A UAV route planning method and system based on airspace grid

Abstract

The present invention discloses a method and system for UAV route planning based on airspace grid, comprising the following steps: constructing an airspace grid UAV flight risk assessment system; constructing a UAV route risk cost model, calculating each risk cost value of the UAV flying in the grid, obtaining the comprehensive risk cost value of the grid, and generating a UAV flight risk grid map in combination with the longitude and latitude of the grid; constructing UAV route constraints; constructing a UAV dynamic route planning model, using the longitude and latitude of the flight start and end points of the flight mission, the UAV flight risk grid map, the route constraints and the preference characteristics of the flight mission as inputs of the UAV dynamic route planning model, and using an improved ant colony algorithm to solve the optimal route path of the UAV; wherein the preference characteristics of the flight mission include priority safety or priority timeliness. The present invention improves the safety and robustness of route planning and the reliability of the results.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and in particular to a method and system for unmanned aerial vehicle (UAV) route planning based on airspace grid. Background Art

[0002] Current research on UAV route planning mostly uses path planning algorithms such as A* algorithm, artificial potential field, ant colony algorithm, etc., to model and optimize algorithms based on environmental factors such as flight obstacles, so as to achieve risk assessment and route planning for UAV operations. Most existing route planning algorithms only consider obstacle avoidance, and rarely conduct comprehensive analysis of factors such as the compliance of UAV flights with laws and regulations, meteorological conditions, UAV performance, aerial obstacles, and ground damage risks to conduct risk assessment and flight route planning. However, the flight risk of UAVs changes dynamically under the combined effect of the above factors. Only route planning from a single dimension will reduce the feasibility and safety of UAV flights. Summary of the invention

[0003] Purpose of the invention: The purpose of the present invention is to provide a UAV path planning method based on airspace grid, which further improves the safety and robustness of path planning and the reliability of the results.

[0004] The second object of the present invention is to provide a UAV route planning system based on airspace grid.

[0005] Technical solution: The present invention discloses a method for UAV route planning based on airspace grid, comprising the following steps:

[0006] (1) Construct an airspace grid drone flight risk assessment system;

[0007] (2) Construct a risk cost model for drone routes, calculate the risk cost of each drone flight within the grid, obtain the comprehensive risk cost of the grid, and generate a drone flight risk grid map based on the longitude and latitude of the grid;

[0008] (3) Constructing the UAV route constraints;

[0009] (4) A dynamic route planning model for UAVs is constructed. The latitude and longitude of the flight start and end points, the UAV flight risk grid map, route constraints, and the preference characteristics of the flight mission are used as the input of the dynamic route planning model for UAVs. An improved ant colony algorithm is used to solve the optimal route path of the UAV. The preference characteristics of the flight mission include prioritizing safety or prioritizing timeliness.

[0010] Among them, step (1) specifically includes the following steps: uniformly discretizing the continuous airspace into regular hexagonal prisms, rectangular parallelepipeds or cube grids with a minimum inscribed circle diameter of the top surface less than 50 meters and a height less than 50 meters, and constructing a drone flight risk assessment system for the safety risk of drone flight in each grid. The drone flight risk assessment system includes four categories of indicators: meteorological conditions, air obstacles, ground losses and route conditions; meteorological conditions include rainfall, wind speed, fog and snowfall indicators; air obstacles include obstacle quantity, obstacle size and obstacle speed indicators; ground losses include ground crowd density, drone size and drone kinetic energy indicators; route conditions include route conflict and route length indicators, and the route length is the flight distance of the drone from the starting point to the end point.

[0011] Preferably, step (2) specifically includes the following steps:

[0012] (2.1) Considering weather conditions, air obstacles, ground losses and flight conditions, a risk cost model function for UAV flight is constructed. :

[0013] ,

[0014] in, is the grid number, They are the risk cost of meteorological conditions, the risk cost of air obstacles, the risk cost of ground losses and the risk cost of route conditions. They are the weights corresponding to the four risk costs;

[0015] (2.2) Calculation of meteorological risk costs , Risk Cost of Air Obstacles , Ground loss risk cost and route risk cost ;

[0016] (2.3) Determine the risk cost of meteorological conditions , Risk Cost of Air Obstacles , Ground loss risk cost and route risk cost The corresponding weights ;

[0017] (2.3.1) The entropy weight method is used to determine the weight of the risk cost. The steps for determining the weight of the risk cost using the entropy weight method are as follows:

[0018] Constructing the risk cost data matrix :

[0019] ,

[0020] In the formula, For the Time Next The risk cost, Indicates the number of sampling time points, represents the number of risk cost samples at each time point. The four risk costs are: weather condition risk cost, air obstacle risk cost, ground loss risk cost and route condition risk cost. The value of is 4;

[0021] Normalization of the original risk cost data,

[0022] ,

[0023] in, is the standardized risk cost data matrix element, and Respectively represent The risk cost is The maximum and minimum values ​​between time points;

[0024] Calculate the entropy value of risk cost,

[0025] ,

[0026] Calculate the weight of risk cost,

[0027] ,

[0028] (2.3.2) The coefficient of variation method is used to determine the weight of risk cost. The coefficient of variation calculation formula is as follows:

[0029] ,

[0030] in is the coefficient of variation, It is The risk cost is The standard deviation between time points, It is The risk cost is The average value among the time points;

[0031] Risk Cost Weight Determined by the following formula:

[0032] ,

[0033] (2.3.3) Game theory combination weighting, is the risk cost weight vector obtained by entropy weight method, is the risk cost weight vector obtained by the coefficient of variation method, , ;

[0034] Construct a basic risk weight vector and The vector weighted combination is the final weight of the risk cost:

[0035] ,

[0036] in is the risk cost weight vector after combination, , is the transposed vector of the risk cost weight vector obtained by the entropy weight method, is the transposed vector of the risk cost weight vector obtained by the coefficient of variation method, and is the weighting coefficient, , ;

[0037] According to the idea of ​​game theory aggregation model, the objective function is established to combine weights and and The goal is to minimize the distance difference and seek the optimal weighting coefficient and ,Right now:

[0038] ,

[0039] According to the differential properties of the matrix, the optimal first-order derivative condition for obtaining the minimum value of the above formula can be transformed into the following linear equations:

[0040] ,

[0041] In the formula, , , , , solve the linear equations to get the weight coefficients and , and normalize it to get and ,

[0042] ,

[0043] Combining the two weight vectors, we can get the risk cost of meteorological conditions. , air obstacle risk cost , ground loss risk cost and route risk cost The corresponding weight , ;

[0044] ,

[0045] (2.4) After calculating the risk costs and risk cost weights of meteorological conditions, air obstacles, ground losses and route conditions, the comprehensive risk cost value of the current grid can be calculated, and the comprehensive risk cost value of each grid can be associated with the longitude and latitude of the grid center point to generate a risk grid map. The risk value of each grid in the map is dynamically updated in real time as time changes.

[0046] Furthermore, the meteorological risk cost in step (2.2) The calculation formula is:

[0047] ,

[0048] in, They are the levels of strong wind, heavy rain, heavy fog and heavy snow. They are the risk weight factors corresponding to strong wind, heavy rain, heavy fog and heavy snow respectively.

[0049] Furthermore, the risk cost of air obstacles in step (2.2) The calculation formula is:

[0050] ,

[0051] in, The number of obstacles identified by the obstacle perception system on the drone. For the The obstacle volume, For the Obstacle speed.

[0052] Preferably, the ground loss risk in step (2.2) refers to the risk of casualties caused by ground personnel being hit by an out-of-control drone, which is determined by the ground population density in the drone flight area, the size of the drone, and the kinetic energy index of the drone. The ground loss risk cost is The calculation formula is as follows:

[0053] ,

[0054] in, is the number of floating population in the area corresponding to the grid, is the total area of ​​the grid area, is the size of the drone, is the drone volume weight, For the kinetic energy of the drone, is the kinetic energy weight of the drone;

[0055] Drone kinetic energy The calculation formula is:

[0056] ,

[0057] in, For drone quality, The speed at which the drone operates.

[0058] Furthermore, the route risk cost in step (2.2) is divided into route length cost and route conflict cost. Assume that the current grid center position coordinates are , the location coordinates of the next grid center in route planning are , the position coordinates of the center of the flight starting point grid are , then the route length cost is:

[0059] ,

[0060] in, The total number of grids that the route passes through from the starting point to the current position;

[0061] Route conflict cost: The real-time flight data of other drones is obtained from the drone flight control platform. The more drones there are around the drone route, the greater the probability of route conflict and the higher the collision risk cost. The six grids on the side of the grid where the drone is located, and the two grids on the bottom and top of the current grid are used as neighbors to check whether there are other drones passing around the drone route. The calculation method of route conflict cost is:

[0062] ,

[0063] in is the number of drones in the grid neighborhood where the drone is located;

[0064] Route Conditions Risk Cost Cost of route length Cost of route conflict sum:

[0065] ,

[0066] Calculate the risk cost of the route .

[0067] Further, step (3) specifically includes the following steps:

[0068] The UAV route constraints include UAV turning angle constraints, energy constraints, flight area constraints and flight speed constraints;

[0069] The UAV turning angle constraint is: Introducing the turning angle restriction coefficient , in order to improve the oscillation problem of the planned route while avoiding obstacles, The calculation formula is:

[0070] ,

[0071] in, is a vector whose size is the position of the drone and obstacle position The Euclidean distance between them is from the drone position to the obstacle position. is the minimum circumscribed circle diameter of the grid, , restriction coefficient The value range is ;

[0072] The turning angle of the planned route of the drone is calculated based on the position relationship between the previous node and the current node on the planned route of the drone, and the maximum allowable turning angle of the drone is multiplied by the turning angle limit coefficient , compared with the turning angle of the planned route to ensure that the drone can achieve a safe turn within a certain distance tolerance.

[0073] ,

[0074] in, is the turning angle of the planned route, Indicates the maximum allowed turning angle of the drone;

[0075] The energy constraints satisfy:

[0076] ,

[0077] in, is the total energy consumption of the planned route; Energy storage for drones;

[0078] The flight area constraints are:

[0079] Assume the real-time coordinates of the UAV are , then:

[0080] ,

[0081] ,

[0082] ,

[0083] in, They are the minimum and maximum values ​​of the horizontal coordinate, the minimum and maximum values ​​of the vertical coordinate, and the minimum and maximum values ​​of the altitude in the specified unmanned airspace;

[0084] The flight speed constraint is:

[0085] ,

[0086] in, , They are respectively the minimum and maximum flight speeds in the specified airspace where the UAV is located.

[0087] Preferably, step (4) specifically includes the following steps:

[0088] (4.1) Three-dimensional modeling of the UAV flight space is performed to form a three-dimensional grid. The three-dimensional grid division method is the same as the grid division method in step (1);

[0089] (4.2) Ant colony algorithm parameter initialization and pheromone concentration initialization;

[0090] (4.3) The ants start to move from the starting point, select the next grid node based on the generated transfer probability and roulette method, and update the pheromone;

[0091] (4.4) Determine whether to add a grid node to the route planning path based on the risk cost value of the grid node, route constraints and flight preference: When the flight preference prioritizes safety, the next node of the flight selects the grid node with the smallest risk cost value that meets the constraints; When the flight preference prioritizes timeliness, the next node of the flight selects the node that meets the constraints, has a risk cost below the threshold and has the smallest route path;

[0092] (4.5) During the flight, the system determines whether to initiate a new round of route planning based on the UAV position and altitude information provided by the onboard navigation system. The logic judgment is based on the UAV altitude deviation and horizontal deviation. When the deviation distance is greater than the set threshold, a new round of route planning is initiated.

[0093] Based on the same inventive concept, the present invention discloses a UAV route planning system based on airspace grid, comprising:

[0094] The risk assessment system module is used to build an airspace grid drone flight risk assessment system;

[0095] The risk grid map module is used to build a risk cost model for drone routes, calculate each risk cost value of drones flying within the grid, obtain the comprehensive risk cost value of the grid, and generate a drone flight risk grid map based on the longitude and latitude of the grid;

[0096] Constraint module, used to construct UAV route constraints;

[0097] The dynamic route planning module is used to build a dynamic route planning model for UAVs. The longitude and latitude of the flight start and end points of the flight mission, the UAV flight risk grid map, route constraints and the preference characteristics of the flight mission are used as the input of the UAV dynamic route planning model, and the improved ant colony algorithm is used to solve the optimal route path of the UAV; the preference characteristics of the flight mission include prioritizing safety or prioritizing timeliness.

[0098] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: first, the present invention divides the airspace into a three-dimensional grid, comprehensively considers the safety, economy and feasibility of UAV flight within the grid, and establishes a UAV flight risk assessment index system within the grid; secondly, a route risk cost model integrating factors such as meteorological conditions, air obstacles, ground damage risks and route length is constructed, and the route risk cost value is combined with the geographical location of the grid to generate a risk grid map; thirdly, in order to solve the problem of fuzziness of indicator weights in the route cost model, a game theory combined weighting model is constructed, and the entropy method and the coefficient of variation method are combined to recalculate the risk cost weight, which makes up for the shortcomings of a single weighting method and helps to make the calculation result of the route cost more reasonable and objective; and a UAV route planning constraint condition integrating UAV flight airspace range restrictions, speed restrictions, maximum allowable turning angles, etc. is constructed to improve the safety and robustness of route planning; thirdly, based on the improved ant colony algorithm, a real-time dynamic route planning mechanism for UAVs is designed. During the flight of the UAV, the data of the grid where the UAV is currently located can be called in real time according to the deviation between the actual coordinates of the UAV and the route planning position to update the route planning. BRIEF DESCRIPTION OF THE DRAWINGS

[0099] Figure 1 is a flow chart of the present invention;

[0100] Figure 2 It is a system principle block diagram of the present invention;

[0101] Figure 3 This is a diagram of the UAV flight risk assessment system of the present invention;

[0102] Figure 4 This is a flow chart of the improved ant colony algorithm of the present invention. DETAILED DESCRIPTION

[0103] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0104] Example 1: Figure 1 As shown, this embodiment discloses a method for UAV route planning based on airspace grid, comprising the following steps:

[0105] (1) Construct an airspace grid drone flight risk assessment system;

[0106] Step (1) specifically includes the following steps:

[0107] The continuous airspace is evenly discretized into regular hexagonal prisms, cuboids or cube grids with a minimum inscribed circle diameter of less than 50m and a height of less than 50m. According to the control requirements of different types of drones in different types of airspace in relevant regulations and combined with the actual operation of drones, a drone flight risk assessment system is constructed for the safety risks of drone flights in each grid. The drone flight risk assessment system includes four categories of indicators: meteorological conditions, air obstacles, ground losses and route conditions. Meteorological conditions include rainfall, wind speed, fog and snowfall indicators. Air obstacles include obstacle quantity, obstacle size and obstacle speed indicators. Ground losses include ground crowd density, drone size and drone kinetic energy indicators. Route conditions include route conflicts and route length indicators. The route length is the flight distance of the drone from the starting point to the end point. The drone flight risk assessment system is as follows: Figure 3 As shown;

[0108] (2) Construct a risk cost model for UAV routes and calculate the risk cost function value of UAV flying within the grid;

[0109] (2.1) Considering weather conditions, air obstacles, ground losses and flight conditions, a risk cost model function for UAV flight is constructed. :

[0110] ,

[0111] in, is the grid number, They are the risk cost of meteorological conditions, the risk cost of air obstacles, the risk cost of ground losses and the risk cost of route conditions. They are the weights corresponding to the four risk costs;

[0112] (2.2) Calculation of meteorological risk costs , Risk Cost of Air Obstacles , Ground loss risk cost and route risk cost ;

[0113] Severe weather conditions, such as strong winds, heavy rain, fog and snow, may have a negative impact on the performance of drones and reduce their countermeasures. For example, strong winds may cause drones to deviate from their planned routes or even lose control. Heavy rain and snow may affect the circuits and electronic components of drone countermeasures, causing performance degradation or failure. The calculation formula is:

[0114] ,

[0115] in, They are the levels of strong wind, heavy rain, heavy fog and heavy snow. They are the risk weight factors corresponding to strong wind, heavy rain, heavy fog and heavy snow respectively.

[0116] The corresponding relationship between the levels of strong wind, heavy rain, heavy fog and heavy snow and the risk weight factors is shown in Table 1.

[0117] Table 1 Correspondence between wind, rain, fog and snow levels and risk weights

[0118]

[0119] During the flight of a drone, obstacles such as birds and other aircraft may suddenly appear in a local area. This sudden information is difficult to predict and may cause a collision, resulting in serious flight safety risks. The risk cost of air obstacles is The calculation formula is:

[0120] ,

[0121] in, The number of obstacles identified by the obstacle perception system on the drone. For the The obstacle volume, For the Obstacle speed.

[0122] Ground loss risk refers to the risk of casualties caused by the uncontrolled drone hitting the ground personnel. It is determined by the ground population density, the size of the drone and the kinetic energy index of the drone in the flight area. The ground loss risk cost is The calculation formula is as follows:

[0123] ,

[0124] in, is the number of floating population in the area corresponding to the grid, is the total area of ​​the grid area, is the size of the drone, is the drone volume weight, For the kinetic energy of the drone, is the kinetic energy weight of the drone;

[0125] Drone kinetic energy The calculation formula is:

[0126] ,

[0127] in, For drone quality, The speed at which the drone operates.

[0128] The route risk cost is divided into route length cost and route conflict cost. Assume that the current grid center position coordinates are , the location coordinates of the next grid center in route planning are , the position coordinates of the center of the flight starting point grid are , then the route length cost is:

[0129] ,

[0130] in, The total number of grids that the route passes through from the starting point to the current position;

[0131] Route conflict cost: The real-time flight data of other drones is obtained from the drone flight control platform. The more drones there are around the drone route, the greater the probability of route conflict and the higher the collision risk cost. The six grids on the side of the grid where the drone is located, and the two grids on the bottom and top of the current grid are used as neighbors to check whether there are other drones passing around the drone route. The calculation method of route conflict cost is:

[0132] ,

[0133] in is the number of drones in the grid neighborhood where the drone is located;

[0134] Route Conditions Risk Cost Cost of route length Cost of route conflict sum:

[0135] ,

[0136] Calculate the risk cost of the route .

[0137] (2.3) Determine the risk cost of meteorological conditions , Risk Cost of Air Obstacles , Ground loss risk cost and route risk cost The corresponding weights ;

[0138] The entropy method and the coefficient of variation method are used to calculate the weight of the risk cost respectively, and the weights obtained by the two weighting methods are weighted and combined based on the game theory algorithm, which can eliminate the deviation caused by a single weighting method and improve the reliability of the UAV flight risk cost calculation;

[0139] (2.3.1) The entropy weight method is used to determine the weight of the risk cost. The entropy weight method is an objective weighting method that can effectively eliminate the influence of subjective factors. The larger the entropy value, the smaller the variation of the risk cost value, the less effective information provided, and the smaller the weight. The steps for determining the risk cost weight using the entropy weight method are as follows:

[0140] Constructing the risk cost data matrix :

[0141] ,

[0142] In the formula, For the Time Next The risk cost, Indicates the number of sampling time points, represents the number of risk cost samples at each time point. The four risk costs are: weather condition risk cost, air obstacle risk cost, ground loss risk cost and route condition risk cost. The value of is 4;

[0143] Normalization of the original risk cost data,

[0144] ,

[0145] in, is the standardized risk cost data matrix element, and Respectively represent The risk cost is The maximum and minimum values ​​between time points;

[0146] Calculate the entropy value of risk cost,

[0147] ,

[0148] Calculate the weight of risk cost,

[0149] ,

[0150] (2.3.2) The coefficient of variation method is used to determine the weight of the risk cost. Since the applicability of the entropy weight method is affected by the dimensionless method, the problem of weight balancing cannot be avoided and sometimes it cannot accurately represent the role of each indicator in the evaluation. Therefore, the coefficient of variation method is introduced to circumvent this problem. The calculation formula of the coefficient of variation is as follows:

[0151] ,

[0152] in is the coefficient of variation, It is The risk cost is The standard deviation between time points, It is The risk cost is The average value among the time points;

[0153] Risk Cost Weight Determined by the following formula:

[0154] ,

[0155] (2.3.3) Game theory combined weighting, with Nash equilibrium as the coordination goal, is a process of mutual comparison and coordination to find the consistency and compromise between the two weighting methods. It can find the maximum common interests of the two, thereby improving the scientific rationality of the risk cost function indicator weighting; is the risk cost weight vector obtained by entropy weight method, is the risk cost weight vector obtained by the coefficient of variation method, , ;

[0156] Construct a basic risk weight vector and The vector weighted combination is the final weight of the risk cost:

[0157] ,

[0158] in is the risk cost weight vector after combination, , is the transposed vector of the risk cost weight vector obtained by the entropy weight method, is the transposed vector of the risk cost weight vector obtained by the coefficient of variation method, and is the weighting coefficient, , ;

[0159] According to the idea of ​​game theory aggregation model, the objective function is established to combine weights and and The goal is to minimize the distance difference and seek the optimal weighting coefficient and ,Right now:

[0160] ,

[0161] According to the differential properties of the matrix, the optimal first-order derivative condition for obtaining the minimum value of the above formula can be transformed into the following linear equations:

[0162] ,

[0163] In the formula, , , , , solve the linear equations to get the weight coefficients and , and normalize it to get and ,

[0164] ,

[0165] Combining the two weight vectors, we can get the risk cost of meteorological conditions. , air obstacle risk cost , ground loss risk cost and route risk cost The corresponding weight , ;

[0166] ,

[0167] (2.4) After calculating the risk costs and risk cost weights of meteorological conditions, air obstacles, ground losses and route conditions, the comprehensive risk cost value of the current grid can be calculated as:

[0168]

[0169] in, is the grid number, They are the risk cost of meteorological conditions, the risk cost of air obstacles, the risk cost of ground losses and the risk cost of route conditions. They are the weights corresponding to the four risk costs;

[0170] The comprehensive risk cost value of each grid is associated with the longitude and latitude of the center point of the grid to generate a risk grid map. The risk value of each grid in the map is dynamically updated in real time as time changes.

[0171] (3) Constructing the UAV route constraints;

[0172] Step (3) The UAV route constraints include UAV turning angle constraints, energy constraints, flight area constraints, and flight speed constraints;

[0173] The UAV turning angle constraint is: Introduce the turning angle limit coefficient , in order to improve the oscillation problem of the planned route while avoiding obstacles, The calculation formula is:

[0174] ,

[0175] in, is a vector whose size is the position of the drone and obstacle position The Euclidean distance between them is from the drone position to the obstacle position. is the minimum circumscribed circle diameter of the grid, , restriction coefficient The value range is ;

[0176] The turning angle of the planned route of the drone is calculated based on the position relationship between the previous node and the current node on the planned route of the drone, and the maximum allowable turning angle of the drone is multiplied by the turning angle limit coefficient , compared with the turning angle of the planned route to ensure that the drone can achieve a safe turn within a certain distance tolerance.

[0177] ,

[0178] in, is the turning angle of the planned route, Indicates the maximum allowed turning angle of the drone;

[0179] The energy constraints satisfy:

[0180] ,

[0181] in, is the total energy consumption of the planned route; Energy storage for drones;

[0182] The flight area constraints are:

[0183] Assume the real-time coordinates of the UAV are , then:

[0184] ,

[0185] ,

[0186] ,

[0187] in, They are the minimum and maximum values ​​of the horizontal coordinate, the minimum and maximum values ​​of the vertical coordinate, and the minimum and maximum values ​​of the altitude required by the Interim Regulations on the Flight Management of Unmanned Aerial Vehicles in unmanned airspace;

[0188] The flight speed constraint is:

[0189] ,

[0190] in, , They are respectively the minimum and maximum flight speeds in the airspace where drones are located in the Interim Regulations on Unmanned Aerial Vehicle Flight Management.

[0191] (4) Construct a UAV dynamic route planning model, using the longitude and latitude of the flight start and end points of the flight mission, the UAV flight risk grid map, flight constraints, and the preference characteristics of the flight mission as the input of the UAV dynamic route planning model, and use the improved ant colony algorithm to solve the optimal route path of the UAV; the preference characteristics of the flight mission include giving priority to the shortest flight time or the highest safety;

[0192] Step (4) specifically includes the following steps:

[0193] (4.1) Three-dimensional modeling of the UAV flight space is performed to form a three-dimensional grid. The three-dimensional grid division method is the same as the grid division method in step (1);

[0194] (4.2) Ant colony algorithm parameter initialization and pheromone concentration initialization;

[0195] (4.3) The ants start to move from the starting point, select the next grid node based on the generated transfer probability and roulette method, and update the pheromone;

[0196] (4.4) Determine whether to add a grid node to the route planning path based on the risk cost value of the grid node, route constraints and flight preference: When the flight preference prioritizes safety, the next node of the flight selects the grid node with the smallest risk cost value that meets the constraints; When the flight preference prioritizes timeliness, the next node of the flight selects the node that meets the constraints, has a risk cost below the threshold and has the smallest route path;

[0197] (4.5) During the flight, the system determines whether to initiate a new round of route planning based on the UAV position and altitude information provided by the onboard navigation system. The logic judgment is based on the UAV altitude deviation and horizontal deviation. When the deviation distance is greater than the set threshold, a new round of route planning is initiated.

[0198] The ant colony algorithm is an intelligent bionic algorithm inspired by the collective path-finding and foraging behavior of ants. The ant colony algorithm simulates the behavior process of ants in selecting the shortest path. It uses a distributed population search strategy to solve the optimal solution to the problem. It integrates grid risk cost evaluation and route constraints into the ant colony algorithm to obtain a safer, more reliable and economical UAV dynamic route planning algorithm.

[0199] Embodiment 2: Based on the same inventive concept, this embodiment 2 discloses a UAV route planning system based on airspace grid, including:

[0200] The risk assessment system module is used to build an airspace grid drone flight risk assessment system;

[0201] In the risk assessment system module, the continuous airspace is evenly discretized into regular hexagonal prisms, cuboids or cube grids with a minimum inscribed circle diameter of less than 50m and a height of less than 50m. According to the relevant regulations on the control of different types of drones in different types of airspace, combined with the actual operation of drones, a drone flight risk assessment system is constructed for the safety risks of drone flights in each grid. The drone flight risk assessment system includes four categories of indicators: meteorological conditions, air obstacles, ground losses and route conditions; meteorological conditions include rainfall, wind speed, fog and snowfall indicators; air obstacles include obstacle quantity, obstacle size and obstacle speed indicators; ground losses include ground crowd density, drone size and drone kinetic energy indicators; route conditions include route conflicts and route length indicators. The route length is the flight distance of the drone from the starting point to the end point. The drone flight risk assessment system is as follows: Figure 3 As shown;

[0202] The risk grid map module is used to build a risk cost model for drone routes, calculate each risk cost value of drones flying within the grid, obtain the comprehensive risk cost value of the grid, and generate a drone flight risk grid map based on the longitude and latitude of the grid;

[0203] The risk grid map module comprehensively considers meteorological conditions, air obstacles, ground losses and route conditions to construct a UAV flight risk cost model function. :

[0204] ,

[0205] in, is the grid number, They are the risk cost of meteorological conditions, the risk cost of air obstacles, the risk cost of ground losses and the risk cost of route conditions. They are the weights corresponding to the four risk costs;

[0206] Calculating the risk cost of weather conditions , Risk Cost of Air Obstacles , Ground loss risk cost and route risk cost ;

[0207] Severe weather conditions, such as strong winds, heavy rain, fog and snow, may have a negative impact on the performance of drones and reduce their countermeasures. For example, strong winds may cause drones to deviate from their planned routes or even lose control. Heavy rain and snow may affect the circuits and electronic components of drone countermeasures, causing performance degradation or failure. The calculation formula is:

[0208] ,

[0209] in, They are the levels of strong wind, heavy rain, heavy fog and heavy snow. They are the risk weight factors corresponding to strong wind, heavy rain, heavy fog and heavy snow respectively.

[0210] The corresponding relationship between the levels of strong wind, heavy rain, heavy fog and heavy snow and the risk weight factors is shown in Table 1 of Example 1.

[0211] During the flight of a drone, obstacles such as birds and other aircraft may suddenly appear in a local area. This sudden information is difficult to predict and may cause a collision, resulting in serious flight safety risks. The risk cost of air obstacles is The calculation formula is:

[0212] ,

[0213] in, The number of obstacles identified by the obstacle perception system on the drone. For the The obstacle volume, For the Obstacle speed.

[0214] Ground loss risk refers to the risk of casualties caused by the uncontrolled drone hitting the ground personnel. It is determined by the ground population density, the size of the drone and the kinetic energy index of the drone in the flight area. The ground loss risk cost is The calculation formula is as follows:

[0215] ,

[0216] in, is the number of floating population in the area corresponding to the grid, is the total area of ​​the grid area, is the size of the drone, is the drone volume weight, For the kinetic energy of the drone, is the kinetic energy weight of the drone;

[0217] Drone kinetic energy The calculation formula is:

[0218] ,

[0219] in, For drone quality, The speed at which the drone operates.

[0220] The route risk cost is divided into route length cost and route conflict cost. Assume that the current grid center position coordinates are , the location coordinates of the next grid center in route planning are , the position coordinates of the center of the flight starting point grid are , then the route length cost is:

[0221] ,

[0222] in, The total number of grids that the route passes through from the starting point to the current position;

[0223] Route conflict cost: The real-time flight data of other drones is obtained from the drone flight control platform. The more drones there are around the drone route, the greater the probability of route conflict and the higher the collision risk cost. The six grids on the side of the grid where the drone is located, and the two grids on the bottom and top of the current grid are used as neighbors to check whether there are other drones passing around the drone route. The calculation method of route conflict cost is:

[0224] ,

[0225] in is the number of drones in the grid neighborhood where the drone is located;

[0226] Route Conditions Risk Cost Cost of route length Cost of route conflict sum:

[0227] ,

[0228] Calculate the risk cost of the route .

[0229] Determine the risk cost of meteorological conditions , Risk Cost of Air Obstacles , Ground loss risk cost and route risk cost The corresponding weights ;

[0230] The entropy method and the coefficient of variation method are used to calculate the weight of the risk cost respectively, and the weights obtained by the two weighting methods are weighted and combined based on the game theory algorithm, which can eliminate the deviation caused by a single weighting method and improve the reliability of the UAV flight risk cost calculation;

[0231] The entropy weight method determines the weight of the risk cost. The entropy weight method is an objective weighting method that can effectively eliminate the influence of subjective factors. The larger the entropy value, the smaller the variation of the risk cost value, the less effective information provided, and the smaller the weight. The steps for determining the risk cost weight using the entropy weight method are as follows:

[0232] Constructing the risk cost data matrix :

[0233] ,

[0234] In the formula, For the Time Next The risk cost, Indicates the number of sampling time points, represents the number of risk cost samples at each time point. The four risk costs are: weather condition risk cost, air obstacle risk cost, ground loss risk cost and route condition risk cost. The value of is 4;

[0235] Normalization of the original risk cost data,

[0236] ,

[0237] in, is the standardized risk cost data matrix element, and Respectively represent The risk cost is The maximum and minimum values ​​between time points;

[0238] Calculate the entropy value of risk cost,

[0239] ,

[0240] Calculate the weight of risk cost,

[0241] ,

[0242] The coefficient of variation method determines the weight of the risk cost. Since the applicability of the entropy weight method is affected by the dimensionless method, the problem of weight balancing cannot be avoided. Sometimes it cannot accurately represent the role of each indicator in the evaluation. Therefore, the coefficient of variation method is introduced to circumvent it. The calculation formula of the coefficient of variation is as follows:

[0243] ,

[0244] in is the coefficient of variation, It is The risk cost is The standard deviation between time points, It is The risk cost is The average value among the time points;

[0245] Risk Cost Weight Determined by the following formula:

[0246] ,

[0247] Game theory combined weighting, with Nash equilibrium as the coordination goal, is a process of mutual comparison and coordination to find the consistency and compromise between the two weighting methods. It can find the maximum common interests of the two, thus improving the scientific rationality of the risk cost function indicator weighting. is the risk cost weight vector obtained by entropy weight method, is the risk cost weight vector obtained by the coefficient of variation method, , ;

[0248] Construct a basic risk weight vector and The vector weighted combination is the final weight of the risk cost:

[0249] ,

[0250] in is the risk cost weight vector after combination, , is the transposed vector of the risk cost weight vector obtained by the entropy weight method, is the transposed vector of the risk cost weight vector obtained by the coefficient of variation method, and is the weighting coefficient, , ;

[0251] According to the idea of ​​game theory aggregation model, the objective function is established to combine weights and and The goal is to minimize the distance difference and seek the optimal weighting coefficient and ,Right now:

[0252] ,

[0253] According to the differential properties of the matrix, the optimal first-order derivative condition for obtaining the minimum value of the above formula can be transformed into the following linear equations:

[0254] ,

[0255] In the formula, , , , , solve the linear equations to get the weight coefficients and , and normalize it to get and ,

[0256] ,

[0257] Combining the two weight vectors, we can get the risk cost of meteorological conditions. , air obstacle risk cost , ground loss risk cost and route risk cost The corresponding weight , ;

[0258] ,

[0259] After calculating the risk costs and risk cost weights of meteorological conditions, air obstacles, ground losses and route conditions, the comprehensive risk cost value of the current grid can be calculated as:

[0260]

[0261] in, is the grid number, They are the risk cost of meteorological conditions, the risk cost of air obstacles, the risk cost of ground losses and the risk cost of route conditions. They are the weights corresponding to the four risk costs;

[0262] The comprehensive risk cost value of each grid is associated with the longitude and latitude of the center point of the grid to generate a risk grid map. The risk value of each grid in the map is dynamically updated in real time as time changes.

[0263] Constraint module, used to construct UAV route constraints;

[0264] The UAV route constraints in the constraint module include UAV turning angle constraints, energy constraints, flight area constraints, and flight speed constraints;

[0265] The UAV turning angle constraint is: Introduce the turning angle limit coefficient , in order to improve the oscillation problem of the planned route while avoiding obstacles, The calculation formula is:

[0266] ,

[0267] in, is a vector whose size is the position of the drone and obstacle position The Euclidean distance between them is from the drone position to the obstacle position. is the minimum circumscribed circle diameter of the grid, , restriction coefficient The value range is ;

[0268] The turning angle of the planned route of the drone is calculated based on the position relationship between the previous node and the current node on the planned route of the drone, and the maximum allowable turning angle of the drone is multiplied by the turning angle limit coefficient , compared with the turning angle of the planned route to ensure that the drone can achieve a safe turn within a certain distance tolerance.

[0269] ,

[0270] in, is the turning angle of the planned route, Indicates the maximum allowed turning angle of the drone;

[0271] The energy constraints satisfy:

[0272] ,

[0273] in, is the total energy consumption of the planned route; Energy storage for drones;

[0274] The flight area constraints are:

[0275] Assume the real-time coordinates of the UAV are , then:

[0276] ,

[0277] ,

[0278] ,

[0279] in, They are the minimum and maximum values ​​of the horizontal coordinate, the minimum and maximum values ​​of the vertical coordinate, and the minimum and maximum values ​​of the altitude required by the Interim Regulations on the Flight Management of Unmanned Aerial Vehicles in unmanned airspace;

[0280] The flight speed constraint is:

[0281] ,

[0282] in, , They are respectively the minimum and maximum flight speeds in the airspace where drones are located in the Interim Regulations on Unmanned Aerial Vehicle Flight Management.

[0283] The dynamic route planning module is used to build a dynamic route planning model for UAVs. The longitude and latitude of the flight start and end points of the flight mission, the UAV flight risk grid map, route constraints and the preference characteristics of the flight mission are used as the input of the UAV dynamic route planning model, and the improved ant colony algorithm is used to solve the optimal route path of the UAV; the preference characteristics of the flight mission include prioritizing safety or prioritizing timeliness.

[0284] In the dynamic route planning module, the UAV flight space is three-dimensionally modeled to form a three-dimensional grid. The three-dimensional grid division method is the same as the grid division method in the risk assessment system module; the ant colony algorithm parameters are initialized, and the pheromone concentration is initialized; the ants start to transfer from the starting point, select the next grid node according to the generated transfer probability and the roulette method, and update the pheromone; according to the risk cost value of the grid node, the route constraint conditions and the flight preference, it is determined whether to add the grid node to the route planning path: when the flight preference is to prioritize safety, the next node of the flight selects the grid node with the smallest risk cost value that meets the constraint conditions; when the flight preference is to prioritize timeliness, the next node of the flight selects the node that meets the constraint conditions, the risk cost is lower than the threshold and the route path is the smallest; during the flight, according to the UAV position and altitude information provided by the airborne navigation system and the decision system whether to start a new round of route planning, the basis for logical judgment is the height deviation and horizontal deviation of the UAV. When the deviation distance is greater than the set threshold, a new round of route planning is started.

Claims

1. A method for UAV route planning based on airspace grid, characterized in that: The steps include: (1) Construct an airspace grid drone flight risk assessment system; The step (1) specifically comprises the following steps: uniformly discretizing the continuous airspace into regular hexagonal prisms, cuboids or cube grids with a minimum inscribed circle diameter of less than 50 m and a height of less than 50 m, and constructing a drone flight risk assessment system for the safety risk of drone flight in each grid, wherein the drone flight risk assessment system comprises four categories of indicators: meteorological conditions, air obstacles, ground losses and route conditions; meteorological conditions comprise indicators of rainfall, wind speed, fog and snowfall; air obstacles comprise indicators of the number of obstacles, obstacle size and obstacle speed; ground losses comprise indicators of ground crowd density, drone size and drone kinetic energy; route conditions comprise indicators of route conflict and route length, wherein the route length is the flight distance of the drone from the starting point to the end point; (2) Construct a risk cost model for drone routes, calculate the risk cost of each drone flight within the grid, obtain the comprehensive risk cost of the grid, and generate a drone flight risk grid map based on the longitude and latitude of the grid; The step (2) specifically includes the following steps: (2.1) Considering weather conditions, air obstacles, ground losses and flight conditions, a risk cost model function for UAV flight is constructed. : ， in, is the grid number, They are the risk cost of meteorological conditions, the risk cost of air obstacles, the risk cost of ground losses and the risk cost of route conditions. They are the weights corresponding to the four risk costs; (2.2) Calculation of meteorological risk costs , Risk Cost of Air Obstacles , Ground loss risk cost and route risk cost ; (2.3) Determine the risk cost of meteorological conditions , Risk Cost of Air Obstacles , Ground loss risk cost and route risk cost The corresponding weights ; (2.3.1) The entropy weight method is used to determine the weight of the risk cost. The steps for determining the weight of the risk cost using the entropy weight method are as follows: Constructing the risk cost data matrix : ， In the formula, For the Time Next The risk cost, Indicates the number of sampling time points, represents the number of risk cost samples at each time point. The four risk costs are: weather condition risk cost, air obstacle risk cost, ground loss risk cost and route condition risk cost. The value of is 4; Normalization of the original risk cost data, ， in, is the standardized risk cost data matrix element, and Respectively represent The risk cost is The maximum and minimum values ​​between time points; Calculate the entropy value of risk cost, ， Calculate the weight of risk cost, ， (2.3.2) The coefficient of variation method is used to determine the weight of risk cost. The coefficient of variation calculation formula is as follows: ， in is the coefficient of variation, It is The risk cost is The standard deviation between time points, It is The risk cost is The average value among the time points; Risk Cost Weight Determined by the following formula: ， (2.3.3) Game theory combination weighting, is the risk cost weight vector obtained by entropy weight method, is the risk cost weight vector obtained by the coefficient of variation method, , ; Construct a basic risk weight vector and The vector weighted combination is the final weight of the risk cost: ， in is the risk cost weight vector after combination, , is the transposed vector of the risk cost weight vector obtained by the entropy weight method, is the transposed vector of the risk cost weight vector obtained by the coefficient of variation method, and is the weighting coefficient, , ; According to the idea of ​​game theory aggregation model, the objective function is established to combine weights and and The goal is to minimize the distance difference and seek the optimal weighting coefficient and ,Right now: ， According to the differential properties of the matrix, the optimal first-order derivative condition for obtaining the minimum value of the above formula can be transformed into the following linear equations: ， In the formula, , , , , solve the linear equations to get the weight coefficients and , and normalize it to get and , ， Combining the two weight vectors, we can get the risk cost of meteorological conditions. , air obstacle risk cost , ground loss risk cost and route risk cost The corresponding weight , ; ， (2.4) After calculating the risk costs and risk cost weights of meteorological conditions, air obstacles, ground losses and route conditions, the comprehensive risk cost value of the current grid can be calculated, and the comprehensive risk cost value of each grid is associated with the longitude and latitude of the grid center point to generate a risk grid map. The risk value of each grid in the map is dynamically updated in real time as time changes; (3) Constructing the UAV route constraints; (4) A dynamic route planning model for UAVs is constructed. The latitude and longitude of the flight start and end points, the UAV flight risk grid map, route constraints, and the preference characteristics of the flight mission are used as the input of the dynamic route planning model for UAVs. An improved ant colony algorithm is used to solve the optimal route path of the UAV. The preference characteristics of the flight mission include prioritizing safety or prioritizing timeliness.

2. The method for UAV route planning based on airspace grid according to claim 1, characterized in that: The risk cost of meteorological conditions in step (2.2) The calculation formula is: ， in, They are the levels of strong wind, heavy rain, heavy fog and heavy snow. They are the risk weight factors corresponding to strong wind, heavy rain, heavy fog and heavy snow respectively.

3. The method for UAV route planning based on airspace grid according to claim 1, characterized in that: The risk cost of air obstacles in step (2.2) The calculation formula is: ， in, The number of obstacles identified by the obstacle perception system on the drone. For the The obstacle volume, For the Obstacle speed.

4. The method for UAV route planning based on airspace grid according to claim 1, characterized in that: The ground loss risk in step (2.2) refers to the risk of casualties caused by the uncontrolled drone hitting the ground personnel. It is determined by the ground population density, the size of the drone and the kinetic energy index of the drone in the flight area. The ground loss risk cost is The calculation formula is as follows: ， in, is the number of floating population in the area corresponding to the grid, is the total area of ​​the grid area, is the size of the drone, is the drone volume weight, For the kinetic energy of the drone, is the kinetic energy weight of the drone; Drone kinetic energy The calculation formula is: ， in, For drone quality, The speed at which the drone operates.

5. The method for UAV route planning based on airspace grid according to claim 1, characterized in that: The route risk cost in step (2.2) is divided into route length cost and route conflict cost. Assume that the current grid center position coordinates are , the location coordinates of the next grid center in route planning are , the position coordinates of the center of the flight starting point grid are , then the route length cost is: ， in, The total number of grids that the route passes through from the starting point to the current position; Route conflict cost: The real-time flight data of other drones is obtained from the drone flight control platform. The more drones there are around the drone route, the greater the probability of route conflict and the higher the collision risk cost. The six grids on the side of the grid where the drone is located, and the two grids on the bottom and top of the current grid are used as neighbors to check whether there are other drones passing around the drone route. The calculation method of route conflict cost is: ， in is the number of drones in the grid neighborhood where the drone is located; Route Conditions Risk Cost Cost of route length Cost of route conflict sum: ， Calculate the risk cost of the route .

6. The method for UAV route planning based on airspace grid according to claim 1, characterized in that: The step (3) specifically includes the following steps: The UAV route constraints include UAV turning angle constraints, energy constraints, flight area constraints and flight speed constraints; The UAV turning angle constraint is: Introducing the turning angle restriction coefficient , in order to improve the oscillation problem of the planned route while avoiding obstacles, The calculation formula is: ， in, is a vector whose size is the position of the drone and obstacle position The Euclidean distance between them is from the drone position to the obstacle position. is the minimum circumscribed circle diameter of the grid, , restriction coefficient The value range is ; The turning angle of the planned route of the drone is calculated based on the position relationship between the previous node and the current node on the planned route of the drone, and the maximum allowable turning angle of the drone is multiplied by the turning angle limit coefficient , compared with the turning angle of the planned route to ensure that the drone can make a safe turn, ， in, is the turning angle of the planned route, Indicates the maximum allowed turning angle of the drone; The energy constraints satisfy: ， in, is the total energy consumption of the planned route; Energy storage for drones; The flight area constraints are: Assume the real-time coordinates of the UAV are , then: ， ， ， in, They are the minimum and maximum values ​​of the horizontal coordinate, the minimum and maximum values ​​of the vertical coordinate, and the minimum and maximum values ​​of the altitude in the specified unmanned airspace; The flight speed constraint is: ， in, , They are respectively the minimum and maximum flight speeds in the specified airspace where the drone is located.

7. The method for UAV route planning based on airspace grid according to claim 1, characterized in that: The step (4) specifically includes the following steps: (4.1) Three-dimensional modeling of the UAV flight space is performed to form a three-dimensional grid. The three-dimensional grid division method is the same as the grid division method in step (1); (4.2) Ant colony algorithm parameter initialization and pheromone concentration initialization; (4.3) The ants start to move from the starting point, select the next grid node based on the generated transfer probability and roulette method, and update the pheromone; (4.4) Determine whether to add a grid node to the route planning path based on the risk cost value of the grid node, route constraints and flight preference: When the flight preference prioritizes safety, the next node of the flight selects the grid node with the smallest risk cost value that meets the constraints; When the flight preference prioritizes timeliness, the next node of the flight selects the node that meets the constraints, has a risk cost below the threshold and has the smallest route path; (4.5) During the flight, the system determines whether to initiate a new round of route planning based on the UAV position and altitude information provided by the onboard navigation system. The logic judgment is based on the UAV altitude deviation and horizontal deviation. When the deviation distance is greater than the set threshold, a new round of route planning is initiated.

8. A UAV route planning system based on airspace grid, characterized in that: include: The risk assessment system module is used to build an airspace grid drone flight risk assessment system; In the risk assessment system module, the continuous airspace is evenly discretized into regular hexagonal prisms, cuboids or cube grids with a minimum inscribed circle diameter of less than 50m and a height of less than 50m. A drone flight risk assessment system is constructed for the safety risk of drone flight in each grid. The drone flight risk assessment system includes four categories of indicators: meteorological conditions, air obstacles, ground losses and route conditions. Meteorological conditions include rainfall, wind speed, fog and snowfall indicators. Air obstacles include obstacle quantity, obstacle size and obstacle speed indicators. Ground losses include ground crowd density, drone size and drone kinetic energy indicators. Route conditions include route conflicts and route length indicators. The route length is the flight distance of the drone from the starting point to the end point. The risk grid map module is used to build a risk cost model for drone routes, calculate each risk cost value of drones flying within the grid, obtain the comprehensive risk cost value of the grid, and generate a drone flight risk grid map based on the longitude and latitude of the grid; The risk grid map module comprehensively considers meteorological conditions, air obstacles, ground losses and route conditions to construct a UAV flight risk cost model function. : ， in, is the grid number, They are the risk cost of meteorological conditions, the risk cost of air obstacles, the risk cost of ground losses and the risk cost of route conditions. They are the weights corresponding to the four risk costs; Calculating the risk cost of weather conditions , Risk Cost of Air Obstacles , Ground loss risk cost and route risk cost ; Determine the risk cost of meteorological conditions , Risk Cost of Air Obstacles , Ground loss risk cost and route risk cost The corresponding weights ; The entropy weight method determines the weight of the risk cost. The steps for determining the risk cost weight by the entropy weight method are as follows: Constructing the risk cost data matrix : ， In the formula, For the Time Next The risk cost, Indicates the number of sampling time points, represents the number of risk cost samples at each time point. The four risk costs are: weather condition risk cost, air obstacle risk cost, ground loss risk cost and route condition risk cost. The value of is 4; Normalization of the original risk cost data, ， in, is the standardized risk cost data matrix element, and Respectively represent The risk cost is The maximum and minimum values ​​between time points; Calculate the entropy value of risk cost, ， Calculate the weight of risk cost, ， The coefficient of variation method is used to determine the weight of risk cost. The calculation formula of the coefficient of variation is as follows: ， in is the coefficient of variation, It is The risk cost is The standard deviation between time points, It is The risk cost is The average value among the time points; Risk Cost Weight Determined by the following formula: ， Game theory combination weighting, is the risk cost weight vector obtained by entropy weight method, is the risk cost weight vector obtained by the coefficient of variation method, , ; Construct a basic risk weight vector and The vector weighted combination is the final weight of the risk cost: ， in is the risk cost weight vector after combination, , is the transposed vector of the risk cost weight vector obtained by the entropy weight method, is the transposed vector of the risk cost weight vector obtained by the coefficient of variation method, and is the weighting coefficient, , ; According to the idea of ​​game theory aggregation model, the objective function is established to combine weights and and The goal is to minimize the distance difference and seek the optimal weighting coefficient and ,Right now: ， According to the differential properties of the matrix, the optimal first-order derivative condition for obtaining the minimum value of the above formula can be transformed into the following linear equations: ， In the formula, , , , , solve the linear equations to get the weight coefficients and , and normalize it to get and , ， Combining the two weight vectors, we can get the risk cost of meteorological conditions. , air obstacle risk cost , ground loss risk cost and route risk cost The corresponding weight , ; ， After calculating the risk costs and risk cost weights of meteorological conditions, air obstacles, ground losses and route conditions, the comprehensive risk cost value of the current grid can be calculated, and the comprehensive risk cost value of each grid can be associated with the longitude and latitude of the center point of the grid to generate a risk grid map. The risk value of each grid in the map is dynamically updated in real time as time changes; Constraint module, used to construct UAV route constraints; The dynamic route planning module is used to build a dynamic route planning model for UAVs. The longitude and latitude of the flight start and end points of the flight mission, the UAV flight risk grid map, route constraints and the preference characteristics of the flight mission are used as the input of the UAV dynamic route planning model, and the improved ant colony algorithm is used to solve the optimal route path of the UAV; the preference characteristics of the flight mission include prioritizing safety or prioritizing timeliness.

Images