CN116610917A - Method for scheduling observation tasks of orbiting satellites based on RNP-PSO algorithm - Google Patents
Method for scheduling observation tasks of orbiting satellites based on RNP-PSO algorithm Download PDFInfo
- Publication number
- CN116610917A CN116610917A CN202310572671.4A CN202310572671A CN116610917A CN 116610917 A CN116610917 A CN 116610917A CN 202310572671 A CN202310572671 A CN 202310572671A CN 116610917 A CN116610917 A CN 116610917A
- Authority
- CN
- China
- Prior art keywords
- observation
- task
- sensor
- algorithm
- satellite
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 239000002245 particle Substances 0.000 claims abstract description 79
- 230000008901 benefit Effects 0.000 claims abstract description 28
- 230000001133 acceleration Effects 0.000 claims abstract description 25
- 230000002441 reversible effect Effects 0.000 claims abstract description 15
- 238000010845 search algorithm Methods 0.000 claims abstract description 12
- 230000006870 function Effects 0.000 claims description 25
- 238000001514 detection method Methods 0.000 claims description 17
- 238000005457 optimization Methods 0.000 claims description 15
- 230000008569 process Effects 0.000 claims description 11
- 238000003780 insertion Methods 0.000 claims description 9
- 230000037431 insertion Effects 0.000 claims description 9
- 230000003044 adaptive effect Effects 0.000 claims description 8
- 238000013507 mapping Methods 0.000 claims description 7
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 238000013461 design Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 1
- 239000008186 active pharmaceutical agent Substances 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- FFBHFFJDDLITSX-UHFFFAOYSA-N benzyl N-[2-hydroxy-4-(3-oxomorpholin-4-yl)phenyl]carbamate Chemical compound OC1=C(NC(=O)OCC2=CC=CC=C2)C=CC(=C1)N1CCOCC1=O FFBHFFJDDLITSX-UHFFFAOYSA-N 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000005070 ripening Effects 0.000 description 1
- 239000000523 sample Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/04—Constraint-based CAD
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Medical Informatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The application relates to an orbital satellite observation task scheduling method based on an RNP-PSO algorithm. The method comprises the following steps: establishing a multi-sensor collaborative observation scheduling model; solving a multi-sensor collaborative observation scheduling model according to an RNP-PSO algorithm to obtain a plurality of initial solutions; constructing an initial population according to a plurality of initial solutions; reverse learning is carried out on the initial population by using a reverse learning algorithm to obtain an initialized population; carrying out greedy search on the initialized population according to an IFPFS algorithm to obtain an elite solution set; updating self-adaptive weights, acceleration constants, neighborhoods and particle disturbance in a particle swarm algorithm, and searching elite solution sets on the basis of preset boundary conditions and termination conditions according to the updated particle swarm algorithm to obtain candidate solutions; and searching the candidate solution according to a heuristic search algorithm with the maximized observation benefit to obtain a final solution. By adopting the method, the accuracy rate of the task scheduling of the orbit satellite observation can be improved.
Description
Technical Field
The application relates to the technical field of satellite task scheduling, in particular to an orbital satellite observation task scheduling method based on an RNP-PSO algorithm.
Background
With the development of space exploration, the number of orbiting satellites is increasing, and space surveillance networks are not able to keep pace with increasingly difficult tasks. If the space surveillance network is not able to fully perform its tasks, the space country will be vulnerable to gaps in information caused by the inability to properly describe the orbiting satellites. Due to the shrinking of budgets and limited resources, there is an urgent need to utilize existing assets without increasing inventory.
While orbiting satellites are in orbit, each time a sensor passes through the sensor's field of view, they receive a detectable time window. A problem with multi-sensor collaborative observation scheduling in a spatial monitoring network is that certain time windows are selected from these detectable time windows for tracking observation, and then directory parameters are refined by orbit determination. These detectable time windows have a large number of discrete distributions. The scheduling result must consider the observation task sequence and the observation time of each orbiting satellite at the same time, and as the RSO orbiting satellites increase, the solution space grows exponentially, so that the accuracy of the orbital satellite observation task scheduling is low.
Disclosure of Invention
Accordingly, in order to solve the above-mentioned problems, it is necessary to provide an orbital satellite observation task scheduling method based on RNP-PSO algorithm, which can improve the accuracy of the orbital satellite observation task scheduling.
An orbital satellite observation task scheduling method based on an RNP-PSO algorithm, the method comprising:
acquiring a sensor to be scheduled, a foundation sensor set and an observation task set;
performing pre-analysis according to resource deployment information of the foundation sensors and a pre-acquired orbit database of the orbit satellite to obtain a detectable arc section of each sensor to the orbit satellite;
setting constraint conditions of multi-sensor collaborative observation scheduling according to the detectable arc segments, task demands and resource information of the foundation sensors; setting the total task income of the orbiting satellite to be scheduled as an objective function of the multi-sensor collaborative observation scheduling problem; establishing a multi-sensor collaborative observation scheduling model by using constraint conditions and an objective function;
solving a multi-sensor collaborative observation scheduling model according to an RNP-PSO algorithm, constructing a multi-orbit satellite and a multi-sensor coding structure, and calculating the multi-sensor collaborative observation scheduling model according to three fitness of observation time, observation opportunity and observation conflict on the basis of the coding structure to obtain a plurality of initial solutions; constructing an initial population according to a plurality of initial solutions;
reverse learning is carried out on the initial population by using a reverse learning algorithm to obtain an initialized population; carrying out greedy search on the initialized population according to an IFPFS algorithm to obtain an elite solution set;
Updating self-adaptive weights, acceleration constants, neighborhoods and particle disturbance in a particle swarm algorithm, and searching elite solution sets on the basis of preset boundary conditions and termination conditions according to the updated particle swarm algorithm to obtain candidate solutions;
searching the candidate solution according to a heuristic search algorithm with maximized observation benefits to obtain a final solution; and finally solving a multi-sensor collaborative observation scheduling scheme.
In one embodiment, the pre-analysis is performed according to the deployment information of the sensor resource on the ground and the pre-acquired orbit database of the orbiting satellite, so as to obtain a detectable arc segment of each sensor to the orbiting satellite, including:
obtaining the orbit number of an orbit satellite according to an orbit satellite orbit database, carrying out orbit prediction in a scheduling period by utilizing an orbit prediction model, and converting the position information of the orbit satellite into a station coordinate system by combining the deployment information of a foundation sensor resource j;
according to the detection constraint of the foundation sensor resource, primarily judging whether the orbiting satellite i passes through the detection range of the sensor j;
and calculating the detectable arc section of the orbiting satellite i according to the detection range and the working time of the sensor j.
In one embodiment, the mission requirement comprises a shortest observation time of the orbiting satellite, a minimum observation frequency of the orbiting satellite, a start-stop time of an observation mission and two adjacent observation times of the same orbiting satellite; the resource information of the foundation sensor comprises foundation sensor conversion time, foundation sensor observation capability and minimum elevation angle of the foundation sensor; utilizing constraint conditions and objective functions to establish a multi-sensor collaborative observation scheduling model, comprising:
utilizing constraint conditions and objective functions to establish a multi-sensor collaborative observation scheduling model as
s.t.
Wherein f represents a task benefit value of the scheduling scheme, y i Representing decision variables, y i =1 means that the task was successfully scheduled, otherwise y i =0. N represents the number of tasks, p i Indicating the benefit obtained by successfully executing the task, TS indicating the start time of the schedule, TE indicating the end time of the schedule, TD indicating the total time of the schedule, i indicating the sequence number of the observation task, j indicating the sequence number of the foundation sensor, L indicating the set of the number of all observable arc segments between the observation task i and the foundation sensor j, L indicating the number of all observable arc segments between the observation task i and the foundation sensor j, M indicating the number of foundation sensors, Representing decision variables->Representing a detectable arc segment, A i,j Representing a set of detectable arcs between the ground based sensor and the orbiting satellite, < >>Representing the earliest start time of an observable arc, +.>Indicating the latest end time of the observed arc section, w i Representing oneThe scheduling results, W represents the scheduling result set, fs i Indicating the start time of the actual task execution, fe i Representing the end time of the actual task execution, t= { T i I=1, 2,..n } represents the set of orbital satellite observation tasks, t i Representing orbital satellite observation tasks, ns i And ne i Representing the earliest start time and the latest end time, w, respectively, of a task request h Representation and w i Different another scheduling result, w sk Represents a scheduling result set, fs h Indicating the start time, tr, of another actual task execution of the same target j Representing sensor switching time, fe h Represents the end time of another actual task execution of the same target, phi represents the empty set, r j Represents a foundation sensor, R represents a foundation sensor set,/->Representing the included angle between the line between the observation resource and the orbiting satellite and the horizontal line, dg j Representing the minimum angular requirement satisfied by the sensor observations, d i Representing the shortest detection time, fr, required to perform task i k Representing the number of observations required by an orbiting satellite, id i Representing the identity, dg, corresponding to a task i Represents the shortest time interval, id, between two adjacent observations of an orbiting satellite s Track-around target identifier and id for indicating task sh Indicating the track-wound target identifier to which another task belongs, os Ai,j Representing the earliest start time of an observable arc segment, oe Ai,j Indicating the observable arc end time.
In one embodiment, the encoding structure of the multi-orbit satellite and the multi-sensor is represented by the first orbit satellite performing the task at the 1 st detectable arc, the second orbit satellite performing the task at the 3 rd detectable arc, the third orbit satellite performing the task at the 5 th detectable arc, the i-th orbit satellite performing the task at the j-th detectable arc, and the last orbit satellite performing the task at the 4-th detectable arc.
In one embodiment, according to a heuristic algorithm of time-first resource cost and a random window algorithm, whether each window can execute the task is sequentially considered, when the window can be executed, the observation starting time and the observation ending time of the window are determined, and the observation resources of the window are occupied; the observation resource is a ground-based sensor resource.
In one embodiment, the observation opportunities are the number of observation opportunities for a task under an observation resource that is equal in value to the number of observable windows for the task under the observation resource during one scheduling period.
In one embodiment, observing conflicts includes no conflicts and conflicts; a collision-free means that if the observation arc segments selected by two tasks do not have an intersection, or belong to different observation resources, the probability of the two tasks colliding is 0; a conflict indicates that there is an intersection of two task selected observation arcs and belongs to the same resource, and then there is a conflict over the two observation time windows.
In one embodiment, updating the adaptive weights, acceleration constants, neighborhoods, and particle disturbances in the particle swarm algorithm includes:
the updating mode of the self-adaptive weight is as follows
Wherein omega max Represents the maximum inertial weight, ω min Representing the minimum inertia weight, T representing the current iteration number, phi representing the cosine function phase, and T representing the maximum iteration number;
the updating mode of the self-adaptive weight is as follows
Wherein c 1 (t) represents c 1 Numerical value at t iterations, c 1max Indicating acceleration constant c 1 Maximum value, c 1min Indicating acceleration constant c 1 Minimum value, c 2 (t) represents c 2 Numerical value at t iterations, c 2max Indicating acceleration constant c 2 Maximum value, c 2min The representation represents the acceleration constant c 2 A minimum value;
the neighborhood updating mode is to define a speed step interval [ a, b ] according to the number of observable windows of each orbiting satellite, wherein a is the shortest step when the algorithm operates, and b is the maximum step when the algorithm operates; when the speed is updated, a traditional PSO optimization algorithm is adopted to update a formula, the speeds are ordered in each generation of particles, and integers in the intervals of [ a, b ] are correspondingly used as integer solutions according to the continuous mapping thought according to the speed;
the updating mode of the particle disturbance is to record each generation of optimal particles according to a tabu list strategy, and then obtain a new solution through inter-particle exchange.
In one embodiment, the preset boundary condition is that when the maximum position of the particle exceeds the number of the detectable arc segments of the orbiting satellite, a numerical value is randomly generated in the range of the number of the detectable arc segments to replace; the termination condition is that if the global optimal observation gain is not increased for 20 consecutive generations, and the result after the particle disturbance is not improved, the algorithm is terminated.
In one embodiment, searching the candidate solution according to a heuristic search algorithm with maximized observation yield to obtain a final solution, including:
The candidate solution is used as input of a heuristic searching algorithm for maximizing the observation yield, and a set of observed orbiting satellites and unobserved orbiting satellites is obtained;
searching whether task insertion can be performed in the set of the non-observed orbit satellites, and if the constraint is met, inserting the task insertion into a scheduling scheme to obtain a final solution.
According to the method for scheduling the orbital satellite observation tasks based on the RNP-PSO algorithm, firstly, constraint conditions of multi-sensor collaborative observation scheduling are set according to the detectable arc segments, task demands and resource information of the foundation sensors; setting the total task income of the orbiting satellite to be scheduled as an objective function of the multi-sensor collaborative observation scheduling problem; the method comprises the steps of establishing a multi-sensor collaborative observation scheduling model by using constraint conditions and objective functions, solving the multi-sensor collaborative observation scheduling model by designing a better RNP-PSO algorithm, correcting by adopting a reverse learning algorithm with minimum observation conflict and highest observation priority as targets to obtain an initial solution with better quality, carrying out greedy search on an initialized population by designing an IFCFS algorithm, improving solving efficiency, updating adaptive weights, acceleration constants, neighborhoods and particle disturbance in a particle swarm algorithm, improving the diversity of the population, global optimizing capability and convergence efficiency of particles in the process of carrying out solution optimization by using the updated particle swarm algorithm, avoiding the incidence of local optimization in the solving process by using the neighborhood updating and particle disturbance updating, finally improving the observation benefit and increasing the utilization rate of observation resources by using a heuristic search algorithm based on the maximization of observation benefit, and improving the scheduling accuracy of satellite orbit satellite observation task in the process of observation by using the multi-sensor collaborative observation scheduling scheme.
Drawings
FIG. 1 is a flow chart of an orbital satellite observation task scheduling method based on an RNP-PSO algorithm in one embodiment;
FIG. 2 is a schematic diagram of an encoding structure in one embodiment;
FIG. 3 is a task insertion diagram in one embodiment;
fig. 4 is an internal structural diagram of a computer device in one embodiment.
Detailed Description
The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
In one embodiment, as shown in fig. 1, there is provided an orbital satellite observation task scheduling method based on RNP-PSO algorithm, including the steps of:
102, acquiring a sensor to be scheduled, a foundation sensor set and an observation task set; and carrying out pre-analysis according to the resource deployment information of the foundation sensor and a pre-acquired orbit database of the orbit satellite to obtain a detectable arc section of each sensor to the orbit satellite.
The detectable arc section needs to be analyzed in advance according to the resource deployment information of the foundation sensor and the prior information of the orbiting satellite, the detectability of each sensor to the orbiting satellite is obtained, and basic input is provided for overall scheduling of the detection plan of each sensor. Firstly, obtaining the orbit number of i from an orbit satellite orbit database, carrying out orbit prediction in a scheduling period by utilizing an SGP4 orbit prediction model, and converting the position information of all orbit satellites into a station coordinate system by combining with the deployment information of a foundation sensor resource j; then, according to the detection constraint of the foundation sensor resource, primarily judging whether i passes through the detection range of the sensor j; next, specific probe information is calculated in consideration of the working time of the resource for the scene where i passes through the resource j.
104, setting constraint conditions of multi-sensor collaborative observation scheduling according to the detectable arc segments, task demands and resource information of the foundation sensors; setting the total task income of the orbiting satellite to be scheduled as an objective function of the multi-sensor collaborative observation scheduling problem; and establishing a multi-sensor collaborative observation scheduling model by using the constraint conditions and the objective function.
Utilizing constraint conditions and objective functions to establish a multi-sensor collaborative observation scheduling model as
s.t.
Equation (1) represents that the objective function of the optimization model is to maximize the total task benefit, and f represents the task benefit value of the scheduling scheme.
Constraint (2) is a task-unique constraint, with each observable window being executed at most once.
The constraint (3-5) is a scheduling period constraint, and related elements related to scheduling are all within a scheduling period, including the start-stop time of an observation task, the start time of an observable task and the start-stop time of actual task execution.
Constraint (6) is an observation resource conversion time constraint, and the same observation resource needs to meet a certain time interval between two adjacent tasks to provide equipment and operators to perform necessary preparation work.
Constraint (7) is an observation capability constraint observation unique constraint, and the same foundation sensor resource can only track one target at a time in consideration of the observation tracking of the RSO orbiting satellite.
The constraint (8) is the minimum elevation constraint of the observation resource, and the minimum elevation constraint is required to be considered when the ground-based sensor observes the RSO orbiting satellite by taking the complex environments such as high mountain and the like into consideration.
The constraint (9) is the shortest observation time constraint of the target, and the actual detected time of the target is not less than the detected duration requirement of the target in consideration of the target track precision.
Constraint (10) is a target minimum observation times constraint, and the target is required to be observed multiple times in one scheduling period in consideration of the orbit determination precision of the RSO orbit-finding satellite.
The constraint (11) is the constraint of two adjacent observation times of the same target, and the two adjacent observation times of the same RSO cannot be too low in consideration of the orbit determination precision of the RSO orbit satellite.
The constraint (12) is a task start-stop time constraint, the actual detection start time of the task should be no earlier than the earliest start time of the detectable arc segment and the earliest start time of the task allowed service, and the actual detection end time of the task should be no later than the latest end time of the detectable arc segment and the latest end time of the task allowed service.
And taking the maximum total income of the task as an objective function, establishing a multi-sensor collaborative observation scheduling model by taking the detectable arc section, the task requirement and the resource information of the foundation sensor into consideration, and enabling the requirement of the track-surrounding target observation to be met and simultaneously releasing redundant sensor resources by taking the observation frequency and the minimum observation time into consideration.
Step 106, solving a multi-sensor collaborative observation scheduling model according to an RNP-PSO algorithm, constructing a multi-orbit satellite and a multi-sensor coding structure, and calculating the multi-sensor collaborative observation scheduling model according to three fitness of observation time, observation opportunity and observation conflict on the basis of the coding structure to obtain a plurality of initial solutions; an initial population is constructed from a plurality of initial solutions.
Step 108, reverse learning is carried out on the initial population by using a reverse learning algorithm to obtain an initial population; and carrying out greedy search on the initialized population according to the IFPFS algorithm to obtain an elite solution set.
Step 110, updating the self-adaptive weight, the acceleration constant, the neighborhood and the particle disturbance in the particle swarm algorithm, and searching the elite solution set on the basis of preset boundary conditions and termination conditions according to the updated particle swarm algorithm to obtain candidate solutions.
Step 112, searching the candidate solution according to a heuristic search algorithm with maximized observation benefits to obtain a final solution; and finally solving a multi-sensor collaborative observation scheduling scheme.
The particle velocity of the conventional PSO algorithm is a real number updated towards the evolution direction when the position is updated, which is not suitable for solving the integer number of the optimal window of the application. If only its velocity is rounded, it will result in discontinuous velocity updates, resulting in poor quality solutions. In addition, the conventional PSO algorithm generates initialization particles randomly, and the optimization process has single particle update, so that 'early ripening' is easy to cause, and an optimal solution cannot be obtained. Finally, in view of the particularity of the multi-sensor collaborative observation scheduling problem in orbiting satellite tracking, the observation time does not need to be performed within the whole window, and the specific observation time needs to be determined while the optimal window is determined. The application provides an RNP-PSO algorithm, which overcomes the defects of the traditional PSO algorithm and is more suitable for the problems of the application.
The RNP-PSO algorithm observes an orbit satellite set S to be observed in a scheduling problem in a multi-sensor collaborative mode, a ground sensor resource set R, an orbit satellite observation task set T and a detectable arc segment set A i,j And a set of scheduling intervals [ TS, TE ]]For input, with the goal of minimum observation collision and highest observation priority, an improved first-come-first-serve algorithm (IFCFS is designed) And correcting by adopting a reverse learning algorithm to obtain an initial solution with better quality, then combining the initial solution with a task planning algorithm (MPA) strategy, calculating the observation benefits corresponding to each particle, and reserving the individual extremum and the global extremum of each generation of particles. And updating the particle speed by adopting an adaptive speed updating strategy, and processing the boundary condition of the particle speed. Further, it is determined whether or not the observation yield increases with the increase of the number of iterations. If the continuous DS generation is not improved, the particles are disturbed, the diversity of the population is improved, otherwise, whether termination conditions are met or not is judged, if not, omega, c are adjusted in a self-adaptive mode 1 ,c 2 And continuously iterating until the algorithm meets the termination condition. Finally, a heuristic search algorithm with maximized observation benefit is provided to improve the quality of the solution.
In the method for scheduling the orbital satellite observation tasks based on the RNP-PSO algorithm, constraint conditions of multi-sensor collaborative observation scheduling are set according to the detectable arc segments, task demands and resource information of the foundation sensors; setting the total task income of the orbiting satellite to be scheduled as an objective function of the multi-sensor collaborative observation scheduling problem; the method comprises the steps of establishing a multi-sensor collaborative observation scheduling model by using constraint conditions and objective functions, solving the multi-sensor collaborative observation scheduling model by designing a better RNP-PSO algorithm, correcting by adopting a reverse learning algorithm with minimum observation conflict and highest observation priority as targets to obtain an initial solution with better quality, carrying out greedy search on an initialized population by designing an IFCFS algorithm, improving solving efficiency, updating adaptive weights, acceleration constants, neighborhoods and particle disturbance in a particle swarm algorithm, improving the diversity of the population, global optimizing capability and convergence efficiency of particles in the process of carrying out solution optimization by using the updated particle swarm algorithm, avoiding the incidence of local optimization in the solving process by using the neighborhood updating and particle disturbance updating, finally improving the observation benefit and increasing the utilization rate of observation resources by using a heuristic search algorithm based on the maximization of observation benefit, and improving the scheduling accuracy of satellite orbit satellite observation task in the process of observation by using the multi-sensor collaborative observation scheduling scheme.
In one embodiment, the pre-analysis is performed according to the deployment information of the sensor resource on the ground and the pre-acquired orbit database of the orbiting satellite, so as to obtain a detectable arc segment of each sensor to the orbiting satellite, including:
obtaining the orbit number of an orbit satellite according to an orbit satellite orbit database, carrying out orbit prediction in a scheduling period by utilizing an orbit prediction model, and converting the position information of the orbit satellite into a station coordinate system by combining the deployment information of a foundation sensor resource j;
according to the detection constraint of the foundation sensor resource, primarily judging whether the orbiting satellite i passes through the detection range of the sensor j;
and calculating the detectable arc section of the orbiting satellite i according to the detection range and the working time of the sensor j.
In one embodiment, the mission requirement comprises a shortest observation time of the orbiting satellite, a minimum observation frequency of the orbiting satellite, a start-stop time of an observation mission and two adjacent observation times of the same orbiting satellite; the resource information of the foundation sensor comprises foundation sensor conversion time, foundation sensor observation capability and minimum elevation angle of the foundation sensor; utilizing constraint conditions and objective functions to establish a multi-sensor collaborative observation scheduling model, comprising:
Utilizing constraint conditions and objective functions to establish a multi-sensor collaborative observation scheduling model as
s.t.
Wherein f represents a task benefit value of the scheduling scheme, y i Representing decision variables, y i =1 means that the task was successfully scheduled, otherwise y i =0. N represents the number of tasks, p i Indicating the benefits obtained by successfully executing the task, TS indicating the start time of the schedule, TE indicating the end time of the schedule, TD indicating the total duration of the schedule, i indicating the observation task number, j tableShowing the sequence numbers of the foundation sensors, L represents the number set of all observable arc segments between the observation task i and the foundation sensors j, L represents the number of all observable arc segments between the observation task i and the foundation sensors j, M is the number of the foundation sensors,representing decision variables->Representing a detectable arc segment, A i,j Representing a set of detectable arcs between the ground based sensor and the orbiting satellite, < >>Representing the earliest start time of an observable arc, +.>Indicating the latest end time of the observed arc section, w i Represents a scheduling result, W represents a scheduling result set, fs i Indicating the start time of the actual task execution, fe i Representing the end time of the actual task execution, t= { T i I=1, 2,..n } represents the set of orbital satellite observation tasks, t i Representing orbital satellite observation tasks, ns i And ne i Representing the earliest start time and the latest end time, w, respectively, of a task request h Representation and w i Different another scheduling result, w sk Represents a scheduling result set, fs h Indicating the start time, tr, of another actual task execution of the same target j Representing sensor switching time, fe h Represents the end time of another actual task execution of the same target, phi represents the empty set, r j Represents a foundation sensor, R represents a foundation sensor set,/->Representing the included angle between the line between the observation resource and the orbiting satellite and the horizontal line, dg j Representing minimum angular requirements satisfied by sensor observations，d i Representing the shortest detection time, fr, required to perform task i k Representing the number of observations required by an orbiting satellite, id i Representing the identity, dg, corresponding to a task i Represents the shortest time interval, id, between two adjacent observations of an orbiting satellite s Track-around target identifier and id for indicating task sh Representing the track-bound target identity to which another task belongs,representing the earliest start time of an observable arc, +.>Indicating the observable arc end time.
In one embodiment, the encoding structure of the multi-orbit satellite and the multi-sensor is represented by the first orbit satellite performing the task at the 1 st detectable arc, the second orbit satellite performing the task at the 3 rd detectable arc, the third orbit satellite performing the task at the 5 th detectable arc, the i-th orbit satellite performing the task at the j-th detectable arc, and the last orbit satellite performing the task at the 4-th detectable arc.
In a specific embodiment, since the multi-sensor collaborative observation scheduling problem is a complex combinatorial optimization problem with NP hard characteristics, the encoding structure of the particles is the key to the algorithm solution. While facilitating description and decoding, each RSO may be searched independently before considering that the PSO algorithm is updated with a single particle velocity update as a neighborhood update. Therefore, the application takes the RSO orbital satellite as traction and takes the task number of each RSO as an index, and constructs the multi-RSO orbital satellite and the multi-sensor coding structure as shown in figure 2. In one scheduling period, each RSO is available in advance with respect to the "detectable arc segments" of all sensor resources. Thus, the coding structure is expressed as a first RSO performing tasks at its 1 st detectable arc, a second RSO performing tasks at its corresponding 3 rd detectable arc, a third RSO performing tasks at its corresponding 5 th detectable arc, an ith RSO performing tasks at its corresponding j th detectable arc, and a last RSO performing tasks at its corresponding 4 th detectable arc.
In one embodiment, according to a heuristic algorithm of time-first resource cost and a random window algorithm, whether each window can execute the task is sequentially considered, when the window can be executed, the observation starting time and the observation ending time of the window are determined, and the observation resources of the window are occupied; the observation resource is a ground-based sensor resource.
In one embodiment, the observation opportunities are the number of observation opportunities for a task under an observation resource that is equal in value to the number of observable windows for the task under the observation resource during one scheduling period.
In one embodiment, observing conflicts includes no conflicts and conflicts; a collision-free means that if the observation arc segments selected by two tasks do not have an intersection, or belong to different observation resources, the probability of the two tasks colliding is 0; a conflict indicates that there is an intersection of two task selected observation arcs and belongs to the same resource, and then there is a conflict over the two observation time windows.
In a specific embodiment, the method starts from observation time, observation opportunities and observation conflicts, constructs fitness functions of several types of criteria, designs an FPFS heuristic greedy search algorithm and obtains a preliminary population. And then combining a reverse learning algorithm to obtain an initial population with higher quality.
The observation time refers to the specific observation time of each sensor resource for the RSO orbiting satellite that needs to be determined after the optimal observable window is obtained. The method for constructing the adaptation degree of the observation time in the application refers to the traditional First coming first service algorithm, and the task which arrives first in a window is observed preferentially. For low-orbit RSO orbiting satellites, which may orbit the earth multiple times in a scheduling period, the relative ground-based sensor may generate multiple detectable windows, and thus multiple detectable arcs for each RSO under observation. For high orbit, the orbit period is the same as the earth rotation period, and if the sensor is not changed in orientation, the detectable area is fixed, namely the window is unique. Thus for high rail RSO, the main concern is the specific observation time within an observable window.
The number of observable time windows of each RSO under different observation resources is different, and selectable execution opportunities between tasks and different observation resources are different. Since the high-rail RSO has a longer observable arc (even the entire scheduling period), to maximize global observation gain, the present application first allocates sensor resources to the low-rail targets and then allocates free resources to the high-rail RSO. The definition of the observation opportunity is as follows: the observation opportunities for a task under an observation resource are numerically equal to the number of observable windows for the task under the observation resource during a scheduling period.
Certain cross relation can exist between the selectable detectable arc segments of the same observation task, and the observation time of each task can automatically slide in the detectable arc segments on the premise of meeting the constraint. Due to the limited ability to observe resources, considering the RSO orbiting satellite tracking problem, each resource can only track one RSO at the same time, and the process must be continuous and complete during the observation process. Thus, the task selects different observable arc segments and the probability of collisions occurring during the actual execution time of the slip is not the same. Considering the actual situation, the observation conflict generally exists in the following two cases:
No conflict: if the observation arc segments selected by the two tasks do not have an intersection by themselves or belong to different observation resources, the probability of the two tasks conflicting is 0.
Possible collisions: there is an intersection of the observation arc segments selected by the two tasks and belonging to the same resource, there may be a conflict over the two observation time windows.
Therefore, the probability of collision between two tasks is related to the two "detectable arc lengths" and the specific required observation time of the RSO orbiting satellite, and based on the analysis and the risk probability knowledge, the application provides a window collision evaluation method considering the RSO orbiting satellite observation time, as shown in the formula (13).
Wherein t is c For the overlap time of two windows, n i ，n J The number of RSO that can be observed in the i/j th window, ts k ，ts q RSO on the i, j th windows respectively i Is a function of the time of observation of (a). t is t i,e ，t i,s The real time and the end time of the ith window, respectively. Equation (13) shows that the degree of collision is proportional to the overlap time of the two windows and inversely proportional to the total length of the windows. In addition, equation (13) is more suitable for the problem of the present application, taking into account the observation time of the RSO orbiting satellite.
The main idea of the method is to compare the estimated values of the current point and the reverse point of the variable, and then determine the optimal value, which is defined as follows:
In 1-dimensional space, let x= [ x ] 1 ,x 2 ,...,x d ]Define the inversion point of x as
In d-dimensional space, let x= [ x ] 1 ,x 2 ,...,x d ]And x is i ∈[a i ,b i ]Define the inversion point of x asWherein->
Let x= [ x ] 1 ,x 2 ,...,x d ]To solve the optimization problem, the fitness function is f (x) for a point in d-dimensional space. The inversion point of x is expressed as Is +.>When->Direction Point->Instead of x, x is reserved otherwise.
Therefore, the application obtains the initial solution after combining the three kinds of fitness of the observation time, the observation conflict and the observation opportunity, and places the initial solution in the initial population. Then, a reverse population is obtained by adopting a reverse learning algorithm, and individuals with high fitness are reserved and used as an initialized population. The premise of obtaining a better initialized population is to obtain an elite solution. The application defines the observation time, the observation conflict and the observation opportunity, and weights the three fitness degrees as benefits. The earlier the observation time, the earlier the observation. The fewer the observed collisions, the more reasonable the scheduling scheme generated. The more opportunities for observation, the more the RSO can be made to be observed later. Based on the theory, inspired by FCFS, the IFPFS algorithm is provided in this section, and the algorithm introduces a greedy search mechanism, so that elite individuals are reserved, and the algorithm is convenient to converge.
After the initial population is generated, the MPA algorithm is used to determine the specific observation time. And determining whether the tasks are executed or not according to the conflicts among the tasks and the observation benefits, and determining the final observation benefits. In selecting a particular observation time, it is necessary to consider whether each window can perform the task in turn. When the window can be executed, determining the observation starting time and the observation ending time of the window, and occupying the observation resources of the window, wherein if the task is not in the window or has conflict with the window, the task is not scheduled.
In the present application, two heuristic rules are designed to determine the specific observation time of each task, one is a heuristic algorithm (TFRC) based on time-first resource cost, and the other is a random window algorithm (RT).
The flow of the TRFC algorithm is as follows: for tasks within each scheduling window, the tasks are first ordered by time and, if the task meets all constraints, the task is assigned to the corresponding window. Since each cycle places the earliest task in the window, no other resources are considered, and therefore, resource cost time prioritizes.
The procedure of the RT algorithm is as follows: for a task within each scheduling window, if the task meets all constraints, the task is randomly placed elsewhere within the window. Since the tasks are randomly placed in each cycle, time and resources may be affected, and thus the task is also called a random window algorithm.
In one embodiment, updating the adaptive weights, acceleration constants, neighborhoods, and particle disturbances in the particle swarm algorithm includes:
the updating mode of the self-adaptive weight is as follows
Wherein omega max Represents the maximum inertial weight, ω min Representing the minimum inertia weight, T representing the current iteration number, phi representing the cosine function phase, and T representing the maximum iteration number;
the updating mode of the self-adaptive weight is as follows
Wherein c 1 (t) represents c 1 Numerical value at t iterations, c 1max Indicating acceleration constant c 1 Maximum value, c 1min Indicating acceleration constant c 1 Minimum value, c 2 (t) represents c 2 Numerical value at t iterations, c 2max Indicating acceleration constant c 2 Maximum value, c 2min The representation represents the acceleration constant c 2 A minimum value;
the neighborhood updating mode is to define a speed step interval [ a, b ] according to the number of observable windows of each orbiting satellite, wherein a is the shortest step when the algorithm operates, and b is the maximum step when the algorithm operates; when the speed is updated, a traditional PSO optimization algorithm is adopted to update a formula, the speeds are ordered in each generation of particles, and integers in the intervals of [ a, b ] are correspondingly used as integer solutions according to the continuous mapping thought according to the speed;
the updating mode of the particle disturbance is to record each generation of optimal particles according to a tabu list strategy, and then obtain a new solution through inter-particle exchange.
In a specific embodiment, the adaptive weights: the inertia weight ω is an important control parameter in the PSO algorithm and can be used to control the development and exploration capabilities of the algorithm. The magnitude of ω indicates how much is inherited to the current velocity of the particle. When ω is larger, the global optimizing capability of the particles is stronger and the local optimizing capability is weaker. When ω is smaller, the global optimizing ability of the particles is weaker and the local optimizing ability is stronger.
Considering the complexity of the MCOS problem, the observation resource generates certain benefits when observed in each window, the cooperative observation among different windows has great influence on the observation benefits, the algorithm has stronger global searching capability in the initial period, and the algorithm needs stronger local optimizing capability in the later period for convergence. Therefore, the present application designs an adaptive cosine weight updating method, as shown in formula (14).
Acceleration constant: acceleration constant c 1 And c 2 Respectively adjust the direction p best And g best The maximum compensation of the directional flight respectively determines the influence of individual particle experience and group experience on the particle running track and reflects the information communication among particle groups. When c 1 When the particle size is larger, the particle has stronger exploration capacity, and is helpful for finding an optimal observation plan. When c 2 When the particle size is larger, the particle has stronger development capability, and is beneficial to algorithm convergence. To maintain good search performance, c should be reduced during the iteration process 1 Increase c 2 In combination with the self-adaptive cosine weight updating method, the section design is thatC of (2) 1 、c 2 The updating method is shown in the formula (15-16).
Neighborhood update: the PSO algorithm of the global version takes the whole population as the neighborhood of the particles, has the advantage of high searching speed, and is easy to sink into local optimization. The particle swarm algorithm of the local version takes individuals with similar positions as the neighborhood of particles, has low convergence speed and is not easy to fall into local optimum.
For the MCOS problem, the application not only corrects the inertia weight and the acceleration constant, but also designs a self-adaptive neighborhood updating method. Conventional PSO algorithms update mainly the velocity and multiplication of random numbers, inertia constants, etc. will appear real, which is not appropriate for optimizing the window (integer) here. If a forced rounding approach is used, this may result in discontinuous speed updates, resulting in poor quality solutions, e.g., [1.1-1.4] all defined to be "1".
Based on the reasons, the application designs a self-adaptive neighborhood updating method, which uses the idea of continuous mapping as reference, and comprises the following main processes: (1) A speed step interval is defined according to the number of observable windows per RSO [ a, b ], where a is the shortest step at which the algorithm is running and b is the largest step at which the algorithm is running. (2) And updating a formula, such as formula (17), by adopting a traditional PSO optimization algorithm when updating the speed. The velocities are then ordered in each generation of particles. (3) According to the speed, the continuous mapping thought is used for reference, the continuous mapping thought is correspondingly used as an integer in the [ a, b ] interval, and the particle position is updated by adopting a formula (18).
v ij (t+1)＝ω×v ij (t)+c 1 ×rand×(p ij (t)-x ij (t))
+c 2 ×rand×(g ij (t)-x ij (t)) (17)
x ij (t+1)＝x ij (t)+v ij (t+1) (18)
From the above analysis, it can be seen that the method overcomes the problem of discontinuous speed updates by considering the speed ordering and mapping it to an integer set, and is more applicable to the problem herein by generating an integer solution.
Particle perturbation: to avoid the algorithm falling into a local optimum, the acceleration constant c 1 And c 2 With proper adjustment, it is still difficult to avoid the local optimal solution, so the application proposes a particle perturbation method, which records each generation of optimal particles through a tabu-table strategy, and then obtains a new solution through inter-particle exchange. The method can improve the diversity of solutions, and the quality of the solutions generated by the method is higher because the optimal particles in each generation are recorded by adopting a tabu table.
In one embodiment, the preset boundary condition is that when the maximum position of the particle exceeds the number of the detectable arc segments of the orbiting satellite, a numerical value is randomly generated in the range of the number of the detectable arc segments to replace; the termination condition is that if the global optimal observation gain is not increased for 20 consecutive generations, and the result after the particle disturbance is not improved, the algorithm is terminated.
In a specific embodiment, when the position or speed of a certain dimension or a plurality of dimensions exceeds a set value, the position of the particles can be limited in a feasible search space by adopting a boundary condition processing strategy, so that the expansion and divergence of the population are avoided, and the blind search of the particles is also avoided, thereby improving the solving efficiency. In the present application, the encoding of the particles is strictly obtained according to the number of each RSO, if the ith RSO has only 5 detectable arc segments in one scheduling period, and the encoding is performed to 7, this may cause solving failure. In order to avoid the above problems, the present application employs a boundary absorption method, i.e., when the maximum position exceeds the number of detectable arc segments of the RSO, a number is randomly generated within the range to be replaced.
In one embodiment, searching the candidate solution according to a heuristic search algorithm with maximized observation yield to obtain a final solution, including:
the candidate solution is used as input of a heuristic searching algorithm for maximizing the observation yield, and a set of observed orbiting satellites and unobserved orbiting satellites is obtained;
searching whether task insertion can be performed in the set of the non-observed orbit satellites, and if the constraint is met, inserting the task insertion into a scheduling scheme to obtain a final solution.
In a specific embodiment, in order to improve the quality of an optimal solution, improve the observation benefit and increase the utilization rate of observation resources, the application provides a heuristic search algorithm based on the maximization of the observation benefit aiming at the specificity of the MCOS problem by taking the scheduling solution obtained through optimization as input. The core idea of the algorithm is to add unsuccessfully scheduled tasks into a scheduling scheme, so that the observation benefit of the scheduling scheme is improved.
As shown in fig. 3, for task 2 to be inserted, it is determined whether the insertion opportunity conflicts with the scheduling scheme, and if there is no conflict, it may be directly added to the scheduling sequence. In summary, the flow is as follows: (1) Taking a solution of an optimization algorithm as input to obtain a set of observed RSO orbiting satellites and unobserved RSO orbiting satellites; (2) Searching whether task insertion can be performed in the RSO orbiting satellite set or not; (3) If the constraint is satisfied, it is inserted into the scheduling result, otherwise it is not observable.
It should be understood that, although the steps in the flowchart of fig. 1 are shown in sequence as indicated by the arrows, the steps are not necessarily performed in sequence as indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited in the present application, and the steps may be executed in other orders. Moreover, at least some of the steps in fig. 1 may include multiple sub-steps or stages that are not necessarily performed at the same time, but may be performed at different times, nor do the order in which the sub-steps or stages are performed necessarily performed in sequence, but may be performed alternately or alternately with at least a portion of other steps or sub-steps of other steps.
It will be appreciated by persons skilled in the art that the architecture shown in fig. 4 is merely a block diagram of some of the architecture relevant to the present inventive arrangements and is not limiting as to the computer device to which the present inventive arrangements are applicable, and that a particular computer device may include more or fewer components than shown, or may combine some of the components, or have a different arrangement of components.
Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in embodiments provided herein may include non-volatile and/or volatile memory. The nonvolatile memory can include Read Only Memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), memory bus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), among others.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples illustrate only a few embodiments of the application, which are described in detail and are not to be construed as limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of protection of the present application is to be determined by the appended claims.
Claims (10)
1. An orbital satellite observation task scheduling method based on an RNP-PSO algorithm is characterized by comprising the following steps:
acquiring a sensor to be scheduled, a foundation sensor set and an observation task set;
performing pre-analysis according to resource deployment information of the foundation sensors and a pre-acquired orbit database of the orbit satellite to obtain a detectable arc section of each sensor to the orbit satellite;
Setting constraint conditions of multi-sensor collaborative observation scheduling according to the detectable arc segments, task demands and resource information of the foundation sensors; setting the total task income of the orbiting satellite to be scheduled as an objective function of the multi-sensor collaborative observation scheduling problem;
establishing a multi-sensor collaborative observation scheduling model by utilizing the constraint conditions and the objective function;
solving the multi-sensor collaborative observation scheduling model according to an RNP-PSO algorithm, constructing a multi-orbit satellite and multi-sensor coding structure, and calculating the multi-sensor collaborative observation scheduling model according to three fitness of observation time, observation opportunity and observation conflict on the basis of the coding structure to obtain a plurality of initial solutions; constructing an initial population according to the plurality of initial solutions;
reverse learning is carried out on the initial population by using a reverse learning algorithm to obtain an initial population; carrying out greedy search on the initialized population according to an IFPFS algorithm to obtain an elite solution set;
updating self-adaptive weights, acceleration constants, neighborhoods and particle disturbance in a particle swarm algorithm, and searching the elite solution set on the basis of preset boundary conditions and termination conditions according to the updated particle swarm algorithm to obtain candidate solutions;
Searching the candidate solution according to a heuristic search algorithm with maximized observation benefits to obtain a final solution; the final solution is a multi-sensor collaborative observation scheduling scheme.
2. The method of claim 1, wherein the pre-analysis based on the ground-based sensor resource deployment information and the pre-acquired orbital satellite orbit database to obtain a detectable arc for each sensor pair of the orbital satellite comprises:
obtaining the orbit number of an orbit satellite according to an orbit satellite orbit database, carrying out orbit prediction in a scheduling period by utilizing an orbit prediction model, and converting the position information of the orbit satellite into a station coordinate system by combining the deployment information of a foundation sensor resource j;
according to the detection constraint of the foundation sensor resource, primarily judging whether the orbiting satellite i passes through the detection range of the sensor j;
and calculating the detectable arc section of the orbiting satellite i according to the detection range and the working time of the sensor j.
3. The method of claim 1, wherein the mission requirement comprises a minimum observation time for an orbiting satellite, a minimum number of observations for an orbiting satellite, a start-stop time for an observation mission, and two adjacent observations for the same orbiting satellite; the resource information of the foundation sensor comprises foundation sensor conversion time, foundation sensor observation capability and minimum elevation angle of the foundation sensor; and establishing a multi-sensor collaborative observation scheduling model by using the constraint condition and the objective function, wherein the multi-sensor collaborative observation scheduling model comprises the following steps:
Establishing a multi-sensor collaborative observation scheduling model as follows by utilizing the constraint condition and the objective function
s.t.
Where f represents a task benefit value of the scheduling scheme, yi represents a decision variable, yi=1 represents that the task is successfully scheduled, otherwise yi=0. N represents the number of tasks, p i Indicating the benefit obtained by successfully executing the task, TS indicating the start time of the schedule, TE indicating the end time of the schedule, TD indicating the total time of the schedule, i indicating the sequence number of the observation task, j indicating the sequence number of the foundation sensor, L indicating the set of the number of all observable arc segments between the observation task i and the foundation sensor j, L indicating the number of all observable arc segments between the observation task i and the foundation sensor j, M indicating the number of foundation sensors,representing decision variables->Representing a detectable arc segment, A i,j Representing a set of detectable arcs between the ground based sensor and the orbiting satellite, < >>Representing the earliest start time of an observable arc, +.>Indicating the latest end time of the observed arc section, w i Represents a scheduling result, W represents a scheduling result set, fs i Indicating the start time of the actual task execution, fe i Representing the end time of the actual task execution, t= { T i I=1, 2,..n } represents the set of orbital satellite observation tasks, t i Representing orbital satellite observation tasks, ns i And ne i Representing the earliest start time and the latest end time, w, respectively, of a task request h Representation and w i Different another scheduling result, w sk Represents a scheduling result set, fs h Representation ofStart time of another actual task execution of the same target tr j Representing sensor switching time, fe h Represents the end time of another actual task execution of the same target, phi represents the empty set, r j Represents a foundation sensor, R represents a foundation sensor set,/->Representing the included angle between the line between the observation resource and the orbiting satellite and the horizontal line, dg j Representing the minimum angular requirement satisfied by the sensor observations, d i Representing the shortest detection time, fr, required to perform task i k Representing the number of observations required by an orbiting satellite, id i Representing the identity, dg, corresponding to a task i Represents the shortest time interval, id, between two adjacent observations of an orbiting satellite s Track-around target identifier and id for indicating task sh Representing the track-bound target identity to which another task belongs,representing the earliest start time of an observable arc, +.>Indicating the observable arc end time.
4. A method according to any one of claims 1 to 3, wherein the code structure of the multi-orbital satellite and the multi-sensor is expressed as that the first orbital satellite performs the task at its 1 st detectable arc, the second orbital satellite performs the task at its corresponding 3 rd detectable arc, the third orbital satellite performs the task at its corresponding 5 th detectable arc, the i-th orbital satellite performs the task at its corresponding j-th detectable arc, and the last orbital satellite performs the task at its corresponding 4 th detectable arc.
5. The method of claim 1, wherein the process of obtaining the observation time comprises:
sequentially considering whether each window can execute the task according to a heuristic algorithm and a random window algorithm of time priority resource cost, and determining the observation start time and the observation end time of the window and occupying the observation resources of the window when the window can be executed; the observation resource is a ground-based sensor resource.
6. The method of claim 5, wherein the observation opportunities are the number of observation opportunities for a task under an observation resource equal in value to the number of observation windows for the task under the observation resource during a scheduling period.
7. The method of claim 6, wherein the observed conflicts comprise no conflicts and conflicts; the conflict-free means that if the observation arc segments selected by the two tasks do not have intersection per se or belong to different observation resources, the probability of the two tasks conflicting is 0; the conflict indicates that the observation arc sections selected by the two tasks have intersection and belong to the same resource, and then the conflict exists on the two observation time windows.
8. The method of claim 1, wherein updating the adaptive weights, acceleration constants, neighborhoods, and particle disturbances in the particle swarm algorithm comprises:
The updating mode of the self-adaptive weight is as follows
Wherein omega max Represents the maximum inertial weight, ω min Representing the minimum inertia weight, T representing the current iteration number, phi representing the cosine function phase, and T representing the maximum iteration number;
the updating mode of the self-adaptive weight is as follows
Wherein c 1 (t) represents c 1 Numerical value at t iterations, c 1max Indicating acceleration constant c 1 Maximum value, c 1min Indicating acceleration constant c 1 Minimum value, c 2 (t) represents c 2 Numerical value at t iterations, c 2max Indicating acceleration constant c 2 Maximum value, c 2min The representation represents the acceleration constant c 2 A minimum value;
the neighborhood updating mode is to define a speed step interval [ a, b ] according to the number of observable windows of each orbiting satellite, wherein a is the shortest step when the algorithm operates, and b is the maximum step when the algorithm operates; when the speed is updated, a traditional PSO optimization algorithm is adopted to update a formula, the speeds are ordered in each generation of particles, and integers in the intervals of [ a, b ] are correspondingly used as integer solutions according to the continuous mapping thought according to the speed;
the updating mode of the particle disturbance is to record each generation of optimal particles according to a tabu list strategy, and then obtain a new solution through inter-particle exchange.
9. The method of claim 1, wherein the predetermined boundary condition is that when the maximum position of the particle exceeds the number of detectable arcs of the orbiting satellite, a value is randomly generated within the range of the number of detectable arcs to be substituted; the termination condition is that if the global optimal observation gain is not increased continuously for 20 generations, the result after the particle disturbance is not improved, and the algorithm is terminated.
10. The method of claim 1, wherein searching the candidate solution according to a heuristic search algorithm that maximizes observed revenue to obtain a final solution comprises:
the candidate solution is used as input of a heuristic searching algorithm for maximizing the observation yield, and a set of observed orbiting satellites and unobserved orbiting satellites is obtained;
searching whether task insertion can be performed in the set of the non-observed orbit satellites, and if the constraint is met, inserting the task insertion into a scheduling scheme to obtain a final solution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310572671.4A CN116610917A (en) | 2023-05-19 | 2023-05-19 | Method for scheduling observation tasks of orbiting satellites based on RNP-PSO algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310572671.4A CN116610917A (en) | 2023-05-19 | 2023-05-19 | Method for scheduling observation tasks of orbiting satellites based on RNP-PSO algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116610917A true CN116610917A (en) | 2023-08-18 |
Family
ID=87675958
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310572671.4A Pending CN116610917A (en) | 2023-05-19 | 2023-05-19 | Method for scheduling observation tasks of orbiting satellites based on RNP-PSO algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116610917A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117168490A (en) * | 2023-11-03 | 2023-12-05 | 四川国蓝中天环境科技集团有限公司 | Road dust accumulation navigation monitoring vehicle route planning method based on mathematical heuristic method |
-
2023
- 2023-05-19 CN CN202310572671.4A patent/CN116610917A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117168490A (en) * | 2023-11-03 | 2023-12-05 | 四川国蓝中天环境科技集团有限公司 | Road dust accumulation navigation monitoring vehicle route planning method based on mathematical heuristic method |
| CN117168490B (en) * | 2023-11-03 | 2024-01-23 | 四川国蓝中天环境科技集团有限公司 | Road dust accumulation navigation monitoring vehicle route planning method based on mathematical heuristic method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Du et al. | A data-driven parallel scheduling approach for multiple agile earth observation satellites | |
| CN109087023B (en) | Multi-satellite observation layered scheduling method and system based on divide-and-conquer strategy | |
| CN116610917A (en) | Method for scheduling observation tasks of orbiting satellites based on RNP-PSO algorithm | |
| CN113852405B (en) | Construction method and device of multi-beam relay satellite task scheduling model | |
| CN108090630B (en) | Multi-satellite emergency task planning method and device | |
| US10762458B1 (en) | Satellite scheduling system | |
| EP3549366B1 (en) | Forcasting time series data | |
| US8199643B2 (en) | Optimization strategies for resource management and course of action analysis | |
| CN111651905A (en) | Agile satellite scheduling method considering time-dependent conversion time | |
| CN114357724A (en) | Dynamic multi-objective optimization-based opportunistic frequency planning method, device and equipment | |
| Gaafar et al. | Reinforcement learning for cognitive radar task scheduling | |
| Saw et al. | Intelligent buses in a loop service: Emergence of no-boarding and holding strategies | |
| Johnson et al. | Hybrid information and plan consensus in distributed task allocation | |
| US7991722B1 (en) | Information request generator | |
| CN114846492A (en) | Method of operating one or more vehicles and/or equipment in a terminal, and related apparatus and system | |
| CN112382081A (en) | Traffic flow prediction method based on multiple tasks | |
| CN113852406B (en) | Multi-beam relay satellite task scheduling method and device | |
| CN115576278A (en) | Multi-agent multi-task layered continuous control method based on temporal equilibrium analysis | |
| EP4266004A1 (en) | User trajectory positioning method, electronic device and computer storage medium | |
| Elboher et al. | A formal metareasoning model of concurrent planning and execution | |
| JP7313323B2 (en) | TIME SERIES FORECAST SYSTEM, TIME SERIES FORECAST METHOD AND COMPUTER PROGRAM | |
| CN111950869A (en) | Iterative solution method and system for improving initial solution of scheduling problem of space measurement and control network | |
| Zhao et al. | An improved local search algorithm with pruning for satellite data transmission scheduling problem | |
| Capponi et al. | A polynomial time algorithm for the multidimensional assignment problem in multiple sensor environments | |
| US20220391249A1 (en) | Cognitive Resource Scheduling |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |