 may also be included in the robot control device (not shown).

[0047] <Variation Example 2> For example, in the above embodiment, the cuboid set conversion unit 101 converts the shape data (3D CAD data) of the robot (or surrounding obstacles) stored in the shape data memory unit 132 into point cloud data, and based on a genetic algorithm, divides the point cloud data into n point cloud data to convert it into a three-dimensional model of a set of cuboids of the robot (or surrounding obstacles), but it is not limited to this. For example, the cuboid set conversion unit 101 can also convert the robot (or surrounding obstacles) into a three-dimensional model of a set of cuboids according to the number of cuboids, the position of each cuboid, the size of each cuboid, etc., specified by the user based on the shape data (3D CAD data).

[0048] Furthermore, the various functions included in an embodiment of the interference detection device 1 can be implemented separately by hardware, software, or a combination thereof. Here, implementation by software means implementation by reading and executing a program by a computer.

[0049] Programs can be stored and supplied to a computer using various types of non-transitory computer-readable media. Non-transitory computer-readable media includes various types of tangible storage media. Examples of non-transitory computer-readable media include magnetic recording media (such as floppy disks, magnetic tapes, and hard disk drives), optical magnetic recording media (such as magneto-optical discs), CD-ROM (Read Only Memory), CD-R, CD-R / W, semiconductor memory (such as masked read-only memory, PROM (Programmable ROM), EPROM (Erasable PROM), flash memory, and RAM). Furthermore, programs can also be supplied to a computer using various types of transient computer-readable media. Examples of transient computer-readable media include electrical signals, optical signals, and electromagnetic waves. Temporary computer-readable media can deliver programs to a computer via wired communication paths such as power lines and fiber optics, or via wireless communication paths.

[0050] Furthermore, the steps describing the process of recording on the recording medium certainly include processing performed in sequence and in time, but not necessarily in time sequence, and also include processing performed in parallel or individually.

[0051] In other words, the interference detection device disclosed herein can be implemented in various forms with the following configuration.

[0052] (1) The interference detection device 1 disclosed herein performs interference detection between the robot and surrounding obstacles. The interference detection device 1 includes: a cuboid set conversion unit 101, which converts the robot and surrounding obstacles into three-dimensional models of cuboid sets respectively; and an interference determination unit 122, which determines whether there is interference between the three-dimensional model of the robot and the three-dimensional model of the surrounding obstacles by simulating the movements of the three-dimensional models of the robot and the surrounding obstacles based on the motion program. Based on this interference detection device 1, the shape of the robot and surrounding obstacles can be modeled with less data, and the accuracy of interference detection can be improved.

[0053] (2) The interference detection device 1 described in (1) may also include an input unit 11, which inputs the number of cuboids of the set of cuboids. The cuboid set conversion unit 101 may also include: a point cloud data conversion unit 112, which converts the shape data of the robot and the surrounding obstacles into point cloud data of the robot and the surrounding obstacles respectively; a gene point cloud data segmentation unit 113, which calculates an evaluation value for each combination of a plurality of cut surfaces, and calculates an evaluation value for each combination of a plurality of cut surfaces regenerated based on a gene algorithm, wherein the plurality of cut surfaces are cut surfaces that divide the point cloud data of the robot and the surrounding obstacles into the number of cuboids input to the input unit 11; and a minimum total volume search unit 114, which searches for the plurality of cut surfaces with the highest evaluation value among the evaluation values ​​of each combination of a plurality of cut surfaces of the robot and the surrounding obstacles. In this way, the interference detection device 1 can be modeled as a set of cuboids that best fit the robot and the surrounding obstacles.

[0054] (3) The interference detection device 1 described in (1) or (2) may also include a cuboid set conversion unit 101, which sets the range to be used in the simulation in the shape data of the surrounding obstacles; and a margin setting unit 115, which sets the margin for the three-dimensional model of the cuboid set of the robot and the surrounding obstacles. In this way, the interference detection device 1 only models the parts of the surrounding obstacles that need to be interfered with, thereby reducing the memory capacity required for simulation.

[0055] (4) In the interference detection device described in (2), the gene point cloud data segmentation unit 113 recalculates the evaluation value of each combination of multiple cut surfaces by a preset number of repetitions. In this way, the interference detection device 1 can generate a three-dimensional model of the optimal set of cuboids for the robot and the surrounding obstacles.

[0056] 1: Interference detection device 10: Control Department 11: Input Section 12: Display Section 13: Memory Department 101: Contains a cuboid set transformation part 102: Simulation Execution Unit 111: Simplified Range Setting Section 112: Point Cloud Data Conversion Department 113: Gene Point Cloud Data Segmentation Department 114: Minimum Total Volume Search Unit 115: Margin Setting Department 121: Coordinate System Conversion Department 122: Interference Detection Unit 131: Motion Recording Memory Department 132: Shape Data Memory Department 200(1),200(2): cuboid d: Measure of margin m: Maximum number of repetitions n: Number of cuboids S1~S4, S21~S29: Steps X, Y, Z: Axes

Claims

1. An interference detection device for detecting interference between a robot and surrounding obstacles, the interference detection device comprising: a cuboid set conversion unit that converts the robot and the surrounding obstacles into three-dimensional models of cuboid sets respectively; an interference determination unit that determines whether there is interference between the three-dimensional models of the robot and the surrounding obstacles by simulating the movements of the three-dimensional models of the robot and the surrounding obstacles based on an action program; and an input unit that inputs the number of cuboids in the cuboid set, the cuboid set conversion unit comprising: a point cloud data conversion unit that converts the shape data of the robot and the surrounding obstacles into point cloud data of each of the robot and the surrounding obstacles respectively; The point cloud data segmentation unit segments the point cloud data of the robot and the surrounding obstacles into the aforementioned number of cuboids input to the aforementioned input unit, and calculates an evaluation value for each combination of the plurality of cut surfaces into the aforementioned number of cuboids. It then regenerates each combination of the aforementioned plurality of cut surfaces based on a genetic algorithm and calculates the aforementioned evaluation value again. The minimum total volume search unit searches for the plurality of cut surfaces with the highest evaluation value among the aforementioned evaluation values ​​of each combination of the aforementioned plurality of cut surfaces of the robot and the surrounding obstacles.

2. The interference detection device as claimed in claim 1, wherein the aforementioned cuboid set conversion unit comprises: a simplified range setting unit that sets the range of the shape data of the surrounding obstacles to be used in the aforementioned simulation; and a margin setting unit that sets a margin for the three-dimensional model of the cuboid set of the robot and the surrounding obstacles.

3. The interference detection device as claimed in claim 1, wherein the aforementioned point cloud data segmentation unit recalculates the aforementioned evaluation value for each combination of the aforementioned plurality of cut surfaces by a predetermined number of repetitions.

Images