Industrial robot multi-target path planning method based on quantum annealing algorithm

Abstract

The invention relates to an intelligent path planning technology, and discloses an industrial robot multi-target path planning method based on a quantum annealing algorithm, and the method comprises the steps: obtaining the environment information of an industrial site; path planning is carried out based on the environment information, and a quantum Hamiltonian is constructed according to an optimization target of a path planning problem; encoding each optional path in the path planning space by adopting quantum bits, and initializing the quantum bits to be in a uniform superposition state; quantum system energy is gradually reduced through a quantum annealing algorithm, and an optimal solution of quantum bits is searched; after quantum annealing is finished, a final path is extracted; carrying out smooth optimization processing on the extracted final path; and generating a motion instruction of the industrial robot according to the optimized final path. The invention further discloses a control system, the industrial robot and a computer readable storage medium. According to the method, the global optimality of path planning is improved while multi-target path planning is efficiently carried out on the industrial robot.

Description

Technical Field

[0001] The present application relates to the technical field of intelligent path planning, and in particular to a multi-objective path planning method for an industrial robot based on a quantum annealing algorithm, a control system, an industrial robot, and a computer-readable storage medium. Background Art

[0002] In modern industrial production, industrial robots are widely used in processes such as material handling and logistics. Path planning, one of the core technologies of industrial robots, directly impacts production efficiency and safety. With the increasing degree of industrial automation, the requirements for robot path planning are also becoming increasingly stringent. Traditional path planning methods primarily focus on minimizing path length. However, in real industrial scenarios, robots must also consider multiple objectives, such as minimizing energy consumption, maximizing obstacle avoidance safety, and prioritizing tasks. These objectives are often mutually constrained; for example, the shortest path may mean high energy consumption or low safety. Furthermore, the complexity and uncertainty of industrial environments, such as automotive manufacturing workshops and logistics warehouses, further exacerbate the difficulty of path planning and place even higher demands on path planning technology.

[0003] Currently, traditional algorithms (such as genetic algorithms and artificial potential field methods) are computationally intensive when dealing with multi-objective optimization problems, making it difficult to meet real-time requirements. These algorithms often require numerous complex calculations, making it difficult for path planning to keep up with the rapid changes in the production environment. For example, in an automotive manufacturing plant, when industrial robots need to move frequently between multiple workstations, traditional genetic algorithms can take seconds or even tens of seconds to plan a path. In actual production, path planning may be completed within hundreds of milliseconds, otherwise it will cause delays in the production process.

[0004] Furthermore, traditional algorithms struggle to strike an effective balance between multiple objectives and are prone to falling into local optimal solutions, preventing them from achieving a globally optimal path planning solution. For example, while artificial potential field methods can quickly plan paths, they are prone to falling into local optimal solutions, causing the robot to oscillate in complex environments and be unable to find a path to the target. While genetic algorithms possess global search capabilities, they struggle to determine appropriate weight coefficients when dealing with multi-objective optimization problems, resulting in poor optimization performance and difficulty achieving an optimal balance between path length, energy consumption, and obstacle avoidance safety.

[0005] The above content is only used to assist in understanding the technical solution of this application and does not constitute an admission that the above content is prior art. Summary of the Invention

[0006] The main purpose of this application is to provide an industrial robot multi-objective path planning method, control system, industrial robot and computer-readable storage medium based on a quantum annealing algorithm, aiming to achieve efficient multi-objective path planning for the industrial robot while improving the global optimality of path planning.

[0007] To achieve the above objectives, the present application provides a multi-objective path planning method for an industrial robot based on a quantum annealing algorithm, comprising the following steps:

[0008] When the industrial robot is running, the environmental information of the industrial site is obtained; wherein the environmental information includes the robot's initial position and target position, as well as the position and shape of obstacles;

[0009] Path planning is performed based on environmental information, and a quantum Hamiltonian is constructed based on the optimization objectives of the path planning problem, including path length, energy consumption, obstacle avoidance safety, and task priority.

[0010] Use quantum bits to encode each optional path in the path planning space, and initialize the quantum bits to a uniform superposition state;

[0011] The quantum annealing algorithm is used to gradually reduce the energy of the quantum system and search for the optimal solution of the quantum bit;

[0012] After the quantum annealing is completed, the state of the quantum bit is measured to obtain the encoding of the final path, and the final path is decoded and extracted;

[0013] Perform smooth optimization on the extracted final path;

[0014] Generate motion instructions for the industrial robot based on the optimized final path.

[0015] To achieve the above objectives, the present application also provides a control system, comprising:

[0016] A data acquisition module is used to obtain environmental information of the industrial site when the industrial robot is running; wherein the environmental information includes the robot's initial position and target position, as well as the position and shape of obstacles;

[0017] The path planning module is used to plan paths based on environmental information and construct a quantum Hamiltonian based on the optimization objectives of the path planning problem. These optimization objectives include path length, energy consumption, obstacle avoidance safety, and task priority. Quantum bits are used to encode each optional path in the path planning space and initialized to a uniform superposition state. A quantum annealing algorithm is used to gradually reduce the energy of the quantum system and search for the optimal solution of the quantum bits. After the quantum annealing is completed, the state of the quantum bits is measured to obtain the encoding of the final path, which is then decoded and extracted.

[0018] The post-processing module is used to perform smooth optimization processing on the extracted final path;

[0019] The robot control module is used to generate motion instructions for the industrial robot based on the optimized final path.

[0020] To achieve the above-mentioned objectives, the present application also provides an industrial robot, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the steps of the above-mentioned industrial robot multi-objective path planning method based on the quantum annealing algorithm are implemented.

[0021] To achieve the above objectives, the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned industrial robot multi-objective path planning method based on the quantum annealing algorithm are implemented.

[0022] The present application provides a multi-objective path planning method, control system, industrial robot and computer-readable storage medium based on a quantum annealing algorithm for industrial robots. The method proposes a multi-objective path planning method for industrial robots based on a quantum annealing algorithm, which simultaneously optimizes multiple objectives such as path length, energy consumption, obstacle avoidance safety and task priority; encodes dynamic interference factors in the industrial site into time-varying quantum perturbation terms, so that the robot can quickly adapt to environmental changes and replan the path; utilizes the quantum bit tunneling effect to perform global search, avoid local optimal solutions, and significantly improve the planning speed, thereby achieving efficient multi-objective path planning for industrial robots while improving the global optimality of path planning. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a schematic diagram of the steps of a multi-objective path planning method for an industrial robot based on a quantum annealing algorithm in one embodiment of the present application;

[0024] Figure 2 This is a schematic diagram of a path planning process based on a quantum annealing algorithm in one embodiment of the present application;

[0025] Figure 3 This is a schematic diagram of the smooth optimization process of the final path in one embodiment of the present application;

[0026] Figure 4 This is a schematic diagram of a control system in one embodiment of the present application;

[0027] Figure 5 Schematic diagram of the internal structure of an industrial robot according to an embodiment of the present application.

[0028] The realization of the objectives, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0029] The embodiments of the present application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be understood as limiting the present application. All other embodiments obtained by ordinary technicians in this field based on the embodiments in the present application without making any creative efforts shall fall within the scope of protection of the present application.

[0030] In addition, if the descriptions of "first", "second", etc. are involved in this application, they are only used for descriptive purposes (such as to distinguish the same or similar features) and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" or "second" may explicitly or implicitly include at least one of such features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in this field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0031] Reference Figure 1 In one embodiment, a multi-objective path planning method for an industrial robot based on a quantum annealing algorithm includes:

[0032] Step S10: When the industrial robot is running, obtain environmental information of the industrial site; wherein the environmental information includes the robot's initial position and target position, and the position and shape of obstacles;

[0033] Step S20: performing path planning based on environmental information and constructing a quantum Hamiltonian according to the optimization objectives of the path planning problem; wherein the optimization objectives include path length, energy consumption, obstacle avoidance safety, and task priority;

[0034] Step S30: Encode each optional path in the path planning space using quantum bits, and initialize the quantum bits to a uniform superposition state;

[0035] Step S40: gradually reducing the energy of the quantum system through a quantum annealing algorithm to search for the optimal solution of the quantum bit;

[0036] Step S50: After the quantum annealing is completed, the state of the quantum bit is measured to obtain the encoding of the final path, and the final path is decoded and extracted;

[0037] Step S60: performing smoothing optimization processing on the extracted final path;

[0038] Step S70: Generate motion instructions for the industrial robot according to the optimized final path.

[0039] In this embodiment, the execution terminal of the embodiment may be an industrial robot, or may be other equipment or devices (such as a control system) that controls the industrial robot.

[0040] As described in step S10, an industrial robot is a multi-jointed manipulator or multi-degree-of-freedom machine device designed for industrial applications. It operates according to pre-programmed programs and can be used through automated control to perform various industrial production activities, such as material handling, welding, and assembly. For example, an industrial robot can be used to manufacture material handling robots in a workshop, possessing certain mobility and manipulation capabilities to complete material transportation tasks within the workshop.

[0041] An industrial site is the actual location where industrial production activities take place, encompassing various elements such as production equipment, materials, and personnel. This is where industrial robots are actually used. For example, if an industrial robot is a material handling robot in a manufacturing workshop, then the industrial site is the manufacturing workshop, the spatial scope where the material handling robot operates.

[0042] Optionally, industrial robots can be equipped with key equipment for environmental information collection: lidar and cameras. These two devices work together to comprehensively and accurately obtain environmental information within the workshop.

[0043] LiDAR continuously emits laser pulses at a fixed frequency. Once these pulses encounter obstacles within the industrial site (such as other robots, material racks, production equipment, etc.), they are reflected back, and the LiDAR receives these reflected signals. By analyzing and processing information such as the time difference between the transmitted and received laser pulses and the laser propagation speed, the position of obstacles within the industrial site can be accurately determined. Furthermore, based on the intensity and distribution of the reflected signals at different angles, the shape of the obstacle can be further depicted, providing specific information on the obstacle's location and shape.

[0044] The camera is responsible for collecting image data around the robot. This image data contains rich visual information. Using advanced image recognition technology, it is analyzed and processed to help identify specific types of obstacles, such as other robots, material racks, and production equipment. Furthermore, the image data also contains environmental characteristics of the industrial site, such as ground texture and lighting conditions. This information helps to more comprehensively understand the overall industrial environment.

[0045] Determine the robot's initial and target positions. For example, the robot's initial position is set to coordinates (0,0), which represents the robot's starting point in the two-dimensional coordinate system of the workshop; the target position is set to coordinates (10,10), which is the robot's destination.

[0046] Combining information from LiDAR and cameras, the specific location and range of each obstacle within the industrial site can be precisely determined. For example, obstacle A occupies the spatial range from coordinates (2,2) to (3,3), and obstacle B occupies the spatial range from coordinates (6,6) to (7,7). This information clearly defines the specific area of ​​the obstacle within the industrial site, providing critical foundational data for subsequent path planning.

[0047] After acquiring the data of the environmental information of the industrial site based on step S10, refer to Figure 2 , then describe steps S20 to S50, the process of path planning based on the quantum annealing algorithm.

[0048] As described in step S20, based on the industrial site environment information obtained in step S10, including the robot's initial position, target position, and the position and shape of obstacles, a map model that can accurately reflect the environment is constructed.

[0049] For example, when using a grid map, the industrial site is divided into uniform grid cells, with the size of each cell determined based on the actual environment and the robot's motion accuracy. Grid cells where obstacles are located are marked as impassable, while other areas are marked as passable. This division not only discretizes the continuous environment but also provides clear boundary and node information for subsequent searches in the path planning space.

[0050] For complex industrial environments, topological maps can be used to further optimize the representation of the path planning space. Topological maps abstract the environment into a network of nodes and edges, where nodes represent key locations and edges represent the connections between nodes. This representation can more efficiently handle large-scale environments and complex path planning problems.

[0051] After building the map model, the boundaries of the path planning space need to be clearly defined. This includes physical boundaries, such as walls and fences at the industrial site, as well as the robot's kinematic boundaries, such as the robot's maximum range of motion and minimum turning radius.

[0052] At the same time, considering the robot's motion characteristics and task requirements, the constraints of the path planning space are determined and a multi-objective function is defined. For example, the robot's maximum speed and acceleration limits, as well as the motion posture requirements at different task stages, are considered. These boundaries and constraints will define the range of feasible paths for the robot in the path planning space, ensuring that subsequent path search is carried out within a reasonable range.

[0053] The quantum Hamiltonian is constructed based on the optimization objectives of the path planning problem, where the optimization objectives include path length, energy consumption, obstacle avoidance safety, and task priority. The steps of constructing the quantum Hamiltonian include:

[0054] Calculate the sub-Hamiltonian corresponding to each optimization objective;

[0055] The weighted sum of each sub-Hamiltonian is performed to generate the quantum Hamiltonian.

[0056] Among them, in the path planning space, the path length can be measured by calculating the sum of the distances between adjacent nodes. In the grid map, the distance between adjacent grid cells can be determined by the grid edge length; in the topological map, the edge length represents the distance between nodes. Construct the sub-Hamiltonian (H) corresponding to the path length p ath) is defined as the sum of the distances of each segment on the path, that is, H p ath=Σd i , where d i is the distance of the i-th segment on the path.

[0057] Energy consumption is closely related to the path selection in the path planning space. In the path planning space, different paths may lead to different robot motion modes, thus affecting energy consumption. Construct the sub-Hamiltonian (H) corresponding to energy consumption. e energy) is related to the robot's movement speed and acceleration and can be expressed as H e energy=∑(E k in+E p ot), where E k in is kinetic energy, E p ot is the potential energy.

[0058] Obstacle avoidance safety is a key indicator to ensure that the robot does not collide with obstacles in the path planning space. It can be measured by calculating the minimum distance between the path and the obstacle. In the path planning space, the obstacle occupies a certain area. The larger the minimum distance between the path and the obstacle, the safer the path. Construct the sub-Hamiltonian (H) corresponding to obstacle avoidance safety. s afety) is calculated based on the distance between the robot and the obstacle, and is defined as H s afety=∑(1 / d obs,i), where (d o bs,i) is the distance between the robot and the i-th obstacle.

[0059] In the path planning space, different tasks may correspond to different target locations and time requirements. Each task is assigned a priority weight. The larger the priority weight, the higher the task priority. For example, urgent orders have a higher priority and need to be completed first. When planning a path, if the robot needs to complete multiple tasks, the impact of task priority in the path planning space needs to be considered. Construct a sub-Hamiltonian (H) corresponding to the task priority. p Priority) is: according to the urgency and importance of the task, different priorities are assigned to different tasks, which is defined as H p riority=∑(w t ask,i), where (w t ask,i) is the priority weight of the i-th task.

[0060] As described in step S30, a quantum bit (qubit) is the fundamental unit of information in quantum computing. Unlike classical bits, it can be in a state not only of 0 or 1 but also a superposition of these two states. This makes quantum computing parallel and efficient when solving certain problems. In the path planning problem, qubits are used to encode the robot's possible paths. Assume there are 10 qubits, each representing a dimension in the path planning space, and their state can be represented as a superposition of |0> and |1>.

[0061] In step S20, a path planning space is constructed based on the environmental information. This space can be composed of a series of nodes and edges connecting these nodes. For example, in a grid map, each grid point is a node, and the connections between adjacent grid points are edges. Different ways of moving from the robot's initial position to the target position constitute different possible paths.

[0062] Assume there are 10 qubits, each representing a dimension in the path planning space. The qubit states |0> and |1> in each dimension, as well as their superposition, can represent different path choices or states in that dimension. For example, in a particular dimension, |0> could indicate that the robot has chosen a specific movement method or passed through a specific area, while |1> represents a different choice. This encoding method decomposes the complex information in the path planning space into individual dimensions, representing different paths through combinations of qubit states.

[0063] Optionally, establish a clear mapping between qubit state combinations and available paths in the path planning space. Because each qubit has multiple states (superpositions), and the number of different state combinations for 10 qubits is large, a set of rules needs to be developed based on the actual conditions of the path planning space to determine the specific path corresponding to each state combination. For example, by pre-analyzing the path planning space, different regions or movement methods can be associated with qubit states, thereby achieving a mapping from qubit states to paths.

[0064] The superposition state of a quantum bit refers to a linear combination of the quantum bit being in the 0 and 1 states at the same time. Its mathematical representation is |ψ>=α|0>+β|1>, where α and β are complex numbers and satisfy |α| 2 +|β| 2 = 1. When all qubits are in a uniform superposition state, it means that every possible combination of qubit states has the same probability of being measured.

[0065] Initialize the qubits to a uniform superposition state: This indicates that a single quantum bit is initially in an equiprobable superposition state of |0> and |1>, that is, the probability of being in the |0> state and the |1> state is 1 / 2.

[0066] For example, for 10 qubits, the initial state is:

[0067]

[0068] in, The tensor product represents the tensor product, meaning that the states of the 10 qubits are independent of each other, and each qubit is in the uniform superposition described above. The initial state of the entire 10-qubit system represents an equally probable superposition of all possible combinations of the 10 qubit states.

[0069] Specifically, each qubit's initial state is an equiprobable superposition of |0> and |1>. This means that in the initial stages of path planning, all possible paths are considered potential solutions. This encoding method enables the quantum annealing algorithm to explore multiple path possibilities simultaneously, laying the foundation for subsequent optimization. Because the qubits are in a superposition state, during the subsequent quantum annealing process, the quantum system can "search" all possible paths simultaneously, rather than evaluating each path sequentially as in classical algorithms. This greatly improves the efficiency of the search for the optimal path.

[0070] As described in step S40, the quantum annealing algorithm is an optimization algorithm based on the principles of quantum mechanics. It simulates the process of a quantum system evolving from a high-energy state to a low-energy state. During this process, the system overcomes energy barriers through quantum tunneling, increasing the chances of finding a global optimal solution rather than being trapped in a local optimal solution. For industrial robot path planning, the energy of the quantum system corresponds to the objective function of the path planning problem; lower energy corresponds to a more optimal path.

[0071] The step of gradually reducing the quantum system energy by a quantum annealing algorithm and searching for the optimal solution of the quantum bit includes:

[0072] Initialize quantum system parameters, including annealing time, annealing temperature, and annealing times;

[0073] Adjusting the tunneling probability of the qubit according to the annealing temperature;

[0074] During iterative training, the states of some qubits are randomly flipped, and the energy of the quantum system in the new state is calculated. If the new energy is lower than the original energy, the new state is accepted.

[0075] After each iteration, the annealing temperature is lowered and the number of annealing times is decreased;

[0076] Detecting whether the annealing end condition is met; wherein the annealing end condition is reaching a preset annealing number or the quantum system energy reaches convergence;

[0077] If not, return to the step of adjusting the tunneling probability of the quantum bit according to the annealing temperature;

[0078] If so, the quantum annealing is determined to be finished.

[0079] As described in step S50, after the quantum annealing algorithm completes the entire annealing process, the quantum system will tend to a stable state with the lowest energy, at which point the state of the qubit can be measured. When the measurement is performed after the quantum annealing ends, the qubit collapses from the superposition state to one of the two eigenstates, |0> or |1>. Through the measurement operation, a binary code sequence consisting of |0> and |1> is obtained. This sequence represents the binary code of the final path. For example, the measurement result may be the sequence |1>, |0>, |1>, |1>, |0>, |0>, |1>, |0>, |1>, |0>.

[0080] In path planning problems, the industrial environment is typically gridded, divided into small cells, each corresponding to a location in the path planning grid. To convert the binary state of a qubit into actual coordinates, a predefined decoding rule is required. In this example, each qubit in the sequence corresponds to a cell in the path planning grid, and |1> and |0> are assigned specific movement directions, with |1> indicating rightward movement and |0> indicating upward movement.

[0081] Before decoding, the initial position of the industrial robot must be determined. This initial position is the starting point of the path, assuming the coordinates of the starting point are (0,0). Movement is performed according to the binary sequence: Starting from the first qubit in the obtained binary sequence, movement is performed according to predefined rules. For the first qubit, if it is |1>, the robot moves one unit to the right from the starting point (0,0) to the coordinate point (1,0); if it is |0>, the robot moves one unit upward from the starting point (0,0) to the coordinate point (0,1). For the second qubit, based on the current coordinate point, if it is |1>, the robot moves one unit to the right; if it is |0>, the robot moves one unit upward.

[0082] For example, if the first bit is |1> and reaches (1,0), and the second bit is |0>, the robot moves one unit upward from (1,0) to the coordinate point (1,1). Similarly, each bit in the binary sequence is processed in turn, and the robot's movement direction is determined based on its state, resulting in a series of coordinate points.

[0083] Generate a path node sequence: Record the coordinate points reached after each movement, forming a path node sequence. For example, for a binary sequence with the measurement results of |1>, |0>, |1>, |1>, |0>, |0>, |1>, |0>, |1>, |0>, the resulting path node sequence after decoding might be: (0,0)→(1,0)→(1,1)→(2,1)→(3,1)→(3,2)→(3,3)→(4,3)→(4,4)→(5,4)→(5,5) (This is just an example; the actual result depends on the specific binary sequence and initial position). This path node sequence represents the actual movement path of the industrial robot in the workshop.

[0084] Through the above steps, the decoding process from the binary state of the quantum bit to the actual movement path of the robot is completed.

[0085] As described in step S60, the final path obtained from the quantum annealing algorithm may contain discontinuities such as broken lines and sharp corners. Therefore, it is necessary to perform smooth optimization on the path to ensure that the industrial robot moves smoothly.

[0086] Optionally, the final path may contain some very close points. These points have little impact on the overall shape of the path, but they increase the complexity of subsequent processing. You can set a distance threshold to merge adjacent points with a distance less than the threshold into a single point, thereby reducing the number of path points.

[0087] Optional smoothing optimization methods include: spline interpolation, Bezier curve method and minimum curvature method.

[0088] After smoothing optimization, the resulting new path point sequence is the final optimized path. These path points are stored in order and used as the basis for generating motion instructions for the industrial robot in the next step.

[0089] As described in step S70, after obtaining the optimized final path, the corresponding motion instructions can be generated in combination with the constraints of the motion parameters of the industrial robot. Assume that the maximum speed of the robot is 1.0m / s and the maximum acceleration is 0.5m / s 2 , the distance between adjacent nodes on the path is 1.0m. Calculate the robot's speed and steering angle on straight and curved segments:

[0090] The straight section speed V1 = 1.0 m / s, the straight section steering angle θ1 = 0°; the curve section speed V2 = 0.8 m / s (adjusted according to the curve radius), the curve steering angle θ2 = 45°.

[0091] The final optimized path is carefully analyzed to determine whether each segment formed by adjacent nodes is a straight line or a curved line. For straight line segments, the robot's speed is set to its maximum speed of 1.0 m / s and its steering angle is set to 0°. For curved segments, the default speed is 0.8 m / s. However, if the curved segment has a specific radius, the speed needs to be adjusted accordingly. The steering angle is also set to 45°.

[0092] For example, in the straight segment of the path (2,2)→(3,3), the industrial robot travels at a speed of 1.0 m / s; in the curved segment of the path (3,3)→(4,4), the robot travels at a speed of 0.8 m / s and turns 45°.

[0093] Based on the above analysis and calculation of each path segment, a complete motion instruction sequence is generated according to the order of the path. Each motion instruction contains the starting and ending nodes of the path segment, the speed setting value, the steering angle setting value, and related information on the speed change (if any).

[0094] The generated motion command sequence is transmitted to the robot control system. Based on these commands, the robot control system adjusts the robot's motion state in real time, ensuring that the robot accurately follows the planned path. During the driving process, the control system continuously monitors the robot's actual speed, position, and steering angle, comparing them with the set values ​​in the motion command. It provides timely feedback and adjustments to ensure that the robot can move stably and accurately along the optimized final path.

[0095] In this way, a multi-objective path planning method based on a quantum annealing algorithm effectively improves the path planning capabilities and adaptability of industrial robots. Leveraging the tunneling effect of quantum bits, the algorithm can globally search for the optimal path solution, avoiding local optima and significantly increasing planning speed, surpassing traditional genetic algorithms. This method comprehensively optimizes multiple objectives, including path length, energy consumption, obstacle avoidance safety, and task priority, resulting in shorter path lengths, lower energy consumption, and significantly improved safety. Furthermore, by encoding dynamic interference factors as time-varying quantum perturbation terms, the robot can quickly adapt to environmental changes, such as sudden obstacles, and promptly replan its path, minimizing production interruptions and improving efficiency and safety. This method has broad application prospects in a wide range of industrial scenarios, including automotive manufacturing workshops and logistics warehouses, providing an efficient solution for industrial robot path planning and significantly promoting the development of industrial automation.

[0096] In one embodiment, a multi-objective path planning method for industrial robots based on a quantum annealing algorithm is proposed, which simultaneously optimizes multiple objectives such as path length, energy consumption, obstacle avoidance safety, and task priority; the dynamic interference factors of the industrial site are encoded as time-varying quantum perturbation terms, so that the robot can quickly adapt to environmental changes and replan the path; the quantum bit tunneling effect is used to perform global search, avoid local optimal solutions, and significantly improve the planning speed, thereby achieving efficient multi-objective path planning for industrial robots while improving the global optimality of path planning.

[0097] In one embodiment, based on the above embodiment, the step of gradually reducing the energy of the quantum system by a quantum annealing algorithm and searching for the optimal solution of the quantum bit includes:

[0098] Initialize quantum system parameters, including annealing time, annealing temperature, and annealing times;

[0099] Adjusting the tunneling probability of the qubit according to the annealing temperature;

[0100] During iterative training, the states of some qubits are randomly flipped, and the energy of the quantum system in the new state is calculated. If the new energy is lower than the original energy, the new state is accepted.

[0101] After each iteration, the annealing temperature is lowered and the number of annealing times is decreased;

[0102] Detecting whether the annealing end condition is met; wherein the annealing end condition is reaching a preset annealing number or the quantum system energy reaches convergence;

[0103] If not, return to the step of adjusting the tunneling probability of the quantum bit according to the annealing temperature;

[0104] If so, the quantum annealing is determined to be finished.

[0105] In this embodiment, the annealing time determines the duration of the entire quantum annealing process. A reasonable annealing time is critical. If the time is too short, the quantum system may not have enough time to fully explore the low-energy state, making it difficult to find the optimal solution; if the time is too long, the computational cost will increase.

[0106] The annealing temperature is related to the tunneling probability of quantum bits. At higher temperatures, quantum bits have a greater probability of tunneling, which helps the system escape from the local optimal solution and search in a wider solution space; at lower temperatures, the tunneling probability decreases, and the system is more inclined to stabilize in a low-energy state.

[0107] The number of annealing cycles specifies the total number of iterations in the entire annealing process. Multiple iterations can increase the probability of finding the optimal solution, and the system will try to find a better state with each iteration.

[0108] For example, when initializing parameter settings, the annealing time may be set to 100 milliseconds, the annealing temperature T=1.0, and the number of annealing times may be set to 100 times.

[0109] The tunneling probability of a qubit is a key mechanism in quantum annealing algorithms that allows it to cross energy barriers and explore different states. Annealing temperature is positively correlated with tunneling probability; higher temperatures increase the probability of qubit tunneling. Tunneling probability can be adjusted using a specific functional relationship, allowing qubits to transition more freely between states at high temperatures, thereby exploring a wider range of solution spaces.

[0110] For example, assume that the tunneling probability Ptunnel=2T / 3+2 / 15; at the first iteration, T=1.0, Ptunnel=0.8; at the 50th iteration, T=0.1, Ptunnel=0.2.

[0111] During iterative training, the states of some qubits are randomly flipped, and the quantum system energy under these new states is calculated. This random flipping allows the quantum system to explore different states in the solution space. By randomly changing the states of some qubits, the quantum system can try different path encoding combinations, which is equivalent to trying different paths in a path planning problem. For example, in the first iteration, the states of qubits 3 and 7 are randomly flipped, changing |0> to |1> and |1> to |0>.

[0112] After flipping the states of some qubits, the energy of the quantum system in the new state is calculated based on the constructed quantum Hamiltonian. This energy corresponds to the objective function value of the path planning problem, with lower energy indicating a better path. If the new energy is lower than the original energy, the path corresponding to the new state is better, and the new state is accepted. This satisfies the goal of finding a low-energy state (i.e., the optimal path).

[0113] As the iterations progress, the annealing temperature is lowered after each iteration. A lower temperature reduces the tunneling probability of the qubit, making the system more likely to stabilize near the currently found low-energy state, which helps converge to a better solution. For example, the temperature is multiplied by 0.95 with each iteration.

[0114] And, after each iteration, the number of annealing times is reduced by 1, and the remaining number of iterations is recorded.

[0115] After each iteration is completed and the annealing temperature and annealing times are updated, it is checked whether the annealing end condition is met.

[0116] Optionally, if the number of iterations reaches a pre-set number of annealing times, it means that the specified search process has been completed and the annealing can be terminated at this time; or, when the energy of the quantum system changes very little in multiple iterations and hardly decreases, it means that the system is close to the optimal solution and has reached a convergence state, and the annealing can also be terminated.

[0117] If the annealing end condition is not met: return to the step of "adjusting the tunneling probability of the quantum bit according to the annealing temperature" and continue the iterative search to further find a better quantum bit state.

[0118] If the annealing end condition is met: the quantum annealing is determined to be ended. At this time, the current state of the quantum bit is likely to encode the optimal solution to the industrial robot path planning problem.

[0119] In practical applications, this process is a dynamic search, with the state of the qubits constantly adjusting to find the optimal solution. For example, in a material handling scenario in an automotive manufacturing plant, as a robot approaches its target location, its path planning needs to be adjusted in real time to avoid other moving robots and material racks. The quantum annealing algorithm, by continuously adjusting the state of the qubits, enables the robot to dynamically find the optimal path, thereby improving production efficiency.

[0120] In one embodiment, based on the above embodiment, after the step of randomly flipping the states of some quantum bits during the iterative training process and calculating the quantum system energy in the new state, the method further includes:

[0121] If the new energy is not lower than the original energy, the energy acceptance probability is calculated and tested to see if it is greater than the random number;

[0122] If the energy acceptance probability is greater than the random number, the new state is accepted;

[0123] If the energy acceptance probability is not greater than the random number, the new state is not accepted;

[0124] The energy acceptance probability is calculated as follows: exp(―ΔE / (kT)); ΔE is the energy difference between the new and old energies, k is the Boltzmann constant, and T is the annealing temperature of the current round.

[0125] In this embodiment, if the new energy is not lower than the old energy, the new state is not rejected directly. Instead, the energy acceptance probability is calculated. The formula for calculating the energy acceptance probability is exp(-ΔE / (kT)), where ΔE is the energy difference between the new and old energies, k is the Boltzmann constant, and T is the annealing temperature of the current round. The larger the energy difference ΔE, the smaller the energy acceptance probability; the higher the annealing temperature T, the greater the energy acceptance probability.

[0126] After calculating the energy acceptance probability, generate a random number between 0 and 1 and compare the energy acceptance probability with the random number:

[0127] If the energy acceptance probability is greater than the random number: Even though the energy of the new state is not lower than the original state, the new state is accepted based on a certain probability. This is to prevent the algorithm from falling into a local optimal solution too early, and to a certain extent allows the system to jump out of the current local optimal area and continue to explore the solution space.

[0128] If the energy acceptance probability is not greater than the random number: the new state is not accepted at this time, the system maintains the original state and continues to search in the neighborhood of the current solution.

[0129] During the algorithm's iterative process, the acceptance probability mechanism allows the system to conduct a more extensive exploration of the solution space. As the temperature gradually decreases, the system gradually transitions from a global search to a local search. At high temperatures, the system has greater freedom to experiment with different states, thereby covering a large portion of the solution space. At low temperatures, the system tends to stabilize near low-energy states and conduct a detailed search of already identified promising areas. This combination of global and local search increases the probability of the system finding the global optimal solution. By randomly jumping between different energy states, the system is able to traverse a wider range of solutions, thereby increasing the chance of discovering the global optimal solution hidden deep within the solution space.

[0130] In one embodiment, based on the above embodiment, the step of constructing the quantum Hamiltonian includes:

[0131] Calculate the sub-Hamiltonian corresponding to each optimization objective;

[0132] The weighted sum of each sub-Hamiltonian is performed to generate the quantum Hamiltonian.

[0133] In this embodiment, a quantum Hamiltonian is constructed based on the objectives of the path planning problem (path length, energy consumption, obstacle avoidance safety, and task priority, etc.). The Hamiltonian is expressed as:

[0134] H=αH p ath+βH e energy+γH s afety+δH p riority;

[0135] Among them, H p riority is the sub-Hamiltonian corresponding to the task priority, H e energy is the sub-Hamiltonian corresponding to energy consumption, H p ath is the sub-Hamiltonian corresponding to the path length, H s afety is the sub-Hamiltonian corresponding to obstacle avoidance safety; δ is the weight of the sub-Hamiltonian corresponding to task priority, β is the weight of the sub-Hamiltonian corresponding to energy consumption, α is the weight of the sub-Hamiltonian corresponding to path length, and γ is the weight of the sub-Hamiltonian corresponding to obstacle avoidance safety.

[0136] Optionally, the weight relationship between the sub-Hamiltonians is: δ<β<α<γ;

[0137] Among them, δ is the weight of the sub-Hamiltonian corresponding to the task priority, β is the weight of the sub-Hamiltonian corresponding to the energy consumption, α is the weight of the sub-Hamiltonian corresponding to the path length, and γ is the weight of the sub-Hamiltonian corresponding to the obstacle avoidance safety.

[0138] Optionally, weights α=0.3, β=0.2, γ=0.4, δ=0.1.

[0139] In one embodiment, based on the above embodiment, the multi-objective path planning method for an industrial robot based on a quantum annealing algorithm further includes:

[0140] After decoding the final path, the system checks whether there is a risk of collision with obstacles based on the workshop map and obstacle information.

[0141] If there is no risk, the final path is extracted.

[0142] In this embodiment, after completing the quantum bit state decoding and obtaining the final path of the industrial robot, it is also possible to carefully check whether this path has the risk of collision with obstacles based on the workshop map and obstacle information.

[0143] The workshop map clearly depicts the layout of the entire industrial site, accurately identifying the boundaries of each area, the location of passages, and other information. The workshop map is presented in a two-dimensional or three-dimensional grid format, with each grid cell representing a specific location within the workshop. Obstacle information includes the specific location and shape of all obstacles within the workshop; the location of obstacles can be precisely represented by coordinates, while the shape can be described by geometric figures (such as rectangles and circles). This information is crucial for determining whether a path is safe.

[0144] Optionally, map each path node in the final path to a grid cell on the shop floor map. For each node on the path, check whether its grid cell and adjacent grid cells are occupied by obstacles. If so, the path node may collide with the obstacle.

[0145] Optionally, analysis can be performed based on the shape of the obstacle and the geometric characteristics of the path. For a path consisting of straight segments, the distance between the path segment and the obstacle geometry can be calculated. If the distance is less than a certain safety threshold, a collision risk is determined. For example, for a rectangular obstacle and a straight path, the shortest distance between the straight line and each side of the rectangle is calculated; for a circular obstacle, the distance from the straight line to the center of the circle is calculated and compared with the radius.

[0146] All nodes in the final path are collided with each other, starting from the path's starting point and continuing until the path's end. If a node is found to be at risk of colliding with an obstacle, the final path is considered to have a problem.

[0147] After comprehensive collision detection, if all nodes on the final path do not collide with obstacles, that is, there is no collision risk, then it can be determined that this path is safe and feasible.

[0148] Once a path is determined to be risk-free, the final, safe path is extracted from the inspection process. This extraction process preserves the complete node sequence of the path for subsequent use. These node sequences can then be used to generate motion commands for the industrial robot, guiding it safely within the workshop. Furthermore, this path serves as the basis for subsequent smoothing optimization to further improve path quality and robot motion efficiency.

[0149] Optionally, if a collision risk is detected on the final path, consider that the collision risk on the original path may be due to improper weighting of certain optimization objectives. For example, excessive emphasis was placed on path length, while obstacle avoidance safety was neglected. In this case, the weight of obstacle avoidance safety can be appropriately increased, while the weights of other objectives (such as path length and energy consumption) can be adjusted accordingly to place greater emphasis on obstacle avoidance. Then, based on the adjusted optimization objectives, the quantum Hamiltonian is reconstructed. The quantum Hamiltonian describes the energy function of the path planning problem, and the new Hamiltonian will be more inclined to generate paths that avoid obstacles. During the construction process, the updated optimization objectives and workshop environment information are combined to ensure that the energy function accurately reflects the quality of the paths. Using the newly constructed quantum Hamiltonian, the quantum annealing algorithm is re-initiated for path planning. The previous steps are repeated, including qubit encoding of each optional path in the path planning space, initializing the qubits to a uniform superposition state, and searching for the optimal solution using the quantum annealing algorithm, to obtain a new final path.

[0150] Alternatively, if a collision risk is detected on the final path, the area of ​​collision risk is defined within the original path. The starting and ending nodes of this area are determined, clarifying the path range that requires adjustment. A path planning algorithm (such as A* or Dijkstra) is used to perform local optimization adjustments within the collision area. The locally optimized subpath is then joined with the unaffected portion of the original path to form a new final path.

[0151] In one embodiment, based on the above embodiment, the step of performing smooth optimization processing on the extracted final path includes:

[0152] The Bezier curve is used to perform smooth optimization on the extracted final path.

[0153] In this embodiment, before using the Bezier curve for smoothing, possible redundant points in the path may be removed, that is, points that are too close and have little impact on the overall shape of the path, thereby reducing the complexity of subsequent processing.

[0154] Reference Figure 3 ,The process of using Bezier curve to smoothly optimize the extracted final path is as follows:

[0155] Divide the extracted path into multiple path segments, and select appropriate control points for each path segment to construct a Bezier curve. For example, for the path segment (2,2)→(3,3)→(4,4), use a quadratic Bezier curve with control points P0(2,2), P1(3,3), and P2(4,4). The curve formula is:

[0156] B(t)=(1―t) 2P0+2t(1-t)P1+t 2 P2, t∈[0,1];

[0157] Within this range, by continuously changing the value of t, a series of points on the Bezier curve can be calculated to obtain a smoothed path.

[0158] For example, when t = 0.5, B(0.5) = 0.25×(2,2) + 2×0.5×0.5×(3,3) + 0.25×(4,4) = (3,3), which can make the path smoother.

[0159] Optionally, after the path smoothing process is completed, the distance between the newly generated smooth path and obstacles may be checked to ensure the safety of the robot operation.

[0160] Optionally, for each point on the path, the Euclidean distance formula is used to calculate the distance between it and each obstacle. The minimum distance between the path and the obstacle is determined and compared with a pre-set safety distance threshold. For example, if the safety distance threshold is set to 0.5m, and the calculated minimum distance between the path and obstacle A is 0.5m, then the path meets the safety requirements. If the minimum distance is less than the safety distance threshold, then there is a safety risk, and the path or Bezier curve control points need to be readjusted, and smoothing and safety checks must be performed again until the safety requirements are met.

[0161] After smoothing and safety checking, if the path ensures both smoothness and safety requirements, the newly generated path point sequence is stored in order. This path is the final path after optimization and smoothing, which can be used as the basis for generating industrial robot motion instructions in the next step.

[0162] In addition, refer to Figure 4 In an embodiment of the present application, a control system Z10 is further provided, including:

[0163] The data acquisition module Z11 is used to obtain environmental information of the industrial site when the industrial robot is running; wherein the environmental information includes the robot's initial position and target position, as well as the position and shape of obstacles;

[0164] Path planning module Z12 is used to plan paths based on environmental information and construct a quantum Hamiltonian based on the optimization objectives of the path planning problem. These optimization objectives include path length, energy consumption, obstacle avoidance safety, and task priority. It uses quantum bits to encode each optional path in the path planning space and initializes the quantum bits to a uniform superposition state. It uses a quantum annealing algorithm to gradually reduce the quantum system energy and search for the optimal solution of the quantum bits. After the quantum annealing is completed, the quantum bit state is measured to obtain the encoding of the final path, which is then decoded and extracted.

[0165] Post-processing module Z13, used for smoothing and optimizing the extracted final path;

[0166] The robot control module Z14 is used to generate motion instructions for the industrial robot according to the optimized final path.

[0167] Optionally, the control system Z10 may be a virtual control system (such as a virtual machine) or a physical device (such as a physical device other than an industrial robot that can execute the corresponding method).

[0168] In addition, an industrial robot is also provided in the embodiment of the present application. The internal structure of the industrial robot can be as follows: Figure 5 As shown, it includes a processor, memory, communication interface, and input interface connected via a system bus. The processor is used to provide computing and control capabilities. The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database is used to store data called by the computer program. The communication interface is used to communicate data with an external terminal. The input interface is used to receive signals input by an external device. When the computer program is executed by the processor, it implements a multi-objective path planning method for an industrial robot based on a quantum annealing algorithm as described in the above embodiment.

[0169] Those skilled in the art will understand that Figure 5 The structure shown in the figure is merely a block diagram of a portion of the structure related to the present invention and does not limit the industrial robot to which the present invention is applied. For example, in some optional embodiments, the industrial robot may further include an output interface (not shown in the figure), and the output interface is also connected to the system bus and is used to output corresponding signals to the peripheral device.

[0170] In addition, this application also provides a computer-readable storage medium, which includes a computer program. When executed by a processor, the computer program implements the steps of the multi-objective path planning method for an industrial robot based on a quantum annealing algorithm as described in the above embodiment. It is understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0171] In summary, the multi-objective path planning method for industrial robots based on the quantum annealing algorithm, control system, industrial robot and computer-readable storage medium provided in the embodiments of this application propose a multi-objective path planning method for industrial robots based on the quantum annealing algorithm, which simultaneously optimizes multiple objectives such as path length, energy consumption, obstacle avoidance safety and task priority; encodes dynamic interference factors in the industrial site into time-varying quantum perturbation terms, so that the robot can quickly adapt to environmental changes and replan the path; utilizes the quantum bit tunneling effect for global search to avoid local optimal solutions and significantly improve the planning speed, thereby achieving efficient multi-objective path planning for industrial robots while improving the global optimality of path planning.

[0172] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided in this application and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM).

[0173] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, apparatus, article, or method comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, apparatus, article, or method. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, apparatus, article, or method comprising the element.

[0174] The above description is only a preferred embodiment of the present application and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A multi-objective path planning method for industrial robots based on quantum annealing algorithm, characterized in that: include: When the industrial robot is running, the environmental information of the industrial site is obtained; wherein the environmental information includes the robot's initial position and target position, as well as the position and shape of obstacles; Path planning is performed based on environmental information, and a quantum Hamiltonian is constructed based on the optimization objectives of the path planning problem, including path length, energy consumption, obstacle avoidance safety, and task priority. Use quantum bits to encode each optional path in the path planning space, and initialize the quantum bits to a uniform superposition state; The quantum annealing algorithm is used to gradually reduce the energy of the quantum system and search for the optimal solution of the quantum bit; After the quantum annealing is completed, the state of the quantum bit is measured to obtain the encoding of the final path, and the final path is decoded and extracted; Perform smooth optimization on the extracted final path; Generate motion instructions for the industrial robot based on the optimized final path.

2. The multi-objective path planning method for industrial robots based on quantum annealing algorithm according to claim 1, characterized in that: The step of gradually reducing the quantum system energy by a quantum annealing algorithm and searching for the optimal solution of the quantum bit includes: Initialize quantum system parameters, including annealing time, annealing temperature, and annealing times; Adjusting the tunneling probability of the qubit according to the annealing temperature; During the iterative training process, the states of some quantum bits are randomly flipped, and the energy of the quantum system in the new state is calculated; if the new energy is lower than the original energy, the new state is accepted; After each iteration, the annealing temperature is lowered and the number of annealing times is decreased; Detecting whether the annealing end condition is met; wherein the annealing end condition is reaching a preset annealing number or the quantum system energy reaches convergence; If not, return to the step of adjusting the tunneling probability of the quantum bit according to the annealing temperature; If so, the quantum annealing is determined to be finished.

3. The multi-objective path planning method for industrial robots based on quantum annealing algorithm according to claim 2, characterized in that: After the step of randomly flipping the states of some quantum bits during the iterative training process and calculating the quantum system energy in the new state, the method further includes: If the new energy is not lower than the original energy, the energy acceptance probability is calculated and tested to see if it is greater than the random number; If the energy acceptance probability is greater than the random number, the new state is accepted; If the energy acceptance probability is not greater than the random number, the new state is not accepted; The energy acceptance probability is calculated as follows: exp(―ΔE / (kT)); ΔE is the energy difference between the new and old energies, k is the Boltzmann constant, and T is the annealing temperature of the current round.

4. The multi-objective path planning method for industrial robots based on quantum annealing algorithm according to claim 1, characterized in that: The steps of constructing the quantum Hamiltonian include: Calculate the sub-Hamiltonian corresponding to each optimization objective; The weighted sum of each sub-Hamiltonian is performed to generate the quantum Hamiltonian.

5. The multi-objective path planning method for industrial robots based on quantum annealing algorithm according to claim 4, characterized in that: The weight relationship between each sub-Hamiltonian is: δ<β<α<γ; Among them, δ is the weight of the sub-Hamiltonian corresponding to the task priority, β is the weight of the sub-Hamiltonian corresponding to the energy consumption, α is the weight of the sub-Hamiltonian corresponding to the path length, and γ is the weight of the sub-Hamiltonian corresponding to the obstacle avoidance safety.

6. The multi-objective path planning method for industrial robots based on quantum annealing algorithm according to claim 1, characterized in that: The multi-objective path planning method for industrial robots based on the quantum annealing algorithm also includes: After decoding the final path, the system checks whether there is a risk of collision with obstacles based on the workshop map and obstacle information. If there is no risk, the final path is extracted.

7. The multi-objective path planning method for industrial robots based on quantum annealing algorithm according to claim 1, characterized in that: The step of performing smooth optimization processing on the extracted final path includes: The Bezier curve is used to perform smooth optimization on the extracted final path.

8. A control system, characterized in that: include: A data acquisition module is used to obtain environmental information of the industrial site when the industrial robot is running; wherein the environmental information includes the robot's initial position and target position, as well as the position and shape of obstacles; The path planning module is used to plan paths based on environmental information and construct a quantum Hamiltonian based on the optimization objectives of the path planning problem. These optimization objectives include path length, energy consumption, obstacle avoidance safety, and task priority. Quantum bits are used to encode each optional path in the path planning space and initialized to a uniform superposition state. A quantum annealing algorithm is used to gradually reduce the energy of the quantum system and search for the optimal solution of the quantum bits. After the quantum annealing is completed, the state of the quantum bits is measured to obtain the encoding of the final path, which is then decoded and extracted. The post-processing module is used to perform smooth optimization processing on the extracted final path; The robot control module is used to generate motion instructions for the industrial robot based on the optimized final path.

9. An industrial robot, characterized in that: The industrial robot includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the steps of the multi-objective path planning method for an industrial robot based on a quantum annealing algorithm are implemented as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the multi-objective path planning method for an industrial robot based on a quantum annealing algorithm according to any one of claims 1 to 7.

Images