Path generation method of fire-fighting reconnaissance robot
By constructing a path cost function and dynamically updating the path point positions, the adaptability and stability issues of path planning in dynamic fire environments are solved, generating a highly adaptable, smooth path that can be executed by the robot.
Patent Information
- Application Number
- CN202511718373.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-21
- Publication Date
- 2026-02-17
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing path planning methods lack adaptability in dynamic fire environments, making it difficult to respond quickly to environmental changes, leading to path failures or unevenness, and failing to balance global optimality and execution stability.
A path cost function is constructed, which combines path smoothness, obstacle avoidance, and ambient temperature. By dynamically updating the path point positions, the path point sequence is optimized to adapt to dynamic environmental changes, and executability is verified and adjusted in real time.
It enables adaptive adjustment of the path in dynamic fire environment, ensuring that the path is smooth, safe and can be executed by the robot, thus improving the adaptability and robustness of path planning.
Smart Images
Figure CN121540151A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of intelligent robots, in particular to a path generation method of a fire reconnaissance robot. BACKGROUND
[0002] With the rapid development of robot technology, intelligent fire reconnaissance robots have gradually become an important tool in fire rescue scenarios. Robots can enter high-risk environments (such as dense smoke, high temperature and collapsed areas) to conduct fire reconnaissance and information collection, effectively reducing the life threat of rescue personnel and improving the timeliness and accuracy of fire information acquisition, thanks to their efficient environmental perception ability, flexible path planning and execution ability. The core task of a fire reconnaissance robot is to plan a path from the starting point to the target point according to the dynamic changes of the fire environment, and the path needs to meet the requirements of efficiency, safety and execution stability. However, the fire environment has the following characteristics, and traditional path planning methods face many challenges in practical application:
[0003] The fire environment usually contains a large number of static obstacles (such as buildings and furniture) and dynamic obstacles (such as collapsed objects and moving equipment), combined with the presence of high-temperature areas and smoke, which makes path planning need to be carried out in a complex environment model.
[0004] The distribution of obstacles and high-temperature areas in the fire will change in real time with the spread of the fire and external intervention, and path planning needs to quickly perceive dynamic changes and adjust the path.
[0005] Fire rescue requires a very high time requirement, and the path planning algorithm needs to generate an optimal path in a short time and can respond to environmental changes in real time to avoid delays that could cause serious consequences.
[0006] The hardware capabilities (such as speed and acceleration limits) of the fire reconnaissance robot pose an executable requirement for path planning, and the path planning needs to ensure that the generated path can be executed by the robot.
[0007] Under the above background, the existing path planning methods have the following limitations:
[0008] Traditional algorithms perform well in static environments, but lack adaptability to dynamic environmental changes, which can easily lead to path failure in actual fire scenes.
[0009] Some fast path generation algorithms have high computational efficiency in dynamic environments, but the generated path is not smooth, which can cause the robot to move unstably or even deviate from the path during execution.
[0010] The path generation method for dynamic fire environment generally lacks dynamic correction mechanism, and cannot quickly adjust the position of path points when the fire environment suddenly changes (such as the movement of obstacles or the spread of high-temperature areas), resulting in insufficient path adaptability and robustness.
[0011] Most traditional path planning methods use local optimization strategy, which cannot consider the global optimality of the path, especially in the multi-factor complex trade-off (such as smoothness, obstacle avoidance and environmental cost), which is easy to fall into local optimum.
[0012] Therefore, the path generation method of the fire reconnaissance robot is provided by the person skilled in the art to solve the above problems. SUMMARY
[0013] In view of the deficiencies of the prior art, the path generation method of the fire reconnaissance robot is provided to solve the problems in the background art.
[0014] To achieve the above purpose, the following technical solutions are adopted: a path generation method of a fire reconnaissance robot, comprising:
[0015] Step 1, according to the requirement of path generation of the fire reconnaissance robot, a path cost function is constructed, the path cost function is composed of multiple influence factors of the path, including the smoothness of the path, the stability of the path execution, the distance between the path point and the obstacle and the environmental temperature of the area through which the path passes, and each influence factor is weighted and combined to comprehensively evaluate the feasibility and cost size of the path point in the environment;
[0016] Step 2, according to the model of the path cost function, the initial path meeting the initial feasibility condition is generated by using the position information of the starting point and the ending point and the distribution of the obstacles, the initial path is composed of a series of path points;
[0017] Step 3, based on the sequence of the initial path points, the path points are adjusted and optimized by combining the path cost function, the position of the path points is dynamically updated, so that the path points gradually reduce the value under the influence factors of the path cost, and the optimized path points always meet the feasibility requirements in the initial path generation process during the adjustment process;
[0018] Step 4, the executable verification is carried out on the optimized path point sequence, the verification content includes whether the path point meets the motion ability and physical constraint condition of the fire reconnaissance robot, the verification is carried out by checking the motion parameters of the path point sequence, including the speed and acceleration, and the path point sequence can be accurately executed while meeting the path cost requirements;
[0019] Step 5, the path point sequence optimized and verified is fused with the real-time updated fire field environment data, the path point sequence is dynamically adjusted according to the new position distribution of the obstacle in the dynamic environment and the change of the high temperature area, and the path optimization requirement is continuously maintained after adjustment, and the requirements of dynamic obstacle avoidance, real-time response and safety of the fire reconnaissance task are met.
[0020] Preferably, the path smoothness in the path cost function is represented by the square sum of the velocity and the square sum of the acceleration of the path point sequence, and the calculation formula of the smoothness part is:
[0021] ,
[0022] wherein, represents the path smoothness cost, represents the position of the path point, is the time interval between the path points, and are the weight factors of the velocity and the smoothness cost,
[0023] and represent the position coordinates of the and the path points in the path point sequence, is the number of initial path points.
[0024] Preferably, the obstacle avoidance cost in the path cost function is determined by the minimum distance from the path point to the obstacle set, and the calculation formula is:
[0025] ,
[0026] wherein, represents the obstacle avoidance cost, represents the minimum distance from the path point to the obstacle set , is the weight factor of the obstacle avoidance cost, is the number of initial path points.
[0027] Preferably, the environmental temperature cost in the path cost function is determined by the temperature distribution value of the area passed by the path point, and the calculation formula is:
[0028] ,
[0029] wherein, represents the environmental temperature cost, is the temperature value of the position where the path point is located, is the weight factor of the environmental temperature cost, The number of initial path points.
[0030] Preferably, the initial spacing of the path points is calculated by the following formula:
[0031] ,
[0032] wherein, is the initial spacing of the path points, and is the end point and the start point of the path, is the number of discrete points of the initial path.
[0033] Preferably, the optimized adjustment of the path points is performed by dynamically updating the positions of the path points, and the weight factor of the path point update satisfies the following dynamic adjustment formula:
[0034] ,
[0035] wherein, represents the weight factor in the current iteration, is the adjustment step length,
[0036] is the partial derivative of the cost function with respect to the weight factor, is the i-th weight factor of the path cost function in the n-th iteration.
[0037] Preferably, when the executability of the optimized path point sequence is verified, the path speed and acceleration are checked by the following conditions:
[0038] ,
[0039] ,
[0040] wherein, and are the speed and acceleration of the path points, and are the maximum speed and maximum acceleration of the fire reconnaissance robot.
[0041] Preferably, in the comprehensive optimization of the path cost function, in order to improve the dynamic environment adaptability of the path, the overall optimization of the path is realized by increasing the dynamic environment adaptation cost item of the path points, and the calculation formula of the dynamic environment adaptation cost of the path points is:
[0042] ,
[0043] wherein, represents the dynamic environment adaptation cost, represents the temperature change amount of the region where the path point i is located. represents the distance variation of the region where the path point i is located to the dynamic obstacle, is a weight factor of the dynamic environment adaptation cost, is the number of path points.
[0044] Preferably, the path point updating step combines the objective function gradient and the dynamic constraint optimization strategy, and the path point updating formula is:
[0045] ,
[0046] wherein, represents the path point position after the kth iteration, represents the path point position after the kth iteration, represents the path point position after the kth iteration, represents the learning rate of path point updating;
[0047] is the partial derivative of the path cost function with respect to the path point ;
[0048] represents the weight factor of the dynamic constraint optimization;
[0049] represents the dynamic environment constraint correction amount of the path point.
[0050] Preferably, the calculation formula of the dynamic environment constraint correction amount of the path point is:
[0051] ,
[0052] wherein, represents the dynamic correction amount of the path point, is a correction step factor, represents the nearest dynamic obstacle position of the path point ,
[0053] represents the center position of the high-temperature region where the path point is located,
[0054] represents the Euclidean distance of the path point to the dynamic obstacle and the high-temperature region.
[0055] The present application provides a path generation method of a fire reconnaissance robot. The method has the following advantages:
[0056] 1、The application introduces a dynamic environment adaptive cost term in path optimization, combines real-time perception of obstacle position changes and high-temperature area diffusion, realizes adaptive adjustment of the path in a dynamic fire environment, and achieves the effect of avoiding dynamic obstacles and keeping away from dangerous areas.
[0057] 2、The application introduces a speed and smoothness cost term of path points in the path cost function, optimizes the change amplitude between path points, realizes the continuity and smoothness of the path, and achieves the effect of improving the execution stability of the path and ensuring the accuracy of the hardware control of the robot.
[0058] 3、The application calculates the dynamic correction amount of the path point, combines the real-time changes of the dynamic obstacle position and the center of the high-temperature area, realizes dynamic optimization of the path point position, and achieves the effect of improving the adaptability and robustness of the path in a complex fire environment.
[0059] 4、The application combines the global search of the multi-factor weight by comprehensively using the path cost function and the particle swarm optimization method, realizes the optimal planning of the path point in the global range, and achieves the effect of improving the comprehensive performance of the path cost function and avoiding local optimization. BRIEF DESCRIPTION OF DRAWINGS
[0060] Figure 1 The flowchart of the application. DETAILED DESCRIPTION
[0061] In order for those skilled in the art to understand the application scheme, the technical solutions in the embodiments of the application will be described clearly and completely below in combination with the drawings of the embodiments of the application. Obviously, the described embodiments are part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, other embodiments obtained by those skilled in the art without creative labor should be within the protection scope of the application.
[0062] The application will be described in detail below in combination with the drawings:
[0063] Embodiment:
[0064] Please refer to the accompanying drawings Figure 1 The embodiment of the application provides a path generation method of a fire reconnaissance robot, which comprises:
[0065] Step 1: According to the requirements of path generation of the fire reconnaissance robot, a path cost function is constructed, the path cost function is composed of multiple influence factors of the path, including the smoothness of the path, the stability of the path execution, the distance between the path point and the obstacle, and the environmental temperature of the area through which the path passes, and each influence factor is weighted and combined to comprehensively evaluate the feasibility and cost size of the path point in the environment.
[0066] Step 2: Based on the model of the path cost function, use the location information of the starting point and the ending point and the distribution of obstacles to generate an initial path that meets the initial feasibility conditions. The initial path is composed of a series of path points.
[0067] Step 3: Based on the sequence of initial path points, optimize and adjust the path points by combining the path cost function. By dynamically updating the location of the path points, the path points gradually reduce the value of the path cost under the influence of the factors, and the optimized path points always meet the feasibility requirements in the initial path generation process during the adjustment process.
[0068] Step 4: Verify the executability of the optimized path point sequence. The verification content includes whether the path points meet the movement ability and physical constraint conditions of the fire reconnaissance robot. The verification is performed by checking the motion parameters of the path point sequence, including speed and acceleration, and ensuring that the path point sequence can be accurately executed while meeting the path cost requirements.
[0069] Step 5: Fuse the path point sequence that has been optimized and verified with the real-time updated fire field environment data. According to the new location distribution of obstacles and the change of high temperature area in the dynamic environment, dynamically adjust the path point sequence, and continue to meet the requirements of path optimization after adjustment, while meeting the needs of dynamic obstacle avoidance, real-time response and safety of the fire reconnaissance task.
[0070] Benefits of Step 1:
[0071] Provides a unified optimization target, facilitating the quantification and optimization of subsequent path generation.
[0072] Integrates multiple fire field characteristic factors into the planning model, improving the accuracy and pertinence of path planning.
[0073] Benefits of Step 2:
[0074] Provides a feasible starting point for path optimization, reducing the computational complexity of path planning.
[0075] Ensures the basic feasibility of the initial path in complex environments, avoiding path generation failure.
[0076] Benefits of Step 3:
[0077] Dynamically optimizes the location of path points, improving the adaptability and execution efficiency of the path in the actual environment.
[0078] Ensures the global optimality of the path in the complex environment of the fire field, avoiding falling into local optimality.
[0079] Benefits of Step 4:
[0080] Improves the practicality of path planning, ensuring that the path can be executed by the robot.
[0081] Solve the problem of path planning not matching the performance of robot hardware, enhance the reliability of the system.
[0082] The benefits of step 5:
[0083] Improve the dynamic response capability of path planning to adapt to the complex changes of fire environment.
[0084] Ensure that the path can meet the safety and real-time requirements after dynamic obstacle avoidance and environmental adjustment.
[0085] In summary, the present application systematically evaluates the smoothness, stability, obstacle avoidance and environmental temperature influence of the path by constructing the path cost function, providing an optimization goal for path generation. On this basis, the initial path is generated and optimized, and the path points gradually reduce the value of the generation, realizing the global optimal path planning. Through the executability verification of the optimized path, it is ensured that the path can be accurately executed by the robot hardware. In addition, by combining the real-time updated fire environment data, the path is dynamically adjusted, so that the path can adapt to the dynamic changes in complex environment.
[0086] The path smoothness in the path cost function is represented by the sum of the square of the velocity and the square of the acceleration of the path point sequence, and the calculation formula of the smoothness part is:
[0087] ,
[0088] wherein, represents the path smoothness cost, represents the position of the path point, is the time interval between path points, and are the weight factors of velocity and smoothness cost,
[0089] and represent the position coordinates of the and the path points in the path point sequence, is the number of initial path points.
[0090] The obstacle avoidance cost in the path cost function is determined by the minimum distance from the path point to the obstacle set, and the calculation formula is:
[0091] ,
[0092] wherein, represents the obstacle avoidance cost, represents the minimum distance from the path point to the obstacle set , is the weight factor of obstacle avoidance cost, The number of initial path points.
[0093] The environmental temperature cost in the path cost function is determined by the temperature distribution value of the area passed by the path point, and its calculation formula is:
[0094] ,
[0095] Wherein, represents the environmental temperature cost, is the temperature value of the location where the path point is located, is the weight factor of the environmental temperature cost, is the number of initial path points.
[0096] The benefits of path smoothness:
[0097] Improve the smoothness of the path, ensure the continuity of the path trajectory, and there is no mutation point;
[0098] Reduce the violent movement of the robot when executing the path, protect the robot hardware, and prolong the service life of the equipment;
[0099] Improve the execution stability of the path, and ensure that the robot can move according to the trajectory.
[0100] The benefits of obstacle avoidance cost:
[0101] Improve the obstacle avoidance ability of the path, and ensure that the path point is always outside the safe distance of the obstacle;
[0102] Reduce the risk of collision between the path point and the obstacle, and improve the safety of the robot operation;
[0103] Ensure the feasibility of path generation in a complex obstacle environment, so that the robot can flexibly respond to changes in obstacle distribution.
[0104] The benefits of environmental temperature cost:
[0105] Improve the avoidance ability of the path to the high-temperature area of the fire scene, and ensure the safety of the robot operation;
[0106] Optimize path planning, so that the path point meets the environmental characteristics of the fire scene, and adapts to the diffusion of dynamic high-temperature areas;
[0107] Reduce the time of the robot running in the high-temperature area, and reduce the risk of equipment damage.
[0108] In summary, path smoothness, obstacle avoidance cost, and environmental temperature cost successively optimize and constrain path planning from the perspectives of trajectory smoothness, safety, and environmental adaptability of the path:
[0109] The path smoothness cost constraint the velocity and acceleration between path points, ensuring the continuity and smoothness of the path trajectory, and improving the stability and accuracy of path execution.
[0110] The obstacle avoidance cost effectively improves the obstacle avoidance ability of the path by calculating the minimum distance from the path point to the obstacle, ensuring that the path point is always far away from the obstacle and reducing the risk of collision.
[0111] The environmental temperature cost combines the temperature distribution value of the area where the path point is located, guiding the path to avoid high-temperature dangerous areas and improving the environmental adaptability and safety of path planning.
[0112] The initial spacing of the path point is calculated by the following formula:
[0113] ,
[0114] wherein, is the initial spacing of the path point, and is the end point and the starting point of the path, is the number of discrete points of the initial path.
[0115] The optimization adjustment of the path point is carried out by dynamically updating the position of the path point, and the weight factor of the path point update satisfies the following dynamic adjustment formula:
[0116] ,
[0117] wherein, represents the weight factor in the current iteration, is the adjustment step,
[0118] is the partial derivative of the cost function to the weight factor, is the value of the i-th weight factor of the path cost function in the n-th iteration.
[0119] The benefits of the initial spacing of the path point:
[0120] Ensure the uniformity of the distribution of path points, so that the initial path has continuity;
[0121] Provide stable input conditions for path optimization, avoid unreasonable distribution of initial path points leading to optimization failure;
[0122] Improve the efficiency of path planning and reduce the computational complexity of path optimization algorithm.
[0123] The benefits of dynamic adjustment of path points:
[0124] Dynamically adjust the position of the path point, gradually optimize the path cost function, and improve the global optimality of the path;
[0125] The efficiency of path point optimization is improved by adjusting the weight factor, reducing the number of iterations;
[0126] The partial derivative information of the cost function is combined to make the path point optimization process accurate and stable.
[0127] In summary, the initial spacing of path points and the dynamic adjustment of path points are designed by the initial conditions of path generation and the links of the optimization process:
[0128] By uniformly distributing path points, a basic feasible path is provided for path planning, ensuring that the initial path has good continuity and feasibility, and providing input conditions for subsequent path optimization.
[0129] By dynamically updating the position of path points and combining the dynamic adjustment of the weight factor, the path cost function is gradually optimized, ensuring that the distribution of path points converges to the optimal solution in the global range, improving the accuracy and efficiency of path planning.
[0130] The combination of the two makes the path planning process have a good starting point and can be dynamically adjusted during the optimization process, ultimately generating a globally optimal, smooth, and adaptive path to complex environments, providing technical support for path generation of fire reconnaissance robots in dynamic fire environments.
[0131] When verifying the executability of the optimized path point sequence, the path speed and acceleration are checked by the following conditions:
[0132] ,
[0133] ,
[0134] where, and are the speed and acceleration of the path point, and are the maximum speed and maximum acceleration of the fire reconnaissance robot.
[0135] The benefits of path speed checking:
[0136] Ensure that the path can be executed within the physical performance range of the fire reconnaissance robot;
[0137] Avoid the loss of control or operation failure of the robot due to the speed of the path exceeding the hardware limit;
[0138] Improve the safety and reliability of path execution.
[0139] The benefits of path acceleration checking:
[0140] Reduce the mechanical load during robot operation and protect the service life of the robot hardware;
[0141] Improve the stability of the execution path and avoid trajectory deviation caused by excessive acceleration;
[0142] It provides paths that conform to robot dynamics constraints, making path planning practical.
[0143] In summary, the executability of the optimized path point sequence was verified. Path velocity and acceleration checks ensured that the generated path met the physical constraints of the fire reconnaissance robot. Specifically:
[0144] Path speed check: By limiting the speed of path points, the path is ensured to be executed within the capabilities of the robot's hardware, avoiding control failure or errors caused by excessive path speed.
[0145] Path acceleration check: By limiting the acceleration of path points, the smoothness of path execution is improved, ensuring that the load on the robot hardware is minimized and the service life of the equipment is extended.
[0146] The above verification process can effectively improve the practicality and reliability of the path planning scheme, provide technical support for the path execution of fire reconnaissance robots in complex fire environments, and ensure the safety and efficiency of robot operation.
[0147] In the comprehensive optimization of the path cost function, to improve the dynamic environmental adaptability of the path, the overall optimization of the path is achieved by adding a dynamic environmental adaptation cost term for path points. The formula for calculating the dynamic environmental adaptation cost of path points is as follows:
[0148] ,
[0149] in, This represents the cost of adapting to a dynamic environment. This represents the temperature change in the area where path point i is located. This represents the change in distance from the area containing path point i to the dynamic obstacle. As a weighting factor for the cost of adapting to the dynamic environment, This represents the number of path points.
[0150] The path point update step combines the objective function gradient and dynamic constraint optimization strategy. The path point update formula is as follows:
[0151] ,
[0152] in, Indicates the first The path point position after the next iteration This represents the position of the path point in the k-th iteration. The learning rate represents the path point update rate;
[0153] The partial derivative of the path cost function with respect to the path point ;
[0154] The weight factor representing dynamic constraint optimization;
[0155] The dynamic environment constraint correction of the path point.
[0156] The dynamic environment constraint correction of the path point The calculation formula is:
[0157] ,
[0158] Wherein, The dynamic correction of the path point, The correction step factor, The center position of the path point The nearest dynamic obstacle position,
[0159] The center position of the path point The high temperature area,
[0160] The Euclidean distance of the path point to the dynamic obstacle and the high temperature area.
[0161] The benefits of dynamic environment adaptive cost:
[0162] The path has the ability to adapt to the dynamic environment, can avoid dynamic obstacles and stay away from high temperature areas;
[0163] Improve the safety and reliability of the path in complex dynamic fire environment;
[0164] Improve the response ability of path planning to real-time fire information, and ensure the feasibility of the path.
[0165] The benefits of the path point dynamic update formula:
[0166] Use gradient information to accelerate the convergence process of path optimization and improve the efficiency of path optimization;
[0167] Combine dynamic environment constraint optimization strategy, so that path point update can adapt to the change of obstacle position and high temperature area in real time;
[0168] Avoid local optimal path and ensure the global optimality of the path cost function.
[0169] The benefits of dynamic environment constraint correction:
[0170] Improve the avoidance ability of the path to dynamic obstacles and high temperature area diffusion;
[0171] Ensure that the path still meets the optimization requirements after dynamic adjustment, enhance the robustness of the path;
[0172] Through the accurate calculation of the dynamic correction amount, the accuracy and adaptability of path adjustment are improved.
[0173] In summary, the optimization method realizes the adaptive adjustment and optimization of the path in the complex fire environment through three parts: dynamic environment adaptive cost, dynamic path point update and dynamic environment constraint correction amount:
[0174] Dynamic environment adaptive cost: by considering the temperature change and dynamic obstacle distance change, the path can adapt to the dynamic change of the fire environment in real time, ensuring the safety and effectiveness of the path in the dynamic environment.
[0175] Dynamic path point update formula: combined with the gradient of the objective function and the dynamic environment constraint optimization strategy, the path point can quickly converge to the global optimum, while dynamically adapting to the change of the fire environment, improving the efficiency and accuracy of path optimization.
[0176] Dynamic environment constraint correction amount: through the calculation of the distance from the path point to the dynamic obstacle and the high temperature area, the dynamic adjustment of the path point is realized, and the obstacle avoidance ability and robustness of the path in the dynamic fire environment are improved.
[0177] Through the above optimization strategy, the path generation method realizes efficient, real-time and adaptive dynamic path planning, which provides guarantee for the task execution of the fire reconnaissance robot in the complex fire environment.
[0178] Although the embodiments of the present application have been shown and described, it can be understood by those skilled in the art that various changes, modifications, replacements and variations can be made to the embodiments without departing from the principles and spirits of the present application, and the scope of the present application is defined by the appended claims and their equivalents.
Claims
1. A path generation method of a fire reconnaissance robot, characterized by, The application relates to a path generation method for a fire reconnaissance robot, and belongs to the field of robot path planning. Step 1: constructing a path cost function according to the requirements of path generation of the fire reconnaissance robot, wherein the path cost function is composed of multiple influencing factors of the path, including the smoothness of the path, the stability of path execution, the distance between the path point and the obstacle and the environmental temperature of the region passed by the path, and each influencing factor is combined by weighting to comprehensively evaluate the feasibility and cost size of the path point in the environment; Step 2: generating an initial path meeting the initial feasibility conditions by using the position information of the starting point and the ending point and the distribution of the obstacle according to the model of the path cost function, wherein the initial path is composed of a series of path points; Step 3: optimizing and adjusting the path points based on the sequence of the initial path points and combining the path cost function, and gradually reducing the value of the path points under the influence of the path cost factors by dynamically updating the position of the path points, wherein the optimized path points always meet the feasibility requirements in the initial path generation process in the adjustment process; Step 4: verifying the executability of the sequence of the optimized path points, wherein the verification contents include whether the path points meet the motion ability and physical constraint conditions of the fire reconnaissance robot, and the verification is performed by checking the motion parameters of the sequence of the path points, including the speed and the acceleration, and the sequence of the path points can be accurately executed while meeting the path cost requirements; Step 5: fusing the sequence of the path points which are optimized and verified with the real-time updated fire field environment data, dynamically adjusting the sequence of the path points according to the new position distribution of the obstacle and the change of the high-temperature region in the dynamic environment, and continuously meeting the requirements of the fire reconnaissance task on dynamic obstacle avoidance, real-time response and safety after the adjustment.
2. The path generation method of a fire reconnaissance robot according to claim 1, wherein The path smoothness in the path cost function is represented by the square sum of the speed and the square sum of the acceleration of the sequence of the path points, and the calculation formula of the smoothness part is as follows: ， wherein denotes a path smoothness cost, denotes a position of a path point, is a time interval between path points, and are weight factors for the speed and smoothness cost, and denotes the position coordinate of the and the path point, is the number of initial path points.
3. The path generation method of a fire reconnaissance robot according to claim 2, wherein The obstacle avoidance cost in the path cost function is determined by the minimum distance from the path point to the obstacle set, and the calculation formula is as follows: ， wherein, represents an obstacle avoidance cost, represents a path point to a minimum distance to a set of obstacles, is a weight factor for the obstacle avoidance cost, is a number of initial path points. 4. The path generation method of a fire reconnaissance robot according to claim 3, wherein, The environmental temperature cost in the path cost function is determined by the temperature distribution value of the region passed by the path point, and the calculation formula is as follows: ， wherein, represents the ambient temperature cost, is the temperature value of the location where the path point is located, is a weight factor for the ambient temperature cost, is the number of initial path points.
5. The path generation method of a fire reconnaissance robot according to claim 1, wherein, The initial interval of the path point is calculated by the following formula: ， wherein, is the initial spacing of the waypoints, and is the end and start point of the path, is the number of discrete points of the initial path.
6. The path generation method of a fire reconnaissance robot according to claim 5, wherein, The optimization adjustment of the path point is performed by dynamically updating the position of the path point, and the weight factor of the path point updating meets the following dynamic adjustment formula: ， wherein, denotes the weight factor in the current iteration, is the adjustment step size, the partial derivative of the cost function with respect to the weight factor, the value of the i-th weight factor of the path cost function in the j-th iteration. the value of the i-th weight factor of the path cost function in the j-th iteration.
7. The path generation method of a fire reconnaissance robot according to claim 1, wherein When the executability of the sequence of the optimized path points is verified, the path speed and the acceleration are checked by the following conditions: ， ， wherein, and are the velocity and acceleration of the path point, and are the maximum velocity and maximum acceleration of the fire reconnaissance robot.
8. The path generation method of a fire reconnaissance robot according to claim 1, wherein, In the comprehensive optimization of the path cost function, in order to improve the dynamic environment adaptability of the path, the overall optimization of the path is realized by increasing the path point dynamic environment adaptation cost item, and the calculation formula of the path point dynamic environment adaptation cost is as follows: ， wherein, represents a dynamic environment adaptation cost, represents a temperature change amount of the region where the path point i is located, represents a distance change amount from the dynamic obstacle of the region where the path point i is located, is a weight factor of the dynamic environment adaptation cost, is the number of path points.
9. The path generation method of a fire reconnaissance robot according to claim 9, wherein, The updating step of the path point combines the gradient of the objective function and the dynamic constraint optimization strategy, and the updating formula of the path point is as follows: ， wherein, represents the path point position after the kth iteration, represents the path point position after the kth iteration, represents the path point position after the kth iteration, represents the learning rate of the path point update; the partial derivative of the path cost function with respect to the path point ; a weight factor representing a dynamic constraint optimization; a dynamic environmental constraint modification amount representing a waypoint.
10. The path generation method of a fire reconnaissance robot according to claim 9, wherein, a dynamic environment constraint correction amount of the path point The calculation formula is: ， wherein, represents a dynamic correction amount of the path point, is a correction step factor, represents a path point the most recent dynamic obstacle position, representative waypoint center position of a high-temperature region represents the Euclidean distance of the waypoint to the dynamic obstacle and the high temperature region.