Autonomous navigation inspection robot path planning method based on quantum particle swarm optimization
By using quantum circuits to generate particle position information in the autonomous navigation inspection robot path planning, and combining with the Hamming distance algorithm, the problem that the quantum particle swarm algorithm is easily trapped in local optimal solutions, and more efficient path planning is achieved.
Patent Information
- Application Number
- CN202510260421.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-06
- Publication Date
- 2025-06-20
AI Technical Summary
Existing quantum particle swarm algorithms are prone to fall into local optimal solutions in the path planning of autonomous navigation and patrol robots, and their application range is limited, so the problem of local optimal cannot be effectively avoided.
The particle swarm algorithm based on quantum circuit is used to randomly generate particle position information through quantum circuits, and the search process is improved using quantum gates and auxiliary inversion bits. Combined with the Hanming distance algorithm, the invalid solution is converted into an effective solution.
It effectively overcomes the problem of local optimal solution, expands the application scope of quantum particle swarm algorithm, and improves the efficiency and effect of path planning.
Smart Images
Figure CN120176671A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of path planning, and more precisely, it relates to a path planning method for an autonomous navigation inspection robot based on a quantum particle swarm algorithm. Background Art
[0002] In the path planning of an autonomous navigation inspection robot, the particle swarm algorithm can be used to optimize the patrol route of the robot and improve the inspection efficiency of the robot. With the development of combinatorial optimization algorithms, the current quantum particle swarm algorithm has also been introduced into the path planning of autonomous navigation inspection robots. The quantum particle swarm algorithm is an algorithm that combines the respective advantages of the traditional particle swarm algorithm and quantum computing methods. After the quantum particle swarm algorithm was proposed, it has received extensive attention in the academic community and has been improved from multiple perspectives such as quantum-inspired algorithms to classical-quantum and full quantum computing.
[0003] Due to the introduction of quantum computing in the path planning of autonomous robots, some inherent defects of the traditional particle swarm algorithm have been significantly improved. For example, the generation of random numbers by quantum circuits is a generation of true random numbers compared to classical-structured computers, which improves the defect that the traditional ant colony algorithm is prone to falling into local optimal solutions. Nevertheless, the existing quantum particle swarm algorithms also have limitations. The most obvious defect is that the current mainstream quantum particle swarm algorithm is a quantum-inspired algorithm, which only uses the idea of quantum computing, but still uses a classical computer to run during actual calculations, thus greatly limiting the application scope of the current quantum ant colony algorithm and being unable to effectively avoid defects such as falling into local optima. Summary of the Invention
[0004] The purpose of the present invention is to propose a path planning method for an autonomous navigation inspection robot based on a quantum particle swarm algorithm in view of the deficiencies of the prior art.
[0005] In a first aspect, there is provided a path planning method for an autonomous navigation inspection robot based on a quantum particle swarm algorithm, including:
[0006] Step 1, define a search space, where the search space is used to represent the environment of the autonomous navigation inspection robot, and determine the positions of the starting point, the ending point, and the obstacles, and then determine the number of bits required for the quantum circuit according to the required solution accuracy;
[0007] Step 2, initialize the particle swarm. Each particle in the particle swarm represents a candidate path from the starting point to the ending point. Use a quantum circuit to randomly generate the position information of each particle, ensuring that the starting point and the ending point are fixed, and the intermediate nodes of the path can be freely adjusted;
[0008] Step 3, design a fitness function, and evaluate the quality of the particle's path according to the fitness function, and record the individual optimal solution and the global optimal solution;
[0009] Step 4: Use the fitness function and iteratively update the particles through a quantum circuit, use the Hamming distance algorithm to transform invalid solutions into the valid solutions closest to them, recalculate the fitness values, and record the individual optimal solutions and the global optimal solutions;
[0010] Step 5: Simultaneously select the maximum number of iterations and the change amplitude of the global optimal solution as the termination conditions. After reaching the termination conditions, output the current optimal path.
[0011] Preferably, in Step 2, the use of the quantum circuit to randomly generate the position information of each particle includes: using a quantum circuit including auxiliary inversion bits, qubit bits, and quantum gates to generate the position information. The auxiliary inversion bits are used to cause mutations in the problem solutions generated by the quantum circuit, the qubit bits are used to generate position coordinates, and the quantum gates are used to set weights for each qubit bit.
[0012] Preferably, in Step 2, the mathematical expression of the quantum gate is:
[0013]
[0014] where θ is the weight parameter, the value range of θ is [0, 1], and Ry represents the quantum gate.
[0015] Preferably, in Step 3, the fitness function is based on the path length. When the path collides with an obstacle, the fitness value is taken as 0, and interpolation is used to optimize the path to obtain a smoother path.
[0016] Preferably, it further includes:
[0017] Step 6: Conduct numerical experiments using combinatorial optimization problems; the numerical experiments include: embedding data into qubit bits in the way of amplitude embedding, and using a quantum computing script to implement distance calculation.
[0018] In a second aspect, a path planning system for an autonomous navigation inspection robot based on a quantum particle swarm algorithm is provided, which is used to execute any of the methods in the first aspect, and includes:
[0019] A definition module, which is used to define the search space. The search space is used to represent the environment of the autonomous navigation inspection robot, determine the positions of the starting point, the ending point, and the obstacles, and then determine the number of bits required for the quantum circuit according to the solution accuracy requirements;
[0020] An initialization module, which is used to initialize the particle swarm. Each particle in the particle swarm represents a candidate path from the starting point to the ending point. Use the quantum circuit to randomly generate the position information of each particle, ensure that the starting point and the ending point are fixed, and the intermediate nodes of the path can be freely adjusted;
[0021] A design module, which is used to design a fitness function, evaluate the quality of the paths of particles according to the fitness function, and record the individual optimal solution and the global optimal solution;
[0022] An update module, which is used to use the fitness function to iteratively update the particles through a quantum circuit, use the Hamming distance algorithm to convert invalid solutions into valid solutions closest to them, recalculate the fitness values, and record the individual optimal solution and the global optimal solution;
[0023] An output module, which is used to select both the maximum number of iterations and the change range of the global optimal solution as termination conditions, and output the current optimal path after reaching the termination conditions.
[0024] In a third aspect, a computer storage medium is provided. A computer program is stored in the computer storage medium; when the computer program runs on a computer, the computer is enabled to execute the method according to any one of the first aspect.
[0025] In a fourth aspect, an electronic device is provided, including:
[0026] A memory, which is used to store a computer program;
[0027] A processor, which is used to execute the computer program to implement the method according to any one of the first aspect.
[0028] The beneficial effects of the present invention are:
[0029] 1. The present invention uses a real quantum circuit to improve the quantum particle swarm algorithm in the path planning problem of autonomous navigation inspection robots, overcoming the problem that the current quantum particle swarm only adopts the idea of quantum computing without introducing a real quantum circuit for calculation, so that the method is prone to falling into a local optimal solution.
[0030] 2. Based on the actual situation of the current quantum computer, the present invention uses a small number of quantum bits to generate the position information of different individuals in the particle swarm, greatly improving the scope of use of the current quantum computer, and can accelerate the speed and effect of data classification.
[0031] 3. The present invention uses quantum computing to improve the Hamming Distance algorithm, which can accelerate the process of converting an unreasonable solution into a solution that meets the requirements. Description of the Drawings
[0032] Figure 1 It is a flowchart of the path planning method for an autonomous navigation inspection robot based on the quantum particle swarm algorithm provided by this application;
[0033] Figure 2 It is a schematic diagram of a quantum circuit for solving path traversal provided by this application;
[0034] Figure 3a Schematic diagram of the quantum Hamming-Distance circuit provided for this application;
[0035] Figure 3b inc provided for this application n Schematic diagram of the specific composition of the gate circuit;
[0036] Figure 4 Schematic diagram of the quantum circuit for using amplitude to embed traditional data provided for this application;
[0037] Figure 5 Schematic diagram of the traversal result of 52 cities provided for this application. Specific implementation manner
[0038] The present invention will be further described below in conjunction with embodiments. The description of the following embodiments is only used to help understand the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several modifications can still be made to the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.
[0039] Embodiment 1:
[0040] To solve the defects of the quantum particle swarm optimization algorithm (QPSO) in the path planning process of autonomous navigation inspection robots, Embodiment 1 of this application provides a path planning method for autonomous navigation inspection robots based on the quantum particle swarm optimization algorithm to overcome the problems of the existing algorithm not using a quantum computer and having fewer quantum bits. In QPSO, each particle represents a candidate path from the starting point to the ending point. Specifically, the path is usually represented by a set of coordinate sequences or nodes, that is, the key points between obstacles. The task of the particle searching in the solution space is to continuously adjust the positions of these nodes to make the path more optimized. In this process, this application introduces a quantum circuit to generate effective solutions, effectively reducing the possibility that the algorithm falls into a local optimal solution during the search process, and further expanding the application scope of the quantum ant colony algorithm in the problem of autonomous robot path planning.
[0041] Specifically, as Figure 1 shown, the method provided in Embodiment 1 of this application includes:
[0042] Step 1, define the search space, where the search space is used to represent the environment of the autonomous navigation inspection robot, and determine the positions of the starting point, the ending point, and the obstacles, and then determine the number of bits required for the quantum circuit according to the solution accuracy requirement.
[0043] Step 2, initialize the particle swarm. Each particle in the particle swarm represents a candidate path from the starting point to the ending point. Use the quantum circuit to randomly generate the position information (coordinates of the path nodes) of each particle, ensuring that the starting point and the ending point are fixed, and the intermediate nodes of the path can be freely adjusted.
[0044] Specifically, Figure 2 is a schematic diagram for solving a quantum circuit by path traversal, where q is a qubit, Ry is a quantum gate, and C0 is the result after measurement. In Figure 2 the qubit q0 is an auxiliary inversion bit, whose function is to make the problem solution generated by the quantum circuit mutate under certain conditions, thereby improving the defect that the ant colony algorithm falls into a local optimal solution.
[0045] Figure 2 In q10 - q1 n are n qubit bits. This circuit is used to generate the position information on the navigation map of the autonomous inspection robot. The number of qubit bits n is determined by the required accuracy. Each time a position coordinate is generated, thus constituting the position information of an individual in the particle swarm. Similar to the heuristic quantum particle swarm algorithm, in the quantum algorithm, it is necessary to generate the global optimal value and the local optimal value that guide the particle swarm to search for the problem. In Figure 2 the given quantum circuit, the Ry gate is used to set weights for each qubit bit. Thus, the Ry quantum gate circuit corresponding to n qubit bits generates a new solution close to the global optimal solution. Among them, the mathematical expression of the Ry quantum gate is:
[0046]
[0047] where the value range of θ is [0, 1]. By making corresponding assignments to it, the Figure 2 weights of the position information generated by the quantum circuit in can be changed. This parameter can enable the Ry gate to implement the function of pheromone update in the traditional ant colony algorithm.
[0048] Step 3: Design a fitness function, evaluate the quality of the particle's path according to the fitness function, and record the individual optimal solution and the global optimal solution.
[0049] Specifically, consider factors such as path length, whether to collide with obstacles, and path smoothness to design the fitness function. The fitness function is based on the path length and takes 0 when encountering an obstacle. After obtaining the current path, interpolation method is used for optimization to obtain a smoother path. Evaluate the quality of the particle's path according to the fitness function, and record the individual optimal solution and the global optimal solution of each particle.
[0050] Step 4: Use the fitness function (i.e., the quantum position update formula of the quantum particle swarm algorithm) and iteratively update the particles through the quantum circuit, and use the Hamming - Distance algorithm to convert the invalid solution into the valid solution closest to it, and recalculate the fitness value, and record the individual optimal solution and the global optimal solution.
[0051] Specifically, as Figure 1As shown, the quantum circuit generates possible movement trajectories, calculates the path length, and updates the optimal path if it is shorter than the global optimal path. In addition, the fitness value is calculated based on the fitness function in step 3, combined with the individual optimal solution and the global optimal solution values at the current step.
[0052] Step 5: At the same time, select the maximum number of iterations and the change range of the global optimal solution as the termination conditions. After reaching the termination conditions, output the current optimal path.
[0053] Embodiment 2:
[0054] Based on Embodiment 1, Embodiment 2 of the present application provides a more specific path planning method for an autonomous navigation inspection robot based on the quantum particle swarm algorithm, including:
[0055] Step 1: Define the search space, which is used to represent the environment of the autonomous navigation inspection robot, determine the positions of the starting point, the ending point, and the obstacles, and then determine the number of bits required for the quantum circuit according to the solution accuracy requirements.
[0056] Step 2: Initialize the particle swarm. Each particle in the particle swarm represents a candidate path from the starting point to the ending point. Use the quantum circuit to randomly generate the position information of each particle, ensuring that the starting point and the ending point are fixed, and the intermediate nodes of the path can be freely adjusted.
[0057] Step 3: Design a fitness function, and evaluate the quality of the particle's path according to the fitness function, and record the individual optimal solution and the global optimal solution.
[0058] Step 4: Use the fitness function and iterate and update the particles through the quantum circuit, and use the Hamming distance algorithm to convert invalid solutions into the closest valid solutions, and recalculate the fitness value, recording the individual optimal solution and the global optimal solution.
[0059] Due to the randomness of the measurement results of quantum computing, when obtaining the problem results generated by the quantum circuit, especially in the initial stage of path search, when the number of runs is relatively small and the probability differences of each identification point are relatively small, solutions that do not meet the path planning requirements may occur. For example, the position information of an individual in the particle swarm does not meet the requirements of the problem to be processed. To solve this problem, the present invention uses the quantum Hamming-Distance method to correct the non-conforming solutions, so as to quickly generate reasonable solutions, thereby improving the efficiency of pheromone update in the initial operation of the program. The quantum circuit is as Figure 3a and Figure 3b shown.
[0060] In Figure 3a R X stores the binary X variable, R VStore the binary V variable, R D Store the operation result of Hamming Distance, r ij Represent the binary bit v i And x j According to the comparison result of Hamming Distance, d ij Represent the reasonable result after being corrected by this algorithm. The custom inc in the figure n The specific structure of the gate circuit is as Figure 3b shown. Compared with the complexity O(M) of the traditional Hamming Distance algorithm, the complexity of the quantum Hamming Distance is O(log2M), and the computational complexity has been significantly improved. Among them, M is the number of Vs.
[0061] Step 5: At the same time, select the maximum number of iterations and the change range of the global optimal solution as the termination conditions. After reaching the termination conditions, output the current optimal path.
[0062] Step 6: In order to verify the effectiveness of this method, use the combinatorial optimization problem for numerical experiments.
[0063] Specifically, for the points in the calculation dataset, the present invention uses the amplitude loading method to embed the traditional data into the qubit bits and uses the initialization feature in the IBM quantum computing script Qiskit to implement the distance calculation. The quantum circuit for specifically calculating the distance is as Figure 4 shown. Among them, the traditional data refers to the corresponding data in the classical particle swarm algorithm.
[0064] In addition, in order to prove the improvement of the present invention on the quantum particle swarm algorithm, use the standard dataset "berlin52" of the traveling salesman problem for testing. This dataset contains 52 points, and the traversal result is as Figure 5 shown.
[0065] At the same time, use the search result of the quantum particle swarm algorithm without using the quantum circuit as a comparison. The specific parameter settings are that the number of particles is 50, the number of iterations is 1000 times, the quantum parameter is 0.5, and the convergence criterion tolerance is 10 -6 , calculate using the same computer, run 10 times respectively, and take the best results of the two methods for comparison. The results show that the optimal path distance of the quantum particle swarm algorithm without introducing the quantum circuit is 27426.37, and the result of using the improved quantum particle swarm algorithm is 9158.71. It can be seen from the results that the result of the improved quantum particle swarm algorithm is significantly better than that of the traditional particle swarm algorithm.
[0066] It should be noted that the same or similar parts in this embodiment and Embodiment 1 can be referred to each other and will not be elaborated in this application.
[0067] Example 3:
[0068] Based on Example 1, Example 3 of the present application provides a path planning system for an autonomous navigation inspection robot based on the quantum particle swarm algorithm, including:
[0069] A definition module, used to define the search space, where the search space is used to represent the environment of the autonomous navigation inspection robot, determine the positions of the starting point, the ending point, and the obstacles, and then determine the number of bits required for the quantum circuit according to the solution accuracy requirement;
[0070] An initialization module, used to initialize the particle swarm. Each particle in the particle swarm represents a candidate path from the starting point to the ending point. The position information of each particle is randomly generated using a quantum circuit, ensuring that the starting point and the ending point are fixed, and the intermediate nodes of the path can be freely adjusted;
[0071] A design module, used to design a fitness function, evaluate the quality of the particle paths according to the fitness function, and record the individual optimal solution and the global optimal solution;
[0072] An update module, used to use the fitness function and iteratively update the particles through a quantum circuit, and use the Hamming distance algorithm to convert invalid solutions into the valid solutions closest to them, recalculate the fitness values, and record the individual optimal solution and the global optimal solution;
[0073] An output module, used to simultaneously select the maximum number of iterations and the change range of the global optimal solution as the termination conditions, and output the current optimal path after reaching the termination conditions.
[0074] It should be noted that the system provided in this embodiment is the system corresponding to the method provided in Example 1. Therefore, for the parts that are the same or similar in this embodiment and Example 1, reference can be made to each other, and details will not be repeated in this application.
[0075] In summary, in the path planning of the autonomous navigation inspection robot of the present application, by introducing quantum behavior, particles are allowed to "jump" randomly in the solution space, thereby increasing the global search ability of the algorithm. At the same time, in order to avoid solutions that do not meet the requirements generated during the measurement of the quantum circuit, the Hamming-Distance algorithm is introduced to generate the valid solution closest to the defective solution on the basis of the defective solution, improving the operation efficiency of the algorithm. Due to the problem of few quantum bits in the current quantum circuit, the quantum circuit is combined with the classical computer memory to store the solutions generated by the quantum circuit each time into the classical register, and together they form a valid solution that can be used for the problem to be calculated.
Claims
1. A path planning method for an autonomous navigation inspection robot based on a quantum particle swarm algorithm, characterized in that: include: Step 1: define a search space, which is used to represent the environment of the autonomous navigation inspection robot, and determine the positions of the starting point, end point and obstacles, and then determine the number of bits required for the quantum circuit according to the solution accuracy requirements; Step 2: Initialize a particle swarm, where each particle of the particle swarm represents a candidate path from the starting point to the end point. Use a quantum circuit to randomly generate the position information of each particle to ensure that the starting point and the end point are fixed, and the intermediate nodes of the path can be freely adjusted. Step 3: Design a fitness function, evaluate the quality of the particle path according to the fitness function, and record the individual optimal solution and the group optimal solution; Step 4: Use the fitness function and quantum circuit to iteratively update the particles, use the Hamming distance algorithm to convert the invalid solution into the closest valid solution, recalculate the fitness value, and record the individual optimal solution and the group optimal solution; Step 5: Select the maximum number of iterations and the change range of the global optimal solution as the termination conditions, and output the current optimal path after the termination conditions are met.
2. The path planning method for an autonomous navigation inspection robot based on a quantum particle swarm algorithm according to claim 1 is characterized in that: In step 2, the use of a quantum circuit to randomly generate position information of each particle includes: using a quantum circuit including an auxiliary inversion bit, a quantum bit, and a quantum gate to generate position information, the auxiliary inversion bit is used to cause a mutation in the problem solution generated by the quantum circuit, the quantum bit is used to generate position coordinates, and the quantum gate is used to set a weight for each quantum bit.
3. The path planning method for an autonomous navigation inspection robot based on a quantum particle swarm algorithm according to claim 2 is characterized in that: In step 2, the mathematical expression of the quantum gate is: Among them, θ is the weight parameter, the value range of θ is [0,1], and Ry represents the quantum gate.
4. The path planning method for an autonomous navigation inspection robot based on a quantum particle swarm algorithm according to claim 3 is characterized in that: In step 3, the fitness function is based on the path length, the fitness value is 0 when the path collides with an obstacle, and the path is optimized using an interpolation method to obtain a smoother path.
5. The path planning method for an autonomous navigation inspection robot based on a quantum particle swarm algorithm according to claim 4 is characterized in that: Also includes: Step 6: Conduct numerical experiments using combinatorial optimization problems; The numerical experiment includes: using amplitude embedding to embed data into quantum bits, and using quantum computing scripts to implement distance calculations.
6. The autonomous navigation inspection robot path planning system based on quantum particle swarm algorithm is characterized by: The method for executing any one of claims 1 to 5 comprises: A definition module is used to define a search space, which is used to represent the environment of the autonomous navigation inspection robot and determine the positions of the starting point, the end point and obstacles, and then determine the number of bits required for the quantum circuit according to the solution accuracy requirements; An initialization module is used to initialize a particle swarm. Each particle in the particle swarm represents a candidate path from the starting point to the end point. The position information of each particle is randomly generated using a quantum circuit to ensure that the starting point and the end point are fixed, and the intermediate nodes of the path can be freely adjusted. A design module is used to design a fitness function, evaluate the quality of the particle path according to the fitness function, and record the individual optimal solution and the group optimal solution; An updating module, used to iteratively update the particles using the fitness function and quantum circuits, and to convert invalid solutions into valid solutions closest to them using a Hamming distance algorithm, and to recalculate the fitness values, and record the individual optimal solution and the group optimal solution; The output module is used to simultaneously select the maximum number of iterations and the change range of the global optimal solution as termination conditions, and output the current optimal path after the termination conditions are met.
7. A computer storage medium, characterized in that: The computer storage medium stores a computer program; when the computer program is executed on a computer, the computer executes any one of the methods described in claims 1 to 5.
8. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the method according to any one of claims 1 to 5.