Differential evolution algorithm-based satellite-ground networking task planning method

Abstract

The invention discloses a satellite-ground networking task planning method based on a differential evolution algorithm, and relates to the field of satellite-ground networking task planning. The method comprises the following steps: firstly, obtaining a list of arc segments to be planned and constraint conditions, and generating a corresponding satellite-ground networking task scheme initial population; then corresponding differential variation individuals are generated; aiming at each individual in the current population and the corresponding differential variation individual, generating a corresponding cross individual; for each individual and the corresponding cross individual in the current population, respectively calculating a corresponding fitness function value, and selecting the individual with the large fitness function value for reservation to generate a new population; and after the steps are iteratively executed, selecting an individual with the maximum fitness function value from the final population as a final satellite-ground networking task scheme. According to the method, the task allocation efficiency can be improved, the optimal utilization of resources is ensured, and a satellite-ground networking task planning scheme conforming to an optimization target is generated.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of satellite-to-ground networking task planning, and particularly relates to a satellite-to-ground networking task planning method based on a differential evolution algorithm, which is used for intelligent planning and optimization of satellite-to-ground networking tasks. BACKGROUND

[0002] The satellite-to-ground networking task aims to realize normal and intelligent management and dynamic optimization configuration of satellite-to-ground links, and is a core link of resource scheduling and link construction in a space-ground integrated network system. The task is usually implemented in a periodic manner, and when the state of networking resources or the external operating environment changes, the system needs to be dynamically adjusted in real time to ensure the continuous availability and communication quality of space-ground links.

[0003] From the perspective of resource allocation, the core goal of the satellite-to-ground networking task is to maximize resource use efficiency to achieve optimal space-ground path capacity and coverage. The scheduling objects cover multi-level system resources, including satellites and their on-board feeder loads, ground stations and their device antennas, etc. Task planning needs to comprehensively arrange the link establishment pairing relationship and order of satellites and ground stations, the establishment time of feeder links, the working scheme of satellites and loads, the working mode of ground stations and antennas, and other multi-dimensional factors.

[0004] Currently, in the field of space-ground integrated network resource scheduling and link construction, a variety of rule-based, heuristic and evolutionary algorithms have been proposed to solve complex resource allocation problems. For example, related research summarizes the application and challenges of rule-based, heuristic and evolutionary algorithms in satellite task scheduling, revealing the limitations of existing methods in multi-objective, multi-constraint and real-time adjustment scenarios. Another research proposes a large neighborhood search algorithm combined with knowledge assistance for satellite-ground link scheduling tasks, which has made some progress in improving scheduling efficiency, but still has problems of insufficient global search ability and poor real-time responsiveness. Another research summarizes the advantages of differential evolution algorithm in continuous and mixed variable optimization, such as fewer parameters and stronger global search ability, providing a theoretical reference for subsequent scheduling optimization. SUMMARY

[0005] Therefore, the present application provides a satellite-to-ground networking task planning method based on a differential evolution algorithm to solve the problems in the prior art. In particular, when dealing with complex and large-scale networking task scenarios, the calculation amount of traditional linear programming methods increases exponentially with the increase of constraint conditions, which cannot guarantee the efficiency and reliability of the system. The method of the present application can improve the efficiency of task allocation, ensure the optimal use of resources, and generate a satellite-to-ground networking task planning scheme that meets the optimization goal.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] A satellite-to-ground networking mission planning method based on differential evolution algorithm includes the following steps:

[0008] Step 1: Obtain the list of arc segments to be planned, the shortest link establishment time, and the antenna switching time;

[0009] Step 2: For the list of arc segments to be planned, generate the initial population of the corresponding satellite-to-ground networking task schemes;

[0010] Step 3: For each individual in the current population, generate its corresponding differentially mutated individual;

[0011] Step 4: For each individual in the current population and its corresponding differentially mutated individual, generate the corresponding crossover individual;

[0012] Step 5: For each individual in the current population and the corresponding crossover individual, calculate the corresponding fitness function value, and select the individual with the larger fitness function value to retain, thereby generating a new population;

[0013] Step 6: Iterate through steps 3 to 5 until the maximum number of iterations is reached. Select the individual with the largest fitness function value from the final population as the final satellite-to-ground networking mission scheme, thus completing the satellite-to-ground networking mission planning based on the differential evolution algorithm.

[0014] Furthermore, the specific method for step 2 is as follows:

[0015] Based on a preset population size, multiple different individuals are randomly generated, each individual being... Bit-bit binary encoded string Let be the number of arc segments to be planned. If the th arc segment... If the binary code is 0, it means that the arc segment is not selected. A binary code of 1 indicates that the arc segment has been selected. .

[0016] Furthermore, step 3 is performed as follows:

[0017] Step 301: For the current population Individuals in Generate its corresponding initial differential mutation individuals. :

[0018]

[0019] in, , where is the scaling factor, and the operations in the formula are all calculations of the bit positions corresponding to different individuals;

[0020] one individual in the population except , the selection principle is:

[0021] one individual in the population except , the selection principle is: , the fitness function value of each individual is calculated respectively, and then the selection probability is randomly generated , if , the individual with the maximum fitness function value is selected as from the selected ; if , the individual with the minimum fitness function value is selected as from the selected ;

[0022] one individual in the population except , , , the selection principle is:

[0023] one individual in the population except , , , the selection principle is: , the fitness function value of each individual is calculated respectively, and then the selection probability is randomly generated , if , the individual with the maximum fitness function value is selected as from the selected ; if , the individual with the minimum fitness function value is selected as from the selected ;

[0024] one individual in the population except , , , , the selection principle is:

[0025] one individual in the population except , , , , the selection principle is: , the fitness function value of each individual is calculated respectively, and then the selection probability is randomly generated , if , the individual with the maximum fitness function value is selected as​​​ ;like Then in the selected Each individual selects the individual with the smallest fitness function value as... ;

[0026] ; ; ; Population size; When selecting the individual with the largest or smallest fitness function value, if there are individuals with the same fitness function value, then one individual with the same fitness function value is randomly selected for subsequent operations.

[0027] Step 302: Based on the initial differential variation individuals Calculate its The mapping probabilities corresponding to each bit of binary encoding :

[0028]

[0029] Further generation of individuals Corresponding differentially mutated individuals :

[0030]

[0031] Represents the initial difference variation individuals The first in Bit-bin encoding Represents an individual Corresponding differentially mutated individuals The first in Bit-bin encoding Indicates that for the first Bit-binary encoding generated Random numbers within.

[0032] Furthermore, the fitness function value is calculated as follows:

[0033] For each individual, first determine whether it meets the constraints of the shortest link establishment time and antenna switching time:

[0034] If the duration of a continuous arc segment corresponding to a certain antenna in an individual is less than the shortest link establishment time, then the individual is invalid and its fitness function value is recorded as 0.

[0035] If an individual needs to switch antennas between two adjacent arc segments, and the interval between adjacent arc segments is less than the antenna switching time, then the individual is invalid and its fitness function value is recorded as 0.

[0036] Otherwise, calculate the fitness function value of the current individual:

[0037]

[0038] wherein, is a weight coefficient, ;

[0039]

[0040]

[0041]

[0042]

[0043]

[0044] denotes the total duration of the current chain establishment networking, denotes the maximum value of the total duration of the chain establishment networking, denotes the minimum value of the total duration of the chain establishment networking, which is 0 by default, denotes the normalized total duration of the current chain establishment networking;

[0045] denotes the number of antennas in the current chain establishment, denotes the maximum value of the number of antennas in the chain establishment, denotes the minimum value of the number of antennas in the chain establishment, which is 1 by default, denotes the normalized number of antennas in the chain establishment;

[0046] denotes the sum of priorities in the current chain establishment, denotes the maximum value of the sum of priorities, denotes the minimum value of the sum of priorities, which is 0 by default, denotes the normalized sum of priorities;

[0047] denotes the duration of single chain establishment, denotes the maximum value of the duration of single chain establishment, denotes the minimum value of the duration of single chain establishment, which is 0 by default, denotes the normalized duration of single chain establishment;

[0048] denotes the number of antenna switching, denotes the maximum value of the number of antenna switching, denotes the number of antenna switching, which is 0 by default, denotes the normalized number of antenna switching.

[0049] Further, the specific manner of step 4 is:

[0050] For the individual in the current population , and the corresponding differential variation individual , the corresponding cross individual is calculated:

[0051]

[0052] Wherein, represents the binary code in the individual , th position, , is the crossover probability, represents the random integer within the generated , ; is the binary code of the cross individual th position.

[0053] Further, the specific manner of step 5 is:

[0054] For the individual in the current population , generate the corresponding next generation individual :

[0055]

[0056] Finally, a new population is generated.

[0057] Due to the adoption of the above technical solutions, the beneficial effects of the present application compared with the prior art are:

[0058] 1、The present application uses differential evolution algorithm to quickly find the optimal solution or approximate optimal solution through unique mutation and crossover operation, which significantly improves the efficiency of task allocation.

[0059] 2、The present application can effectively avoid the problem of algorithm falling into local optimal solution through the random selection and global optimization mechanism of differential evolution algorithm, and ensure the global optimization of satellite-terrestrial networking task allocation.

[0060] 3、The present application can reasonably allocate satellite-terrestrial link resources considering multiple factors, maximize resource utilization, and ensure timely completion of tasks while avoiding resource waste.

[0061] 4、The method of the present application can flexibly cope with large-scale complex satellite-terrestrial networking task scenarios, including multiple satellites, multiple stations, multiple constraint conditions, etc. ​BRIEF DESCRIPTION OF DRAWINGS

[0062] Figure 1 A flowchart of the overall process of a satellite-to-ground networking task planning method based on a differential evolution algorithm in an embodiment of the present application. DETAILED DESCRIPTION

[0063] The present application will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0064] A satellite-to-ground networking task planning method based on a differential evolution algorithm, as shown in FIG. 1, includes the following steps: Figure 1

[0065] Step 1, obtaining a list of to-be-planned arc segments, a minimum chain building time length, and an antenna switching time length;

[0066] Step 2, generating an initial population of satellite-to-ground networking task schemes corresponding to the list of to-be-planned arc segments;

[0067] Step 3, generating a differential mutation individual corresponding to each individual in the current population;

[0068] Step 4, generating a corresponding crossover individual for each individual in the current population and the corresponding differential mutation individual;

[0069] Step 5, calculating the corresponding fitness function value for each individual in the current population and the corresponding crossover individual, respectively, and selecting the individual with a larger fitness function value for reservation to generate a new population;

[0070] Step 6, iteratively executing steps 3-5 until the maximum number of iterations is reached, selecting the individual with the largest fitness function value in the final population as the final satellite-to-ground networking task scheme, and completing the satellite-to-ground networking task planning based on the differential evolution algorithm.

[0071] Further, the specific manner of step 2 is as follows:

[0072] Based on a preset population size, a plurality of different individuals are randomly generated, each individual being a binary code string, wherein n is the number of to-be-planned arc segments, and if the binary code at the i th position is 0, it represents that the i th arc segment is not selected, and if the binary code at the i th position is 1, it represents that the i th arc segment is selected. Specifically, as shown in Table 1:

[0073]

[0074] Further, the specific manner of step 3 is as follows:

[0075] Step 301, for each individual in the current population, generating a differential mutation individual corresponding to the individual;​​​​ Individuals in Generate its corresponding initial differential mutation individuals. :

[0076]

[0077] in, , where is the scaling factor, and the operations in the formula are all calculations of the bit positions corresponding to different individuals;

[0078] For population Except The selection principle for the individual other than the one mentioned above is:

[0079] In population Except Any other For each individual, calculate its corresponding fitness function value, and then randomly generate the selection probability. ,like Then in the selected Each individual selects the individual with the largest fitness function value as... ;like Then in the selected Each individual selects the individual with the smallest fitness function value as... ;

[0080] For population Except , The selection principle for the individual other than the one mentioned above is:

[0081] In population Except , Any other For each individual, calculate its corresponding fitness function value, and then randomly generate the selection probability. ,like Then in the selected Each individual selects the individual with the largest fitness function value as... ;like Then in the selected Each individual selects the individual with the smallest fitness function value as... ;

[0082] For population Except , , The selection principle for the individual other than the one mentioned above is:

[0083] In population Except , , Any other For each individual, calculate its corresponding fitness function value, and then randomly generate the selection probability. ,like Then in the selected Each individual selects the individual with the largest fitness function value as... ;like Then in the selected Each individual selects the individual with the smallest fitness function value as... ;

[0084] ; ; ; Population size; When selecting the individual with the largest or smallest fitness function value, if there are individuals with the same fitness function value, then one individual with the same fitness function value is randomly selected for subsequent operations.

[0085] Specifically, in this embodiment, ;

[0086] Step 302: Based on the initial differential variation individuals Calculate its The mapping probabilities corresponding to each bit of binary encoding :

[0087]

[0088] Further generation of individuals Corresponding differentially mutated individuals :

[0089]

[0090] Represents the initial difference variation individuals The first in Bit-bin encoding Represents an individual Corresponding differentially mutated individuals The first in Bit-bin encoding Indicates that for the first Bit-binary encoding generated Random numbers within.

[0091] Furthermore, the fitness function value is calculated as follows:

[0092] For each individual, firstly, it is judged whether it meets the constraints of the minimum link establishment time length and the antenna switching time length:

[0093] If the duration of a certain continuous arc corresponding to a certain antenna in the individual is less than the minimum link establishment time length, the individual is invalid, and the fitness function value of the individual is recorded as 0;

[0094] If the antenna needs to be switched between two adjacent arcs in the individual, and the interval time between the adjacent arcs is less than the antenna switching time length, the individual is invalid, and the fitness function value of the individual is recorded as 0;

[0095] Otherwise, the fitness function value of the current individual is calculated:

[0096]

[0097] Wherein, is a weight coefficient, ; are numbers between 0 and 1;

[0098]

[0099]

[0100]

[0101]

[0102]

[0103] In this embodiment, ;

[0104] denotes the total time length of the link establishment networking this time, denotes the maximum value of the total time length of the link establishment networking, denotes the minimum value of the total time length of the link establishment networking, which is 0 by default, denotes the total time length of the link establishment networking this time after normalization processing;

[0105] Specifically, that is, the total time length of the arc corresponding to "1" in the individual, that is, the total time length of the N arcs;

[0106] denotes the number of antennas for link establishment this time, denotes the maximum value of the number of antennas for link establishment, denotes the minimum value of the number of antennas for link establishment, which is 1 by default, denotes the number of antennas for link establishment after normalization processing;

[0107] Specifically, each arc segment corresponds to its own antenna. That is, the number of antenna types corresponding to all "1"s in an individual; That is, the number of antenna types corresponding to N arc segments;

[0108] This represents the sum of priorities for this chain establishment. This represents the maximum value of the sum of priority levels. The minimum value of the sum of priorities is 0 by default. This represents the sum of priorities after normalization.

[0109] Specifically, each arc segment has its own priority; the smaller the priority value, the higher the priority.

[0110] Indicates the duration of a single chain establishment. This represents the maximum duration of a single chain establishment. The minimum time for a single chain establishment is 0 by default. This represents the duration of a single chain establishment after normalization.

[0111] Specifically, That is The value is divided by the number of all "1"s in the individual. That is Divide the value by N;

[0112] Indicates the number of antenna switching times. This indicates the maximum number of antenna switching attempts. This indicates that the default number of antenna switching attempts is 0. This represents the number of antenna switching operations after normalization.

[0113] Specifically, This refers to the number of times the antenna is switched between arc segments within an individual;

[0114] Furthermore, step 4 is specifically implemented as follows:

[0115] For the current population Individuals in and its corresponding differentially mutated individuals. Calculate the corresponding crossover individuals :

[0116]

[0117] in, Represents an individual The first in Bit-bin encoding , is the crossover probability, represents an individual generated a random integer within the range of 0 and 1; is the crossover individual the first bit binary code of the individual. ensuring that at least one bit is from , avoiding = .

[0118] In this embodiment, CR = 0.5;

[0119] Further, the specific way of step 5 is:

[0120] For each individual in the current population , generate the corresponding next generation individual :

[0121]

[0122] Finally, generate a new population.

[0123] Specifically, in this embodiment, there are population , each individual is a real number vector. Each time, randomly pick several individuals from the population for comparison; as shown in Table 2:

[0124]

[0125] The next generation population obtained is:

[0126]

[0127] Those skilled in the art will realize that the embodiments described are for the purpose of helping the reader understand the principles of the present application, and should be understood as not limiting the scope of protection of the present application to the embodiments described. The present application can have various modifications and changes for those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the scope of protection of the claims of the present application.

Claims

1. A satellite-to-ground networking mission planning method based on differential evolution algorithm, characterized in that, Includes the following steps: Step 1: Obtain the list of arc segments to be planned, the shortest link establishment time, and the antenna switching time; Step 2: For the list of arc segments to be planned, generate the initial population of the corresponding satellite-to-ground networking task schemes; Step 3: For each individual in the current population, generate its corresponding differentially mutated individual; Step 4: For each individual in the current population and its corresponding differentially mutated individual, generate the corresponding crossover individual; Step 5: For each individual in the current population and the corresponding crossover individual, calculate the corresponding fitness function value, and select the individual with the larger fitness function value to retain, thereby generating a new population; Step 6: Iterate through steps 3 to 5 until the maximum number of iterations is reached. Select the individual with the largest fitness function value from the final population as the final satellite-to-ground networking mission scheme, thus completing the satellite-to-ground networking mission planning based on the differential evolution algorithm.

2. The satellite-to-ground networking mission planning method based on differential evolution algorithm according to claim 1, characterized in that, The specific method for step 2 is as follows: Based on a preset population size, multiple different individuals are randomly generated, each individual being... Bit-bit binary encoded string Let be the number of arc segments to be planned. If the th arc segment... If the binary code is 0, it means that the arc segment is not selected. A binary code of 1 indicates that the arc segment has been selected. .

3. The satellite-to-ground networking mission planning method based on differential evolution algorithm according to claim 2, characterized in that, The specific method for step 3 is as follows: Step 301: For the current population Individuals in Generate its corresponding initial differential mutation individuals. : in, , where is the scaling factor, and the operations in the formula are all calculations of the bit positions corresponding to different individuals; For population Except The selection principle for the individual other than the one mentioned above is: In population Except Any other For each individual, calculate its corresponding fitness function value, and then randomly generate the selection probability. ,like Then in the selected Each individual selects the individual with the largest fitness function value as... ;like Then in the selected Each individual selects the individual with the smallest fitness function value as... ; For population Except , The selection principle for the individual other than the one mentioned above is: In population Except , Any other For each individual, calculate its corresponding fitness function value, and then randomly generate the selection probability. ,like Then in the selected Each individual selects the individual with the largest fitness function value as... ;like Then in the selected Each individual selects the individual with the smallest fitness function value as... ; For population Except , , The selection principle for the individual other than the one mentioned above is: In population Except , , Any other For each individual, calculate its corresponding fitness function value, and then randomly generate the selection probability. ,like Then in the selected Each individual selects the individual with the largest fitness function value as... ;like Then in the selected Each individual selects the individual with the smallest fitness function value as... ; ; ; ; Population size; When selecting the individual with the largest or smallest fitness function value, if there are individuals with the same fitness function value, then one individual with the same fitness function value is randomly selected for subsequent operations. Step 302: Based on the initial differential variation individuals Calculate its The mapping probabilities corresponding to each bit of binary encoding : Further generation of individuals Corresponding differentially mutated individuals : Represents the initial difference variation individuals The first in Bit-bin encoding Represents an individual Corresponding differentially mutated individuals The first in Bit-bin encoding Indicates that for the first Bit-binary encoding generated Random numbers within.

4. The satellite-to-ground networking mission planning method based on differential evolution algorithm according to claim 3, characterized in that, The fitness function value is calculated as follows: For each individual, first determine whether it meets the constraints of the shortest link establishment time and antenna switching time: If the duration of a continuous arc segment corresponding to a certain antenna in an individual is less than the shortest link establishment time, then the individual is invalid and its fitness function value is recorded as 0. If an individual needs to switch antennas between two adjacent arc segments, and the interval between adjacent arc segments is less than the antenna switching time, then the individual is invalid and its fitness function value is recorded as 0. Otherwise, calculate the fitness function value of the current individual: in, These are weighting coefficients. ; This indicates the total duration of the chain establishment and network formation. This represents the maximum total time required for establishing a network. The minimum value for the total time required to establish a network connection is 0 by default. This represents the total time for this chain establishment and networking process after normalization. This indicates the number of antennas used in this link establishment. This represents the maximum number of antennas that can establish a link. The minimum number of antennas required to establish a link is 1 by default. This represents the number of linked antennas after normalization. This represents the sum of priorities for this chain establishment. This represents the maximum value of the sum of priority levels. The minimum value of the sum of priorities is 0 by default. This represents the sum of priorities after normalization. Indicates the duration of a single chain establishment. This represents the maximum duration of a single chain establishment. The minimum time for a single chain establishment is 0 by default. This represents the duration of a single chain establishment after normalization. Indicates the number of antenna switching times. This indicates the maximum number of antenna switching attempts. This indicates that the default number of antenna switching attempts is 0. This represents the number of antenna switching operations after normalization.

5. The satellite-to-ground networking mission planning method based on differential evolution algorithm according to claim 4, characterized in that, The specific method for step 4 is as follows: For the current population Individuals in and its corresponding differentially mutated individuals. Calculate the corresponding crossover individuals : in, Represents an individual The first in Bit-bin encoding , For crossover probability, Represents an individual generated A random integer within; For cross-individuals The Bit-binary encoding.

6. The satellite-to-ground networking mission planning method based on differential evolution algorithm according to claim 5, characterized in that, The specific method for step 5 is as follows: For the current population Individuals in Generate the corresponding next generation of individuals : This eventually leads to the formation of a new population.

Images