Area coverage strategy optimization method and device based on multi-agent collaboration

Abstract

The present invention provides a method and device for optimizing a regional coverage strategy based on multi-agent collaboration, which relates to the field of regional coverage technology. The method and device obtain the current solution space of a pre-constructed coverage strategy optimization model, wherein the current solution space corresponds to the current regional coverage strategy, and the coverage strategy optimization model is constructed based on coverage area state parameters and agent state parameters in the regional coverage strategy; the current coverage benefit is obtained based on the current solution space, and the current solution space of the coverage strategy optimization model is updated based on the current coverage benefit; the updated current solution space is used as the new current solution space, and the above steps are repeated until a preset number of iterations is reached or a preset convergence condition is reached, and a target regional coverage strategy for the area to be covered is determined based on the current solution space finally obtained, thereby solving the technical problem that an effective energy-saving coverage strategy cannot be formed for regional coverage tasks involving collaboration of multiple agents in the prior art.

Description

Technical Field

[0001] The present invention relates to the field of regional coverage technology, and in particular to a regional coverage strategy optimization method and device based on multi-agent collaboration. Background Art

[0002] In an open scene, when there are a large area or multiple task areas to be covered, a single intelligent agent is constrained by its fuel volume and cannot independently complete the coverage task of a large area or multiple task areas to be covered. Therefore, a certain number of intelligent agents are required to cooperate to complete the coverage task of a large area or multiple task areas to be covered, while ensuring that the energy consumption generated in the process of area coverage is minimized and the maximum coverage benefit is obtained, where energy consumption refers to the motion energy consumption generated by the intelligent agent in the process of performing the coverage task.

[0003] In the prior art, multiple agents are used to coordinate and cooperate to complete coverage tasks of large areas or multiple areas to be covered, and the coverage actions of a single agent are optimized for energy saving, so as to reduce the energy consumption generated during the regional coverage process. However, the energy saving benefits brought by this method are limited and cannot meet the requirements of energy saving benefits in actual application scenarios. Therefore, how to form an effective energy-saving coverage strategy for regional coverage tasks coordinated by multiple agents is a technical problem that needs to be solved urgently by technicians in related fields. Summary of the invention

[0004] The present invention provides a method and device for optimizing regional coverage strategy based on multi-agent collaboration, which is used to solve the defect that the prior art cannot form an effective energy-saving coverage strategy for regional coverage tasks involving collaboration of multiple agents, and realize energy-saving optimization of the regional coverage process and the regional coverage strategy, thereby meeting the requirements for energy-saving benefits in actual application scenarios.

[0005] The present invention provides a method for optimizing regional coverage strategy based on multi-agent collaboration, comprising: obtaining a current solution space of a pre-constructed coverage strategy optimization model, the current solution space corresponding to the current regional coverage strategy, the coverage strategy optimization model being constructed based on coverage area state parameters and agent state parameters in the regional coverage strategy; obtaining a current coverage benefit based on the current solution space, and updating the current solution space of the coverage strategy optimization model based on the current coverage benefit; using the updated current solution space as a new current solution space, and repeating the above steps until a preset number of iterations is reached or a preset convergence condition is reached, and determining a target area coverage strategy for the area to be covered based on the current solution space finally obtained.

[0006] According to a method for optimizing regional coverage strategy based on multi-agent collaboration provided by the present invention, the current coverage benefit is obtained based on the current solution space, including: obtaining the current regional segmentation result of the area to be covered based on the current solution space; obtaining the current coverage path length of the area to be covered based on the current regional segmentation result; and obtaining the current coverage benefit obtained by the agent performing the corresponding coverage task of the area to be covered based on the current coverage path length.

[0007] According to a method for optimizing area coverage strategy based on multi-agent collaboration provided by the present invention, the current area segmentation result of the area to be covered is obtained based on the current solution space, including: obtaining the area segmentation method of the area to be covered and the number of agents and agent capability parameters of the agents corresponding to the area to be covered based on the current solution space; obtaining the expected sub-area area of ​​each agent corresponding to the area to be covered based on the area to be covered and the agent capability parameters; obtaining multiple segmentation lines based on the area segmentation method, the number of agents and the area to be covered, and adjusting the positions of the segmentation lines based on the expected sub-area area; dividing the area to be covered into multiple target sub-segmentation areas based on the adjusted multiple segmentation lines, and using the multiple target sub-segmentation areas as the current area segmentation result.

[0008] According to a method for optimizing an area coverage strategy based on multi-agent collaboration provided by the present invention, the current coverage path length of the area to be covered is obtained based on the current area segmentation result, including: obtaining the area width of each of the target sub-segmented areas and the first driving distance of the agent in each of the target sub-segmented areas; for each target sub-segmented area, obtaining the minimum number of turns of the agent based on the area width corresponding to the target sub-segmented area, and obtaining the second driving distance of the agent outside the target sub-segmented area based on the minimum number of turns; obtaining the current area driving distance of the agent in the area to be covered based on the first driving distance corresponding to each of the target sub-segmented areas; obtaining the current area driving distance outside the area to be covered based on the second driving distance corresponding to each of the target sub-segmented areas; obtaining the current coverage path length based on the driving distance in the current area and the driving distance outside the current area, and the current coverage path length includes the driving distance in the current area and the driving distance outside the current area.

[0009] According to a method for optimizing area coverage strategy based on multi-agent collaboration provided by the present invention, the current coverage benefit obtained by the agent when performing the coverage task corresponding to the area to be covered is obtained based on the current coverage path length, including: obtaining the driving distance of the agent within the current area within the area to be covered and the driving distance outside the current area outside the area to be covered based on the current coverage path length; obtaining the current information benefit obtained by the agent when performing the coverage task corresponding to the area to be covered based on the driving distance within the current area; obtaining the current coverage energy consumption consumed by the agent when performing the coverage task corresponding to the area to be covered based on the driving distance within the current area and the driving distance outside the current area; and obtaining the current coverage benefit based on the current information benefit and the current coverage energy consumption.

[0010] According to a regional coverage strategy optimization method based on multi-agent collaboration provided by the present invention, the current solution space of the coverage strategy optimization model is updated based on the current coverage benefit, including: obtaining the last coverage benefit, and determining the current optimization direction based on the current coverage benefit and the last coverage benefit; obtaining the initial coverage benefit and the initial global moving step, and determining the current global moving step based on the initial coverage benefit, the last coverage benefit and the initial global moving step; based on the current global moving step and the current optimization direction, globally optimizing the current solution space; obtaining the current local moving step, and locally optimizing the current solution space after the global optimization update based on the current local moving step and the current optimization direction.

[0011] According to a method for optimizing regional coverage strategy based on multi-agent collaboration provided by the present invention, after determining the target regional coverage strategy based on the current solution space, the method further includes: determining competition constraints for obtaining coverage tasks corresponding to the area to be covered based on the target regional coverage strategy; obtaining each agent group competing for the coverage task corresponding to the area to be covered and agent capability data of each agent group based on the competition constraints; obtaining a regional competition index value for each agent group based on the agent capability data, and determining the target agent group corresponding to the area to be covered based on the regional competition index value.

[0012] According to a method for optimizing regional coverage strategy based on multi-agent collaboration provided by the present invention, the regional competition index value of each of the agent groups is obtained based on the agent capability data, including: obtaining the group coverage time of each of the agent groups to complete the coverage task corresponding to the area to be covered; obtaining the group travel time of each of the agent groups from the current position to the area to be covered; obtaining the group coverage benefit obtained by each of the agent groups when completing the coverage task corresponding to the area to be covered; and obtaining the regional competition index value of each of the agent groups based on the group coverage time, the group travel time and the group coverage benefit.

[0013] The present invention also provides a regional coverage strategy optimization device based on multi-agent collaboration, including: a data acquisition module, used to obtain the current solution space of a pre-constructed coverage strategy optimization model, the current solution space corresponds to the current regional coverage strategy, and the coverage strategy optimization model is constructed based on the coverage area state parameters and agent state parameters in the regional coverage strategy; a first optimization module, used to obtain the current coverage benefit based on the current solution space, and update the current solution space of the coverage strategy optimization model based on the current coverage benefit; a second optimization module, used to use the updated current solution space as the new current solution space, and repeat the above steps until a preset number of iterations is reached or a preset convergence condition is reached, and determine the target area coverage strategy for the area to be covered based on the current solution space finally obtained.

[0014] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements any of the above-described methods for optimizing regional coverage strategies based on multi-agent collaboration.

[0015] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described methods for optimizing regional coverage strategies based on multi-agent collaboration.

[0016] The present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned methods for optimizing regional coverage strategies based on multi-agent collaboration.

[0017] The regional coverage strategy optimization method and device based on multi-agent collaboration provided by the present invention updates the current solution space of the coverage strategy optimization model based on the current coverage benefit corresponding to the current regional coverage strategy, so that the coverage area state parameters in the coverage strategy are adapted to the agent state parameters, thereby maximizing the current coverage benefit, and improving the energy-saving optimization effect by energy-saving optimization of the regional coverage process and the regional coverage strategy, and obtaining the target regional coverage strategy with the best energy-saving effect based on the current solution space of the optimized coverage strategy optimization model, thereby solving the technical problem that it is impossible to form an effective energy-saving coverage strategy for the regional coverage task of multiple agent collaboration in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are 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.

[0019] Figure 1 This is one of the flow charts of the regional coverage strategy optimization method based on multi-agent collaboration provided by the present invention;

[0020] Figure 2 This is the second flow chart of the regional coverage strategy optimization method based on multi-agent collaboration provided by the present invention;

[0021] Figure 3 This is the third flow chart of the regional coverage strategy optimization method based on multi-agent collaboration provided by the present invention;

[0022] Figure 4a This is the fourth flow chart of the regional coverage strategy optimization method based on multi-agent collaboration provided by the present invention;

[0023] Figure 4b is a schematic diagram of obtaining the region width of a target sub-segmented region in an embodiment of the present invention;

[0024] Figure 5 This is the fifth flow chart of the regional coverage strategy optimization method based on multi-agent collaboration provided by the present invention;

[0025] Figure 6 This is the sixth flow chart of the regional coverage strategy optimization method based on multi-agent collaboration provided by the present invention;

[0026] Figure 7 This is the seventh flow chart of the regional coverage strategy optimization method based on multi-agent collaboration provided by the present invention;

[0027] Figure 8 This is the eighth flow chart of the regional coverage strategy optimization method based on multi-agent collaboration provided by the present invention;

[0028] Fig. 9 It is a structural schematic diagram of the regional coverage strategy optimization device based on multi-agent collaboration provided by the present invention;

[0029] Fig.10 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] Combine the following Figure 1-Figure 8 The regional coverage strategy optimization method based on multi-agent collaboration of the present invention can be applied to regional coverage tasks in application scenarios such as rapid search of post-disaster areas, sweeping net delivery of items, and forest fire prevention. The agents include unmanned vehicles, drones, or other unmanned systems.

[0032] like Figure 1 As shown, the present invention provides a regional coverage strategy optimization method based on multi-agent collaboration, including: step S1, obtaining the current solution space of a pre-constructed coverage strategy optimization model, the current solution space corresponds to the current regional coverage strategy, and the coverage strategy optimization model is constructed based on the coverage area state parameters and agent state parameters in the regional coverage strategy.

[0033] The coverage area status parameter includes at least one of the area of ​​the area to be covered, intelligence data and importance. The agent status parameter includes at least one of the number of agents, the scanning width of the agent and the driving speed of the agent.

[0034] Step S2, based on the current solution space, obtain the current coverage benefit, and based on the current coverage benefit, update the current solution space of the coverage strategy optimization model. The current coverage benefit represents the coverage benefit value obtained by the agent in the process of performing the corresponding coverage task of the area to be covered.

[0035] Step S3, taking the updated current solution space as the new current solution space, and repeating the above steps until a preset number of iterations is reached or a preset convergence condition is reached, and determining a target area coverage strategy for the area to be covered based on the current solution space finally obtained.

[0036] In one embodiment, the current coverage gain of each iteration process is obtained, and a coverage gain change curve is generated based on the current coverage gain of multiple iteration processes. When the coverage gain change curve converges, it is determined that the preset convergence condition is met. Alternatively, when the current coverage gain tends to a stable coverage gain, it is determined that the preset convergence condition is met.

[0037] The above steps S1 to S3 update the current solution space of the coverage strategy optimization model based on the current coverage benefit corresponding to the current area coverage strategy, so that the coverage area state parameters in the coverage strategy are adapted to the agent state parameters, thereby maximizing the current coverage benefit, and improving the energy-saving optimization effect by optimizing the area coverage process and the area coverage strategy, and obtaining the target area coverage strategy with the best energy-saving effect based on the current solution space of the optimized coverage strategy optimization model, thereby solving the technical problem that it is impossible to form an effective energy-saving coverage strategy for the area coverage task of multiple agents collaborating in the prior art.

[0038] In one embodiment, the coverage area state parameters include the area segmentation method of the area to be covered in the coverage strategy, and the agent state parameters include the agent capability parameters and the number of agents corresponding to the area to be covered in the coverage strategy.

[0039] In one embodiment, Figure 2 As shown, the above step S2 includes steps S21 to S23, wherein: step S21, based on the current solution space, obtains the current region segmentation result of the area to be covered. Step S22, based on the current region segmentation result, obtains the current coverage path length of the area to be covered. Step S23, based on the current coverage path length, obtains the current coverage benefit obtained by the agent performing the corresponding coverage task of the area to be covered.

[0040] In one embodiment, Figure 3 As shown, the above step S21 includes steps S211 to S214, wherein: step S211, based on the current solution space, obtains the region segmentation method of the area to be covered and the number of agents and agent capability parameters of the agents corresponding to the area to be covered. Step S212, based on the area to be covered and the agent capability parameters, obtains the expected area of ​​the sub-region of each agent corresponding to the area to be covered.

[0041] Furthermore, based on the area of ​​the area to be covered and the agent capability parameter of each agent, the expected area of ​​the sub-area of ​​each agent is calculated, wherein the calculation method of the expected area of ​​the sub-area is shown in the following formula (1):

[0042]

[0043] in, S represents the expected area of ​​the sub-region of agent i. j represents the area of ​​the region j to be covered. θ represents the agent capability parameter, θ∈(0,1).

[0044] Step S213, based on the region segmentation method, the number of agents and the area to be covered, multiple segmentation lines are obtained, and the positions of the segmentation lines are adjusted based on the expected area of ​​the sub-area. Specifically, the area to be covered is divided into multiple sub-areas based on the number of agents and the region segmentation method to obtain multiple segmentation lines. The position of the segmentation line corresponding to each sub-area is adjusted so that the area of ​​each sub-area is within the error range of the expected area of ​​the corresponding sub-area.

[0045] Step S214: segment the area to be covered into a plurality of target sub-segmented areas based on the adjusted plurality of segmentation lines, and use the plurality of target sub-segmented areas as the current area segmentation result.

[0046] In the above steps S211 to S214, the expected sub-region area of ​​each agent corresponding to the sub-region is set through the agent capability parameters of the agent, multiple segmentation lines are obtained based on the region segmentation method, the number of agents and the area to be covered, and the position of the segmentation line is adjusted based on the expected area of ​​the sub-region, so as to link the region segmentation process of the area to be covered with the agent capability parameters and the number of agents, thereby allocating agents with different numbers of agents and agent capability parameters to different areas to be covered, and adapting the area to be covered to the agent capability parameters and the number of agents, so as to maximize the current coverage benefit obtained by the agent when performing the coverage task corresponding to the area to be covered, and then obtain the optimal current area coverage strategy, so as to form an effective energy-saving coverage strategy for the area coverage task of multiple agents collaborating.

[0047] In one embodiment, different shapes of the area to be covered will produce different area coverage path lengths, and different area segmentation methods for the same area to be covered will produce different area coverage path lengths. Therefore, a straight line passing through the center point of the area to be covered is defined as the reference for the area segmentation method of the area to be covered. That is, the area segmentation method is set to the baseline slope of the baseline segmented by the center point of the area to be covered, so that the coverage strategy optimization model (or coverage strategy optimization space) can be abstracted as Λ=(n,θ), as shown in formula (2) and formula (3):

[0048]

[0049]

[0050] Among them, n represents the number of agents assigned to each sub-area corresponding to the area to be covered; θ represents the baseline slope of the baseline segmented by the center point of the area to be covered, θ∈[0,pi].

[0051] In one embodiment, the above step S213 includes steps 21 to 24, wherein: Step 21: Rotate the area to be covered based on the baseline slope of the center point segmentation baseline and the first rotation direction so that the center point segmentation baseline of the area to be covered is parallel to the horizontal coordinate axis. The center point segmentation baseline is vertically segmented to obtain multiple segmentation lines and multiple sub-areas, and the segmentation lines are perpendicular to the center point segmentation baseline.

[0052] Step 22: For one of the segmentation lines, obtain the length of the intersection of the segmentation line and the area to be covered, the area of ​​the left sub-area of ​​the segmentation line, and the current displacement of the segmentation line. Based on the sub-area area of ​​the left sub-area of ​​the segmentation line, the length of the intersection of the segmentation line and the area to be covered, the number of agents, and the area of ​​the area to be covered, obtain the adjustment displacement of the segmentation line. The position adjustment method of the segmentation line is shown in formula (4):

[0053]

[0054] in, Indicates the area of ​​the sub-region on the left side of the cutting line, Length(Cutline i ∩P) represents the length of the intersection of the dividing line and the area to be covered, L represents the number of agents (that is, the number of divisions of the area to be covered), and S represents the area of ​​the area to be covered.

[0055] Step 23: Adjust the position of the segmentation line based on the adjustment displacement, and obtain the left sub-region of the segmentation line based on the segmentation line after the position adjustment. Determine whether the sub-region area of ​​the left sub-region is within the error range of the expected area of ​​the corresponding sub-region; if the sub-region area of ​​the left sub-region is within the error range of the expected area of ​​the corresponding sub-region, use the left sub-region of the segmentation line as a target sub-segmentation region; if the sub-region area of ​​the left sub-region is within the error range of the expected area of ​​the corresponding sub-region, repeat the above step 22 until the sub-region area of ​​the left sub-region is within the error range of the expected area of ​​the corresponding sub-region.

[0056] Step 24: Repeat the above steps 22 to 23 until the positions of all segmentation lines are adjusted and all target sub-segmentation areas of the area to be segmented are obtained. The target sub-segmentation area is rotated based on the baseline slope of the center point segmentation baseline and the second rotation direction to obtain a rotated target sub-segmentation area; the first rotation direction is opposite to the second rotation direction.

[0057] In one embodiment, in actual application, there is a situation where the vertex of the area to be covered is narrow, resulting in the segmentation line falling outside the area to be covered after displacement adjustment. Therefore, the original displacement of each segmentation line before displacement adjustment, the current displacement of each segmentation line after displacement adjustment, and the maximum displacement and minimum displacement of the area to be covered are obtained, and the current displacement of the first segmentation line is updated based on the original displacement, the maximum displacement, and the minimum displacement. The updating process is shown in the following formula (5):

[0058]

[0059] Among them, c i represents the original displacement of each segmentation line before adjustment, c i +d represents the current displacement of each segmentation line after displacement adjustment, x min represents the minimum displacement of the area to be covered, x max represents the maximum displacement of the area to be covered, and x represents the current displacement of the first segmentation line after the update.

[0060] In one embodiment, Figure 4a As shown, the above step S22 includes steps S221 to S224, wherein: step S221, obtains the area width of each target sub-division area and the first driving distance of the intelligent agent in each target sub-division area.

[0061] In one embodiment, the area width of each target sub-segmentation area is obtained based on the "point-edge" width calculation method. Specifically, the distance from each point in the target sub-segmentation area to other edges (i.e., other straight lines do not include the edge where the point is located) is calculated in turn, and the distances between the point and two adjacent edges are compared in turn until the distance difference between the point and the two adjacent edges is less than 0, and the distance from the point to one of the edges is used as the span corresponding to the point, and the spans corresponding to all points are recorded, and the minimum span among all spans is used as the area width of the target sub-segmentation area. Alternatively, the distance from each edge to other vertices (i.e., other vertices do not include the vertex in the edge) is calculated, and the distance from the edge to two adjacent vertices is compared in turn until the distance difference between the edge and the two adjacent vertices is less than 0, and the distance from the edge to one of the vertices is used as the span corresponding to the edge, and the spans corresponding to all edges are recorded, and the minimum span among all spans is used as the area width of the target sub-segmentation area.

[0062] Furthermore, the straight line equation of each edge in the target sub-segmentation area is obtained. Since the shape of the target sub-segmentation area is a polygon, the polygon is used to represent the target sub-segmentation area in the following formula. The straight line equation of any edge in the polygon is shown in the following formula (6):

[0063] (y i+1 -y i )x-(xi+1 -x i )y+x i+1 y i -x i y i+1 =0 (6)

[0064] Among them, y i+1 Represents the ordinate of the end point of the i-th edge in the polygon, y i represents the vertical coordinate of the vertex of the i-th edge in the polygon, x i+1 represents the horizontal coordinate of the end point of the i-th edge in the polygon, x i Represents the horizontal coordinate of the vertex of the i-th edge in the polygon.

[0065] In one embodiment, an edge is randomly selected as the starting point, C is added to the line equation of the edge, and the distances C between two adjacent vertices to the edge are calculated and compared until the difference between the two distances Δ<=0. Then, the point farthest from the edge can be found, thereby obtaining the span of the edge. The calculation process is shown in the following formula (7):

[0066]

[0067] Among them, C i,j Represents a vertex (x j ,y j ) to the first distance of the edge, C i,j+1 Represents a vertex (x j+1 ,y j+1 ) to the second distance of the edge, Δ represents the distance difference between the first distance and the second distance. In the case of Δ<=0, the second distance C i,j+1 is the span of the edge.

[0068] Similarly, until the spans corresponding to all edges are found, and the minimum span among the spans of all edges in the target sub-segmentation area is obtained, the minimum span is used as the area width of the target sub-segmentation area.

[0069] In one embodiment, Figure 4bAs shown, the process of calculating the area width of the target sub-segmentation area in the present invention is further explained by taking the shape of the target sub-segmentation area as a pentagon as an example. Calculate the distance from vertex A to other sides in the pentagon, that is, substitute the coordinates of vertex A into the straight line equation of side 3 respectively, and obtain the third distance from vertex A to side 3. Substitute the coordinates of vertex A into the straight line equation of side 4 respectively, and obtain the fourth distance from vertex A to side 4. Calculate the distance difference between the fourth distance and the third distance. When the distance difference is less than or equal to 0, use the fourth distance as the span corresponding to vertex A. Similarly, find the spans corresponding to all vertices, and use the minimum span of all spans as the area width of the target sub-segmentation area. Furthermore, obtaining the first driving distance of the intelligent agent in each target sub-segmentation area can be expressed by the following formula (8):

[0070]

[0071] Among them, S in represents the first driving distance of the agent in the target sub-segmentation area, represents the length of the lth flight trajectory of the agent in the target sub-segmentation area, S area represents the area of ​​the target sub-segmentation region, and w represents the turning radius.

[0072] Step S222, for each target sub-segment area, obtain the minimum number of turns of the agent based on the area width corresponding to the target sub-segment area, and obtain the second driving distance of the agent outside the target sub-segment area based on the minimum number of turns.

[0073] Furthermore, the minimum number of turns of the agent is obtained based on the area width corresponding to the target sub-segmentation area. The minimum number of turns can be expressed by the following formula (9):

[0074]

[0075] Among them, n turn represents the minimum number of turns, K represents the area width of the target sub-segmentation area, and w represents the turning radius.

[0076] Furthermore, based on the minimum number of turns, the second driving distance of the agent outside the target sub-divided area is obtained. The second driving distance can be expressed by the following formula (10):

[0077] S out =n turn S turn (10)

[0078] Among them, S out represents the second driving distance of the agent outside the target sub-segmentation area, n turn represents the minimum number of turns, Sturn Represents the distance traveled by the agent when turning.

[0079] It can be seen that for a specific target sub-segmentation area, its area S area is a fixed value. For the agent, its turning radius w and the driving distance S when turning turn is a constant, therefore, by determining the minimum area width K of the target sub-division area, the minimum number of turns and the second driving distance of the agent outside the target sub-division area can be obtained.

[0080] Step S223, based on the first driving distance corresponding to each target sub-divided area, obtain the driving distance of the intelligent agent in the current area within the area to be covered; based on the second driving distance corresponding to each target sub-divided area, obtain the driving distance outside the current area outside the area to be covered.

[0081] Step S224, based on the driving distance in the current area and the driving distance outside the current area, the current coverage path length is obtained, and the current coverage path length includes the driving distance in the current area and the driving distance outside the current area.

[0082] In one embodiment, Figure 5 As shown, the above step S23 includes steps S231 to S234, wherein: step S231, based on the current coverage path length, obtains the driving distance of the agent in the current area within the area to be covered and the driving distance outside the current area outside the area to be covered. Step S232, based on the driving distance in the current area, obtains the current information benefit obtained by the agent performing the corresponding coverage task of the area to be covered.

[0083] The longer the coverage path of the agent in the area to be covered, the larger the coverage area, and the more current information benefits are obtained. The current information benefits are calculated as shown in the following formula (11):

[0084]

[0085] Among them, R b represents the current information gain, represents the distance traveled by the agent in the current area within the area to be covered, w i j represents the agent scan width, v i j represents the driving speed of the agent, S j represents the area to be covered, ω j It indicates the intelligence information of the area to be covered, and strategy indicates the method of dividing the area to be covered.

[0086] Step S233, based on the driving distance in the current area and the driving distance outside the current area, obtain the current coverage energy consumption consumed by the intelligent agent to perform the coverage task corresponding to the area to be covered.

[0087] The energy consumption of the intelligent agent during the coverage process is mainly the consumption caused by movement, including the total driving distance inside and outside the area to be covered. Therefore, the current calculation method of coverage energy consumption is shown in formula (12):

[0088]

[0089] Among them, R c Indicates the current coverage energy consumption, Indicates the distance traveled by the agent in the current area within the area to be covered, It represents the distance traveled by the agent outside the current area to be covered, a1, a2, a3 represent weight parameters whose sum is 1, n represents the number of agents, and strategy represents the regional segmentation method of the area to be covered.

[0090] Step S234, based on the current information gain and the current coverage energy consumption, obtain the current coverage gain. The current coverage gain represents the net coverage gain value obtained by the agent in the process of performing the corresponding coverage task of the area to be covered. The current coverage gain increases with the increase of the current information gain and increases with the decrease of the current coverage energy consumption.

[0091] In one embodiment, the current coverage benefit is proportional to the current information benefit, and the current coverage benefit is inversely proportional to the current coverage energy consumption. Therefore, the calculation method of the current coverage benefit is shown in the following formula (13):

[0092]

[0093] Among them, R t represents the current coverage benefit, R b represents the current information gain, R c Indicates the current coverage energy consumption, c j represents the importance of the area to be covered, and n represents the number of agents.

[0094] In one embodiment, a regional coverage strategy is set according to the importance of the area to be covered. When the importance of the area to be covered is greater than a preset threshold, the area to be covered is covered in a full coverage manner. When the importance of the area to be covered is less than or equal to the preset threshold, the area to be covered is covered in a sparse coverage manner.

[0095] In one embodiment, Figure 6As shown, the above step S2 also includes steps S24 to S27, wherein: step S24, obtains the last coverage gain, and determines the current optimization direction based on the current coverage gain and the last coverage gain.

[0096] Further, when the current coverage gain is greater than the previous coverage gain, the current optimization direction is determined to be the first optimization direction; and when the current coverage gain is less than the previous coverage gain, the current optimization direction is determined to be the second optimization direction.

[0097] Step S25, obtaining the initial coverage gain and the initial global moving step length, and determining the current global moving step length based on the initial coverage gain, the last coverage gain and the initial global moving step length. The initial global moving step length indicates the initial moving step length of the global moving search. The current global moving step length indicates the current moving step length of the global moving search.

[0098] In one embodiment, in order to fully retain high-quality solutions and enable the global mobile search algorithm to converge quickly, the current global mobile step length β in the global mobile search process is adaptively adjusted based on the optimization process, so that the obtained solution has strong diversity in the early iteration process, so that the diversity of solutions can be gradually reduced in the later local mobile search process, and thus the local mobile search can be better performed. The calculation method of the current global mobile step length is shown in the following formula (14):

[0099]

[0100] in, represents the initial coverage gain. is the coverage benefit of the t-1 generation, that is, the last coverage benefit. β0 represents the initial global moving step. β represents the current global moving step.

[0101] Step S26: Based on the current global moving step size and the current optimization direction, the current solution space is globally optimized and updated.

[0102] In one embodiment, the Lévy flight based on the global adaptive step size adjustment mechanism randomly performs a global search update on the current solution space. As a unique feature of the cuckoo optimization algorithm, the global Lévy flight is used to change the global moving step size over a large range during the optimization process, effectively avoiding the optimization process from falling into a local optimum. The global moving search process based on the Lévy flight is shown in formula (15):

[0103]

[0104] Where: represents the i-th solution of the t+1th generation. represents the ith solution of the tth generation. α represents the step length, and the value of α is usually 1. Lévy distribution represents the random optimization direction, that is, the current optimization direction in the present invention, and its moving step length obeys the Lévy distribution. The global mobile search process in the present invention is shown in formula (16), that is, the cuckoo's continuous jump forms a random walk process:

[0105]

[0106] in, represents the i-th solution of the t+1th generation. represents the ith solution of the tth generation. α represents the step size, and the value of α is usually 1. β represents the current global moving step size. u and v follow a normal distribution.

[0107] Furthermore, based on the current global moving step size and the current optimization direction, the current solution space is globally optimized and updated, and solutions in the current solution space that are less than a preset probability value are discarded.

[0108] Step S27, obtaining the current local moving step length, and performing local optimization update on the current solution space after global optimization update based on the current local moving step length and the current optimization direction.

[0109] The current solution space after global optimization update is updated based on the current local moving step size and the current optimization direction. The local moving search process is shown in formula (17):

[0110]

[0111] in, represents the i-th solution of the t+1th generation. represents the i-th solution of the t-th generation, α represents the step size, and the value of α is usually 1. and Represents two random number sequences, H represents the Heaviside function, ε represents a random number, and pa represents the probability of being discovered in the cuckoo algorithm.

[0112] In one embodiment, before step S24, the method further includes: each row of the current solution space is a solution, and the current solution space includes multiple solutions. Obtain the current coverage benefit corresponding to each solution in the current solution space, compare the current coverage benefits of any two solutions, and keep the solution with the higher current coverage benefit of the two solutions, and so on, until each solution in the current solution space is traversed to obtain the current solution space after preliminary optimization. This embodiment achieves preliminary optimization of the current solution space by retaining solutions with higher current coverage benefits. Its optimization process is similar to elite retention in genetic algorithms, to ensure that subsequent global mobile searches and local mobile searches are always within the range of local optimal solutions, so that the optimal solution can be retained to the next generation, thereby reducing the risk of the optimal solution being expelled from the population.

[0113] In one embodiment, Figure 7 As shown, the regional coverage strategy optimization method based on multi-agent collaboration provided by the present invention also includes steps S4 to S6, wherein: Step S4, based on the target area coverage strategy, determines the competitive constraints for obtaining the coverage task corresponding to the area to be covered. The competitive constraints include the minimum coverage time and the maximum coverage benefit of the area to be covered and the required number of agents.

[0114] Step S5, based on the competition constraint condition, obtaining each agent group competing for the coverage task corresponding to the area to be covered and the agent capability data of each agent group.

[0115] Furthermore, each agent determines whether the competition constraint is satisfied based on its own agent capability data, and the agent participates in the competition for the coverage task corresponding to the area to be covered when the competition constraint is satisfied. Each agent communicates with each other and forms an agent group with adjacent agents. In addition, the agent group determines whether its own agent capability data satisfies the competition constraint, and the agent group participates in the competition for the coverage task corresponding to the area to be covered when the competition constraint is satisfied.

[0116] For example, when the agent determines that the coverage time required for completing the corresponding coverage task of the area to be covered is less than the minimum coverage time of the area to be covered, it will not participate in the competition for the corresponding coverage task of the area to be covered. When the agent determines that the coverage time required for completing the corresponding coverage task of the area to be covered is greater than or equal to the minimum coverage time of the area to be covered, it determines whether the proportion of the maximum coverage benefit of the area to be covered to the total coverage benefit reaches the preset proportion threshold, and when the proportion reaches the preset proportion threshold, it participates in the competition for the corresponding coverage task of the area to be covered, that is, the agent tends to perform the coverage task corresponding to the area to be covered with higher coverage benefits. Similarly, the agent group needs to determine whether the number of its own agents reaches the competition constraints such as the number of agents required for the area to be covered, and participate in the competition for the corresponding coverage task of the area to be covered when the competition constraints are met.

[0117] Step S6, based on the agent capability data, the regional competition index value of each agent group is obtained, and based on the regional competition index value, the target agent group corresponding to the area to be covered is determined. Specifically, based on the regional competition index values ​​of multiple agent groups, the agent group with the highest regional competition index value is determined as the target agent group corresponding to the area to be covered.

[0118] Furthermore, the regional competition index value of each agent group is obtained based on the agent capability data and the preset competition index function, and the target agent group corresponding to the area to be covered is determined based on the regional competition index value. The preset competition index function includes the coverage time index parameter and coverage benefit index parameter of the agent group completing the coverage task corresponding to the area to be covered and the driving time index parameter of the agent group driving from the current position to the area to be covered.

[0119] In one embodiment, Figure 8 As shown, the above step S6 includes steps S61 to S64, wherein: step S61, obtains the group coverage time of each agent group completing the corresponding coverage task of the area to be covered. Step S62, obtains the group travel time of each agent group traveling from the current position to the area to be covered. Step S63, obtains the group coverage benefit obtained by each agent group completing the corresponding coverage task of the area to be covered. Step S64, based on the group coverage time, group travel time and group coverage benefit, obtains the regional competition index value of each agent group.

[0120] It should be noted that the group coverage time is the above-mentioned coverage time index parameter, the proportion parameter of the group coverage revenue to the total coverage revenue is the above-mentioned coverage revenue index parameter, and the group driving time is the above-mentioned driving time index parameter.

[0121] Furthermore, the preset competition index function also includes a value allocation parameter of the area to be covered and a group travel time, wherein the value allocation parameter represents a value attribute allocated to the area to be covered, and the preset competition index function can also be called a bidding function. The regional competition index value is calculated based on the preset competition index function, the group coverage time, the group travel time, the group coverage income and the value allocation parameter, wherein the preset competition index function is shown in the following formula (18):

[0122]

[0123] Among them, t ij represents the group travel time of the agent group from the to-be-covered area i to the to-be-covered area j, λ1 represents the weight corresponding to the group travel time, and λ i represents the corresponding weight, λ j Represents the corresponding weight, and λ1+λ2+λ3=1. represents the value allocation parameter of the area j to be covered, V j Represents the coverage benefit indicator parameter of the area j to be covered.

[0124] Furthermore, the calculation method of the group coverage time and the group coverage benefit is shown in the following formula (19) and formula (20):

[0125]

[0126]

[0127] in, It represents the shortest coverage path length of agent i to complete the coverage task in the area to be covered j. i represents the average driving speed of agent i. j R is the group coverage time for the agent group to complete the corresponding coverage task of the area to be covered j. t Indicates the total revenue covered. V represents the group coverage benefit obtained by the agent group when completing the corresponding coverage task of the area to be covered j. j Represents the coverage benefit indicator parameter of the area j to be covered.

[0128] A specific embodiment is provided below to further illustrate the area coverage strategy optimization method based on multi-agent collaboration provided by the present invention, which specifically includes the following steps:

[0129] Step 1: Obtain the current solution space of the pre-built coverage strategy optimization model. The current solution space corresponds to the current regional coverage strategy. The coverage strategy optimization model is constructed based on the coverage area state parameters and agent state parameters in the regional coverage strategy. Based on the current solution space, obtain the current regional segmentation result of the area to be covered. Based on the current regional segmentation result, obtain the current coverage path length of the area to be covered. Based on the current coverage path length, obtain the current coverage benefit obtained by the agent performing the corresponding coverage task of the area to be covered.

[0130] Step 2: Obtain the last coverage gain, and determine the current optimization direction based on the current coverage gain and the last coverage gain; obtain the initial coverage gain and the initial global moving step, and determine the current global moving step based on the initial coverage gain, the last coverage gain and the initial global moving step; based on the current global moving step and the current optimization direction, perform a global optimization update on the current solution space; obtain the current local moving step, and perform a local optimization update on the current solution space after the global optimization update based on the current local moving step and the current optimization direction.

[0131] Step 3: Use the updated current solution space as the new current solution space, repeat the above steps 1 to 2 until the preset number of iterations is reached or the preset convergence condition is reached, and determine the target area coverage strategy for the area to be covered based on the current solution space finally obtained. Determine the competition constraint conditions for obtaining the coverage task corresponding to the area to be covered based on the target area coverage strategy; based on the competition constraint conditions, obtain each agent group competing for the coverage task corresponding to the area to be covered and the agent capability data of each agent group; obtain the regional competition index value of each agent group based on the agent capability data, and determine the target agent group corresponding to the area to be covered based on the regional competition index value.

[0132] The following is a description of the area coverage strategy optimization device based on multi-agent collaboration provided by the present invention. The area coverage strategy optimization device based on multi-agent collaboration described below and the area coverage strategy optimization method based on multi-agent collaboration described above can be referenced to each other.

[0133] like Fig. 9 As shown, the present invention also provides a regional coverage strategy optimization device based on multi-agent collaboration, and the regional coverage strategy optimization device 100 based on multi-agent collaboration includes: a data acquisition module 10, which is used to obtain the current solution space of a pre-constructed coverage strategy optimization model, the current solution space corresponds to the current regional coverage strategy, and the coverage strategy optimization model is constructed based on the coverage area state parameters and agent state parameters in the regional coverage strategy. A first optimization module 20 is used to obtain the current coverage benefit based on the current solution space, and update the current solution space of the coverage strategy optimization model based on the current coverage benefit. A second optimization module 30 is used to use the updated current solution space as a new current solution space, and repeat the above steps until a preset number of iterations is reached or a preset convergence condition is reached, and determine the target area coverage strategy for the area to be covered based on the current solution space finally obtained.

[0134] In one embodiment, the first optimization module 20 includes: a region segmentation unit, which is used to obtain the current region segmentation result of the area to be covered based on the current solution space; a path planning unit, which is used to obtain the current coverage path length of the area to be covered based on the current region segmentation result; and a benefit acquisition unit, which is used to obtain the current coverage benefit obtained by the agent performing the corresponding coverage task of the area to be covered based on the current coverage path length.

[0135] In one embodiment, the region segmentation unit includes: a segmentation data acquisition subunit, which is used to acquire the region segmentation method of the area to be covered and the number of agents and agent capability parameters of the agents corresponding to the area to be covered based on the current solution space. An expected area acquisition subunit, which is used to acquire the expected sub-region area of ​​each agent corresponding to the area to be covered based on the area to be covered and the agent capability parameters. A segmentation adjustment acquisition subunit, which is used to acquire multiple segmentation lines based on the region segmentation method, the number of agents and the area to be covered, and adjust the positions of the segmentation lines based on the expected area of ​​the sub-region. A segmentation result acquisition subunit, which is used to segment the area to be covered into multiple target sub-segmentation areas based on the adjusted multiple segmentation lines, and use the multiple target sub-segmentation areas as the current region segmentation results.

[0136] In one embodiment, the path planning unit includes: a first distance acquisition subunit, which is used to acquire the area width of each target sub-divided area and the first driving distance of the agent in each target sub-divided area. A second distance acquisition subunit, which is used to acquire the minimum number of turns of the agent based on the area width corresponding to the target sub-divided area for each target sub-divided area, and acquire the second driving distance of the agent outside the target sub-divided area based on the minimum number of turns. A third distance acquisition subunit, which is used to acquire the driving distance of the agent in the current area within the area to be covered based on the first driving distance corresponding to each target sub-divided area; and acquire the driving distance outside the current area outside the area to be covered based on the second driving distance corresponding to each target sub-divided area. A coverage path acquisition subunit, which is used to obtain the current coverage path length based on the driving distance in the current area and the driving distance outside the current area, and the current coverage path length includes the driving distance in the current area and the driving distance outside the current area.

[0137] In one embodiment, the benefit acquisition unit includes: a driving distance acquisition subunit, which is used to acquire the driving distance of the agent in the current area to be covered and the driving distance outside the current area to be covered based on the current coverage path length. An information benefit acquisition subunit, which is used to acquire the current information benefit obtained by the agent performing the coverage task corresponding to the area to be covered based on the driving distance in the current area. A coverage energy consumption acquisition subunit, which is used to acquire the current coverage energy consumption consumed by the agent to perform the coverage task corresponding to the area to be covered based on the driving distance in the current area and the driving distance outside the current area. A coverage benefit acquisition subunit, which is used to acquire the current coverage benefit based on the current information benefit and the current coverage energy consumption.

[0138] In one embodiment, the first optimization module 20 further includes: a direction determination unit, which is used to obtain the last coverage gain, and determine the current optimization direction based on the current coverage gain and the last coverage gain. A step length acquisition unit, which is used to obtain the initial coverage gain and the initial global moving step length, and determine the current global moving step length based on the initial coverage gain, the last coverage gain and the initial global moving step length. A global optimization unit, which is used to perform global optimization update on the current solution space based on the current global moving step length and the current optimization direction. A local optimization unit, which is used to obtain the current local moving step length, and perform local optimization update on the current solution space after the global optimization update based on the current local moving step length and the current optimization direction.

[0139] In one embodiment, the regional coverage strategy optimization device 100 based on multi-agent collaboration further includes: a constraint acquisition module, which is used to determine and acquire competition constraints for the coverage task corresponding to the area to be covered based on the target area coverage strategy. A competition constraint module, which is used to acquire each agent group competing for the coverage task corresponding to the area to be covered and the agent capability data of each agent group based on the competition constraints. A group determination module, which is used to acquire the regional competition index value of each agent group based on the agent capability data, and determine the target agent group corresponding to the area to be covered based on the regional competition index value.

[0140] In one embodiment, the group determination module includes a coverage time acquisition unit, a driving time acquisition unit, a coverage benefit acquisition unit and a competition index acquisition unit, wherein: the coverage time acquisition unit is used to acquire the group coverage time of each agent group completing the corresponding coverage task of the area to be covered. The driving time acquisition unit is used to acquire the group driving time of each agent group traveling from the current position to the area to be covered. The coverage benefit acquisition unit is used to acquire the group coverage benefit obtained by each agent group completing the corresponding coverage task of the area to be covered. The competition index acquisition unit is used to acquire the regional competition index value of each agent group based on the group coverage time, the group driving time and the group coverage benefit.

[0141] Fig.10 An example of a physical structure diagram of an electronic device is shown in FIG. Fig.10As shown, the electronic device may include: a processor 1010, a communication interface 1020, a memory 1030 and a communication bus 1040, wherein the processor 1010, the communication interface 1020 and the memory 1030 communicate with each other through the communication bus 1040. The processor 1010 may call the logic instructions in the memory 1030 to execute the regional coverage strategy optimization method based on multi-agent collaboration, the method comprising: obtaining the current solution space of the pre-constructed coverage strategy optimization model, the current solution space corresponds to the current regional coverage strategy, and the coverage strategy optimization model is constructed based on the coverage area state parameters and the agent state parameters in the regional coverage strategy; obtaining the current coverage benefit based on the current solution space, and updating the current solution space of the coverage strategy optimization model based on the current coverage benefit; using the updated current solution space as the new current solution space, and repeating the above steps until the preset number of iterations is reached or the preset convergence condition is reached, and determining the target area coverage strategy of the area to be covered based on the final current solution space.

[0142] In addition, the logic instructions in the above-mentioned memory 1030 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk.

[0143] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the regional coverage strategy optimization method based on multi-agent collaboration provided by the above methods. The method includes: obtaining the current solution space of a pre-constructed coverage strategy optimization model, the current solution space corresponds to the current regional coverage strategy, and the coverage strategy optimization model is constructed based on the coverage area state parameters and agent state parameters in the regional coverage strategy; obtaining the current coverage benefit based on the current solution space, and updating the current solution space of the coverage strategy optimization model based on the current coverage benefit; using the updated current solution space as the new current solution space, and repeating the above steps until a preset number of iterations is reached or a preset convergence condition is reached, and determining the target area coverage strategy for the area to be covered based on the current solution space finally obtained.

[0144] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the regional coverage strategy optimization method based on multi-agent collaboration provided by the above-mentioned methods, the method comprising: obtaining a current solution space of a pre-constructed coverage strategy optimization model, the current solution space corresponding to the current regional coverage strategy, the coverage strategy optimization model being constructed based on the coverage area state parameters and the agent state parameters in the regional coverage strategy; obtaining a current coverage benefit based on the current solution space, and updating the current solution space of the coverage strategy optimization model based on the current coverage benefit; using the updated current solution space as a new current solution space, and repeating the above steps until a preset number of iterations is reached or a preset convergence condition is reached, and determining the target area coverage strategy for the area to be covered based on the current solution space finally obtained.

[0145] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. A person of ordinary skill in the art can understand and implement it without creative labor. Through the description of the above implementation methods, a person of ordinary skill in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, by hardware. Based on such an understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., and includes a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or some parts of the embodiment. Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing regional coverage strategy based on multi-agent collaboration, characterized in that: include: Acquire a current solution space of a pre-constructed coverage strategy optimization model, wherein the current solution space corresponds to a current regional coverage strategy, and the coverage strategy optimization model is constructed based on coverage area state parameters and agent state parameters in the regional coverage strategy; the agent includes an unmanned vehicle, an unmanned aerial vehicle, or other unmanned system; the coverage area state parameter includes at least one of the area of ​​the area to be covered, intelligence data, and importance; the agent state parameter includes at least one of the number of agents, the scanning width of the agent, and the driving speed of the agent; Acquire a current coverage benefit based on the current solution space, and update a current solution space of the coverage strategy optimization model based on the current coverage benefit; The updated current solution space is used as the new current solution space, and the above steps are repeated until a preset number of iterations is reached or a preset convergence condition is reached, and a target area coverage strategy for the area to be covered is determined based on the current solution space finally obtained; The obtaining of the current coverage benefit based on the current solution space includes: Based on the current solution space, obtaining a current region segmentation result of the area to be covered; Based on the current region segmentation result, obtaining the current coverage path length of the area to be covered; Based on the current coverage path length, obtaining the current coverage benefit obtained by the agent performing the coverage task corresponding to the area to be covered; The obtaining, based on the current solution space, a current region segmentation result of the region to be covered includes: Based on the current solution space, obtaining a region segmentation method of the area to be covered and the number of agents and agent capability parameters corresponding to the agents in the area to be covered; Based on the area to be covered and the agent capability parameter, obtaining an expected area of ​​a sub-area of ​​each agent corresponding to the area to be covered; Acquire multiple segmentation lines based on the area segmentation method, the number of agents, and the area to be covered, and adjust the positions of the segmentation lines based on the expected area of ​​the sub-area; Based on the adjusted multiple segmentation lines, the area to be covered is segmented into multiple target sub-segmentation areas, and the multiple target sub-segmentation areas are used as the current area segmentation results; The updating of the current solution space of the coverage strategy optimization model based on the current coverage benefit includes: Obtain the last coverage gain, and determine the current optimization direction based on the current coverage gain and the last coverage gain; Acquire an initial coverage gain and an initial global movement step length, and determine a current global movement step length based on the initial coverage gain, the last coverage gain and the initial global movement step length; Based on the current global moving step size and the current optimization direction, globally optimizing and updating the current solution space; The current local moving step length is obtained, and based on the current local moving step length and the current optimization direction, a local optimization update is performed on the current solution space after the global optimization update.

2. The method for optimizing regional coverage strategy based on multi-agent collaboration according to claim 1 is characterized in that: The obtaining, based on the current region segmentation result, the current coverage path length of the region to be covered includes: Obtaining the area width of each of the target sub-divided areas and the first driving distance of the agent in each of the target sub-divided areas; For each target sub-divided area, obtaining a minimum number of turns of the agent based on the area width corresponding to the target sub-divided area, and obtaining a second driving distance of the agent outside the target sub-divided area based on the minimum number of turns; Based on the first driving distance corresponding to each of the target sub-divided areas, the driving distance of the agent in the current area within the area to be covered is obtained; based on the second driving distance corresponding to each of the target sub-divided areas, the driving distance outside the current area outside the area to be covered is obtained; The current covered path length is obtained based on the driving distance in the current area and the driving distance outside the current area, and the current covered path length includes the driving distance in the current area and the driving distance outside the current area.

3. The method for optimizing regional coverage strategy based on multi-agent collaboration according to claim 1 is characterized in that: The obtaining, based on the current coverage path length, a current coverage benefit obtained by the agent performing the coverage task corresponding to the area to be covered includes: Based on the current coverage path length, obtaining the driving distance of the intelligent agent within the area to be covered and the driving distance outside the area to be covered; Based on the driving distance in the current area, obtaining the current information benefit obtained by the agent performing the coverage task corresponding to the area to be covered; Based on the driving distance in the current area and the driving distance outside the current area, obtaining the current coverage energy consumption consumed by the agent to perform the coverage task corresponding to the area to be covered; Based on the current information benefit and the current coverage energy consumption, the current coverage benefit is obtained.

4. The method for optimizing regional coverage strategy based on multi-agent collaboration according to claim 1, characterized in that: After determining the target area coverage strategy based on the current solution space, the method further includes: Determine the competitive constraint conditions for obtaining the corresponding coverage task of the area to be covered based on the target area coverage strategy; Based on the competition constraint condition, acquiring each agent group competing for the coverage task corresponding to the area to be covered and agent capability data of each of the agent groups; The regional competition index value of each of the agent groups is obtained based on the agent capability data, and the target agent group corresponding to the area to be covered is determined based on the regional competition index value.

5. The method for optimizing regional coverage strategy based on multi-agent collaboration according to claim 4 is characterized in that: The step of obtaining the regional competition index value of each of the agent groups based on the agent capability data includes: Obtaining the group coverage time for each of the intelligent agent groups to complete the coverage task corresponding to the area to be covered; Obtaining the group travel time of each of the intelligent agent groups from the current position to the area to be covered; Obtaining the group coverage benefit obtained by each of the intelligent agent groups for completing the coverage task corresponding to the area to be covered; Based on the group coverage time, the group travel time and the group coverage benefit, the regional competition index value of each of the agent groups is obtained.

6. A regional coverage strategy optimization device based on multi-agent collaboration, characterized in that: include: A data acquisition module is used to acquire the current solution space of a pre-constructed coverage strategy optimization model, wherein the current solution space corresponds to the current regional coverage strategy, and the coverage strategy optimization model is constructed based on the coverage area state parameters and agent state parameters in the regional coverage strategy; the agent includes an unmanned vehicle, an unmanned aerial vehicle or other unmanned system; the coverage area state parameters include at least one of the area of ​​the area to be covered, intelligence data and importance; the agent state parameters include at least one of the number of agents, the scanning width of the agent and the driving speed of the agent; A first optimization module, configured to obtain a current coverage benefit based on the current solution space, and update a current solution space of the coverage strategy optimization model based on the current coverage benefit; A second optimization module is used to use the updated current solution space as a new current solution space, and repeat the above steps until a preset number of iterations is reached or a preset convergence condition is reached, and determine a target area coverage strategy for the area to be covered based on the current solution space finally obtained; The obtaining of the current coverage benefit based on the current solution space includes: Based on the current solution space, obtaining a current region segmentation result of the area to be covered; Based on the current region segmentation result, obtaining the current coverage path length of the area to be covered; Based on the current coverage path length, obtaining the current coverage benefit obtained by the agent performing the coverage task corresponding to the area to be covered; The obtaining, based on the current solution space, a current region segmentation result of the region to be covered includes: Based on the current solution space, obtaining a region segmentation method of the area to be covered and the number of agents and agent capability parameters corresponding to the agents in the area to be covered; Based on the area to be covered and the agent capability parameter, obtaining an expected area of ​​a sub-area of ​​each agent corresponding to the area to be covered; Acquire multiple segmentation lines based on the area segmentation method, the number of agents, and the area to be covered, and adjust the positions of the segmentation lines based on the expected area of ​​the sub-area; Based on the adjusted multiple segmentation lines, the area to be covered is segmented into multiple target sub-segmentation areas, and the multiple target sub-segmentation areas are used as the current area segmentation results; The updating of the current solution space of the coverage strategy optimization model based on the current coverage benefit includes: Obtain the last coverage gain, and determine the current optimization direction based on the current coverage gain and the last coverage gain; Acquire an initial coverage gain and an initial global movement step length, and determine a current global movement step length based on the initial coverage gain, the last coverage gain and the initial global movement step length; Based on the current global moving step size and the current optimization direction, globally optimizing and updating the current solution space; The current local moving step length is obtained, and based on the current local moving step length and the current optimization direction, a local optimization update is performed on the current solution space after the global optimization update.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, it implements the area coverage strategy optimization method based on multi-agent collaboration as described in any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the area coverage strategy optimization method based on multi-agent collaboration as described in any one of claims 1 to 5.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, it implements the area coverage strategy optimization method based on multi-agent collaboration as described in any one of claims 1 to 5.

Images