Improved advanced unmanned aerial vehicle path planning method based on snake optimizer

Abstract

The invention relates to an advanced unmanned aerial vehicle path planning method based on snake optimizer improvement. By integrating digital elevation information and threat information, an equivalent environment model is generated to simulate a real environment, and Hammersley sequence initialization, a Levy flight strategy and a ghost adversarial learning strategy are introduced, so that unmanned aerial vehicle path planning is realized. The problem of path planning of the unmanned aerial vehicle in complex terrains is solved.

Description

Technical Field

[0001] The present invention relates to an advanced UAV path planning method based on an improved snake optimizer. Background Art

[0002] With the rapid development of unmanned aerial vehicle technology, drones have been widely used in a variety of complex and dangerous tasks including search and rescue, environmental monitoring, and military reconnaissance. Especially in earthquake-stricken areas, due to the complex terrain and severe road damage, the risk of land transportation is extremely high, so using drones to transport materials has become the first choice. However, to ensure the smooth execution of drone tasks in these complex environments, the development of path planning technology plays a vital role. To this end, studying an efficient path planning algorithm that can adapt to complex terrain is crucial to improving the success rate and efficiency of drone mission execution. Summary of the invention

[0003] The purpose of the present invention is to improve the path planning effect of UAVs in disaster areas with complex terrain, and to provide an advanced UAV path planning method based on improved snake optimizer (ISO). By building an environmental model and integrating threat information and digital elevation data, a more efficient path planning is provided for UAV flight missions in high-risk areas.

[0004] To achieve the above-mentioned purpose, the technical solution of the present invention is: an advanced UAV path planning method based on an improved snake optimizer, which generates an equivalent environmental model by integrating digital elevation information and threat information to simulate the real environment, and uses an improved snake optimizer for UAV path planning, thereby realizing UAV path planning in complex terrain.

[0005] In one embodiment of the present invention, the environment model is constructed in the following manner:

[0006] (1) Establish digital elevation information, that is, abstract the planning space into a three-dimensional digital map with elevation information;

[0007] Assume that the elevation value of each point in space is a function of the coordinates of the point on the horizontal plane and the height on the horizontal plane. The height on the horizontal plane is set to H. min , when the height of the drone is lower than the height on the horizontal plane, flight is prohibited; for a point in space, its horizontal coordinates are (x, y), and the elevation information is expressed as:

[0008] Z1(x,y)=f(x,y)+H min

[0009] Among them, Z1(x,y) represents the elevation value of point (x,y), f(x,y) represents the functional relationship between the point and the horizontal coordinate, and H minIt is the minimum flight altitude limit for drones;

[0010] (2) The equivalent transformation formula of the threat model is as follows:

[0011]

[0012] Where Z2(x,y) represents the elevation value of the ith threat point, N is the total number of threat points, and h i Indicates the impact range of the threat point, (x i0 ,y i0 ) is the center coordinate of the threat point, x si and si Represent the slopes along the X-axis and Y-axis respectively;

[0013] The 3D digital map is combined with the threat model to generate an equivalent environment model. The elevation information of the environment model is expressed as:

[0014] Z(x,y)=max(Z1(x,y),Z2(x,y)).

[0015] In one embodiment of the present invention, the flight trajectory of the drone is evaluated by a cost function, and the optimal trajectory with the minimum cost is selected; considering the fuel consumption cost and the threat cost, the trajectory cost evaluation function is established as follows:

[0016]

[0017] Among them, J is the trajectory optimization objective function, L is the length of the UAV flight path, and J f (s) represents the fuel consumption cost, J t (s) represents the terrain threat cost, δ f and δ t is the weight value of the cost;

[0018] The fuel consumption cost is related to the flight path length as follows:

[0019] J f =c1·L

[0020] Where c1 is the proportionality constant between path length and fuel consumption;

[0021] The terrain threat cost is composed of a mountain model, and the collision probability of a drone in mountainous terrain is expressed as:

[0022]

[0023] Among them, h is the flight altitude of the UAV, H is the height of the obstacle, and d T is the distance at which the UAV is affected by the mountainous terrain, d Tmax and d TminIndicates the maximum impact distance and minimum impact distance of the drone at different altitudes; R T Indicates the extension radius of the mountain at the current height;

[0024] Assuming there are n threat costs to the drone, the terrain threat cost of a trajectory is calculated by summing the mountain threats over the entire trajectory:

[0025] J TkT =(p T (d kT ,1)+p T (d kT ,2)+p T (d kT ,3)+p T (d kT ,4)+p T (d kT ,5)) / 5

[0026] Where k represents the number of the threat point currently being calculated

[0027]

[0028] Among them, J TkT is the threat cost of each threat point, and n represents the number of threat points.

[0029] In one embodiment of the present invention, the snake optimizer is specifically implemented as follows:

[0030] 1) Initialization

[0031] First, the snake optimizer SO generates a uniformly distributed random initial population, which is obtained by the following formula:

[0032] x i =x min +rand×(x max -x min )

[0033] Among them, x i represents the position of the i-th individual, rand is a random number between [0,1], x min and x max are the lower and upper bounds respectively;

[0034] 2) Divide the population

[0035] Assume that the population is divided into two groups: males and females, each accounting for half of the population; use the following two formulas and to divide the population:

[0036] N m =N / 2

[0037] Nf =NN m

[0038] Where N is the number of individuals in the population, N m Refers to the number of male snakes, N f It refers to the number of female snakes;

[0039] 3) Define temperature and food

[0040] Temperature temp is defined by the following formula:

[0041]

[0042] Among them, t is the current iteration number, T is the maximum iteration number;

[0043] The amount of food Q is obtained by the following formula;

[0044]

[0045] Where c1 is a constant;

[0046] 4) When food is low, it is the exploration phase

[0047] If there is insufficient food, i.e., Q < 0.25, the snakes search for food by selecting any random location and update their positions; the exploration phase is modeled as follows:

[0048] x i,m (t+1)=x rand,m ±c2×A m ×((X max -X min )×rand+X min )

[0049] Among them, x i,m is the position of the ith male snake, x rand,m is the position of a random male snake, c2 is a constant, A m The ability of male snakes to find food is calculated as follows:

[0050]

[0051] Among them, f rand,m For x rand,m The fitness value, f i,m is the fitness value of the ith snake in the male population;

[0052] x i,f (t+1)=x rand,f ±c2×A f ×((X max -X min )×rand+Xmin )

[0053] Among them, x i,f is the position of the ith female snake, x rand,f is the position of a random female snake, A f The ability of a female snake to find food is calculated as follows:

[0054]

[0055] Among them, f rand,f is x rand,f The fitness value, f i,f is the fitness value of the i-th snake in the female population;

[0056] 5) Development phase when food is sufficient

[0057] If there is enough food and the temperature is adequate, that is, Q>0.25 and Temp>0.6, the snake will only move toward the food;

[0058] x i,j (t+1)=x food ±c3×Temp×rand×(x food -x i,j (t))

[0059] Among them, x i,j is the position of the individual, that is, the male or female snake, x food is the position of the optimal individual, c3 is a constant. If there is sufficient food but insufficient temperature, that is, Q>0.25 and Temp<0.6, the snake will be in fighting mode or mating mode;

[0060] 5.1) Struggle Mode

[0061] x i,m (t+1)=x i,m (t)±c3×FM×rand×(x best,f -x i,m (t))

[0062] Among them, x best,f is the position of the best individual in the female population, and FM is the fighting ability of male snakes

[0063] x i,f (t+1)=x i,f (t+1)±c3×FF×rand×(x best,m -x i,f (t+1))

[0064] Among them, x best,m is the position of the best individual in the male population, and FF is the fighting ability of female snakes;

[0065] The calculation formulas for FM and FF are as follows:

[0066]

[0067] Among them, f best,f is the fitness value of the best individual in the female population, f best,m is the fitness value of the best individual in the male population, f i is the fitness value of the individual;

[0068] 5.2) Mating pattern

[0069] x i,m (t+1)=x i,m (t)±c3×M m ×rand×(Q×x i,f (t)-x i,m (t))

[0070] x i,f (t+1)=x i,f (t)±c3×M f ×rand×(Q×x i,m (t)-x i,f (t))

[0071] Among them, M m and M f are the mating abilities of male and female snakes, respectively, and the calculation formula is as follows:

[0072]

[0073] If the eggs hatch, use them to replace the worst male and female snakes;

[0074] x worst,m =x min +rand×(x max -x min )

[0075] x worst,f =x min +rand×(x max -x min )

[0076] Among them, x worst,m is the worst individual in the male population, x worst,f The worst individual in the female population.

[0077] In one embodiment of the present invention, an initialization population method including a Hammersley sequence, a Levy flight strategy, and a ghost adversarial learning strategy are introduced to improve the snake optimizer to obtain an improved snake optimizer.

[0078] In one embodiment of the present invention, a method for initializing a population using a Hammersley sequence is introduced, that is, a Hammersley sequence is used to generate a uniformly distributed population in step 1) of initializing the snake optimizer.

[0079] In one embodiment of the present invention, the Levy flight strategy is introduced, that is, in step 5) of the snake optimizer when food is sufficient, that is, in the development stage, the position update formula of the individual, that is, the male snake or the female snake, is replaced, and the specific implementation is as follows:

[0080] The following formula is used to generate random numbers that match Levy flight:

[0081]

[0082] In the formula, u and v are variables that obey the normal distribution, that is, And σ u ,σ v Defined as:

[0083]

[0084] Where Γ is the standard gamma function, and the value range of β is [0,2];

[0085]

[0086] Where δ is the step size scaling factor, Represents a tensor product operation.

[0087] In one embodiment of the present invention, the Levy flight strategy is introduced, that is, in step 5) of the snake optimizer when food is sufficient, i.e., in the development stage, a special adversarial learning mechanism based on ghost confrontation is introduced to enhance the global search capability of the algorithm, which is specifically implemented as follows:

[0088] After the position of the individual, that is, the male or female snake, is updated, a ghost solution is generated at the position of the optimal individual. The update formula is as follows:

[0089]

[0090] ub and lb are the upper and lower bounds of the exploration area, respectively;

[0091] where x i,x is the projection length of the current solution on the X-axis, is the projection intensity of the imaging solution on the x-axis, x food is the position of the optimal individual, is the generated ghost solution, The calculation formula is as follows:

[0092]

[0093] x i is the original position of the individual, is the generated imaging solution;

[0094] After calculating the ghost solution, evaluate the fitness value of the ghost solution. If the ghost solution is better than the original individual, use the ghost solution to replace the original individual.

[0095] The present invention also provides an improved advanced UAV path planning system based on a snake optimizer, comprising a memory, a processor, and computer program instructions stored in the memory and capable of being executed by the processor. When the processor executes the computer program instructions, the method steps described above can be implemented.

[0096] The present invention also provides a computer-readable storage medium, on which computer program instructions that can be executed by a processor are stored. When the processor executes the computer program instructions, the method steps described above can be implemented.

[0097] Compared with the prior art, the present invention has the following beneficial effects: by introducing Hammersley sequence initialization, Levy flight strategy and ghost adversarial learning strategy, the performance of the present invention in complex environments is significantly better than that of traditional algorithms, and the intelligence level of the UAV path planning algorithm can be improved. DETAILED DESCRIPTION

[0098] The technical solution of the present invention is described in detail below.

[0099] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0100] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0101] The present invention provides an advanced UAV path planning method based on an improved snake optimizer, which integrates digital elevation information and threat information to generate an equivalent environmental model to simulate the real environment, and uses an improved snake optimizer for UAV path planning, thereby realizing the path planning of UAV in complex terrain.

[0102] The following is the specific implementation process of the present invention.

[0103] Based on the improved SO algorithm, this paper proposes an optimization scheme for path planning of UAVs in complex environments. By constructing an environmental model and integrating threat information and digital elevation data, this paper aims to provide a more efficient path planning method for UAV flight missions in high-risk areas.

[0104] 1. UAV mathematical model

[0105] The environment in which drones are located when performing tasks is very complex, and environmental modeling is the basis of drone path planning. In this invention, in addition to the elevation information based on the two-dimensional map, the threat of complex terrain is also considered. By integrating digital elevation information and threat information, an equivalent environmental model is generated to simulate the real environment.

[0106] 1.1 Environmental Model

[0107] The method of establishing digital elevation information is to abstract the planning space into a three-dimensional digital map with elevation information. It is assumed that the elevation value of each point in the space is a function of the coordinates of the point on the horizontal plane and the height of the horizontal plane. The height of the horizontal plane is set to H min , that is, when the height of the drone is below the horizontal plane, it is prohibited to fly. For a point in space, its horizontal coordinates are (x, y), and the elevation information can be expressed as:

[0108] Z1(x,y)=f(x,y)+H min

[0109] Among them, Z1(x,y) represents the elevation value of point (x,y), f(x,y) represents the functional relationship between the point and the horizontal coordinate, and H min It is the minimum flight altitude limit for drones.

[0110] The equivalent transformation formula of the threat model is as follows:

[0111]

[0112] Among them, Z2(x,y) represents the elevation value of the current threat point, N is the total number of threat points, and h i Indicates the impact range of the threat point, (x i0 ,y i0 ) is the center coordinate of the threat point, x si and si Represents the slope along the X-axis and Y-axis respectively.

[0113] According to the above model, the 3D elevation digital map is combined with the threat model using information fusion technology to generate an equivalent environmental model. The elevation information of the environmental model can be expressed as:

[0114] Z(x,y)=max(Z1(x,y),Z2(x,y))

[0115] 2.2 Trajectory constraints and cost function

[0116] The flight trajectory of the drone can be evaluated by the cost function, and the optimal trajectory with the minimum cost is selected. This paper considers the fuel consumption cost and threat cost. The trajectory cost evaluation function is established as follows:

[0117]

[0118] Among them, J is the trajectory optimization objective function, L is the length of the UAV flight path, and J f (s) represents the fuel consumption cost, J t (s) represents the terrain threat cost, δ f and δ t is the weight value of the cost.

[0119] The fuel consumption cost is mainly related to the flight path length, expressed as:

[0120] J f =c1·L

[0121] The threat cost is composed of a mountain model, and the collision probability of a drone in mountainous terrain is expressed as:

[0122]

[0123] Finally, the terrain threat cost of a trajectory can be calculated by summing the mountain threats over the entire trajectory:

[0124]

[0125] 2. Improved SO algorithm

[0126] Snake Optimizer (SO) is a new type of optimization algorithm based on swarm intelligence. It draws on the movement characteristics and behavior patterns of snakes when hunting to solve complex optimization problems. Like other swarm intelligence algorithms (such as ant colony algorithm, particle swarm algorithm, etc.), Snake Optimizer finds the global optimal solution by simulating the behavior of snakes in the environment.

[0127] 2.1 Initialization

[0128] First, SO will generate a uniformly distributed random population. The initial population can be obtained by the following formula:

[0129] x i =x min+rand× (x max -x min )

[0130] Among them, x i represents the position of the i-th individual, rand is a random number between [0,1], x min and x max are the lower and upper bounds of the problem respectively.

[0131] 2.2 Dividing the population

[0132] Assume that the population is divided into two groups: males and females, each accounting for half of the population. Use the following two formulas and to divide the population.

[0133] N m =N / 2

[0134] N f =NN m

[0135] Where N is the number of individuals in the population, N m Refers to the number of male snakes, N f Refers to the number of female snakes.

[0136] 2.3 Defining Temperature and Food

[0137] Temperature temp can be defined by the following formula.

[0138]

[0139] Where t is the current iteration number and T is the maximum iteration number. The amount of food Q can be obtained using the following formula.

[0140]

[0141] Here, c1 is a constant with a value of 0.5.

[0142] 2.4 When food is insufficient (exploration phase)

[0143] If there is insufficient food (the threshold is 0.25), that is, Q<0.25, the snakes will search for food by selecting any random location and update their positions. The exploration phase is modeled as follows:

[0144] x i,m (t+1)=x rand,m ±c2×A m ×((X max -X min )×rand+X min )

[0145] Among them, x i,m is the position of the ith male snake, x rand,m is the position of a random male snake, Am The ability of male snakes to find food is calculated as follows:

[0146]

[0147] Among them, f rand,m For x rand,m The fitness value, f i,m is the fitness value of the ith snake in the male population, and c2 is a constant with a value of 0.05.

[0148] x i,f (t+1)=x rand,f ±c2×A f ×((X max -X min )×rand+X min )

[0149] Among them, x i,f is the position of the ith female snake, x rand,f is the position A of a random female snake f The ability of a female snake to find food is calculated as follows:

[0150]

[0151] Among them, f rand,f is x rand,f The fitness value, f i,f is the fitness value of the ith snake in the female population.

[0152] 2.5 When food is sufficient (development stage)

[0153] If there is enough food and the temperature is sufficient (threshold is 0.6), that is, Q>0.25 and Temp>0.6, the snake will only move towards the food.

[0154] x i,j (t+1)=x food ±c3×Temp×rand×(x food -x i,j (t))

[0155] Among them, x i,j is the position of the individual (male or female), x food is the position of the optimal individual, c3 is a constant with a value of 2. If there is sufficient food but insufficient temperature, that is, Q>0.25 and Temp<0.6, the snake will be in fighting mode or mating mode.

[0156] 2.5.1 Struggle Mode

[0157] x i,m (t+1)=x i,m(t)±c3×FM×rand×(x best,f -x i,m (t))

[0158] Among them, x best,f is the position of the best individual in the female population, and FM is the fighting ability of male snakes

[0159] x i,f (t+1)=x i,f (t+1)±c3×FF×rand×(x best,m -x i,f (t+1))

[0160] Among them, x best,m is the position of the best individual in the male population, and FF is the fighting ability of female snakes.

[0161] The calculation formulas for FM and FF are as follows:

[0162]

[0163] Among them, f best,f is the fitness value of the best individual in the female population, f best,m is the fitness value of the best individual in the male population, f i is the fitness value of the individual.

[0164] 2.5.2 Mating pattern

[0165] x i,m (t+1)=x i,m (t)±c3×M m ×rand×(Q×x i,f (t)-x i,m (t))

[0166] x i,f (t+1)=x i,f (t)±c3×M f ×rand×(Q×x i,m (t)-x i,f (t))

[0167] Among them, M m and M f are the mating abilities of male and female snakes, respectively, and the calculation formula is as follows:

[0168]

[0169] If the eggs hatch, use them to replace the worst male and female snakes.

[0170] x worst,m =x min+rand×(x max -x min )

[0171] x worst,f =x min +rand×(x max -x min )

[0172] Among them, x worst,m is the worst individual in the male population, x worst,f The worst individual in the female population.

[0173] 2.6 Improved Snake Optimizer (ISO)

[0174] In order to improve the application effect of SO algorithm in complex environments, this paper proposes several important improvement strategies, including the initialization population method of Hammersley sequence, Levy flight strategy, and ghost adversarial learning strategy.

[0175] 2.6.1 Hammersley sequence initialization population

[0176] In metaheuristic algorithms, the distribution of the initial population has an important impact on the convergence speed and search efficiency of the algorithm. Traditional random initialization methods may lead to uneven distribution of individuals, affecting the performance of the algorithm. This paper uses the Hammersley sequence to generate a uniformly distributed population to ensure that the population covers the solution space more evenly, thereby improving the search efficiency and convergence speed.

[0177] 2.6.2 Levy flight strategy

[0178] Levy flight is a random walk model with long-tail distribution characteristics, which is suitable for describing the movement pattern of particles in complex search spaces. By introducing the Levy flight strategy, the algorithm's ability to jump out of the local optimal solution can be enhanced, thereby improving the global search performance. This paper uses the following formula to generate random numbers that conform to Levy flight.

[0179]

[0180] The generated random numbers are used to replace the Gaussian distribution random numbers to expand the exploration area of ​​the population. After introducing the Levy flight strategy, the position update formula during direct feeding is as follows

[0181]

[0182] Among them, δ is the step size scaling factor, which is taken as 0.7 in this paper. Represents the tensor product operation. It is used to replace the original algorithm

[0183]

[0184] 2.6.3 Learning Strategy Based on Ghost Adversarial

[0185] In order to improve the search efficiency and avoid falling into the local optimum, this paper enhances the global search capability of the algorithm by introducing a special adversarial learning mechanism based on ghost adversarial learning.

[0186] In the development stage of 2.5, after updating its own position, the individual will generate a ghost solution based on the optimal solution position. The update formula is as follows:

[0187]

[0188] where x i,x is the projection length of the current solution on the X-axis, is the projection intensity of the imaging solution on the x-axis, x food is the global optimum, Solution for the generated ghost. The calculation formula is as follows:

[0189]

[0190] k1=(1+(t / T) 0.5 ) 10

[0191] After calculating the ghost solution, evaluate the fitness value of the ghost solution. If the ghost solution is better than the original individual, use the ghost solution to replace the original individual.

[0192] 3. The process of algorithm for drone path planning

[0193] The core of the algorithm's optimization of the path through the cost function is to comprehensively evaluate the impact of different factors on the drone's path planning, and then select the flight path with the lowest total cost. Specifically, the cost function mainly includes the following two contents: fuel consumption cost and terrain threat cost. The algorithm balances these two costs by constantly adjusting the drone's flight path and finally finds the optimal path.

[0194] 3.1. Optimization of fuel consumption costs

[0195] The fuel consumption cost is proportional to the path length of the drone flight and is expressed as:

[0196] J f =c1·L

[0197] Among them, c1 is the proportional constant of path length and fuel consumption, and L is the length of the drone's flight path. A longer flight path means higher fuel consumption, so the algorithm will try to reduce the length of the flight path during the optimization process to reduce fuel consumption.

[0198] 3.2. Optimization of terrain threat cost

[0199] The terrain threat cost mainly considers the flight safety of UAVs in complex terrain. Assuming that there are multiple threat points (such as mountains) on the path, the impact range of the threat points is quantified through the threat model. The expression of terrain threat cost is:

[0200]

[0201] The threat cost J of each threat point TkT It is determined by the flight altitude of the drone, the height of the mountain, and the distance between the drone and the threat point. If the drone is closer to the threat point, the risk of collision will increase, thereby increasing the threat cost.

[0202] 3.3 Comprehensive Optimization Strategy

[0203] The overall cost function of the drone path is expressed as:

[0204]

[0205] Among them, δ f and δ t Represent the weights of fuel consumption cost and threat cost respectively. The algorithm adjusts the flight path through continuous iterations to minimize the total cost J. During the optimization process, the algorithm will find a balance between a shorter path and a lower threat.

[0206] Optimization process:

[0207] Initialize paths: First, generate a set of initial paths and calculate the fuel consumption cost and threat cost of each path.

[0208] Iterative Optimization: Based on different candidate paths, the algorithm will gradually update the position of the drone. During the update process, fuel consumption and threat cost will be optimized at the same time.

[0209] Balancing global and local search: The algorithm gradually finds the optimal path by striking a balance between global exploration (reducing path length) and local search (avoiding threat areas).

[0210] Convergence judgment: When the convergence conditions of the algorithm (such as the cost function no longer changes significantly) are met, the current optimal path is output.

[0211] In this way, the algorithm can find the best flight path for the drone in complex environments that reduces fuel consumption and avoids threats.

[0212] The present invention also provides an improved advanced UAV path planning system based on a snake optimizer, comprising a memory, a processor, and computer program instructions stored in the memory and capable of being executed by the processor. When the processor executes the computer program instructions, the method steps described above can be implemented.

[0213] The present invention also provides a computer-readable storage medium, on which computer program instructions that can be executed by a processor are stored. When the processor executes the computer program instructions, the method steps described above can be implemented.

[0214] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0215] The above is only a preferred embodiment of the present invention, and does not limit the present invention in other forms. Any technician familiar with the profession may use the above disclosed technical content to change or modify it into an equivalent embodiment with equivalent changes. However, any simple modification, equivalent change and modification made to the above embodiment according to the technical essence of the present invention without departing from the technical solution of the present invention still belongs to the protection scope of the technical solution of the present invention.

Claims

1. An advanced UAV path planning method based on snake optimizer improvement, characterized in that: By integrating digital elevation information and threat information, an equivalent environmental model is generated to simulate the real environment, and an improved snake optimizer is used for UAV path planning, thereby realizing the path planning of UAV in complex terrain.

2. The advanced UAV path planning method based on snake optimizer improvement according to claim 1 is characterized in that: The environmental model is constructed in the following manner: (1) Establish digital elevation information, that is, abstract the planning space into a three-dimensional digital map with elevation information; Assume that the elevation value of each point in space is a function of the coordinates of the point on the horizontal plane and the height on the horizontal plane. The height on the horizontal plane is set to H. min , when the height of the drone is lower than the height on the horizontal plane, flight is prohibited; for a point in space, its horizontal coordinates are (x, y), and the elevation information is expressed as: Z1(x,y)=f(x,y)+H min Among them, Z1(x,y) represents the elevation value of point (x,y), f(x,y) represents the functional relationship between the point and the horizontal coordinate, and H min It is the minimum flight altitude limit for drones; (2) The equivalent transformation formula of the threat model is as follows: Where Z2(x,y) represents the elevation value of the ith threat point, N is the total number of threat points, and h i Indicates the impact range of the threat point, (x i0 ,y i0 ) is the center coordinate of the threat point, x si and si Represent the slopes along the X-axis and Y-axis respectively; The 3D digital map is combined with the threat model to generate an equivalent environment model. The elevation information of the environment model is expressed as: Z(x,y)=max(Z1(x,y),Z2(x,y)).

3. The advanced UAV path planning method based on snake optimizer improvement according to claim 2 is characterized in that: The flight trajectory of the drone is evaluated by the cost function, and the optimal trajectory with the minimum cost is selected; considering the fuel consumption cost and threat cost, the trajectory cost evaluation function is established as follows: Among them, J is the trajectory optimization objective function, L is the length of the UAV flight path, and J f (s) represents the fuel consumption cost, J t (s) represents the terrain threat cost, δ f and δ t is the weight value of the cost; The fuel consumption cost is related to the flight path length as follows: J f =c1·L Where c1 is the proportionality constant between path length and fuel consumption; The terrain threat cost is composed of a mountain model, and the collision probability of a drone in mountainous terrain is expressed as: Among them, h is the flight altitude of the UAV, H is the height of the obstacle, and d T is the distance at which the UAV is affected by the mountainous terrain, d Tmax and d Tmin Indicates the maximum impact distance and minimum impact distance of the drone at different altitudes; R T Indicates the extension radius of the mountain at the current height; The terrain threat cost of a trajectory is calculated by summing the mountain threats over the entire trajectory: Where k represents the number of the threat point currently being calculated in, is the threat cost of each threat point, and n represents the number of threat points.

4. The advanced UAV path planning method based on snake optimizer improvement according to claim 1 is characterized in that: The specific implementation process of the snake optimizer is as follows: 1) Initialization First, the snake optimizer SO generates a uniformly distributed random initial population, which is obtained by the following formula: x i =x min +rand×(x max -x min ) Among them, x i represents the position of the i-th individual, rand is a random number between [0,1], x min and x max are the lower and upper bounds respectively; 2) Divide the population Assume that the population is divided into two groups: males and females, each accounting for half; use the following two formulas and to divide the groups: N m =N / 2 N f =N-N m Where N is the number of individuals in the population, N m Refers to the number of male snakes, N f It refers to the number of female snakes; 3) Define temperature and food Temperature temp is defined by the following formula: Among them, t is the current iteration number, T is the maximum iteration number; The amount of food Q is obtained by the following formula; Where c1 is a constant; 4) When food is low, it is the exploration phase If there is insufficient food, i.e., Q < 0.25, the snakes search for food by selecting any random location and update their positions; the exploration phase is modeled as follows: x i,m (t+1)=x rand,m ±c2×A m ×((X max -X min )×rand+X min ) Among them, x i,m is the position of the ith male snake, x rand,m is the position of a random male snake, c2 is a constant, A m The ability of male snakes to find food is calculated as follows: Among them, f rand,m For x rand,m The fitness value, f i,m is the fitness value of the ith snake in the male population; x i,f (t+1)=x rand,f ±c2×A f ×((X max -X min )×rand+X min ) Among them, x i,f is the position of the ith female snake, x rand,f is the position of a random female snake, A f The ability of a female snake to find food is calculated as follows: Among them, f rand,f is x rand,f The fitness value, f i,f is the fitness value of the i-th snake in the female population; 5) Development phase when food is sufficient If there is enough food and the temperature is adequate, that is, Q>0.25 and Temp>0.6, the snake will only move toward the food; x i,j (t+1)=x food ±c3×Temp×rand×(x food -x i,j (t)) Among them, x i,j is the position of the individual, that is, the male or female snake, x food is the position of the optimal individual, c3 is a constant. If there is sufficient food but insufficient temperature, that is, Q>0.25 and Temp<0.6, the snake will be in fighting mode or mating mode; 5.1) Struggle Mode x i,m (t+1)=x i,m (t)±c3×FM×rand×(x best,f -x i,m (t)) Among them, x best,f is the position of the best individual in the female population, and FM is the fighting ability of male snakes x i,f (t+1)=x i,f (t+1)±c3×FF×rand×(x best,m -x i,f (t+1)) Among them, x best,m is the position of the best individual in the male population, and FF is the fighting ability of female snakes; The calculation formulas for FM and FF are as follows: Among them, f best,f is the fitness value of the best individual in the female population, f best,m is the fitness value of the best individual in the male population, f i is the fitness value of the individual; 5.2) Mating pattern x i,m (t+1)=x i,m (t)±c3×M m ×rand×(Q×x i,f (t)-x i,m (t)) x i,f (t+1)=x i,f (t)±c3×M f ×rand×(Q×x i,m (t)-x i,f (t)) Among them, M m and M f are the mating abilities of male and female snakes, respectively, and the calculation formula is as follows: If the eggs hatch, use them to replace the worst male and female snakes; x worst,m =x min +rand×(x max -x min ) x worst,f =x min +rand×(x max -x min ) Among them, x worst,m is the worst individual in the male population, x worst,f The worst individual in the female population.

5. The advanced UAV path planning method based on snake optimizer improvement according to claim 4 is characterized in that: The snake optimizer is improved by introducing the initialization population method including the Hammersley sequence, the Levy flight strategy, and the ghost adversarial learning strategy.

6. The advanced UAV path planning method based on snake optimizer improvement according to claim 5 is characterized in that: The Hammersley sequence is introduced as a population initialization method, that is, the Hammersley sequence is used to generate a uniformly distributed population in the initialization of step 1) of the snake optimizer.

7. The advanced UAV path planning method based on snake optimizer improvement according to claim 5 is characterized in that: The Levy flight strategy is introduced, that is, in step 5 of the snake optimizer, when food is sufficient, that is, in the development stage, the position update formula of the individual, that is, the male or female snake, is replaced. The specific implementation is as follows: The following formula is used to generate random numbers that match Levy flight: In the formula, u and v are variables that obey the normal distribution, that is, And σ u ,σ v Defined as: Where Γ is the standard gamma function, and the value range of β is [0,2]; The generated random numbers are used to replace the Gaussian distribution random numbers to expand the exploration area of ​​the population. After the Levy flight strategy is introduced, the position update formula of the individual, that is, the male or female snake, is as follows: Where δ is the step size scaling factor, Represents a tensor product operation.

8. The advanced UAV path planning method based on snake optimizer improvement according to claim 5 is characterized in that: The Levy flight strategy is introduced, that is, in step 5 of the snake optimizer when food is sufficient, i.e., in the development stage, a special adversarial learning mechanism based on ghost confrontation is introduced to enhance the global search ability of the algorithm. The specific implementation is as follows: After the position of the individual, that is, the male or female snake, is updated, a ghost solution is generated at the position of the optimal individual. The update formula is as follows: ub and lb are the upper and lower bounds of the exploration area, respectively; where x i,x is the projection length of the current solution on the X-axis, is the projection intensity of the imaging solution on the x-axis, x food is the position of the optimal individual, is the generated ghost solution, The calculation formula is as follows: k1=(1+(t / T) 0.5 ) 10 x i is the original position of the individual, is the generated imaging solution; After calculating the ghost solution, the fitness value of the ghost solution is evaluated. If the ghost solution is better than the original individual, the ghost solution is used to replace the original individual.

9. An advanced UAV path planning system based on snake optimizer, characterized in that: The method comprises a memory, a processor and computer program instructions stored in the memory and capable of being executed by the processor. When the processor executes the computer program instructions, the method steps as claimed in any one of claims 1 to 8 can be implemented.

10. A computer-readable storage medium having stored thereon computer program instructions that can be executed by a processor, and when the processor executes the computer program instructions, the method steps according to any one of claims 1 to 8 can be implemented.

Images