Unmanned aerial vehicle three-dimensional path planning method and system based on fusion mechanism

Abstract

The invention discloses an unmanned aerial vehicle three-dimensional path planning method and system based on a fusion mechanism. Firstly, a PSO population is initialized through an improved Circle mapping and reverse learning mechanism, and then dynamic switching between a PSO algorithm and an HO algorithm is monitored and guided by using a fitness change function. Wherein the PSO algorithm is mainly responsible for global search, and the HO algorithm focuses on local fine tuning and deep search. Specifically, global search is carried out by using the PSO algorithm, and once the PSO algorithm locates a better area in the global search, the HO algorithm is switched to carry out deep search. The HO algorithm is good at local development and fine search, and can further optimize and find a better solution in a better region, so that the fast global search capability of the PSO algorithm and the fine local search advantage of the HO algorithm are effectively combined, and the faster convergence speed and the better solution quality are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle three-dimensional path planning, and in particular to a method and system for unmanned aerial vehicle three-dimensional path planning based on a fusion mechanism. Background Art

[0002] In recent years, with the rapid development of the low-altitude economy, drones have been widely used in the fields of cargo transportation, wildfire detection, etc. However, the three-dimensional path planning of drones faces complex environmental challenges. Drone trajectory path planning is essentially an optimization problem. Traditional single optimization algorithms often have poor convergence effects and are difficult to meet the requirements of high efficiency. Summary of the invention

[0003] The present invention provides a method and system for three-dimensional path planning of unmanned aerial vehicles based on a fusion mechanism, thereby solving the technical problems in the prior art that a single algorithm in three-dimensional path planning of unmanned aerial vehicles is prone to fall into local optimal solutions and has insufficient global search capabilities, thereby achieving the technical effect of obtaining faster convergence speed and better solution quality.

[0004] The present invention provides a three-dimensional path planning method for an unmanned aerial vehicle based on a fusion mechanism, comprising:

[0005] Step 1: Initialize the PSO population using the improved Circle sequence combined with the reverse learning mechanism as the initial planning path of the UAV;

[0006] Step 2: Use the PSO algorithm to perform a global search for the three-dimensional path of the UAV. In the process of the global search for the three-dimensional path, the formula Calculate the fitness function value Df(t); where f(t) is the fitness function, f(i) is the mapping value of the Df(t) function at point i, i is the variable that changes with iteration, t is the current iteration number, iter is the maximum iteration number, and M is a comparison constant;

[0007] Step 3: If the fitness function value Df(t) is less than the preset switching comparison threshold Threshold, switch to the HO algorithm to perform a three-dimensional path local search on the path obtained by the three-dimensional path global search, otherwise continue to perform a three-dimensional path global search;

[0008] Step 4: During the local search of the three-dimensional path, if the fitness function value Df(t) is equal to or greater than the switching comparison threshold Threshold, switching to the PSO algorithm to continue the global search of the three-dimensional path;

[0009] Step 5: Repeat steps 2 to 4 until the algorithm reaches the optimal fitness or the maximum number of iterations, and use the last searched three-dimensional path as the planned UAV trajectory path.

[0010] Specifically, the improved Circle sequence combined with the reverse learning mechanism to initialize the PSO population as the initial planning path of the UAV includes:

[0011] By formula Calculate the value X of the n+1th chaotic particle n+1 , generate a Circle chaotic sequence; where X n is the value of the nth chaotic particle;

[0012] By the formula Xnew = k × (Ub-Lb)-X n+1 Reverse learning is performed as the initial population solution Xnew of the PSO population; wherein k is a random vector obeying a normal distribution, and Ub and Lb are the upper and lower limits of the decision space, respectively.

[0013] Specifically, through the formula The switching comparison threshold Threshold is obtained by calculation; wherein Tefitness is the theoretical optimal fitness, and Maxiter is the preset maximum number of iterations.

[0014] Specifically, in the process of local search of three-dimensional paths, the formula Calculate the position of the male hippopotamus of the i-th generation in, is the position of the male hippopotamus of the i-th generation with j hippopotamus individuals, w(j) is the scaling factor of the cosine decrease, and x ij is the i-th generation hippo population with j position variables, y 1 is a random number on [0,1], I 1 is a random number on [1,2], D hippo is the position of the historical optimal hippo population, and the position sequence of male hippos obtained by iterative calculation is the three-dimensional path of the drone obtained by local search.

[0015] Specifically, through the formula The scaling factor w(j) is calculated; wherein w max is the maximum value of the preset scaling factor, w min is the preset minimum value of the scaling factor, j is the current number of iterations, j max is the maximum number of iterations.

[0016] The present invention also provides a three-dimensional path planning system for unmanned aerial vehicles based on a fusion mechanism, comprising:

[0017] The path initialization module is used to initialize the PSO population using the improved Circle sequence combined with the reverse learning mechanism as the initial planning path of the UAV;

[0018] The three-dimensional path planning module is used to use the PSO algorithm to perform a global search for the three-dimensional path of the UAV. In the process of the three-dimensional path global search, the formula The fitness function value Df(t) is calculated; wherein, f(t) is the fitness function, f(i) is the mapping value of the Df(t) function at point i, i is a variable that is performed with iteration, t is the current number of iterations, iter is the maximum number of iterations, and M is a comparison constant; if the fitness function value Df(t) is less than the preset switching comparison threshold Threshold, the HO algorithm is switched to perform a local three-dimensional path search on the path obtained by the global search of the three-dimensional path, otherwise the global search of the three-dimensional path continues; during the local search of the three-dimensional path, if the fitness function value Df(t) is equal to or greater than the switching comparison threshold Threshold, the PSO algorithm is switched to continue the global search of the three-dimensional path until the algorithm reaches the optimal fitness or reaches the maximum number of iterations, and the three-dimensional path finally searched is used as the planned UAV track path.

[0019] Specifically, the path initialization module includes:

[0020] The chaotic sequence generation unit is used to generate the chaotic sequence through the formula Calculate the value X of the n+1th chaotic particle n+1 , generate a Circle chaotic sequence; where X n is the value of the nth chaotic particle;

[0021] The initial path planning unit is used to calculate the path through the formula Xnew = k × (Ub-Lb)-X n+1 Reverse learning is performed as the initial population solution Xnew of the PSO population; wherein k is a random vector obeying a normal distribution, and Ub and Lb are the upper and lower limits of the decision space, respectively.

[0022] Specifically, it also includes:

[0023] Switch comparison threshold calculation module to use the formula The switching comparison threshold Threshold is obtained by calculation; wherein Tefitness is the theoretical optimal fitness, and Maxiter is the preset maximum number of iterations.

[0024] Specifically, the three-dimensional path planning module is specifically used to use the PSO algorithm to perform a global search for the three-dimensional path of the UAV. During the global search of the three-dimensional path, the formula The fitness function value Df(t) is calculated; wherein, f(t) is the fitness function, f(i) is the mapping value of the Df(t) function at point i, i is a variable that is performed with iteration, t is the current number of iterations, iter is the maximum number of iterations, and M is a comparison constant; if the fitness function value Df(t) is less than the preset switching comparison threshold Threshold, the HO algorithm is switched to perform a local search of the three-dimensional path on the path obtained by the global search of the three-dimensional path, otherwise the global search of the three-dimensional path continues; during the local search of the three-dimensional path, if the fitness function value Df(t) is equal to or greater than the switching comparison threshold Threshold, the PSO algorithm is switched to continue the global search of the three-dimensional path until the algorithm reaches the optimal fitness or reaches the maximum number of iterations, and the three-dimensional path finally searched is used as the planned UAV track path; during the local search of the three-dimensional path, the formula is used Calculate the position of the male hippopotamus of the i-th generation in, is the position of the male hippopotamus of the i-th generation with j hippopotamus individuals, w(j) is the scaling factor of the cosine decrease, and x ij is the i-th generation hippo population with j position variables, y 1 is a random number on [0,1], I 1 is a random number on [1,2], D hippo is the position of the historical optimal hippo population, and the position sequence of male hippos obtained by iterative calculation is the three-dimensional path of the drone obtained by local search.

[0025] Specifically, it also includes:

[0026] Scaling factor calculation module, used to calculate the The scaling factor w(j) is calculated; wherein w max is the maximum value of the preset scaling factor, w min is the preset minimum value of the scaling factor, j is the current number of iterations, j max is the maximum number of iterations.

[0027] One or more technical solutions provided in the present invention have at least the following technical effects or advantages:

[0028] First, the PSO population is initialized through the improved Circle mapping and reverse learning mechanism, and then the fitness change function is used to monitor and guide the dynamic switching between the PSO algorithm and the HO algorithm. Among them, the PSO algorithm is mainly responsible for global search, while the HO algorithm focuses on local fine-tuning and deep search. Specifically, the PSO algorithm is used for global search first. Once the PSO algorithm locates a better area in the global search, it switches to the HO algorithm for deep search. The HO algorithm is good at local development and fine search, and can further optimize in the better area to find a better solution, thereby effectively combining the fast global search capability of the PSO algorithm and the fine local search advantage of the HO algorithm, thereby achieving faster convergence speed and better solution quality.

[0029] In addition, the present invention also has the following advantages:

[0030] 1. The random initialization method of the original algorithm was improved. The PSO population was initialized using the improved Circle mapping combined with the reverse learning mechanism, which reduced the impact of the random population on the search results and made the distribution of the initialized population more even, thereby improving the convergence speed.

[0031] 2. A fitness function is proposed to monitor the fitness changes to achieve dynamic switching between the PSO algorithm and the HO algorithm.

[0032] 3. A scaling factor is added to the hippo population update phase to control the aggregation level of the hippo population and enhance its local search capability. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A flowchart of a three-dimensional path planning method for a UAV based on a fusion mechanism provided in an embodiment of the present invention;

[0034] Figure 2 It is the trajectory planning diagram of various algorithms in urban terrain;

[0035] Figure 3 for Figure 2 2D graph of ;

[0036] Figure 4 Convergence curves of each algorithm. DETAILED DESCRIPTION

[0037] The embodiments of the present invention provide a method and system for three-dimensional path planning of unmanned aerial vehicles based on a fusion mechanism, thereby solving the technical problems in the prior art that a single algorithm in three-dimensional path planning of unmanned aerial vehicles is prone to fall into local optimal solutions and has insufficient global search capabilities, thereby achieving the technical effect of obtaining faster convergence speed and better solution quality.

[0038] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0039] like Figure 1 As shown, the three-dimensional path planning method for a UAV based on a fusion mechanism provided by an embodiment of the present invention includes:

[0040] Step 1: Initialize the PSO population using the improved Circle sequence combined with the reverse learning mechanism as the initial planning path of the UAV;

[0041] This step is explained in detail. The improved Circle sequence is used in combination with the reverse learning mechanism to initialize the PSO population as the initial planning path of the drone, including:

[0042] By formula Calculate the value X of the n+1th chaotic particle n+1 , generate a Circle chaotic sequence; where X n is the value of the nth chaotic particle. In the three-dimensional path planning problem, it represents the position of the nth particle in the initial population.

[0043] By the formula Xnew = k × (Ub-Lb)-X n+1 Reverse learning is performed as the initial population solution Xnew of the PSO population; where k is a random vector obeying a normal distribution, and Ub and Lb are the upper and lower limits of the decision space, respectively.

[0044] Step 2: Use the PSO algorithm (particle swarm optimization algorithm) to perform a global search for the three-dimensional path of the drone. In the process of the global search for the three-dimensional path, the formula The fitness function value Df(t) is calculated; where f(t) is the fitness function, which is used to guide the switching between the PSO algorithm and the HO algorithm (Hippo optimization algorithm); f(i) is the mapping value of the Df(t) function at point i, i is a variable that progresses with the iteration, indicating the inflection point of the Df(t) function, t is the current number of iterations, iter is the maximum number of iterations, and M is a comparison constant used to ensure that the PSO algorithm dominates in the early stage; (i, f(i)) satisfies: when there are more than two non-differentiable points on the left side of point t, (i, f(i)) is the second non-differentiable point on the left; when there is one and only one non-differentiable point on the left side of point t, (i, f(i)) is that non-differentiable point; when there is no non-differentiable point on the left side of point t, increase the value of iter to meet the first two conditions.

[0045] In this embodiment, iter is one tenth of the maximum number of iterations maxiter.

[0046] Step 3: If the fitness function value Df(t) is less than the preset switching comparison threshold Threshold, it indicates that the global search is gradually converging and the PSO algorithm may be close to the local optimum. At this time, the HO algorithm needs to be used for local search of the three-dimensional path. Then, the HO algorithm is switched to perform local search of the three-dimensional path on the path obtained by the global search of the three-dimensional path. Otherwise, the global search of the three-dimensional path continues.

[0047] Specifically, through the formula The switching comparison threshold Threshold is calculated; where Tefitness is the theoretical optimal fitness and Maxiter is the preset maximum number of iterations.

[0048] In the process of local search of three-dimensional path, the formula Calculate the position of the male hippopotamus of the i-th generation in, is the position of the male hippopotamus of the i-th generation with j hippopotamus individuals, w(j) is the scaling factor of the cosine decrease, and x ij is the i-th generation hippo population with j position variables, y 1 is a random number on [0,1], I 1 is a random number on [1,2], D hippo is the position of the historical optimal hippo population, and the position sequence of male hippopotamus obtained by iterative calculation is the three-dimensional path of the UAV obtained by local search. In the early stage of iteration, a larger w(j) value can increase the proportion of the current optimal hippo in the population, while in the later stage of iteration, a smaller w(j) value will reduce the impact of the current optimal solution on the update of the hippo population, thereby enhancing the local optimization ability of the hippo optimization algorithm in the UAV path planning.

[0049] Specifically, through the formula The scaling factor w(j) is calculated; where w max is the maximum value of the preset scaling factor, w min is the preset minimum value of the scaling factor, j is the current number of iterations, j max is the maximum number of iterations.

[0050] In this embodiment, w max and w min Take 0.9 and 0.4 respectively.

[0051] Step 4: During the local search of the three-dimensional path, if the fitness function value Df(t) is equal to or greater than the switching comparison threshold Threshold, it means that the global search has made significant progress and the PSO algorithm is still effectively exploring the solution space. At this time, the PSO algorithm should be used to continue to maintain the global search capability, and then switch to the PSO algorithm to continue the global search of the three-dimensional path;

[0052] Step 5: Repeat steps 2 to 4 until the algorithm reaches the optimal fitness or the maximum number of iterations, and use the last searched three-dimensional path as the planned UAV trajectory path.

[0053] In addition, an embodiment of the present invention further provides a three-dimensional path planning system for a UAV based on a fusion mechanism, including:

[0054] The path initialization module is used to initialize the PSO population using the improved Circle sequence combined with the reverse learning mechanism as the initial planning path of the UAV;

[0055] Specifically, the path initialization module includes:

[0056] The chaotic sequence generation unit is used to generate the chaotic sequence through the formula Calculate the value X of the n+1th chaotic particle n+1 , generate a Circle chaotic sequence; where X n is the value of the nth chaotic particle. In the three-dimensional path planning problem, it represents the position of the nth particle in the initial population.

[0057] The initial path planning unit is used to calculate the path through the formula Xnew = k × (Ub-Lb)-X n+1 Reverse learning is performed as the initial population solution Xnew of the PSO population; where k is a random vector obeying a normal distribution, and Ub and Lb are the upper and lower limits of the decision space, respectively.

[0058] The three-dimensional path planning module is used to use the PSO algorithm to perform a global search for the three-dimensional path of the UAV. In the process of the global search for the three-dimensional path, the formula The fitness function value Df(t) is calculated; wherein, f(t) is the fitness function, f(i) is the mapping value of the Df(t) function at point i, i is a variable that is carried out with iteration, t is the current number of iterations, iter is the maximum number of iterations, and M is a comparison constant; if the fitness function value Df(t) is less than the preset switching comparison threshold Threshold, the HO algorithm is switched to perform a local search of the three-dimensional path on the path obtained by the global search of the three-dimensional path, otherwise the global search of the three-dimensional path continues; during the local search of the three-dimensional path, if the fitness function value Df(t) is equal to or greater than the switching comparison threshold Threshold, the PSO algorithm is switched to continue the global search of the three-dimensional path until the algorithm reaches the optimal fitness or reaches the maximum number of iterations, and the three-dimensional path obtained by the last search is used as the planned UAV trajectory path.

[0059] In this embodiment, iter is one tenth of the maximum number of iterations maxiter.

[0060] Specifically, the three-dimensional path planning module is used to use the PSO algorithm to perform a global search for the three-dimensional path of the UAV. In the process of the global search for the three-dimensional path, the formula The fitness function value Df(t) is calculated; where f(t) is the fitness function, f(i) is the mapping value of the Df(t) function at point i, i is a variable that is carried out with iteration, t is the current number of iterations, iter is the maximum number of iterations, and M is a comparison constant; if the fitness function value Df(t) is less than the preset switching comparison threshold Threshold, the HO algorithm is switched to perform a local search of the three-dimensional path on the path obtained by the global search of the three-dimensional path, otherwise the global search of the three-dimensional path continues; during the local search of the three-dimensional path, if the fitness function value Df(t) is equal to or greater than the switching comparison threshold Threshold, the PSO algorithm is switched to continue the global search of the three-dimensional path until the algorithm reaches the optimal fitness or reaches the maximum number of iterations, and the three-dimensional path obtained by the last search is used as the planned UAV trajectory path; during the local search of the three-dimensional path, the formula is used Calculate the position of the male hippopotamus of the i-th generation in, is the position of the male hippopotamus of the i-th generation with j hippopotamus individuals, w(j) is the scaling factor of the cosine decrease, and x ij is the i-th generation hippo population with j position variables, y 1 is a random number on [0,1], I 1 is a random number on [1,2], D hippo is the position of the historical optimal hippo population, and the position sequence of male hippos obtained by iterative calculation is the three-dimensional path of the drone obtained by local search.

[0061] To calculate the scaling factor, also include:

[0062] Scaling factor calculation module, used to calculate the Calculate the scaling factor w(j); where w max is the maximum value of the preset scaling factor, w min is the preset minimum value of the scaling factor, j is the current number of iterations, j max is the maximum number of iterations.

[0063] In this embodiment, w max and w min Take 0.9 and 0.4 respectively.

[0064] To calculate the handover comparison threshold, also include:

[0065] Switch comparison threshold calculation module to use the formula The switching comparison threshold Threshold is calculated; where Tefitness is the theoretical optimal fitness and Maxiter is the preset maximum number of iterations.

[0066] In order to analyze the three-dimensional path planning effect of the embodiment of the present invention, the algorithm (PSO-HO algorithm) provided by the embodiment of the present invention is compared with the existing SBOA algorithm, BWO algorithm, RUN algorithm, SO algorithm and GWO algorithm.

[0067] Specifically, this embodiment simulates an urban environment. The characteristics of an urban environment are that buildings are densely packed and have a small distribution range. Based on this characteristic, specific parameters are set. The map size is set to [1000*1000*30], the number of obstacles is set to 20, and the starting point and the end point are set to (50, 50, 1) and (1000, 1000, 8) respectively. The experimental results are as follows: Figure 2 , Figure 3 and Figure 4 As shown. Figure 4 It can be found that the PSO-HO algorithm, RUN algorithm and BWO algorithm all show a relatively smooth convergence process. The rapid fitness reduction of these three algorithms at the beginning of the iteration shows their excellent global search ability and can quickly approach the global optimal solution. Among them, the fitness of the PSO-HO algorithm decreases the fastest and maintains a low fitness value throughout the iteration process, which shows that the PSO-HO algorithm has the best overall performance, which is attributed to its effective fusion of the global search ability of PSO and the local development ability of HO. In contrast, the fitness of the SBOA algorithm, SO algorithm and GWO algorithm fluctuates greatly during the entire iteration process and has poor convergence, which reflects that these algorithms still have significant room for improvement in global search and local development capabilities.

[0068] In addition, the performance comparison results of the PSO-HO algorithm provided by the embodiment of the present invention and the existing SBOA algorithm, BWO algorithm, RUN algorithm, SO algorithm and GWO algorithm are shown in Table 1.

[0069]

[0070] Table 1 Performance comparison results between algorithms

[0071] As can be seen from Table 1, the convergence effect of the PSO-HO algorithm can be improved by up to 14.8% compared with other algorithms, and the average convergence times are reduced by up to 71.8%. The PSO-HO hybrid algorithm shows a strong superiority and is suitable for three-dimensional path planning of UAVs.

[0072] In summary, in order to solve the problem that a single algorithm is prone to fall into the local optimal solution and the global search ability is insufficient in the three-dimensional path planning of unmanned aerial vehicles, the embodiment of the present invention proposes a hybrid strategy combining the particle swarm optimization algorithm (PSO algorithm) and the hippopotamus optimization algorithm (HO algorithm). The PSO-HO hybrid algorithm effectively integrates the fast global search ability of the PSO algorithm and the fine local search advantage of the HO algorithm through a dynamic switching mechanism, thereby obtaining a faster convergence speed and better solution quality. Specifically, the PSO population is first initialized by an improved Circle mapping and reverse learning mechanism, and then the fitness change function is used to monitor and guide the switching between the PSO algorithm and the HO algorithm to optimize the synergistic effect of global and local search. In the HO search stage, the aggregation of the hippopotamus population is adjusted by introducing a cosine-type decreasing scaling factor to enhance its local search ability. The experimental results show that the fusion algorithm proposed in the embodiment of the present invention performs well in both convergence effect and speed, fully verifying its application potential in the three-dimensional path planning of unmanned aerial vehicles.

[0073] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0074] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0075] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0076] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0077] The parts not described in detail in the embodiments of the present invention are all known technologies to those skilled in the art. Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions of the present invention, which should be included in the scope of the claims of the present invention.

Claims

1. A three-dimensional path planning method for unmanned aerial vehicles based on a fusion mechanism, characterized in that: include: Step 1: Initialize the PSO population using the improved Circle sequence combined with the reverse learning mechanism as the initial planning path of the UAV; Step 2: Use the PSO algorithm to perform a global search for the three-dimensional path of the UAV. In the process of the global search for the three-dimensional path, the formula Calculate the fitness function value Df(t); where f(t) is the fitness function, f(i) is the mapping value of the Df(t) function at point i, i is the variable that changes with iteration, t is the current iteration number, iter is the maximum iteration number, and M is a comparison constant; Step 3: If the fitness function value Df(t) is less than the preset switching comparison threshold Threshold, switch to the HO algorithm to perform a three-dimensional path local search on the path obtained by the three-dimensional path global search, otherwise continue to perform a three-dimensional path global search; Step 4: During the local search of the three-dimensional path, if the fitness function value Df(t) is equal to or greater than the switching comparison threshold Threshold, switching to the PSO algorithm to continue the global search of the three-dimensional path; Step 5: Repeat steps 2 to 4 until the algorithm reaches the optimal fitness or the maximum number of iterations, and use the last searched three-dimensional path as the planned UAV trajectory path.

2. The method for three-dimensional path planning of an unmanned aerial vehicle based on a fusion mechanism as claimed in claim 1, characterized in that: The method of using the improved Circle sequence and the reverse learning mechanism to initialize the PSO population as the initial planning path of the UAV includes: By formula Calculate the value X of the n+1th chaotic particle n+1 , generate a Circle chaotic sequence; where X n is the value of the nth chaotic particle; By the formula Xnew = k × (Ub-Lb)-X n+1 Reverse learning is performed as the initial population solution Xnew of the PSO population; wherein k is a random vector obeying a normal distribution, and Ub and Lb are the upper and lower limits of the decision space, respectively.

3. The method for three-dimensional path planning of an unmanned aerial vehicle based on a fusion mechanism as claimed in claim 1, characterized in that: By formula The switching comparison threshold Threshold is obtained by calculation; wherein Tefitness is the theoretical optimal fitness, and Maxiter is the preset maximum number of iterations.

4. The method for three-dimensional path planning of an unmanned aerial vehicle based on a fusion mechanism as claimed in claim 1, characterized in that: In the process of local search of three-dimensional path, the formula Calculate the position of the male hippopotamus of the i-th generation in, is the position of the male hippopotamus of the i-th generation with j hippopotamus individuals, w(j) is the scaling factor of the cosine decrease, and x ij is the i-th generation hippo population containing j position variables, y1 is a random number on [0,1], I1 is a random number on [1,2], D hippo is the position of the historical optimal hippo population, and the position sequence of male hippos obtained by iterative calculation is the three-dimensional path of the drone obtained by local search.

5. The method for three-dimensional path planning of an unmanned aerial vehicle based on a fusion mechanism as claimed in claim 4, characterized in that: By formula The scaling factor w(j) is calculated; wherein w max is the maximum value of the preset scaling factor, w min is the preset minimum value of the scaling factor, j is the current number of iterations, j max is the maximum number of iterations.

6. A three-dimensional path planning system for unmanned aerial vehicles based on a fusion mechanism, characterized in that: include: The path initialization module is used to initialize the PSO population using the improved Circle sequence combined with the reverse learning mechanism as the initial planning path of the UAV; The three-dimensional path planning module is used to use the PSO algorithm to perform a global search for the three-dimensional path of the UAV. In the process of the three-dimensional path global search, the formula The fitness function value Df(t) is calculated; wherein, f(t) is the fitness function, f(i) is the mapping value of the Df(t) function at point i, i is a variable that is performed with iteration, t is the current number of iterations, iter is the maximum number of iterations, and M is a comparison constant; if the fitness function value Df(t) is less than the preset switching comparison threshold Threshold, the HO algorithm is switched to perform a local three-dimensional path search on the path obtained by the global search of the three-dimensional path, otherwise the global search of the three-dimensional path continues; during the local search of the three-dimensional path, if the fitness function value Df(t) is equal to or greater than the switching comparison threshold Threshold, the PSO algorithm is switched to continue the global search of the three-dimensional path until the algorithm reaches the optimal fitness or reaches the maximum number of iterations, and the three-dimensional path finally searched is used as the planned UAV track path.

7. The three-dimensional path planning system for unmanned aerial vehicles based on the fusion mechanism as claimed in claim 6, characterized in that: The path initialization module includes: The chaotic sequence generation unit is used to generate the chaotic sequence through the formula Calculate the value X of the n+1th chaotic particle n+1 , generate a Circle chaotic sequence; where X n is the value of the nth chaotic particle; The initial path planning unit is used to calculate the path through the formula Xnew = k × (Ub-Lb)-X n+1 Reverse learning is performed as the initial population solution Xnew of the PSO population; wherein k is a random vector obeying a normal distribution, and Ub and Lb are the upper and lower limits of the decision space, respectively.

8. The three-dimensional path planning system for unmanned aerial vehicles based on the fusion mechanism as claimed in claim 6, characterized in that: Also includes: Switch comparison threshold calculation module to use the formula The switching comparison threshold Threshold is obtained by calculation; wherein Tefitness is the theoretical optimal fitness, and Maxiter is the preset maximum number of iterations.

9. The three-dimensional path planning system for unmanned aerial vehicles based on the fusion mechanism as claimed in claim 6, characterized in that: The three-dimensional path planning module is specifically used to use the PSO algorithm to perform a global search for the three-dimensional path of the UAV. During the global search of the three-dimensional path, the formula The fitness function value Df(t) is calculated; wherein, f(t) is the fitness function, f(i) is the mapping value of the Df(t) function at point i, i is a variable that is performed with iteration, t is the current number of iterations, iter is the maximum number of iterations, and M is a comparison constant; if the fitness function value Df(t) is less than the preset switching comparison threshold Threshold, the HO algorithm is switched to perform a local search of the three-dimensional path on the path obtained by the global search of the three-dimensional path, otherwise the global search of the three-dimensional path continues; during the local search of the three-dimensional path, if the fitness function value Df(t) is equal to or greater than the switching comparison threshold Threshold, the PSO algorithm is switched to continue the global search of the three-dimensional path until the algorithm reaches the optimal fitness or reaches the maximum number of iterations, and the three-dimensional path finally searched is used as the planned UAV track path; during the local search of the three-dimensional path, the formula is used Calculate the position of the male hippopotamus of the i-th generation in, is the position of the male hippopotamus of the i-th generation with j hippopotamus individuals, w(j) is the scaling factor of the cosine decrease, and x ij is the i-th generation hippo population containing j position variables, y1 is a random number on [0,1], I1 is a random number on [1,2], D hippo is the position of the historical optimal hippo population, and the position sequence of male hippos obtained by iterative calculation is the three-dimensional path of the drone obtained by local search.

10. The three-dimensional path planning system for unmanned aerial vehicles based on the fusion mechanism as claimed in claim 9, characterized in that: Also includes: Scaling factor calculation module, used to calculate the The scaling factor w(j) is calculated; wherein w max is the maximum value of the preset scaling factor, w min is the preset minimum value of the scaling factor, j is the current number of iterations, j max is the maximum number of iterations.

Images