An iterative adaptive large neighborhood search method for large-scale communication satellite tracking and control scheduling problems

Abstract

This paper proposes an iterative adaptive large neighborhood search method for the large-scale communication satellite TT&C scheduling problem, aiming to rapidly solve it. Unlike traditional TT&C problems, this one considers intersatellite links to facilitate the rapid completion of TT&C tasks. In this method, an adaptive large neighborhood search algorithm improves a heuristic initial solution. A frequent pattern mining method is then embedded into the iterative mechanism to guide the algorithm's search process. On the one hand, poor structures are mined from low-quality solutions to guide the adaptive large neighborhood search algorithm's removal process; on the other hand, advantageous structures are mined from high-quality solutions to guide the construction of new solutions. Experimental test results at various task scales demonstrate that the iterative adaptive large neighborhood search algorithm can effectively solve the communication satellite TT&C scheduling problem, outperforming three state-of-the-art algorithms.

Description

Technical field:

[0001] The present invention relates to the field of satellite measurement and control technology, and in particular to an iterative adaptive large neighborhood search method for large-scale communication satellite measurement and control scheduling problems. Background technology:

[0002] A network of high-speed communications satellites operating in low orbits forms a low-orbit internet constellation. With their broadband communications capabilities, low latency, and global coverage, these constellations are an indispensable complement to terrestrial networks. For example, Starlink and OneWeb provide global broadband access to support global communications missions, including mobile services, emergency communications, and emerging services such as multimedia and the internet.

[0003] To improve the coverage and sustainability of low-orbit internet constellations, their size continues to expand. Furthermore, the trend is for LEO internet constellations to be jointly managed with other constellations, such as remote sensing constellations. Consequently, extensive communication between satellites and ground stations is required to ensure the proper operation of the constellation. To ensure that satellite-to-ground communication requests for LEO internet constellations are met and to improve their efficiency, this paper extends the Satellite Range Scheduling Problem (SRSP) to communication satellite constellations and proposes a large-scale Communication Satellite Range Scheduling Problem (CSRSP).

[0004] According to Luo's research, the TT&C problem primarily involves selecting the appropriate ground antenna and time window for satellite requests to maximize total profit. However, the scheduling process for communication satellite TT&C scheduling differs from traditional TT&C problems. From a constellation perspective, low-Earth orbit internet constellations possess onboard processing and inter-satellite forwarding capabilities. Each satellite establishes four inter-satellite links: two links connecting adjacent satellites within the same orbital plane, and two links connecting satellites in adjacent orbital planes. These inter-satellite links increase the opportunities for TT&C missions and compensate for resource constraints. TT&C missions can transmit satellite data directly to ground stations via satellite-to-ground links, or forward this data to adjacent satellites via inter-satellite links. The combination and sequence of satellite-to-ground links in low-Earth orbit internet constellations are pre-arranged, and the time periods during which satellite-to-ground links are established are defined as arcs, within which TT&C missions must be executed. Regarding resource types, each satellite and ground antenna is equipped with two types of payloads: millimeter wave (Q / V band) and microwave (S band). Therefore, there are two types of arcs that can be allocated as resources for performing TT&C tasks, but only one of the two arcs generated simultaneously by the same satellite and ground antenna can be used as an effective resource. In terms of mission type, there are many types of TT&C tasks, each with specific resource requirements, which complicates the scheduling process.

[0005] The establishment of intersatellite links between satellites provides more opportunities for mission execution, thereby expanding the knowledge space. At the same time, different types of resources and tasks also increase the complexity of constraints. Based on the characteristics of the communication satellite TT&C scheduling problem, this paper establishes an integer programming model. Then, an iterative adaptive large neighborhood search algorithm (IALNS) is proposed. This algorithm extends the adaptive large neighborhood search algorithm (ALNS) by incorporating the iterative local search (ILS) algorithm into ALNS to solve the TT&C scheduling problem for communication satellites in low-Earth orbit internet constellations. Summary of the invention:

[0006] To solve the above technical problems, the present invention proposes an iterative adaptive large neighborhood search method for large-scale communication satellite measurement, control and scheduling problems, which specifically includes the following steps:

[0007] S1. Obtain input data, including arc segment data and mission data. The arc segment data is the result of satellite-to-ground link scheduling. An arc segment is considered an executable arc segment for a mission if it meets the following conditions: 1) the arc segment type matches the mission type; 2) the arc segment belongs to the same satellite as the mission satellite or has an intersatellite link with the mission satellite; 3) the mission executable time is within the arc segment range.

[0008] S2. Generate three initial solutions based on the three heuristic rules. The resulting solutions are scheduling plans, where the scheduling plans include the tasks to be executed, the execution arcs of the tasks, and the execution times.

[0009] S3. Randomly input the three initial solutions into three threads and run ALNS in parallel on each thread. Each thread runs the following steps:

[0010] S3.1. Based on the weights of each operator, remove and insert operators according to a roulette wheel;

[0011] If the conditions for activating the iterative mechanism 1 of the frequent pattern mining method are not met, proceed to S3.3. Otherwise, the ALNS removal process is assisted by the frequent pattern mining method. The specific steps of the iterative mechanism 1 based on the frequent pattern mining method are as follows:

[0012] S3.2.1. Data conversion: convert each partial solution task sequence in the low-quality solution set, each of which is a single task, into a sequence of each task consisting of two adjacent tasks;

[0013] S3.2.2. Use the FP-Growth algorithm to mine frequent patterns. Build an FP-tree based on the dataset. To mine frequent patterns, the algorithm traverses paths starting from leaf nodes and generates a conditional pattern tree. Frequent itemsets are then mined from these paths.

[0014] S3.2.3. Remove the mined maximum frequent pattern set from the solution;

[0015] S3.3. Use the remove operator to remove the solution, delete the corresponding task, and clear the low-quality solution set;

[0016] S3.4. Use the insertion operator to repair the solution and obtain a new solution, including sorting the unscheduled tasks and inserting them sequentially using the insertion algorithm;

[0017] If the new solution's payoff is greater than the current solution, accept it. Otherwise, decide whether to accept the new solution according to the simulated annealing criterion.

[0018] S3.6. If the new solution is better than the optimal solution, it is recorded as an elite solution. If the profit value of the new solution is more than 1% lower than the current solution, it is recorded as a low-quality solution. The elite solution will be continuously replaced and updated during the iteration process.

[0019] S3.6. If the termination condition is met, terminate and output the elite solution set; otherwise, return to S3.1;

[0020] S4. If the termination condition is met, the process terminates; otherwise, iterative mechanism 2 starts.

[0021] S5. After the termination condition is met, the obtained solution is used to solve the communication satellite scheduling problem and is applied to determine the tasks to be executed, the execution arc of the tasks, and the execution time.

[0022] Furthermore, the mission data includes scenarios with different numbers of missions. The number of missions for each satellite included in the scenarios ranges from 50 to 150, increasing in increments of 10. All missions are randomly generated based on actual conditions.

[0023] Furthermore, the ALNS is an ALNS algorithm search process assisted by an iterative mechanism.

[0024] Furthermore, the ALNS includes an operator adaptation layer, a neighborhood operator, a new solution, and a preset iterative model processing process.

[0025] Furthermore, the preset iterative model includes the processes of recording low-quality solutions, judging disturbance conditions, converting data, mining frequent patterns, and judging termination criteria.

[0026] Furthermore, the specific steps of the iterative mechanism 2 are as follows: S4.1. The elite solution sets output by the three threads are combined and data converted, and each solution sequence of a single task in the elite solution set is converted into a sequence of two adjacent tasks;

[0027] S4.2. Use the FP-Growth algorithm to mine frequent patterns. Build an FP-tree based on the dataset. To mine frequent patterns, traverse paths starting from leaf nodes and generate a conditional pattern tree. Then, mine frequent itemsets from these paths.

[0028] S4.3. Construct two new solutions based on the mined maximal frequent pattern set. First, insert the task in the frequent pattern. Then, use the optimal solution and a randomly selected solution from the elite solution set to guide the insertion process and obtain two new solutions.

[0029] S4.4. The two new solutions obtained and the optimal solution are updated as the initial solution and enter the next round of parallel ALNS together, returning to S3.

[0030] Furthermore, the deleting of the corresponding tasks is specifically to sort the tasks in the solution according to the removal operator and remove them in sequence until a specified number is reached.

[0031] According to another aspect of the present invention, a system for iterative adaptive large neighborhood search for large-scale communication satellite measurement, control and scheduling problems is provided. The system includes multiple functional modules, and the functional modules implement any of the above-mentioned methods.

[0032] According to another aspect of the present invention, there is further provided an electronic device, comprising:

[0033] at least one processor, and a memory communicatively coupled to the at least one processor;

[0034] The memory stores instructions that can be executed by at least one of the processors, and the instructions are executed by at least one of the processors so that the at least one processor can execute any of the above-mentioned methods.

[0035] According to another aspect of the present invention, a computer-readable storage medium is provided, on which computer instructions are stored. The computer instructions are used to enable the computer to execute any of the above-mentioned methods. Description of the drawings:

[0036] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific 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 any creative work.

[0037] Figure 1 This is a flow chart of the iterative adaptive large neighborhood search algorithm in the present invention. Specific implementation method:

[0038] The following description of exemplary embodiments of the present application is made in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0039] Barbulescu et al. demonstrated that satellite TT&C scheduling on a single antenna corresponds to a traditional scheduling problem—minimizing the number of delayed jobs on a single machine. This conclusion has been extended to general TT&C problems. Extensive research has been conducted to automate TT&C scheduling. Mathematical programming, metaheuristics, and machine learning algorithms are the three main approaches to this problem.

[0040] Mathematical programming algorithms can ensure optimal solutions under certain assumptions. The challenge lies in applying operations research knowledge to construct a linear programming model for the problem. Gooley developed a mixed-integer programming model to effectively generate an initial 24-hour satellite tracking and control schedule, achieving over 91% of satellite scheduling requests. Linares introduced a mixed-integer linear programming model to optimize antenna-satellite scheduling (a type of tracking and control problem encountered in Earth observation satellite operations), improving efficiency by balancing time allocation and reducing conflicts.

[0041] However, the computational cost of mathematical programming is often too high to be applied to large-scale practical problems. Therefore, many researchers have begun to study metaheuristic algorithms and machine learning algorithms to solve these problems.

[0042] Metaheuristic algorithms are guided by experience or intuition and guide the search process through neighborhood operators or evolutionary operators. Chen introduced a genetic algorithm with population perturbation and elimination to optimize multi-satellite measurement and control scheduling, effectively improving profits and task completion rates. Wang proposed a multi-objective differential evolution algorithm based on space partitioning and adaptive selection strategy to solve multi-objective measurement and control problems. Machine learning algorithms aim to identify and extract hidden patterns from large data sets and are becoming increasingly popular in measurement and control problems. Ou effectively solved the measurement and control problem by combining deep reinforcement learning with heuristics. Liang integrated a mask processing mechanism into deep reinforcement learning for rapid rescheduling of measurement and control problems, and demonstrated optimization performance and computational efficiency through experimental results.

[0043] Most of the research on measurement and control problems focuses on remote sensing satellites, while communication satellite constellations have received less attention. Wu considered the problem of scheduling communication tasks on intersatellite links and proved that the problem is NP-complete. Gu combined the ground measurement and control resources of the constellation with the intersatellite links and studied how to comprehensively utilize the intersatellite links and ground resources in the field of measurement and control. Although these studies include intersatellite links, they do not consider multiple types of payloads and tasks. Therefore, based on the characteristics of communication satellite constellations, the present invention uniformly defines the measurement and control problems involving multiple types of payloads and tasks on intersatellite links as communication satellite measurement and control scheduling problems. In short, any measurement and control problem with a communication satellite constellation as a background or taking into account certain characteristics of a communication satellite constellation can be regarded as a communication satellite measurement and control scheduling problem and can be solved by the algorithm of the present invention.

[0044] One of the hallmarks of the proposed algorithm, ALNS, is its strategy of continuously iteratively enhancing solutions. This involves destroying parts of existing solutions and then reconstructing them, aiming to find improved solutions with each iteration. Typically, a set of destruction and repair operators is employed, with the choice of operators adaptively determined as the algorithm iterates. In the field of satellite scheduling, Liu et al. and He et al. extended and applied ALNS to single and multiple agile satellite scheduling problems, respectively. Liu et al. proposed a data-driven parallel ALNS algorithm for inter-satellite link scheduling in low-Earth orbit communication satellite constellations. The design of an iterative mechanism in ALNS is motivated by the concept of inter-satellite link scheduling (ILS), which has been successfully applied to satellite scheduling problems. This concept, by providing sufficient perturbations, helps the algorithm escape local optima and reduces ineffective searches. When the algorithm reaches a local optimum, an iterative mechanism is initiated. Frequent pattern mining methods are used within this iterative mechanism to overcome the limitations of blind search and enable the algorithm to explore a larger solution space in a targeted and informed manner.

[0045] When a satellite orbits the earth, it will pass over the ground station many times. When the ground antenna and the satellite beam overlap each other, a satellite-to-ground link can be established between the ground antenna and the satellite. The time period for establishing the satellite-to-ground link is defined as an arc segment. In the communication satellite measurement and control scheduling problem, both the satellite and the ground antenna are equipped with two types of payloads: millimeter wave band (Q / V band) and microwave band (S band). Therefore, when the conditions for establishing a satellite-to-ground link are met, two different types of arc segments will be generated at the same time. The arc segment generated by the Q / V band is called the feeding arc segment, and the arc segment generated by the S band is called the measurement and control arc segment, such as Figure 1 As shown. Only one of the feed arc and TT&C arc generated simultaneously by the same satellite and the same ground antenna can be used as a valid resource, so the two are in conflict with each other. The combination and order of satellite-to-ground links are pre-arranged, which means that the arcs are pre-generated and can be allocated for the execution of TT&C tasks. Tasks generated by a satellite need to be executed on the arcs of satellites that have intersatellite links with the satellite or with it. In short, the TT&C scheduling problem of communication satellites is to assign different types of TT&C tasks to specific types of arcs without violating constraints and determine the execution order and time of the tasks.

[0046] In order to study the characteristics of large-scale and complex communication satellite measurement and control scheduling problems, this paper introduces some reasonable assumptions to standardize this problem:

[0047] 1) The transmission of satellite data is limited to a single-layer intersatellite link, that is, the data of a satellite can only be transmitted to the satellite with which it has an intersatellite link, and no secondary transmission is allowed.

[0048] 2) The task conversion time is a fixed constant, and these constants are the maximum time required to ensure the normal operation of the equipment.

[0049] 3) Sufficient power, storage, and other resources are available to ensure mission completion, regardless of unforeseen circumstances.

[0050] 4) Temporary tasks are not considered. Once a task starts executing, there will be no interruption or resource preemption.

[0051] Based on the problem description and the above assumptions, an integer programming model is proposed. The model and related variables are defined as follows.

[0052] Variable Definition

[0053] R is the set of tasks, R = {i = 1, ..., |r|}

[0054] O arc set, O = {k = 1, ..., |o|}

[0055] d i Duration of task i

[0056] p i Profit of task i

[0057] The executability of task i on arc k is that task i can be executed on arc k. but otherwise es i ,le i The earliest start time and the latest end time of task i

[0058] st k ,et k The measurement and control start time and end time of arc segment k

[0059] T task transition time

[0060] k0 is the conflicting arc of arc k (i.e., k and k0 are obtained by using the same antenna and the same satellite at the same time with different payloads)

[0061] M is a predetermined integer

[0062] Decision variables:

[0063] If tasks i and j are executed in arc k and j is executed immediately after i = 1, otherwise = 0

[0064] If task i is executed in segment k = 1, otherwise = 0

[0065] t i The start time of task i

[0066] Objective function:

[0067]

[0068] constraint:

[0069]

[0070]

[0071] The objective function (1) represents the maximum total benefit of the scheduling task.

[0072] Constraint (2) represents the target uniqueness constraint.

[0073] Constraint (3) is the flow balance constraint and connects the decision variables and

[0074] Constraints (4) and (5) indicate that the task sequence of each arc segment starts from the virtual start task s and ends with the virtual end task e.

[0075] Constraint (6) represents the transition time constraint.

[0076] Constraints (7) and (8) represent arc segment time constraints.

[0077] Constraints (9) and (10) represent task time constraints.

[0078] Constraint (11) indicates that tasks can only be scheduled on executable arcs.

[0079] Constraint (12) indicates that conflicting arcs cannot be used at the same time, that is, if a task has been executed on an arc, its conflicting arc cannot execute any task.

[0080] Constraint (13) represents the value range of the decision variable.

[0081] In the above algorithm model, there is a preprocessing process.

[0082] Preprocessing process:

[0083] The preprocessing process includes calculating the executable arc of each task. An arc can be considered as an executable arc of a task if it meets the following conditions:

[0084] 1) The arc type matches the task type.

[0085] 2) The satellite to which this arc belongs is the same as the mission satellite or there is an intersatellite link between the satellite and the mission satellite.

[0086] 3) The task execution time is within the arc segment range.

[0087] To clearly illustrate this preprocessing process, the present invention provides a simple example involving eight arcs and eight tasks. If all three conditions are met, the arc is considered executable for the task. For example, arc A1 is of the same type as task T1, the satellites for both arc A1 and task T1 are S1, and the time range for task T1 is within the time range of arc A1. Therefore, A1 meets all the necessary conditions to be an executable arc for T1.

[0088] To quickly construct an initial solution, the present invention employs three heuristics: maximum priority, earliest start time, and random construction. For each task, its executable arcs are traversed. If the task satisfies all the constraints, the task is inserted into the current schedule and the traversal is terminated.

[0089] The adaptive large neighborhood search algorithm is specifically:

[0090] ALNS has strong adaptability and self-calibration capabilities, demonstrating strong performance in many combinatorial optimization problems. ALNS uses a pair of heuristic operators to explore the solution space by destroying and repairing the current solution, respectively. We design six removal operators and four insertion operators that define how each task is arranged in the schedule.

[0091] The adaptive large neighborhood search algorithm includes a removal operator and an insertion operator process. The algorithm effectively searches the solution space to find a better solution. Its core is a combination of destruction and repair. The solution is first destroyed using the removal operator, then repaired using the insertion operator to obtain a new solution. If the new solution obtained does not improve after multiple attempts during the algorithm's search, the first iteration mechanism is activated to assist in the removal process.

[0092] The remove operator is specifically designed to allow new insertions and thus obtain a new schedule. It first uses the remove operator to remove some tasks, disrupting the current solution. The number of tasks removed is determined by the removal rate r, which represents the percentage of tasks removed from the current schedule to the total number of scheduled tasks.

[0093] Random Removal: Randomly selects tasks and removes them until a specified number is reached. This operator is the simplest removal heuristic and has a certain degree of blindness, but it helps obtain solutions with different structures and can help the algorithm escape local optima.

[0094] Priority removal: Sort candidate tasks in ascending order of priority. Tasks with lower priorities will be removed first until the specified number of removals is reached.

[0095] Execution Opportunity Removal: Tasks may have multiple executable arcs and, therefore, multiple execution opportunities within the scheduling scope. Tasks with more execution opportunities are removed first, ensuring that less flexible tasks are retained. If tasks have the same number of executable arcs, the lower-priority task is removed first.

[0096] Conflict removal: The task's conflict degree is defined as the average overlap time between a task and other unscheduled tasks. Tasks with greater conflict degrees are more likely to hinder the insertion of other tasks. Therefore, tasks with greater conflict degrees are removed first.

[0097] Priority removal per unit time: Sort tasks from high to low according to the ratio of priority to duration. Tasks with low priority and long execution time will be removed first.

[0098] Arc Priority Removal: According to Pisinger, it is more efficient to remove tasks that are geographically or temporally similar. Tasks scheduled on the same arc have similar execution times. To ensure temporal similarity for removed tasks, this operator selects arcs with high-priority tasks and removes all tasks in the selected arcs.

[0099] Specifically, the insertion operator sorts unscheduled tasks and deleted tasks through different heuristic methods, and uses an effective fast insertion algorithm to insert tasks in sequence.

[0100] Priority insertion: According to He, introducing randomness can diversify the insertion process: the priority is regarded as a heuristic value p. By using p*(1+r) as the heuristic value (where r is a random value in the range [0,1]), a random component is added without neglecting the heuristic.

[0101] Execution Opportunity Insertion: This operator is used to preferentially insert tasks with fewer execution opportunities. Candidate tasks are sorted in ascending order based on the number of their executable arcs.

[0102] Conflict Insertion: This operator is similar to Conflict Removal in that tasks with less conflict are more likely to be inserted into the current solution.

[0103] Unit time priority insertion: This operator sorts candidate tasks in descending order according to the ratio of their priority to their duration.

[0104] After ranking candidate tasks using the heuristics described above, we use the fast insertion algorithm (Algorithm 1) to quickly insert tasks. This method traverses each task's executable arc until it finds an insertion location. If a task has already been executed on a conflicting arc, the task cannot be inserted into the current arc. Otherwise, it traverses all tasks executing in the current arc and calculates their slack based on their latest finish time. A task is successfully inserted if it satisfies its constraints and its subsequent tasks do not violate those constraints.

[0105] The iterative mechanism design process includes a frequent pattern mining process, and the frequent pattern mining process includes a data conversion process, FP-Growth, and a frequent pattern mining perturbation process.

[0106] The design of the iteration mechanism is a core component of the algorithm. By mining relevant information from historical solutions, the algorithm makes more informed decisions. Therefore, the present invention embeds frequent pattern mining methods into the iteration mechanism to improve the quality of the solution.

[0107] Frequent pattern mining process

[0108] Frequent pattern mining can guide the removal of solutions from ALNS by mining low-quality structures from low-quality partial solutions, and can also guide the construction of new solutions by mining superior structures from high-quality solutions. The implementation of this method is described in detail below.

[0109] 1. Transform the data

[0110] For the communication satellite tracking and control scheduling problem, the order in which tasks are executed greatly affects the scheduling results. Therefore, the combination of adjacent tasks is defined as a frequent pattern of the problem and named "knowledge".

[0111] 2.FP-Growth

[0112] The FP-Growth algorithm can reduce the costs associated with generating and testing candidate item sets, thereby accelerating the frequent pattern mining process. Therefore, compared with other frequent pattern mining algorithms such as Apriori, this algorithm is more suitable for large datasets.

[0113] The FP-Growth algorithm mines frequent patterns by storing a dataset in a tree structure called a frequent pattern tree. It first calculates the frequency of each item. It then calculates its support, which is the ratio of the item's occurrence to the total number of items in the dataset. Items with a support greater than a minimum support s are retained. When creating a header table, the items are sorted in descending order of frequency. Then, an FP-tree is constructed from the dataset. To mine frequent patterns, the algorithm traverses paths starting from leaf nodes and generates a conditional pattern tree. Frequent item sets are then mined from these paths.

[0114] 3. Frequent Pattern Mining Perturbations

[0115] If the payoff of a new solution is at least 1% lower than that of the current solution, it is considered a low-quality solution. Tasks present in the new solution but not in the current solution may be the cause of the low-quality solution, and these tasks constitute a low-quality partial solution. Such low-quality partial solutions are recorded during the local search process.

[0116] If ALNS fails to improve the solution within M1 iterations and the number of recorded partial solutions reaches m, the frequent pattern mining method described above is used to mine low-quality structures from the low-quality partial solutions. These low-quality structures are then used as knowledge to guide the removal process. The removal ratio (denoted by R) is increased to expand the search neighborhood. If low-quality structures exist in the solution, they are first removed, followed by the removal operator. After each mining process, the recorded partial solutions are cleared.

[0117] 4. Frequent pattern mining builds new solutions

[0118] Frequent pattern mining can also uncover other useful information, such as specific patterns that frequently appear in high-quality solutions. After each round of parallel ALNS search, frequent pattern mining is applied to discover subsequences of adjacent tasks that appear frequently in high-quality solutions. Based on the mined patterns, two new solutions are constructed. First, the tasks in the frequent pattern are inserted. Then, the optimal solution and a randomly selected solution from the elite solution set are used to guide the insertion process, respectively, to obtain two new solutions. All insertions are performed using the fast insertion algorithm (Algorithm 1). Finally, these two new solutions, along with the optimal solution, are fed into the next round of parallel ALNS.

[0119] 5. Data Verification

[0120] During the data validation process, 11 scenarios with varying numbers of missions were constructed. As shown in Table 1, the number of missions per satellite in these scenarios ranged from 50 to 150, increasing in increments of 10. All missions were randomly generated based on the actual situation. The specific steps for mission generation are as follows:

[0121] Step 1: Read the arc segment data and initialize the problem parameters.

[0122] Step 2 generates n missions for each satellite.

[0123] Step 2-1 randomly generates an integer between 1 and 10 as the priority pi of task i.

[0124] In step 2-2, a random integer between 40 and 50 is generated as the duration ti of task i.

[0125] Steps 2-3 randomly generate 0 or 1 as the task type.

[0126] In steps 2-4, randomly select a segment k of the satellite. stk and etk represent the start and end times of segment k, respectively. Generate an integer in the interval [stk, etk-ti] as the earliest start time for task i, esi. Generate an integer in the interval [esi+ti+10, esi+ti+30] as the latest end time for task i, lei.

[0127] Step 2-5 returns to execute steps 2-1 to 2-4 in sequence to generate the next task, and repeats the process until the number of generated tasks reaches n.

[0128] Table 1 Experimental settings

[0129]

[0130] 6. Parameter settings

[0131] In IALNS, the number of solutions m for frequent pattern mining is set to 10; the minimum support s is set to 0.2; the removal rates r and R are set to 0.1 and 0.3, respectively; the annealing coefficient c is set to 0.95; the maximum number of ALNS generations is set to 500; the fragment size of each ALNS generation is set to 30; the number of unimproved iterations with perturbation enabled is set to 100; the maximum number of unimproved iterations is set to 1000; and the parameters for operator weight updates and score increments follow the settings of Liu et al. The parameters of the comparison algorithm that are not present in this paper are kept consistent with those in the literature.

[0132] 7. Analysis of experimental results

[0133] Scheduling results in different scenarios

[0134] This paper compares several advanced algorithms for solving satellite scheduling problems (including measurement and control and satellite imaging scheduling) to demonstrate the performance of the proposed IALNS. These algorithms include a cluster-based genetic algorithm (C-BGA), a frequent pattern-based parallel search (FPBPS), and an adaptive large neighborhood search (ALNS). The k-means clustering method is combined with the genetic algorithm.

[0135] Table 2 shows the average objective function values ​​and CPU runtimes for five runs of the algorithms, with the best values ​​and runtimes for these algorithms highlighted in bold. The table shows that IALNS outperforms the other algorithms in most scenarios. However, in the C1 scenario, which has the smallest mission scale, IALNS slightly underperforms FPBPS. Compared to the other algorithms, C-BGA, applied to low-orbit internet constellation systems, performs worst in small-scale scenarios (C1-C4), with the longest runtime and the worst objective function values. This may be due to the longer clustering process and greater randomness in the search neighborhood of C-BGA compared to the other algorithms. C-BGA performs better in larger-scale scenarios. Furthermore, as the mission scale increases, the performance gap between IALNS and the other compared algorithms becomes significant, especially when compared to ALNS. Overall, the proposed IALNS offers advantages over other algorithms for solving the communication satellite TT&C scheduling problem.

[0136] Table 2 Comparison results of IALNS with C-BGA, FPBPS, and ALNS

[0137]

[0138]

[0139] Validity test of IALNS:

[0140] Effectiveness of IALNS components

[0141] IALNS consists of two parts: ALNS and an iterative mechanism based on frequent pattern mining. The effectiveness of each component of the algorithm will be analyzed. The specific results are shown in Table 3, which lists the average values ​​of the initial solution generated by the greedy heuristic algorithm, the solution improved by ALNS, and the final solution after further improvement by the iterative mechanism. In the table, △1% and △2% represent the percentage of improvement obtained using ALNS and the iterative mechanism, respectively. In addition, the last column shows the improvement of IALNS compared to the initial solution (expressed in △%). It can be seen that the ALNS algorithm can improve the initial solution generated by the greedy heuristic algorithm by 19.54%, and then further optimized by the iterative mechanism, the benefit value is increased by an average of 1.82%. The IALNS algorithm can improve the initial solution by an average of 21.74%. In addition, the present invention also observed that in scenarios with larger task scales, the improvement ratio is greater, which means that the larger the task scale, the more effective ALNS and the iterative mechanism based on frequent pattern mining are.

[0142] Table 3 Detailed results of IALNS

[0143]

[0144]

[0145] Effectiveness of ALNS adaptive layer and heuristic operators

[0146] We propose several heuristic operators for removing and inserting operators in the ALNS framework. To demonstrate the effectiveness of the adaptation layer and each operator, we consider several configurations, including removing the adaptation layer (corresponding to iterative large neighborhood search (ILNS)) and removing one operator. We test the scenario C11, which has the largest number of tasks.

[0147] A summary of these experiments is shown in the table. R1-R6 represent random removal, priority removal, execution opportunity removal, conflict removal, priority-to-execution-time ratio removal, and priority arc removal, respectively. I1-I4 represent priority insertion, minimum execution opportunity insertion, minimum conflict insertion, and priority-to-execution-time ratio insertion, respectively. -avg.Δ% represents the average percentage decrease in benefit compared to the IALNS algorithm for different configurations. avg.t(s) represents the average running time for different configurations.

[0148] In the first six configurations, we investigated the impact of the removal operator. Configurations seven through ten represent the impact of the insertion operator. The final configuration demonstrates the impact of the adaptation layer. These test results demonstrate that the adaptation layer, as well as all removal and insertion operators, performed well and played a significant role in improving IALNS performance. Configuration 1 exhibited the largest percentage drop, indicating that the random removal operator was most effective.

[0149] Table 4 Detailed configuration and experimental results

[0150]

[0151] Use Friedman test to compare algorithm performance

[0152] According to previous research (cite{Joaquin2011, OJW2023}), the Friedman test can detect significant differences between various results at a significance level of 0.05, thus verifying the effectiveness of the proposed algorithm. Table 5 shows the statistical results of the Friedman test for the four comparison algorithms based on the profit value of the scheduled tasks. Among the four algorithms, IALNS ranked first, followed by FPBPS. These results demonstrate that the proposed IALNS algorithm outperforms other algorithms in terms of the indicators evaluated by the Friedman test and can effectively address the communication satellite TT&C scheduling problem.

[0153] Table 5 Ranking of four compared algorithms

[0154]

[0155]

[0156] Algorithm stability analysis

[0157] Scenarios with larger task sizes better reflect algorithm stability. Algorithm stability can be assessed by the length of the bins, the difference between percentiles, the upper and lower whiskers, and the presence of outliers. As can be seen, none of these algorithms exhibit outliers, with IALNS and C-BGA demonstrating good stability in both C10 and C11.

[0158] The present invention has the following technical effects:

[0159] Compared to other heuristic algorithms, the adaptive nature of the Adaptive Large Neighborhood Search Algorithm (ALNS) and the iterative local search (ILS) algorithm's iterative mechanism effectively prevent the algorithm from becoming trapped in local optima. However, traditional iterative mechanisms lack guidance and direction, making it difficult to find better solutions through perturbations, especially for the extremely complex communication satellite tracking and control scheduling problem. Therefore, it is necessary to leverage information from historical solutions to guide the algorithm's perturbation process.

[0160] Two iterative mechanisms are embedded in Parallel ALNS to provide sufficient perturbations and prevent the algorithm from falling into local optima. Both iterative mechanisms utilize frequent pattern mining: one mechanism identifies specific structures that frequently appear in low-quality partial solutions to assist in the ALNS removal process. The other mechanism mines superior structures from the elite solutions obtained by all threads to guide the construction of new solutions, which are then used as the initial solutions for the next round of ALNS.

[0161] This paper explores the telemetry and control scheduling problem for communication constellations based on low-orbit internet constellations, taking into account the satellite's onboard processing and inter-satellite forwarding capabilities, as well as the matching of different types of resources and tasks. This paper provides an in-depth discussion of this issue, which will be helpful for other researchers to conduct subsequent research on this issue.

[0162] By integrating the concepts of ILS into ALNS, two iterative mechanisms were designed using frequent pattern mining methods. This method mines knowledge about the problem and intelligently guides the search process. This fast and simple algorithmic framework can effectively solve the communication satellite scheduling problem in low-orbit internet constellations.

[0163] Based on the above content, the present application provides an iterative adaptive large neighborhood search method for large-scale communication satellite measurement and control scheduling problems, specifically comprising the following steps:

[0164] S1. Obtain input data, including arc segment data and mission data. The arc segment data is the result of satellite-to-ground link scheduling. An arc segment is considered an executable arc segment for a mission if it meets the following conditions: 1) the arc segment type matches the mission type; 2) the arc segment belongs to the same satellite as the mission satellite or has an intersatellite link with the mission satellite; 3) the mission executable time is within the arc segment range.

[0165] S2. Generate three initial solutions based on the three heuristic rules. The resulting solutions are scheduling plans, where the scheduling plans include the tasks to be executed, the execution arcs of the tasks, and the execution times.

[0166] S3. Randomly input the three initial solutions into three threads and run ALNS in parallel on each thread. Each thread runs the following steps:

[0167] S3.1. Based on the weights of each operator, remove and insert operators according to a roulette wheel;

[0168] If the conditions for activating the iterative mechanism 1 of the frequent pattern mining method are not met, proceed to S3.3. Otherwise, construct a new solution based on the frequent pattern mining method. The specific steps of the iterative mechanism 1 of the frequent pattern mining method are as follows:

[0169] S3.2.1. Data conversion: convert each partial solution task sequence in the low-quality solution set, each of which is a single task, into a sequence of each task consisting of two adjacent tasks;

[0170] S3.2.2. Use the FP-Growth algorithm to mine frequent patterns. Build an FP-tree based on the dataset. To mine frequent patterns, the algorithm traverses paths starting from leaf nodes and generates a conditional pattern tree. Frequent itemsets are then mined from these paths.

[0171] S3.2.3. Remove the mined maximum frequent pattern set from the solution;

[0172] S3.3. Use the remove operator to remove the solution, delete the corresponding task, and clear the low-quality solution set;

[0173] S3.4. Use the insertion operator to repair the solution and obtain a new solution, including sorting the unscheduled tasks and inserting them sequentially using the insertion algorithm;

[0174] If the new solution's payoff is greater than the current solution, accept it. Otherwise, decide whether to accept the new solution according to the simulated annealing criterion.

[0175] S3.6. If the new solution is better than the optimal solution, it is recorded as an elite solution. If the profit value of the new solution is more than 1% lower than the current solution, it is recorded as a low-quality solution. The elite solution will be continuously replaced and updated during the iteration process.

[0176] S3.6. If the termination condition is met, terminate and output the elite solution set; otherwise, return to S3.1;

[0177] S4. If the termination condition is met, the process terminates; otherwise, iterative mechanism 2 starts.

[0178] S5. After the termination condition is met, the obtained solution is used to solve the communication satellite scheduling problem and is applied to determine the tasks to be executed, the execution arc of the tasks, and the execution time.

[0179] Furthermore, the mission data includes scenarios with different numbers of missions. The number of missions for each satellite included in the scenarios ranges from 50 to 150, increasing in increments of 10. All missions are randomly generated based on actual conditions.

[0180] Furthermore, the ALNS is an ALNS algorithm search process assisted by an iterative mechanism.

[0181] Furthermore, the ALNS includes an operator adaptation layer, a neighborhood operator, a new solution, and a preset iterative model processing process.

[0182] Furthermore, the preset iterative model includes the processes of recording low-quality solutions, judging disturbance conditions, converting data, mining frequent patterns, and judging termination criteria.

[0183] Furthermore, the specific steps of the iterative mechanism 2 are as follows: S4.1. The elite solution sets output by the three threads are combined and data converted, and each solution sequence of a single task in the elite solution set is converted into a sequence of two adjacent tasks;

[0184] S4.2. Use the FP-Growth algorithm to mine frequent patterns. Build an FP-tree based on the dataset. To mine frequent patterns, traverse paths starting from leaf nodes and generate a conditional pattern tree. Then, mine frequent itemsets from these paths.

[0185] S4.3. Construct two new solutions based on the mined maximal frequent pattern set. First, insert the task in the frequent pattern. Then, use the optimal solution and a randomly selected solution from the elite solution set to guide the insertion process and obtain two new solutions.

[0186] S4.4. The two new solutions obtained and the optimal solution are updated as the initial solution and enter the next round of parallel ALNS together, returning to S3.

[0187] Furthermore, the deleting of the corresponding tasks is specifically to sort the tasks in the solution according to the removal operator and remove them in sequence until a specified number is reached.

[0188] According to another aspect of the present invention, a system for iterative adaptive large neighborhood search for large-scale communication satellite measurement, control and scheduling problems is provided. The system includes multiple functional modules, and the functional modules implement any of the above-mentioned methods.

[0189] According to another aspect of the present invention, there is further provided an electronic device, comprising:

[0190] at least one processor, and a memory communicatively coupled to the at least one processor;

[0191] The memory stores instructions that can be executed by at least one of the processors, and the instructions are executed by at least one of the processors so that the at least one processor can execute any of the above-mentioned methods.

[0192] According to another aspect of the present invention, a computer-readable storage medium is provided, on which computer instructions are stored. The computer instructions are used to enable the computer to execute any of the above-mentioned methods.

[0193] The above specific embodiments do not constitute a limitation on the scope of protection of this application. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application shall be included within the scope of protection of this application.

[0194] In summary, although the basic structure, principles, and methods of the present invention are specifically described through the above examples, the specific implementation of the present invention is not limited to these examples. For those skilled in the art to which the present invention relates, several simple deductions or substitutions can be made without departing from the concept of the present invention.

Claims

1. An iterative adaptive large neighborhood search method for the large-scale communication satellite tracking and control scheduling problem, comprising the following steps: S1. Obtain input data, including segment data and task data; Arc data is the result of satellite-to-ground link scheduling. An arc can be used as an executable arc for a mission if it satisfies the following conditions simultaneously: 1) the arc type matches the mission type; 2) The satellite to which the arc belongs is the same as the mission satellite or has an intersatellite link with the mission satellite; 3) The mission executable time is within the arc range; S2. Generate three initial solutions based on the three heuristic rules. The resulting solutions are scheduling plans, where the scheduling plans include the tasks to be executed, the execution arcs of the tasks, and the execution times. S3. Randomly input the three initial solutions into three threads and run ALNS in parallel on each thread. Each thread runs the following steps: S3.

1. Based on the weights of each operator, remove and insert operators according to a roulette wheel; If the conditions for activating the iterative mechanism 1 of the frequent pattern mining method are not met, proceed to S3.

3. Otherwise, the ALNS removal process is assisted by the frequent pattern mining method. The specific steps of the iterative mechanism 1 based on the frequent pattern mining method are as follows: S3.2.

1. Data conversion: convert each partial solution task sequence in the low-quality solution set, each of which is a single task, into a sequence of each task consisting of two adjacent tasks; S3.2.

2. Use the FP-Growth algorithm to mine frequent patterns. Build an FP-tree based on the dataset. To mine frequent patterns, the algorithm traverses paths starting from leaf nodes and generates a conditional pattern tree. Frequent itemsets are then mined from these paths. S3.2.

3. Remove the mined maximum frequent pattern set from the solution; S3.

3. Use the remove operator to remove the solution, delete the corresponding task, and clear the low-quality solution set; S3.

4. Use the insertion operator to repair the solution and obtain a new solution. This involves sorting the unscheduled tasks and inserting them sequentially using the insertion algorithm. If the new solution's payoff is greater than the current solution, accept it. Otherwise, decide whether to accept the new solution according to the simulated annealing criterion. S3.

6. If the new solution is better than the optimal solution, it is recorded as an elite solution. If the profit value of the new solution is more than 1% lower than the current solution, it is recorded as a low-quality solution. The elite solution will be continuously replaced and updated during the iteration process. S3.

6. If the termination condition is met, terminate and output the elite solution set; Otherwise, return to S3.1; S4. If the termination condition is met, the process terminates; otherwise, iterative mechanism 2 starts. S5. After the termination condition is met, the obtained solution is used to solve the communication satellite scheduling problem and is applied to determine the tasks to be executed, the execution arc of the tasks, and the execution time.

2. The method according to claim 1, wherein: The mission data includes scenarios with different numbers of missions. The number of missions for each satellite included in the scenarios ranges from 50 to 150, increasing in increments of 10. All missions are randomly generated based on actual conditions.

3. The method according to claim 1, wherein: The ALNS includes an operator adaptation layer, a neighborhood operator, a new solution, and a preset iterative model processing process.

4. The method according to claim 3, wherein: The preset iterative model includes the processes of recording low-quality solutions, judging disturbance conditions, converting data, mining frequent patterns, and judging termination criteria.

5. The method according to claim 1, wherein: The specific steps of the iterative mechanism 2 are as follows: S4.

1. The elite solution sets output by the three threads are combined and data converted, and each solution sequence of a single task in the elite solution set is converted into a sequence of two adjacent tasks; S4.

2. Use the FP-Growth algorithm to mine frequent patterns. Build an FP-tree based on the dataset. To mine frequent patterns, traverse paths starting from leaf nodes and generate a conditional pattern tree. Then, mine frequent itemsets from these paths. S4.

3. Construct two new solutions based on the mined maximal frequent pattern set. First, insert the task in the frequent pattern. Then, use the optimal solution and a randomly selected solution from the elite solution set to guide the insertion process, respectively, to obtain two new solutions. S4.

4. The two new solutions obtained and the optimal solution updated as the initial solution enter the next round of parallel ALNS and return to S3.

6. The method according to claim 1, wherein: The deleting of the corresponding tasks specifically involves sorting the tasks in the solution according to the removal operator and removing them in sequence until a specified number is reached.

7. An iterative adaptive large neighborhood search system for large-scale communication satellite tracking and control scheduling problems, characterized by: The system includes multiple functional modules, and the functional modules implement the method according to any one of claims 1 to 6.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor, and a memory communicatively coupled to the at least one processor; The memory stores instructions that can be executed by at least one of the processors, and the instructions are executed by at least one of the processors to enable the at least one processor to execute the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that The storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the method according to any one of claims 1 to 6.

Images