Three-dimensional space unmanned aerial vehicle fire extinguishing path planning method based on improved genetic algorithm

Abstract

The invention belongs to the technical field of unmanned aerial vehicle fire extinguishing, and aims to solve the problem of low fire extinguishing efficiency caused by unreasonable path planning of unmanned aerial vehicle fire extinguishing in a complex environment at present. The invention provides a three-dimensional space unmanned aerial vehicle fire extinguishing path planning method based on an improved genetic algorithm, and the method comprises the following steps: giving environment parameters, building a three-dimensional simulation environment based on the environment parameters, and randomly setting an ignition point in the three-dimensional simulation environment; giving physical parameters of the unmanned aerial vehicle carrying the fire extinguishing bomb, building an unmanned aerial vehicle three-dimensional simulation model based on the physical parameters, and performing behavior simulation in a three-dimensional simulation environment; constructing an improved genetic algorithm based on a multi-point crossover strategy and a self-adaptive mutation rate; and by taking minimization of the fire extinguishing time as a target, performing unmanned aerial vehicle fire extinguishing path planning in a three-dimensional simulation environment by using an improved genetic algorithm. According to the method, the total fire extinguishing time is remarkably shortened, and the path quality and the algorithm robustness are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle fire extinguishing, and in particular relates to a three-dimensional space unmanned aerial vehicle fire extinguishing path planning method based on an improved genetic algorithm. Background Art

[0002] At present, the application of drones has involved aerial photography, agricultural and forestry plant protection, geological exploration, power inspection, oil and gas pipeline inspection, highway accident management, forest fire inspection, polluted environment investigation, emergency rescue and relief, disaster relief, coastline inspection and other fields.

[0003] Fire refers to a sudden disaster caused by an uncontrolled combustion process. In the past, forest fire protection was carried out by personnel climbing mountains to patrol and put out fires, which was highly dangerous and could not carry too much fire-fighting materials, let alone heavy fire-fighting equipment. At present, drones are used to carry cameras, i.e. monocular vision devices, to detect fire scenes and carry fire-extinguishing bombs to extinguish fires.

[0004] However, in complex environments, UAV firefighting currently suffers from unreasonable path planning, which leads to low firefighting efficiency. Summary of the invention

[0005] In order to solve at least one of the above-mentioned technical problems existing in the prior art, the present invention provides a three-dimensional space UAV fire-fighting path planning method based on an improved genetic algorithm.

[0006] The present invention is implemented by the following technical solution: a three-dimensional space UAV fire-fighting path planning method based on an improved genetic algorithm, comprising the following steps:

[0007] Given environmental parameters, a three-dimensional simulation environment is built based on the environmental parameters, and ignition points are randomly set in the three-dimensional simulation environment;

[0008] Given the physical parameters of a drone carrying fire extinguishing bombs, a 3D simulation model of the drone is built based on the physical parameters, and the behavior simulation is performed in a 3D simulation environment;

[0009] Construct an improved genetic algorithm based on multi-point crossover strategy and adaptive mutation rate;

[0010] With the goal of minimizing the fire extinguishing time, an improved genetic algorithm is used to plan the UAV fire extinguishing path in a three-dimensional simulation environment.

[0011] Preferably, the environmental parameters include but are not limited to the shape of the mountain, and the shape of the mountain is simulated based on a combination of a sine function, a cosine function and a Gaussian function to obtain a three-dimensional simulation environment model.

[0012] Preferably, the expression of the three-dimensional simulation environment model is:

[0013]

[0014] In the formula, Z represents the elevation of the corresponding point on the horizontal plane, A 1 and A 2 represents the amplitude, (x 1 ,y 1 )(x 2 ,y 2 ) represents the center of the mountain, σ x , σ y Indicates the width of the mountain.

[0015] Preferably, the mutation rate of individual fitness and population diversity is comprehensively considered, and a combined adaptive mutation rate is introduced into the genetic algorithm. The expression is:

[0016]

[0017] Among them, α is the weight coefficient, which controls the influence of the two mutation rates. To consider the adaptive mutation rate of individual fitness, is the adaptive mutation rate that takes population diversity into account.

[0018] Preferably, a multi-point crossover strategy is adopted in the genetic algorithm, and the multi-point crossover performs gene exchange at multiple positions.

[0019] Preferably, the steps of using the improved genetic algorithm to plan the UAV firefighting path include:

[0020] An initial population containing a set number of individuals is randomly generated, each of which represents the fire-fighting path and fire-fighting bomb distribution of all drones;

[0021] The fitness of each individual is calculated based on the total fire extinguishing time, path length, energy consumption, and fire extinguishing success rate;

[0022] Select individuals with high fitness based on the roulette strategy;

[0023] Apply multi-point crossover strategy to generate new offspring individuals;

[0024] Mutate the offspring individuals according to the combined adaptive mutation rate;

[0025] Add the newly generated offspring individuals to the population to replace the individuals with low fitness;

[0026] Repeat the above steps until the predetermined number of generations is reached or the increase in fitness is less than the set value, and the final drone fire extinguishing path is obtained.

[0027] Preferably, the fitness function is expressed as follows:

[0028]

[0029] Where D(i) is the total fire extinguishing distance of the i-th UAV, T(i) is the total fire extinguishing time of the i-th UAV, and C(i) is the penalty value added by the i-th UAV due to collision.

[0030] Preferably, each drone fire-fighting path includes a visit sequence from the starting point to each fire source point.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] The present invention aims at the problem of UAV path planning in complex environments, takes minimizing the fire-fighting time as the goal, and realizes the path planning of UAVs in three-dimensional space based on an improved genetic algorithm with multi-point crossover operation and adaptive mutation rate. In order to improve the search efficiency and convergence speed of the algorithm, multi-point crossover operation is introduced, and the mutation probability is dynamically adjusted according to individual fitness and population diversity in combination with the elite strategy and adaptive mutation rate. Compared with the traditional genetic algorithm, the improved genetic algorithm not only significantly shortens the total fire-fighting time, but also improves the path quality and algorithm robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0034] Figure 1 is a flow chart of the present invention;

[0035] Figure 2 It is a three-dimensional simulation environment model diagram of the present invention;

[0036] Figure 3 It is a schematic diagram of the distribution of fire source points of the present invention;

[0037] Figure 4 This is a schematic diagram of the path of the 1# UAV after optimization of the present invention;

[0038] Figure 5 This is a schematic diagram of the path of the 2# UAV after optimization of the present invention;

[0039] Figure 6 This is a schematic diagram of the path of the 3# UAV after optimization of the present invention;

[0040] Figure 7 This is a schematic diagram of the path of the 4# UAV after optimization according to the present invention. DETAILED DESCRIPTION

[0041] In conjunction with the drawings in the embodiments of the present invention, the technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other implementation methods obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0042] It should be noted that the structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification so that people familiar with this technology can understand and read them. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantive technical significance. Any structural modification, change in proportional relationship or adjustment of size should fall within the scope of the technical contents disclosed in the present invention without affecting the effects and purposes that can be achieved by the present invention. It should be noted that in this specification, relational terms such as first and second are only used to distinguish one entity from several other entities, and do not necessarily require or imply any actual relationship or order between these entities.

[0043] The present invention provides an embodiment: Figures 1 to 7 As shown, a three-dimensional space UAV fire-fighting path planning method based on an improved genetic algorithm includes the following steps: given environmental parameters, a three-dimensional simulation environment is built based on the environmental parameters, and fire points are randomly set in the three-dimensional simulation environment; given physical parameters of a UAV carrying a fire-extinguishing bomb, a three-dimensional simulation model of the UAV is built based on the physical parameters, and behavior simulation is performed in the three-dimensional simulation environment; an improved genetic algorithm based on a multi-point crossover strategy and an adaptive mutation rate is constructed; with the goal of minimizing the fire-fighting time, the improved genetic algorithm is used to plan the UAV fire-fighting path in the three-dimensional simulation environment.

[0044] In this embodiment, based on the characteristics of many mountains, in order to improve the authenticity of the simulation environment, a mountain terrain for drone path planning is built. The mountain terrain is simplified, and a combination of sine function, cosine function and Gaussian function is used to simulate the shape of the mountain to obtain a three-dimensional simulation environment model.

[0045] Taking the environmental model of two mountains as an example, we first use the sine and cosine combination function to simulate the shape of the mountain.

[0046] Z 1 =A 1 sin(k x x)cos(k y y)+

[0047] A 2 cos(k x x)sin(k y y)

[0048] Where Z 1 A represents the elevation value of the corresponding point on the horizontal plane under the combined function of sine and cosine. 1 and A 2 represents the amplitude, k x , k y It's the frequency.

[0049] The Gaussian function is expressed as:

[0050]

[0051] Among them, Z 2 represents the elevation value of the corresponding point on the horizontal plane under the Gaussian function, C represents the peak height, σ x , σ y Indicates the width of the mountain.

[0052] Combining the above models can obtain a mountain model, that is, a three-dimensional simulation environment model:

[0053]

[0054] In the formula, Z represents the elevation of the corresponding point on the horizontal plane, A 1 and A 2 represents the amplitude, (x 1 ,y 1 )(x 2 ,y 2 ) represents the center of the mountain, σ x , σ y Indicates the width of the mountain.

[0055] In the standard operation of genetic algorithms, excellent individuals may be destroyed due to random selection, crossover and mutation. The elite strategy can ensure that the best individuals in the population will not be lost during the evolution process. The elite strategy is to directly retain a part of the best individuals in each generation of the population, without crossover or mutation operations on these individuals, and directly copy them to the next generation of the population. The purpose of this is to ensure that individuals with high fitness can be retained, thereby maintaining the high-quality genes of the population.

[0056] Assuming the population size is N, first, the fitness of each individual in the population is evaluated. The elite retention ratio is p, and according to the results of the fitness evaluation, the top N with the highest fitness are selected. e N individuals serve as elites. e The number of elites reserved for a pre-set number can be calculated by the following formula:

[0057]

[0058] in Indicates rounding up.

[0059] These elite individuals are then directly copied to the next generation of the population. By retaining the best individuals, the algorithm can quickly converge to a better solution.

[0060] In order to improve the search efficiency of the algorithm and avoid premature convergence, the adaptive mutation rate (AMR) is introduced so that the mutation rate can be dynamically adjusted according to individual fitness or population diversity.

[0061] Assume that in the tth generation population, the fitness of individual i is The best fitness and the worst fitness of the entire population are and Adaptive mutation rate of individual i It can be expressed as:

[0062]

[0063] in and represent the maximum mutation rate and the minimum mutation rate respectively.

[0064] From the above formula, we can see that the worse the fitness of an individual, the higher its mutation rate, which can increase the search space; and the better the fitness of an individual, the lower its mutation rate, in order to maintain its excellent characteristics.

[0065] Population diversity can be measured by genotype differences. Let the population diversity index be D t , then the adaptive mutation rate considering population diversity is It is expressed as:

[0066]

[0067] Among them, D min and D max are the minimum and maximum values ​​of population diversity, respectively.

[0068] Assume G i and G j There are two genotypes, and the H function is used to assist in calculating the population diversity index D t :

[0069]

[0070] The H function is a function that measures the genotype difference between two individuals, and l is the length of the genotype. and The individual G i and G j The gene at the kth position.

[0071] is an indicator function, when When ,δ=0; when , δ=1.

[0072] By formula It can be seen that when the population diversity is low, the mutation rate is high; when the population diversity is high, the mutation rate is low.

[0073] Considering the mutation rate of individual fitness and population diversity, a combined adaptive mutation rate is designed. for:

[0074]

[0075] Among them, α is the weight coefficient, which controls the influence of the two mutation rates. To consider the adaptive mutation rate of individual fitness, is the adaptive mutation rate that takes population diversity into account.

[0076] This combined mutation rate method can explore the solution space to a larger extent in the initial stage, improving the global search capability, and can optimize the solution more meticulously in the later stage to avoid falling into the local optimum.

[0077] Crossover is the process of exchanging some genes of two parent individuals to generate new offspring individuals, thereby combining the excellent characteristics of the parent individuals. This paper mainly studies the path planning of three-dimensional drones, where individuals are a series of three-dimensional coordinate points and adopts a crossover strategy of multi-point crossover.

[0078] Multi-point crossover performs gene exchange at multiple locations. For ease of calculation, R 1 ,R 2 Represent odd and even gene fragments respectively to improve diversity. Let k 1 ,k 2 ,...,k m For multiple crossover points, the offspring individual C 1 and C 2 It is expressed as:

[0079] C 1 =(R 1 [1],…,R 1 [k 1 ],R 2 [k 1 +1],

[0080] …,R 2 [k 2 ],R 1 [k 2 +1],…)

[0081] C2 =(R 2 [1],…,R 2 [k 1 ],R 1 [k 1 +1],

[0082] …,R 1 [k 2 ],R 2 [k 2 +1],…)

[0083] The above two formulas represent the exchange strategies for odd segments and even segments respectively. Multiple intersection points are selected between two parent individuals, and the segments between the intersection points are exchanged to generate new offspring individuals.

[0084] In this embodiment, four drones are used, and the flight area is as follows: Figure 2 The three-dimensional space of 20km x 20km x 2km shown in Figure 3 As shown in the figure, 12 fire points are randomly set on the mountain surface, the initial population is set to 100, the maximum number of iterations is 500 generations, the initial mutation rate is 0.1, the adaptive mutation rate range is 0.01-0.2, the elite individual ratio is 0.1, and the crossover probability is 0.8. To prevent collisions between drones, the minimum distance D between drones is set. s = 10m, if the distance between the two drones is d ij <D s , then it is considered that a collision has occurred. In the fitness evaluation, if a collision is detected, a penalty mechanism is added to the fitness value to reduce the fitness of the path. The maximum flight speed of the drone is 60km / h, the flight time is 1h, each drone can carry up to 5 fire bombs, and the bombing time for each fire source is 1 minute.

[0085] Specifically, the steps of using the improved genetic algorithm to plan the UAV firefighting path include:

[0086] Initialize the population: Randomly generate an initial population containing a set number of individuals, each of which represents the fire-fighting path and fire-fighting bomb distribution of all drones;

[0087] Fitness calculation: The fitness of each individual is calculated based on the total fire extinguishing time, path length, energy consumption and fire extinguishing success rate; the fitness function is expressed as follows:

[0088]

[0089] Where D(i) is the total fire extinguishing distance of the i-th UAV, T(i) is the total fire extinguishing time of the i-th UAV, and C(i) is the penalty value added by the i-th UAV due to collision.

[0090] Selection operation: select individuals with high fitness based on the roulette strategy;

[0091] Crossover operation: Apply multi-point crossover strategy to generate new offspring individuals;

[0092] Mutation operation: mutate the offspring individuals according to the combined adaptive mutation rate;

[0093] Population update: add newly generated offspring individuals to the population to replace individuals with low fitness;

[0094] Iteration process: Repeat the above steps until the predetermined number of generations is reached or the increase in fitness is less than the set value, and the final drone fire-fighting path is obtained.

[0095] The path after 500 iterations is as follows Figures 4 to 7 As shown, each path contains a visit sequence from the starting point to each fire source point.

[0096] 30 independent experiments were conducted, and the fire extinguishing time and fire extinguishing distance were recorded in each experiment. The stability and robustness of the traditional genetic algorithm and the present invention were compared. The results are shown in Table 1.

[0097] Table 1 Experimental results statistics

[0098]

[0099] This paper aims to solve the problem of UAV path planning in complex environments, minimize the fire-fighting time, and realize the path planning of UAV in three-dimensional space based on the improved genetic algorithm with multi-point crossover operation and adaptive mutation rate. Compared with the traditional genetic algorithm, the improved genetic algorithm not only significantly shortens the total fire-fighting time, but also improves the path quality and algorithm robustness.

[0100] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A three-dimensional space UAV fire-fighting path planning method based on improved genetic algorithm, characterized in that: The following steps are involved: Given environmental parameters, a three-dimensional simulation environment is built based on the environmental parameters, and ignition points are randomly set in the three-dimensional simulation environment; Given the physical parameters of a drone carrying fire extinguishing bombs, a 3D simulation model of the drone is built based on the physical parameters, and the behavior simulation is performed in a 3D simulation environment; Construct an improved genetic algorithm based on multi-point crossover strategy and adaptive mutation rate; With the goal of minimizing the fire extinguishing time, an improved genetic algorithm is used to plan the UAV fire extinguishing path in a three-dimensional simulation environment.

2. The three-dimensional space UAV fire-fighting path planning method based on improved genetic algorithm according to claim 1 is characterized by: The environmental parameters include but are not limited to the shape of the mountain. The shape of the mountain is simulated based on a combination of a sine function, a cosine function and a Gaussian function to obtain a three-dimensional simulation environment model.

3. The three-dimensional space UAV fire-fighting path planning method based on improved genetic algorithm according to claim 2 is characterized by: The expression of the three-dimensional simulation environment model is: Where Z represents the elevation of the corresponding point on the horizontal plane, A1 and A2 represent the amplitude, (x1, y1)(x2, y2) represents the center of the mountain, and σ x , σ y Indicates the width of the mountain.

4. The three-dimensional space UAV fire-fighting path planning method based on improved genetic algorithm according to claim 1 is characterized by: Taking into account the mutation rate of individual fitness and population diversity, a combined adaptive mutation rate is introduced into the genetic algorithm. The expression is: Among them, α is the weight coefficient, which controls the influence of the two mutation rates. To consider the adaptive mutation rate of individual fitness, is the adaptive mutation rate that takes population diversity into account.

5. The three-dimensional space UAV fire-fighting path planning method based on improved genetic algorithm according to claim 1 is characterized by: In the genetic algorithm, a multi-point crossover strategy is used, and multi-point crossover performs gene exchange at multiple positions.

6. The three-dimensional space UAV fire-fighting path planning method based on improved genetic algorithm according to claim 1 is characterized by: The steps of using improved genetic algorithm for UAV firefighting path planning include: An initial population containing a set number of individuals is randomly generated, each of which represents the fire-fighting path and fire-fighting bomb distribution of all drones; The fitness of each individual is calculated based on the total fire extinguishing time, path length, energy consumption, and fire extinguishing success rate; Select individuals with high fitness based on the roulette strategy; Apply multi-point crossover strategy to generate new offspring individuals; Mutate the offspring individuals according to the combined adaptive mutation rate; Add the newly generated offspring individuals to the population to replace the individuals with low fitness; Repeat the above steps until the predetermined number of generations is reached or the increase in fitness is less than the set value, and the final drone fire extinguishing path is obtained.

7. The three-dimensional space UAV fire-fighting path planning method based on improved genetic algorithm according to claim 6 is characterized by: The expression of the fitness function is as follows: Where D(i) is the total fire extinguishing distance of the i-th UAV, T(i) is the total fire extinguishing time of the i-th UAV, and C(i) is the penalty value added by the i-th UAV due to collision.

8. The three-dimensional space UAV fire-fighting path planning method based on improved genetic algorithm according to claim 6 is characterized by: Each drone firefighting path contains a visit sequence from the starting point to each fire source point.

Images