Unmanned aerial vehicle obstacle avoidance path planning method suitable for closed space fire

Abstract

The invention relates to path planning, in particular to an unmanned aerial vehicle obstacle avoidance path planning method suitable for a closed space fire, and the method comprises the steps: converting a three-dimensional structure, path nodes and paths in a closed space into corresponding mathematical models; designing a target function of unmanned aerial vehicle obstacle avoidance path planning; designing constraint conditions of unmanned aerial vehicle obstacle avoidance path planning; constructing an unmanned aerial vehicle obstacle avoidance path planning model according to the target function and the constraint condition of unmanned aerial vehicle obstacle avoidance path planning; solving the unmanned aerial vehicle obstacle avoidance path planning model by adopting a quantum shift green bird obstacle avoidance path optimization algorithm to obtain an unmanned aerial vehicle obstacle avoidance path planning scheme; according to the technical scheme provided by the invention, the defect that the obstacle avoidance path of the unmanned aerial vehicle is difficult to plan efficiently and accurately in a closed space fire environment in the prior art can be effectively overcome.

Description

Technical Field

[0001] This invention relates to path planning, and more specifically to a method for obstacle avoidance path planning for unmanned aerial vehicles (UAVs) applicable to fires in confined spaces. Background Technology

[0002] In confined space fire scenarios, the complex spatial structure, numerous obstacles, and harsh environmental factors such as smoke and high temperatures pose significant challenges to traditional rescue and detection methods. Drones, with their high flexibility and ability to access hazardous areas, have become an important tool for fire rescue and detection in confined spaces. However, the key to achieving efficient drone operations in confined space fire environments lies in planning safe and effective obstacle avoidance paths for them.

[0003] Existing UAV obstacle avoidance path planning methods have many limitations when facing complex environments. On the one hand, traditional methods struggle to accurately convert complex three-dimensional structures within confined spaces into mathematical models, resulting in planned paths that fail to accurately reflect the actual environment. On the other hand, the design of objective functions and constraints during path planning is inadequate, lacking comprehensive consideration of various limiting factors for UAVs in confined fire environments, such as collision risk and energy consumption.

[0004] In terms of algorithm application, common path planning algorithms are insufficient in both global search and local exploration capabilities. Some algorithms rely on pseudo-random number generation perturbations during the global exploration phase, resulting in insufficient randomness and breadth of the search, making it difficult to cover the entire feasible path region of the enclosed space and easily leading to local optima. During the local exploration phase, when encountering complex obstacles, there is a lack of effective mechanisms to allow path nodes to break through the limitations imposed by obstacles in existing algorithms, "escape" from the currently obstacle-bound local area, and explore potentially better paths on the other side of the obstacle. Furthermore, it is difficult to quickly find the optimal path, failing to meet the high demands for timeliness and accuracy in enclosed space fire rescue. Therefore, there is an urgent need for a drone obstacle avoidance path planning method suitable for enclosed space fires to solve the above problems. Summary of the Invention

[0005] (a) Technical problems to be solved In view of the above-mentioned shortcomings of the existing technology, the present invention provides a drone obstacle avoidance path planning method applicable to fires in confined spaces, which can effectively overcome the shortcomings of the existing technology in that it is difficult to plan drone obstacle avoidance paths efficiently and accurately in fire environments in confined spaces.

[0006] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: A method for planning obstacle avoidance paths for drones applicable to fires in confined spaces includes the following steps: S1. Transform the three-dimensional structure, path nodes, and path within the enclosed space into a corresponding mathematical model; S2. Design the objective function for drone obstacle avoidance path planning; S3. Design the constraints for the obstacle avoidance path planning of the UAV; S4. Based on the objective function and constraints of the UAV obstacle avoidance path planning, construct the UAV obstacle avoidance path planning model; S5. The obstacle avoidance path optimization algorithm of Quantum Kingfisher is used to solve the obstacle avoidance path planning model of UAV and obtain the obstacle avoidance path planning scheme of UAV. In the quantum kingfisher obstacle avoidance path optimization algorithm: In the global exploration phase, quantum noise is introduced to replace pseudo-random number generation of quantum perturbations, and quantum entanglement simulation is performed to simulate the perching / hovering behavior of the kingfisher, thereby enhancing the global search capability. In the local development phase, the quantum tunneling effect is introduced, allowing path nodes to traverse obstacles with quantum tunneling probability. Combined with quantum gradient descent, the path is guided to converge toward the optimal direction to simulate the diving behavior of the kingfisher. At the same time, the symbiotic behavior of the kingfisher is simulated through quantum state superposition, achieving local escape and fine optimization.

[0007] Preferably, in S1, the three-dimensional structure, path nodes, and path within the enclosed space are transformed into corresponding mathematical models, including: S11. Divide the enclosed space into a three-dimensional grid, labeling each grid cell as 0 or 1, where 0 represents an obstacle and 1 represents passage. Generate a binary matrix. , Let L be the dimension of the binary matrix E, and let L, W, and H be the length, width, and height of the enclosed space, respectively. S12. For each path node, the corresponding three-dimensional coordinates are encoded using 2 qubits: ; in, p is the i-th path node in the path i The quantum state corresponding to the three-dimensional coordinates, , , , These are the ground states of quantum computing. , , , The probability amplitude represents the weight of each quantum computing ground state; S13. For each path, it is represented as a superposition of quantum states: ; in, Let P be the quantum state corresponding to path P. Let N represent the tensor product, where N is the number of path nodes in path P.

[0008] Preferably, the objective function for planning the obstacle avoidance path of the UAV in S2 includes: The objective function for drone obstacle avoidance path planning is expressed by the following formula: ; Where F(P) is the fitness value of path P. , , All are weighting coefficients, and Length(P) is the path length of path P: ; For adjacent path nodes p in path P i p i+1 The L2 norm between adjacent path nodes p i p i+1 The Euclidean distance between them; Risk(P) represents the collision risk of path P: ; p is the i-th path node in path P. i With the j-th obstacle O j The L2 norm between them, i.e., the path node p i With obstacle O j The Euclidean distance between them, d th M represents the number of obstacles, which is the preset safe distance threshold. Energy(P) represents the energy consumption of path P. ; For adjacent path nodes p in path P i p i+1 The radians of the turning angle between them, where a and b are both energy consumption coefficients.

[0009] Preferably, the constraints for designing the obstacle avoidance path planning for the UAV in S3 include: S31. Determine the maximum turning angle in radians between adjacent path nodes in the path. : ; S32. Determine the shortest flight path length d that the UAV needs to fly straight before changing attitude. min : ; S33. Transform the constraints in S31 and S32 into phase penalties for quantum states, and combine this with the objective function of UAV obstacle avoidance path planning to obtain the quantum penalty objective function: ; Among them, F penalty (P) represents the final fitness value of path P, g i (P) represents the i-th constraint on path P, where i=1. When i=2, , where n is the number of constraints on path P. This is the penalty coefficient.

[0010] Preferably, in S5, the Quantum Kingfisher obstacle avoidance path optimization algorithm is used to solve the UAV obstacle avoidance path planning model to obtain the UAV obstacle avoidance path planning scheme, including: The quantum penalty objective function of UAV obstacle avoidance path planning is solved by using the quantum kingfisher obstacle avoidance path optimization algorithm, and a UAV obstacle avoidance path planning scheme that satisfies all constraints is obtained.

[0011] Preferably, the step of using the quantum kingfisher obstacle avoidance path optimization algorithm to solve the quantum penalty objective function of the UAV obstacle avoidance path planning, and obtaining a UAV obstacle avoidance path planning scheme that satisfies all constraints, includes: S51. Randomly generate an initial population of a certain size within the search space. Each kingfisher in the initial population represents a potential solution. The quantum state of each kingfisher is represented by the superposition of quantum states. Combining quantum randomness enhances diversity and avoids getting trapped in local optima. S52. In the global exploration phase, quantum noise is introduced to replace pseudo-random number generation of quantum perturbation, and quantum entanglement simulation is performed to simulate the perching / hovering behavior of the kingfisher and enhance the global search capability. S53. In the local development phase, the quantum tunneling effect is introduced, allowing path nodes to traverse obstacles with quantum tunneling probability, and combined with quantum gradient descent to guide the path to converge in the optimal direction, in order to simulate the diving behavior of the kingfisher. S54. Furthermore, in the local development stage, the symbiotic behavior of the kingfisher is simulated by quantum state superposition to achieve local escape and fine optimization; S55. Observe the current population, collapse the quantum state of the kingfisher individual into a classical solution, and use the quantum penalty objective function of the UAV obstacle avoidance path planning to calculate the final fitness value of the kingfisher individual, record and update the global optimal solution. S56. Determine whether the iteration termination condition is met. If the iteration termination condition is not met, return to S52. Otherwise, output the current global optimal solution as the UAV obstacle avoidance path planning scheme that satisfies all constraints.

[0012] Preferably, in S52, during the global exploration phase, quantum noise is introduced to replace pseudo-random number generation of quantum perturbations, and quantum entanglement simulation is performed to simulate the perching / hovering behavior of the kingfisher, thereby enhancing the global search capability, including: S521. Introduce quantum noise to replace pseudo-random number generation quantum perturbation: ; in, , Let p be the i-th path node in the path represented by the quantum state of individual kingfisher k. k,i The position at the t-th and t+1-th iterations The i-th path node p in the path represented by the quantum state of a randomly selected kingfisher individual l. l,i At the position in the t-th iteration, T is a dynamically adjusted parameter that decreases non-linearly with the iteration number t: ; T max Where is the maximum number of iterations, and BF is the jump factor. The control parameter is used to adjust search compensation and balance global exploration and local exploitation capabilities. QuantumNoise is the quantum noise term, generated by a quantum random number generator. The quantum noise intensity is used to control the degree of influence of quantum perturbations on path planning. S522. Treat adjacent path nodes as quantum entangled pairs and maintain population diversity through quantum entanglement operations.

[0013] Preferably, in S53, during the local development phase, a quantum tunneling effect is introduced, allowing path nodes to traverse obstacles with a quantum tunneling probability. This is combined with quantum gradient descent to guide the path towards the optimal direction, simulating the diving behavior of a kingfisher, including: S531. Introducing the quantum tunneling effect, allowing path nodes to traverse obstacles with quantum tunneling probability: ; Among them, P k,i Let p be the i-th path node in the path represented by the quantum state of individual kingfisher k. k,i The quantum tunneling probability, i.e., the path node p k,i The probability of passing through the nearest obstacle, where ObstacleDistance is the path node p. k,i Euclidean distance to the nearest obstacle The tunneling width is used to control the decay rate of the quantum tunneling probability; S532. Combining quantum gradient descent to guide the path towards the optimal direction: ; in, The quantum penalty objective function for obstacle avoidance path planning for UAVs is defined in the path represented by the quantum state of individual kingfisher k at the i-th path node p. k,i The gradient at position t during the t-th iteration. The learning rate controls the step size for each gradient update.

[0014] Preferably, in S54, further, during the local development phase, the symbiotic behavior of the kingfisher is simulated through quantum state superposition to achieve local escape and fine optimization, including: Simulating the symbiotic behavior of kingfishers through quantum state superposition: ; Where, p new Let p1 and p2 be the positions of the new path nodes, where p1 and p2 are the positions of two randomly selected path nodes from the current path, respectively. Let be the phase factor, and let be the rotation operator on the complex unit circle: , representing the phase shift of path node p2 in the quantum superposition state, where i is the imaginary unit. is the phase angle, used to control the relative phase between path nodes p1 and p2 in a quantum superposition state.

[0015] (III) Beneficial Effects Compared with existing technologies, the UAV obstacle avoidance path planning method for fires in confined spaces provided by this invention has the following advantages: 1) Improve the quality of path planning and achieve refined optimization. In the local development phase, the quantum tunneling effect is introduced, allowing path nodes to traverse obstacles with quantum tunneling probability. This breaks the absolute restriction of obstacles on the path in traditional path planning algorithms. At the same time, combined with quantum gradient descent to guide the path to converge in the optimal direction, the path can be fine-tuned based on precise gradient information, so that the path can be continuously improved in a better direction. In addition, by simulating the symbiotic behavior of kingfishers through quantum state superposition, local escape and fine optimization are further realized, making the planned UAV obstacle avoidance path more reasonable and efficient. Under the premise of meeting obstacle avoidance requirements, the path length is shortened to the greatest extent and energy consumption is reduced, thus improving the quality of path planning. 2) Enhance search efficiency, taking into account both global and local factors. In the global exploration phase, quantum noise is introduced to replace pseudo-random numbers in generating quantum perturbations, and quantum entanglement simulation is performed to simulate the perching / hovering behavior of kingfishers. Quantum noise has unique randomness and uncertainty, and compared with traditional pseudo-random numbers, it can generate more diverse and exploratory perturbations, thereby enhancing the algorithm's global search capability, expanding the search range, and facilitating the discovery of the global optimum. In the local development phase, the combination of quantum tunneling effect and quantum gradient descent enables the algorithm to quickly converge to the optimum in local regions. This combination of global exploration and local development takes into account both the breadth and depth of the search, greatly improving the algorithm's search efficiency, and thus enabling the discovery of high-quality UAV obstacle avoidance paths in a short time. 3) Enhance environmental adaptability and effectively cope with complex enclosed spaces. The fire environment in confined spaces is complex and changeable, with numerous obstacles and uncertainties. This invention utilizes a quantum kingfisher obstacle avoidance path optimization algorithm, which simulates the behavior of a kingfisher using quantum mechanics principles. This makes the algorithm more flexible and adaptable. The quantum tunneling effect allows path nodes to traverse obstacles, and the superposition of quantum states enables local escape. These characteristics allow the algorithm to quickly adjust its path in complex environments, avoid obstacles, and find feasible routes. Even in the face of emergencies or environmental changes, the algorithm can react in a timely manner and replan its path, ensuring the safe and stable flight of UAVs in confined space fire environments. This improves the operational capabilities and reliability of UAVs in complex environments. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0017] Figure 1 This is a schematic diagram of the process of the present invention; Figure 2 This is a flowchart illustrating the process of using the Quantum Kingfisher Obstacle Avoidance Path Optimization Algorithm to solve the obstacle avoidance path planning scheme for UAVs in this invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0019] The following describes the specific process of the drone obstacle avoidance path planning method for fires in confined spaces provided by this invention, using specific examples (e.g.) Figure 1 (as shown) and technical effects.

[0020] 1. Transform the three-dimensional structure, path nodes, and paths within the enclosed space into corresponding mathematical models.

[0021] 1) Divide the enclosed space into a three-dimensional grid, with each grid cell labeled 0 or 1, where 0 represents an obstacle and 1 represents passage. Generate a binary matrix. , Let L be the dimension of the binary matrix E, and let L, W, and H be the length, width, and height of the enclosed space, respectively. 2) For each path node, the corresponding three-dimensional coordinates are encoded using 2 qubits: ; in, p is the i-th path node in the path i The quantum state corresponding to the three-dimensional coordinates, , , , These are the ground states of quantum computing. , , , The probability amplitude represents the weight of each quantum computing ground state; 3) For each path, it can be represented as a superposition of quantum states: ; in, Let P be the quantum state corresponding to path P. Let N represent the tensor product, where N is the number of path nodes in path P.

[0022] II. Design the objective function for UAV obstacle avoidance path planning.

[0023] The objective function for drone obstacle avoidance path planning is expressed by the following formula: ; Where F(P) is the fitness value of path P. , , All are weighting coefficients, and Length(P) is the path length of path P: ; For adjacent path nodes p in path P i pi+1 The L2 norm between adjacent path nodes p i p i+1 The Euclidean distance between them; Risk(P) represents the collision risk of path P: ; p is the i-th path node in path P. i With the j-th obstacle O j The L2 norm between them, i.e., the path node p i With obstacle O j The Euclidean distance between them, d th M represents the number of obstacles, which is the preset safe distance threshold. Energy(P) represents the energy consumption of path P. ; For adjacent path nodes p in path P i p i+1 The radians of the turning angle between them, where a and b are both energy consumption coefficients.

[0024] III. Constraints on the design of UAV obstacle avoidance path planning.

[0025] 1) Determine the maximum turning angle in radians between adjacent path nodes in the path. : ; 2) Determine the shortest flight path length d that the UAV needs to fly straight before changing attitude. min : ; 3) Transform the constraints in 1) and 2) into phase penalties of quantum states, and combine them with the objective function of UAV obstacle avoidance path planning to obtain the quantum penalty objective function: ; Among them, F penalty (P) represents the final fitness value of path P, g i (P) represents the i-th constraint on path P, where i=1. When i=2, , where n is the number of constraints on path P. This is the penalty coefficient.

[0026] IV. Based on the objective function and constraints of UAV obstacle avoidance path planning, construct a UAV obstacle avoidance path planning model.

[0027] Fifth, the obstacle avoidance path optimization algorithm of Quantum Kingfisher is used to solve the obstacle avoidance path planning model of UAV, and the obstacle avoidance path planning scheme of UAV is obtained.

[0028] The technical solution of this application, in the quantum kingfisher obstacle avoidance path optimization algorithm: In the global exploration phase, quantum noise is introduced to replace pseudo-random number generation of quantum perturbations, and quantum entanglement simulation is performed to simulate the perching / hovering behavior of the kingfisher, thereby enhancing the global search capability. In the local development phase, the quantum tunneling effect is introduced, allowing path nodes to traverse obstacles with quantum tunneling probability. Combined with quantum gradient descent, the path is guided to converge toward the optimal direction to simulate the diving behavior of the kingfisher. At the same time, the symbiotic behavior of the kingfisher is simulated through quantum state superposition, achieving local escape and fine optimization.

[0029] Step five specifically includes: The quantum kingfisher obstacle avoidance path optimization algorithm is used to solve the quantum penalty objective function of UAV obstacle avoidance path planning, resulting in UAV obstacle avoidance path planning schemes that satisfy various constraints, such as... Figure 2 As shown: 1) Randomly generate an initial population of a certain size within the search space. Each kingfisher in the initial population represents a potential solution. The quantum state of each kingfisher is represented by the superposition of quantum states. Combining quantum randomness enhances diversity and avoids getting trapped in local optima. 2) During the global exploration phase, quantum noise is introduced to replace pseudo-random number generation for quantum perturbation, and quantum entanglement simulation is performed to simulate the perching / hovering behavior of the kingfisher, thereby enhancing the global search capability: Introducing quantum noise to replace pseudo-random number generation of quantum perturbations: ; in, , Let p be the i-th path node in the path represented by the quantum state of individual kingfisher k. k,i The position at the t-th and t+1-th iterations The i-th path node p in the path represented by the quantum state of a randomly selected kingfisher individual l. l,i At the position in the t-th iteration, T is a dynamically adjusted parameter that decreases non-linearly with the iteration number t: ; T max Where is the maximum number of iterations, and BF is the jump factor. The control parameter is used to adjust search compensation and balance global exploration and local exploitation capabilities. QuantumNoise is the quantum noise term, generated by a quantum random number generator. The quantum noise intensity is used to control the degree of influence of quantum perturbations on path planning. Treat adjacent path nodes as quantum entangled pairs and maintain population diversity through quantum entanglement operations; 3) In the local development phase, the quantum tunneling effect is introduced, allowing path nodes to traverse obstacles with quantum tunneling probability. This is combined with quantum gradient descent to guide the path to converge in the optimal direction, simulating the diving behavior of a kingfisher. Introducing the quantum tunneling effect allows path nodes to traverse obstacles with quantum tunneling probability: ; Among them, P k,i Let p be the i-th path node in the path represented by the quantum state of individual kingfisher k. k,i The quantum tunneling probability, i.e., the path node p k,i The probability of passing through the nearest obstacle, where ObstacleDistance is the path node p. k,i Euclidean distance to the nearest obstacle The tunneling width is used to control the decay rate of the quantum tunneling probability; Combining quantum gradient descent to guide the path towards the optimal direction: ; in, The quantum penalty objective function for obstacle avoidance path planning for UAVs is defined in the path represented by the quantum state of individual kingfisher k at the i-th path node p. k,i The gradient at position t during the t-th iteration. The learning rate controls the step size for each gradient update. 4) Furthermore, in the local development stage, the symbiotic behavior of the kingfisher is simulated through quantum state superposition to achieve local escape and fine optimization: Simulating the symbiotic behavior of kingfishers through quantum state superposition: ; Where, p new Let p1 and p2 be the positions of the new path nodes, where p1 and p2 are the positions of two randomly selected path nodes from the current path, respectively. Let be the phase factor, and let be the rotation operator on the complex unit circle: , representing the phase shift of path node p2 in the quantum superposition state, where i is the imaginary unit. is the phase angle, used to control the relative phase between path nodes p1 and p2 in a quantum superposition state; 5) Observe the current population, collapse the quantum state of individual kingfishers into classical solutions, and use the quantum penalty objective function of UAV obstacle avoidance path planning to calculate the final fitness value of individual kingfishers, record and update the global optimal solution; 6) Determine whether the iteration termination condition is met. If the iteration termination condition is not met, return to step 2. Otherwise, output the current global optimal solution as the UAV obstacle avoidance path planning scheme that satisfies all constraints.

[0030] In the technical solution of this application, quantum tunneling effect is introduced in the local development stage, allowing path nodes to traverse obstacles with quantum tunneling probability. This breaks the absolute restriction of obstacles on the path in traditional path planning algorithms. At the same time, combined with quantum gradient descent to guide the path to converge in the optimal direction, the path can be fine-tuned based on accurate gradient information, so that the path can be continuously improved in a better direction. In addition, by simulating the symbiotic behavior of kingfishers through quantum state superposition, local escape and fine optimization are further realized, making the planned UAV obstacle avoidance path more reasonable and efficient. Under the premise of meeting obstacle avoidance requirements, the path length is shortened to the maximum extent, energy consumption is reduced, and the path planning quality is improved.

[0031] In the global exploration phase, quantum noise is introduced to replace pseudo-random numbers in generating quantum perturbations, and quantum entanglement simulation is performed to simulate the perching / hovering behavior of kingfishers. Quantum noise has unique randomness and uncertainty, and compared with traditional pseudo-random numbers, it can generate more diverse and exploratory perturbations, thereby enhancing the algorithm's global search capability, expanding the search range, and facilitating the discovery of the global optimum. In the local development phase, the combination of quantum tunneling effect and quantum gradient descent enables the algorithm to quickly converge to the optimum in local regions. This combination of global exploration and local development takes into account both the breadth and depth of the search, greatly improving the algorithm's search efficiency, and thus enabling the discovery of high-quality UAV obstacle avoidance paths in a short time.

[0032] The fire environment in confined spaces is complex and changeable, with numerous obstacles and uncertainties. This invention utilizes a quantum kingfisher obstacle avoidance path optimization algorithm, which simulates the behavior of a kingfisher using quantum mechanics principles. This makes the algorithm more flexible and adaptable. The quantum tunneling effect allows path nodes to traverse obstacles, and the superposition of quantum states enables local escape. These characteristics allow the algorithm to quickly adjust its path in complex environments, avoid obstacles, and find feasible routes. Even in the face of emergencies or environmental changes, the algorithm can react in a timely manner and replan its path, ensuring the safe and stable flight of UAVs in confined space fire environments. This improves the operational capabilities and reliability of UAVs in complex environments.

[0033] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for planning obstacle avoidance paths for unmanned aerial vehicles (UAVs) in confined space fires, characterized in that: Includes the following steps: S1. Transform the three-dimensional structure, path nodes, and path within the enclosed space into a corresponding mathematical model; S2. Design the objective function for drone obstacle avoidance path planning; S3. Design the constraints for the obstacle avoidance path planning of the UAV; S4. Based on the objective function and constraints of the UAV obstacle avoidance path planning, construct the UAV obstacle avoidance path planning model; S5. The obstacle avoidance path optimization algorithm of Quantum Kingfisher is used to solve the obstacle avoidance path planning model of UAV and obtain the obstacle avoidance path planning scheme of UAV. In the quantum kingfisher obstacle avoidance path optimization algorithm: In the global exploration phase, quantum noise is introduced to replace pseudo-random number generation of quantum perturbations, and quantum entanglement simulation is performed to simulate the perching / hovering behavior of the kingfisher, thereby enhancing the global search capability. In the local development phase, the quantum tunneling effect is introduced, allowing path nodes to traverse obstacles with quantum tunneling probability. Combined with quantum gradient descent, the path is guided to converge toward the optimal direction to simulate the diving behavior of the kingfisher. At the same time, the symbiotic behavior of the kingfisher is simulated through quantum state superposition, achieving local escape and fine optimization.

2. The UAV obstacle avoidance path planning method for fires in confined spaces according to claim 1, characterized in that: S1 transforms the three-dimensional structure, path nodes, and paths within the confined space into corresponding mathematical models, including: S11. Divide the enclosed space into a three-dimensional grid, labeling each grid cell as 0 or 1, where 0 represents an obstacle and 1 represents passage. Generate a binary matrix. , Let L be the dimension of the binary matrix E, and let L, W, and H be the length, width, and height of the enclosed space, respectively. S12. For each path node, the corresponding three-dimensional coordinates are encoded using 2 qubits: ； in, p is the i-th path node in the path i The quantum state corresponding to the three-dimensional coordinates, , , , These are the ground states of quantum computing. , , , The probability amplitude represents the weight of each quantum computing ground state; S13. For each path, it is represented as a superposition of quantum states: ； in, Let P be the quantum state corresponding to path P. Let N represent the tensor product, where N is the number of path nodes in path P.

3. The UAV obstacle avoidance path planning method for fires in confined spaces according to claim 2, characterized in that: The objective function for designing obstacle avoidance path planning for UAVs in S2 includes: The objective function for drone obstacle avoidance path planning is expressed by the following formula: ； Where F(P) is the fitness value of path P. , , All are weighting coefficients, and Length(P) is the path length of path P: ； For adjacent path nodes p in path P i p i+1 The L2 norm between adjacent path nodes p i p i+1 The Euclidean distance between them; Risk(P) represents the collision risk of path P: ； p is the i-th path node in path P. i With the j-th obstacle O j The L2 norm between them, i.e., the path node p i With obstacle O j The Euclidean distance between them, d th M represents the number of obstacles, which is the preset safe distance threshold. Energy(P) represents the energy consumption of path P. ； For adjacent path nodes p in path P i p i+1 The radians of the turning angle between them, where a and b are both energy consumption coefficients.

4. The UAV obstacle avoidance path planning method for fires in confined spaces according to claim 3, characterized in that: The constraints for designing obstacle avoidance path planning for UAVs in S3 include: S31. Determine the maximum turning angle in radians between adjacent path nodes in the path. : ； S32. Determine the shortest flight path length d that the UAV needs to fly straight before changing attitude. min : ； S33. Transform the constraints in S31 and S32 into phase penalties for quantum states, and combine this with the objective function of UAV obstacle avoidance path planning to obtain the quantum penalty objective function: ； Among them, F penalty (P) represents the final fitness value of path P, g i (P) represents the i-th constraint on path P, where i=1. When i=2, , where n is the number of constraints on path P. This is the penalty coefficient.

5. The UAV obstacle avoidance path planning method for fires in confined spaces according to claim 4, characterized in that: In S5, the Quantum Kingfisher obstacle avoidance path optimization algorithm is used to solve the UAV obstacle avoidance path planning model, resulting in UAV obstacle avoidance path planning schemes, including: The quantum penalty objective function of UAV obstacle avoidance path planning is solved by using the quantum kingfisher obstacle avoidance path optimization algorithm, and a UAV obstacle avoidance path planning scheme that satisfies all constraints is obtained.

6. The UAV obstacle avoidance path planning method for fires in confined spaces according to claim 5, characterized in that: The quantum kingfisher obstacle avoidance path optimization algorithm is used to solve the quantum penalty objective function of UAV obstacle avoidance path planning, resulting in UAV obstacle avoidance path planning schemes that satisfy various constraints, including: S51. Randomly generate an initial population of a certain size within the search space. Each kingfisher in the initial population represents a potential solution. The quantum state of each kingfisher is represented by the superposition of quantum states. Combining quantum randomness enhances diversity and avoids getting trapped in local optima. S52. In the global exploration phase, quantum noise is introduced to replace pseudo-random number generation of quantum perturbation, and quantum entanglement simulation is performed to simulate the perching / hovering behavior of the kingfisher and enhance the global search capability. S53. In the local development phase, the quantum tunneling effect is introduced, allowing path nodes to traverse obstacles with quantum tunneling probability, and combined with quantum gradient descent to guide the path to converge in the optimal direction, in order to simulate the diving behavior of the kingfisher. S54. Furthermore, in the local development stage, the symbiotic behavior of the kingfisher is simulated by quantum state superposition to achieve local escape and fine optimization; S55. Observe the current population, collapse the quantum state of the kingfisher individual into a classical solution, and use the quantum penalty objective function of the UAV obstacle avoidance path planning to calculate the final fitness value of the kingfisher individual, record and update the global optimal solution. S56. Determine whether the iteration termination condition is met. If the iteration termination condition is not met, return to S52. Otherwise, output the current global optimal solution as the UAV obstacle avoidance path planning scheme that satisfies all constraints.

7. The UAV obstacle avoidance path planning method for fires in confined spaces according to claim 6, characterized in that: In S52, during the global exploration phase, quantum noise is introduced to replace pseudo-random number generation for quantum perturbations, and quantum entanglement simulation is performed to simulate the perching / hovering behavior of the kingfisher, enhancing global search capabilities, including: S521. Introduce quantum noise to replace pseudo-random number generation quantum perturbation: ； in, , Let p be the i-th path node in the path represented by the quantum state of individual kingfisher k. k,i The position at the t-th and t+1-th iterations The i-th path node p in the path represented by the quantum state of a randomly selected kingfisher individual l. l,i At the position in the t-th iteration, T is a dynamically adjusted parameter that decreases non-linearly with the iteration number t: ； T max Where is the maximum number of iterations, and BF is the jump factor. The control parameter is used to adjust search compensation and balance global exploration and local exploitation capabilities. QuantumNoise is the quantum noise term, generated by a quantum random number generator. The quantum noise intensity is used to control the degree of influence of quantum perturbations on path planning. S522. Treat adjacent path nodes as quantum entangled pairs and maintain population diversity through quantum entanglement operations.

8. The UAV obstacle avoidance path planning method for fires in confined spaces according to claim 7, characterized in that: In S53, during the local development phase, the quantum tunneling effect is introduced, allowing path nodes to traverse obstacles with quantum tunneling probability. This is combined with quantum gradient descent to guide the path towards the optimal direction, simulating the diving behavior of a kingfisher, including: S531. Introducing the quantum tunneling effect, allowing path nodes to traverse obstacles with quantum tunneling probability: ； Among them, P k,i Let p be the i-th path node in the path represented by the quantum state of individual kingfisher k. k,i The quantum tunneling probability, i.e., the path node p k,i The probability of passing through the nearest obstacle, where ObstacleDistance is the path node p. k,i Euclidean distance to the nearest obstacle The tunneling width is used to control the decay rate of the quantum tunneling probability; S532. Combining quantum gradient descent to guide the path towards the optimal direction: ； in, The quantum penalty objective function for obstacle avoidance path planning for UAVs is defined in the path represented by the quantum state of individual kingfisher k at the i-th path node p. k,i The gradient at position t during the t-th iteration. The learning rate controls the step size for each gradient update.

9. The UAV obstacle avoidance path planning method for fires in confined spaces according to claim 8, characterized in that: Furthermore, in S54, during the local development phase, the symbiotic behavior of the kingfisher is simulated through quantum state superposition to achieve local escape and fine-grained optimization, including: Simulating the symbiotic behavior of kingfishers through quantum state superposition: ； Where, p new Let p1 and p2 be the positions of the new path nodes, where p1 and p2 are the positions of two randomly selected path nodes from the current path, respectively. Let be the phase factor, and let be the rotation operator on the complex unit circle: , representing the phase shift of path node p2 in the quantum superposition state, where i is the imaginary unit. is the phase angle, used to control the relative phase between path nodes p1 and p2 in a quantum superposition state.

Images